Analogous experiments have 
been done in vision. In an 
experiment that Dan Gutman 
and I did, overlapping visual 
figures of novel shapes in 
differing colors were presented 
to subjects at a fairly rapid rate. 
The subject was to rate all red 
(or for other subjects, green) 
figures on a scale of aesthetic 
preference—–a slight deception 
intended to focus attention on 
one figure only in each 
overlapping pair. When all pairs 
in the series had been shown, 
the subject was given a 
recognition test to determine 
whether the unattended figures 
as well as the attended ones 
could be recognized. The results 
indicated significant 
recognition of the attended 
figures but performance no 
better than chance for the 
unattended ones. Follow-up 
experiments showed that even 
directly after a given slide had 
been seen there was no 
evidence for any memory of the 
unattended figure, even if it 
was the outline of a very 
familiar shape.

    Why does the perception of 
shape require attention? This is 
a matter about which 
investigators can disagree, but I 
believe it can be explained in 
terms of the description process 
that I suggested is the basis of 
form perception. Such 
description occurs for the 
attended figure in each pair in 
the foregoing experiment but 
not for the unattended figure. 
Here we have a more drastic 
consequence of the absence of 
attention than was suggested in 
the cases of initial failure to 
discriminate among members of 
a class of objects. In these 
cases, attention is given to 
each member when it is seen, 
but it is focused on its global 
properties and not on its 
details.