Abstract
The perceived elevation of a visual target at an observer's true eye level changes systematically with the pitch of the visual field: it appears to be raised in a pitched topbackward visual field and lowered in a pitched topforward visual field. However, the accuracy of manual pointing or heightmatching to a perceptually mislocalized visual target increases linearly with hand-to-body distance, approaching complete accuracy at full arm extension, with errors equal to the perceptual mislocalization for the hand close to the body. In the present experiments, we investigated bimanual interaction with the two hands held at different hand-to-body distances throughout a trial on manual tasks to a common visual target under pitch induction. Observers in darkness monocularly viewed a visual target at 70 cm distance in their median plane either 12° above, 12° below, or at eye level with a 50°-long line pitched either −30° (topbackward), or 20° (topforward) at 25° horizontal eccentricity. The manual heightmatching to the target was measured by a Polhemus 3-Space search coil with the unseen hand either in the midfrontal plane or at 20 cm or 40 cm in front of the midfrontal plane. The manual heightmatch to the visual target with either hand alone changed linearly with manual distance for each inducer pitch; manual heightsetting-vs-pitch functions were parallel for different target heights. With the two hands at different distances simultaneously, the height of the manual match to the target by the second hand to make a setting was influenced by the setting of the first hand. The average bimanual transfer of a setting from the baseline independence level approximated 44%, whether the first match was with the left or right hand. Hand dominance played an important role: 68% transfer was measured for the dominant to nondominant hand, 26% transfer was measured for the nondominant hand to the dominant hand.
Supported by NSF grant BCS-06-16654.