Given their press, you would think that mirror neurons solved some great big problems with understanding the brain. But the fuss seemed to me to be premature. We do not need more ‘magic’ ideas; we need explanatory ones. Now it appears that mirror neurons may be part of a useful system.

Research Digest has a posting on research by R. Mukamel and group that uses recording from single cells in humans (here).

Most of the 1177 cells that were recorded showed a response either to the execution of an action or the sight of that action, not both. However, there was a significant subset of ‘mirror’ neurons in the front of the brain, including the supplementary motor area, and in the temporal lobe, including the hippocampus, that responded to the sight and execution of the very same actions.

So it seems humans do have them.

Critics could argue that rather than having mirror properties, these cells were responding to a concept. For example, according to this argument, a cell that responded to the sight of a smile and the execution of a smile, was actually being activated by the smile concept. Mukamel’s group reject that argument. They had a control condition in which the words for actions appeared on a screen, rather than those actions being seen or performed. The postulated mirror neurons responded to the sight and execution of an action, but not the word.

This logic implies that there are no non-verbal concepts. It is just a semantic quibble. We could call them pseudo-concepts or motor-cepts or the like. There is no reason that mirror neurons could not be responding to a motor action as if it were a concept, in a concept-ish sort of way.

Another potential criticism is that the execution-related activity of a postulated mirror neuron is triggered by the sight of one’s own action, rather than by motor-output per se. However, this can’t explain the mirror neurons that responded both to the sight of a given facial expression and one’s own execution of that facial expression (although proprioceptive feedback could still be a potential confound).

So it is not just a simple sensory identification of a motor act.

Mirror neurons make functional sense in relation to empathy and imitative learning, but a drawback could be unwanted imitation and confusion regarding ownership over actions. The researchers uncovered another subset of cells that could help reduce these risks - these cells were activated by the execution of a given movement but inhibited by the sight of someone else performing that same movement (or vice versa).

There appears to be quite a separate system for identifying the ownership of actions that does not necessarily include these cells. We do not need to postulate that mirror neurons are involved in that and if they are not involved, why would they confuse ownership. Research would have to be done to show if they were actually tied to the ownership identification.

‘Taken together,’ the researchers concluded, ‘these findings suggest the existence of multiple systems in the brain endowed with neural mirroring mechanisms for flexible integration and differentiation of the perceptual and motor aspects of actions performed by self and others.’

‘flexible integration and differentiation of the perceptual and motor aspects..’ of something sounds like definition of a concept to me.