There is now a substantial body of research demonstrating that pretend play bears a plausible relationship to children's understanding of mental states. One possible reason for this relationship is that pretend play encourages children to appreciate the distinction between actual and represented events. However, as Harris (2000) notes, if this were true then most types of pretend play should be a stimulus for theory of mind development whereas to date there is convincing empirical support only for a relationship between role (sociodramatic) play and mental state understanding.
At the theoretical level, a role play-theory of mind relationship appears to offer support for simulation theory. One way to view role play is to think of it as a process by which children imagine themselves doing what others would do in a particular situation, and then adopting the cognitive-emotional disposition of the characters they portray (Harris, 2000). Over time, experienced role players find themselves involved in different play scenarios that call for simulation of how others would respond to a particular circumstance. Regularly imagining the world as another person sees it has the potential to become an important wedge in moving children toward the realization that mental states guide people's actions (Harris, 2000).
Lillard (2002) notes two allied features of role play that might also contribute to theory of mind development. One is that during role play children not only simulate the thoughts and emotions of the characters they enact but they also have the opportunity to observe their play partners doing much the same thing. In addition, role play often requires a good deal of "out of frame" work, such as first negotiating who will play what role or selecting the appropriate props. Effectively, out of frame work becomes a form of stage direction in which the children gain another opportunity to discuss mental states, e.g., the motivations and intentions of the characters they enact.
Although there is both theoretical and empirical support for a role play-theory of mind relationship two important caveats are worth noting. First, the data are correlational, and most studies simply show concurrent relationships between some form of role play and a variety of theory of mind measures.3 Second, as noted previously, there are a number of alternative explanations for theory of mind development that might either covary with role play (e.g., family talk about thoughts and feelings) or arguably assume a more primary position (e.g., maternal sensitivity). These alternative explanations raise the question of the most productive way forward in pretend play-theory of mind research.
One possible strategy is to continue the in situ correlational studies, which offer the advantage of ecologically valid 'real world' observations, but to embrace the research designs adopted in a number of studies examining the relationship between theory of mind and either language or attachment. Two design features seem particularly promising. One involves the use of time-lagged analyses capable of examining bi-directional relationships. For example, Jenkins and Astington (1999) and Ruffman et al. (2002) used multiple data points to establish a relationship between a measure of either child or maternal language at any early time point and theory of mind at later time points, as well as analyses that ruled out the reverse explanations. Another promising design involves several variables entered as predictors in multiple regression to establish not only the presence of a relationship but also the amount of variance accounted for by each of the predictors. A number of studies have used multiple regression effectively in establishing relationships between theory of mind and language (Jenkins & Astington, 1999; Cutting & Dunn, 1999), parenting style (Ruffman et al., 1999), and attachment (Meins et al., 1998; Steele et al., 1999).
A second possible strategy is to use observational studies as the basis for training studies that have the best chance for establishing true causal relationships. For example, one can ask whether training on variables that emerge from correlational studies as possible predictors of theory of mind leads directly to improvement in mental state understanding. With respect to role play, there is an established literature on training studies that provides the outline of a way forward. Smilansky (1968) noted that infrequent and poorer quality role play was common among economically disadvantaged children, but that these children could be trained effectively to engage in robust role play. Smilansky's (1968) work inspired a number of subsequent studies which demonstrated that fantasy (role) play training led to improvements on intelligence, problem solving, and perspective taking (Rosen, 1974; Satz & Johnson, 1974). In a recent and more directly relevant study, Dockett (1998) found that during the course of a ten week preschool session, 4-year-olds who were given role play training showed significant improvement on both immediate and delayed theory of mind posttests in contrast to controls who participated in the standard curriculum and showed no posttest gains.
The potential benefit from the combination of correlational and training studies is substantial. We may well learn, however, that no single variable, including role play, is by itself a strong predictor of theory of mind. Lillard's (2001a) twin earth model predicts as much suggesting that there are multiple determinants of theory of mind. Harris (in press) makes a related point by suggesting that research on role play forms a natural bridge with research on the pragmatic functions of language, particularly the conversations that children have with caregivers. Certain parental/conversational styles, such as asking children to imagine how others feel, mirror the processes of imagining the perspective of others that unfold during role play. Together the two can form a "virtuous circle" that improves children's belief understanding as well as their appreciation of the feelings and emotions that make-up the social world they inhabit (Harris, in press). The search for multiple contributors to the development of theory of mind seems both warranted and consistent with the evidence we have to date on how children come to appreciate that minds matter.
Was this article helpful?