I've been looking at this Ted talk on interspecies internet, and it appears that they have selected a small number of species of animals (dolphin, elephant, and some ape(?)) and are intent on developing interfaces for these animals to be able to connect with other animals over the internet.

From the talk, it appeared as if one of the criteria for animal selection was whether or not an animal recognizes itself in the mirror. This got me thinking - there are many more cats that have access to technology like the internet or TV or cats that have been interacting with touch screen displays. Yet apparently cats cant recognize their own reflection as themselves.

What is so special about being able to recognize one's own reflection? Is it a true measure of self awareness? Does it only appear in animals with larger brains?


3 Answers 3


I think it is a necessary test of self-awareness. In order to make the connection that "I am the being in the mirror" and that every action I do will be reflected in the mirror (or, vice versa, every action I see the being in the mirror take is a reflection of my own action), one must first have a concept of self. One requires a constant present-tense (at least) story of oneself: "I am doing this. Now I am doing that. Oh look, I am doing this and that guy is doing the exact same thing." And then there's the ultimate realization, "Ah ha! That's me!". That realization would make no sense if the individual has no sense of "me"!

[speculation] On the other hand, perhaps it is too strict. It seems that it would unfairly favor social species, since it requires not only recognizing another member of the species (and being able to associate the anatomy of another member of the species with one's own anatomy), but it may also require the recognition of intent. That latter requirement is complicated by differentiating between intent and compulsion: at some point, one must realize that the being in the mirror isn't trying to copy but rather it always, inevitibly reflects one's motions. So, perhaps a cat is self-aware, in a sense that it has a constant story in its head of what it is doing and what it intends to do, but it cannot easily ascribe intent to other members of the species and thus cannot figure out that the cat in the mirror is automatically mirroring all of its movements. [/speculation]

Herein lies the rub: there is no perfect test for self-awareness; we can cast doubt on any outcome. The best would be if these animals could communicate with us by language. In lieu of that, we have to think of such tests that demonstrate with reasonable confidence that the species carries all of the characteristics that we attribute to self-awareness. I think another good test would be determining whether the species can plan individual actions in the future; in order to say, "first I will do this, then that; finally, that will allow me to do that", it seems like it would reasonably require an understanding of self and the consequences of one's actions. An example would be birds solving sequential puzzles.


The mirror test is definitely useful, but clearly biased toward animals with social intelligence (a notion of self is only useful if there's a notion of other) and also toward highly visual animals. A mole-rat, no matter how intelligent, is unlikely to use a mirror for much of anything. Using the same test for animals with very different sensory modalities is bound to fail. For example, see the comments of the following YouTube video. The user Sluchesi notes:

sticking a piece of post-it paper on my tom cat's head where he couldn't see it without a mirror and while petting him, waited until I was sure he didn't notice it (he didn't respond to it), showed him the mirror and he instantly started trying to get the paper off his head (not the image in the mirror, his own), shaking it and using his paws.

A second user (M. Ransom) did the same test and verified it. Not quite to the level of scientific proof that you would want to see, but certainly worth a follow-up study by someone who focuses on the topic. If true, it would indicate that cats probably do have a conception that the reflection in the mirror is them, but place little value on introspection or visual self-concepts. This might indicate that what cats (and many other animals) lack is not the ability to have a model of self, but to have a model of some other being modeling yourself (i.e., 2nd-order model of self). This is actually a quite complicated level of modeling: even humans are quite bad at guessing what other people think about them. However, basically all humans try to do it, even if their guess is often mirror-bias (i.e., other people think about them like they think about themselves). Given that this is nearly at the limit of average human cognition (3rd order models of other/self are rare and usually limited to strategic games like poker), it seems reasonable that most animals lack this ability.

Having a model of another's conception of oneself is necessary for a being to care what they look like, for instance. After all, if you can't comprehend that the world attaches importance to how you look, you would be unlikely to develop mechanisms to attach importance to it either. In that second case, the mirror test may be accurate (i.e., it measures higher order thought), but it is in fact measuring something very different (and more advanced) than what it has been assumed to measure.

  • $\begingroup$ The sticky note on head is a good experiment, I will try it! $\endgroup$
    – Alex Stone
    Dec 1, 2014 at 13:37
  • 1
    $\begingroup$ An interesting anecdote from the book "Merle's Door". The dog didn't at first appear to recognize itself in small mirrors, but when the human took the dog to an exercise studio with large wall mirrors, the dog saw himself and his human together, and from then on appeared to understand mirrors. He just needed that extra bit of social context... $\endgroup$
    – jamesqf
    Aug 4, 2017 at 19:36

To the answer above: "I am the being in the mirror", one has to ask themselves: Is it so strange that a cat seeing it's own paw in front of it and recognizing it as it's his OWN, and not chasing it's own tail because it knows it's his and he's not a kitten anymore?

So how much different is that when they come to understand, especially after you hold the cat several times in front of the mirror with you, making the association and he's no longer weirded out? Many cats making this transition look in curiosity and awe and move on, never really caring that much about it -- much like kids.

Bring a cat into a bathroom with a mirrored wall behind the sinks, when he's dealing with something where the mirror is in his line of sight. Walk in behind. When the cat turns around to look at you to play instantly, instead of trying to play with the mirror -- it recognizes that it shows whats behind him. I have observed this.

If a cat recognizing it's own tail out of it's "perif" is nothing special, neither is recognizing that it's him in the mirror. He may not be cognitively advanced to think deeper than "Yep, that's me, I have THIS look. Oh, am I getting fat?" -- they care less about that.

A good test would be to have a cat with whom you can put something on his head and he doesn't care -- unless it's something he HATES by its LOOK. Once you've established that over and over 100%, and have been able to put it on him without him caring without him SEEING it -- put him in front of the mirror and see if he tries to take it off.

A red dot isn't going to do squat. They don't give a shoot about things like that. If you want to know whether they know IT'S THEM, put them in position where there's something they DON'T WANT on them -- and they try to get it off... where it's something they don't mind on them, they don't try to get it off after seeing themselves in the mirror.

Not fool-proof, though. They may not mind it on them because it's purely how it looks. They can just walk away from the mirror and they don't have to worry about it.

But back to the original point: I don't think it's a big deal that they can recognize that it's themselves. To UTILIZE that for deeper thinking in making connections is another matter.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .