Physiological evidence suggests there are numerous regions within visual cortex that have a pronounced retinotopic bias toward one vertical hemifield, or code exclusively for a single vertical hemifield. The strongest evidence of this kind is for a processing bias toward the lower visual field within the dorsal pathway. Given the evidence that global motion perception in
humans is critically dependent on dorsal mechanisms, it was predicted that the lower visual field would prove to be functionally specialised for processing global motion. This hypothesis was tested in a series of five experiments. In
Experiments 1-3, observers discriminated motion direction in partially coherent random dot kinematograms, under several stimulus conditions: in isolation, with dynamic noise or a static texture distractors simultaneously present at
fixation , in bidirectional 'transparent' motion stimuli, and in multisegmented kinematograms containing contrasting directions of motion. The lower visual field was found to be more sensitive to global motion in all conditions, except
when a static dot texture was present in the visual field , when no hemifield effect was observed. In Experiment 4, observers viewed a texture 'patch' composed of numerous line elements in which they discriminated global orientation. No vertical hemifield effect was found , both with and without similar distractor stimuli at fixation, suggesting that the lower field advantage previously found was not due simply to enhanced global processing, regardless of the class of stimulus. Experiment 5 had observers covertly track a single, rapidly moving object that abruptly vanished, then spatially localise
the vanishing point. Accuracy was found to be significantly higher in the upper visual field. The results suggest that the upper and lower visual fields may be functionally specialised for processing local and global motion, respectively.
Finally, a case study is presented in which the performance of an adult albino (JK) is compared with age-matched normals. JK demonstrated a reversed asymmetry for global motion, suggesting that the quality of visual input in infancy might determine vertical hemifield effects for motion perception.