Object encoding but not action understanding in the the macaque medial reach-to-grasp network

The vision of an action evokes intriguing discharges in a rich network of cortical areas. Some cortical neurons, called ‘mirror’ neurons, discharge both during action execution and action observation (Rizzolatti et al. 1996, Cogn Brain Res). In the present study, we looked for mirror neurons in macaque medial parietal area V6A, an area of the reach-to-grasp network (Galletti and Fattori, 2018 Cortex) never explored to date in this regard. We recorded neural activity of 100 V6A neurons of two male Macaca fascicularis during grasping action and during observation of the same action performed by the experimenter. The overhealming majority of neurons (86/100) were modulated only when monkey executed the action. A minority (14%) showed mirror features, discharging also during observation of actions performed by experimenter. However, differently from classic mirror neurons, V6A mirror neurons responded also to passive object observation and showed dissimilar responses when monkey performed the action and when it observed the same action performed by experimenter. In other words, these neurons are able to encode the visual features of objects, but do not seem to be involved in action understanding, being the neural representations during execution and observation highly dissimilar, so excluding the possibility of creating an internal representation of other agent’s actions. We have also found that in these neurons, the neural representation of the object changed according to whether grasping was allowed/performed or not, and whether object was or not target of another agent’s grasp. In other words, rather than coding the agent’s observed action, V6A neurons appear to encode the pragmatic relevance of the object. In summary, area V6A is well equipped to monitor own actions, but it is not able to build an internal representation of observed actions, while encoding the pragmatic relevance of object in the action to be performed.