Accuracy level of pointing movements performed during slow passive whole-body rotations

Abstract.

Seated observers requested to detect low-velocity passive rotations show a high motion-detection threshold. However, when standing on a slowly rotating platform, their equilibrium is preserved, suggesting that cognitive sensing and sensorimotor reactions do not share the same central processes. The present experiments investigated the ability of observers seated on a slowly rotating chair in total darkness to indicate with their hand the position of briefly flashed targets (Experiment 1) and to indicate the subjective horizon with an outstretched arm (Experiment 2) or with a target driven by a joystick (Experiment 3). The overall hypothesis stated that egocentric coding of the position of a target should not be affected by sensing or not-sensing body rotation (Experiment 1), while geocentric positioning may (Experiments 2 and 3). Our data partially supported the hypothesis. Subjects pointed accurately to the memorized targets (Experiment1), whereas misperception of body orientation was a source of inaccuracy for actions referred to a geocentric frame (Experiments 2 and 3). More interestingly, subjects' perceptions changed as a single, smooth, and monotonic function of tilt, independent of whether the perception of body orientation was present or not.