부산대학교 도서관

The discovery, in the monkey, of "mirror" neurons, which fire in response both to the performance and to the observation of specific actions, has prompted extensive research into their properties, and into the possible functions of a putative mirror neuron system in humans. Little is known, however, about how such neurons acquire their matching properties. This thesis addresses this question using a variety of techniques. Imitation is one of the key processes thought to be subserved by the mirror neuron system Chapter 3 shows that automatic imitation effects are separable from spatial compatibility effects. This establishes automatic imitation effects as suitable targets for experimental manipulations of mirror neuron system function. Strengthening this conclusion, Chapter 4 indicates that automatic imitation effects can be delayed by repetitive theta burst transcranial magnetic stimulation (TMS) of the inferior frontal gyrus, an area homologous with the premotor F5 mirror neuron area in the macaque. In Chapter 5, single-pulse TMS is used to produce motor evoked potentials (MEPs). In an action observation experiment, an automatic muscle-specific "mirror" effect is shown: the size of the MEP in a given muscle is sensitive to the identity of the muscle that would be used to perform the observed movement. It is then demonstrated that this effect can be reversed following a period of incompatible sensorimotor training. This result is built upon in Chapter 6: it is shown behaviourally that incompatible sensorimotor training can reduce automatic imitation effects, and, using functional magnetic resonance imaging, that it can reverse neural responses to observed actions in the human mirror neuron system. It is concluded that sensorimotor learning can reconfigure the human mirror neuron system, and that it is, therefore, a mechanism through which the mirror neuron system can acquire its ability to match observed with performed actions.