Some people have no mind's eye, and most find out by accident
Asked to picture an apple, a small percentage of people see nothing whatsoever. Many reach middle age assuming “visualise” was a figure of speech that everyone else was also using loosely, and discover otherwise in a conversation that goes strangely.
Adam Zeman named the condition aphantasia in 2015. It is not a memory deficit: most aphantasics recognise faces, recall facts about scenes, and describe objects accurately. What is absent is the experience. There is now an objective correlate — the pupil constricts when most people imagine a bright light, and in aphantasics it does not.
Two people can use the same word for forty years, in full agreement, and mean incompatible interior experiences, with no signal anywhere that anything is wrong. It sets a hard floor on how much of another mind you are entitled to assume.
Your pupil contracts to real light and, to a smaller degree, to imagined light. The response is involuntary and cannot be performed on request, which makes it a rare objective probe of a private experience. In people reporting aphantasia the imagery response is flat while the response to actual light is normal — the eye works, the picture is not being made.
Aphantasics perform normally on many visual-memory tasks and are over-represented in some technical fields. Several report that the discovery reframed their whole biography — why they disliked certain novels, why guided meditation seemed absurd — without changing anything about their capability. The deficit is in the experience, not the information.
Self-report is a poor instrument for a private experience.
Some of the variance may be people interpreting the question differently rather than experiencing differently: “do you see it?” invites a literal and a metaphorical reading, and there is no way to check which one a respondent used. The pupillometry work is the strongest answer to this, and it rests on small samples.
A large pre-registered study finding the pupil marker does not separate self-reported groups.