We tested the ability of a blindsight patient, GY,to identify in which of t
wo locations a target was presented in a spatial two-alternative forced-cho
ice paradigm (spatial 2AFC). On each trial the subject was asked to make a
second manual response indicating whether he had had any awareness of an ev
ent occurring during the trial. A. cue, presented at the fixation location,
could signal the 0.4 s period over which the target appeared within the 10
s duration of each trial. Targets of three contrasts, 93, 43 and 22% were
used. We found that GY's ability to discriminate the location of targets in
his blind field remained significantly above chance, with and without cuei
ng, for each contrast. Cueing, did, however, significantly improve his perf
ormance for low contrast targets. When he performed a similar task with nea
r threshold contrast targets in his spared visual field his discrimination
was at chance unless the presentation of targets was cued, despite his repo
rting more awareness for these stimuli than he did for low-contrast stimuli
in his blind field. These results are compared with those previously repor
ted in monkeys who received lesions to their visual cortices as infants or
adults. We conclude that (1) GY's blindsight is qualitatively different fro
m near-threshold normal vision. (2) In common with infant-lesioned monkeys
his blindsight remains even in the absence of temporal cues. (3) Residual v
ision is subject to modulation by attentional processes, or arousal, associ
ated with temporal cueing. (C) 1999 Elsevier Science Ltd. All rights reserv
ed.