Sensory-motor control and adaptation in cooperative action

Abstract

The objective of this thesis is to present how implicit, nonverbal cues support coordinated action between two people. A set of five experiments shows how people utilise sensory information about a partner in order to control goal-directed action involving synchronous vertical hand movement. Chapter 1 reviews recent developments in studies of interpersonal interaction and highlights control issues that are critical in cooperative action. In Chapter 2, an experimental study addresses the importance of haptic feedback about object dynamics for learning behavioural characteristics of another person who manipulated the same object from the other side. In Chapter 3, the contributions of feedback and feedforward control are assessed as the interaction of the task partners are experimentally controlled using a humanoid robot which serves the role of task partner. Chapter 4 proposes error-based learning as a model of cooperative action wherein subsequent motor response is regressed on current error in order to improve coordination between partners. Chapter 5 shows that adaptation rate of a participant is modulated with respect to the rate of the task partner so that the net adaptation of the two partners becomes optimal using a computer simulated task partner. In Chapter 6, this joint adaptation model is applied to continuous movement to demonstrate the generalisability of the model. The last chapter discusses the contribution of the empirical chapters and reviews the theoretical and methodological contribution of the thesis as a whole to the field of cognitive neuroscience. In conclusion, the thesis provides strong evidence that movement characteristics of a partner expressed both within and across trials determine the way a person engages in a joint task.