Pretend Acts Using Social Cues

Children engage in pretend play from around 18 months of age, rarely confusing their “pretend” world with the “real” world. It has been recently suggested that various social cues and behavioural regularities presented by parents during their own pretend play enhance the ability of children to quarantine pretend acts from real acts. However, a question that has received little empirical investigation is the possibility that children, like their parents, convey that their actions are “only pretend” through producing particular social cues during symbolic play. The aim in this paper was therefore to identify if children convey social cues differently depending on whether they are engaged in functional or pretend play. The present study also sought to explore whether or not children’s “pretend” social cues vary when their pretence is spontaneously generated or when they are imitating an experimenter. To address these aims, fifty-two children aged from 26 to 43 months of age were videotaped engaging in free-play before and after an adult modelled a series of pretend scenarios. Play bouts were categorized as pretend or functional, and post-modelling pretend play was further categorized as being either spontaneous or imitated. Instances of social gaze, both with and without expressions of positive affect were recorded. In line with the hypothesis, during pretend play children engaged in more acts of social gaze, both with and without positive affect than they did during functional play. In addition, the social gaze in imitated pretend play differed from spontaneous pretend play in that it was more often accompanied by positive affect. These results indicate that young children use social cues to communicate that their actions are “only pretend” and suggests that the use of these social cues can be enhanced through imitation. Together, these findings emphasise the social and cultural foundations of pretend play. * E-mail address: [email protected], Phone: +61 7 3365 6805, Fax: +61 7 3365 4466. School of Psychology, University of Queensland, Brisbane, QLD, 4072, Australia. Angela Randell and Mark Nielsen 128 When children pretend, they create a fascinating world where objects can stand in for other unrelated objects, inanimate items can be given life, and all manner of things can be imagined even if there is nothing physically available to take their place. As children’s imagination develops, pretend sequences with objects become complex and elaborate, incorporating fantasy, characters, roles and places. Pretence, however, is not merely a curious activity of childhood. It is also considered to play a significant role in children’s socialcognitive development. The important role that is attributed to pretend play is evident in its associations with other critical developmental achievements such as language (Lyytinen, Poikkeus & Laakso, 1997; McCune, 1995; McCune-Nicolich, 1981; Veneziano, 2002), social competence (Haight, Masiello, Dickson, Huckeby & Black, 1994; Lindsey & Mize, 2000), and in particular, with theory of mind (Astington & Jenkins, 1995; Mitchell & Neal, 2005a,b; Nielsen & Dissanayake, 2000; Schwebel, Rosen & Singer, 1999; Suddendorf & FletcherFlinn, 1996; Taylor & Carlson, 1997; Taylor, Carlson, Maring, Gerow, & Charley, 2004; Youngblade & Dunn, 1995). The relevance of pretend play to children’s social functioning is further evident in the literature on children with autism in whom a lack of pretence is seen as a core deficit (Baron-Cohen, 1987; Baron-Cohen, Leslie & Frith, 1985; DSM-IV-R, 2000). Exploring the development of pretend play thus provides opportunity to uncover insights into broader aspects of children’s social and cognitive development. Pretending by its very definition requires acting symbolically or non-literally with objects in a way that is contrary to reality (Garvey, 1990). Thus, somewhat challenging to developmental theorists is that pretence emerges during the second year of life, when children are still learning to construct their own concept of the world and reality. It would seem that young children are somehow able to concurrently build on their knowledge of the real world (incorporating information about what objects are, what one can do with them, etc.) and engage in the somewhat opposing activity of pretend play (where objects are treated in a manner that may be radically different from their intended purpose). The question therefore arises: What allows pretend play to develop at a time when children are still learning about the real world? The answer to this question has remained contentious, and many of the arguments put forth have focussed on inner, cognitive mechanisms. For example, in an influential paper on this topic Leslie (1987) controversially argued that children are able to pretend because they truly understand the mentalistic, representational nature of their acts during pretend play. More recently, Nichols and Stich (2000) have purported that a “possible world box” in an individual’s head allows pretence to occur. Other theorists, in contrast, maintain that children are not aware of mental representation during pretence, at least not in the early stages of pretend play development (e.g. Lillard, 1993, 1994, 1996; Harris, Lillard & Perner, 1994; Rosen, Schwebel & Singer, 1997). These theorists predominantly argue that early pretending involves the use of a non-representational framework in which children understand pretence only as a characteristic physical action, or “acting as if”. This position would suggest that children use a “theory of behaviour” rather than a “theory of mind” when they initially engage in pretending, and that children only adopt mental/representational interpretations of pretence later in their development (Harris, Lillard & Perner, 1994). While it is interesting to speculate over the cognitive abilities underpinning the development of pretence, other theorists have shifted from this focus on inner, individual characteristics of pretending towards a broader focus on the social and cultural context of Children’s Communication of Pretend Acts Using Social Cues 129 pretending. For these theorists, pretend play is to be understood within the social context in which it develops (e.g. Rakoczy, Tomasello, & Striano, in press; Tomasello, 1999). Importantly, this view purports that pretend play is an inherently social activity, and the development of pretend play in young children relies heavily on social support (Striano, Tomasello & Rochat, 2001). Perhaps the strongest evidence for this account is the finding that children, after witnessing an adult perform pretend actions, tend to copy the pretend actions they see, perform a greater number of pretend actions and perform more complex pretend actions (Bretherton, O’Connell, Shore & Bates, 1984; Fenson & Ramsay, 1981; Jackowitz & Watson, 1980; Lyytinen, 1989; Ungerer, Zelazo, Kearsley & O'Leary, 1981). The social/cultural foundations of children’s early pretence are also evident in recent research indicating that mothers facilitate children’s understanding of pretence by using social cues when they themselves pretend in front of their children. Lillard and Witherington (2004) reported that mothers smile and laugh more, look at their children more frequently and longer, talk more and exaggerate their movements more when they pretend to have a snack in front of their children compared to when they really have a snack in front of their children. Lillard and Witherington (2004) argue that these behaviours serve to help signal to the child that the pretend actions are not to be taken literally. Additionally, the sequence of activities in which children look at their mother, their mother acts then smiles, and then children either engage in a pretend behaviour or smile, occurs significantly more often in mothers’ pretend behaviours than in their real behaviours. This pattern of looking and smiling gives an indication that children do indeed pay attention to their mother’s pretend cues, as it seems that they engage in increased social referencing and joint attention when watching their mothers pretend. Therefore, these social behaviours enacted by mothers appear to act as a signal for children to differentiate pretence from reality. Moving onwards then from children’s interpretation of others’ pretence to children’s generation of their own pretence: It seems reasonable to expect that children will, like their mothers, communicate that their pretend actions are not to be interpreted literally. Indeed, a number of authors have argued that when children engage in pretend play they produce particular social-behavioural patterns to convey their pretence. Piaget (1962) for example claimed that smiling and laugher were significant cues to his children’s pretence. Similarly, Nicolich (1977) argued that children appear playful and aware of their pretending when initiating pretence, and McCune (1995) listed sound effects, exaggerated gestures, and facial expressions as behaviours adopted by children to differentiate literal and pretend behaviours. However, these observations were based on anecdotal evidence and intuitive notions of childhood pretence. In terms of empirical support for children adopting particular social features to convey their pretence, only limited evidence has so far been forthcoming. In the research that exists to date, young children’s communication of pretence has been compared to their communication during more “practical” actions (i.e. functional/instrumental actions). Striano, Tomasello and Rochat (2001, Experiment 2) found that children aged 2 to 3 years looked more frequently and for longer toward an experimenter after producing pretend acts than after performing functional acts. Rakoczy, Tomasello and Striano (2005) found that 2-year-old children both looked and smiled more toward an experimenter when pretending with a novel object than they did when performing instrumental actions with a novel object. These findings thus offer some evidence that children’s early acts of pretend play involve some unique forms of social interaction. Angela Randell and Mark Nielsen 130 However, the afore-cited studies were limited in several ways. Due to the fact that the researchers were interested in documenting how children learn to pretend with novel objects, in each of these studies children were restricted to acting with sometimes rather unusual objects (e.g. a black installation tube) and object combinations (e.g. hair brush and pegboard). They were also limited to using only two objects at a time. In addition, children had been subjected to quite lengthy periods of testing (related to other components of the experiments) prior to play. A likely outcome of the unusualness and paucity of objects available, and the decline of interest associated with long testing times, children in these studies produced very few acts of pretence. The low levels of pretence exhibited by the children in these studies limits the conclusions that can be drawn, especially with regard to the generalisability of the findings to children’s day-to-day play, where numerous objects are available, many of which are highly familiar and custom-made for “pretending”. Possibly related to the fact that the settings of the studies previously mentioned were not like everyday play, Rakoczy et al. (2005) found mixed results regarding whether or not children produce more smiling during pretend play compared to instrumental actions. That is, this effect was evident in one study but not in another. The reason for this mixed result is unclear. Possibly, children’s smiling during pretend actions produced with unusual objects in a rather unfamiliar environment may be less consistent than their smiling in naturalistic play settings. The role of children's social smiles in different contexts is therefore still in need of further empirical verification. Specifically, an examination of children’s smiling in pretend play contexts that more closely replicate those of children’s day-to-day playing experiences is required. In short, whether or not young children convey through social cues that they treat pretence as a particularly social activity (compared to other more practical forms of action) remains to be empirically verified. The primary purpose of the present study was to investigate how young children (2 to 3 year olds) display social behaviour during pretend play compared to their social behaviour during non-pretend play. Specifically, the study explored whether children would engage in more social gaze (i.e. looking towards an experimenter), both with and without expressions of positive affect (smiling and laughing) during pretend play, than they would during functional play. It was hypothesized that children would indeed engage in more social gaze, both with and without expressions of positive affect, during pretend play than they would during functional play. Previously, the social-cultural perspective of pretend play development has been mentioned, alongside research findings that young children’s pretence is enhanced or “scaffolded” by adults. Namely, pretend play is an inherently social activity and the production of pretend actions, be they novel or imitated, increases after witnessing an adult performing (or “modelling”) pretend actions. At the more extreme end of the social-cultural learning argument some have suggested that pretend actions may emerge entirely as a result of cultural learning (e.g., Striano et al, 2001). Evidence for this view is taken from studies demonstrating that young children imitate pretence, and before the age of 2 years, appear to rely almost exclusively on imitation to produce pretend acts (e.g. Striano et al, 2001). Given the potentially crucial role of modelling by adults, it would be reasonable to expect that if modelling is able to increase the overall quantity of pretend play actions performed by children, it may also influence the quality of pretend play actions. Some evidence so far does suggest that the quality of play is enhanced with modelling, in that children tend to enact more complex pretend scenarios and use less physically similar objects in object substitution Children’s Communication of Pretend Acts Using Social Cues 131 after adult modelling (for a review, see Mitchell, in press). It may be that after witnessing an adult modelling pretence, children are primed to enter into the pretend frame with the other person, which may lead them to produce more social cues during their play after modelling, particularly in pretence and particularly in their imitated pretend play actions. It was therefore further hypothesised that the overall frequency of pretend actions, as well as the proportionate elicitation of social cues during pretend actions, would increase after children have witnessed an adult pretending. It was also hypothesised that pretend acts directly imitated by children would be associated with a greater proportion of social gaze than non-imitated pretend actions. To explore these hypotheses, children were invited to play by an experimenter and were given an opportunity for free-play with some dolls, a doll house and a large set of objects including some typically used for pretending by children (e.g. dolls, replica objects) and some not typically used for pretending (e.g. cardboard, hairclips). A modelling phase was introduced where the experimenter enacted several pretend scenarios to the children. Children were then given another opportunity for free-play after adult modelling. Pretend play and functional play actions before and after modelling was coded, and social behaviours during each play type were compared.