Mirror neurons, discovered by Giacomo Rizzolatti and colleagues in the early 1990s in the premotor cortex of macaque monkeys, fire both when the animal performs a specific action (e.g., grasping a peanut) and when it observes another individual performing the same action. This "mirroring" property has generated enormous interest and controversy, with claims that mirror neurons underlie action understanding, imitation, empathy, language evolution, and that their dysfunction contributes to autism.
Evidence
In monkeys, mirror neurons have been directly recorded in premotor cortex (F5) and inferior parietal cortex. In humans, indirect evidence from fMRI, TMS, and EEG (mu suppression) suggests a mirror neuron system in homologous regions, though individual mirror neurons have not been directly demonstrated in humans. The human mirror system is activated during action observation, imitation, and understanding of others' actions and intentions.
The mirror neuron concept has been criticized for overinterpretation. Gregory Hickok and others argue that the evidence does not support claims about mirror neurons underlying language or empathy, that action understanding can occur without motor simulation, and that the "broken mirror" theory of autism is not well supported. The core finding — neurons with both motor and observation properties — is established, but the broader theoretical claims remain contentious.