Contribution of ankle, knee, and hip joints to the perception threshold for support surface rotation.

The purpose of the present experiment was to investigate the extent to which subjects can perceive, at very slow velocities, an angular rotation of the support surface about the medio-lateral axis of the ankle, knee, hip, or neck joint when visual cues are not available. Subjects were passively displaced on a slowly rotating platform at .01, .03, and .05 deg/sec. The subjects' task was to detect movements of the platform in four different postural conditions allowing body oscillations about the ankle, knee, hip, or neck joint. In Experiment 1, subjects had to detect backward and forward rotation (pitching). In Experiment 2, they had to detect left and right rotations of the platform (rolling). In Experiment 3, subjects had to detect both backward/forward and left/right rotations of the platform, with the body fixed and the head either fixed or free to move. Overall, when the body was free to oscillate about the ankle, knee, or hip joints, a similar threshold for movement perception was observed. This threshold was lower for rolling than for pitching. Interestingly, in these postural conditions, an unconscious compensation in the direction opposite to the platform rotation was observed on most trials. The threshold for movement perception was much higher when the head was the only segment free to oscillate about the neck joint. These results suggest that, in static conditions, the otoliths are poor detectors of the direction of gravity forces. They also suggest that accurate perception of body orientation is improved when proprioceptive information can be dynamically integrated.