Abstract
Space perception depends on both external depth information and the visual system' intrinsic bias. The latter plays a larger role in reduced-cue environments. In total darkness, it constructs our perceptual space such that a dimly-lit target is perceived at the intersection between the projection line from the eye to the target and the implicit slant surface delineating the intrinsic bias. Given the reliance on the intrinsic bias, it would be strategic for the intrinsic bias to be independent of the observer' egocenter (body). This idea will be validated if the intrinsic bias is updated when the observer walks to a different location in the dark. Thus our main experiment tested observers at three viewing-locations: (i) original location; (ii) forward location (observer walked forward from the original location by 1.25m); (iii) backward location (walked 1.25m backward). After arriving at each viewing-location, observers stood for ∼17s before a 5Hz flickering target (0.22 deg) was displayed for 1s at one of six predetermined target-locations (2–7m). Observers reported the perceived target location (distance and height) using the blind walking-gesturing task. We found that judged target locations from all three viewing-locations are fitted by a single, rather than three separate, slant curve (surface). This indicates that when one moves to a different viewing-location, the intrinsic bias is not tagged with the body but remains at the original location prior to moving, i.e., the intrinsic bias is updated. Our further experiments revealed the characteristics of the updating operation. We found that space updating: (1) though less perfect, occurs even when observers moved a longer distance (2.5m) from the original location; (2) is restricted to a viewing-location with the same facing-direction as the original location (does not occur when the facing-direction is orthogonal); (3) is also revealed using the verbal report task.