Can we have direct access to other minds by “motor resonance”? A critical reassessment

Abstract

The serendipitous discovery of mirror neurons in cognitive neuroscience has recently revived interest in the old philosophical “problem of other minds”. The remarkable thing about the mirror neurons is that they fire not only when an individual is performing an action, but also when that individual passively observes a similar action performed by another individual. It is widely assumed that mirror neurons provide a “motor resonance” mechanism by which we can understand the actions of others by directly mapping the actions of other people onto our own motor system. This mechanism seems to bridge the gap between the self and the other, thus providing a direct access to other minds. Here we systematically review the motor resonance theory, discuss its current problems and challenges, and provide suggestions for future research. We firstly introduce the framework of the theory with its definitions, scopes and methods. In particular, we clarify several ambiguous notions in the literature such as “intra- vs. inter-personal resonance” and “movement vs. action”. Then we outline empirical evidence in support of the theory from two bodies of research: Non-invasive neuro-stimulation (e.g. transcranial magnetic stimulation) and clinical studies with action understanding deficits (e.g. apraxia, Broca’s aphasia and autism). These studies suggest the motor resonance is the necessary and sufficient condition for direct access to other minds. Next, we summarize major criticisms of motor resonance theory in recent years. Ever since mirror neurons were discovered, there has been a great debate on its functions and ontogeny. Recent studies question its “direct-mapping” property by showing that sensory representation of an observed action can be arbitrarily mapped to the motor representation of a different action after sensorimotor training. This suggests the motor resonance may merely be a byproduct of associative learning, rather than an evolutionary selected mechanism for action understanding. In contrast, the Genetic hypothesis suggests motor resonance provides a unique direct-mapping mechanism for direct access to other minds, thus functions as an adaptation for action understanding. This claim has been supported by neonate imitation research as well as evidence showing early stage activity ( Other criticisms challenge the directness of motor resonance by showing that motor resonance can be modulated by multiple contextual, social and cultural factors. This implies the mirror neuron system may intricately interact with other brain regions during motor resonance. Therefore, some suggests action understanding should be examined in the framework of “extended mirror neuron system” which includes other functional routes (other than motor resonance) such as the mentalizing system. We discuss three directions for future research. First, we believe most debates on motor resonance are “theoretical” instead of “empirical” issues. Theoretical dilemma such as “strict congruent vs. broadly congruent mirror neurons” cannot reconcile with the core “direct-mapping” mechanism; different levels of inference (e.g. low-level kinematics vs. high-level intention) should be separately theorized. More precise definitions and taxonomy in this field will allow for better assessments and hence a more fruitful empirical research. Second, the field could greatly benefit from investigating the “extended mirror neuron system”. These distribute networks interact with classical frontal-parietal circuit during motor resonance and facilitate overall understanding of other’s behavior. Future research is needed to elucidate the relationship between them. Finally, due to the technical limitations of current neuroimaging methods, it is challenging to obtain direct evidence that unequivocally proves or rejects motor resonance theory. We would encourage researchers to employ advanced technique with excellent spatial and temporal resolution (e.g. intracranial electroencephalography) to directly examine the processing sequence as well as the implicated brain regions during motor resonance.