Self-rotation estimate about the vertical axis.

The aim of this paper was to examine the accuracy of passive whole-body self-rotation estimate around the earth-vertical (yaw) axis. Subjects were required to reach an imposed angle of +/- 90 degrees, 180 degrees or 360 degrees (outward way), and then to rotate back to the initial position (return way), through passive rotation controlled by the subject, in darkness. During these rotations, only the semi-circular canals were stimulated. On the outward way, subjects were required to elaborate the correct imposed angle, without any external reference. In average, subjects undershot the expected angles on outward way, which suggests an overestimate of self-rotation, and the error increased with increasing expected angle. On the return way, subjects had to rotate back to the starting point, through a path integration process. There was no effect of the imposed angle on the error, and the variability was lower than on the outward way. The data suggests that the subjects could construct the internal representation of a virtual starting point: the initial body position, which became a goal to reach clearer than the outward rotation angle, which the subjects also have defined, in an environment deprived of any external spatial reference.