The classical mirror neuron theory of action understanding asserts that when we observe an action, the representations that are engaged for performing it, are automatically activated. In order to do gain information about the role of the simulation in action understanding a state dependent TMS experiment has been carried out. The fundamental idea is to adapt a neural population in the motor system and then testing the effects of this adaptation when participants categorize visually presented actions. The second aim of the present work is to find a paradigm, or a particular cognitive set, that does not allow the simulation process to take place when the participants are observing actions. This step will be important in testing whether the simulation process is necessary in order to understand a visually presented action.