Broken mirrors
Do children with autism have difficulty imitating the actions of others? The answer is hotly debated in the field.
Proponents of the so-called broken mirror hypothesis contend that people with autism have faulty mirror neurons, cells that are typically active when mimicking others. Problems with the mirror neuron system could explain many of the social deficits linked to autism, these researchers say.
However, research exploring this idea has had mixed results. Some studies show children with autism have trouble imitating people, but others suggest they behave similarly to controls.
A new study, published 18 July in Journal of Autism and Developmental Disorders, suggests that problems with imitation may be limited to emotional contexts, and may explain the seemingly contradictory results.
Imitation is thought to be an essential part of human interaction and communication. People naturally mimic the facial expressions of those around them, even if not asked to do so.
In the new study of 15 high-functioning children with autism and 15 controls, researchers found that both groups are equally able to imitate a finger movement performed by someone with a neutral facial expression. However, the controls perform much better when the person doing the finger task has an emotional expression rather than a neutral one. The autism group, in contrast, shows no change.
In a second experiment, the same groups of children had to name different facial expressions and then mimic them. Children with autism are able to name emotions linked to facial expressions almost as well as controls are. (They are worse at naming happy expressions, a difference that may be the result of weaker language skills in these children.) But the children with autism have a much more difficult time imitating fearful facial expressions than controls do.
Taken together, the results suggest that children with autism can imitate movements and recognize facial expressions as well as controls can, but they have problems when these two skills are put together — imitating others in a social context.
A better understanding of the imitation deficits may help improve behavioral treatments. For example, some therapies employ imitative play, in which the therapist mimics the child’s action. Research suggests this approach can improve eye contact and other measures of social responsiveness.
Comments
Any explanation of why an emotional expression should improve imitation of finger movements?
Thanks for your question, Jon. I'm looking around for an answer and will post once I find one.
Hi Jon,
I believe imitation is influenced by the emotionality conveyed by the model to imitate. In other words, emotions influence and boost the learning of new behaviors (imitation leads to the acquisition of a new behavior). We found that the emotionality of the model to imitate facilitated imitation velocity in subjects without autism, but not in autistic children (actually they are slow down in the emotional imitation, though not statistically significant, as if emotions are distracting them or worsening their performance). This lack of emotion-imitation integration and facilitation may partially explain why these children tend to isolate and block their development of social abilities that are normally based on mutual influence and imitation.
Thanks Alessandro. Intriguing.
I'd assumed from the description that it was a positive emotion. In fact the emotional cue in all cases was (a pretty scary looking) fearful face. Why would a fearful face enhance the ability to imitate (in typically developing children)? Intuitively, I'd expect it to be the opposite - which is the pattern the autistic kids actually do show!
There are several explanations I can think of for why emotional faces "prime" imitation. One is that the emotion somehow primes the brain for action in general. Another is that the emotional face attracts attention to the centre of the screen, which is where the hand then appears. A third possibility comes from the fact that participants had to wait until the movement was completed before imitating it and perhaps the emotional face affected the inhibition of response rather than imitation per se. In none of these is it clear that the effect is specific to imitation. Have you or others looked at whether the same effect obtains if there's a non-imitation task to perform (e.g., responding to a number on the screen or something similar)? If not, that might be a useful control condition for future studies.
As you say, the autistic kids did show an effect of emotion but in the opposite direction. Although statistically non-significant, the effect was actually numerically larger than the effect in the control group. This can *only* be because the effect was more variable in the autistic kids (ie there were some kids who showed a "normal" effect of emotion and others who showed a large effect in the opposite direction). My guess (and it is purely a guess) is that the effect size for a given individual might depend on their overall RT. In other words, there were some kids who were quick to respond and showed the "normal effect" and others who were slow to respond and showed an exaggerated effect in the "wrong" direction.
Sorry for the barrage of questions. As I said, it's very intriguing!