The role of learning in the development of imitation systems, both in normal development and autism spectrum disorders.

In our highly social culture, it has been proposed that the ability to imitate is invaluable for skills acquisition and other social abilities which require us to understand another person‘s perspective, such as empathy. The origin of imitation systems is hotly debated. One hypothesis suggests that these systems evolve through natural selection to support other social abilities. Alternatively, these systems may develop through learning, such that actions are executed at the same time as observing matching actions, for example, through self-observation. The proposed research will distinguish these two hypotheses, by looking at the effects of training, both on the cortical systems mediating imitation in a normally developing population, and on behavioural imitation effects in a population where there is likely to be less experience which can lead to the development of imitation systems, namely individuals with autism spectrum disorders. If these systems evolve through natural selection, there will be little effect of training, but if they develop through experience, training will have dramatic effects on imitation and its supporting systems. Understanding of the origin of imitation systems has important implications for treatment of disorders where these systems are impaired, such as autism spectrum disorders and apraxia.

The ability to imitate is invaluable, allowing us to learn new skills from other people, and perhaps supporting other social abilities which require us to understand another person‘s perspective, such as empathy. The origin of imitation systems, namely systems which will translate observed actions into matching executed actions, is highly controversial. Under one hypothesis, these systems evolve through natural selection to support higher sociocognitive functions (phylogenetic hypothesis). Another hypothesis suggests that these systems develop through learning, and that the learning process is driven by experience in which specific actions - those to which the system will subsequently be responsive - are concurrently observed and executed (ontogenetic hypothesis).

The first strand of work will address whether the phylogenetic or ontogenetic hypothesis can better explain a human bias in imitation, such that observed human actions are imitated more than observed non-biological movements. The experiments in this strand of work will use electroencephalography and magnetoencephalography to record cortical activity when participants observe human or non-biological stimuli, before and after training where they execute actions while observing non-biological movements. Before training, both hypotheses predict activation of imitation systems when observing human, but not non-biological, movements. After training, the phylogenetic hypothesis predicts the same pattern of results, but the ontogenetic hypothesis predicts activation of imitation systems under both observation conditions.

The second strand of work will ask whether a group who are likely to have less experience of the type that can lead to the representation of some movements in imitation systems, namely individuals with autism spectrum disorders, exhibit less imitation of these movements than a control group, and whether this impairment can be reduced through training. The ontogenetic hypothesis would predict that any impairment reduces following this training, in contrast to the phylogenetic hypothesis, which would predict that any imitation impairment present before training is also present after training.

Understanding of the origin of imitation systems is of great significance to those researching the function of imitation systems, in addition to those researching functions in which imitation systems have been implicated e.g. empathy and theory of mind. This work also has important implications for individuals where imitation systems may be impaired e.g. those with autism spectrum disorders and apraxia.