64
G. Riva, M.T. Anguera, B.K. Wiederhold and F. Mantovani (Eds.)
From Communication to Presence: Cognition, Emotions and Culture towards the
Ultimate Communicative Experience. Festschrift in honor of Luigi Anolli
IOS Press, Amsterdam, 2006, (c) All rights reserved - http://www.emergingcommunication.com
covert imitation may involve two further mechanisms: selective attention and
memory integration.
First, the capture process does not operate on entire perceptual states but only on
components of them. On viewing someone catching an apple, for example, the brain
does not capture the entire scene. Instead, as attention focuses on the specific action,
such the moving hand, an associative area captures the neural state that represents it.
Second, once attention selects a component of experience, a memory of the
component becomes integrated with memories of similar components, via content
addressable memory. When focusing attention on the moving hand, for example, the
active neural state in the visual system becomes integrated with similar visual
patterns captured previously.
In this occasion, a subset of mirror neurons produces one particular simulation in
the visual system. The content of the simulation depends on which subsets of stored
information become active. Possible outcomes are a given instance, an average of
several instances, or a variety of other possibilities. Moreover, re-enactments
typically occur on multiple modalities simultaneously, producing a multimodal
simulation of the action including not only sensory states but also motor and mental
states. In this view a simulator is a “distributed collection of modality-specific
memories captured across a category’s instances” [58].
Neuro-physiological evidences support the Covert Imitation Theory. On one side,
the mirror neurons within the F5c-PF functional cluster of premotor neurons already
discharge in early phases of the movement [84]. Moreover, the pre-motor cortex, the
posterior parietal cortex, and the cerebellum are activated during action generation,
action imagination, and action observation [66, 85].
There is a main criticism to this view coming from Gergely and Csibra [86, 87].
Gergely and colleagues showed that a novel response - illuminating a box by
touching it with the head - imitatively learned from the demonstration of a human
model is retained by infants in spite of the availability and production of more readily
accessible and rational response alternatives - the use of the hands - that also produce
the same effect [88]. This suggests that imitative learning of novel actions is a
qualitatively different process in humans than the imitative copying of new and
reinforcing behavior of observed conspecifics that has been demonstrated in several
other animal species. Specifically, it suggests the existence of some specific
processes selecting what to imitate.
Another issue is raised by Lyons and colleagues [89]. They note that macaque
monkeys, who have mirror neurons, simply do not imitate. So, what are mirror
neurons for in monkeys? In their opinion, their mirror neuron system is tuned to
extract the goal structure of observed action, as opposed to the lower-level kinematic
features of the action. In other words, “mirror neurons enable non-human primates to
infer the intentions of other agents” (p 231).
The difference between humans and primates appears to be related to the level of
intentional granularity: the human mirror system is capable of extracting not only
high-level goals (do x) but also more subtle, subsidiary goals (do x in manner y). This
could account for human ability to reproduce not only the overall results of observed
actions but also the specific means that were used to achieve them.