← Blog

Right Brain · Reading

Why You See Faces in Things

A plug socket looks startled. The front of a car looks smug. A crease in a curtain watches you from across the room, and you have to look twice to un-see it. This is pareidolia, and the usual explanation — an overactive imagination — has it backwards. Nothing is being imagined. A very fast, very committed piece of visual machinery has found the pattern it is built to find, in a place that happens not to contain one. Take in the whole. Before the parts.

By ComunAI · 20 July 2026 · ~6 min read

The face system is fast, and it is cheap to trigger

Faces get treatment no other object category receives. The trigger is remarkably crude: two dots above a line, arranged roughly right, is enough. That is why :-) works, why headlights read as eyes, and why you cannot look at a three-hole socket neutrally once someone has pointed it out.

The tuning makes sense when you consider the asymmetry of errors. Missing a real face — a person, an expression, an approach — is expensive. Briefly seeing a face in a rock costs you nothing but a second glance. A system built under that asymmetry should be set to fire early and forgive false alarms, and that is what the behaviour looks like. Moderate (We'd flag that the evolutionary framing is a reasonable interpretation rather than a directly tested result — the behaviour is well documented; the why is inference.)

Face-like objects get face-like processing

The more interesting question is whether a face-like object is merely interpreted as a face after the fact, or actually handled by face machinery from the start. The evidence leans towards the second. Work by Nouchine Hadjikhani and colleagues using magnetoencephalography found that face-like objects elicited an early, face-characteristic response in face-selective cortex at around 165 milliseconds — the same early window in which real faces produce their signature response.¹ Strong

That timing matters. A sixth of a second is too early for deliberation. Whatever is happening, you are not reasoning your way to a face in the curtain; the face arrives already assembled, and your knowledge that it's a curtain shows up late and changes nothing — the same sealed-off quality we described in Why Optical Illusions Fool Everyone.

Faces are processed as wholes, not as parts

The defining feature of face perception is that it is holistic: the face is handled as a single integrated pattern rather than a list of features. Two classic demonstrations make this vivid, and both are easy to verify on yourself.

• The inversion effect. Turn a photograph upside down and recognition falls off far more steeply for faces than for other objects — the disproportionate cost is the finding, established by Robert Yin.² Strong Upside down, the holistic route stops working and you are reduced to laborious feature-by-feature comparison.
• The Thatcher illusion. Take a face, invert just the eyes and mouth within it, and leave the whole image upside down. It looks broadly normal. Turn it the right way up and it becomes grotesque — the same image, transformed by orientation alone. Peter Thompson's original demonstration remains one of the most striking results in the field.³ Strong
• The composite effect. Align the top half of one face with the bottom half of another and the top becomes noticeably harder to identify — the halves fuse into a new whole you cannot easily un-fuse. Misalign them and the difficulty drops away.⁴ Strong

Each of these is the same point from a different angle: you do not see eyes, nose and mouth and then compute a face. You see a face, and the parts are downstream of it. That ordering — whole first, parts second — is the thread running through the Gestalt grouping principles we covered in Gestalt Perception: How the Mind Groups What You See.

Where the argument is still live

There is a region of the ventral visual pathway that responds much more strongly to faces than to other objects — the fusiform face area, identified in the imaging work of Nancy Kanwisher and colleagues.⁵ Strong That the region exists and is face-preferring is not seriously disputed.

What it means is disputed, and we'd rather show you the seam than paper over it. One reading is that faces are a special category with dedicated machinery. A competing account, associated with Isabel Gauthier and Michael Tarr, holds that the region is really a hub for fine-grained expertise at telling apart very similar members of a familiar category — faces being the category almost everyone is expert in, with bird experts and car experts showing related patterns.⁶ Contested Decades in, this remains genuinely unresolved. Anyone who tells you it's settled is telling you about their preferred camp, not the literature.

What this does and doesn't tell you about you

Two things worth saying plainly, because the internet says otherwise.

Seeing faces in objects is ordinary. It is a normal property of a normally tuned visual system, not a sign of anything about your personality, creativity, or mental state. Any quiz claiming to read your character from what you see in an ambiguous image is entertainment.

Face recognition ability varies a lot between people — the range runs from developmental prosopagnosia at one end to unusually strong recognisers at the other Moderate — and where you sit on it is not a verdict on your attention or effort. If difficulty recognising people is causing you real problems in daily life, that is a conversation for a clinician, not a puzzle app.

And the limit we always restate: there is no good evidence that playing perception games improves your everyday face recognition, or transfers to any real-world ability. Broad transfer is the weak link in this whole literature, and we went through it in Do Brain-Training Games Actually Work? Contested

What this means for Right Brain

Right Brain is named for a metaphor, not a hemisphere — the open, whole-first way of looking, as against the narrow, part-by-part one. Faces are the clearest case in all of perception of the whole arriving before the parts, which is why they are such a satisfying thing to catch yourself doing. The offer is not a score and not a claim about your abilities: a few quiet minutes in which the machinery becomes visible, and the socket on the wall looks back.