Abstract
Zihl & von Cramon (1980) asked patients with lesions of primary visual cortex to report light onset in their blind field, either by blinking, by pressing a button, or by saying “Yes”. Detection with the manual or eyelid response was more reliable than verbal report. This suggests that some response modalities have better access to residual visual pathways than others. We therefore explored whether or not contrast sensitivity in normal subjects would vary as a function of response modality—since different visuomotor pathways might be expected to have different CSFs. Subjects viewed a central fixation target indicating the beginning of a trial. Stimuli were 0.5° circular Gabor patches at 5 contrasts and 5 spatial frequencies appearing centrally or 5° above or below fixation. Following a variable temporal delay (0.5–1.5 s) stimuli were flashed for 150 ms. Detection latency was measured by 3 response modalities: 1) conventional—subjects released a “ready” key and were instructed to press a dummy key, 2) pointing—subjects released a key and pointed to the target on the display screen and 3) verbal—subjects gave a verbal response for stimulus detection. As expected, response latency increased with higher spatial frequencies and lower contrasts and was shorter for stimuli in the central visual field than in the upper or lower visual fields. There was no difference, however, in response latency for the upper and lower fields. Verbal response latencies were uniformly longer than the other two response modalities but there was no difference in reaction time between pointing and conventional response modalities. These data suggest that in normal subjects, unlike “blindsight” patients, response modality has no effect on stimulus detection—perhaps because subjects have to decide whether or not a stimulus is present before responding. Goal selection may depend more on ventral stream mechanisms, whereas on-line control of action may be mediated more by dorsal stream mechanisms.