Commentary on Michael A. Arbib

Word counts
Abstract 55
Main text 582
References 43
Entire text (total) 682

Linguistic skills and motor skills

Masao Ito
RIKEN Brain Science Institute, Wako, Saitama 351-0198, Japan
+81-48467-6984
masao@brain.riken.jp
http://common.brain.riken.jp/en/research/inter/lb_main_e.aspx

Abstract: Dr. Arbib's hypothesis sets future targets for neuroscience to clarify neural mechanisms of language. My specific comment is on the relationship of the mirror neuron mechanism for linguistic skills and the cerebellar internal model mechanism for motor skills. These are distinct from each other, but both are important in learning skills in movement and language.

Dr. Arbib’ hypothesis sets targets for neuroscience which aims at clarifying neural mechanisms of language. Neural mechanisms of language might be formulated by accumulation of transitional changes in brain's structure and function occurring at each evolutionary stage. A transitional change of brain function at each stage of evolution must be underlain by a change in a genetic code responsible for formation of a neuronal circuit of the brain. An evolutionary expansion of the brain may accompany twofold changes. First, addition of a new brain area elaborates the control system structure of neuronal circuits, and consequently its operation. Second, introduction of novel cellular (such as a new type of synaptic plasticity) or molecular processes (such as regulation by a neuropeptide) may modify the architecture, and consequently its operation, of a local neuronal circuit. These changes are obvious future targets of neuroscientific research.

A basic assumption in Dr. Arbib’s hypothesis is that linguistic skills originated from the mirror system for grasping motor actions. On the other hand, I have been working with the internal forward model hypothesis in explaining neuronal mechanisms for motor skills (Ito, 2001, 2004). How are these two set of mechanisms related?

For the following discussions, let’s adopt a classic feedback control system scheme to illustrate the voluntary control mechanism of hand movement (Fig. 1). Mirror neurons in F5 premotor area should form a pre-stage to the controller that involves motor cortex. The motor cotrex acts on a motor apparatus such as a hand. Mirror neurons should then receive instructions of an intended movement generated within the own brain or sensory feedback from the performed movement. The pathway through which mirror neurons receive information of an observed performance of another body should be a unique addition to the classic control system scheme for movement. This additional pathway is for receiving instruction for imitating another body’s performance.

When the motor cortex moves a hand repeatedly responding to instruction of a plan and referring to sensory feedback of the actual hand movement, an internal model simulating the dynamics of the hand will be formed in the cerebellum by a learning mechanism based on synaptic plasticity, long-term depression (Ito, 2001). The motor cortex then becomes able to control the hand by referring to internal feedback through the internal model of the hand, even if there is no sensory feedback from he hand. This explains why we can perform an exercised movement accurately and smoothly even at a fast speed or without visual feedback. The cerebellar learning to form an internal model will be conducted similarly in movements induced either by the instruction of a plan arising from the own brain or by the instruction to mimic another body’s performance (Fig. 1). The internal model mechanism of the cerebellum is therefore separate from the mirror neuron mechanism, and both are important mechanisms operating in parallel for motor learning.

The scheme of Fig. 1 for motor learning in general may apply to the linguistic skill learning if the motor apparatus is replaced by components of language represented in Wernicke’s area, and F5 and the motor cortex are replaced by Broca’s area (Fig. 2). Two sensory systems in Fig. 1 may merge to form one system corresponding to the inferior parietal area PF that contains neurons behaving like mirror neurons. If Fig. 1 is the case, such PF mirror neurons are driven by both feedback from own actual movement and observation of other body’s performance, but not by the instruction of a plan generated within own brain.

References

Ito M. Cerebellar long-term depression –characterization, signal transduction and functional roles, Physiol. Rev. 81: 1143-1195, 2001

Ito, M. Bases and implications of learning in the cerebellum – adaptive control and internal model mechanism. In: Creating coordination in cerebellum. Progr. Brain Res. (in Press) 2004.