David Barrettvisual experiences (and presumably, those of the studied monkeys), yet they make none to the
higher-level cells.
Moving to fMRI brain scans in humans, we find much the same pattern of results. During
Rapid Eye Movement (REM) sleep, when we do sometimes have visual experiences, activa-
tions are found in V3 (an intermediate level visual brain area) but not in V1. Color afterimages
associate with activity in the intermediate levels, but not in V1. The interpretation of bistable
figures—such a figure can be experienced in two different ways, like how a Necker cube can be
experienced as facing up or facing down—from subjects correlated with activity in intermedi-
ate levels. And, finally, illusory colors and illusory contours seem to invoke only intermediate
areas. This heightened activity in intermediate areas, correlated with our visual experiences,
suggests strongly that it is it is the intermediate level of processing where consciousness arises.
Humans with behavioral deficits also provide interesting evidence for Prinz’s intermediate
representation (IR) hypothesis. Supposing it is the intermediate level of processing that houses
conscious representations, we should expect three findings: (1) that damage to early visual areas
will largely destroy visual consciousness (since damage here means information does not get to
intermediate levels); (2) that damage to intermediate levels means a total loss of visual conscious-
ness (since the areas where the conscious arises will have been destroyed); and (3) that damage to
higher-level areas will not destroy consciousness (since the processing for conscious experiences
occurs earlier in the sequence). There is much evidence for (1). Many people have had damage
to V1, which resulted in blindness. There is still the phenomenon of ‘blindsight’ (Weiskrantz
1986), where individuals with V1 damage still retain the ability to navigate their environments
and retain some small level of visual acuity, but it is well known that this occurs through sub-
cortical projections from the retina to higher levels that bypass the usual route through V1. Since
blindsighters have no visual experiences (it is blindsight, after all), this does not bother the inter-
mediate level hypothesis. Evidence for (2) exists in abundance, as well. Since the intermediate
level in vision is fractionated into different areas that process specific information, there should
exist specific kinds of blindnesses to correspond to the different processing areas. One thus finds
patients with brain damage who have a form of color blindness known as achromatopsia, those
with a motion blindness called akinetopsia (where one experiences the world visually as a series
of still frames), and those with something like form blindness called apperceptive agnosia (who,
for instance, cannot accurately copy the shapes of pictures they see). The evidence for (3) also
exists. What one would predict from Prinz’s position is that damage to the higher levels would
lead only to inability to recognize and classify objects, not necessarily to experience them from
particular perspectives. This is exactly what happens in patients. They are said to suffer from
associative agnosia: they can see objects and draw them faithfully, but without any ability to
recognize what the object is.
To move beyond vision for a moment, there is ample evidence for the intermediate hypothesis
in other sensory modalities. Physiologically the brain areas that support audition and touch are
organized hierarchically, just as the areas that support vision. There is a primary auditory cortex
and a primary somatosensory cortex, and later processing areas just as with vision (‘the belt’ for
audition, and ‘S2’—the touch analogue for V2). More interestingly, one also finds the same sorts
of behavioral deficits, as with vision, in these other sense modalities. Damage to the lower levels
renders patients deaf or unable to feel. With the primary somatosensory cortex, information about
different body parts is processed in different areas, so usually there are deficits for sensation in
distinct areas of the body. In the auditory stream, at intermediate levels, one finds deafness just for
particular kinds of sounds. As one finds brain damage in these areas, one finds the same appercep-
tive/associative distinction as with vision. Some auditory deficits leave patients with the inability
to recognize sounds, although they are able to match any pairs of sounds as the same. The same