No evidence for throttling of perceptual resolution: comments on “Partial blindness…” by Kim & Chong (2023)
Andrew Haun · 2024
In the original experiment, an observer is tasked with detecting gaussian blur of a single face in a single-interval yes/no task, where on some trials the face is blurred and on some trials the face is not blurred. In some conditions there is only a single face on the screen, while in other conditions there are multiple distractor faces on the screen. The authors found that sensitivity to face blur was worse in the multiple-face condition. They interpreted this result as meaning that in the multiple-face condition, the resolution of the face percepts is reduced because of limited perceptual bandwidth. The notion that spatial resolution of perception might be variable depending on bandwidth considerations has been proposed before, but no mechanism for such a phenomenon has been proposed. However, there is no need: there is a simpler explanation for these results that does not require any kind of perceptual resolution throttling. The explanation is that by attending to multiple possible targets, the observer must be pooling extra noise, reducing their signal-noise ratio for the target. A basic SDT simulation shows that the results of the study can be replicated closely without any adjustment of “blur sensitivity” or “perceptual resolution”.