A Review of Theory of Mind and Autism

Marcella D. Baghetti

Emotions play a significant role in daily interactions with others. Emotions concern what is experienced and what is communicated (Dennis, Lockyer, & Lazenby, 2000; Rieffe, Terwogt, & Stockmann, 2000). We purposefully exhibit emotions to influence others behavior. We also react and respond to the emotions of others; perhaps even anticipate and adapt our behavior accordingly (Rieffe et al.).

Dennis et al. (2000) describes emotional communication as the automatic, spontaneous expression of affect; and emotive communication is the conscious, strategic modification of emotional expression for social communication. Facial expressions express both emotional and emotive communication (Dennis et al.). Distinct facial expressions are associated with different emotions conveying happiness, surprise, sadness, anger, fear, and disgust (Dennis et al.). A key component in understanding relationships is what children understand about themselves and other people, an understanding of "other minds" (Bauminger & Kasari, 1999). This "theory of mind" is the ability to construct people in terms of internal mental states such as their beliefs, desires, intentions, imagination, emotions, etc. (Baron-Cohen, 2001; Baron-Cohen, Leslie, & Frith, 1985; Frith, 2000). Having a theory of mind is having the ability to reflect on the contents of ones own and others minds (Baron-Cohen). This ability has been extensively studied in both normal and abnormal development (Baron-Cohen; Muris, Steernemen, Meesters, Merckelback, Horselenberg, Hogen, & Dongen, 1999). Specifically, theory of mind has been comprehensively studied in relation to those diagnosed with autism. This literature review will address recent findings in this area.

As already described theory of mind is the ability of normal children to attribute mental states, such as intentions, beliefs, desires, and so forth, to themselves and other people, thus having the ability to understand and predict behavior (Buitelaar & van der Wees, 1997). Muris et al. (1999) suggests that theory of mind is a precondition for the understanding of social environment and for engaging in socially competent behavior. Interactions between people involve attributions about other peoples mental states at different levels. First order belief describes what people think about real events, whereas second order beliefs pertain to what children think about other peoples thoughts (Baron-Cohen, 2001; Bauminger & Kasari, 1999; Muris et al.).

A well known characteristic of Autism is the lack of emotionally based contact with others (Rieffe et al., 2000). It has been proposed that autistic children are socially impaired precisely because they lack a theory of mind (Muris et al., 1999). Autism is a rare, but severe developmental disorder of the brain that exists from birth throughout life (Baron-Cohen et al., 1985; Carpenter, Pennington, & Rogers, 2001; Firth, 2000; Holroyd & Baron-Cohen, 1993; Rutter & Bailey, 1994). There is a spectrum of autistic disorders of different types and severity. Difficulty understanding other minds is a core cognitive feature of autism spectrum conditions (Baron-Cohen, 2001). The cognitive and social impairments of autistic children include difficulties in the expression and understanding of emotion, as well as problems in discerning facial affect (Dennis et al., 2000). It has been found that autistic children find immediate social environments to be unpredictable and incomprehensible; often they are said to, in some sense, treat people and objects alike (Baron-Cohen et al., 1985).

Theory of mind has been found to account for an increase in understanding the many perplexing features of autism (Frith, 2000). Researchers suggest that the majority of individuals with autism are mind-blind (Frith; Rieffe et al., 2000). Mind-blindness means that children are unable to attribute mental states, such as dreaming, hoping, thinking, believing and wanting in others or in oneself (Frith; Rieffe et al.). This lack of theory of mind makes it probable that an inability to attribute mental states must lead to a profound lack of insight into the nature of normal social interactions. This lack of insight would lead to limited social relationships (Rutter & Bailey, 1994). Autistic children lack not only social insight; they also lack normal communication skills, and have problems with playful imagination (Frith, 2000). A common element in these impairments seems to be a lack of understanding that people can think thoughts that are interestingly different from reality and interestingly different from each other (Frith).

Links Between Theory of Mind and Autism

Baron-Cohen et al. (1985) tested theory of mind on 61 autistic, Down Syndrome, and normal preschool children. To assess theory of mind, Baron-Cohen et al. employed the Sally-Ann experiment. In this experiment Sally and Ann are two protagonist dolls. The first task involved a naming question to ensure the children knew which doll was which. Sally placed a marble in her basket and left the scene. The marble was then transferred by Ann and hidden in her box. When Sally returned the experimenter asked a critical belief question: "Where will Sally look for her marble?" (Baron-Cohen et al., p. 41). Baron-Cohen et al. state that if the children point to the previous location of the marble, Sallys basket, they pass the belief question because they understand the dolls now false belief. If they point to the marbles current location, they fail the question by not taking into account the dolls belief. These conclusions are acceptable if the two control questions are answered correctly: "Where is the marble really?" (Reality question); "Where is the marble in the beginning?" (Memory question) (Baron-Cohen et al., p. 41).

Results strongly supported the hypothesis that autistic children as a group fail to employ a theory of mind. Baron-Cohen et al. (1985) found that every child in the experiment correctly answered the control questions, leading to the conclusion that the children all know and implicitly believed the marble was put somewhere else after Sally left. The group difference emerged with the critical question of "Where will Sally look?" after she returns (Baron-Cohen et al.). Autistic children were found to answer this question in a distinctively different way from the others. The Downs Syndrome and normal preschool children answered by pointing to where the marble was in the first place. Thus, they must have "appreciated that their own knowledge of where the marble actually was and the knowledge that could be attributed to the doll were different" (Baron-Cohen et al., pp. 42-43). It seems that they predicted the dolls behavior on the basis of the dolls belief. On the other hand, the autistic children answered by pointing consistently to where the marble really was. As a result Baron-Cohen et al. concluded that the autistic children did not appreciate the difference between their own and the dolls knowledge. In all Baron-Cohen et al. found evidence that autistic children lack the ability to appreciate that other peoples beliefs might differ from their own. They further argued that a theory of mind deficit might comprise the core of autism.

An expansion of these earlier findings was a study conducted by Rieffe et al. (2000) assessing autistic children understanding of atypical emotions. Participants consisted of high-functioning children from the autism spectrum (mean age of 9-years), 6-year-old controls, and 10-year-old controls. Results confirmed prior findings that children from the autistic spectrum sometimes explain other peoples emotions in terms of desires but rarely in terms of beliefs with regard to typical emotions but not atypical emotions (Rieffe et al.). In the case of typical emotions, happiness on receiving a present, autistic participants were found to attribute fewer mental states (desires and beliefs) than a much younger control group of 6-year-olds (Rieffe et al.). In the case of atypical emotions, the autistic group performed as well as the 10-year-old control group.

Rieffe et al.s (2000) explanation for atypical emotions show that children from the autistic spectrum indeed have the capacity to mind read, with respect to both desires and beliefs, although they do not always use this capacity in the same manner normally developing children do. Furthermore, Rieffe et al. claims that the mind-reading capacity of high-functioning children from the autistic spectrum might be basically intact; unused in everyday circumstances but not necessarily defective. This is valid in that autistic children lack curiosity for others unless they need something from another person. They play solitarily with little regard for those around them. But when something unexpected happens high-functioning children with autism know where to find a caretaker to provide comfort (Rieffe et al.).

Theory of Mind and Social Cognitive Domains

It has been shown that autistic children are limited in their understanding of mental states and expression and understanding of emotion (Baron-Cohen et al., 1985; Rieffe et al., 2000). Little has been done to understand how these children understand emotions. Dennis et al. (2000) studied how eight high-functioning children with autism and eight age and gender matched controls with similar oral language development understand real and deceptive emotion in narratives.

Children with autism were found to have limited understanding of socially derived emotion. High-functioning children with autism were able to understand emotion labels (i.e. could match happy and sad with appropriate facial expressions) and relate those expressions to prototypical situations (Dennis et al, 2000). However, they were less able than controls to indicate the real emotions story characters feel, the deceptive emotions they express in the face, or the social reasons prompting a deceptive facial expression. These results should be interpreted cautiously in light of the small sample size, which creates a tension with generalizing. Nonetheless it appears that facial expressions are learned codes for particular emotions, but not forms of social communication that, through deception, can modify the beliefs of the person to whom the expression is directed (Dennis et al.).

Carpenter et al. (2001) examined a less researched area of what autistic children understand about simpler mental states such as intentions. This study tested the understanding of others unfulfilled intentions in 2 1/2 to 5-year-old children with autism and a control group of children with developmental delays. Carpenter et al. presented participants with an adults action for which the intention did not match the reality of the situation.

Findings revealed that no significant between-group differences on any measure involving the understanding of others intentions. Although within-group differences were found to suggest that children with autism may have a slightly less complex understanding of others intentions than do other children, it was clear to Carpenter et al. (2001) that any deficits these children showed were not as marked as those typically shown on traditional theory of mind tasks. Replications of this study are needed with a larger sample size and more diverse age group to determine whether similar data will result. Additionally it would be beneficial to have an extra control group of normal developing children with which to compare understanding of others intentions both fulfilled and unfulfilled.

Buitelaar and van der Wees (1997) examined relationships between deficits in the various social cognitive domains in children with autistic disorder, children with pervasive developmental disorder not otherwise specified (PDDNOS), in psychiatric control, and in normal children. Buitelaar and van der Wees matched these groups person-to-person by age and verbal IQ. The ages of the participants were older than the aforementioned studies; the clinical group children were 8-18 years old, the normal children 8-13 years old. The tests involved tasks for matching and context recognition of emotional expression and a set of first and second order theory of mind tasks.

Results suggested that theory of mind and emotion recognition functioning proved to be better integrated in the non-PDD children than the PDD children, but significant correlations were found between theory of mind and emotion recognition measures in the PDD children (Buitelaar and van der Wees, 1997). Overall, these findings suggest that theory of mind and emotion recognition functioning is based on a common underlying ability, consistent with other evidence that demonstrates the affective and cognitive processes are interdependent in shaping a persons social understanding and emotional reactions (Buitelaar and van der Wees).

Theory of Mind in High-Functioning Autistic Children

Thus far the majority of studies examining theory of mind and autism have focused on a fixed time period. Serra, Loth, van Geert, Hurkens, and Minderaa (2002) employed a longitudinal study that investigated the development of theory of mind knowledge in children with lesser variants of autism (PPDNOS) over a period considered to be critical for theory of mind development, 3 to 5 years of age.

Participants consisted of 11 children with PPDNOS, 23 normally developing children for cross-sectional comparison and 13 normally developing children for longitudinal comparison. The groups were comparable in verbal and non-verbal mental age. Two storybooks containing identical tasks, but different protagonists, were used to investigate various aspects of the children's theory of mind (Serra et al., 2002). Results indicated that children with PDDNOS had specific difficulties in understanding their ability to predict other peoples emotions based on situational cues, desires, and beliefs. Their ability to predict actions from beliefs and desires were relatively intact. Compared to the normally developing children, these children achieved lower levels of theory of mind knowledge, both at the time of initial assessment and six months later.

Serra et al. (2002) concluded that the data suggests theory of mind development in children with PDDNOS is both delayed and unexpected. The growth pattern of theory of mind skills in children with PDDNOS seemed, according to Serra et al., to be qualitatively different from the growth pattern found in the group of normally developing children.

Research has shown that high functioning autistic children can develop some theory of mind with age. Through a longitudinal study, with a 7 to 8 year follow-up period, Holroyd and Baron-Cohen (1993) tried to determine how much of a theory of mind an autistic individual develops. Results based on 17 autistic participants revealed that the majority of people with autism, 60-70%, had little development of a theory of mind. The minority, 20-30%, development was thought to possibly extend to the equivalent of a 3 to 4 year old level by the teenage years (Holroyd & Baron-Cohen). However, given the relatively small sample size carefulness should be used in interpreting these results. It would be beneficial to replicate this study with a larger sample size consisting of participants from various degrees of the autism spectrum.

Validity and Reliability of Theory of Mind

Numerous researchers utilize various theory of mind tasks to evaluate children, adolescents, and adults who are normal developing, developmentally delayed, or are on the autistic spectrum. The results of these findings, while mostly consistent, have also found inconsistencies or lack of significant correlations (Baron-Cohen, 2001). It is only natural and scientific to question the reliability and validity of the instruments used to assess theory of mind. Previous studies, some reviewed here, have used more traditional assessments of theory of mind such as, the Sally and Ann test (Baron-Cohen et al., 1985), the smarties test, tests of emotion recognition (perception of emotion from facial expressions), role taking test, and the John and Mary test assessing children's understanding of second order beliefs (Baron-Cohen).

Muris et al. (1999) provided one of the first attempts to investigate the reliability and validity of the theory of mind test. Their research was broken into four studies assessing both normally developing children and children with pervasive developmental disorders (PDDs). The first main result was that the theory of mind test was found to be a reliable instrument with good internal consistency, sufficient test-retest reliability, and high interrater reliability (Muris et al). Secondly, Muris et al. found that theory of mind test scores increased with age, indicating the test is receptive to developmental progression. Young children only succeeded on theory of mind items that utilize basic domains of theory of mind, whereas older children also passed items that measure the more advanced areas of theory of mind. Thirdly, Muris et al. found evidence that supports the concurrent validity of the theory of mind test. Theory of mind test scores were found to positively correlate with the performance on several of other theory of mind tasks (i.e. tests of emotion recognition, understanding first and second-order beliefs, and role taking). Lastly, Muris et al. stated that children with a PDD performed worse on the test than children with other psychiatric disorder. This suggests that the theory of mind test possess discriminant validity. Overall, these results indicate that the theory of mind test is a reliable and valid instrument that can be employed to measure various aspects of theory of mind (Muris et al).

Muris et al. (1999) stated that theory of mind tests can be used in three ways. The first is that the test can be used to screen children for deficits in theory of mind. Secondly, because this test is informative about the developmental phase of children's theory of mind, it allows clinicians to tailor their intervention to specific problems of each child. Thirdly, the theory of mind test can be used to evaluate the efficacy of theory of mind in training programs.

Conclusions

Given the findings in the aforementioned studies, theory of mind tests do have a reliable and valid place in assessing developmentally delayed children as proposed by Muris et al. However, theory of mind does not go without its limitations. It does not address the reasons for the varying degree of severity of symptoms in autism. Moreover, as proposed by Frith (2000), another limitation is that the theory of mind hypothesis only addresses the triad of impairments. There are other features of autism that are highly persistent and handicapping which are in need of explanation. Moreover, research needs to resolve how the various core features of autism and the presumed underlying cognitive components relate to each other (Frith). To do this it is essential to find out whether some of the same features are found in other disorders.

Future Directions

Studies thus far have provided ample and relevant information regarding the issue of theory of mind. Ultimately however, studies like Serra et al. (2002) emphasize the importance of conducting longitudinal studies. In doing so, greater understanding of theory of mind in children at different stages of development can be assessed and compared more thoroughly. Additionally, the specific emotion problems of children with lesser variants or higher functioning autism need further investigation. As proposed by Serra et al. "could the difficulty to predict emotions from desires and beliefs be related to the inability to predict emotions from situation cues?" (p. 895). Since theory of mind involves an individuals ability to take another's perspective, additional research should look at the relation between theory of mind and social behavior. This research would lead to a better understanding of whether theory of mind development in children with pervasive disorders amounts to a delay, deficit or combination of the two (Serra et al.).

Furthermore, future research in theory of mind should take an additional look at normally developing children, as they pass through the stages set forth by Piaget. Due to the lack of research on older, normally developing children and adolescents, more research is needed on their ability to accomplish theory of mind tests for a deeper understanding of theory of mind and what it validly explains. Additionally, more comparison studies of normally developing and autistic children are needed to address similarities and differences in gender and age as they progress through developmental stages.

Research conducted in more naturalistic settings would be beneficial in ensuring whether higher functioning children with autism are showing genuine understanding about minds or are demonstrating effective strategies for success on tasks. Most importantly, future research should focus on other methods for studying false belief and theory of mind abilities in autism. Given the persistent problems in social functioning among autistic children and the potential link between theory of mind abilities and social skills, it is important to have a better understanding of these skills in autism (Bauminger & Kasari, 1999).

As proposed by Frith (2000), the explanation of autism at the cognitive level needs to be complemented by the explanation at a biological level. While the research presented here suggests links between theory of mind and autism, it is still unknown if it will provide the complete answer to the riddle of autism. It has, however, begun to bridge the gap between social and cognitive deficits that characterize autism (Rutter & Bailey, 1994). Ultimately, future research in this area will hopefully refine techniques for studying theory of mind skills across the life span and make further movement in understanding the underlying mechanisms essential for theory of mind (Baron-Cohen, 2001).

References

Baron-Cohen, S. (2001). Theory of mind and autism: A review. International Review of research in Mental Retardation, 23, 169-184.

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

Bauminger, N., & Kasari, C. (1999). Brief report: Theory of mind in high-functioning children with autism. Journal of Autism and Developmental Disorders, 29, 81-86.

Buitelaar, J. K., & van der Wees, M. (1997). Are the deficits in the decoding of affective cues and internalizing abilities independent? Journal of Autism and Developmental Disorders, 27, 539-556.

Carpenter, M., Pennington, B. F., & Rogers, S. J. (2001). Understanding of others intentions in children with autism. Journal of Autism and Developmental Disorders,
31, 589-599.

Dennis, M., Lockyer, L., & Lazenby, A. L. (2000). How high-functioning children with autism understand real and deceptive emotion. Autism, 4, 370-381.

Frith, U. (2000). Cognitive explanations of autism. In K. Lee (Ed.), Childhood cognitive development: The essential readings (pp. 324-337). Malden, MA: Blackwell Publishers.

Holroyd, S., & Baron-Cohen, S. (1993). Brief report: How far can people with autism go in developing a theory of mind? Journal of Autism and Developmental Disorders, 23, 379-385).

Muris, P., Steerneman, P., Meesters, C., Merckelbach, H., Horselenberg, R., van den Hogen, T., & Van Dongen, L. (1999). The TOM Test: A new instrument for assessing theory of mind in normal children and children with pervasive developmental disorders. Journal of Autism and Developmental Disorders,29, 67-80.

Rieffe, C, Terwogt, M. M., & Stockmann, L. (2000). Understanding atypical emotions among children with autism. Journal of Autism and Developmental Disorders, 30,
195-203.

Rutter, M., & Bailey, A. (1994). Thinking and relationships: Mind and brain (some reflections on theory of mind and autism). In S. Baron-Cohen, H. Tager-Flusberg, et al. (Eds.), Understanding other minds: Perspectives from autism (pp. 481-504). London: Oxford University Press.

Serra, M., Loth, F. L., van Geert, P. L., Hurkens, E., & Minderaa, R. B. (2002). Theory
of mind in children with lesser variants of autism: A longitudinal study. Journal of Child Psychology and Psychiatry, 43, 885-900.