Separate visual representations for perception and action revealed by saccadic eye movements

Some 30 years ago, Trevarthen [1] introduced the idea of two separate visua l systems, a focal system for fine motor acts and an ambient system for gro ss body movements such as ambulation. More recent developments indicating a natomically and physiologically separate pathways in primate vision [2] hav e led to a different idea of separate visual systems, one for conscious per ception and one for action [3]. It has received empirical support from seve ral studies showing that pointing, reaching, and grasping can remain accura te while the perceived position or size of objects is subject to illusory d istortion [4-6], However, much of this evidence has been challenged on the grounds of methodological flaws, particularly failure to match perfectly th e conditions for verbal and motor tasks and failure to replicate results [7 -10]. Here we take advantage of the strong compression of perceived positio n that occurs around the time of saccadic eye movements [11, 12], Under nor mal lighting conditions, stimuli flashed briefly over a wide range of spati al positions just before saccadic onset are neither seen nor reached for in their veridical positions, but are compressed toward the saccadic target. We validate the idea of separate systems by showing that, in the dark, subj ects are able to point accurately to the correct target position, even thou gh their verbal reports are still subject to compression.