Motion of a different kind 
definitely does appear to be an 
important source of 
information about distance 
very early in an animal’s life. 
The experiments on the alarm 
reactions of infants to the 
"looming effect" (discussed in 
Chapter 2) suggest that there is 
an innate preference to 
interpret expanding and 
contracting retinal images of 
an object as changes in the 
object’s distance rather than 
changes in its size.

    Although these studies 
suggest that the capacity to 
perceive depth is innate, 
learning may still play a role in 
the development of depth 
perception. We have seen that 
the perceptual system is 
capable of a kind of learning: 
after a short period of exposure 
to conflicting cues (the 
trapezoid that looks like a 
rectangle at a slant vs. 
stereopsis), observers 
recalibrated the depth implied 
by a given degree of retinal 
disparity. Experiments in 
which observers view the world 
through prisms, such as one 
conducted by Arien Mack and 
Deanna Chitayat, have yielded 
similar results. Their subjects 
wore prisms over each eye that 
tilted the images slightly in 
opposite directions. As a result, 
a vertical rod appeared to be 
sloping slightly toward the 
observer. But after a period 
walking around while wearing 
the prisms, this distorting 
effect wore off. When the 
prisms were removed, vertical 
lines in the scene for which 
there was no disparity now gave 
the impression that they were 
sloping away. Thus, even 
information provided by retinal 
disparity, for which there is an 
innate physiological basis, 
seems to be subject to learned 
modification.