Six experiments were conducted in naive human participants to examine
any facilitation produced in manual reaction time (RT) by the interpos
ition of a temporal gap between a warning signal and an imperative sig
nal. Peripheral visual stimuli and monoaural auditory stimuli were use
d as targets. Participants showed a facilitation of RTs to the targets
for both auditory and visual stimuli in the five experiments in which
RTs were the dependent variable. In addition, the gap effect increase
d over successive blocks of trials, suggesting learning. RTs were faci
litated only when the gap had predictive value and was salient. Using
a variable temporal gap or a visual warning stimulus did not change th
e facilitation in RTs. A further experiment demonstrated that the gap
can be perceived by the participants. The dissociation between a learn
ed and a non-learned component in the gap effect suggests that the tem
poral gap induces two independent processes: warning and disengagement
of attention. (C) 1998 Elsevier Science B.V.