Cognitive load of navigating without vision when guided by virtual sound versus spatial language

A vibrotactile N-back task was used to generate cognitive load while participants were guided along virtual paths without vision. As participants stepped in place, they moved along a virtual path of linear segments. Information was provided en route about the direction of the next turning point, by spatial language ("left," "right," or "straight") or virtual sound (i.e., the perceived azimuth of the sound indicated the target direction). The authors hypothesized that virtual sound, being processed at direct perceptual levels, would have lower load than even simple language commands, which require cognitive mediation. As predicted, whereas the guidance modes did not differ significantly in the no-load condition, participants showed shorter distance traveled and less time to complete a path when performing the N-back task while navigating with virtual sound as guidance. Virtual sound also produced better N-back performance than spatial language. By indicating the superiority of virtual sound for guidance when cognitive load is present, as is characteristic of everyday navigation, these results have implications for guidance systems for the visually impaired and others.     

Knowledge About Hand Shaping and Knowledge About Objects

Our two experiments investigated associations between cognitive representations of objects and hand-shape categories. Hand configurations were partitioned according to prehensility and the size of the contacting surface, resulting in the classes: pinch, poke, palm, and clench. Experiment 1 elicited object names in response to configuration-name cues, provided ratings of the relevance of each configuration to a set of objects, and probed for the functions determining such relevance. Cueing with a configuration class elicited an associated object category with substantial intersubject agreement, and vice versa. Both the object categories and the functions associated with the four hand-configuration classes differed substantially, although the same object could be associated to some extent with multiple configurations, given variations in function. Experiment 2 elicited the names of hand-configuration classes in response to unfamiliar forms, which varied systematically in depth and the size of the projecting picture-plane surface. The modal response, response time, and degree of intersubject agreement were directly related to these variables. These structural variables, however, did not adequately predict shaping responses to real objects, as ascertained from Experiment 1. The results have implications for cognitive representation of motor categories and hand shaping in response to objects.