Reciprocal error behavior in estimated body position and subjective visual vertical

This study investigated the reciprocal relation between estimation of body tilt and visual vertical by using self-controlled passive body tilts at constant velocity (slow tilts with no semicircular canal activation) or constant acceleration (rapid tilts with canal activation). In both conditions, the visual vertical was overestimated in the luminous line setting paradigm, whereas body tilt was underestimated in the position estimation paradigm. These errors were larger after slow than rapid tilts. During slow tilts, the range of actually reached positions was on average underestimated by about 25% with respect to the desired positions. Interestingly, there were no significant differences in the estimated positions for tilts in the roll and pitch plane. Most remarkably, in the range of +/-45 degrees the resulting means of position and luminous line setting errors of the velocity and acceleration paradigms as a function of the desired roll positions were close to zero. Furthermore, the resulting means of the two paradigms showed a high correlation in the tested range of +/-90 degrees. We conclude that: (a). the otoliths provide the main information for the spatial reference for both the estimation of body positions and the luminous line settings, at least in the range of about +/-45 degrees where the resulting mean errors between the two paradigms are close to zero, and (b). coactivation of semicircular canals improves the estimations.