Neither the richness of optic flow, nor the precision of heading judgements, predicts walking trajectories
When an observer translates through space, a pattern of image motion, “optic flow”, is projected onto the back of each eye. If the translation is forwards, a radial pattern of optic flow results, with the center of the pattern specifying the direction of translation. It is commonly assumed that humans use optic flow in the visual guidance of walking and that the contribution of optic flow is proportional to its richness. A further assumption is that “heading” judgements (judgements of the direction of translation) and guidance of walking rely on the same visual information and thus the processes involved in the latter can be studied using the former. These assumptions underpin a very extensive body of psychophysical, behavioral, neurophysiological, clinical, computational modelling and imaging research. We measure the form of walking trajectories (using a standard perturbation design in Experiment 1) and the precision of heading judgements (Experiment 2) in four different visual environments. We find that neither the richness of optic flow nor the precision of heading judgements predicts walking trajectories. These results challenge the widely held assumptions about optic flow, perception of heading and the visual guidance of walking.