Abstract
Some believe that palm boards are more accurate measures of perceived slope than are verbal reports. But the apparent accuracy of palm boards is accidental. Haptic perception of surface orientation by hand and by foot is distorted, similarly to visual distortions of slope at a distance. This has nothing to do with the dorsal system: For a small surface sloped 30 deg, viewed within reach, we have found that palm-board matching results in settings of 18 deg, whereas matching with an unseen free hand is quite accurate. Evidently proprioceptive information is altered by contact with a supporting surface. In contrast to purely visual accounts of slope misperception (frontal tendency) and in contrast to accounts appealing to behavioral potential our investigations of slope suggest a more general principle that surfaces (large and small) appear more vertical as perceptual information about them becomes less reliable or less well-calibrated. In studies of frontal tendency in visual slope perception, for example, we have controlled for viewing distance and found that surfaces viewed more frontally are indeed subject to less distortion than surfaces viewed obliquely, but slope overestimation still occurs at distances of about 1.7 m for small surfaces viewed with gaze fully frontal. Haptic perception of the slopes of ramps in sighted and blind individuals is extremely exaggerated. Ramps feel steeper than they look. The only case of haptic slope perception that we find to be fairly accurate (via verbal judgments) is for surfaces explored by a single finger, so that computations of slope could be made from calibrated position information. For this task, slope perception was accurate for sighted blindfolded participants, but congenitally blind observers showed a vertical tendency, probably because they lack near space calibration from vision: Blind observers were more accurate than sighted at retracing the slope without feedback.