Cue combination or co-specificity? Effects of exploration on attunement in dynamic touch

The co-specificity hypothesis states that intention to perceive some property, attunement to information, and exploration are co-implicated in the perception of that property (Riley, Wagman, Santana, Carello, & Turvey, 2002). The research reported here attempts to expand and fit this hypothesis into the theory of direct learning (Jacobs & Michaels, 2007), while challenging theories that classify information as multiple cues that are weighted, combined, or assigned probability distributions. In a dynamic touch task, participants made judgments and were given feedback about the height or diameter of unseen, wielded pipes. Participants were assigned to either a free-exploration condition, or a restricted exploration condition in which a gimbal limited explorations to the axes that yielded the greatest (Ixx) and least (Iyy) resistance to rotational acceleration (in the height condition); or, only the axis that yielded the least resistance to rotational acceleration (in the diameter condition). In the latter condition, participants were given the optimal information variable for width, whereas in the former case, participants were forced to combine the information variables that serve as the functional basis for height perception. ^ A six-block design (pretest, 3 feedback blocks, post-test, exploration-test) was used, with each of 20 objects presented once per block. Feedback was provided only during the feedback blocks. In the exploration-test block, participants were tested in the other condition; restricted participants were tested in the free exploration, and vice-versa. Although the optimal information variable for either height can be computed as a weighted function of I1 and I3, participants in the restricted-exploration condition did not report height as accurately as did participants in the free exploration condition. In fact, the judgments of restricted explorers consistently correlated most highly with which supports the hypothesis that the perceptual system does not combine or weight information cues piecemeal. Exploratory movements in all conditions were tracked. In the free-exploration condition there was evidence that changes in information use were accompanied by changes in exploration, which supported the co-specificity hypothesis.