Blindsight in the lab: eye on awareness.

Damage to certain parts of the brain's visual cortex produces a puzzling disturbance of consciousness known as blindsight. In such cases, people deny having seen objects, moving lights, or other visible cues but can nevertheless guess correctly the location, orientation, and form of these "unseen" stimuli.

Blindsight can be induced in volunteers with healthy brains, scientists now report in the Sept. 28 Nature. Further research with this new technique may yield insights into the anatomy of conscious visual experience, assert F. Christopher Kolb and Jochen Braun, both neuroscientists at the California Institute of Technology in Pasadena.

Kolb and Braun's procedure will also make possible "a much finer-grained analysis" of blindsight, writes Alan Cowey, a psychologist at the University of Oxford in England, in an accompanying comment. For instance, investigators can examine whether sights that do not pierce consciousness nonetheless trigger subtle eye movements that later guide guesses about what was seen.

The researchers asked three participants to look at a series of moving dot displays on a computer screen. After viewing each display for one-quarter of a second, volunteers tried to identify which quadrant of the screen contained a patch of "target" dots moving in a different direction from the rest.

One display contained pairs of dots that moved away from each other along a diagonal. A few pairs moved apart along the opposite diagonal, within a target area that changed quadrants from one presentation to another. These displays created the impression of a single field of flickering dots that obscured the target patch. Volunteers reported simply guessing at target locations.

In a second display, diagonal lines of single dots moved in alternating, opposite directions. Observers easily picked out the location of a few extra dots moving in a different direction.

Although the second task seemed much easier than the first, participants correctly noted the target's position on nearly three-quarters of the trials in both displays. Random guessing would yield only one in four correct responses.

Similar findings emerged for a second set of displays. On these trials, observers wore polarizing goggles that allowed researchers to present different images to the right and left eyes. Volunteers looked at two arrays of diagonally oriented bars simultaneously, one with each eye. When their right and left eyes saw bars oriented in opposite directions, they reported not seeing target areas bearing a few differently aligned bars. But again volunteers' guesses regarding target locations proved highly accurate.

The kinds of displays that induce blindsight may fail to stimulate adequately parts of the cortex that take secondary roles in handling visual information and that probably orchestrate visual awareness, Kolb and Braun suggest. These areas may maintain indirect connections to the primary visual cortex that help to guide behavior in the absence of awareness, they propose.