Clearly, then, there is 
still much to learn about the 
role of motion cues in depth 
perception. My own view is that 
the perceptual system strives to 
solve the problem of what is 
signified by the transforming 
retinal image created by 
motion. If the image is that of a 
connected object (as in the 
kinetic depth effect), a pattern 
of shifting eccentric circles (as 
in the stereokinetic effect), or 
a gradient of texture motion, it 
is likely that the perceptual 
system will readily conceive of 
the transforming image as 
differing perspective 
projections over time of a 
three-dimensional array. But if 
the pattern is a relatively 
impoverished one of 
unconnected elements, that 
solution is not so readily 
available--unless it is the 
observer, not the display, that 
is moving. 
 
    From this point of view, 
depth perception is not an 
inevitable solution to the 
problem posed by the 
transforming stimulus, but it is 
the preferred one. Why it is 
preferred we do not know. 
Although progress has been 
made in isolating the cues that 
govern our perception of the 
third dimension, little is 
known about what happens 
inside the head once the 
stimulus information that 
constitutes a cue is registered 
on the retina. Experiments 
designed to investigate this 
issue suggest that the answer 
may lie in some preference of 
the mind for solutions that are 
analogous to the principle of 
parsimony in science. Consider 
further the example given of 
the kinetic depth effect. Why 
does the transforming image 
yield an impression of a rod 
rotating in depth rather than 
simply an object 
simultaneously changing 
length and orientation? The 
answer could be that, for a 
perceptual system that faces a 
situation of ambiguity and 
"knows" about perspective 
foreshortening, the solution of 
a rigid object rotating in the 
third dimension is the one that 
most elegantly accounts for the 
facts. Later in this chapter, we 
will see a similar issue arise 
with regard to the pictorial 
cues.