Conversation in fMRI: DMN Activates During Live Mentalizing
Using fMRI data from participants engaged in long live conversations, this study reveals that while language production and comprehension show distinct regional asymmetries, comprehension uniquely recruits the default mode network and frontoparietal control network, suggesting enhanced self-reflective and Theory of Mind processing during listening compared to speaking.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Language is often thought of as a tool for thinking, a way to organize our own ideas. But many scientists now believe language evolved primarily for something else: for talking to one another. If this is true, then the most natural place to study how our brains handle language is not in a quiet room where a person listens to recorded sentences, but in the messy, fast-paced reality of a live conversation. In a real chat, we are not just decoding words; we are guessing what the other person is thinking, planning our next move, and reacting to their tone and gestures, all while the clock is ticking. To understand how the brain manages this complex dance of speaking and listening, researchers need to look beyond the classic language centers and examine the broader networks that handle social understanding and mental control.
A team of researchers at the KTH Royal Institute of Technology and Stockholm University set out to explore this by placing people inside a brain scanner while they had real conversations. They wanted to see if the brain works differently when we are speaking compared to when we are listening, and which parts of the brain light up to handle the social and mental demands of the moment. They focused on two major brain systems that usually work in the background: one network that helps us understand the minds of others and imagine different scenarios, and another that helps us stay focused and manage complex tasks. By recording brain activity during long, ten-minute discussions about ethical dilemmas, they captured a rare glimpse of the brain in its natural social habitat.
The study involved thirty participants who engaged in three separate ten-minute conversations with a partner. The partner was a researcher's assistant who acted as a friend, guiding the discussion through various ethical questions. The setup was designed to be as natural as possible, with the participant seeing and hearing the partner in real time, allowing for spontaneous back-and-forth dialogue. While the participant spoke, the researchers recorded their brain activity. When the partner spoke, the participant listened, and the researchers recorded that activity as well. This allowed the team to compare the brain's state during production (speaking) and comprehension (listening) within the same person, under the same conditions.
The results confirmed some long-held ideas about how the brain handles language. When people were speaking, a specific area in the lower left part of the forehead was more active, while listening triggered stronger activity in the upper and middle parts of the temples on both sides. This aligns with the known division of labor in the language network, where one side of the brain tends to handle the motor planning of speech, and the other handles the interpretation of sounds and meaning. However, the study went much further than just looking at these language centers. It revealed a striking difference in how the brain's broader social and control networks behaved during the two tasks.
When the participants were listening, a large network known as the default mode network became highly active. This network is often associated with daydreaming, thinking about oneself, or imagining what others are thinking. The researchers found that during comprehension, the brain was not just passively receiving words; it was actively building a mental model of the conversation, tracking the partner's beliefs, and planning what to say next. This suggests that while listening, the brain has the mental space to engage in deep social reflection and to anticipate the flow of the dialogue. In contrast, when the participants were speaking, this same network was less active. The act of formulating a sentence, managing the timing of the turn, and coordinating gestures appeared to take up so much of the brain's resources that it suppressed these broader, internal processes. The speaker was too busy managing the immediate demands of the conversation to engage in the deeper, self-reflective thinking that the listener could afford to do.
The study also looked at a second network responsible for cognitive control, which helps us switch between tasks and stay focused. This network showed a similar pattern, becoming more active during listening than during speaking, though the effect was less pronounced than with the default mode network. This indicates that listening requires a significant amount of mental effort to integrate new information, manage the social context, and prepare for the next turn. Interestingly, a specific sub-network within the default mode system, which is linked to understanding the minds of others, remained active throughout the entire conversation, whether the person was speaking or listening. This suggests that the ability to infer what another person is thinking is a continuous, underlying current in any social interaction, never fully turning off.
One of the most stable findings was that a specific area in the upper middle part of the forehead was consistently more active when people were speaking. This region, which sits near the motor areas that control the body, seems to be crucial for the higher-level planning required to produce speech in a social setting. It appears that the brain recruits this area to help organize the complex task of turning a thought into a spoken sentence while keeping an eye on the listener's reaction. This finding held true regardless of how long the conversation lasted, suggesting it is a fundamental part of how we speak to one another.
The researchers also noted that the brain's activity was not simply a matter of being "on" or "off." The differences between speaking and listening were subtle but distinct, reflecting the different pressures each role places on the mind. Speaking is a bottleneck; it requires rapid, uninterrupted execution and tight coordination with the listener's feedback. Listening, on the other hand, allows for a broader allocation of resources, enabling the brain to construct a rich picture of the situation, the speaker's intent, and the direction the conversation might take. The study suggests that the brain does not just process language; it constantly shifts its focus between the immediate mechanics of speech and the deeper social and mental work of understanding, depending on whether it is the one talking or the one listening.
By using long conversations instead of short, isolated sentences, the study captured a more complete picture of how the brain works in the real world. The findings suggest that the brain's language system is deeply intertwined with networks that handle social cognition and mental control. It is not just a machine for decoding words, but a dynamic system that adapts its resources to the demands of the moment. When we speak, we focus on the mechanics of delivery; when we listen, we open up to the full complexity of the social world. This research helps us understand that conversation is not just an exchange of information, but a complex, coordinated dance of two minds, each using different parts of their brain to stay in step.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.