Development of Self-Other Understanding - 1
Running head: Development of Self-Other Understanding
Mental State Discourse, Theory of Mind, and the Internalization of Self – Other Understanding
Douglas K. Symons
Psychology Department, AcadiaUniversity, Wolfville, Nova Scotia, CanadaB4P 2R6
Contact information:
Tel: (902) 585-1571
Fax: (902) 585-1078
Note. This research was supported by a grant from the Social Sciences and Humanities Research Council of Canada, as well as a Visiting Scholar Travel Grant from the University of Queensland, Brisbane, Australia. G. Moran, C. Peterson, V. Slaughter, and three anonymous reviewers are thanked for their comments during the preparation of this manuscript. Correspondence can be addressed to Dr. Doug Symons, Psychology Department, AcadiaUniversity, Wolfville, Nova Scotia, CanadaB4P 2R6, email: .
Abstract
This paper examines the social origins of self – other understanding in young children. The proposition is that mental state discourse is related to the development of social understanding, and evidence for this proposal comes from cross-sectional, longitudinal, and training studies that relate discourse about mental states in a variety of interactional contexts to children’s acquisition of theory of mind. Vygotsky’s concept of internalization is used to account for how representational models of self and other as distinct mental agents may originally develop from this kind of discourse. This social account of theory of mind development has implications for language about mental states, levels of internalization,the cultural context of self and other, and social competence. However, this precludesneither cognitive- nor neurological - based accounts of theory of mind development, which are at other levels of functional explanation.
Key Words: theory of mind, discourse, internalization, mental state language, attachment, Vygotsky, self/other understanding, social development, language
Mental State Discourse, Theory of Mind, and the Internalization of Self – Other Understanding
1. Introduction
A fundamental question in developmental psychology is how young children come to understand the mental states of self and others. The ability to predict and explain people’s behavior with reference to mental states defines a child’s theory of mind (e.g., Dunn, 2000; Flavell, 2000; Slaughter & Repacholi, 2003), and an important developmental accomplishment is when a child comes to understand that others can have thoughts and feelings that differ from her or his own (see Flavell, 2000; Wellman, Cross, & Watson, 2001). This social understanding is dependent on an awareness of mental states, which are defined here as the thoughts and feelings that children could ascribe to either social partners or the self.
Initial research on the acquisition of theory of mind focused primarily on cognitive factors and models of representation (e.g., Gopnik, 1993; Perner, 1991; Wellman, 1990). Cognitive factors in theory of mind development, such as the executive functions of attention, memory, and impulse control, remain crucial (e.g., Carlson & Moses, 2001; Perner & Lang, 1999; Hughes, 2002). However, researchers have increasingly focused on experiential factors that contribute to individual differences in theory of mind acquisition (see Dunn, 1988; Nelson, Plesa Skwerer, Goldman, Henseler, Presler, & Walkenfeld, 2003; Repacholi & Slaughter, 2003; Meins, 1997). This focus has built upon a research tradition that has integrated cognitive and social approaches to developmental theory (for examples, see Bretherton & Beeghly, 1982; Hinde & Stephenson-Hinde, 1987; Kaye, 1982; Lewis & Brooks-Gunn, 1979; Schaffer, 1984; Thompson, 1993). For example, theory of mind has been examined as a consequent of early attachment relationships between children and parents,given the role of security as an experiential factor (e.g., Fonagy, Steele, Steele, & Holder, 1997; Sroufe & Waters, 1977) and the basis of internal working models for how self and other behave within close relationships (Bowlby, 1969/1982; Thompson & Raikes, 2003). More recent research has examined specific social interaction histories with parents, siblings, and peers that predict children’s theory of mind, and an important feature of early experience is discourse about mental states within these social interactions.
The goal of the current paper is to review research that relates mental state discourse to children’s theory of mind.The thesis is that discourse between young children and significant persons (i.e., family members and friends) about the thoughts and feelings of others in early social relationships facilitates the development of theory of mind as indicated by understanding false beliefs. This review therefore addresses recent research suggesting that mental state discourse plays an important developmental role in the origins of self – other understanding in children (see Bretherton, 1999; Carpendale & Lewis, in press; Meins, 1997; Harris, 1999). The proposed mechanism for this process is Vygotsky’s (1978; Lloyd Fernyhough, 1999)internalization, within which reasoning about mental states becomes internalized in the child from participation in interpersonal discourse about thoughts and feelings about self and others.
This paper first examines the theoretical basis of the thesis, and defines constructs to be examined. The concept of internalization will be introduced, which draws upon convergent developmental theory from many theorists including Baldwin, Piaget, and Bowlby. These dominant theorists all addressed the development of self – other understanding, although their theoretical differences are better known (see Chapman, 1991; Glassman, 1994; Tudge & Winteroff, 1993). Methodology commonly used in developmental research to examine self – other constructs will presented and false belief paradigms will be described, given that much of the evidence to be presented comes from a theory of mind research tradition. The focus of this paper is on how children internalize the notion that self and other exist as distinct mental agents, which is fundamental to theory of mind and can be assessed with false belief tasks. Mental state language will be defined, and examples of language which reflects beliefs, desires, and emotions will be given.
The second objective of this paper is to review evidence that mental state discourse is related to self – other understanding. The relation between various measures of attachment security and theory of mind in young children will be examined, given that attachment is proposed to be an important socio-emotional context for understanding self and other. A growing body of literature examines a variety of family factors related concurrently and prospectively to the development of mental state understanding within a self – other approach. Discourse about the mental states of others within a number of interactional contexts will be reviewed, including play, discipline, and joint reading. Mental state discourse and pretense with peers is another area of relevant social experience that can also be related to theory of mind in children. Finally, training studies address whether discourse about the mental states of others contributes to social understanding of children using experimental methodology.
The third section of this paper is devoted to the broader implications of this research base. These are the importance of language for the development of self – other understanding, different levels of internalization, the cultural context of self – other understanding, and social competence. While the current review is of social factors in the development of theory of mind, this does not preclude cognitive nor neurological accounts, which are at other levels of functional explanation.
2. Theoretical Approaches and ConstructsRelevant to Self – Other Understanding
2.1. Internalisation of self and other
Discourse about mental states of the self and other has a basis in psychological theories of self that date back to the origins of psychology as a distinct discipline. For example, J.M. Baldwin (1897, 1903) first suggested that sense of others develops in tandem with a sense of self, and wrote that the social self is born when a child realizes that “other people’s bodies …. have experiences in them such as mine have. They are also me’s” (Baldwin, 1911, p. 321). This proposal can be used to support a conceptualization of a child’s understanding of mind within a social context, rather than from a child analogy of his or her own mind to that of others (for review, see Muller & Runions, 2003). Baldwin proposed that the social self comes about by an interaction of constitutional processes and social constructs, and one mechanism for this interaction is the reciprocal imitation of others carried out by infants and toddlers as they consolidate a sense of self (see Lee, 1982; Parker, 1993). In a similar fashion, Cooley (1902) suggested that others helped define the self as though one were looking into a social mirror, and coined the phrase “looking-glass self” to describe this process. Throughout its history, self research has therefore defined the development of the self in relation to thoughts about and actions of others.
Baldwin’s approach is echoed by Jean Piaget (1995) in some of his descriptions of the development of conscious thought about the self. Piaget suggested that conceptual development underlying a sense of self and personality is affected by social and cultural context, and social contact is what brings about consciousness of the self (see Ferrari, Pinard, & Runions, 2001). This has been referred to as reflective consciousness, in which the self is mature when it can reflect upon itself at the stage of formal operations (see Pons & Harris, 2001; Smith, 1998). On the other hand, a majority of Piaget’s work suggests that consciousness develops as a result of maturation and individual activity, so it is acknowledged that Piaget did not always emphasize social factors in the development of the self (see Ferrari et al., 2001; Ferreiro, 2001; Lourenco & Machado, 1996). Nevertheless, the importance of interpersonal processes to the development of thoughts about self and other have been frequently emphasised (Cicchetti & Beeghly, 1990; Hinde, Finkenauer, & Auhagen, 2001; Stern, 1985).
Developmentally, children are initially egocentric, but over time become aware of the perspective of the hearer and exchange ideas with her or him (see Piaget, 1926). This is only possible once it is realized that others may have thoughts and emotions that differ from the self. John Flavell (1985) emphasized the interpersonal aspects of children’s cognitions during social situations, such as reactions children have when exposed to the emotional arousal of others. This understanding of the thoughts and feelings of others has become known as mentalizing, mind-reading, and theory of mind (e.g., Astington, 1993; Flavell, 2000; Wellman, 1990). Of course, infants are also born with basic biological, cognitive, and perceptual predispositions that facilitate communication, socialization, and ultimately the development of self – other understanding, but social experience allows these developmental processes to unfold.
Predicting the thoughts and feelings of others is also important to children’s social behavior within an attachment theory approach. Bowlby (1969/1982) initially described internal working models of behaviour, for example, as representations for how self and other function within relationships across the lifespan (see also Bretherton, 1991; Thompson & Raikes, 2003). Such models are a natural consequence of the human ability to construct mental representations of the social world (e.g., Bretherton, Ridgeway, & Cassidy 1990; Main, Kaplan, & Cassidy, 1985), but can only proceed once a child overcomes egocentrism, and is capable of “seeing things from a viewpoint other than [her or] his own” (Bowlby, 1969/1982, p. 352). There are many parallels between internal working models, mental event representation from a cognitive perspective (see Nelson, 1999), and Piaget’s schemas as representational frameworks of understanding of the social behavior of others (Piaget, 1962, 1969; Baldwin, 1992). In fact an argument could be made that it would be hard to distinguish between these constructs. Bowlby drew upon Piaget’s work in his own theorizing, and suggested that schemas of relationships later in life contain expectations that self and other will act in certain ways, and these are based on an early social history with a primary caregiver (see also Bretherton & Munholland, 1999; Sroufe & Waters, 1977).
Internal working models have been proposed to develop from an interaction between representational advances and social experience. The latter include sensitive parenting, an infant balancing attachment behaviour and exploration, and language discourse (Grossman, 1999; Meins, 1997; van IJzendoorn, 1995). Thompson (1999) has described four representational systems that are encompassed by working models. These are: (1) social expectations of the attributes of significant others; (2) event representations of attachment-related experiences; (3) autobiographical memories related to personal narratives and self understanding; and, (4) understanding of other people and their psychological characteristics, including thoughts, motives, and intentions. The latter two points suggest that self – other understanding within personal relationships is fundamental to internal working models.
Vygotsky (1978, 1987) described how children internalize cultural values, ways of thinking, and social norms through social interaction (see also Wertsch, 1985). He proposed that higher-order thought originates in the internalizationof external social relationships and meanings, not by merely copying the external into the internal, but by recoding what is known about the external into the internal (see Astington, 1999; Frawley, 1997). In other words, interpsychological activity ultimately becomes intrapsychological activity in the child. In this fashion, Vygotsky’s work has influenced those who believe that children may come to understand their social world by hearing directly and indirectly about their social world from mature social partners such as by engaging in collaborative conversations about the world around them and internalizing rules by which the world functions (see Fivush, 1991; Moore, 1996; Nicolopoulou, 1993; Scholnick & Wing, 1992; Tomasello, Kruger, & Ratner, 1993: Wertsch, 1985). A critical aspect of the human social world is mental state reasoning.
Bruner (1986) discussed how children internalize rules of language and dialogue, which have powerful influences on what becomes internal speech of children. Discourse between children and more advanced mentors becomes a potential source of information about the social world that children can internalize into thoughts of their own (see Reiber & Carton, 1987). Internalization has also been used to explain how early social relationships affect future social relationships (Bretherton & Beeghly, 1982), which is fundamental to the essential properties of relationships (see Sroufe and Fleeson, 1986). At other functional levels, internalisation has also been applied to transmission of culture (Markus & Kitayama, 1991; Tomasello et al., 1993), gender roles (Bussey & Bandura, 1999), and parental standards for behaviour (Grusec & Kuczynski, 1997; Kochanska & Thompson, 1997).
A social – cognitive theory of the origins of self and other requires a mechanism of transmission between parents and children (see van IJzendoorn, 1995). Meins (1997) suggested that children may internalize representations of self and other through mentalistic discourse with parents and others, and wrote: “By focusing on young children as individuals with minds, caregivers may nurture their children’s developing understanding of themselves and others as mental agents” (p. 145). Meins (1999) has proposed a model in which a child’s mentalising abilities are a developmental consequent of mentalistic discourse with caregivers. In a complementary fashion, Harris (1999) proposed that coherent discourse about emotional states leads to children understanding emotional states in others, and in his direct model, discourse leads to child’s understanding of emotion, but attachment status is does not play a causal role in this pathway, and is a separate product of discourse and caregiver sensitivity. The proposition that discourse about mental states becomes internalized into self – other reasoning would be supported by empirical links between mental state discourse and performance of theory of mind tasks.
2.2. False belief paradigms and self – other understanding
Theory of mind researchers have focused on the thoughts of children about the previous thoughts of themselves or the thoughts of others regarding either the properties of objects, or other state of reality (for review, see Flavell, 2000). This orientation was reflected Piaget’s focus on the thoughts of children about the properties of objects from a cognitive perspective (see Ferrari et al., 2001; Chapman, 1991). For example, some false belief tasks within theory of mind research have examined the thoughts children previously had about either the identities of misleading objects or containers that have unexpected contents. Other tasks have focused on the thoughts children have about others who have experienced an object unexpectedly transferred within a task (Gopnik & Astington, 1988; Perner, Leckham, & Wimmer, 1987; Riggs & Robinson, 1995; Wimmer & Perner, 1983). The essential test questions address whether participant children can distinguish between knowledge they now have about stimulus characteristics, and either thoughts they themselves previously had or thoughts story characters could possibly have. If they make this distinction, then participants are said to have an understanding of the “self” as a mental agent whose thoughts and feelings can either change over time or can be different from those of an “other”. This is because children distinguish between their current thoughts and either their past thoughts or those of others in order to correctly approach tasks.
There are debates on the influence of false belief task parameters on performance, as well as age of emergence of successful task completion (e.g., Freeman & Lacohee, 1995; Miller, 2000; Wellman et al., 2001). These are both important issues beyond the scope of this paper. As discussed in Wellman’s meta-analysis (Wellman et al., 2001), however, false belief tasks are being increasingly used as a marker of theory of mind understanding in research on individual differences in social – cognitive development. There is a general consensus that successful completion of false belief tasks can be taken as an indicator that a child understands that their current thoughts and feelings can differ from what they experienced at another time or that held by another, and these tasks have some concurrent and predictive validity (for review, see Slaughter & Repacholi, 2003).
Many studies look for above chance performance on theory of mind tasks, and interpret statistically above chance performance as an indicator that a child “has” a theory of mind. There is little doubt that passing over 65%of a sufficient number of tasks,which have a chance probability of 50%, is “above chance”. However, there is still interesting variability in performance above 65%. As adults paying attention to false belief tasks typically would score 100%, some researchers are interested in children who do not show 100% mastery in tasks, or put another way, individual difference questions in the rate of acquisition or consolidation of a sense of “other”, rather than its emergence. This approach characterizes the development of social understanding as a continuous as opposed to stage-like process (e.g., Tomasello, 1999), particularly when operationalised as an aggregate score on battery of tasks (see for example Hughes, Adlam, Happe, Jackson, Taylor, & Caspi, 2000; Rushton, Brainerd, & Pressley, 1983). The current proposal is that children gradually gain experience with the minds of self and others through social discourse during childhood, and this facilitates passing false belief tasks.