Intermediate Level Theory of ConsciousnessAdditionally, Wu (2013) claims that body-centered representations—of the kind it seems
reasonable to suppose Prinz would categorize as the perspective-dependent IRs that can
become conscious—are localized in the ventral interparietal area. This area is one which Prinz
claims to count as part of high-level visual processing. Furthermore, Wu also notes evidence
from an apperceptive agnosic, who cannot see shapes and objects, but who can see textures and
colors. Her visual experience must be mostly like seeing an animal camouflaged against the
background; color and texture is uniform and consciously experienced, but there is no individu-
ation of the boundaries of the animal itself. These object/shape representations are the kinds that
Prinz identifies as populating consciousness, yet the evidence shows this woman has damage to
the lateral occipital complex, an area Prinz identifies as a part of high-level visual processing.
Prinz can always choose to redraw the lines between low, intermediate, and high-level (vis-
ual) processing. Especially given the reliable difference across modalities between associative and
apperceptive agnosia, Prinz has some justification for holding out hope for a better remapping. It
is out of the scope of the chapter to do justice to the literature on the contents of consciousness,
so I leave any controversy surrounding (1) to the side.
Claim 2Claim (2) is a hotly contested claim in philosophy (Mole 2008; Wu 2014), psychology (Kentridge
et al. 2008) and neuroscience (Lamme 2003). For the most part, scrutiny is reserved for the suffi-
ciency claim. That is, many researchers discuss evidence against the idea that all cases of attention
imply cases of consciousness; there can be, according to this crowd, attention in the absence of
consciousness. Below I will focus only on the sufficiency claim, but it is worth mentioning the
necessity claim does come under attack, too. Some think that there can be consciousness in the
absence of attention (see Block 2013; van Boxtel et al. 2011). Setting the necessity claim aside,
what are the reasons for thinking that attention is not sufficient for consciousness?
The most popular evidence for this claim comes from Kentridge et al. (2008). There they
review evidence of a patient with blindsight, who is totally blind to particular locations in his
visual fields, but can still make above-chance forced judgments about stimuli that appear in those
locations. When a visible cue is presented (to the hemifield which is still conscious), the abil-
ity of the patient in question to make discriminations about stimuli in the blind visual field is
enhanced. More precisely, when the target stimulus is validly cued (that is, when the cue indicates
correctly where the stimulus will appear) there is enhanced performance compared to an invalid
cue. This is an example of the famous Posner experimental paradigm I mentioned above, which
is thought to be a measure of attentional capture (by the cue). Hence the evidence for attention
(focused on the target in the blind hemifield) without consciousness (of that target).
Prinz’s response is that this experiment only captures spatial attention, so that attention is
allocated only a particular location in space and not any particular object. The enhancement in
performance is explained by a shift in gaze, which results from that capture of spatial attention.
Given the shift in gaze, more receptors will be allocated on the target. Whatever remaining
processing resources left in the patient’s V1 can then act on whatever is found in the gazed-at
location. When the cue is misleading with respect to the target, gaze shifts away from where the
target is located, where fewer receptors in the periphery will mean less processing of the target.
This ultimately means attenuated task performance.
Norman et al. (2013), however, respond to exactly this line of criticism of the evidence
from the original blindsight patient. They reprise the Posner technique, this time using invalid
cues that appear in different locations with respect to the cue. In either case, the target appears
equidistant from the (invalid) cue. This should mean, if Prinz is right, that performance will be