Developmental origins of theory of mind in the brain

Li, Shuting (2025): Developmental origins of theory of mind in the brain: neural correlates of false belief understanding from infancy to early childhood. Dissertation, LMU München: Fakultät für Psychologie und Pädagogik

Abstract

An essential component of human social interaction and communication is the ability to comprehend mental states, encapsulated in the concept of “Theory of Mind” (Harman, 1978; Perner, 1991). Theory of Mind refers to understanding mental states such as beliefs, desires, and intentions and their influence on behavior (Perner, 1991; Wellman et al., 1990). Over the past 40 years, the exploration of how children develop a Theory of Mind has been a central focus in developmental psychology (Rakoczy, 2022; Sodian et al., 2020). A key aspect of Theory of Mind that has received extensive attention is the false belief understanding—the understanding that others may hold beliefs about the world that differ from one’s own knowledge of reality. Seminal empirical research and a comprehensive meta-analysis demonstrated that children typically begin to attribute false beliefs to themselves and others in the fourth years of life (Wellman et al., 2001; Wimmer & Perner, 1983) when they start to acquire a proper notion of mental representation (Perner, 1991). These traditional findings of a marked age-related trend support the view that conceptual changes in false belief understanding occur at approximately 4 years of age (e.g., Wellman et al., 2001). However, the claim that the false belief understanding undergoes a dramatic conceptual change at this age has been contested (Baillargeon et al., 2010). Behavioral evidence challenging the conceptual change perspective has been demonstrated in both implicit and low-demands explicit false belief tasks. Implicit false belief tasks involve inferring children’s understanding of an agent’s false belief from their spontaneous behaviors as they observe scenes unfold (Baillargeon et al., 2010; Onishi & Baillargeon, 2005), while explicit false belief tasks require participants to provide direct responses to questions about an agent’s false belief (Baron-Cohen et al., 1985; Wellman et al., 2001; Wimmer & Perner, 1983). For instance, nonverbal spontaneous-response paradigms revealed that infants (under 2 years old) and toddlers (ages 2–3 years) exhibit implicit reasoning about false beliefs (e.g., Clements & Perner, 1994; Schneider et al., 2012; Scott & Baillargeon, 2017; Southgate et al., 2007). Moreover, recent research utilizing a behaviorally low-demands false belief task suggests that toddlers aged 2.5 years can grasp explicit false beliefs when task processing demands related to response generation and inhibitory control are reduced properly (Grosso et al., 2019; Setoh et al., 2016; Sodian et al., 2024). If infants and toddlers already exhibit sensitivity to others’ false beliefs, it is plausible that the cortical regions involved in false belief understanding are engaged in these socio-cognitive functions, not only during toddlerhood but potentially as early as infancy. Earlier approaches to exploring false belief understanding were primarily focused on the behavioral level, whereas recent approaches have further investigated its neuroscientific basis. Over the last two decades, significant progress has been made toward identifying a specialized neural system associated with false belief understanding, primarily focusing on adults and children at 6 to 12 years of age (e.g., Gweon et al., 2012; Meinhardt et al., 2011; Sommer et al., 2010). Neuroimaging studies revealed that the brain regions implicated in false belief understanding are primarily located in the frontal regions (e.g., Geangu et al., 2013; Sommer et al., 2010) and the posterior regions, often measured in parietal or parieto-occipital regions (e.g., Grosse Wiesmann et al., 2020; Perner et al., 2006). Researchers have begun exploring the neural correlates of this socio-cognitive ability in children under 6 to address important open questions in the development of false belief understanding. Although such studies remain insufficient, they provide important insights (e.g., Grosse Wiesmann et al., 2020; Richardson et al., 2018; Richardson & Saxe, 2020). For example, Richardson et al. (2018) and Moraczewski et al. (2018) observed that in children aged 3 to 6, the temporal dynamics of the mentalizing network closely mirrored those seen in adults, contrasting with other control networks. Grosse Wiesmann et al. (2017b) reported increased connectivity in tracts surrounding mentalizing regions, including the medial prefrontal cortex (MPFC) and the temporoparietal junction (TPJ), in 3- and 4-year-olds. Furthermore, Hyde et al. (2018) demonstrated that the right TPJ is preferentially active when children are thinking about others’ thoughts, even in infants as young as 7 months. Existing magnetic resonance imaging (MRI) studies have aimed to “localize” neural regions involved in false belief reasoning with high spatial resolution. However, due to the limited temporal resolution of MRI techniques, these studies offer only a partial understanding of how these regions contribute to the cognitive processes underlying false belief understanding. To gain further insights into the temporal dynamics of these processes, electroencephalogram (EEG) techniques, which provide higher temporal precision, may offer a complementary approach. A notable event-related potential (ERP) study has examined belief reasoning in children around 4 years of age, a developmental milestone associated with the emergence of traditional explicit false belief understanding (Liu et al., 2009b). Liu et al. (2009b) compared neural responses between children aged 4-6 years and adults during tasks involving belief attribution to story protagonists and reality judgments. Children were grouped into “passers” and “failers” based on an independent assessment of their behavioral false belief competence. A late waveform was observed over frontal regions only for child “passers,” with a less localized and more diffuse scalp distribution than in adults. Children who failed the behavioral false belief task showed no systematic differentiation between belief and reality conditions on the neural level. To date, this study provides preliminary evidence of brain-behavior connections in acquiring false belief understanding in early childhood. However, Liu et al. (2009b) did not differentiate between the false and true beliefs, treating them as a combined condition in the ERP belief task. Given that children demonstrated explicit false belief competence in a low-demands false belief task, it is necessary and critical to investigate whether a specialized neural system supports false belief processing in younger children. This dissertation presents a first step in extending the line of Liu et al.’s (2009b) study to children under the age of 3 years; it takes a more direct and rigorous approach to examining the neural basis of false belief understanding. Furthermore, it is important to acknowledge that task-dependent and task-independent neural techniques differ fundamentally in their approach and should not be equated when interpreting findings across studies. Task-dependent ERP techniques involve recording the time-locked brain activity to a specific external stimulus designed to elicit a false belief judgment (Sabbagh, 2013). In contrast, task-independent techniques, like resting-state EEG, measure brain activity without externally induced stimuli, focusing instead on intrinsic neural processes (MacLean et al., 2012). They highlight the degree of integration processes within the brain (i.e., brain coherence, Aykan et al., 2021) or hemispheric asymmetry in brain activity (e.g., Licata et al., 2015), which are relevant for understanding socio-cognitive and emotional processes. Differences in frontal and parietal asymmetry might indicate varying levels of engagement in ToM-related processes, which are critical for understanding false beliefs (N. A. Fox et al., 1995; Stewart et al., 2011). For example, a resting-state source-localized EEG study provided evidence supporting the right-hemispheric lateralization of brain activity linked to false belief understanding (Sabbagh et al., 2009). Sabbagh et al. (2009) identified individual differences in alpha oscillations in the dorsal MPFC and several right hemisphere areas—including the TPJ, precentral gyrus, cuneus, and inferior temporal cortex—were associated with representational Theory of Mind performance in 4-year-old children. Therefore, beyond examining task-dependent neural correlates of false belief understanding in toddlers, investigating task-independent neural correlates at around 4 years of age or even younger may facilitate the study of intrinsic brain networks without the influence of explicit tasks. This task-independent neural method could offer an additional understanding of how individual differences in resting-state brain activity during early childhood are associated with the subsequent development of false belief understanding (e.g., Sabbagh et al., 2009). This dissertation employed both task-dependent and task-independent methods to investigate the neural correlates of false belief understanding from infancy to early childhood. This dissertation aims to address two overarching research questions that advance our understanding of false belief reasoning in early childhood: (1) What neural-behavioral connections underpin false belief understanding in toddlers under 3 years of age? (2) How does the specialized neural system associated with false belief understanding emerge? Specifically, does this specialized neural system develop because of behavioral false belief competence, or is this system already functional and active before behavioral competence becomes observable? To address these questions, this thesis conducted three empirical studies among children aged 14 to 52 months, and the results are presented in subsequent chapters (Chapter 2 and Chapter 3). Study 1 is a cross-sectional investigation into the brain-behavior connections associated with false belief understanding in 33- to 36-month-old toddlers. A multi-trial explicit false belief task was designed to be compatible with the task-dependent ERP methodology. Toddlers were grouped into passers and failers according to their performance on a low-demands behavioral false belief task (see Setoh et al., 2016). This study aims to examine the neural basis of explicit false belief understanding in this age group and explore potential brain-behavior connections associated with it. The findings from Study 1 revealed distinct neural patterns associated with false belief competence in 33- to 36-month-old children. Specifically, children who demonstrated false belief competence exhibited a more bilaterally diffused occipital positive late waveform in the false belief condition compared to their peers without false belief competence. Additionally, a late negative waveform, predominantly over right-lateralized frontocentral sites, consistently differentiated the false belief from the true belief condition, regardless of performance on low-demands behavioral false belief task. These findings provide evidence of neural correlates associated with false belief competence in children under 3 years old. Furthermore, they highlight a developmental pattern linking occipital positive late waveforms to the development of false belief understanding. Studies 2 and 3 adopted a longitudinal design to investigate the relationship between resting-state EEG alpha asymmetry in infants and toddlers and their false belief understanding at around 3 and 4 years of age. This investigation distinguished between explicit and implicit false belief understanding. Building on evidence suggesting that cortical network supporting false belief understanding may begin to develop as early as 3 years of age (e.g., Grosse Wiesmann et al., 2020; Richardson et al., 2018; Richardson & Saxe, 2020), and the potential right-hemispheric lateralization of brain activity associated with false belief understanding in infants and toddlers (Hyde et al., 2018; Sabbagh et al., 2009), following exploratory research questions are posed: (1) Is there a correlation between brain asymmetric activity observed during resting-state EEG recording and false belief understanding? (2) Does resting-state EEG asymmetry precede the development of representational false belief understanding, and can it predict later behavioral performance in false belief tasks? Employing a longitudinal design, Study 2 assessed resting-state EEG alpha asymmetry across frontal and parietal electrode sites at 34 months, explicit false belief understanding at 52 months, and implicit false belief understanding at both time points. Study 3 analyzed data from another independent longitudinal dataset to test the generality of the relationship between resting-state EEG alpha asymmetry (assessed at 14 months) and explicit false belief understanding at 51 months old. Results from both studies showed that better explicit false belief understanding at age 4 was associated with greater right than left frontal activity at 14 and 34 months. Better implicit false belief understanding was only cross-sectionally associated with greater relative right than left parietal activity at 34 months. These findings suggest resting-state EEG alpha asymmetry may be a stable early-developing neural marker in explicit false belief understanding. Additionally, these findings tentatively suggest that implicit false belief understanding may not follow a monotonic development across childhood and that implicit and explicit false belief understanding may develop based on partly distinct neural mechanisms. Overall, this dissertation makes novel contributions to identifying the task-dependent and task-independent neural correlates of false belief understanding from infancy to early childhood through electrophysiological techniques. In Study 1, toddlers’ behavioral competencies in a low-demands false belief task were significantly correlated with neural responses linked to false belief understanding. This study is the first to investigate task-dependent ERP correlates of false belief understanding in children aged 3 years and younger, raising the possibility that a sensitive neural system supporting false belief understanding may emerge early in development. Studies 2 and 3 further indicate that resting-state EEG alpha asymmetry assessed at infancy and toddlerhood may serve as an early neural precursor to subsequent explicit false belief understanding at 4 years of age—the typical age when explicit false belief understanding emerges. These studies offer preliminary evidence for a neural marker that reliably predicts individual differences in false belief understanding longitudinally. Additionally, comparisons between Study 1 and subsequent Studies 2 and 3 suggest that the brain–behavior connections associated with false belief understanding in preschool-aged children are not merely the result of the short-term acquisition of behavioral competence; Instead, they may reflect the early developmental (ontogenetic) and possibly evolutionary (phylogenetic) emergence of the Theory of Mind network. Collectively, these studies bridge a critical gap between the extensive research on the early behavioral development of false belief understanding and its corresponding neural profiles. Brain regions associated with Theory of Mind exhibit early signs of functional specialization even before children succeed in explicit false belief tasks. However, these findings are exploratory and require further validation, primarily through future research exploring the neural mechanisms underpinning both implicit and explicit false belief understanding in children younger than 4 years of age across multiple developmental stages.

• Dokumententyp: Dissertationen (Dissertation, LMU München)
• Keywords: Theory of Mind, false belief understanding, event-related potential (ERP), resting-state EEG alpha asymmetry, longitudinal study, toddlers
• Themengebiete: 100 Philosophie und Psychologie > 150 Psychologie
• Fakultäten: Fakultät für Psychologie und Pädagogik
• Sprache der Hochschulschrift: Englisch
• Datum der mündlichen Prüfung: 21. Februar 2025
• 1. Bericht­erstatter:in: Sodian, Beate