Goal-Driven Vs. Mimetic Imitation in School Children with Typical Development and Autism Spectrum Disorders

Background:  Imitation is known to be affected in people with Autism Spectrum Disorders (ASD), but it remains unclear whether these problems should be taken as a primary deficit in autism, or whether they are a consequence of some other core deficits. This study started from Hamilton's (2008) proposal on the existence of two routes for imitation, and from her suggestion that a deficit in a mimetic route could account for the imitation pattern typically found in children with autism.

Objectives:  Taking advantage of the end-state comfort (ESC) effect (Rosenbaum et al., 1990), which allowed us to model different action patterns fulfilling the same goals, but differing importantly in the efficiency of the observed action patterns, we designed a variant of the grasp imitation task reported in Hamilton et al. (2007), and analyzed whether a large group of children and adolescentes with typical development (TD) emulated the goal, or reproduced the observed action patterns, even when they were not consistent with the ESC effect. We also tested a group of participants with high functional autism or Asperger syndrome, and analyzed their differences with respect to the control sample.

Methods: The task required participants to imitate the action of a model who grasped a wooden bar held horizontally over two cradles, and who grop it by inserting one of its two extremes on a metal pen holder. We manipulated the salience of the goal by using either a uniformly white bar, or a black-and-white bar, that made the goal of inserting one extreme of the bar in the container much more salient. The modeled grip (overhand vs. underhand) and the modeled end state (thumb down vs. thumb up) were also manipulated randomly between trials, to assess whether participants tended to emulate the goals, or to  mimetically reproduce even the less functional modeled patterns.  

Results:  The results indicated that even the youngest children tested (7 to 9 years old) were affected by the ESC effect, and that participants with ASD also showed these planning effects. With age, participants with TD were progressively more  prone to replicate even the less functional actions performed by the model. Participants with ASD emulated the observed goals to the same extent than did the control group, but they tended to imitate the specific actions less frequently when they ended in uncomfortable end states.

Conclusions: The results are consistent with the claim that participants with ASD are able to emulate the goals achieved by a model in an explicit imitation task, but that they show a delayed pattern of mimetic imitation, which is consistent with the dual-route model of imitation proposed in Hamilton (2008)