Representing the properties of object classes in manipulation and weight perception
MetadataShow full item record
The ability to accurately predict object weight is essential for skilled manipulation and recent studies suggest that such prediction is based, in part, on learned size-weight maps associated with families of objects. Weight prediction based on size-weight maps is also involved when judging weights; there is strong evidence that weight judgments are biased by expected weight, based on size. This bias is revealed by the size-weight illusion (SWI) whereby the smaller of two equally weighted and otherwise similar objects is judged to be heavier because it is heavier than expected based on its size. The overall aim of the current set of studies was to examine how size-weight maps for different families of objects are organized and represented at the perceptual and sensorimotor levels. We found that distinct and independent size-weight maps, used to predict weight, were used when lifting objects and judging their weights. At the perceptual level, interference between size-weight maps for the different sets of cubes was observed; participants could learn the inverted size-weight relationship for the green cubes when experienced alone but not when experienced along with the black cubes with a normal size-weight relationship. However, about half of participants learned to scale lift forces accurately for both sets of cubes indicating that the sensorimotor system can learn, without interference, opposite size-weight maps. We further investigated why not all participants learned to accurately scale their lift forces and found that learning to lift objects with different and arbitrary size weight maps involves visuomotor working memory resources. Moreover, an outside task that steals attentional resources can interfere even after previous learning of the size-weight maps.