Our visual system not only mediates information about the visual environmen
t but is capable of generating pictures of nonexistent worlds: afterimages,
illusions, phosphenes, etc. We are "aware" of these pictures just as we ar
e aware of the images of natural, physical objects. This raises the questio
n: is the neural correlate of consciousness (NCC) of such images the same a
s that of images of physical objects? images of natural objects have some p
roperties in common with afterimages (e.g., stability of verticality) but t
here are also obvious differences (e.g., images maintain size constancy, wh
ereas afterimages follow Emmert's Law: when seen while screens at different
distances are observed, an afterimage looks larger, the greater the distan
ce of the screen). The differences can be explained by differences in the r
etinal extent of images and afterimages, which favors the view that both ha
ve the same NCC. It seems reasonable to assume that before neural activity
can produce awareness, all the computations necessary for a veridical repre
sentation of, e.g., an object, must be completed within the neural substrat
e and that information characteristic of a particular object must be availa
ble within the NCC. Given these assumptions, it can be shown that no retino
topic (in a strict sense) cortical areas can serve as the NCC, although som
e type of topographic representation is necessary. It seems also to be unli
kely that neurons classified as cardinal cells alone can serve as NCC. (C)
1999 Academic Press.