Mirror neuron research has come a long way since the early 1990s, and many theorists are now stressing the heterogeneity and complexity of the sensorimotor properties of fronto--parietal circuits. However, core aspects of the initial 'mirror mechanism' theory, i.e. the idea of a symmetric encapsulated mirroring function translating sensory action perceptions into motor formats, still appears to be shaping much of the debate. This article challenges the empirical plausibility of the sensorimotor segregation implicit in the original mirror metaphor. It is proposed instead that the teleological organization found in the broader fronto--parietal circuits might be inherently sensorimotor. Thus the idea of an independent 'purely perceptual' goal understanding process is questioned. Further, it is hypothesized that the often asymmetric, heterogeneous and contextually modulated mirror and canonical neurons support a function of multisensory mapping and tracking of the perceiving agents affordance space. Such a shift in the interpretative framework offers a different theoretical handle on how sensorimotor processes might ground various aspects of intentional action choice and social cognition. Mirror neurons would under the proposed "social affordance model" be seen as dynamic parts of larger circuits, which support tracking of currently shared and competing action possibilities. These circuits support action selection processes-but also our understanding of the options and action potentials that we and perhaps others have in the affordance space. In terms of social cognition 'mirror' circuits might thus help us understand not only the intentional actions others are actually performing-but also what they could have done, did not do and might do shortly.
The caricature view of mirror neurons and the ensuing debateMirror neurons were initially discovered by Giacomo Rizzolatti and colleagues in Parma (di Pellegrino et al., 1992[1]. The core finding that inspired the name was that certain single cells in the macaque premotor cortex were modulated both by the execution of some goal--directed actions and by the perception of others performing similar actions. Such action--sensitive sensorimotor modulations were also found in parietal areas known to be highly anatomically and functionally interconnected with the premotor area F5. Additionally, indirect behavioral and imaging findings suggested homologue modulations in humans. The Parma group soon hypothesized that these fronto--parietal neurons together formed a mirror system by which