The attentional blink has been attributed to capacity limitations at a cent
ral level of processing. We tested whether failure to identify the target w
ould eliminate the blink. Two agnostic patients were presented with streams
of letters, which they were able to identify, and streams of pictures, whi
ch they were unable to identify. The dual-task involved identification of a
target and detection of a probe. With letters the duration of the blink wa
s equivalent to that of the control subjects. A prolonged blink was observe
d in both patients for pictures irrespective of whether the target was iden
tified. This result indicates that failure to identify the target does neve
rtheless mobilize attentional resources sufficiently to prevent detection o
f a second target stimulus. NeuroReport 11:2775-2780 (C) 2000 Lippincott Wi
lliams & Wilkins.