Theory Of Mind Research Paper

1. Theory Of Mind And Cognitive Psychology

It is all too easy to overlook or take for granted our everyday cognitive abilities, even the most spectacular of them. The ability to think about thinking is a case in point. While our knowledge of cause and effect, space, time, number, object, and so forth have been the focus of intense philosophical scrutiny since classical times, the highly abstract concepts that form cornerstones of our social intelligence have until recently received much less attention. One of the most surprising aspects of theory of mind ability is just how early in life it begins to develop. One might have thought that concepts as abstract as desire, pretense, or belief could only be acquired late in life, long after many other more concrete concepts, extensive general knowledge, and sophisticated reasoning abilities are in place. It would be very hard to explain to a young child what a belief is, given that one cannot point at a belief or pick one up or produce an intelligible definition. Fortunately, one does not need to do any of these things because young children acquire the relevant concepts spontaneously. In fact, mental state concepts emerge rapidly and without formal tuition during the preschool period (Leslie 1987, 2000, Perner et al. 1987, Wellman 1990), appear to be universal across cultures (Avis and Harris 1991), and can be acquired even by children with a clinically diminished IQ (Baron-Cohen et al. 1985, Tager-Flusberg et al. 1998).

According to US psychologist, Henry Wellman, the concept of desire develops around two years of age (Wellman 1990). A related sensitivity to intention or goal of action seems to emerge even earlier, as shown by the 18-month-old child’s disposition to imitate the action an actor intended to perform but, due to error, did not actually perform (Meltzoff 1995). The Scottish psychologist, Alan Leslie, has shown that the concept of pretending begins to be used by children between 18 and 24 months of age. Strikingly, when children first begin to pretend play by themselves, they simultaneously acquire the ability to recognize that another person is pretending and can even share pretenses with others (Leslie 1987, 1994).

The question at what age the concept of belief is first acquired has been more controversial and has attracted a great deal of study. It is generally accepted that to show a grasp of the concept of belief, the child should understand that what someone believes might be false. Working in Austria, Heinz Wimmer and Josef Perner developed a seminal task in which children are required to predict the behavior of a story character by attributing a false belief to the character (Wimmer and Perner 1983). They tested children between three and six years of age and found that it was not until the children were six years old that a majority of their subjects passed their task.

Subsequently, a British team, working in London, and comprising Simon Baron-Cohen, Alan Leslie, and Uta Frith, simplified the Wimmer and Perner task and found that the vast majority of four year olds whom they tested passed by attributing a false belief to the story character (Baron-Cohen et al. 1985). Since then, such simplified tasks have been found to produce successful performance in four year olds. The child is told a short story, with the aid of pictures or props, in which a character, Sally, has a marble. She places this marble in a basket and covers it. Sally then leaves and goes outside. While she is gone, naughty Ann enters the room, finds the marble in the basket, and moves it to a nearby box, and hides it there. The child is then asked two ‘control’ questions to test for basic comprehension: ‘Where did Sally put the marble in the beginning?’ and ‘Where is the marble now?’ Finally, the child is asked either directly about Sally’s belief, ‘Where does Sally think the marble is?’ or the child is asked to predict Sally’s behavior, ‘Where will Sally look for the marble?’ In order to answer either of the latter questions, the child must attribute to Sally a belief that is different from the child’s own belief, and that the child considers to be false.

Baron-Cohen et al. (1985) also found that older children with Down’s syndrome who had an average IQ of 64 (in the mildly retarded range) mostly passed this test of false belief. However, the majority of a third group of children, children with autism, failed the false belief task. The group of children with autism had an average IQ of 82 and were thus in the low normal range. Their failure could not therefore be accounted for by mental retardation.

Wimmer and Perner developed a further false belief task in which the child is shown a container which normally contains a well-known candy, and asked what is inside (Perner et al. 1987). After the child answers with the name of the candy, the container is opened and the child is shown that there is in fact a pencil inside. The pencil is replaced and the container closed once more. The child is then told that another person will be shown the container in the same way, and asked to predict what the other person will say is inside. Their findings showed that most four-year-olds could pass this task while most three-year-olds failed. Perner et al. (1989) found that a group of children with autism, despite having verbal mental ages in excess of seven years, nevertheless mostly failed this task. Children with specific language impairment, matched on verbal ability with the autistic group, almost all passed. Subsequently, these patterns have been widely replicated: Normally developing four-year-old children and older moderate to mildly retarded children mostly pass false belief tasks, and most children with autism fail (see Happe 1995 for a review).

2. Theories Of Theory Of Mind—Agreements And Disagreements

The existence of these early developing yet highly abstract concepts raises the question: What is their origin? Are these concepts innate and, if so, how is that possible? Alternatively, are these concepts acquired and, if so, how is that possible? Either way, we have a challenge. If innate, how could such concepts be built into a cognitive system? If acquired, how could such concepts be gained by children with limited experience and ability? The concept of belief seems as abstract as the concept of the electron, and there seems to be as little direct everyday evidence for the existence of beliefs as there is for the existence of electrons. Thus, the child’s theory of mind offers an intriguing case study of the nature of abstract concepts and their origins.

So far, two main answers to the above questions have been proposed. The first answer assumes that possession of abstract concepts depend upon possession of abstract knowledge. This reflects the traditional assumption that a concept is really a name for a packet of knowledge. In the case of theory of mind concepts, the knowledge packets are said to be commonsense theories (Gopnik and Meltzoff 1997, Perner 1991, Wellman 1990). This view of abstract concepts is referred to as the ‘theory theory.’ The second model assumes that possession of abstract concepts can depend upon a cognitive mechanism, rather than upon knowledge. This view is known as the theory of mind mechanism (ToMM) view (BaronCohen 1995, Frith and Frith 1999, Leslie 1987, 1992, 2000). The principle focus of debate has been around the concepts of pretense and belief. Before examining the knowledge versus mechanism dispute, we should outline points on which the two views agree.

First of all, while interpretations of the data differ, there are few disputes about the data themselves. Second, it is not disputed that even young children’s theory of mind concepts are mentalistic rather than behavioristic in character. Behaviorist philosophers and psychologists have argued that mental states do not really exist and that talk of mental states is really just a confused way of talking about behavior. However, our common sense, rightly or wrongly, is undeniably committed to the existence of mental states and, whether or not mental state concepts are ‘sound,’ we certainly possess and employ such concepts. We naturally assume that Sally’s behavior will be caused by her desire to find the marble in combination with her belief concerning its whereabouts. We use such belief-desire predictions ubiquitously and to good effect everyday. This suggests that our mental states concepts do refer to something real, probably to complex cognitive properties of agents. If concepts like pretending and believing are mentalistic rather than behavioral, then they cannot be a simple summary of sensory experience. And that will have implications for how these concepts can be acquired. A third point that is not generally in dispute is that the child’s theory of mind has some kind of innate basis in a disposition to employ mentalistic concepts. And finally, there is no dispute that in the course of development, the child acquires more and more knowledge about people, about mental states, and about the role mental states play in causing people’s behavior.

Beyond the four points above, however, there is much controversy. As we have indicated, the disagreements center on the nature of abstract concepts and the role of concepts in development. If a concept is the ‘name’ of a body of knowledge, then the only way to acquire a given concept is to acquire the appropriate body of knowledge. Another way to look at this is that possession of a given body of knowledge determines possession of a given concept. So, if this body of knowledge changes, then so too will the associated concept. Since everyone agrees that children’s knowledge changes in the course of development, then so too, on this view, must their concepts change. Hence, a major topic in the study of development is the study of conceptual change (Carey 1985). A further implication of this widespread view is that if a given concept is innate, then its associated body of knowledge must also be innate. Often, it is assumed that if something is innate then it cannot change with development. However, some theory theorists have proposed that some knowledge (specifically, knowledge of some theories) is both innate and revisable (Gopnik and Wellman 1994, 1995). On this account, both the child’s innate and acquired concepts can and do change.

According to the ToMM view, concepts do not name bodies of knowledge but instead directly represent properties in the world. This means that the representational relation between a concept (in the head) and the property (in the world) that it designates is subserved by a causal mechanism of some sort. For example, the concept ‘red’ is typically established and sustained by the neurocognitive mechanisms of color vision. Without color vision, it is hard to attend to and learn about redness. But nothing you will learn about redness changes the meaning of the concept ‘red,’ namely, the property of redness. On this view, concepts do not depend upon knowledge; on the contrary, they are a prerequisite for obtaining knowledge. Therefore, conceptual development is not the same thing as knowledge development. Consequently, innate concepts do not require innate knowledge, and acquired concepts do not depend upon acquiring knowledge. Instead, on this view, we have the job of inquiring into the nature of the neurocognitive mechanisms that establish and sustain conceptual representations.

The above contrasting assumptions about the nature of concepts lead to very different interpretations of theory of mind development.

3. Early Development, Normal And Abnormal

As mentioned earlier, 18-month-old infants will imitate the goal of an action rather than simply the movements comprising the action. It is possible that they do so because infants have knowledge of a theory of goal-directed action. If so, the ‘theory of goals’ they have knowledge of needs to be discovered and specified. Alternatively, infants may not have knowledge of a theory at all; their imitation may instead simply reflect the operation of a neurocognitive mechanism. The ‘mirror neurons’ discovered by Rizzolatti in his investigations of cells in the ventral premotor cortex (area F5) in monkeys are suggestive in this regard. Neurons with similar properties are apparently involved in human imitation (Iacoboni et al. 1999). Such mechanisms with their sensitivity to the ‘hidden’ goals of an actor’s actions may form the innate basis for the concept of desire.

Early pretending provides another example of an early social sensitivity where it is hard to imagine what the content of the infant’s theory, in this case ‘theory of pretending,’ might be. Leslie (1987) drew attention to a key and long ignored aspect of early pretending, namely, that when the ability for solitary pretending first emerges, between 18 and 24 months, so does the ability to recognize pretending in other people. Leslie went on to argue that pretending should be understood as part of our theory of mind ability. Instead of proposing that the infant possessed innate knowledge of a theory, he postulated the existence of a specialized neurocognitive mechanism, ToMM, that forms the innate basis for theory of mind concepts, including the concept of pretending. A number of empirical predictions followed from the link between pretending and ToMM, including the prediction that autistic children, who do not develop spontaneous pretending, would be specifically impaired in their understanding of belief.

As we saw, this prediction has been confirmed. However, this confirmation in itself does not prove the link between autism and an impaired neurocognitive mechanism like ToMM. One alternative possibility is that children with autism have some kind of general processing difficulty, for example, with language, or with abstract reasoning, or with working memory, or with some other general factor that is demanded by false belief tasks. Any of these possibilities would provide an alternative explanation to the idea of a specific impairment caused by an impaired ToMM.

To test the alternative general processing explanation for autistic performance on false belief, Leslie and Thaiss (1992) adapted tasks devised by Zaitchik (1990). Zaitchik had developed a task in which a Polaroid photograph is taken of an object in a certain location. The object is then moved to a new location. Following control questions about where the object was when the photograph was taken and where the object is now, the child is asked to say where the object is in the photograph. The child does not get to see the photograph. After all, the child in the Sally and Ann task does not get to see Sally’s belief! Zaitchik found that preschool children tend to fail or pass both the photographs task and the Sally and Ann task about the same age. This is not surprising given how similar both tasks are from the point of view of general problem solving. When Leslie and Thaiss tested a group of high-functioning autistic adolescents and a group of normally developing four-year-old children on similar pairs of photographs and false belief tasks, they found that the four year olds performed well on both sets of tasks. The autistic subjects, as expected, performed poorly on the false belief tasks, but their performance on the photograph tasks was nearly perfect. This result too has been replicated and extended. The failure of autistic children on false belief tasks cannot be attributed to the general problemsolving demands of the tasks. The photograph tasks make the same or similar demands and the autistic subjects passed these.

So far, we have considered only comparisons between autistic subjects who fail false belief tasks and normally developing four year olds or nonautistic developmentally disordered subjects who pass. But we should also compare autistic subjects with normally developing three-year-olds who also fail. Do the latter two groups fail false belief tasks for the same reason? The ToMM model predicts that they fail for different reasons. Only in the case of autism is ToMM neurologically impaired; in the typically developing threeyear-old, ToMM should be intact, limited only by immaturity. One way to test this prediction is to examine whether ways that help three year olds to pass false belief tasks also help autistic subjects. Siegal and Beattie (1991) found a very simple way to help three year olds achieve better performance on the Sally and Ann task. Instead of asking, ‘Where will Sally look for her marble?,’ they asked ‘Where will Sally look first for her marble?’ Surian and Leslie (1999) found that while this small change did indeed help three-year-olds to calculate Sally’s false belief, it had no effect on autistic performance, which remained just as poor. Other manipulations which help typically developing threeyear-olds but which do not help older children with autism have been reported by Roth and Leslie (1991, 1998).

Finally, although the existence of the neurocognitive mechanism ToMM was originally postulated on the basis of purely cognitive findings, recent advances in brain imaging techniques are beginning to reveal the neural systems basis of this instinct (Frith and Frith 1999).

4. Summary

The ability to detect the hidden mental states of other people is a remarkable cognitive power. It is tempting to assume that the idea of mental states must rest upon the same sort of foundation as many other highly abstract ideas, like electrons or inflation—that is, upon the capacity for scientific reasoning and theorizing. However, the speed and precocity with which ideas about mental states emerge in normal development, together with the patterns of normal and abnormal development reviewed above, suggest a different picture. The basic theory of mind that typically emerges during the preschool period is based upon an in-built mechanism that provides an intuitive and instinctual sensitivity to the mental states of other people.

Bibliography:

1. Avis J, Harris P 1991 Belief-desire reasoning among Baka children: Evidence for a universal conception of mind. Child Development 62: 460–7
2. Baron-Cohen S 1995 Mindblindness: An Essay on Autism and Theory of Mind. MIT Press, Cambridge, MA
3. Baron-Cohen S, Leslie A M, Frith U 1985 Does the autistic child have a ‘theory of mind’? Cognition 21: 37–46
4. Carey S 1985 Conceptual Change in Childhood. MIT Press, Cambridge, MA
5. Frith C D, Frith U 1999 Interacting minds—A biological basis. Science 286: 1692–5
6. Gopnik A, Meltzoff A N 1997 Words, Thoughts, and Theories. MIT Press, Cambridge, MA
7. Gopnik A, Wellman H M 1994 The theory theory. In: Hirschfeld L, Gelman S A (eds.) Mapping the Mind: Domain Specificity in Cognition and Culture. Cambridge University Press, New York, pp. 257–93
8. Gopnik A, Wellman H M 1995 Why the child’s theory of mind really is a theory. In: Davies M, Stone T (eds.) Folk Psychology: The Theory of Mind Debate. Blackwell, Oxford, UK and Cambridge, MA, pp. 232–58
9. Happe F G E 1995 The role of age and verbal ability in the theory of mind task performance of subjects with autism. Child Development 66: 843–55
10. Iacoboni M, Woods R P, Brass M, Bekkering H, Mazziotta J C, Rizzolatti G 1999 Cortical mechanisms of human imitation. Science 286: 2526–8
11. Leslie A M 1987 Pretense and representation: The origins of ‘theory of mind.’ Psychological Review 94: 412–26
12. Leslie A M 1992 Pretense, autism, and the ‘theory of mind’ module. Current Directions in Psychological Science 1: 18–21
13. Leslie A M 1994 Pretending and believing: Issues in the theory of ToMM. Cognition 50: 211–38
14. Leslie A M 2000 ‘Theory of mind’ as a mechanism of selective attention. In: Gazzaniga M S (ed.) The New Cognitive Neurosciences, 2nd edn. MIT Press, Cambridge, MA, pp. 1235–47
15. Leslie A M, Thaiss L 1992 Domain specificity in conceptual development: Neuropsychological evidence from autism. Cognition 43: 225–51
16. Meltzoff A N 1995 Understanding the intentions of others: Reenactment of intended acts by 18-month-old children. Developmental Psychology 31: 838–50
17. Perner J 1991 Understanding the Representational Mind. MIT Press, Cambridge, MA
18. Perner J, Frith U, Leslie A M, Leekam S R 1989 Exploration of the autistic child’s theory of mind: Knowledge, belief and communication. Child Development 60: 689–700
19. Perner J, Leekam S R, Wimmer H 1987 Three-year-olds’ difficulty with false belief: The case for a conceptual deficit. British Journal of Developmental Psychology 5: 125–37
20. Premack D, Woodruff G 1978 Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences 4: 515–26
21. Roth D, Leslie A M 1991 The recognition of attitude conveyed by utterance: A study of preschool and autistic children. British Journal of Developmental Psychology 9: 315–30
22. Roth D, Leslie A M 1998 Solving belief problems: Toward a task analysis. Cognition 66: 1–31
23. Siegal M, Beattie K 1991 Where to look first for children’s knowledge of false beliefs. Cognition 38: 1–12
24. Surian L, Leslie A M 1999 Competence and performance in false belief understanding: A comparison of autistic and normal three-year-old children. British Journal of Developmental Psychology 17: 141–55
25. Tager-Flusberg H, Boshart J, Baron-Cohen S 1998 Reading the windows to the soul: Evidence of a domain specificity sparing in Williams syndrome. Journal of Cognitive Neuroscience 10: 631–9
26. Wellman H M 1990 The Child’s Theory of Mind. MIT Press, Cambridge, MA
27. Wimmer H, Perner J 1983 Beliefs about beliefs: Representation and constraining function of wrong beliefs in young children’s understanding of deception. Cognition 13: 103–28
28. Zaitchik D 1990 When representations conflict with reality: The preschooler’s problem with false beliefs and ‘false’ photographs. Cognition 35: 41–68