Optic flow induces nonvisual self-motion aftereffects

There is strong evidence of shared neurophysiological substrates for visual and vestibular processing that likely support our capacity for estimating our own movement through the environment [1, 2]. We examined behavioral consequences of these shared substrates in the form of crossmodal aftereffects. In particular, we examined whether sustained exposure to a visual self-motion stimulus (i.e., optic flow) induces a subsequent bias in nonvisual (i.e., vestibular) self-motion perception in the opposite direction in darkness. Although several previous studies have investigated self-motion aftereffects [3-6], none have demonstrated crossmodal transfer, which is the strongest proof that the adapted mechanisms are generalized for self-motion processing. The crossmodal aftereffect was quantified using a motion-nulling procedure in which observers were physically translated on a motion platform to find the movement required to cancel the visually induced aftereffect. Crossmodal transfer was elicited only with the longest-duration visual adaptor (15 s), suggesting that transfer requires sustained vection (i.e., visually induced self-motion perception). Visual-only aftereffects were also measured, but the magnitudes of visual-only and crossmodal aftereffects were not correlated, indicating distinct underlying mechanisms. We propose that crossmodal aftereffects can be understood as an example of contingent [7] or contextual adaptation [8, 9] that arises in response to correlations across signals and functions to reduce these correlations in order to increase coding efficiency. According to this view, crossmodal aftereffects in general (e.g., visual-auditory [10] or visual-tactile [11]) can be explained as accidental manifestations of mechanisms that constantly function to calibrate sensory modalities with each other as well as with the environment.