Many students of perception 
believe that experience plays a 
major role in the genesis of 
depth perception in the case of 
the pictorial cues. For example, 
linear perspective—–the 
converging projection of 
parallel contours receding into 
the distance—–is far more 
prevalent in the constructed 
environment of modern society 
than it is in the more natural 
environment in which Homo 
sapiens evolved. Therefore it is 
not likely that linear 
perspective would have evolved 
as an innate sign of depth.

    But how would we have 
learned to perceive depth on 
the basis of linear perspective? 
Simply by moving around in the 
world and discovering that 
what had appeared to be 
converging lines in a frontal 
plane are actually receding 
parallel lines? This would 
assume that knowledge about 
the world can affect our 
perceptions of it, which, as we 
have seen, is contrary to the 
nature of perception. A better 
explanation is the following. 
When, as children, we first 
view parallel lines in depth and 
receive the image of converging 
lines, we also have available 
other sensory information 
about depth from cues such as 
retinal disparity, convergence, 
and accommodation. These 
physiological cues produce 
veridical perception of the 
parallel lines. At this point, 
linear perspective is not 
functioning as a cue. But it is 
present. Therefore, we can 
associate the converging 
pattern with parallel lines in 
depth. Because of this 
association, later on, the 
converging pattern by itself can 
evoke the interpretation of 
parallel lines in depth. It thus 
would have become a learned 
cue.