Illusions Across Species: Do Animals See Optical Illusions?

If optical illusions were simply quirks of human visual wiring, you'd expect them to disappear entirely in other species. They don't. Researchers studying animal cognition have found that a striking range of animals - primates, birds, fish, and even insects with compound eyes wired nothing like our own - can be fooled by illusions strikingly similar to the ones that fool humans. That cross-species overlap is one of the more compelling pieces of evidence that illusions aren't cultural artifacts or human quirks, but the visible signature of genuinely ancient, deeply conserved strategies for processing visual information efficiently.

Primates: the closest comparison

Given how closely related they are to humans, it's unsurprising that non-human primates show many of the same illusion susceptibilities we do, though testing them requires clever non-verbal methods since researchers can't simply ask a monkey what it sees. Instead, scientists train animals to make choices - reaching toward the "larger" of two stimuli, for instance - and then present illusion displays to see whether the animal's choices get systematically biased the same way a human's verbal reports would be. Studies using this approach have found that rhesus macaques and chimpanzees show size-perception biases on illusions in the Ebbinghaus and Müller-Lyer family that parallel human responses, suggesting the depth- and context-dependent size-inference machinery discussed in why do optical illusions happen has roots that predate the human lineage by a considerable margin.

Birds and the limits of similarity

Bird vision runs on a different chassis than mammalian vision in several respects - birds generally have a far higher density of color-sensitive cone types, including sensitivity to ultraviolet wavelengths humans can't perceive at all - yet susceptibility to illusions still shows up. Pigeons, extensively studied in operant-conditioning vision research because they can be trained to peck at targets for food rewards, have demonstrated susceptibility to some brightness- and size-contrast illusions in patterns that echo human perception, though not always identically. This partial overlap is itself informative: where bird and human responses diverge, it often points to genuine differences in how their visual systems weight particular cues, offering researchers a way to reverse-engineer which parts of an illusion depend on general-purpose contrast processing (likely shared broadly across vertebrates) versus which parts depend on more human- or primate-specific depth inference habits, like the corner-cue interpretation thought to partly drive the Müller-Lyer illusion.

Fish and an unexpectedly small brain doing big inference

Perhaps the most surprising findings come from fish. Researchers have documented that certain fish species can be trained to respond to illusion displays in ways that mirror human size- and brightness-illusion responses, including versions of the Ebbinghaus-type size-contrast effect. Fish brains are organized very differently from mammalian brains and lack a layered visual cortex analogous to ours, which makes the parallel illusion susceptibility particularly interesting: it suggests that certain perceptual shortcuts - comparing an object's size or brightness relative to its immediate surroundings rather than judging it in isolation - may be such a generally useful computational strategy for any visual system operating under noisy, resource-limited conditions that evolution has arrived at similar solutions independently, or very early, across a wide range of vertebrate lineages.

Insects: illusions without a cortex at all

Perhaps the strongest evidence that illusions reflect general computational principles rather than mammal-specific neural hardware comes from insects. Honeybees, whose visual systems are built from compound eyes and a comparatively tiny brain with no cortex whatsoever, have been shown in behavioral experiments to be susceptible to certain motion- and contrast-based illusions when trained to forage toward specific visual targets. Since insect and vertebrate visual systems evolved largely independently and are built from entirely different neural architectures, shared illusion susceptibility here is best explained not by shared brain structure but by shared computational problems: any visual system that needs to extract edges, motion, and relative contrast efficiently from limited neural resources tends to converge on broadly similar shortcuts, and those shortcuts produce broadly similar failure modes when pushed to their limits.

What comparative research adds to the picture

Testing illusions across species does more than satisfy curiosity about what a bee or a goldfish experiences. It's a genuine research tool: when an illusion fools species as distantly related as primates, pigeons, fish, and bees, that convergence is strong evidence the underlying mechanism is a fundamental solution to a fundamental problem in visual processing, rather than an accident of human neuroanatomy. When an illusion fails to transfer to another species, that gap can help isolate exactly which part of the mechanism is specific to more sophisticated, experience-dependent processing - the kind of cultural and developmental calibration discussed in do optical illusions work the same for everyone and illusions and child development. Illusions, in other words, aren't just entertaining tricks played on human eyes. Across the animal kingdom, they're recurring evidence of how any visual system, built from any kind of neural hardware, tends to solve the same basic problem of extracting a useful picture of the world from limited, ambiguous input.