Towards Collaborative Pretence and Collective Intentionality: Metacommunication In the Pretend Play of Children with Autism

Background:

Impairments in joint attention, social imitation and social-emotional reciprocity mean that children with autism find spontaneous social play very difficult (Jordan 2003; Wolfberg 2009). Carpendale and Lewis (2004) indicate an association between children’s false belief recognition and their engagement in metacommunication in pretend play. However social pretend play is by nature a metacommunicative activity: for play to evolve successfully, the players must communicate in the creation, expansion and sustainment of a shared make-believe world (Giffin 1984).

Objectives:

The aim of this exploratory study was to investigate engagement in the social pretend play of children with autism – in particular, the extent to which these children could initiate and sustain a shared pretend scenario and make use of metacommunicative strategies to achieve this. We hypothesized that the children would not seek to engage in collaborative pretend play.

Methods:

The corpus consisted of 30 hours of videotaped free play of child – adult dyads. 5 English-speaking children with autism (ages 3;6 – 7;2) participated in the study, all of whom had undergone a team assessment from a recognised child mental health service.  We identified 64 pretend play sequences in the corpus. The data was approached from a conversational interaction perspective involving detailed qualitative examination of the play sequences in their linguistic and non-linguistic context, to determine whether the children sought to recruit the adult into their play. Specifically, we analysed verbal and nonverbal metacommunicative strategies employed by the children, using an adaptation of Giffin’s (1984) coding scheme for typically developing children’s play. We also coded the data according to our own continuum of engagement in social play.

Results:

While only one child engaged in true collaborative pretend play, all of the children shared the pretence to varying degrees: some children did not actively seek collaboration in the pretence, but still interacted with the adult play partner, while for others the adult was invited to share in the pretence but actively discouraged from making creative contributions.

All of the children used metacommunicative strategies applicable in both solitary and collaborative play (e.g. sound effects). However, only three children demonstrated the ability to utilize and respond to metacommunicative strategies exclusive to collaborative settings: negotiation of the pretence (within or outside the play frame) and sharing in the creative process. Even the most able children exhibited a lack of flexibility in shifting in and out of the play frame and incorporating proposals by the interlocutor into the play.

Conclusions:

While research shows children with autism are capable of pretend play (e.g. Jarrold, Boucher & Smith 1996), social pretence has received less attention. Our detailed analysis of metacommunication shows children participating in pretend play interactions in a directive fashion, but with varying levels of engagement and collaboration, from positioning the adult as audience to clearer attempts to establish collective intentionality. Crucially, even the most competent children displayed atypical behaviours in metacommunication. The results support a nuanced continuum model of collaborative pretend play ability.