Blink Rate: Suppression During the Lie, Flood After It

Sharon Leal and Aldert Vrij ran an experiment in 2008 that is unusually clean, and its result is unusually specific.

Participants either lied or told the truth in response to questions, while their eye blinks were recorded. The finding came in two parts. During the deceptive answers, blink rate went down. Immediately after the deceptive answers, in the period following the lie, blink rate went sharply up — well above baseline, and above the post-answer rate of truth-tellers.

Down, then up. The pattern is the finding, and the level at any single moment is not.

The explanation runs through cognitive load, which is the organizing idea of Vrij's entire research program and the subject of chapter 108. Spontaneous blinking is generated by a brainstem mechanism, but it is suppressed under high attentional demand — when you are concentrating hard on something, you blink less. Lying is typically more cognitively demanding than telling the truth: the liar has to construct a plausible account, monitor it for consistency with what the listener already knows, suppress the true version, and monitor their own behavior for signs of the effort. All of that competes for the same resources, and blinking gives way.

The rebound afterwards is the release. Load drops, the suppressed blinks arrive in a cluster, and the effect is large enough to be visible without instrumentation if you know to look for it.

This is a better-founded cue than most of the ones in this part, and it is important to state immediately why it still will not do what people want it to do.

The pattern indexes cognitive load, not deception. Any question that is hard to answer produces it. Someone struggling to remember a genuine event, doing arithmetic, choosing words carefully because the subject is delicate, or answering in a second language will show the same profile. This is precisely why Vrij's own work does not use load as a cue to be observed passively — it uses it as a variable to be manipulated. If lying is more effortful than truth-telling, then increasing the difficulty of the task widens the gap between liars and truth-tellers, and that is a technique rather than a tell: ask for the account in reverse order, ask unexpected questions, require drawing rather than telling.

The second limitation is measurement. Reliable blink counting means video, a frame rate, and someone counting, or an eye tracker. An interviewer forming an impression in real time about whether someone blinked less than usual, when they do not know the person's usual, is not doing what Leal and Vrij did. They are generating an impression and then finding evidence for it, which is the mechanism that produces chapter 52.

So the honest use is narrow. In recorded material with a decent baseline, a suppression-then-flood pattern time-locked to a specific answer is worth noting as a marker that the answer was effortful. Which tells you where to dig, and nothing about what you will find.