Evidence for a Vestibular Contribution to Object Motion Prediction

Humans can predict an objects motion better if its movements are consistent with gravity. Here we investigate whether this may be due to an internalized strong Earth gravity prior or to vestibular cues reporting instantaneous information about gravity. These two directions can be separated using virtual reality by providing strong visual cues to the direction of up which may or may not be aligned with true gravity. Participants were presented with a ball travelling on a parabola path simulated with either downward acceleration created by simulated Earths gravity (1g) or inverted gravity (-1g) resulting in the ball curving upwards. In both types of trial, the ball disappeared at between 57.5% and 75% of its full trajectory - after it had started its descent in the case of 1g or ascent in the case of -1g. Participants pressed a mouse button when they judged the ball to have got back to the height at which it was launched. Participants were either standing or supine. There were no differences in the estimated time to reach the indicated level between the 1g and -1g simulations, however, we found an overestimation of the perceived time for the ball to reach target height when observers were lying supine compared to when they were standing upright independent of the gravity condition simulated. A control experiment confirmed that this was not due to a general slowing of reaction times while lying supine versus while upright. To explore whether these observations might reflect posture-related changes in vestibular activity, participants completed the task under both simulated gravity conditions while seated upright in the presence of disruptive galvanic vestibular stimulation (dGVS) or during sham stimulation. As when lying supine, the perceived time for the ball to reach the target height was significantly longer in the presence of dGVS compared to during sham stimulation. Overall, participants were no better at anticipating 1g motion compared to -1g motion, but we provide compelling evidence that the state of vestibular signalling can impact the prediction of object motion.