Summary: What makes us human? For decades scientists have debated whether our ability to use language and our capacity to understand other people’s minds—often called Theory of Mind—arise from a single shared cognitive source or from separate systems. A new study using functional MRI (fMRI) in young children finds clear evidence for distinct, non-overlapping brain systems for language and mentalizing, indicating these abilities have separate neural origins from early development.
Researchers scanned the brains of children as young as three while they listened to spoken sentences and while they watched brief silent cartoons. The results show that the brain regions that support language and those that support Theory of Mind are spatially distinct within the superior temporal lobe, already lateralized to opposite hemispheres in early childhood. Resting-state connectivity analyses confirmed that each region communicates with the rest of the brain in its own characteristic way. Longitudinal data further indicate that this separation is stable across childhood, although the networks become more interactive in adulthood as the skills are used together in complex social communication.
Key Facts
- Hemispheric separation: Both functions are centered in the superior temporal lobe, but language is anchored in the left hemisphere while Theory of Mind is anchored in the right.
- No developmental overlap: Contrary to the idea that early childhood brains are initially “messy” and later specialize, the study found these regions are distinct even in three-year-olds and show no spatial overlap.
- Connectivity fingerprints: Resting-state scans revealed unique connectivity patterns for each region. These distinct “fingerprints” demonstrate that language and mentalizing systems interact with other brain areas in different ways.
- Stable wiring across childhood: Longitudinal analyses tracking the same children over time found that the regional specialization and connectivity patterns remain consistent rather than emerging from a single shared system.
- Adult integration: Although the regions are separate in children, the networks show increased cross-talk in adults, suggesting that as people mature they integrate language and social cognition more frequently to support complex communication.
Source: Ohio State University
New evidence shows language and Theory of Mind have separate neural beginnings in childhood, mirroring the adult brain.
This research challenges theories proposing that language and the ability to infer others’ mental states share a common developmental origin. Instead, the findings support a model in which the human brain contains distinct, specialized circuits for these two foundational capacities from an early age.
In the study, the team mapped brain activity in response to language and to scenarios requiring mental-state inference. During language tasks children listened to coherent sentences and to nonsense-word controls; during Theory of Mind tasks they viewed silent cartoons designed to elicit mentalizing versus control scenes that elicited basic responses such as perceiving pain. High-resolution imaging across both hemispheres showed that the clusters of activation for language and for Theory of Mind were spatially separate, with no detectable overlap at the voxel level.
Lead author Kelly Hiersche, a doctoral student in the lab, noted that the findings extend established adult results into developmental stages: the same dissociation seen in adults already exists in young children. Senior author Zeynep Saygin, associate professor of psychology, emphasized the evolutionary implication: the processors that support speech and those that support mentalizing appear to have been dissociated early in our species’ history, leaving distinct traces in human development.
To deepen the analysis, researchers examined resting-state connectivity—the pattern of spontaneous communication between brain regions when a person is not performing a task. A region’s connectivity fingerprint helps reveal how it functions within broader networks. Predictive modeling demonstrated that the connectivity fingerprints for language and Theory of Mind regions are distinct and largely non-overlapping in children, and these patterns remain stable over time within individuals.
Longitudinal scans allowed the team to observe development within the same children. The data did not support a model in which early overlap resolves into separation; rather, the specialization is already present at the youngest ages sampled and persists through childhood. When comparing children to adults, the researchers did note increased interaction between the networks in adulthood, suggesting that while the systems are distinct, they become more coordinated as people learn to use language and empathy together in social contexts.
These results argue against the idea that language emerges from more general social-processing circuits and instead support distinct neural mechanisms for language and Theory of Mind. The findings have implications for how we interpret developmental delays: a child who is a late talker, for example, may still have typical Theory of Mind development because the underlying systems are separate.
Funding: This research was supported by the U.S. National Science Foundation, the Alfred P. Sloan Foundation, the National Institutes of Health, and Ohio State University’s College of Arts and Sciences, Center for Brain Injury Recovery and Discovery, and Women in Philanthropy award.
Key Questions Answered:
A: Not according to this research. Because the brain architecture for language and Theory of Mind is discrete, a delay in one domain does not necessarily imply a delay in the other. The systems are distinct and can develop at different rates.
A: Researchers present scenarios where a character holds a false belief—for example, searching for a toy in the wrong place because it has been moved. Understanding the character’s mistaken belief requires the viewer to adopt the character’s perspective, engaging mentalizing processes.
A: As social life becomes more complex, adults frequently combine language and social reasoning—for persuasion, comfort, instruction, and negotiation. Increased communication between the networks likely reflects the behavioral need to integrate language and empathy during mature social interactions.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- Journal paper reviewed in full.
- Additional context added by staff.
About this theory of mind and neurodevelopment research news
Author: Emily Caldwell
Source: Ohio State University
Contact: Emily Caldwell – Ohio State University
Image: The image is credited to Neuroscience News
Original Research: Open access. “Functional dissociation of language and theory of mind in the developing superior temporal lobe” by Kelly J. Hiersche, David E. Osher & Zeynep M. Saygin. DOI: 10.1038/s42003-026-10040-2
Abstract
Functional dissociation of language and theory of mind in the developing superior temporal lobe
Language and Theory of Mind (the ability to infer others’ mental states) are both essential for human communication, but their developmental origins have been debated. Do these functions occupy distinct neural substrates in the superior temporal lobe with opposite lateralization, as in adults, or do they emerge from shared substrates early in development?
This study examined neural specificity and connectivity fingerprints in a cohort of children ages 3–9 (54 sessions, 42 subjects) and adults (28 subjects). Children displayed distinct neural responses for language and Theory of Mind in the superior temporal lobe, mirroring adult patterns. Cross-sectional and longitudinal analyses revealed no evidence that the functions begin overlapping and later separate; instead, their neural identities are present early and remain stable across childhood.
Connectivity fingerprints predicting future language or Theory of Mind activation were nearly identical to concurrent fingerprints and were largely non-overlapping across domains. While both systems continue to refine toward an adult-like state, they are already remarkably distinct in early development. These results challenge models that propose language develops from general social processors and instead support separate neural origins for linguistic and mental-state reasoning abilities.