cognitive science
and more
A bit about pupil size, attention, and inhibition

Let's try a little trick.

  1. Take a piece of cardboard and punch a small hole in it, no bigger than say two millimeters. If you're in a bar—this is one of those bar tricks—a beer coaster will do just fine.
  2. Cover your right eye with the piece of cardboard so that you can see through the hole. Make sure that no light gets through, except for through the hole.
  3. Cover your left eye with your hand. Make sure that no light gets through at all.
  4. Now uncover your left eye and watch what happens to the hole: It shrinks!
  5. Now recover your left eye with your hand, and watch the hole become bigger again.
A beer coaster.

This trick allows you to see your own pupillary light response. When you remove your hand from your left eye, the light that suddenly enters the eye causes your pupils to constrict. (Your right pupil as well as your left, because pupils always act together). And this, in turn, causes the little hole to appear smaller. In other words, the apparent size of the hole directly reflects the size of your pupil! (If you're brave you can try to figure out the optics behind this effect. It took me a while.)

All of this was just an elaborate introduction to tell you what you already know: Pupils respond to light. Brightness causes pupils to constrict, and darkness causes pupils to dilate. But what you may not know (unless you read my previous post) is that you don't need to look at something bright for your pupils to constrict. Just paying attention (without looking) is enough. In a way, when you pay attention to something, your pupils respond as if you were looking directly at it.

In a paper that just appeared in Journal of Vision, my coauthors and I studied this phenomenon in more detail. I'm very excited about this study, so I wanted to share the main result in a blog.

Our experiment was very simple. Participants kept their eyes fixated in the center of a display, and identified a target stimulus that appeared on the left or right side of the display. Just before the target appeared, there was a brief movement on the left or right side of the display. This movement was not relevant for the task (i.e. it did not predict the location or identity of the target), but participants were nevertheless unable to ignore it: Movement automatically attracts attention, whether you want it to or not.

So far, so just another good-old-fashioned cuing paradigm. But we added a little twist: Half of the display was bright, the other half was dark. Therefore, attention was sometimes drawn toward brightness (when the movement occurred on the bright side), and sometimes toward darkness. And because we know from previous studies that attention affects the pupillary light response, this brightness/darkness manipulation allowed us to study what happens to attention in this type of experiment.

So let's take a look at the results. The movement occurred at time 0. The target was presented 2.5 seconds later. And we're looking at what happened to the size of the pupil in between:

As you can see, about 0.4 s after the movement, the pupil started to dilate in both conditions, independent of where the movement had occurred. This is typical: Any kind of stimulation (sound, touch, vision, etc.) evokes a dilation of the pupil. You also see that, aside from this initial bump, the pupil generally dilates throughout the trial. Again, this is typical for this type of experiment, and is related to the effort that participants invest in the task. However, for our purpose, the interesting thing is the difference between the movement-on-bright (orange) and movement-on-dark (gray) conditions. This difference is shown as the blue line at the bottom of the graph.

From about 0.4 to 0.9 s after the movement, the pupil was smaller when attention had been drawn to the bright side, compared to the dark side. This shows that when you attend to (without looking at) something bright, your pupil constricts as if you were looking directly at the brightness. This had already been shown before by several groups, including ours.

But, after about 1 s, this pattern flips around: Now the pupil was larger when attention had been drawn to the bright side! So that's odd—why would that be? We believe that this reflects a phenomenon called inhibition of return. The idea is that attention is automatically drawn to the movement, but participants quickly see that this was a false alarm. Therefore, they inhibit this location for a little while after, so that their attention doesn't unnecessarily gets drawn there again. And this inhibition causes a kind of inverse-attention effect, as if participants were looking away from the inhibited location.

In summary, we have used pupil size to study exactly what happens when attention is automatically drawn to something. And we've been able to show the initial shift of attention (the green part in the graph above), and the inhibitory phase that follows (the red part). This is one of these little studies that gets a cognitive psychologist very excited!

  • Mathôt, S., Dalmaijer, E., Grainger, J., & Van der Stigchel, S. (2014). The pupillary light response reflects exogenous attention and inhibition of return. Journal of Vision, 14(14), 7. doi:10.1167/14.14.7