4.2.3 Metarepresentation and pretend play
Alan Leslie (1991) has suggested that understanding mental states such as false belief requires the sophisticated skill of ‘de-coupling’ or disengaging (mentally speaking) from the truth of a situation (e.g. ‘The car is in Park Street’), in order to hold in mind an idea that differs from this reality (‘Jane thinks the car is in Mount Street’). This capacity is known as metarepresentation and is seen as a crucial element of language understanding.
Leslie argues that typically developing children display simple de-coupling at around eighteen months, when they start enacting pretend play. In his words, when a child puts a banana to his/her ear, pretending that it is a telephone, s/he is temporarily disengaging from the reality (‘This object in my hand is a banana’) in order to indulge in the pretence (I'll pretend that ‘this object is a telephone’). The principle that a person's belief/knowledge about a situation depends partly on what perceptual information has been available to them. Children typically grasp this between 36 and 48 months. This ‘simpler’ metarepresentational skill may act as a developmental precursor for understanding that one's own or other people's thoughts can be hypothetical or different from reality. Figure 5 illustrates how acts of both pretending and false belief about a situation can be seen as metarepresentations.
There ia a similar idea of metacommunication, which refers to how children communicate their disengagement from reality in social pretend play. Leslie's primary emphasis is on characterising what thought processes are necessary for pretence – whether solitary or social.
There is much experimental and observational evidence that children with autism fail to develop early ‘pre-mind-reading’ skills. Box 5 summarises some of this evidence in relation to gaze following, proto-declarative pointing and pretend play.
Box 5: Developmental pre-cursors of theory of mind
Leekam et al . (1997) tested whether children with autism would spontaneously follow, with their eyes, an experimenter sitting opposite them, who changed her head direction to look at a toy. The children showed significant impairment in this gaze monitoring task compared with control participants. In another study, Baron-Cohen (1989) tested whether children with autism would use ‘protodeclarative’ pointing to indicate an object of interest, with similar results.
Baron-Cohen (1987) gave autistic children a range of toys and observed how they played with them. The children engaged in as much ‘functional’ play, such as ordering or stacking bricks, as a control group, but showed much less ‘symbolic’ or pretend play than the control group, such as using a brick as a cup, a box as a car.
These experimental results were confirmed by a survey conducted by Baron-Cohen et al . (1992). A questionnaire asking about the presence of the above behaviours was completed by the health visitors and parents of three groups of children: 20 autistic children; 20 younger siblings, aged around 18 months; and another group of 50 toddlers aged 18 months. All of the 50 ‘normal’ toddlers had the key behaviours, while a majority of the autistic children lacked them. Among the younger siblings, who were considered genetically ‘at risk’ of developing autism (see Section 5 ), one child lacked the key behaviours and subsequently received a diagnosis of autism.
The ToM ideas considered here provide an introduction to an extensive body of related theories and research findings. We now turn to some key difficulties.
Gaze following/gaze monitoring : The skill of following where someone is looking and looking there too in order to share attention. Infants typically develop this behaviour at around eight months.
Proto-declarative pointing : The skill of pointing to indicate an object of interest, as opposed to pointing in order to ask for an object to be fetched. Infants typically develop this behaviour at around twelve months.
Seeing leads to knowing : The principle that a person's belief/knowledge about a situation depends partly on what perceptual information has been available to them. Children typically grasp this between 36 and 48 months.
Metarepresentation : Process of disengaging from reality in order to think about one's own or another person's thoughts. Considered necessary for false belief, pretence, etc.