• Research Summary (2000 ~ 2010)

    Jun Tani
    Lab. for Behavior and Dynamic Cognition, RIKEN BSI

    The main objective of our research has been to understand the mechanisms of higher-order cognitive brain functions with particular focus on the problems of compositionality. Compositionality means that the whole can be derived by combining reusable parts; for example, complex goal-directed actions can be generated by combining reusable behavior primitives into specific sequences according to Arbib’s motor schemata theory. Although it is considered in conventional cognitive science that compositionality can be achieved by means of symbol representation and manipulation of the representations at a higher cognitive level, our motivation has been to investigate an alternative process whereby compositionality could be developed in the neural dynamics of distributed activity through accumulative learning of sensory-motor experiences.

    Read More (pdf)