The characterizing feature of the primary visual cortex of primates is the presence of neurons sensitive to stimulus orientation. Regardless of whether the orientation-sensitive neurons are determined genetically or acquired by experience, or both, they are at the basis of the functional organization of primates' visual system (Hubel & Wiesel Reference Hubel and Wiesel1998; Marr Reference Marr1982). The same is true for mirror neurons (MNs). Regardless of whether their properties are determined genetically, acquired by experience, or both, they represent the neural substrate of a fundamental mechanism that transforms sensory information into a motor format (the mirror mechanism). The functions of the mirror mechanism vary from action understanding, to imitation, to empathy, and even, in birds, to song recognition (Rizzolatti & Sinigaglia Reference Rizzolatti and Sinigaglia2010). Their function depends on their anatomical location. Thus, contrary to the view advanced in the present target article, the problem of how MNs originate is utterly irrelevant as far as their function is concerned. It is an interesting problem, but it has little to do with the function of the mirror mechanism.
The claim that mirror neurons are just “another type” of association neurons misses their characterizing, unique property, which is that of giving a motor format to sensory stimuli. This misunderstanding can be also found in an interesting paper on mirror neurons by Damasio and Meyer (Reference Damasio and Meyer2008). They claimed that the parieto-frontal mirror neurons are neural ensembles included in higher-order association areas called “non-local convergence-divergence zones” that collect information from lower-order visual, auditory, and somatosensory association areas, and signal back to those areas. Action understanding depends on the activation of this network. This proposal had the merit of highlighting the role of top-down connections in action understanding. It overlooked, however, as done in the present target article, the fact that parieto-frontal mirror neurons are motor neurons. When MNs discharge, they “ignite” motor schemata similar to those endogenously activated during motor imagery and, within limits, during actual motor act execution. In other words, my motor schemata are activated during the observation of similar motor schemata of others. This provides a neurophysiological account of the mechanism underlying action understanding “from inside” (Rizzolatti & Sinigaglia Reference Rizzolatti and Sinigaglia2010): “a first person process, where the self feels like an actor, rather than a spectator” (Jeannerod Reference Jeannerod1997, p. 95; emphasis added). This appears to be a function that only the mirror mechanism is able to mediate.
Mirror neurons have top-down effects (Damasio & Meyer Reference Damasio and Meyer2008, see also Kilner et al. Reference Kilner, Friston and Frith2007b). In other words, following MN activation, signals go not only toward other motor areas, but also backwards to lower-order visual, auditory, and somatosensory areas. This top-down activation binds the understanding of what a person is doing (e.g., grasping), encoded in the motor cortex, with the visual details of that action. An interesting possibility is that the top-down mechanism also has another function, which is: to be the neural substrate of a learning mechanism that starts from motor centers rather than from the environment. An elegant experiment by Van Elk et al. (Reference Van Elk, van Schie, Hunnius, Vesper and Bekkering2008) illustrates this point well. EEG was recorded during observation of action videos in 14- to 16-month old infants. Desynchronization of the movements-related rhythms (e.g., mu rhythms) was found for the observation of crawling, but not for the observation of walking. Furthermore, the size of the effect was strongly related to the infant's own crawling experience. The authors concluded that experience of one's own actions is closely related to how actions of others are perceived.
Cook et al. dismiss the experiments showing that human neonates are able to copy actions done by others (Meltzoff & Moore Reference Meltzoff and Moore1977). Their argument is the following. The best-documented imitative action is tongue protrusion, but even this act “lacks the specificity … of imitation” (target article, sect. 6.1, para. 1). In addition, this behavior can also be elicited when infants observe a mechanical “tongue” or disembodied mouth. It is hard for me to conceive how the mirror mechanism of a neonate might have a neurological maturity such as to provide a precise copy of tongue protrusion. Occasionally this could happen, but the potential act encoded in a newborn must be, for maturational reasons, just “protruding.” Note also that grasping MNs generalize across the observed actions having the same goal. For example, in both monkeys and humans the observation of a grasping robot arm is effective in triggering mirror neurons (Gazzola et al. Reference Gazzola, Rizzolatti, Wicker and Keysers2007, Peeters et al. Reference Peeters, Simone, Nelissen, Fabbri-Destro, Vanduffel, Rizzolatti and Orban2009; Rochat et al. Reference Rochat, Caruana, Jezzini, Escola, Intskirveli, Grammont, Gallese, Rizzolatti and Umiltà2010), exactly as does the “mechanical tongue” in the example mentioned above. My hypothesis is that tongue protrusion in newborns is an effect mediated by a mirror mechanism similar to that described for crawling by Van Elk et al. (Reference Van Elk, van Schie, Hunnius, Vesper and Bekkering2008). Action comes first and links motor centers with sensory centers. Once these connections are established (or reinforced), the external information can flow in a forward direction, from stimuli to actions. Hence the imitation.
The characterizing feature of the primary visual cortex of primates is the presence of neurons sensitive to stimulus orientation. Regardless of whether the orientation-sensitive neurons are determined genetically or acquired by experience, or both, they are at the basis of the functional organization of primates' visual system (Hubel & Wiesel Reference Hubel and Wiesel1998; Marr Reference Marr1982). The same is true for mirror neurons (MNs). Regardless of whether their properties are determined genetically, acquired by experience, or both, they represent the neural substrate of a fundamental mechanism that transforms sensory information into a motor format (the mirror mechanism). The functions of the mirror mechanism vary from action understanding, to imitation, to empathy, and even, in birds, to song recognition (Rizzolatti & Sinigaglia Reference Rizzolatti and Sinigaglia2010). Their function depends on their anatomical location. Thus, contrary to the view advanced in the present target article, the problem of how MNs originate is utterly irrelevant as far as their function is concerned. It is an interesting problem, but it has little to do with the function of the mirror mechanism.
The claim that mirror neurons are just “another type” of association neurons misses their characterizing, unique property, which is that of giving a motor format to sensory stimuli. This misunderstanding can be also found in an interesting paper on mirror neurons by Damasio and Meyer (Reference Damasio and Meyer2008). They claimed that the parieto-frontal mirror neurons are neural ensembles included in higher-order association areas called “non-local convergence-divergence zones” that collect information from lower-order visual, auditory, and somatosensory association areas, and signal back to those areas. Action understanding depends on the activation of this network. This proposal had the merit of highlighting the role of top-down connections in action understanding. It overlooked, however, as done in the present target article, the fact that parieto-frontal mirror neurons are motor neurons. When MNs discharge, they “ignite” motor schemata similar to those endogenously activated during motor imagery and, within limits, during actual motor act execution. In other words, my motor schemata are activated during the observation of similar motor schemata of others. This provides a neurophysiological account of the mechanism underlying action understanding “from inside” (Rizzolatti & Sinigaglia Reference Rizzolatti and Sinigaglia2010): “a first person process, where the self feels like an actor, rather than a spectator” (Jeannerod Reference Jeannerod1997, p. 95; emphasis added). This appears to be a function that only the mirror mechanism is able to mediate.
Mirror neurons have top-down effects (Damasio & Meyer Reference Damasio and Meyer2008, see also Kilner et al. Reference Kilner, Friston and Frith2007b). In other words, following MN activation, signals go not only toward other motor areas, but also backwards to lower-order visual, auditory, and somatosensory areas. This top-down activation binds the understanding of what a person is doing (e.g., grasping), encoded in the motor cortex, with the visual details of that action. An interesting possibility is that the top-down mechanism also has another function, which is: to be the neural substrate of a learning mechanism that starts from motor centers rather than from the environment. An elegant experiment by Van Elk et al. (Reference Van Elk, van Schie, Hunnius, Vesper and Bekkering2008) illustrates this point well. EEG was recorded during observation of action videos in 14- to 16-month old infants. Desynchronization of the movements-related rhythms (e.g., mu rhythms) was found for the observation of crawling, but not for the observation of walking. Furthermore, the size of the effect was strongly related to the infant's own crawling experience. The authors concluded that experience of one's own actions is closely related to how actions of others are perceived.
Cook et al. dismiss the experiments showing that human neonates are able to copy actions done by others (Meltzoff & Moore Reference Meltzoff and Moore1977). Their argument is the following. The best-documented imitative action is tongue protrusion, but even this act “lacks the specificity … of imitation” (target article, sect. 6.1, para. 1). In addition, this behavior can also be elicited when infants observe a mechanical “tongue” or disembodied mouth. It is hard for me to conceive how the mirror mechanism of a neonate might have a neurological maturity such as to provide a precise copy of tongue protrusion. Occasionally this could happen, but the potential act encoded in a newborn must be, for maturational reasons, just “protruding.” Note also that grasping MNs generalize across the observed actions having the same goal. For example, in both monkeys and humans the observation of a grasping robot arm is effective in triggering mirror neurons (Gazzola et al. Reference Gazzola, Rizzolatti, Wicker and Keysers2007, Peeters et al. Reference Peeters, Simone, Nelissen, Fabbri-Destro, Vanduffel, Rizzolatti and Orban2009; Rochat et al. Reference Rochat, Caruana, Jezzini, Escola, Intskirveli, Grammont, Gallese, Rizzolatti and Umiltà2010), exactly as does the “mechanical tongue” in the example mentioned above. My hypothesis is that tongue protrusion in newborns is an effect mediated by a mirror mechanism similar to that described for crawling by Van Elk et al. (Reference Van Elk, van Schie, Hunnius, Vesper and Bekkering2008). Action comes first and links motor centers with sensory centers. Once these connections are established (or reinforced), the external information can flow in a forward direction, from stimuli to actions. Hence the imitation.
ACKNOWLEDGMENTS
This work was funded by ERC grant COGSYSTEMS to Giacomo Rizzolatti.