Spatial and temporal factors in visual-auditory interaction

In two experiments, saccadic eye movements were investigated in a bimodal focused attention task with visual targets and auditory distractors under various spatial and temporal conditions. Di®erent spatial e®ects on saccadic reaction time could be ob-served depending on whether the auditory distractor preceded or followed the visual target. While vertical interstimulus distance a®ected reaction time only when the auditory signal was presented ¯rst, e®ects of horizontal displacement seemed not to depend on the temporal interstimulus relationship. In an auditory localization experiment, eye movements to the perceived sound source were measured. In con-trast to visually guided saccades, trajectories of eye movements evoked by auditory stimuli were frequently curved. Further analysis showed that vertical eye movements often start with a delay and are corrected once or twice. Both results indicate that di®erent mechanisms for azimuth and elevation cue analysis exist in the auditory system. These mechanisms are also involved in auditory-oculomotor and visual-auditory sensory integration. Latencies to target stimuli are usually signi¯cantly smaller if an additional (non-informati-ve) accessory stimulus is presented in close temporal and spatial relationship with the target. Various psychophysical and physiological studies have suggested di®erent expla-nations for this intersensory facilitation e®ect (IFE), for example, attentional or warning e®ects or multisensory information integration. In humans, interaction between the visual and the auditory system is of special impor-tance, with vision usually dominating perception while audition seems most important for the detection of warning signals. Temporal aspects of various visual-auditory interaction e®ects, particularly with respect to reaction times, have been investigated since the early 60ies. Several models reaching from simple statistical facilitation to attentional e®ects have been considered to explain the ¯ndings (for an early review see Nickerson 1973). More recently, quantitative analyses of the e®ect of spatial interstimulus relations on sac-cadic reaction time (SRT) have been performed (Frens, Van Opstal, and Van der Willige