The mirror test is an experiment designed to measure whether an animal recognizes its own reflection as itself rather than as another animal. Developed by psychologist Gordon Gallup Jr. in the late 1960s, it remains the most widely used method for studying self-awareness in nonhuman species. The basic idea is simple: place a mark on an animal’s body where it can only be seen in a mirror, then watch whether the animal uses the reflection to investigate the mark on its own body.
How the Test Works
The mirror test, formally called the mark test or mirror self-recognition test (MSR), follows a structured sequence with built-in controls. First, an animal is given time to acclimate to a mirror. Many animals initially react to their reflection as if it were a stranger, displaying social or aggressive behaviors. Over time, some begin using the mirror to explore parts of their body they can’t normally see, which is the first sign they may understand what they’re looking at.
The critical step comes next. A researcher places a colored mark on a part of the animal’s body that it cannot see without the mirror, typically the face or head. The mark must be something the animal can’t feel, so it can only discover the mark visually. In the classic version, researchers use odorless, non-irritating dye. More recent fieldwork on wild baboons has used laser pointers projected onto the cheek or ear, which avoids the need to capture and handle the animal.
The experiment also includes control conditions. In one, a mark is placed somewhere the animal can see without a mirror, confirming the animal cares about marks on its body at all. In another, a mark is placed on a non-visible body part while the animal is away from the mirror, establishing a baseline for how often it touches that area normally. If the animal touches or investigates the marked spot significantly more often when it can see the mark in the mirror, it passes the test.
Which Animals Pass
The list of species that reliably pass the mirror test is shorter than most people expect. The three great apes (chimpanzees, orangutans, and gorillas) are considered the most convincingly self-aware by this measure. Bonobos have also passed in multiple independent studies. Beyond primates, bottlenose dolphins have demonstrated mirror self-recognition across several experiments, with one study finding that dolphins show signs of self-recognition at an earlier age than either humans or chimpanzees.
Asian elephants and European magpies have each shown evidence of passing, though these results have been replicated less extensively. The most surprising and controversial addition to the list is the cleaner wrasse, a small tropical fish. In studies published from 2019 onward, cleaner wrasses appeared to use mirrors to inspect and scrape off colored marks injected under their skin. Not all colors triggered this response: brown marks, which resemble the parasites these fish naturally remove from other fish, provoked scraping behavior, while blue and green marks did not. This suggests the fish were reacting specifically to what looked like a parasite on their own body rather than demonstrating a general sense of “that’s me in the mirror.”
That distinction sits at the heart of the ongoing debate. Critics, including Gallup himself, have argued that the fish may simply be responding to a combination of physical sensation and visual input rather than truly recognizing their reflection. Supporters counter that requiring spontaneous, untrained responses to purely visual marks sets a bar that may be unreasonably high for a species with very different sensory biology.
When Humans Develop Self-Recognition
In Western populations, children typically begin passing the mirror test between 18 and 24 months of age. This milestone is considered a benchmark for an emerging self-concept, the point where a child starts to understand themselves as a distinct individual. Before this age, infants tend to treat their reflection as another baby, reaching toward it or trying to look behind the mirror.
But this timeline isn’t universal. Children as old as 6 in rural non-Western societies often do not pass the mark test, and researchers believe this reflects differences in cultural experience with mirrors and with being scrutinized by adults, not differences in cognitive ability. A child who has rarely encountered a mirror, or who has been raised in a culture where touching your own face in response to a strange mark isn’t a natural reaction, may simply not perform the expected behavior. This is one of the strongest pieces of evidence that failing the test does not equal lacking self-awareness.
Why Failing Doesn’t Mean No Self-Awareness
The mirror test has drawn substantial criticism for what a “fail” result actually tells us. The core concern is that the test is heavily biased toward species that rely on vision and that are motivated to touch or investigate marks on their bodies. A species could be fully self-aware and still fail because it doesn’t care about a dye spot, because it navigates the world through smell or echolocation rather than sight, or because it lacks the motor skills to touch the marked area.
Researchers have also pointed out that the test’s binary pass-or-fail structure misses a lot of nuance. Rhesus macaques, for instance, have been trained to use mirrors to guide their movements and investigate marks on their faces, behavior that looks identical to passing. Some scientists argue this shows the monkeys had the cognitive capacity all along but lacked experience with the specific sensory-motor mapping that mirrors require. Others maintain that trained responses don’t count as genuine self-recognition because the behavior needs to arise spontaneously to reflect true self-awareness.
There is also the problem of anthropocentrism, the assumption that the human way of demonstrating self-awareness (visually inspecting yourself) is the universal standard. Many species experience the world primarily through senses other than vision, which means a mirror is simply the wrong tool for measuring what they know about themselves.
Tests Built for Other Senses
To address the visual bias, researchers have developed alternatives tailored to how different species actually perceive the world. The most well-known is the “olfactory mirror” test designed for dogs. Dogs fail the traditional mirror test consistently, which isn’t surprising given that they navigate their social world through scent rather than sight.
In the olfactory version, dogs are presented with canisters containing different odor samples, including their own urine, their own urine with an unfamiliar scent added to it, and the urine of other dogs. Across 36 dogs tested, a clear pattern emerged: dogs spent more time investigating their own odor when it had been modified with an additional scent than when it was unaltered. This mirrors the logic of the mark test. Just as a chimpanzee investigates a new mark on its face, dogs investigate a change to their own scent “image.” Dogs also spent more time sniffing the odors of other dogs than their own unmodified odor, which is consistent with how they behave in the real world, where your own smell is old news.
A follow-up experiment confirmed that the dogs weren’t just attracted to novelty. When presented with the added scent by itself (without their own urine), dogs showed less interest than they did in the combined modified sample. The unfamiliar smell was only interesting when it was attached to their own scent signature, suggesting the dogs recognized something had changed about “them.”
What the Test Actually Measures
The mirror test was originally conceived as a way to operationalize the concept of self, to turn a philosophical question into something that could be measured in a lab. It succeeded in creating a repeatable experiment, but decades of research have made clear that it measures something narrower than “self-awareness” in the broadest sense. What it reliably detects is the ability to map visual information from a mirror onto one’s own body and act on it. That ability correlates with self-awareness, but the two aren’t identical.
Failing the test could mean an animal lacks self-awareness, or it could mean the animal doesn’t rely on vision, doesn’t care about marks, hasn’t encountered mirrors before, or doesn’t have hands to touch its face. Passing the test is more informative: an animal that spontaneously uses a mirror to investigate a hidden mark on its body almost certainly understands, on some level, that the reflection belongs to it. But even passing leaves open questions about the depth of that understanding and whether it resembles anything like the rich, narrative self-awareness humans experience.