Reading Minds: What Neuroscience Tells Us About Theory of Mind

Imagine you are four years old. A researcher shows you two dolls: Sally and Anne. Sally places a marble in a basket, then leaves the room. While she is gone, Anne moves the marble to a box. Sally comes back. The researcher turns to you and asks: where will Sally look for her marble?

If you are four, there is a good chance you will point to the box, where the marble physically is. If you are five, you will probably point to the basket, where Sally left it, and where she believes it still is. This deceptively simple scenario, first formalised by Baron-Cohen, Leslie, and Frith in 1985, is not testing memory, attention, or language. It is testing whether a child understands that another person can hold a belief about the world that differs from reality, and from their own theory of mind.

What is theory of mind?

The term was first introduced by Premack and Woodruff in 1978, in a paper asking whether chimpanzees could attribute mental states to others. Their answer was cautious. The question, however, proved transformative.

Theory of mind (ToM) refers to the capacity to understand that others have beliefs, desires, intentions, and emotions that may differ from our own. It allows us to predict behaviour, navigate social situations, recognise deception, feel empathy, and understand why someone might act against their own apparent interests. Researchers distinguish between first-order ToM, or understanding what someone else believes, and second-order ToM, which involves reasoning about what one person thinks another person believes. Both are considered essential components of fully developed social cognition.

How does it develop?

The developmental story of ToM begins earlier than most people expect. Before the age of four, children typically fail the false belief task. When asked where Sally will look for her marble, they indicate where the marble is, not where Sally left it. They cannot yet separate their own knowledge of the world from someone else's perspective. Wimmer and Perner’s landmark study established this pattern clearly, showing that the shift in understanding false beliefs occurs reliably around the age of four to five.

"(...) different nodes of the ToM network are differentially engaged depending on whether a person is reading the scenario, forming a belief attribution, or preparing a response"

But ToM does not emerge from nowhere; it develops gradually, with children progressively acquiring the ability to interpret a range of mental states across the preschool years. Simpler precursors appear much earlier: joint attention, gaze following, and the understanding of intention are all present in infancy and toddlerhood and are thought to scaffold the more complex capacities that emerge later.

Second-order ToM develops later (typically between the ages of 6 and 8) and continues to be refined well into adolescence. The implication is that theory of mind is not a single cognitive switch, but a set of related abilities that build on one another across development.

What drives this development is still debated. Some researchers argue for a dedicated, domain-specific mechanism, known as the Theory of Mind Mechanism (ToMM) that matures on a biological schedule. This is the modularity account: ToM is hardwired, and development reflects its unfolding rather than its construction. Others favour the theory account, which proposes that children build an implicit understanding of how minds work gradually through observation and experience, in much the same way a scientist builds a theory through accumulating evidence. A third view, simulation theory, suggests that we understand others’ minds not by applying rules but by mentally putting ourselves in another person's position, running a kind of internal simulation. Each makes different predictions, and none has yet decisively won.

What happens in the brain?

Neuroimaging has given us a detailed picture of which brain regions are recruited when we reason about other minds. A network of regions is consistently implicated in ToM tasks: the temporoparietal junction (TPJ), the superior temporal sulcus (STS), the precuneus, the temporal poles, and the medial prefrontal cortex (mPFC). These regions do not all do the same thing. The TPJ, particularly on the right side, appears to be especially important for representing the beliefs and perspectives of others (distinct from one's own) while the mPFC is thought to be involved in general social reasoning and monitoring the difference between self and other perspectives.

"The understanding of intention are all present in infancy and toddlerhood and are thought to scaffold the more complex capacities that emerge later"

A major meta-analysis by Schurz and colleagues covering dozens of neuroimaging studies confirmed a core network active across all ToM tasks, centred on the mPFC and TPJ, while also revealing that different types of ToM tasks recruit additional regions depending on the demands involved. A 2024 fMRI study went further, examining distinct phases of belief processing and finding that different nodes of the ToM network are differentially engaged depending on whether a person is reading the scenario, forming a belief attribution, or preparing a response.

What this tells us is that theory of mind is not a single, localised function, but a dynamic process, distributed across networks, and sensitive to the specific cognitive demands of the social situation.

When theory of mind breaks down

Understanding the neural basis of ToM has important clinical implications, because ToM does not always develop typically and can be selectively impaired. For example, in autism spectrum disorder, difficulties with ToM are consistently reported. Neuroimaging studies and meta-analyses have shown reduced activation in key ToM regions, including the mPFC, anterior cingulate cortex, amygdala, and TPJ, when autistic individuals perform ToM tasks compared to neurotypical controls. This does not mean that autistic people lack empathy or social interest; the relationship between ToM as measured in the lab and social experience in everyday life is considerably more complicated. 

In schizophrenia, ToM impairments are thought to contribute significantly to the social difficulties characterizing the condition. fMRI research has shown reduced recruitment of the mPFC during ToM tasks in individuals with schizophrenia, and correlations with real-world social functioning. 

ToM can also be disrupted by neurodegeneration. In Alzheimer's disease, ToM deficits emerge as the condition progresses, reflecting the gradual loss of brain regions involved in social cognition. In frontotemporal dementia, which disproportionately affects frontal and temporal lobes, changes in social behaviour and empathy can appear early and be among the most prominent features of the disease.

These clinical observations matter not only for understanding these conditions, but for developing better interventions. If ToM difficulties in schizophrenia are linked to reduced mPFC recruitment, can targeted cognitive training improve both the neural signal and social outcomes? Early evidence suggests it can, and this is an active area of research.

Future directions

Despite decades of research, fundamental questions remain. The most basic is definitional: what exactly is ToM? Is it a dedicated cognitive module, as Leslie proposed? A simulation, where we model others' minds? Or a set of implicit rules about how minds work that we apply to predict behaviour? Each account makes different predictions about brain organisation, developmental trajectory, and patterns of impairment, and no consensus has yet emerged.

"In Alzheimer's disease, ToM deficits emerge as the condition progresses, reflecting the gradual loss of brain regions involved in social cognition."

A more recent distinction, between hot and cold ToM, has added further complexity. Cold ToM refers to the explicit, deliberate reasoning about others' beliefs that the false belief task captures. Hot ToM involves the same capacity applied in emotionally charged, real-world situations. The two do not always correlate, and individuals who pass laboratory ToM tasks can still struggle significantly in social contexts that engage emotional and motivational systems alongside purely cognitive ones.

Looking ahead, two directions seem particularly promising. The first is the integration of ToM research with artificial intelligence. By understanding how machines can or cannot model mental states, and what the limitations of current AI ToM are, it can also tell us about the uniqueness of human social cognition. The second is the study of individual differences in ToM across the lifespan in typically developing adults, an area that has been relatively understudied compared to the developmental and clinical literature.

Theory of mind is, in many ways, the cognitive ability that makes human social life possible. The capacity to step outside our own perspective and model what someone else believes, intends, or feels is so automatic that we rarely notice it until it goes wrong. What neuroscience has given us is a picture of how this capacity is built, how it unfolds, and what its disruption means for social cognition. The question of how exactly the brain does it remains open.

References

1. Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does the autistic child have a "theory of mind"? Cognition, 21(1), 37–46.

2. Golec-Staśkiewicz, K., Wojciechowski, J., Haman, M., Wolak, T., Wysocka, J., & Pluta, A. (2024). Unveiling the neural dynamics of the theory of mind: a fMRI study on belief processing phases. Social Cognitive and Affective Neuroscience, 19(1), nsae095.

3. Leslie, A. M. (1987). Pretense and representation: The origins of "theory of mind." Psychological Review, 94(4), 412–426.

4. Premack, D., & Woodruff, G. (1978). Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences, 1(4), 515–526.

5. Schurz, M., Radua, J., Aichhorn, M., Richlan, F., & Perner, J. (2014). Fractionating theory of mind: A meta-analysis of functional brain imaging studies. Neuroscience & Biobehavioral Reviews, 42, 9–34.

6. van Neerven, T., Bos, D. J., & van Haren, N. E. (2021). Deficiencies in theory of mind in patients with schizophrenia, bipolar disorder, and major depressive disorder: A systematic review of secondary literature. Neuroscience & Biobehavioral Reviews, 120, 249–261.

7. Wellman, H. M., Cross, D., & Watson, J. (2001). Meta-analysis of theory-of-mind development: The truth about false belief. Child Development, 72(3), 655–684.

8. Wimmer, H., & Perner, J. (1983). Beliefs about beliefs: Representation and constraining function of wrong beliefs in young children's understanding of deception. Cognition, 13(1), 103–128.

9. Yeung, E. K. L., Apperly, I. A., & Devine, R. T. (2024). Measures of individual differences in adult theory of mind: A systematic review. Neuroscience & Biobehavioral Reviews, 157, 105481.

---

This article was written by Laura Bilbao Broch and edited by Julia Dabrowska, with graphics produced by Saba Keshan. If you enjoyed this article, be the first to be notified about new posts by signing up to become a WiNUK member (top right of this page)! Interested in writing for WiNUK yourself? Contact us through the blog page and the editors will be in touch.