We investigated ability to understand goals and attribute intentions in the context of two imitation studies in low-functioning, nonverbal children with autism (L-F CWA), a population that is rarely targeted by research in the domain. Down syndrome children (DSC) and typically developing children (TDC) were recruited to form matched comparison groups. In the two sets of simple action demonstrations only contextual indicators of the model's intentions were manipulated. In the Head touch experiment the model activated a button on a toy by pushing it with the forehead, whereas in the Hidden box experiment the model used a ball with a magnet to lift a box out of its container. Both actions were unusual and non-affordant with regards to the objects involved, none of the children in the baseline condition produced them. L-F CWA imitated the experimenter exactly, regardless of the model's intention. TDC showed appreciation of the model's intention by imitating her actions selectively. DSC reproduced only the intentional action as often as they imitated the experimenter exactly. It is concluded that L-F CWA attributed goals to the observed model, but did not show an appreciation of the model's intentions even in these simplified, nonverbal contexts.
Children interpret and predict others’ actions on the basis of the mental states they attribute to the actor. Autism is associated with a specific cognitive deficit in inferring and representing mental states, as documented by seminal studies showing difficulties with false belief tasks (Baron-Cohen, 1995, Baron-Cohen et al., 1985, Leslie and Thaiss, 1992, Perner et al., 1989 and Sodian and Frith, 1994) and with pretend play (Wing, Gould, Yeates, & Brierley, 1977). Mental states, however, vary in nature; beliefs, desires, goals, intentions, emotions as well as perceptions have been proposed in literature (Frith et al., 1991, Luo, 2011, Premack and Woodruff, 1978, Saxe et al., 2004 and Vivanti et al., 2011). Goals, for instance may be considered as mental states that are more ‘transparent’ or observable in behaviour than beliefs and desires, at least under a broad mentalising theory (Hamilton, 2009). Still, relatively few studies have investigated goal understanding in autism and most of these studies have involved high-functioning children with autism (H-F CWA). One of the aims of this study therefore was to assess goal understanding on the other end of the autistic spectrum, in low-functioning children with autism (L-F CWA). A further issue with existing reports is that they do not distinguish in all cases between an understanding of a goal as an internal state and the understanding of the visible outcome of a goal directed action, without inferring an intentional mental state. Therefore the other aim of our study was to better control this possibility by using imitation paradigms where the children's responses may indicate for us the different intentional mental states they attribute to the model across the experimental conditions.
