Adaptation-Induced Blindness Is Orientation-Tuned and Monocular

Abstract

We examined the recently discovered phenomenon of Adaptation-Induced Blindness (AIB), in which highly visible gratings with gradual onset profiles become invisible after exposure to a rapidly flickering grating, even at very high contrasts. Using very similar stimuli to those in the original AIB experiment, we replicated the original effect across multiple contrast levels, with observers at chance in detecting the gradual onset stimuli at all contrasts. Then, using full-contrast target stimuli with either abrupt or gradual onsets, we tested both the orientation tuning and interocular transfer of AIB. If, as the original authors suggested, AIB were a high-level (perhaps parietally mediated) effect resulting from the 'gating' of awareness, we would not expect the effects of AIB to be tuned to the adapting orientation, and the effect should transfer interocularly. Instead, we find that AIB (which was present only for the gradual onset target stimuli) is both tightly orientation-tuned and shows absolutely no interocular transfer, consistent with a very early cortical locus.

Keywords: adaptation-induced blindness; awareness; orientation tuning.

Figure 1.

Results from Experiment 1, testing detection of gradually-onset target stimuli after adaptation to fast (8 Hz) flicker at the full range of target contrasts. (a) individual results; (b) mean results across the six participants (error bars show ± 1 standard error). As this was a 4AFC task, chance detection threshold was .25. There was no statistical difference from chance at any of the contrast levels, suggesting full adaptation-induced blindness.

Figure 2.

A schematic illustration of the procedure for Experiments 2 and 3. Observers first adjusted the stereoscope using the fixation squares, which then disappeared. They adapted to four counterphasing high-contrast gratings in one eye and were then tested either in the same or in the other eye, with gradual-onset or abrupt-onset target gratings. Because images were viewed through a mirror stereoscope, images on the left were seen by the left eye and images on the right by the right eye. Target gratings were always at full contrast, and the number of incorrect judgments was recorded. Responses were always 4AFC, with participants choosing which of four locations contained the target grating.

Figure 3.

Percentage of incorrect responses for full contrast target stimuli after adaptation to counterphasing gratings, for targets which were either 0° or 90° of relative orientation from the adapting stimuli, with either gradually onset (a) or abrupt (b) targets, tested either in the same eye as the adapting stimulus (dark grey) or in the other eye (light grey). Shaded areas ∼ 1 standard error, lines show standard deviations, and individual grey dots are individual data points. Solid lines show the means and dotted lines the medians. All five subjects completed all conditions.

Figure 4.

The orientation tuning of the percentage of error rates at full contrast. Red data points show percent of errors for stimuli presented in the same eye as the adapting stimuli; blue points show those presented in the unadapted eye, and green show the control stimuli presented without adaptors. The solid line shows a Gaussian fit to the data, fitted in ProFit Version 6.2.11, using a Leverburg-Marquardt algorithm; the fit was given two free parameters (amplitude and bandwidth), with a fixed mean of 0 and baseline of 0. Standard deviation of the fit was 7.76°, with an amplitude of 37.4%.