In order to experience something through our sense of touch, we usually have to actually ´touch´, in other words, to actively move the fingers for a certain time. During this period of time, several sequential movements generate new sensory information (Gibson, 1962; Klatzky &
Lederman, 1999). Haptic perception can, therefore, be considered as a dynamic process in which the sensory basis for perception and movements is continuously updated. In the past, several studies investigated how sensory information from multiple sources is integrated into a final percept (e.g., Ernst &
Banks, 2002). Other studies examined movement control in haptic exploration (e.g., Klatzky &
Lederman, 1987; Kaim &
Drewing, 2011). However, existing literature on natural haptic exploration did not consider dynamic developments in movements and perception over the entire process. Within my thesis, I aimed to overcome these limitations by studying the sequential nature of the haptic perceptual integration and the online adjustments of movements in natural haptic exploration.Across the first two studies, I investigated how sequentially gathered sensory information is integrated into a unified percept for two central haptic dimensions, softness and texture. First, in Study I, participants compared two textures after exploring them one after the other with varying numbers of exploration movements. The integration of the sensory information from sequential movements resulted to be more complex than predicted by an optimal integrator model which is known from the integration of simultaneous information (MLE, e.g., Ernst &
Bülthoff, 2004). Second, Study II focused on the contributions of individual sequential movements for softness judgments. The psychophysical results of this study were well in agreement with neurophysiologic literature on decision-making (e.g., Deco, Rolls, &
Remo, 2010) and - again - not consistent with a simple MLE model. In order to account for the temporal dynamics of the sequential exploration process, I developed a Kalman filter model (Kalman, 1962) as an expanded optimal integrator model. Predictions from this model resulted to be consistent with the empirical data. In sum, the model incorporates online comparisons between a memory representation of the first object and the current sensory information about the second object during each movement over the second object (see e.g., Romo &
Salinas, 2003). The memory representation of the first object, however, is additionally assumed to decay during the exploration of the second object (see e.g., Murray, Ward &
Hockley, 1975). Studies III and IV investigated whether sequentially gathered sensory information impact the control of key movement parameters for softness and texture perception. Specifically, Study III examined peak indentation forces during the process of softness exploration. The results revealed that sensory information had less impact on the executed movements than predictive information. However, the impact of sensory information was moderated by motivation, which is in line with models on optimal movement control (e.g., Todorov &
Jordan, 2002). Study IV, focused on movement directions during the process of texture exploration. Within this study, I developed a novel method that allows directly comparing the use of sensory signals in movement control to its use in perception. The results indicated that sensory signals are incorporated in movement control and that this can improve perception. In sum, movements were reported to be adjusted over the exploration process with the goal to optimize haptic perception while minimizing motor costs. Taken together, the presented thesis expands the exciting literature by demonstrating that due to the sequential gathering of sensory information perception and movements continuously evolve and mutually influence each other in a process of natural haptic exploration.
Verknüpfung zu Publikationen oder weiteren Datensätzen
