Under these conditions, the 
only basis for perceiving the 
spot’s location is our knowledge 
of where the eyes are looking. 
If the eyes were to be slowly 
drifting when viewing a spot, 
we could be tracking a spot that 
was moving slowly. Perhaps 
that explains why suggestion 
can be effective. On hearing 
that the spot is moving, say to 
the right, we can imagine that 
we are tracking it to the right 
when, in fact, the eyes remain 
stationary.

    Some investigators have 
argued that the autokinetic 
effect results from actual eye 
movements. The idea behind 
this argument is simply that, 
with eye movement, the image 
of the spot displaces over the 
retina and that retinal 
displacement causes the 
illusion. Such a theory is 
inadequate because, as we have 
seen, stationary objects do not 
appear to move every time the 
eyes move. Position constancy 
is achieved, presumably 
because retinal displacement is 
discounted when the 
perceptual system "knows" that 
it is caused by eye movement. 
For an eye-movement theory of 
the autokinetic effect to be 
tenable, it would have to be 
maintained that the eyes move 
but that the brain does not 
"know" that they are moving. 
One difficulty for this theory is 
that it must also predict that an 
entire stationary scene will 
appear to move, not just an 
isolated spot. I suggest that it is 
not eye movement that causes 
the illusion but the illusory 
misperception of eye behavior. 
The eyes are stationary but are 
misperceived to be tracking the 
(stationary) spot because the 
perceptual system "believes" 
for some reason—–whether it be 
suggestion or self-suggestion—–
that the spot is drifting slowly 
across the field.