One line of evidence comes 
from studies of 6-to-12-week-
old human infants conducted 
by T. G. R. Bower, then at 
Edinburgh University. In 
experiments on size constancy, 
Bower conditioned infants to 
turn their heads slightly in 
response to a 12-inch cube at a 
distance of 3.3 feet, using as a 
reward a peek-a-boo 
appearance of a hidden 
experimenter. Bower then 
moved the cube farther away, to 
9.9 feet, to test for the presence 
of constancy in the infants. 
Bower’s method thus 
circumvented the difficulties 
alluded to above of doing 
experiments on perception 
with very young infants. The 
infants were not required to 
move around in the 
environment and not required 
to learn a difficult size-
discrimination problem.

    Based on a well-known fact 
about conditioning discovered 
by Ivan Pavlov, Bower reasoned 
that, if the cube continued to 
look much the same to the 
infant, the conditioned 
response should remain in full 
force; to the extent that it 
looked different, the response 
should be weak or absent. 
Bower found that infants 
indeed continued to respond 
vigorously when the same cube 
was tested at the greater 
distance, although its visual 
angle was thereby reduced to 
one-third its initial size. The 
responsiveness to the 12-inch 
cube at that distance was not 
quite as great as to that cube 
when it remained at the 
original distance of 3.3 feet, 
perhaps because the infants 
were aware that the distance 
had changed. They responded 
only weakly, however, when a 
cube three times larger, 36 
inches in size, was placed at 
such a distance (9.9 feet) that 
its visual angle remained the 
same as that of the initial 
object. The results imply that 
infants perceive size much as 
we do, on the basis not of visual 
angle alone but of constancy.