Motor Learning and Motor Control Mechanisms in an Haptic Dyad

The word “dyad” defines the interaction between two human or cybernetic organisms. During such interaction, there is an organized flow of information between the two elements of the dyad, in a fully bidirectional manner. With this mutual knowledge they are able to understand the actual state of the dyad as well as the previous states and, in some cases, to predict a response for possible scenarios. In the studies presented in this thesis we aim to understand the kind of information exchanged during dyadic interaction and the way this information is communicated from one individual to another not only in a purely dyadic context but also in a more general social sense, namely dissemination of knowledge via physical and non-physical interpersonal interactions. More specifically, the focus of the experimental activities will be on motor learning and motor control mechanisms, in the general context of embodied motor cognition. Solving a task promotes the creation of an internal representation of the dynamical characteristics of the working environment. An understanding of the environmental characteristics allows the subjects to become proficient in such task. We also intended to evaluate the application of such a model when it is created and applied under different conditions and using different body parts. For example, we investigated how human subjects can generalize the acquired model of a certain task, carried out by means of the wrist, in the sense of mapping the skill from the distal degrees of freedom of the wrist to the proximal degrees of freedom of the arm (elbow & shoulder), under the same dynamical conditions. In the same line of reasoning, namely that individuals solving a certain task need to develop an internal model of the environment, we investigated in which manner different skill levels of the two partners of a dyad interfere with the overall learning/training process. It is known indeed that internal models are essential for allowing dyadic member to apply different motor control strategies for completing the task. Previous studies have shown that the internal model created in a solo performance can be shared and exploited in a dyadic collaboration to solve the same task. In our study we went a step forward by demonstrating that learning an unstable task in a dyad propitiates the creation of a shared internal model of the task, which includes the representation of the mutual forces applied by the partners. Thus when the partners in the dyad have different knowledge levels of the task, the representation created by the less proficient partner can be mistaken since it may include the proficient partner as part of the dynamical conditions of the task instead of as the assistance helping him to complete the experiments. For this reason we implemented a dyadic learning protocol that allows the naïve subject to explore and create an accurate internal model, while exploiting, at the same time, the advantage of working with an skilled partner.