A series of (un)predictable events: Influences of somatosensory predictions and action control on goal-directed movements

Abstract

Every day, we move through our environment and interact with it. After switching off the alarm clock and drinking our first coffee, our day is filled with a wide variety of activities. For all these movements, the brain must select appropriate actions, formulate motor plans and process feedback from the body’s sensory systems. This entire process of executing a goal-directed movement is influenced by a multitude of factors. The aim of this dissertation is to investigate the influence sensorimotor predictions on perception and action in the context of goal-directed movements. A specific emphasis will be placed on sensorimotor predictions based on direct feedback signals, sensorimotor memories, and action control processes. The research is based on three studies and aims to elucidate the intricate interplay of sensory feedback, task relevance, and motor planning in executing precise and adaptable movements in varied contexts. The first study explores how task-relevant feedback signals and their intensity modulate predictions based on efference copies during reaching. As a target is approached, there is a greater probability of physical contact with a surface, which in turn makes the tactile signals at the end of a movement more predictable. Feedforward prediction mechanisms can explain a resulting reduction of tactile sensitivity. However, previous research yielded contradictory findings, prompting the question of whether the relevance of feedback signals exerts a greater influence on the modulation of tactile perception than their mere predictability. My findings demonstrate that tactile suppression dynamically adjusts to the necessity of processing anticipated feedback, especially when the feedback signal is weak yet crucial for task completion. This highlights the brain's ability to selectively enhance sensory processing based on task demands. The first study focusses on predictions based on current information, whereas my second study examines the influence of predictions derived from past sensorimotor memories on tactile perception and movement kinematics during the interaction with objects that possess uncertain properties. In the presence of uncertainty regarding object properties, humans tend to plan their movements based on their most recent experiences with that object. For example, when placing their fingers and grasping an object to prevent tilting, they usually repeat their grasping configuration in anticipation of constant object properties. In this instance, it is probable that one must compensate for errors in movement, thereby rendering incoming somatosensory signals crucial for the process of motor adjustments. My results indicate that while predictions based on sensorimotor memories significantly impact movement kinematics on a trial-by-trial basis, they do not similarly affect tactile sensitivity. This suggests that tactile perception necessitates a more stable informational foundation for sensorimotor predictions to exert an effective influence. Moreover, explicit knowledge about the change in object properties had no effect on either kinematic behavior or tactile sensitivity. The third study aims to gain a deeper understanding of the extent to which predictions based on past somatosensory memories influence goal-directed movements. In this study, the focus is on a cognitive approach, specifically action control and predictions made on the basis of event files. A recent approach, the Binding and Retrieval in Action Control (BRAC) framework, aims to provide an overarching theory to explain results from different action control paradigms. This is achieved by using two core principles, feature binding and retrieval. My investigation revealed that previous experiences influence sensorimotor predictions, which in turn shape current motor planning. These results support the BRAC framework and highlight the importance of past interactions in shaping future actions, even in the context of complex motor tasks. It appears that goal-directed movements may serve as an effective means of integrating features of varying task-relevance and temporally varying accessibility. Overall, this dissertation contributes to a deeper understanding of the mechanisms underlying goal-directed movements, emphasizing the importance of sensorimotor predictions based on current and past information and their influence on motor planning and execution.