PubMed1999
A review of recent experiments in patients with labyrinthine and neurological disorders assessing the subjective postural vertical (SPV) and the subjective visual vertical (SVV) is presented. The SPV was measured with subjects (Ss) seated in a motorized flight simulator tilting at 1.5 deg/s in roll and pitch; the Ss' task was to indicate when they entered and left self-verticality. The SVV was measured by Ss adjusting a straight line to what they perceived as gravitational upright. Clear dissociations between the SVV and SPV were found, for example, patients with acute unilateral vestibular disorders had marked tilts of the SVV toward the side of the lesion but a "lean" (bias, tilt) of the SPV was never found. Dissociations of the SPV and SVV could also be induced in normal subjects by roll-plane visual motion stimuli: the SVV was tilted in the direction of motion, but the SPV was not. Prolonged lateral body tilt did, however, bias the SVV (the "A" effect) and the SPV, but these effects are likely to be mediated by somatosensory rather than otolithic input. Evidence for the latter came from (i) findings in patients with absent vestibular function, who showed an enhanced "A" effect, and (ii) from a patient with a thalamic infarction, who showed absence of the "A" effect when leaning on the hemihypesthetic side. In separate experiments where normal Ss indicated space-vertical and space-horizontal with saccadic eye movements, we found differences between these percepts, that is, subjective external space lost orthogonality. The findings in these various experiments can be interpreted if we abandon the idea of a single, "internal representation" of verticality. Different sensory modalities convey different and sometimes conflicting messages about verticality. Otolithic and somatosensory signals can have opposite sign effects during verticality estimates while tilted. In man, somatosensory cues have a prominent role in verticality perception.