Perspective is a powerful 
force in creating the 
impression of depth, even 
when other cues work against 
it. A luminous trapezoid viewed 
in a frontal plane, for example, 
when seen with one eye in a 
dark room will look like a 
rectangle at a slant. Even when 
such a figure is viewed with 
both eyes, which provides 
information from stereopsis, 
perspective strongly affects 
what is perceived, as Barbara 
Gillam, then at the Australian 
National University, and 
William Epstein and his 
associates at the University of 
Wisconsin have shown. 
Moreover, Epstein has 
demonstrated that continued 
binocular viewing of objects in 
the environment with a special 
lens in front of one eye that 
alters retinal disparity will 
result in an adaptive 
recalibration of retinal 
disparity. Ordinarily, retinal 
disparity and perspective yield 
congruent information, such 
as, for example, that a 
rectangle is slanted back from 
the frontal plane by 30 degrees. 
In the experiment, however, 
the two eyes are made to 
provide discrepant 
information: the perspective 
information that objects are in 
a frontal plane is at odds with 
the retinal disparity 
information that they are 
slanted backwards in depth. 
Apparently, the subjects learn 
to associate unconsciously a 
given disparity with some new 
depth, not the depth the 
disparity signified before, 
because, in a subsequent test 
for depth based on disparity 
using two luminous dots, the 
subjects saw depth under 
conditions of zero disparity 
and, conversely, no depth 
when some actual disparity was 
present. Thus, in only minutes, 
a pictorial cue such as linear 
perspective is sufficiently 
potent to reeducate the mind 
about stereopsis.