Saccadic reaction times (SRTs) to suddenly appearing targets are influenced
by neural processes that occur before and after target presentation. The m
ajority of previous studies have focused on how posttarget factors, such as
target attributes or changes in task complexity, affect SRTs. Studies of p
retarget factors have focused on how prior knowledge of the timing or locat
ion of the impending target, gathered through cueing or probabilistic infor
mation, affects SRTs. Our goal was to investigate additional pretarget fact
ors to determine whether SRTs can also be influenced by the history of sacc
adic and visual activity even when these factors are spatially unpredictive
as to the location of impending saccadic targets. Monkeys were trained on
two paradigms. In the saccade-saccade paradigm, monkeys were required to fo
llow a saccadic target that stepped from a central location, to an eccentri
c location, back to center, and finally to a second eccentric location. The
stimulus-saccade paradigm was similar, except the central fixation target
remained illuminated during presentation of the first eccentric stimulus; t
he monkey was required to maintain central fixation and to make a saccade t
o the second eccentric stimulus only on disappearance of the fixation point
. In both paradigms, the first eccentric stimulus was presented at the same
, opposite, or orthogonal location with respect to the final target locatio
n in a given trial. We measured SRTs to the final target under conditions i
n which all parameters were identical except for the location of the first
eccentric stimulus. In the saccade-saccade paradigm, we found that the SRT
to the final target was slowest when it was presented opposite to the initi
al saccadic target, whereas in the stimulus-saccade paradigm the SRT to the
final target was slowest when it was presented at the same location as the
initial stimulus. In both paradigms, these increases in SRTs were greatest
during the shortest intervals between presentation of successive eccentric
stimuli, yet these effects remained present for the longest intervals empl
oyed in this study. SRTs became faster as the direction and eccentricity of
the two successive stimuli became increasingly misaligned from that which
produced the maximal SRT slowing in each paradigm. The results of the stimu
lus-saccade paradigm are similar to the phenomenon of inhibition of return
(IOR) in which human subjects are slower to respond to stimuli that are pre
sented at previously cued locations. We interpret these findings in terms o
f overlapping representations of visuospatial and oculomotor activity in th
e same neural structures.