191 — Humans quickly learn to blink strategically in response to environmental task demands

It’s pretty obvious that when you blink, you lose the ability to see. And though we don’t usually suffer from signal loss when blinking, a lot of high-frequency blinking can start to impair our ability to focus on a target or perform other vision-based tasks.

Hoppe, Helfmann & Rothkopf wanted to find out if blink rate is tied to how reliant on vision we must be to complete a task. (For example, spotting a short-lived stimulus requires a lot of visual attention, whereas listening to music can be done with the eyes completely closed.)

This paper answers two questions. First: How does blink rate change in response to different requirements placed on visual attention? And second: What is the justification for the (previously well-known) distribution of interblink intervals in average conditions?

The setup: A dot travels smoothly and slowly around an imaginary circle on a screen. It is (or non-randomly) replaced with a face for ~50ms. This is short enough that a single blink can cause the subject to miss the stimulus.

For certain viewers, the angle at which the face appeared were consistent across views: This meant that they could eventually learn when to expect the stimulus. The authors used this learning process to characterize the ways in which the subjects delayed blinks (so they didn’t miss the stimulus) and then compensated by blinking more (during times when no stimulus was expected). The subjects tended to blink-compensate more after the event than before, even though the probability of missing the event was symmetric around the angle of incidence along the circle.