Once that critical trial had 
occurred and subjects had been 
questioned about seeing 
anything else, they might on 
subsequent trials expect to see 
something else and thus devote 
some attention to looking for it. 
That is why we could only have 
the one critical trial. We 
compensated for the 
comparatively little data 
obtained from each subject by 
using hundreds of subjects. And 
with each subject we ran trials 
after the critical one, on the 
last of which we instructed the 
subject to ignore the task of 
comparing the lengths of the 
lines of the cross. Thus subjects 
had no primary task and 
undoubtedly expected the 
critical stimulus to be 
presented. Since the time 
allowed was the same brief 
period, responses on this last 
trial informed us of what could 
be seen in the brief interval 
when attention was deployed.

    I have already alluded to one 
of our results in discussing 
perceptual grouping: When a 
Gestalt array of elements was 
unexpectedly presented along 
with the cross, grouping on the 
basis of proximity or similarity 
did not occur. A more striking 
finding was that subjects often 
were completely unaware that 
anything aside from the cross 
had been presented. We refer to 
this as inattentional 
blindness. Even more 
surprising was that when we 
presented the cross elsewhere 
on the retina than in the 
central fovea, and on the 
critical trial presented the 
unexpected stimulus precisely 
where the subject was fixating 
(i.e., foveally), not only did 
inattentional blindness occur, 
but its incidence increased, to 
more than 60 percent of the 
subjects. This result suggests 
that subjects had learned to 
disregard or inhibit objects in 
the central region of vision 
because, on the first few trials, 
the cross was always centered 
off the fovea.