How does "mirroring" support joint action?

The discovery of mirror neurons (e.g., Gallese et al., 1996) has reignited interest in theories that postulate a tight functional link between perception and action. According to these theories, perception and action share a common representational code, with actions coded in terms of the distal perceptual effects that they produce (Prinz, 1997). Accordingly, action representations should be activated when perceiving actions or perceiving the perceptual effects generated by actions. Mirror neurons have been regarded as a neural substrate implementing this functional principle. Perhaps themost important implication of common coding is that it establishes a social link between actor and observer that supports action understanding (Rizzolatti and Sinigaglia, 2010) and/or prediction (Wilson and Knoblich, 2005). How might close perceptioneaction links help people to perform actions together? In the sections that follow,we discuss three possible functions of mirroring for joint action: 1) supporting temporal coordination in real time, 2) enabling seamless integration of one’s own and others’ actions in joint action planning, and 3) enabling groups to imitate the coordinated actions of other groups. Close perceptioneaction links support temporal coordination by enabling predictions about others’ actions without the need for a separate perceptual prediction mechanism. Emulator theories postulate that forward models in the motor system automatically generate continuous real-time predictions about the expected sensory consequences that will result from performing an action (Wolpert and Kawato, 1998). Combining this postulate with the assumption of an automatic perception action match leads to the hypothesis that the same forwardmodels used in action execution can also be used to generate predictions about the future course of actions that others are currently performing (Csibra, 2008; Kilner et al., 2007; Knoblich and Flach, 2001; Wilson and Knoblich, 2005).