We measured human ocular torsion (OT) monocularly (using video) and bi
nocularly (using search coils) while sinusoidally accelerating (0.7 g)
five human subjects along an earth-horizontal axis at five frequencie
s (0.35, 0.4, 0.5, 0.75, and 1.0 Hz). The compensatory nature of OT wa
s investigated by changing the relative orientation of the dynamic (li
near acceleration) and static (gravitational) cues. Four subject orien
tations were investigated: (1) Y-upright - acceleration along the inte
r aural (y) axis while upright; (2) Y-supine - acceleration along the
y-axis while supine; (3) Z-RED - acceleration along the dorsoventral (
z) axis with right ear down; (4) Z-supine - acceleration along the z-a
xis while supine. Linear acceleration in the Y-upright, Y-supine and Z
-RED orientations elicited conjugate OT. The smaller response in the Z
-supine orientation appeared disconjugate. The amplitude of the respon
se decreased and the phase lag increased with increasing frequency for
each orientation. This frequency dependence does not match the freque
ncy response of the regular or irregular afferent otolith neurons; the
refore the response dynamics cannot be explained by simple peripheral
mechanisms. The Y-upright responses were larger than the Y-supine resp
onses (P<0.05). This difference indicates that OT must be more complic
ated than a simple low-pass filtered response to interaural shear forc
e, since the dynamic shear force along the interaural axis was identic
al in these two orientations. The Y-supine responses were, in turn, la
rger than the Z-RED responses (P<0.01). Interestingly, the vector sum
of the Y-supine responses plus Z-RED responses was not significantly d
ifferent (P=0.99) from the Y-upright responses. This suggests that, in
this frequency range, the conjugate OT response during Y-upright stim
ulation might be composed of two components: (1) a response to shear f
orce along the y-axis (as in Y-supine stimulation), and (2) a response
to roll tilt of gravitoinertial force (as in Z-RED stimulation).