Perceptual grounding of spatial frames of reference in communication and action

Spatial cognition is a basic aspect of interactions between humans as well as between humans and robotic systems. Representations of spatial relations are likely to play an important role for joint attention, joint action, cooperation and competition. Such spatial representations in human-human or human-robot interaction are in many cases coordinated via language. According to a widely-accepted proposal, spatial frames of reference are classified with respect to the question of whether the reference object (binary localizations) or a third object as viewing point (ternary localizations) determines the orientation of the reference frame; this classification regards the type of reference object as irrelevant. In this project we want to corroborate the position that the type of viewing point used might be very relevant for the characterization of a reference frame because different viewing points may be based on different (partly modality-dependent) types of perceptual information. Previous studies have shown that, in an egocentric frame of reference, the space to which humans assign the label “in front” is more extended and more differentiated than the space assigned the labels “left”, “right”, or “back”. One aim of our project is to compare this situation to situations in which different frames of reference might be applied depending on viewing point, perception, movement and intention. We want to examine the hypothesis that speaker-related vs. object-related frames of reference differ regarding their mental representation in experiments using psycholinguistic and action-related behavioural paradigms in which mental representations of the spatial constellations used in the experiments are measured to reveal the cognitive structures underlying the application of the different frames of reference. Special attention will be paid to the influence of different modalities of perception and of (intended) movement involved in situations in which the reference object is a person, a robot, an object with an intrinsic front side (e.g. a car), or an object with no intrinsic front side (e.g. a ball), situations in which the participant has to perform an action related to the reference object (e.g. grasp), and situations in which the participant’s body posture changes relative to the reference object (e.g. sitting down, lying down, turning the head). From the data obtained in the experiments we want to derive adequate models of representation of spatial relations in different frames of reference.[view:groupmembers==196]