Theory of mind is the developing capacity to understand that people act in light of mental states, including desires, perceptions, intentions, knowledge, emotions, and beliefs. A child uses this understanding to explain why another person reaches for a toy, looks for an object in the wrong place, feels disappointed despite receiving a gift, or interprets the same event differently. The essential developmental achievement is recognizing that another person's representation of a situation can differ from both reality and the child's own representation. Research on theory of mind therefore examines how children understand minds, how that understanding changes, and how language, experience, cognition, and culture contribute to social reasoning (Wellman, Cross, and Watson, 2001; Flavell, 1999).
Theory of mind develops through multiple, partially overlapping achievements across infancy, preschool, and middle childhood. Many children succeed on conventional verbal false-belief tasks during the preschool years, often around ages four to five, while precursors of understanding goals, attention, and preferences appear earlier. More complex reasoning about what one person thinks another person believes continues developing through school age. These ages describe study-level trends, not diagnostic deadlines or a fixed timetable for each child (Wellman and Liu, 2004; Hughes and Devine, 2015).
What Theory of Mind Means in Developmental Psychology
Mental states, representations, and behavior
Theory of mind, often abbreviated ToM and also called mental-state understanding or mentalizing in some research traditions, denotes an ability to attribute psychological states to people and use those attributions to interpret their actions. A child who says that a sibling is searching under the couch because she thinks her lost book is there has made a belief-based explanation. A child who realizes that two friends can want different games recognizes that desires vary across individuals. A child who understands that someone who did not see a surprise cannot know about it recognizes the limits of informational access. These accomplishments involve related judgments, but they are distinct achievements rather than one interchangeable skill (Flavell, 1999; Wellman and Liu, 2004).
The phrase “theory” does not mean that a child consciously constructs a formal scientific theory. It refers to a structured explanatory understanding of the invisible mental processes that help make observable behavior intelligible. Social reasoning includes direct perception of expression and action, memories of relationships, inference about unseen goals, culturally learned expectations, and increasingly explicit representations of conflicting perspectives. No single test or experimental paradigm encompasses all these processes.
Belief, knowledge, desire, intention, and emotion
Desires concern preferred outcomes: one person wants strawberries, another wants grapes. Beliefs concern what a person takes to be true, including beliefs that happen to be mistaken. Knowledge additionally involves informational access and what someone can reasonably know. Intentions concern the goals toward which actions are directed, while emotions are affective responses whose causes depend on situations, expectations, relationships, and interpretations. A child's progress includes integrating these different states: disappointment arises when a desired event fails, surprise follows an unexpected event, and an apparently puzzling action may become reasonable from another person's mistaken point of view (Flavell, 1999).
Belief reasoning has been especially influential because it tests whether children can represent someone else's interpretation of the world even when the child knows that interpretation is false. It provides a rigorous window into representational understanding. Nevertheless, a preschooler's performance on that window cannot be equated with the entirety of empathy, honesty, kindness, social adjustment, attachment, or moral development.
Theory of mind, perspective taking, empathy, and social cognition
Perspective taking is a broader family of skills. Visual perspective taking involves recognizing that two observers can see different sides of the same object. Cognitive perspective taking concerns another person's information or interpretation. Affective perspective taking concerns another person's emotional experience. Empathy commonly includes processes of affective resonance or concern, together with mechanisms of understanding. Theory of mind often contributes to cognitive perspective taking, but emotion sharing, compassionate action, and successful mental-state inference can vary independently. A child can correctly infer that a friend feels excluded and still need guidance about what to do next; another child may care deeply about distress while struggling to explain why it arose.
Social cognition additionally encompasses reading social cues, recognizing relationships and norms, interpreting intentions, memory for people, learning from interaction, and evaluating reputations. In this larger territory, theory of mind denotes a distinctive explanatory construct. Research on broader social development in children examines multiple domains of interaction; the present article concentrates on mental-state reasoning and its developmental evidence.
Historical Foundations and Contemporary Explanations
The emergence of belief reasoning as a research question
Developmental psychologists have long investigated children's understanding of perspectives, knowledge, and appearances. The modern theory-of-mind literature gave special prominence to tasks in which children must predict actions based on another person's false belief. Wimmer and Perner's influential unexpected-transfer paradigm and subsequent unexpected-contents tasks created reproducible experimental situations in which the child's knowledge could be placed in conflict with the protagonist's belief. Researchers then examined which developmental mechanisms explain the striking preschool transition and why modifying task wording, memory load, or social context changes performance (Wellman, Cross, and Watson, 2001).
This research established false-belief understanding as a central empirical question while also encouraging a wider conception of children as interpreters of intentional action. The history of the field is therefore a history of improving explanatory models and measurement rather than discovering a single age at which social understanding suddenly appears.
Theory-theory, simulation, and specialized-processing accounts
Theory-theory accounts propose that children gradually revise organized concepts about how the mind works. A younger child may rely strongly on observable reality when predicting another person's behavior; later the child explicitly represents that a person's belief can misrepresent reality. Simulation accounts emphasize using one's own psychological processes to imaginatively model another person's point of view, while taking account of differences between self and other. Specialized-processing accounts consider whether early-developing mechanisms support sensitivity to intentional agents or belief-like information. Contemporary work also considers Bayesian learning, domain-general executive resources, language-mediated representations, and interactive social experience. These models propose partly different mechanisms and levels of explanation rather than mutually exclusive lists of childhood skills (Flavell, 1999; Asakura and Inui, 2016).
The strongest empirical conclusions concern observable developmental patterns, task conditions, and measured relationships among language, control, and mental-state reasoning. Claims that a single underlying theoretical mechanism has been conclusively established exceed what the diverse results presently support. Different tasks can recruit different combinations of conceptual knowledge, attention, linguistic processing, and social expectations.
A developmental-systems interpretation
Theory-of-mind development can be investigated as an outcome of coordinated processes: maturation of attention and memory, language learning, growing experience with distinct preferences, conversations about perspectives, opportunities for disagreement and repair, and the child's own participation in relationships. Children influence the kinds of conversations and interactions adults offer them. Caregivers, siblings, peers, and teachers respond to the child's interests and communicative skills, which in turn change developmental opportunities. This transactional perspective helps explain individual variability without reducing development to one parenting practice or one isolated cognitive module. Identifying a plausible developmental pathway, however, does not by itself establish its causal contribution.
Development Across Infancy, Toddlerhood, and Preschool
Infancy: intentional action, attention, and early expectations
During infancy, researchers observe developing attention to people, sensitivity to goal-directed movement, and responses to what others see or attend to. These abilities prepare children to participate in shared activity and information exchange. Infants may differentiate familiar actions from unexpected actions and follow another person's gaze; such outcomes provide evidence of early social processing. They do not automatically prove possession of the explicit, flexible belief concepts demonstrated by older children on verbal tasks. Research into serve and return interactions and joint attention concerns related developmental foundations when the linked article is available, but social engagement and full false-belief reasoning should remain conceptually distinct.
An influential 2005 experiment interpreted longer looking by 15-month-old infants at certain events as sensitivity to an actor's mistaken belief (Onishi and Baillargeon, 2005). This launched an extensive debate: some researchers argue that infant looking patterns index early representational belief understanding; others emphasize alternative explanations, unmeasured expectations, and difficulties replicating the results across paradigms. A preregistered replication program found serious robustness and convergent-validity problems among several implicit anticipatory-looking tasks (Kulke et al., 2018). A later review describes mixed evidence and discontinuities between implicit and explicit measures (Poulin-Dubois et al., 2023). Consequently, early sensitivity to agents is established more securely than the proposition that infants possess the same mature belief concept measured in school-age children.
Toddlerhood: preferences, desires, and informational access
In the second and third years, everyday conversation and play increasingly reveal appreciation of goals, preferences, and emotion-related outcomes. A toddler may choose one snack for themselves while giving another person the item that person prefers. Understanding that different individuals have different desires is an important step because it loosens the assumption that one's own preference is universal. Children also begin to appreciate that seeing, hearing, being present, and being told something affect what another person knows. These developments depend partly on how an experiment communicates the task and how the child can express an answer. Success on a simple desire task and difficulty on a belief task is a coherent developmental profile, not inconsistency (Wellman and Liu, 2004).
Emerging language supports more detailed contrasts between states: want versus have, know versus guess, think versus see. Daily routines supply situations in which another person's desire differs from the child's own, a person does not know where an object was moved, or two people describe the same event differently. Adults can name these differences in ordinary language without turning them into tests.
Ages three to five: the classic false-belief transition
In a standard location-change task, a character puts an object in a box and leaves. The object is moved to a drawer without the character seeing. The child knows the new location and is asked where the character will first look upon returning. Predicting the original box requires reasoning from the character's mistaken belief rather than from the child's own accurate knowledge. Children who answer with the drawer may be relying on reality, misunderstanding the question, losing track of the scenario, or struggling to inhibit the correct-location response. The task therefore measures performance that integrates mental-state understanding with additional demands.
A foundational meta-analysis reviewed the age-related pattern across false-belief studies and found a robust developmental change during the preschool period, while confirming that task manipulations affect difficulty (Wellman, Cross, and Watson, 2001). The conventional shorthand that children “develop theory of mind at age four” captures a rough transition in explicit false-belief task performance. It oversimplifies earlier desire and knowledge understanding, later growth, cultural variation, and the substantial difference between passing one laboratory task and understanding people across unfamiliar social settings.
Ages five to seven: integration and conversational flexibility
As children move through the early elementary years, many become better at coordinating beliefs with intentions, emotional outcomes, and contrasting sources of information. They can explain that a person is surprised because a prior expectation proved wrong or that a friend acted incorrectly because a message was misunderstood. They can increasingly identify when someone knows something only because they saw it and when another person's claim rests on an incorrect assumption. Language and task conditions continue to influence how well these capacities are expressed. Advances also allow children to recognize more subtle distinctions between apparent emotion and experienced emotion, although the pace differs across contexts and individuals.
Middle childhood: second-order beliefs and social interpretation
Middle childhood brings opportunities to reason about nested mental states: “Maya thinks that Leo believes the surprise party is tomorrow.” Such second-order representations support interpretations of misunderstandings, strategic communication, irony, social exclusion, and reputational concerns. They also place heavier demands on language and working memory. More advanced mental-state understanding is neither reducible to passing a second-order puzzle nor guaranteed to translate into tactful behavior during conflict. In a longitudinal study following 137 children across approximately four years, individual differences in theory of mind were moderately stable, and earlier theory-of-mind performance predicted later teacher-rated social competence; the same study did not find longitudinal links between theory of mind and executive function despite concurrent associations (Devine et al., 2016).
Research therefore treats development beyond preschool as substantively important. A child can pass basic false-belief tests yet still refine the ability to coordinate multiple perspectives in emotionally charged, ambiguous, or culturally unfamiliar situations (Hughes and Devine, 2015).
How Researchers Measure Children's Understanding of Minds
Unexpected-location tasks
Unexpected-location or change-of-location tasks present a protagonist whose information becomes outdated while the child observes a change. The question probes whether the child predicts behavior from the protagonist's perspective. Researchers often add control questions about where the object originally was and where it is now, helping distinguish memory or comprehension problems from the target belief judgment. Performance depends on the character's visibility, linguistic clarity, whether the protagonist actually witnessed the transfer, and how the child understands the response request. The famous Sally-Anne format is one example of this task family, not a clinical diagnostic instrument.
Unexpected-contents and appearance-reality tasks
In an unexpected-contents task, a familiar package contains an unexpected item. After discovering the contents, the child is asked what another person, who has not looked inside, will think the package contains. Some versions ask what the child themselves initially thought, probing memory for a prior false belief. Appearance-reality tasks distinguish how an object looks from what it actually is. These measures reveal the importance of representing alternatives to immediately available reality, but they are sensitive to vocabulary, memory, conversational expectations, and the particular structure of the question (Wellman, Cross, and Watson, 2001).
Sequences and scales of mental-state tasks
One response offers a narrow snapshot. Task batteries examine understanding of diverse desires, diverse beliefs, knowledge access, false belief, and hidden emotion, producing a more informative developmental profile. Wellman and Liu demonstrated a broadly ordered progression across several tasks in preschool-aged children. Later cultural and population studies indicate that the precise ordering of particular skills can change according to communicative context and social experience (Wellman and Liu, 2004; Shahaeian et al., 2011). A scale can be useful for research while still requiring careful validation before anyone treats an individual score as an assessment of overall social capacity.
Nonverbal, implicit, and naturalistic measures
Researchers also measure looking time, anticipatory eye movements, spontaneous pointing, facial reactions, story interpretation, and behavior in real interactions. These methods address different questions and sometimes reduce explicit speech demands. Yet “nonverbal” does not mean “free of interpretation.” Eye movements can reflect attention, associative learning, perceptual novelty, or task expectations. A child's spontaneous helping behavior can be influenced by motivation, familiarity, safety, and local norms. The most defensible approach triangulates tasks and avoids assigning one unobserved psychological mechanism to every outward response (Kulke et al., 2018; Poulin-Dubois et al., 2023).
Reliability, validity, and developmental interpretation
Assessment requires clarity about what an instrument measures, how consistently it measures it, and whether comparisons are meaningful across age, language, disability, and culture. False-belief tasks can be informative when administered under controlled conditions, but differences in comprehension, response modality, fatigue, and familiarity with narrative testing alter conclusions. A test that predicts average developmental patterns may still be a poor stand-alone measure for an individual child's capacities. Researchers should report population characteristics, task adaptations, scoring, reliability, and the uncertainty of interpretations rather than silently treating a wrong answer as a stable personal deficit.
Language and Conversation as Developmental Resources
Vocabulary for mental experience
Words such as think, guess, know, hope, pretend, forget, and wonder make distinctions among mental states communicable. Children can learn to explain why people differ by comparing what each saw or heard and what each consequently believed. Vocabulary is one component of a wider communicative system: receptive comprehension, sentence structure, narratives, conversational inference, and the ability to understand that an utterance expresses someone else's perspective all contribute. In a major meta-analysis, language ability and false-belief understanding were reliably associated across studies; the relationship cannot be interpreted as proof that vocabulary instruction alone produces full mental-state understanding (Milligan, Astington, and Dack, 2007).
Syntax, embedded propositions, and competing perspectives
Certain language constructions permit a proposition to be embedded as someone else's belief: “She thinks that the keys are upstairs,” even when the keys are downstairs. These sentences distinguish the truth of what is believed from the truth of the world. Researchers propose that mastery of relevant syntax may provide representational tools that facilitate explicit belief reasoning. At the same time, children who better understand perspectives may participate in more complex conversations, which improves language. Statistical associations, longitudinal prediction, and language-focused training support related but not identical hypotheses about the causal direction of this relationship (Milligan, Astington, and Dack, 2007).
Mental-state talk in families and classrooms
Talking about why someone was confused, disappointed, excited, or mistaken can bring invisible mental processes into shared reflection. A longitudinal study linked caregivers' early use of mental-state language to children's later theory-of-mind performance after adjusting for several measured variables (Ruffman, Slade, and Crowe, 2002). Although its authors proposed a causal interpretation, an observational design cannot fully exclude unmeasured family and child characteristics. The study more securely supports the conclusion that the conversational environment is a consequential developmental correlate with plausible routes for influence.
Adults can encourage explanation without requiring polished answers. “Why does the character think the bag is empty?” or “What does your friend know about the plan?” invites children to connect access to information with belief. Discussions remain more useful when they follow the child's attention, use comprehensible language, and allow disagreement and correction rather than evaluating the child continuously. Research on language development in children provides a separate, broader account of receptive and expressive growth.
Executive Function, Attention, and Cognitive Control
Inhibition and the problem of privileged knowledge
False-belief tasks put the child's correct knowledge into competition with the protagonist's mistaken knowledge. Answering from the other's point of view may require suppressing a salient response based on reality. Executive function encompasses inhibitory control, working memory, and flexible shifting among rules or representations; these functions influence the task's cognitive load. A meta-analysis of 102 studies, including 9,994 children aged three to six, found moderate associations between executive function and false-belief understanding and evidence of longitudinal asymmetry favoring earlier executive function as a predictor of subsequent false-belief understanding (Devine and Hughes, 2014).
The finding does not imply that the two constructs are interchangeable. Executive demands differ across tests. A child can know that another person holds an inaccurate belief and still struggle to answer a lengthy question under pressure, especially when memory and inhibition are taxed. Conversely, success on a simple laboratory response does not guarantee flexible social inference outside the task.
Working memory and cognitive flexibility
Complex stories often require holding several events and identities in mind while updating whose information changed. Second-order beliefs are particularly demanding because a child must represent beliefs nested within beliefs. Flexibility is needed to move between the child's own perspective, the protagonist's perspective, and the objective sequence of events. Strengths or difficulties in these processes can shape observed performance. Developmental research needs designs that separate conceptual competence from the resources needed to express it, and studies of self-regulation in early childhood address these regulatory processes as their own scientific intent.
Are executive functions causes or companions?
The evidence varies by age, task, and design. Meta-analytic developmental data support meaningful relationships, but one middle-childhood study reported no longitudinal link between executive function and theory of mind across its four-year interval (Devine et al., 2016). This difference is informative: preschool transition mechanisms may be different from sources of variation after children have mastered basic belief judgments. Training executive skills in isolation should not be promoted as a universally established way to improve understanding of other people's minds.
Families, Peers, and Everyday Interaction
The role of conversation and shared experiences
Children encounter other perspectives in shared meals, stories, play, household routines, and peer interaction. When two people disagree about what happened, children can discover that witnesses possess different information. When a child imagines what a friend expected, mental-state explanations become useful for conflict repair. Caregiver responsiveness makes these exchanges accessible by fitting language and expectations to the child's developmental level. Such practices are plausible and often educationally valuable, but the research does not support claiming that any one parental style uniquely determines a child's theory-of-mind trajectory. Family circumstances, individual temperament, language access, opportunities for interaction, and socioeconomic constraints intersect.
Siblings, disagreements, and differing viewpoints
Sibling and peer relationships provide frequent examples of conflicting goals and incomplete information. Negotiating a shared game requires noticing that others may prefer different rules or misunderstand an agreement. These experiences may support social reasoning, and children who are already skilled at explaining intentions may elicit different forms of collaboration. Claims that having a particular number of siblings causes superior theory of mind require stronger control of family composition and selection than simple group comparisons allow. Family diversity is an ordinary part of the developmental context.
Pretend play and narrative imagination
Pretend play creates temporary worlds in which one object represents another and characters can act according to invented information. A child may establish that a wooden block is a telephone and then explain why a puppet mistakenly believes someone is at home. Storybooks similarly enable children to compare what a narrator, character, and reader know. These activities offer opportunities to exercise symbolic and mental-state interpretation. Evidence that they necessarily produce broad transferable developmental advantages is more limited; children with stronger language and social understanding may also engage in richer pretend narratives. For a separate analysis of play mechanisms and outcome evidence, see play and child development.
Peer acceptance and social competence
Theory of mind can help children anticipate consequences, interpret misunderstandings, and select socially appropriate responses. A meta-analysis of 20 studies involving 2,096 children found a small positive association between theory-of-mind scores and peer popularity, r = .19 (Slaughter et al., 2015). The effect indicates a modest population-level relationship; it does not establish that performance causes popularity or that a socially unpopular child lacks understanding of mental states. Longitudinal work also associates individual differences in theory of mind with subsequent school social competence, even while leaving potential mediators and contextual factors open (Devine et al., 2016). Social acceptance additionally depends on classroom culture, friendship opportunities, discrimination, relationships, and responses from peers and adults.
Culture, Language Access, and Developmental Diversity
Cross-cultural regularities and differences
Studies from multiple societies reveal that understanding mistaken beliefs emerges across cultures, supporting the relevance of social-representational reasoning to human development. Yet sequences and ages vary. Some communities emphasize differences in desires early, while others place greater conversational emphasis on knowledge, informational access, or social obligations. Research in five cultures found broad developmental convergence in false-belief tasks, while cross-cultural task-sequence studies documented variation in the relative order of several mental-state concepts (Callaghan et al., 2005; Shahaeian et al., 2011).
Cultural comparisons must account for translation, the pragmatics of being questioned by an unfamiliar adult, how knowledge claims are expressed, and familiarity with picture books or hypothetical stories. A task can privilege one communicative convention while ostensibly testing a universal psychological ability. Neither a common average trend nor a difference between groups justifies ranking cultures by overall social sophistication.
Deaf and hard-of-hearing children and accessible interaction
Language access offers an especially important perspective on research findings. Hearing status alone does not identify a child's linguistic environment. Early, fluent access to spoken or signed language and to conversation about mental states can differ substantially from delayed language access. Research comparing children who were deaf, autistic, or typically developing found both differences in timing and differences in the pattern of theory-of-mind task achievement among groups (Peterson, Wellman, and Liu, 2005). These findings should be interpreted with attention to language history, sampling, task suitability, and heterogeneity, rather than as a judgment about intelligence, caring, or personal worth.
Accessible conversation is a practical priority. A child who can participate fluently in signed or spoken dialogue has more opportunities to discuss perceptions, mistaken expectations, and points of view. Assessment should be linguistically appropriate and should distinguish limited access to an unfamiliar testing language from the target cognitive construct.
Autism and the limits of deficit explanations
Autistic children and adults show substantial variation in communication, language, task interpretation, social motivation, and mental-state reasoning. Historical research often treated false-belief task differences as evidence of a single distinctive autism mechanism. That account has undergone extensive criticism concerning specificity, replication, test validity, and assumptions about the relationship between task performance and everyday interaction. A 2026 critical review concludes that available theoretical accounts do not provide a sufficient explanation of theory of mind in autism and identifies validity problems in existing measures (Long, Catmur, and Bird, 2026); earlier analysis likewise documents the weaknesses of claims that autistic people categorically lack theory of mind (Gernsbacher and Yergeau, 2019).
A poor score on a conventional false-belief task is neither a diagnosis of autism nor evidence that a child cannot care about others or experience empathy. Reciprocal interaction involves understanding on the part of every participant, and mismatches in communication can be bidirectional. Developmental and clinical assessment should focus on the child's actual communication, support needs, adaptive functioning, history, and strengths using appropriate multidisciplinary methods. Disorder-specific diagnostic guidance belongs to its own clinical scientific domain, not to an isolated theory-of-mind exercise.
Temperament, disability, and individual variation
Children differ in conversational ease, attention, anxiety in test settings, response speed, and willingness to engage with unfamiliar adults. Some children prefer observing before speaking. Some have language or sensory differences that require adapted materials. Neurodevelopmental conditions can influence task performance, but theory of mind itself is a research construct and an evolving set of abilities rather than a diagnostic category. Researchers and practitioners should compare patterns across tasks and everyday contexts and avoid conflating differences in outward expression with an absence of internal understanding.
Social Understanding, Emotion, and Moral Life
Interpreting emotional outcomes
To interpret another person's emotions, a child may need to integrate events with the person's beliefs and desires. Two people can receive the same news and feel differently because they expected different outcomes. A person may smile politely while feeling disappointed or may be frightened by a harmless object because they believe it is dangerous. These situations connect mental-state understanding with emotion knowledge. Yet emotional awareness and the ability to regulate one's own arousal have distinct developmental determinants. A child who understands another's feeling may still become overwhelmed and require adult co-regulation during conflict.
Deception, jokes, and irony
Understanding an intentional lie requires considering that a speaker may try to induce a belief they do not share. Understanding a joke or ironic remark can require recognizing what a speaker intends beyond literal words. These are demanding applications of perspective reasoning, language pragmatics, and knowledge of local conventions. Success varies by familiarity, linguistic form, and the child's ability to track multiple interpretations. Knowing that someone could deceive does not predict a child's moral character, and an advanced ability to infer intentions can be used in both cooperative and manipulative contexts.
Morality and empathy
Moral evaluation includes concerns about harm, fairness, obligation, and care, along with interpretation of intentions and outcomes. Theory of mind may help a child distinguish accidental harm from deliberate harm, but moral judgment is shaped by multiple social and affective mechanisms. It is scientifically more accurate to examine specific relations among mental-state reasoning, emotion, social goals, and observed behavior than to label theory of mind as the source of kindness or conscience. Research on emotion regulation in children has a different canonical focus on managing emotion and behavior.
Can Adults Support Theory-of-Mind Development?
Everyday practices grounded in developmental principles
Age-appropriate conversation creates opportunities for children to compare perspectives. During a story, an adult can ask what a character knows, why a character expects something, and what new information would change that expectation. During a misunderstanding, an adult can explain that two people saw different parts of an event. During pretend play, adults can follow the child's invented scenario, ask why a character acts, and model revisions when new facts appear. Such approaches make mental states discussable while respecting the child's autonomy, interests, language, and cultural context.
For toddlers, simple language about wants and what people can see is appropriate. For preschoolers, stories about hidden information and mistaken expectations become more accessible. For school-age children, discussions can include mixed emotions, divergent interpretations, and what one person thinks another person knows. Encouraging explanations matters more than obtaining an adult-approved answer at every moment. Narratives and disagreements should not become frequent informal examinations; supportive conversations can be pleasant in their own right.
What controlled training studies show
The most direct intervention evidence examines whether specific, structured theory-of-mind tasks and discussion improve performance on related assessments. A meta-analysis of 45 controlled experiments across 32 papers, involving 1,529 children, found an overall moderate-to-large training effect on measured outcomes (Hedges' g = 0.75, 95% CI 0.60–0.89) (Hofmann et al., 2016). This supports the conclusion that measured theory-of-mind performance is amenable to instruction and experience under studied conditions.
The result does not warrant declaring that an ordinary household activity of unspecified duration will yield the same effect. Studies varied in age, design, comparison condition, targeted skill, and intervention content. Gains on closely related false-belief questions may be larger than gains in naturalistic empathy, sustained peer relationships, or long-term functioning. To infer broader benefits, research needs independent outcome measures, longer follow-up, and evidence of transfer beyond the trained format.
Conversational scaffolding without performance pressure
A constructive adult response can begin with curiosity: “What did she think was going to happen?” The adult can then trace the evidence: “She was outside when the toys were moved, so she still thinks they are by the door.” The child's answer can be explored rather than marked as morally good or bad. Scaffolding can involve drawings, puppets, gestures, signed language, or simplified narratives. Children gain from access to explanation and meaningful participation, not from repeatedly being told they are behind because they answered a hypothetical question differently.
When social support should be more individualized
When a child regularly struggles with reciprocal communication, understanding language, coping in social settings, or sustaining valued relationships, caregivers can discuss the whole developmental picture with a pediatrician or appropriately qualified specialist. The focus should be the child's functional experience, communication access, developmental history, and available supports. Individualized assessment may include language, hearing, attention, learning, or social communication as clinically appropriate. No particular false-belief response is a stand-alone reason for diagnosis or a substitute for comprehensive evaluation. When to seek evaluation for child developmental concerns examines assessment decisions separately.
What the Evidence Establishes and What Remains Unresolved
Established findings
Across a broad literature, many children show measurable developmental changes in explicit false-belief understanding during preschool years; multiple other mental-state concepts appear in a partly ordered developmental progression. Language ability correlates with false-belief performance. Executive function is moderately associated with belief understanding during early childhood. Theory-of-mind measures have modest associations with aspects of social competence and peer relationships. Controlled training can improve targeted performance, though size and generality depend on intervention and outcome definitions (Wellman, Cross, and Watson, 2001; Milligan, Astington, and Dack, 2007; Devine and Hughes, 2014; Hofmann et al., 2016).
Preliminary or limited findings
Specific mechanisms linking family conversation, pretend play, social exposure, and later mental-state reasoning remain difficult to isolate causally. Training gains may not generalize to untrained social environments. Social competence reflects numerous processes outside belief understanding. The relative contributions of language structure, executive control, social interaction, and conceptual change probably differ by age and task. More multilingual, culturally diverse, disability-inclusive, longitudinal, and preregistered research is needed for stronger generalizations.
Contested interpretations
The claim that infant looking behavior demonstrates the same belief representation as explicit preschool tasks remains contested because replications and convergent validity have been inconsistent (Kulke et al., 2018; Poulin-Dubois et al., 2023). Universal deficit accounts of autistic theory of mind likewise face substantial theoretical and measurement challenges (Long, Catmur, and Bird, 2026). Rather than treating these disagreements as reasons to abandon the construct, they define the questions that more discriminating methods must resolve.
Correlation, prediction, and causation
A concurrent association does not specify causal direction; longitudinal prediction improves temporal ordering but cannot by itself remove every unmeasured confounder. Randomized comparisons can support causal effects on measured outcomes, yet their conclusions are bounded by the intervention, population, time horizon, and outcome instrument. In theory-of-mind research these distinctions matter especially because the predictor and outcome may share language, executive, or task-format requirements. A training study that teaches one kind of story question and then tests a similar question is informative about learning, but its ecological implications require separate evidence.
Questions Parents and Educators Commonly Ask
At what age do children develop a theory of mind?
Its components emerge at different points. Early sensitivity to goals and preferences appears before preschool, and many children achieve conventional verbal false-belief task success around ages four to five. More sophisticated understanding of nested beliefs, misunderstandings, and socially complex perspectives continues through school age. A single birthday threshold has no clinical validity (Wellman and Liu, 2004; Hughes and Devine, 2015).
What is the simplest example of false-belief understanding?
Imagine a child watches Sam put a ball in a closet. Sam leaves, and someone moves the ball to a drawer. If asked where Sam will look first, a child demonstrating the task's target understanding answers “the closet” because Sam has not seen the move. A different answer may reflect several sources of task difficulty and should not be interpreted in isolation.
Is theory of mind the same thing as empathy?
No. Theory of mind refers to understanding or inferring mental states; empathy includes affective and interpersonal processes. They often interact, but accurate reasoning about a person's thoughts neither guarantees emotional concern nor exhausts the capacity to care.
Can a child understand beliefs and still struggle socially?
Yes. Social participation relies on language, attention, affect regulation, sensory context, reciprocal opportunity, peer behavior, culture, and environmental support. A child may reason well about fictional characters yet need substantial assistance navigating rapid or stressful real-life exchanges. Social success cannot be inferred from one cognitive test.
Does failing a false-belief task indicate autism, a developmental disorder, or a lack of empathy?
No. False-belief performance is affected by age, task construction, language access, memory, executive demands, anxiety, and familiarity. It is not a diagnostic test for autism or any other condition. Contemporary methodological analyses caution against treating a single theory-of-mind task as evidence of a universal trait or deficit (Long, Catmur, and Bird, 2026).
Do children learn theory of mind simply by playing with others?
Social play offers occasions to negotiate imagined situations, preferences, and misunderstandings. Whether any particular play activity causes broad, enduring improvement is a separate empirical question. Controlled interventions suggest that specific mental-state discussions can enhance targeted measures, but everyday play is shaped by many factors and is valuable for reasons beyond test-score gains (Hofmann et al., 2016).
Does bilingualism delay mental-state understanding?
Bilingual children develop mental-state reasoning within their linguistic and social environments. A child's performance in one test language can be affected by proficiency, instructions, and cultural pragmatics. Evidence linking language ability to theory-of-mind measures does not establish that exposure to two languages causes a deficit. Assessment needs to consider the child's entire language repertoire rather than treating a test in one language as an exhaustive measure of understanding.
How should adults respond when a child misreads someone's intentions?
Clarify what each person knew, wanted, or expected, using language the child can access. Explain the sequence of events and invite the child to propose another interpretation. In emotionally intense situations, support regulation before demanding a complex explanation. The aim is increasing flexible understanding and successful communication, while respecting the child's perspective.
Synthesis: Social Understanding as a Developing Capacity
Theory of mind gives developmental psychology a precise way to study how children connect behavior to the unseen perspectives from which it arises. Understanding different desires, informational access, mistaken beliefs, hidden emotions, and nested perspectives develops through interdependent cognitive and social processes. The empirical record supports robust changes in explicit preschool false-belief reasoning, substantial relations with language and executive function, developmental growth beyond preschool, and some benefits of targeted instruction. It also reveals important limits of infant implicit measures, cross-cultural measurement, simplistic deficit narratives, and predictions based on one task.
For families and educators, the central practical implication is to make mental states understandable through accessible language, stories, shared interaction, and patient explanation. For researchers, the central obligation is to measure the intended construct carefully and separate performance, interpretation, and causal claims. Both perspectives recognize that children's understanding of minds develops in relationships and across different developmental pathways.
Related Articles
Cluster Child Development and Parenting — The published research chapter situates social cognition within the broader science of development and caregiving.
Language Development in Children: Stages, Milestones, and Interaction — Language comprehension, conversation, and developmental variation relevant to mental-state reasoning.
Play and Child Development: Learning, Relationships, and Evidence — Symbolic play, social interaction, learning mechanisms, and evidentiary limitations.
Self-Regulation in Early Childhood: Attention, Emotion, and Behavior — Executive processes and supported regulation relevant to perspective-taking tasks.
Social Development in Children: Relationships, Cooperation, and Conflict — Broader peer relationships and social competence beyond belief reasoning.
References
Asakura, N., and Inui, T. (2016). A Bayesian framework for false belief reasoning in children: A rational integration of theory-theory and simulation theory. Frontiers in Psychology, 7, 2019. https://doi.org/10.3389/fpsyg.2016.02019
Callaghan, T., Rochat, P., Lillard, A., Claux, M. L., Odden, H., Itakura, S., Tapanya, S., and Singh, S. (2005). Synchrony in the onset of mental-state reasoning: Evidence from five cultures. Psychological Science, 16(5), 378–384. https://doi.org/10.1111/j.0956-7976.2005.01544.x
Devine, R. T., and Hughes, C. (2014). Relations between false belief understanding and executive function in early childhood: A meta-analysis. Child Development, 85(5), 1777–1794. https://doi.org/10.1111/cdev.12237
Devine, R. T., White, N., Ensor, R., and Hughes, C. (2016). Theory of mind in middle childhood: Longitudinal associations with executive function and social competence. Developmental Psychology, 52(5), 758–771. https://doi.org/10.1037/dev0000105
Flavell, J. H. (1999). Cognitive development: Children's knowledge about the mind. Annual Review of Psychology, 50, 21–45. https://doi.org/10.1146/annurev.psych.50.1.21
Gernsbacher, M. A., and Yergeau, M. (2019). Empirical failures of the claim that autistic people lack a theory of mind. Archives of Scientific Psychology, 7(1), 102–118. https://pmc.ncbi.nlm.nih.gov/articles/PMC6959478/
Hofmann, S. G., Doan, S. N., Sprung, M., Wilson, A., Ebesutani, C., Andrews, L. A., Curtiss, J., and Harris, P. L. (2016). Training children's theory-of-mind: A meta-analysis of controlled studies. Cognition, 150, 200–212. https://doi.org/10.1016/j.cognition.2016.01.006
Hughes, C., and Devine, R. T. (2015). Individual differences in theory of mind from preschool to adolescence: Achievements and directions. Child Development Perspectives, 9(3), 149–153. https://doi.org/10.1111/cdep.12124
Kulke, L., von Duhn, B., Schneider, D., and Rakoczy, H. (2018). Is implicit theory of mind a real and robust phenomenon? Results from a systematic replication study. Psychological Science, 29(6), 888–900. https://doi.org/10.1177/0956797617747090
Long, E. L., Catmur, C., and Bird, G. (2026). The theory of mind hypothesis of autism: A critical evaluation of the status quo. Psychological Review, 133(4), 1006–1024. https://doi.org/10.1037/rev0000532
Milligan, K., Astington, J. W., and Dack, L. A. (2007). Language and theory of mind: Meta-analysis of the relation between language ability and false-belief understanding. Child Development, 78(2), 622–646. https://doi.org/10.1111/j.1467-8624.2007.01018.x
Onishi, K. H., and Baillargeon, R. (2005). Do 15-month-old infants understand false beliefs? Science, 308(5719), 255–258. https://doi.org/10.1126/science.1107621
Peterson, C. C., Wellman, H. M., and Liu, D. (2005). Steps in theory-of-mind development for children with deafness or autism. Child Development, 76(2), 502–517. https://doi.org/10.1111/j.1467-8624.2005.00859.x
Poulin-Dubois, D., Goldman, E. J., Meltzer, A., and Psaradellis, E. (2023). Discontinuity from implicit to explicit theory of mind from infancy to preschool age. Cognitive Development, 65, 101273. https://doi.org/10.1016/j.cogdev.2022.101273
Ruffman, T., Slade, L., and Crowe, E. (2002). The relation between children's and mothers' mental state language and theory-of-mind understanding. Child Development, 73(3), 734–751. https://doi.org/10.1111/1467-8624.00435
Shahaeian, A., Peterson, C. C., Slaughter, V., and Wellman, H. M. (2011). Culture and the sequence of steps in theory of mind development. Developmental Psychology, 47(5), 1239–1247. https://doi.org/10.1037/a0023899
Slaughter, V., Imuta, K., Peterson, C. C., and Henry, J. D. (2015). Meta-analysis of theory of mind and peer popularity in the preschool and early school years. Child Development, 86(4), 1159–1174. https://doi.org/10.1111/cdev.12372
Wellman, H. M., Cross, D., and Watson, J. (2001). Meta-analysis of theory-of-mind development: The truth about false belief. Child Development, 72(3), 655–684. https://doi.org/10.1111/1467-8624.00304
Wellman, H. M., and Liu, D. (2004). Scaling of theory-of-mind tasks. Child Development, 75(2), 523–541. https://doi.org/10.1111/j.1467-8624.2004.00691.x