“Look up and to the right, they’re lying.” You may have heard this one. It comes from NLP’s eye accessing cues, and it is among the most widely repeated and most thoroughly tested claims in the field.
The claim
NLP teaches that where a person’s eyes move when they think shows which sense they are using:
- Up and left: visual memory
- Up and right: visual construction (imagining something new)
- Level, left or right: remembered or constructed sounds
- Down and left: inner dialogue
- Down and right: feelings and bodily sensations
(The pattern is described for a typical right-handed person and is reversed for some others.)
A popular extension says that looking toward “construction” when asked about a past event means the person is making it up, in other words lying.
What studies found
Lie detection (Wiseman and colleagues, 2012). In a paper in PLoS ONE this team ran three studies:
- They recorded people lying and telling the truth and coded their eye movements. The movements did not match the NLP pattern.
- They told one group of volunteers about the supposed eye-movement cues and left another group uninformed, then asked both to spot lies. The informed group did no better.
- They analysed real press conferences in which people appealed for missing relatives, some of whom turned out to be lying. Again, no link between eye direction and lying.
The authors concluded that the evidence does not support the claim.
The wider theory. Earlier reviews of the research on NLP’s representational-system theory also found little support.
What eye movements do show
Eyes do move when we think, and that part is real. Research on gaze aversion (such as work by Glenberg and colleagues in 1998) found that people tend to look away when answering a hard question, apparently to reduce distraction and focus inward. Harder questions led to more looking away.
But this is about how hard we are thinking, not which sense we are using or whether we are lying.
Why the belief persists
- It is simple and satisfying: an easy rule to apply.
- Confirmation bias. We notice the times it seemed right.
- Confident teachers. The claim is delivered as fact.
- Lie detection is genuinely hard. Most people, including professionals, perform only slightly better than chance, so a quick rule is tempting.
Be careful
Treating a nervous glance as proof of dishonesty can hurt people. Anxiety, shyness, memory effort and cultural norms about eye contact all change where people look.
The bottom line
Eye accessing cues are a neat idea that fails when tested. You can’t read minds or detect lies by following someone’s gaze, and it is wise to doubt any course that says you can.
Next in the series: Visual, Auditory, Kinaesthetic: Does Your “Representational System” Matter?
This article is for general information and is not medical advice.