The “Yellow Snow” Test for Self-Recognition

The Mirror Self-Recognition Test

The mirror self-recognition (MSR) test has long been used to assess whether an animal is self-aware, whether it has a sense of self. In the classic version of the test, a colored mark is placed on an animal’s body in such a way that it can only be seen in a mirror. To pass the test, the animal must spontaneously use the mirror to detect the mark and then scratch or otherwise direct activity toward it, thereby indicating recognition of the image in the mirror as itself and not some other animal.

Not only is self-recognition considered to be an indication of higher cognitive functioning, but it has also been seen as a potential springboard to even more sophisticated abilities, such as being able to attribute mental states to other individuals (sometimes referred to as “theory of mind”).

To date, only a relatively few animals have passed the MSR test, including certain primates, dolphins, elephants, and, as we saw in a prior post, magpies.

But is the test itself biased? We humans rely heavily on our eyesight and may naturally – anthropocentrically – have settled on a test that is based on visual interpretation.

What about animals who rely more on their sense of smell – dogs, for instance? Well, Marc Bekoff, Professor Emeritus of Ecology and Evolutionary Biology at the University of Colorado, Boulder, wondered about this too.

The Yellow Snow Test

Over a five year period, Bekoff performed a study1 in which he diligently tracked the behavior of his own dog, Jethro, when Jethro encountered clumps of snow saturated with his own and other dogs’ urine (“yellow snow”) while walking freely along a bicycle path in Colorado on winter mornings.

Snow pile, snow pile, on the ground, who’s the finest smelling hound? (photo: Walter Jeffries)

Bekoff would wait until Jethro (a neutered male German Shepherd and Rottweiler mix) or other known female and male dogs urinated on snow, and then scoop up the clump of yellow snow as soon as Jethro was elsewhere and did not see him pick it up or move it (Bekoff used clean gloves each time and took other precautions to minimize odor and visual cues). Bekoff then moved the yellow snow varying distances down the path so that Jethro would run across the displaced urine: (i) within about 10 seconds, (ii) between 10-120 seconds later, or (iii) between 120-300 seconds later. After Jethro arrived, Bekoff recorded how long he sniffed at the yellow snow, whether he urinated over it using the typical male raised-leg posture, and whether urination immediately followed the sniffing (“scent marking”).

After compiling and statistically analyzing the data, Bekoff found that Jethro paid significantly less attention to his own displaced urine than he did to the displaced urine of other dogs. For example, when Jethro encountered the yellow snow within 10 seconds, he sniffed for longer than 3 seconds only about 10% of the time when it was his own urine, compared to over 80% of the time when it was other dogs’ urine. (Jethro did tend to have longer sniffs at his own urine when he arrived after more than 10 seconds, but in all scenarios he still sniffed significantly longer at the other dogs’ urine.) Likewise, he very rarely urinated over or scent-marked his own yellow snow, but frequently did so with the yellow snow of other dogs, particularly other males.

The following table summarizes the data collected (note that the reference to “120-150s” in the Arrives column appears to be erroneous, and should instead read “120-300s”):

In sum, Jethro’s behavior clearly demonstrates that he was able to discriminate the scent of his own urine from that of other dogs. Of course this is just one set of tests on one dog, but would it surprise anyone if other dogs showed similar abilities?

Is there a fundamental difference between an animal recognizing its own image in a mirror and one recognizing its own scent in yellow snow? There certainly are different cognitive process involved (Bekoff himself has suggested that the yellow snow test may be more indicative of a sense of “mine-ness” in dogs than of a sense of “I-ness”2).

At a minimum, though, the yellow snow test stands as a useful warning that we humans need to be careful not to make quick judgments about animal intelligence or cognitive capacity (or lack thereof) based on tests that are well-suited to humans, but that fail to match the skills and abilities of the particular animal.


1Bekoff, M. (2001). Observations of scent-marking and discriminating self from others by a domestic dog (Canis familiaris): tales of displaced yellow snow Behavioural Processes, 55 (2), 75-79 DOI: 10.1016/S0376-6357(01)00142-5.

2Bekoff, M. Considering Animals—Not “Higher” Primates. Zygon 38, 229-245 (2003).

A Reflection of Intelligence?

What is it that makes us most “human,” distinguishing us from other animals?

One common response is that we have a sense of self, an ability to recognize ourselves as being separate and distinct from other individuals. Thus, when I look in the mirror, I know that I am looking at myself and not a dashingly handsome stranger. You know this as well; when you gaze in the mirror, you realize that the incredibly attractive person peering back is you!

How about other animals? They don’t “get” mirrors, do they? When they look at mirrors, don’t they either stare blankly or, at best, act as if they have seen another animal? (Hey, what’s that other beast doing on my turf? I wonder if it’s friendly….) Right?


Increasingly, we are finding that the answer is that other animals know exactly who is staring back at them.

First, we found out that certain great apes, like chimpanzees, can recognize themselves in mirrors. Ok, we all know that primates are smart; I’ll buy that. Then, it was bottlenose dolphins. All right, Flipper was pretty darned smart, plus dolphins have those big melon-shaped heads. I suppose that makes sense. Next, Asian elephants. Really? That’s sounds a bit odd. They do have those big eyes, but still…. I guess if you say so. Most recently, birds. Hey, now, wait a minute!

That’s right, magpies are the newest – and only non-mammal – member of the mirror self-recognition club.

Magpies with colored stickers recognize themselves in the mirror (photo credit: Helmut Prior, Goethe University)

As published in PLoS Biology1, researcher Helmut Prior and his colleagues affixed a red, yellow or black mark to feathers on the throats of five magpies (the black marks were basically a “control”: since they were the same color as the surrounding feathers they were essentially invisible to the magpies, thereby allowing the researchers to see whether any magpie behavior during the tests was the result of feeling, rather than seeing, the marks). The colored spots were positioned so that they could not be seen by the magpies unless they were looking in a mirror. When the researchers added a mirror to the cage, certain of the magpies noticed the colored spots in the mirrors and displayed “mark-directed” behavior, swiping at the marks with their beaks or scratching at them with their feet, and then checking in the mirror to determine whether they had successfully removed them. The magpies did not attempt to remove the black spots, which they couldn’t see in the mirror. Here’s a YouTube video of one of the magpies during the testing:

Also, you can read nice summaries of the research and mirror self-recognition testing in ScienceNOW2 and NewScientist3 online magazines.

This research is particularly notable given the differences between the neural anatomy of birds and mammals. Science Daily4 describes the significance:

These findings not only indicate that non-mammalian species can engage in self-recognition behaviour, but they also show that self-recognition can occur in species without a neocortex. This area is thought to be crucial to self-recognition in mammals, and its absence in this case suggests that higher cognitive skills can develop independently along separate evolutionary lines.

Mammals and birds have developed vastly different brain structures, and future studies will be able to further examine how these structures converge to produce similar cognitive abilities.

So, the magpie, lacking a neorcortex area in its brain and with an evolutionary history that diverged from ours 300 million years ago, shares our ability to look into a mirror and see itself.

The point here is not that birds can think like humans – they undoubtedly think like birds. Rather, the lesson is that we need to be very careful in labeling ourselves as special, as having exclusive abilities and intellectual talents. The more we study other animals, the more have found – and the more we will continue to find – how much we have in common, how much we share. Increasingly, I think we will find that our claims regarding uniquely human abilities are just not true, that they are simply smoke and mirrors.


1Prior, H., Schwarz, A., & Güntürkün, O. (2008). Mirror-Induced Behavior in the Magpie (Pica pica): Evidence of Self-Recognition PLoS Biology, 6 (8) DOI: 10.1371/journal.pbio.0060202.

2ScienceNOW, “The Magpie in the Mirror,” August 19, 2008.

3NewScientist, “Mirror test shows magpies aren’t so bird-brained,” August 19, 2008.

4ScienceDaily, “Mirror Self-Recognition In Magpie Birds,” August 19, 2008.

%d bloggers like this: