Skip to main content

Crows take a look in the mirror

One question in an animal cognition is whether animals other than humans have the ability to recognise themselves. A classic way of testing this, established in 1970 by Gordon Gallup, is the `mirror test'.

This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American


One question in an animal cognition is whether animals other than humans have the ability to recognise themselves. A classic way of testing this, established in 1970 by Gordon Gallup, is the ‘mirror test’.

At first glance this might seem a rather straightforward test: get an animal to look in the mirror and see whether it seems to recognise itself, or responds as if it were looking at someone else. However, there are more factors to consider than are at first apparent. For example, whether the animal in question has ever experienced anything akin to mirrors before in its life will greatly impact on how it responds to a mirror. Humans which have been blind and then had sight restored to them, or children that have never encountered a mirror before do not recognise themselves when looking in one, and react as if they were seeing another person. However, before not too long people recognise that it is in fact themselves they are seeing.

Similarly, when chimpanzees first look in the mirror they react as if they were seeing another chimpanzee. However, they too then realise that it is in fact themselves they are seeing. Once they become familiar with the mirror they then use it to see parts of their body they have never seen before (yes, predictably the anal-genital area – and some people doubt that these are our closest living relatives). They also used the mirror for more respectable activities, like cleaning their eyes, nose, and teeth.


On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


To test whether chimps really recognise themselves in the mirorr, researchers first llowed chimps to become accustomed to their appearance in a mirror. They then drew a coloured mark on to the forehead of the chimp in question. When the chimps were not provided with a mirror, they did not touch the coloured-in area on their heads. However, when they were given mirror, they did start touching this area, indicating that they were aware how mirrors work.

The ‘mark test’ has now been carried out in a variety of animals, and it seems that a number of animals also have mirror self-recognition: other great apes (bonobos, orang-utans and gorillas), bottle-nose dolphins, an Asian Elephant, and two Eurasian magpies. Animals that do not seem to be able to recognise themselves in a mirror include some monkey species, fish, gibbons, sea lions, dogs, cats and parrots.

Whether the fact that an animal can recognise itself in a mirror means that it has self-recognition akin to humans in something which has been long debated, and cannot be answered by the ‘mark test’. However, a number of animals which have not been able to recognise themselves, have nonetheless showed interesting and varied results when looking at other things in the mirror. Other mirror tasks apart from the ‘mark test’ include ones where animals have to use the mirror to find a hidden reward. In order to do this, they potentially need to understand how the object in the mirror is a reflection of an object in the real word.

In a recent study, scientists used this kind of task to see whether ten New Caledonian crows were able to use mirrors. First, crows were given a mirror which they were encouraged to interact with by leaving food (cubes of meat) next to it. They were exposed to the mirror for 10 minutes a day, for up to six days. The crows were able to walk behind the mirror to see if there was another individual there. At first, as the scientists predicted, when the crows first saw themselves in the mirror they reacted aggressively, as if to another individual. This aggressive behaviour did not decrease over the series of ten trials of which the crows were tested, and the crows did not show any behaviour indicating that they worked out that they were looking at themselves. However, this was a relatively short set of trials, and it could be that if they were given longer to experience mirrors they would have learned how they work.

After the initial aggressive response, the crow would then frequently examine the back of the mirror, and search behind the mirror, as if for the ‘other’ individual. This is also what children and non-human primates have been found to do. This exploration increased the more time they spent with the mirror.

Two crows who had this mirror experience, and two new crows who had not experienced mirrors were then tested on a task where they had to use a mirror to retrieve food. At the start of this task, birds were given a mirror placed on the floor, with meat off it (which they readily ate). A perch was then placed above the mirror (so that they could sit on the perch and look down into the mirror). The birds were then given meat to eat off the mirror and off of the perch, so they could get used to the set-up. The researchers then hung the meat underneath the perch on bits of string, and the birds learned to lean off the perch to retrieve the food hanging on the string, which hung above the mirror on the ground. Finally, two boxes were placed on top of the mirror, below the bird’s perch. The way that these boxes were built meant that they could be used to obstruct the crow’s view of the food. At first, the crow could see both the meat hanging below the perch, and its reflection in the mirror. However, over time the researchers altered the boxes so that the birds could no longer see the real meat, and instead could only see which box it was in if they used the reflection of the meat.

After training with this set-up, the birds were then tested with a set-up where they had four boxes, only one of which contained hidden food. They were given 30 trials to ‘solve the task’ (find the hidden food using only the reflection to guide them). Although it was not immediate, all four of the birds were able to learn to use the mirrors to find the hidden meat, becoming better at doing so as the trials went on. Unexpectedly, the birds which had not had previous experience with mirrors actually solved the task faster than the two which had.

So, does this study show that these crows understand mirrors? Well, probably not. Given the number of errors birds made in the final test before learning to find the food, it seems more likely that the birds learned a ‘rule’, such as ‘moving towards the mirror image of food and searching there leads to real food.’ However, this is not to say that with more experience the crows may come to learn how mirrors reflect objects in the real world.

Photo credits: crow by Marko_K, chimpanzee photo from Povinelli et al. (1993).

References:

Medina, F. S., et al., (2011) New Caledonian crows’ responses to mirrors, Animal Behaviour, doi:10.1016/j.anbehav.2011.07.033

Pearce, J.M. (1997) Animal learning and cognition: An introduction.Psychology Pr.

Povinelli, D.J., Rulf, A.B., Landau, K.R. & Bierschwale, D.T. (1993) Self-recognition in chimpanzees Pan troglodytes: Distribution, ontogeny, and patterns of emergence. Journal of Comparative Psychology 107(4): 347.