Individual Differences in Theory of Mind

Implications for Typical and Atypical Development

Betty Repacholi, Virginia Slaughter

• Publisher: Psychology Press
• Year: 2004
• ISBN:
  9781135432331
• Edition: 1
• Topic: Psychology
• Subtopic: History & Theory in Psychology
• Index: Psychology

Over the last fifteen years, developmentalists, cognitive scientists, philosophers, educators and clinicians have considered the acquisition of a theory of mind - the capacity to predict and explain behavior on the basis of internal, subjective mental states - to be one of the crucial cognitive achievements of early childhood. This volume represents the first collection of work to address, empirically and conceptually, the topic of individual differences in theory of mind. It is also unique because it takes the reader beyond the preschool years, to explore theory of mind development in late childhood and adulthood.

1

Introduction: Individual Differences in Theory of Mind

What Are We Investigating?

VIRGINIA SLAUGHTER

School of Psychology, University of Queensland

BETTY REPACHOLI

Psychology Department, University of Washington

Since the early 1980s, the acquisition of a theory of mind, defined as the ability to predict and explain people's behavior with reference to mental states, has been seen as a crucial watershed in social-cognitive development: one that changes the toddler from a literal observer of human behavior to a folk psychologist, capable of making complex mental-state attributions, engaging in elaborate social and communicative games, and even deception. Theory of mind ability, variously referred to as mentalizing, mind reading, or belief-desire psychology, has traditionally been viewed as the foundation for our adultlike understanding of the social world. As Wellman (1990) expressed it:

Why is achievement of a theory of mind important?...belief-desire psychology is our framework theory of persons. As a framework theory it dictates our basic ontology, our parsing of personal action and thought into its most basic categories. And it dictates our causal-explanatory infrastructure, our basic grasp of how to go about making sense of ourselves and others. In short, belief-desire psychology frames our worldview. (p. 328)

Thus, individual differences in theory of mind ability presumably have important implications for children's and adults’ everyday social interactions. This view is most strongly stated in the theory of mind deficit account of autism (e.g., Baron-Cohen, Leslie, & Frith, 1985; Baron-Cohen, Tager-Flusberg, & Cohen, 1993), where it is argued that the atypical social and communicative behavior of individuals with autism results from their failure to acquire a theory of mind.

Given the centrality of theory of mind as a construct in developmental psychology over the last 15 years, it seems natural that the field should turn to investigations of how it relates to other aspects of development, and indeed, there is already a substantial body of research that does just that (see Bartsch & Estes, 1996, for an early review). This research literature suggests that there are measurable individual differences in the mind-reading abilities of young children, and that these individual differences are correlated in theoretically predictable ways with other variables. We can identify at least three major classes of individual differences studies: those that examine the role of family variables in theory of mind development; those exploring the connection between theory of mind and other cognitive constructs; and those that link theory of mind with various social outcome measures (see Bartsch & Estes, 1996, and Keenan, this volume, for a somewhat different way of classifying this literature).

Studies that have linked theory of mind ability with family variables often focus on identifying predictors of mindreading ability. This class of studies has demonstrated relations between elements of family structure or interaction and mind-reading ability in children, with theory of mind skill being predicted by such variables as family talk about mental states (Dunn, Brown, Slomkowski, Tesla, & Youngblade, 1991; Meins, Fernyhough, Wainwright, Gupta, Fradley, & Tuckey, 2002); number of siblings (Perner, Ruffman, & Leekam, 1994; Ruffman, Perner, Naito, Parkin, & Clements, 1998); socioeconomic status (Holmes, Black, & Miller, 1996; Hughes, Deater-Deckard, & Cutting, 1999); maternal education (Cutting & Dunn, 1999); parenting styles (Ruffman, Perner, & Parkin, 1999; Vinden, 1997); and attachment security (Meins, Fernyhough, Russell, & Clark-Carter, 1998; Symons & Clark, 2000). In general, these studies suggest that children's theory of mind development is enhanced when their family environment somehow draws attention to the fact that people's behavior is based on mental states like belief, desire, and emotions, and that such mental states are unique to individuals.

Mind-reading ability has also been shown to correlate with various cognitive constructs. For instance, significant relations have been obtained between theory of mind and executive functioning (Carlson & Moses, 2001; Frye, Zelazo, & Palfai, 1995; Hughes, 1998; Perner & Lang, 1999), creativity (Suddendorf & Fletcher-Flinn, 1999), fantasy (Taylor & Carlson, 1997), and moral reasoning (Dunn, Cutting, & Demetriou, 2000). In general, children who score highly on measures of theory of mind also perform positively on these other cognitive measures. The most robust relation, replicated across numerous studies, is that between theory of mind and language (Astington & Jenkins, 1999; Cutting & Dunn, 1999; Eisenmajer & Prior, 1991; Jenkins & Astington, 1996; Happé, 1995; Hughes & Cutting, 1999). Some evidence has been found for a minimum threshold of verbal ability that is necessary for children to pass one or more theory of mind tasks (Jenkins & Astington, 1996). Consistent with this finding, children with autism who pass theory of mind tasks tend to have higher levels of verbal ability than those who fail such tasks (Eisenmajer & Prior, 1991; Happé, 1995). The nature of the link between theory of mind and language is not entirely clear, but it is unlikely that this association is simply due to the linguistic demands of the mind-reading tasks (Jenkins & Astington, 1996). Moreover, Astington and Jenkins (1999) recently provided evidence suggesting that language plays a fundamental role in the development of a theory of mind. In their longitudinal study, earlier language ability predicted later theory of mind performance (controlling for earlier theory of mind), but earlier theory of mind did not predict later language ability (controlling for earlier language).

Another class of studies has linked theory of mind to various social outcome measures, including social behavior and other indicators of social competence. These studies demonstrated associations between mind-reading ability and socially competent behavior (Lalonde & Chandler, 1995; Spatz & Cassidy, 1999), the quality of children's peer relationships (see reviews in Sutton and Repacholi et al., this volume), and clinical status (Baron-Cohen, Tager-Flusberg, & Cohen, 2000; Charman, Carroll, & Sturge, 2001; Happé & Frith, 1996; Hughes, Dunn, & White, 1998). This body of research suggests that children with relatively good mind-reading skills enjoy more successful social relationships than those who are less adept mind readers. However, in recent years, evidence has been accumulating (see Sutton and Repacholi et al., this volume) that an advanced theory of mind does not guarantee social success. Indeed, some children (e.g., ringleader bullies; Sutton, Smith, & Swettenham, 1999) and adults (e.g., psychopaths; Blair et al., 1996) consistently use their mind-reading skills for antisocial purposes. Thus, the relationship between theory of mind and social outcome measures is much more complex than previously thought.

So what is it that we are investigating when we try to understand individual variability in theory of mind? What do we mean by “individual differences” in this developmental domain? Presumably, we want to capture something about the way(s) in which individuals vary as mind readers: how they differ in their prediction and explanation of behavior based on mental states. To date, researchers have tended to rely on experimental tasks to assess individual differences in theory of mind, with the bulk of the research focusing on 3- to 5-year-old children or individuals (children and adults) diagnosed with some form of autistic spectrum disorder. This work has generated a wide variety of experimental tasks designed to tap participants’ understanding of an array of mental states and related constructs, including visual perspectives, desires, intentions, pretense and imagination, states of knowledge and degrees of knowledge certainty, remembering and forgetting, recognition of appearances versus reality, understanding of verbal ambiguity, and understanding of deception. Many of these tasks follow a standard structure: Participants are presented with a scenario in which a protagonist is described in some situation, and then they are asked to explain or predict the protagonist's behavior (based on an inferred mental state) or to directly report the protagonist's mental state. Because there is a clear developmental sequence in the typically developing child's acquisition of a theory of mind between the ages of 2½ and 5, the majority of tasks target this age range. Preschoolers’ capacity to correctly infer mental states in these tasks serves as the basis for measuring individual differences in mind reading.

The most widely used theory of mind measure is the false-belief task, in which accurate predictions about another person's behavior can be derived by referring to that person's mental (mis)representation (i.e., their false belief). This task is often referred to as the “litmus test” for assessing children's understanding of the mind, although a number of authors have noted its limitations and called for a reduction in its use (Astington, 2001; Bloom & German, 2000; Chandler & Hala, 1994; Gopnik, Slaughter, & Meltzoff, 1994). Putting these criticisms aside for the moment, the false-belief task has some admirable characteristics that deserve mention. First, it is quick and easy to administer and it is engaging for children. These features are not trivial, particularly when studying individual differences in theory of mind where research protocols demand assessments of multiple constructs. Second, a recent meta-analysis (Wellman, Cross, & Watson, 2001) demonstrated that the task is psychometrically sound. For instance, it is now well standardized. The Wellman et al. (2001) meta-analysis of the false-belief task performance of thousands of children across seven different countries demonstrated a strong age effect whereby the majority of children failed the task at age 3 but passed by age 4½. Furthermore, the task has acceptable reliability, with test-retest and alternate forms reliabilities as high as .80 (Hughes et al., 2000; Wellman et al., 2001). There has been a good deal of debate about the validity of the false belief task. While some have suggested that it is not a valid measure of theory of mind ability (Bloom & German, 2000; Chandler & Hala, 1994; Leslie, 2000), the task appears to have good construct validity, having been developed out of a philosophical analysis of the conditions under which one could be certain that an individual was appealing to mental states when predicting human behavior (Dennett, 1978; Wimmer & Perner, 1983). In addition, as noted before, false-belief task performance has been shown to be correlated in predictable ways with a number of cognitive constructs (e.g., creativity—Suddendorf & Fletcher-Flinn, 1999), behavioral (e.g., social skills— Lalonde & Chandler, 1995), and other variables (e.g., popularity within the peer group—Peterson & Siegal, 2002), thereby demonstrating some concurrent and predictive validity.

In addition to the standard false-belief task and similar tools measuring preschoolers’ ability to attribute other types of mental states (e.g., desires, emotions, etc.), there is also a large literature that analyzes mental state term usage in naturalistic language transcripts as indicators of individual children's theory of mind (Bartsch & Wellman, 1995). To date, however, only a few measures have been developed to assess mind reading in older children. Perhaps the most well known of these is the “second-order” false-belief task, which requires prediction of behavior or a mental state based on what a story protagonist thinks another person thinks (Perner & Wimmer, 1985). This task is not passed by typically developing children until age 7 or 8, and is therefore considered to be an advanced test of theory of mind. Although these are the most widely used theory of mind measures, a number of other tools have been developed over the past 15 years, and many of these have come from the autism literature (see Baron-Cohen, 2001, for a review). Examples include interpreting or sequencing cartoon pictures that portray a mind-reading episode (Baron-Cohen, Leslie, & Frith, 1986); interpreting utterances in terms of the intention behind them (Baron-Cohen, O'Riordan, Stone, Jones, & Plaisted, 1999); creating or identifying deception (Hala, Chandler, & Fritz, 1991); interpreting facial expression, direction of eye gaze, or voice modulation in terms of the underlying mental state (Baron-Cohen, Wheelwright, Hill, Raste, & Plumb, 2001; Rutherford, Baron-Cohen, & Wheelwright, 2002); and making mentalistic attributions about the behavior of moving geometric forms (Abell, Happé, & Frith, 2000). As with the standard false-belief task and the like, these measures typically provide information about individual differences in terms of correct versus incorrect attribution of mental states. Some (e.g., Abell et al., 2000) also provide information about the extent to which individuals appeal to different types of mental states, thereby providing richer information about differences in mind reading.

In summary, then, there are many techniques currently available to assess theory of mind. However, a number of issues related to measuring and conceptualizing theory of mind, issues that are particularly relevant to the project of studying individual differences, have been recently noted and deserve discussion. The standard false-belief task and the various other tasks mentioned here typically capture a developmental trend such that younger children either fail or perform at chance levels, whereas older children pass the task, giving an adultlike answer at better than chance levels. In other words, what has traditionally been measured is simply earlier versus later acquisition of the ability to predict another person's mental state or to explain behavior with reference to mental states. Thus, individual differences are captured in age of onset of the particular mind-reading skill measured by a given task. More recently, researchers have moved in the direction of theory of mind test batteries that include multiple tasks, assess understanding of a variety of mental states, and enable a composite score to be created (see both Peterson and Tager-Flusberg, this volume, for comprehensive developmental assessments using a range of theory of mind tasks). However, even in these designs, it is still the case that the dependent variable is age of onset. Although age of onset provides useful information that allows us to understand when basic theory of mind concepts are mastered, from the perspective of investigating individual differences it is quite limited. All we can do is measure variation in the extent to which individuals pass tasks relatively early or late.

The fact that children's performance on false-belief and similar mentalstate attribution tasks is highly reliable suggests that we do a good job assessing this aspect of mind reading. However, all of the tasks, both standard and advanced measures, are plagued by ceiling effects in samples where mind-reading skill is relatively good. For typically developing children above the age of 7 or 8, and for adults, there are as yet few assessments that reveal significant and meaningful individual differences in mind reading (Happé Advanced Test of Theory of Mind [Happé, 1994] and Baron-Cohen's various mind-reading test batteries [see Baron-Cohen, 2001, for an overview], all initially developed for use with high-functioning autistic individuals, are notable exceptions). This is partly because few researchers have speculated about the changes that might occur in mentalizing ability during middle to late childhood (see, however, Bosacki & Astington, 1999, and O'Connor & Hirsch, 1999), and also because there is a lack of consensus as to what constitutes a mature theory of mind. On the other hand, it is obvious from our everyday social interactions that even adults differ in their mentalizing skills, so devising appropriate theory of mind measures for older children and adults should be a high priority. Having said that, we need to be very careful about relying on any one measure; mind reading is not a unidimensional construct, and it would be shortsighted to believe that one type of task could capture all of its complexity.

The majority of mind-reading tasks assess what individuals are capable of doing in a specific experimental context but the results are generalized to everyday life. For instance, younger children who fail the standard false-belief task are considered by some to be incapable of predicting behavior based on a false belief, whereas older children who pass the task are thought to have acquired this capacity. This conclusion is not sha...

Table of contents

• Front Cover
• Half Title
• Title Page
• Copyright
• Dedication
• Contents
• About the Editors
• Contributors
• 1 Introduction Individual Differences in Theory of Mind: What Are We Investigating?
• 2 Sometimes Necessary, Never Sufficient: False-Belief Understanding and Social Competence
• 3 Bypassing Empathy: A Machiavellian Theory of Mind and Sneaky Power
• 4 Theory of Mind, Machiavellianism, and Social Functioning in Childhood
• 5 ToM Goes to School: Social Cognition and Social Values in Bullying
• 6 Individual Differences in Theory of Mind: The Preschool Years and Beyond
• 7 Did Cain Fail to Represent the Thoughts of Abel Before He Killed Him?: The Relationship Between Theory of Mind and Aggression
• 8 The Social Face of Theory of Mind: The Development of Concepts of Emotion, Desire, Visual Perspective, and False Belief in Deaf and Hearing Children
• 9 Exploring the Relationship Between Theory of Mind and Social- Communicative Functioning in Children with Autism
• 10 Mind Reading and Social Functioning in Children with Autistic Disorder and Asperger’s Disorder
• 11 Theory of Mind and Social Dysfunction: Psychotic Solipsism Versus Autistic Asociality
• 12 Theory of Mind in Autism and Schizophrenia: A Case of Over- optimistic Reverse Engineering
• 13 Synthesis: Psychological Understanding and Social Skills
• Author Index
• Subject Index