Early selective attention effects on cutaneous and acoustic blink reflexes

In two experiments, selective modulation of the human blink reflex was examined by direct­ ing subjects to judge duration of the startling or nonstartling member of an acoustic-cutaneous stimulus pair. The startling stimulus was acoustic in the first experiment and cutaneous in the second. In both experiments, magnitude and onset latency were facilitated when the subjects attended to rather than away from the reflex-eliciting stimulus, although changes in cutaneous reflex size were insignificant. However, a nonselective inhibition of blink magnitude on warned relative to unwarned trials, associated with latency facilitation of the cutaneous reflex, was a stronger effect. The existence of both selective, concordant and nonselective, discordant changes in latency and magnitude could not be explained by a single mechanism. Heart rate changes also suggested that warning initiated some process in addition to attention to intake. The predictability of warned events may playa larger role than has previously been recognized. Can attention modulate the early phases of in­ formation processing or, instead, is the sensory pro­ cessing of ignored stimuli identical to that of attended stimuli? Behavioral evidence has not been able to resolve the issue in favor of either early­ selection (Broadbent, 1958; Treisman, 1969) or late­ selection theories (Deutsch & Deutsch, 1963; Norman, 1968). Studies of evoked brain potentials in the human, however, have commonly supported the early-selection view (e.g., Desmedt & Robertson, 1977; Eason, Oakley, & Flowers, 1983; Hillyard, Hink, Schwent, & Picton, 1973). The general finding has been that potentials evoked by stimuli toward which attention is directed differ from the potentials evoked by unattended stimuli as early as 50 to 100 msec. On the basis of such early differences,. Hillyard (1981) interpreted the findings as being dependent upon a tonic prestimulus set rather than an active recognition of each stimulus, and Naatanen (1982) similarly suggested that a selective state was "already prevailing" at the moment of stimulus de­ livery. The state was assumed to produce a "tem­ porary bias somewhere in the sensory system" (Naatanen, 1982, p. 629).