Why Do We See Optical Illusions?

You stare at two lines that are clearly different lengths — then measure them and discover they are identical. You watch a static image pulse and shimmer as if it is moving. You look at a dress in a photograph and see unmistakably white and gold — while the person next to you sees just as unmistakably blue and black. Optical illusions are not mere curiosities or party tricks. They are controlled experiments that reveal the hidden machinery of human visual perception with remarkable precision.

Every optical illusion is, in a sense, a window into the gap between physical reality and perceived reality. That gap — the distance between what is actually out there and what the brain constructs — is larger, more systematic, and more revealing than most people ever realize. Understanding why we see optical illusions is understanding how the brain builds the world we experience.

source/credit: pexels@CottonbroStudio

Vision Is Not Photography

The most important thing to understand about optical illusions is that they expose a fundamental truth about the visual system: the eyes do not work like cameras, and the brain does not passively record what the eyes send it. Vision is an active, constructive process — a collaboration between the retina, the visual cortex, and a vast network of brain regions that interpret, predict, and fill in the visual scene based on prior knowledge, context, and expectation.

The visual system evolved not to produce a perfectly accurate representation of the physical world, but to produce a useful one — rapidly, efficiently, and with sufficient accuracy for survival and action. The shortcuts and assumptions built into this system are extraordinarily effective under normal circumstances. Optical illusions are the specific conditions under which those shortcuts produce systematic errors — revealing the rules the visual brain operates by.

The Visual Processing Pathway

Before exploring specific illusions, it helps to understand how visual information travels through the brain.

From Eye to Brain

Light enters the eye and is focused by the lens onto the retina — a thin layer of photoreceptors lining the back of the eye. Cone cells in the central fovea handle color and fine detail. Rod cells in the peripheral retina handle low-light and motion detection. Retinal ganglion cells collect signals from these photoreceptors and send them along the optic nerve to the brain.

The optic nerves from both eyes meet at the optic chiasm, where signals from the left visual field of both eyes are routed to the right hemisphere and vice versa. Signals travel to the lateral geniculate nucleus (LGN) of the thalamus — a crucial relay station — and then onward to the primary visual cortex (V1) in the occipital lobe at the back of the brain.

Two Visual Processing Streams

From V1, visual information splits into two major processing pathways:

• The ventral stream — traveling from V1 toward the temporal lobe. Responsible for object recognition, color processing, and identifying what something is. Sometimes called the "what" pathway.
• The dorsal stream — traveling from V1 toward the parietal lobe. Responsible for spatial processing, motion detection, and guiding action — determining where something is and how to interact with it. Sometimes called the "where/how" pathway.

Different types of optical illusions exploit different stages and streams of this processing pathway — which is why understanding what kind of illusion you are looking at reveals which aspect of visual processing it is targeting.

Why Does the Brain Take Shortcuts?

The visual system faces an enormous computational challenge: it must extract meaningful, actionable information from a constant flood of raw sensory data — rapidly, in real time, with limited processing resources. To do this efficiently, the brain relies on a set of built-in assumptions about how the world is structured — heuristics derived from evolutionary experience and individual learning.

These heuristics include assumptions such as:

• Light comes from above
• Objects that are higher in the visual field are usually farther away
• Edges and contours define object boundaries
• Similar colors belong to the same object
• Surfaces that face the light source are lighter than surfaces in shadow
• Objects that partially overlap are at different depths

These assumptions are correct the vast majority of the time in natural environments — which is why they have been preserved by evolution. Optical illusions are specifically constructed — or occasionally occur naturally — to violate these assumptions in ways that expose them.

Types of Optical Illusions and Their Neuroscience

Optical illusions are not a single phenomenon but a diverse family of perceptual effects, each targeting different stages and mechanisms of the visual system.

Geometric Illusions — When Size and Length Are Deceived

Geometric illusions exploit the brain's use of contextual cues to judge size, length, and spatial relationships.

The Müller-Lyer illusion is one of the most studied in all of perceptual psychology. Two lines of identical length appear dramatically different depending on whether arrowheads at their ends point inward or outward. The brain interprets the context of the arrows as a depth cue — in three-dimensional space, inward-pointing arrows resemble the inside corner of a room (typically closer), while outward-pointing arrows resemble the outside corner of a building (typically farther). The brain applies size constancy — adjusting perceived size based on inferred distance — and produces an erroneous length judgment as a result.

The Ponzo illusion similarly exploits depth perception: two identical horizontal lines placed across converging lines (resembling railway tracks receding into the distance) appear to be different sizes — the upper one seeming larger because the converging lines signal that it is farther away, and the brain automatically scales perceived size upward for more distant objects.

Crucially, these illusions persist even when you know they are illusions — because they operate at a level of visual processing that precedes conscious awareness and is not accessible to rational correction. Knowing the lines are equal does not make them look equal. This is one of the strongest pieces of evidence that visual processing is not fully under top-down cognitive control.

Ambiguous Figures — When the Brain Cannot Decide

Ambiguous figures are images that contain two or more equally valid perceptual interpretations and no stable single solution. The brain alternates between them involuntarily.

The Necker Cube — a wireframe drawing of a cube with no depth cues — can be perceived as facing upper-left or lower-right. The brain cannot settle on a single interpretation because neither is more consistent with the sensory data than the other, and so it cycles between them — typically every few seconds.

The Rubin's Vase alternates between a vase in the foreground and two faces in profile — a classic illustration of figure-ground segregation: the visual system's process of deciding which part of a scene is the object (figure) and which is the background (ground). Only one can be figure at a time, and the brain flips between the two solutions.

The duck-rabbit illusion — popularized by philosopher Ludwig Wittgenstein — can be seen as either a duck facing left or a rabbit facing right. Research has shown that whether you see the duck or rabbit first is influenced by the time of year — people are more likely to see the rabbit near Easter and the duck at other times — demonstrating that even high-level conceptual context (seasonal associations) influences low-level perceptual interpretation.

Impossible Objects — When the Brain Cannot Reconcile

Impossible objects are two-dimensional drawings that present locally consistent visual cues that cannot be globally reconciled into a coherent three-dimensional structure. The brain's visual system interprets local regions of the image correctly but fails to integrate them into a globally consistent object — producing a paradoxical perceptual experience.

The Penrose triangle (the "impossible triangle") and the Penrose stairs (an "impossible" staircase that perpetually ascends or descends) were developed by mathematician Roger Penrose and became celebrated after the artist M.C. Escher incorporated them into his iconic lithographs. Each local corner or section of these figures is geometrically plausible — the impossibility only emerges when the brain attempts to integrate all sections into a coherent whole.

Motion Illusions — Static Images That Appear to Move

Motion illusions produce the compelling perception of movement in images that are completely static. They exploit the visual system's motion-detection machinery — particularly cells in area V5/MT of the visual cortex that are specifically sensitive to movement.

High-contrast repeating patterns — particularly those with specific spatial frequency, orientation, and color relationships — can produce rhythmic activation of V5/MT cells in the absence of actual motion, generating the perception of pulsing, rotating, or flowing movement. The Rotating Snakes illusion by Akiyoshi Kitaoka is among the most striking contemporary examples, producing strong rotation percepts in a static image.

Research has shown that these illusions are stronger in peripheral vision — where motion- sensitive rod cells and V5/MT inputs are more dominant — and weaker when the pattern is viewed directly in central vision. Fatigue of direction-selective neurons is also thought to contribute: when cells responding to one direction of motion become fatigued, cells responding to other directions become relatively more active, producing an apparent shift in perceived motion direction.

Color and Brightness Illusions — When Context Overrides Reality

Color and brightness illusions reveal how strongly context — rather than absolute physical properties — determines perceived color and luminance.

The Checker Shadow illusion by Edward Adelson of MIT is one of the most compelling demonstrations. Two squares on a checkerboard — one in light, one in shadow — appear dramatically different in shade. They are physically identical in luminance. The brain compensates for the inferred shadow by lightening the perceived shade of the square it believes to be in shadow — a process called lightness constancy. This compensation is normally invaluable for perceiving object colors accurately across different lighting conditions, but in this carefully constructed image, it produces a striking error.

The same principle — the brain computing color and brightness relative to context rather than in absolute terms — underlies a wide range of color illusions and is central to understanding one of the most discussed optical illusions in internet history.

The Dress — Why People Saw Different Colors

In February 2015, a photograph of a dress posted on social media divided the world into two apparently irreconcilable camps. Some people saw the dress as unmistakably white and gold. Others saw it just as unmistakably as blue and black. The photograph generated millions of social media posts and became one of the most intensely scientifically analyzed images in the history of perceptual psychology.

Why Did People Disagree?

The dress photograph was ambiguous in a very specific and scientifically interesting way: the lighting conditions in the image were sufficiently ambiguous that the brain could not reliably determine the illuminant — the color and intensity of the light source illuminating the dress. And because perceived color depends critically on the inferred illuminant, different brains — using different assumptions about the lighting — arrived at genuinely different color percepts from identical pixel data.

The key mechanism is color constancy — the visual system's ability to perceive object colors as stable across dramatically varying illumination conditions. A red apple looks red whether viewed under warm incandescent light or cool daylight, even though the actual wavelengths reflecting off its surface differ considerably. The brain achieves this by estimating the color of the illuminant and then subtracting it from the surface reflection — a process called chromatic adaptation.

The Two Camps — A Matter of Illuminant Assumption

Research by Bevil Conway and Karl Gegenfurtner and others established that the two camps of dress perceivers were making different unconscious assumptions about the illuminant:

• People who saw white and gold were implicitly assuming the dress was illuminated by a bluish light source (such as overcast daylight or shadow) — and subtracting that blue illuminant from the surface, arriving at white and gold as the underlying object color
• People who saw blue and black were implicitly assuming the dress was illuminated by a warm yellowish-orange light source (such as incandescent or artificial indoor light) — and subtracting that warm illuminant, arriving at blue and black as the object color

The actual dress — purchased and photographed under studio conditions — was confirmed to be blue and black. But the point of scientific interest is not which perception was "right" — it is that the same physical image produced two genuinely different and internally consistent perceptual experiences in different brains, based on different unconscious assumptions about illumination.

Why Did Different People Make Different Assumptions?

Research identified several factors associated with which color camp a person fell into:

• Chronotype — people who identified as early risers (morning types) were more likely to see white and gold, possibly because they have more experience processing objects in natural outdoor daylight. Night owls were more likely to see blue and black, possibly reflecting more experience with artificial indoor illumination.
• Age — older individuals tended more toward white and gold, consistent with age-related changes in the lens of the eye (which yellows with age, potentially shifting chromatic adaptation baselines)
• Individual differences in chromatic adaptation — reflecting genuine variation in how strongly the visual system discounts different illuminants

The dress became a landmark demonstration that color perception is not a simple readout of wavelength — it is a constructive inference about object surface properties, shaped by prior experience, individual neurobiology, and unconscious assumptions about the visual world.

What Optical Illusions Reveal About the Brain

Beyond their entertainment value, optical illusions have contributed substantially to scientific understanding of visual processing. Each category of illusion reveals a different principle:

• Geometric illusions reveal that size and length perception are not absolute but relative — computed in the context of depth and distance cues that the brain infers from the scene
• Ambiguous figures reveal that perception is not a passive readout of sensory data but an active competition between multiple hypotheses, with the winning hypothesis being the one the brain momentarily finds most plausible
• Impossible objects reveal that local and global visual processing operate somewhat independently — and that global integration can fail even when local processing is accurate
• Motion illusions reveal the specific computational mechanisms of motion detection — and the conditions under which those mechanisms can be triggered without actual motion
• Color and brightness illusions reveal that perceived color and luminance are computed relative to context, not absolute physical values — a design feature that serves color constancy at the cost of occasional striking errors
• The dress reveals that even high-level perceptual outcomes — the basic color of an object — are shaped by individual differences in prior experience, neurobiology, and unconscious inferential assumptions

Individual Differences in Optical Illusion Susceptibility

Not everyone experiences optical illusions with the same strength, and these differences are scientifically meaningful.

Cultural Differences

One of the most famous findings in cross-cultural psychology is that the Müller-Lyer illusion is significantly weaker in individuals raised in non-Western, non-urban environments — particularly in cultures with less exposure to the right-angle architecture and perspective-dominated visual environment of Western cities. People raised in circular dwellings and open natural landscapes develop different visual heuristics about straight lines and angles — and are therefore less susceptible to illusions that exploit assumptions built from a carpentered-world visual diet.

This finding, originally reported by Marshall Segall, Donald Campbell, and Melville Herskovits in 1966 and extensively replicated since, is powerful evidence that visual perception is not entirely hard-wired but is partially shaped by the visual environment in which the brain develops.

Age

Susceptibility to many geometric illusions decreases with age — children are often more strongly affected than adults. This is thought to reflect the progressive refinement and calibration of the visual system's depth and size heuristics through accumulated visual experience over the lifespan.

Clinical Conditions

Certain clinical populations show altered susceptibility to specific illusions:

• People with autism spectrum disorder (ASD) tend to show reduced susceptibility to many context-dependent illusions — including the Müller-Lyer and hollow face illusion. This has been interpreted as reflecting a tendency toward more "local" visual processing with less weighting of top-down contextual predictions — consistent with theories emphasizing atypical predictive processing in ASD
• People with schizophrenia show a striking resistance to the hollow face illusion — the tendency to perceive a concave mask of a face as convex (because faces are almost always convex). This resistance may reflect reduced top-down prior knowledge overriding bottom-up sensory signals — a pattern consistent with broader theories of predictive processing disruption in schizophrenia

Afterimages — When Vision Persists After the Stimulus Is Gone

Afterimages are perceptual phenomena closely related to optical illusions — the persistence of a visual impression after the stimulus that caused it has been removed.

Negative Afterimages

Stare at a bright red shape for 30 seconds, then look at a white surface — you will see a cyan (blue-green) afterimage. This is a negative afterimage, and it arises from the adaptation of opponent-process color channels in the visual system. Cone cells responding to red wavelengths become fatigued after sustained stimulation. When you look at a neutral white surface — which stimulates all color channels equally — the fatigued red channel responds weakly while the opponent cyan channel responds normally, producing the impression of cyan where the red stimulus had been.

Motion Aftereffects — The Waterfall Illusion

Stare at a waterfall for 30–60 seconds, then look at a static scene — the rocks and surrounding landscape will appear to drift upward. This is the motion aftereffect — also called the waterfall illusion or motion after-effect (MAE). Direction-selective neurons in area V5/MT that have been responding to downward motion for an extended period become fatigued. When the moving stimulus is removed, the neurons responding to the opponent direction (upward) are momentarily more active by comparison — producing the perception of upward movement in a static scene.

Motion aftereffects have been used as a powerful research tool in visual neuroscience — allowing researchers to probe the specificity, duration, and transfer characteristics of motion-sensitive neural populations with considerable precision.

When Visual Disturbances Are Not Illusions

Optical illusions are normal, universal features of healthy visual perception that occur because of how the visual system is constructed. They require no medical attention. However, certain visual disturbances may superficially resemble optical illusions but warrant professional evaluation.

Signs to Discuss With a Doctor

• Persistent visual distortions — straight lines appearing wavy or bent, or objects appearing distorted in size — that are new, persistent, and occur in everyday viewing rather than in response to a specific constructed stimulus. These may indicate conditions affecting the retina (such as macular degeneration or central serous retinopathy) or the visual cortex
• Visual snow — a persistent overlying static or "snow" pattern in vision — may indicate visual snow syndrome, a neurological condition distinct from optical illusions
• Persistent afterimages that linger for minutes or hours — possible palinopsia, which can be associated with migraine, medication effects, or neurological conditions
• Sudden loss of part of the visual field or sudden onset of floaters and flashes — warrants urgent ophthalmological assessment
• Visual hallucinations — seeing things that are not present, particularly if persistent, detailed, and not associated with a known trigger such as fever or medication

FAQ

Why do optical illusions still work even when you know they are illusions?

Because the mechanisms that produce optical illusions operate at stages of visual processing that precede and are largely inaccessible to conscious cognitive control. The Müller-Lyer lines still look different in length even when you have just measured them and know they are equal — because the size and depth heuristics producing the illusion are computed automatically and pre-consciously by early visual processing areas. This demonstrates a fundamental architecture of the visual system: top-down knowledge and bottom-up sensory processing are parallel streams that do not fully override each other, and many illusions are generated at a level where rational correction simply cannot reach.

Why did some people see The Dress as white and gold and others as blue and black?

Because the photograph was ambiguous about the color of the light source illuminating the dress, and different people made different unconscious assumptions about that illuminant. The brain computes perceived color by estimating the illuminant and subtracting it from the surface reflection — a process called chromatic adaptation or color constancy. People who assumed a bluish daylight illuminant perceived the dress as white and gold. People who assumed a warm artificial illuminant perceived it as blue and black. Factors including chronotype, age, and individual variation in chromatic adaptation contributed to which assumption different brains defaulted to. The actual dress was blue and black.

Are some people more susceptible to optical illusions than others?

Yes, and the differences are scientifically meaningful. Cultural background significantly affects susceptibility to geometric illusions — people raised in non-Western, non-urban environments are measurably less susceptible to illusions like the Müller-Lyer that exploit assumptions built from rectangular architecture. Age also matters — children are typically more strongly affected than adults. Clinical populations show characteristic differences: people with autism spectrum disorder tend to show reduced susceptibility to context-dependent illusions, while people with schizophrenia show reduced susceptibility to the hollow face illusion — both reflecting underlying differences in how strongly top-down predictions shape bottom-up sensory processing.

What is the hollow face illusion and why is it so powerful?

The hollow face illusion is the experience of perceiving a concave mask of a human face as convex — seeing it as a normal protruding face rather than an indentation. It is one of the most powerful and robust illusions known, working even when the viewer knows it is hollow and can see it rotating. It works because the brain's prior knowledge about faces — that faces are always convex — is so strong that it overrides the genuine bottom-up sensory evidence of concavity. The top-down prediction about face shape is assigned extremely high precision and simply wins the competition with the physical sensory data. Its failure in schizophrenia is one of the most important pieces of evidence for predictive processing theories of that condition.

Can optical illusions reveal anything about mental health?

Yes — researchers use specific illusions as non-invasive probes of how the brain weights top-down predictions against bottom-up sensory evidence. The hollow face illusion, the Müller-Lyer, and several others show characteristic and replicable differences in people with schizophrenia, autism spectrum disorder, bipolar disorder, and major depression. These differences are consistent with broader theories about how predictive processing is altered in these conditions — with schizophrenia associated with reduced top-down precision weighting and autism associated with enhanced weighting of bottom-up sensory signals relative to prior predictions. Illusion research is therefore not merely perceptual psychology — it is contributing to fundamental understanding of psychiatric neuroscience.

References

• Gregory RL: Eye and Brain — the psychology of seeing, visual heuristics and illusion mechanisms (1966, updated review 2022)
• Adelson EH: Lightness perception and lightness illusions — the checker shadow illusion (1995, updated analysis 2022)
• Conway BR and Gegenfurtner KR: The neuroscience of color perception — color constancy and the dress illusion (2015, follow-up studies 2023)
• Segall MH, Campbell DT, Herskovits MJ: Cultural differences in the perception of geometric illusions (1966, replication and meta-analysis 2023)
• Predictive processing and reduced illusion susceptibility in autism spectrum disorder: a systematic review (2023)
• Hollow face illusion and top-down processing failure in schizophrenia: neuroimaging evidence (2022)
• Kitaoka A: Rotating snakes and related peripheral drift illusions — mechanisms and V5/MT involvement (2022)
• Motion aftereffects and direction-selective neural adaptation in area V5/MT (2023)
• The dress illusion: individual differences in color perception, chronotype, and chromatic adaptation (2015, updated 2022)

This article is for educational purposes only and is not a substitute for professional medical advice, diagnosis, or treatment. If you are experiencing persistent, unexplained visual disturbances that are not clearly associated with a known optical illusion, please consult a qualified ophthalmologist or neurologist.