The role of the mirror mechanism (MM) in social cognition is still a matter of debate. Much disagreement is primarily due to different notions of action and of action understanding. On the standard view, an action is directed to a goal in virtue of the action's being appropriately related to some high-order mental states (e.g., beliefs, desires, and intentions) representing this goal. The relation between actions and their goals is traditionally assumed to be largely independent of the motor processes and representations underpinning action execution. These processes and representations allegedly concern motor features such as joint displacements or muscle contractions only. However, two decades of cognitive and neuroscientific research on the cortical motor system has repeatedly challenged this view. Neurophysiological and behavioral evidence (Rizzolatti & Sinigaglia Reference Rizzolatti and Sinigaglia2010) has demonstrated that motor processes may involve motor representations of action goals (e.g., to grasp, to place, etc.), and not only kinematic or dynamic components of actions. This suggests that beliefs, desires, and intentions are neither primitive, nor the only bearers of intentionality in action. Indeed, motor representation is enough to ground the directedness of an action to its goal (Gallese Reference Gallese2000; Butterfill & Sinigaglia Reference Butterfill and Sinigaglia2014).

The MM indicates that this holds not only for action execution, but also for action observation. The activation of the MM can be selectively related to the observed action goal regardless of its kinematics, dynamics, and the body effector involved (Gallese & Sinigaglia Reference Gallese and Sinigaglia2011; Rizzolatti & Sinigaglia Reference Rizzolatti and Sinigaglia2010). This supports the claim that the MM may play a role in action understanding. By recruiting her own motor representation of the goal to which the observed action is directed, the observer may understand what the agent is doing without needing any high-order processing.

Claiming that the MM plays a role in action understanding of course does not imply that action understanding is overall solely explained by the MM. Understanding an action is a complex process. It involves at least representing to which (proximal and distal) goals the action is directed; identifying which beliefs, desires, and intentions specify reasons explaining why the action happened; and realizing how those reasons are linked to the agent and to her action. The MM enables the representation of the goals of others' actions by taking advantage of one's own motor cognition. The richer a person's own motor cognition is, the greater her sensitivity to another's action and the better her ability to grasp the goal to which that action is directed. Consistently, action understanding deficits occur following specific impairments in the recruitment of the motor representation of action goals (for a review, see Gallese & Sinigaglia Reference Gallese and Sinigaglia2011).

The fact that a motor-based action understanding doesn't match any “established category of psychological functioning” (target article, sect. 9.1) is not per se a good reason to reject it. On the contrary, it provides a new empirically and theoretically sound framework to investigate basic aspects of social cognition. Furthermore, differently from what Cook et al. maintain, it enables us to shed new light on the various components (e.g., perceptual, motor, and mentalizing) of action understanding, as well as on their relationship. To this extent, it is worth noting that even high-order purely sensory mechanisms, like those characterizing extrastriate cortices such as the Superior Temporal Sulcus (STS) region, are insufficient to represent the goal of a given action at the same general level as the MM does. In addition, there is no evidence to date that the STS responds both to proximal goals (e.g., grasping a piece of food) and to distal action goals (for eating or for throwing away), as the MM does (Gallese & Sinigaglia Reference Gallese and Sinigaglia2011; Rizzolatti & Sinigaglia Reference Rizzolatti and Sinigaglia2010).

Pertaining to the relationship between the motor components of action understanding and mentalizing, many studies have demonstrated that the MM kicks in when people have to understand to which goal an observed action is directed (de Lange et al. Reference de Lange, Spronk, Willems, Toni and Bekkering2008; Liepelt et al. Reference Liepelt, von Cramon and Brass2008a). When people must determine the reasons why an agent performed a given action, additional activations of cortical regions such as the mesial anterior frontal cortex, the anterior cingulate cortex, and the temporo-parietal junction – typically considered to belong to the so-called mentalizing network – are detected (de Lange et al. Reference de Lange, Spronk, Willems, Toni and Bekkering2008; Liepelt et al. Reference Liepelt, von Cramon and Brass2008a; Van Overwalle Reference Van Overwalle2009). In spite of many theoretical attempts to integrate these different components of action understanding, so far there is neither convincing evidence about the mentalizing specificity of these activations, nor a theoretically coherent and empirically motivated explanation of how the “mentalizing network” might work. Ironically, such controversial aspects of the neurobiological bases of mentalizing have not attracted so much debate, certainly not as much as the role of the MM in social cognition.

Finally, the functional properties of the MM and its involvement in action understanding are not captured by either a strictly genetic or a purely associative account about its origin. We hypothesize (Gallese et al. Reference Gallese, Rochat, Cossu and Sinigaglia2009) that an innate rudimentary MM is already present at birth, which can then be flexibly modulated by motor experience and gradually enriched by visuomotor learning. Indeed, such a system could be an ideal candidate for the neural underpinning of neonatal facial imitation in humans and nonhuman primates. Recent neurophysiological evidence also suggests there is an inborn rudimentary form of action mirroring in neonate rhesus macaques (Vanderwert et al. Reference Vanderwert, Fox and Ferrari2013). Differently from the associative account proposed by Cook et al., our hypothesis entails a primacy of motor experience in the development of the MM and its contribution to understanding others' actions. This primacy is supported by several studies showing a causal link between the ability to produce an action and the ability to understand it (see, among others: Cannon & Woodward Reference Cannon and Woodward2012; Kanakogi & Itakura Reference Kanakogi and Itakura2011).