SCALING-UP SENSORIMOTOR SYSTEMS - CONSTRAINTS FROM HUMAN INFANCY

Work in human infancy and behavior-based robotics that grounds intelli gent abilities in sensorimotor exchanges between a system and its envi ronment shares recurrent problems of when, whether, and how scaling up from basic to supposedly higher abilities is possible. An action-base d model of the infant is introduced that converges with features of in dependently motivated animat models exploiting emergent functionality and challenges alternatives that invoke conceptual representations. Ad aptive change routinely exhibited in infants' everyday activities outs trips the scaling-up potential of current robotic systems and clarifie s effective principles obeyed by naturally intelligent systems. A gene ral form is outlined to subject-environment interaction that ''enginee rs'' restructuring of early abilities in the direction of greater anti cipation (considered an upper boundary for the competence of concept-f ree human and animat systems); and an action-based account of the phen omena is provided. This emphasizes the relationship between representa tion and situated inference and the role of reciprocal constraints bet ween cognitive and physical-motor mechanisms. Finally, this article qu estions how far typical self-organizing connectionist networks take us toward understanding a system that is capable of mapping recurrent vi able patterns of activity into more permanent adaptive changes.