The mirror test and cats


Categories: Personal GeekStuff

One of the standard tests for self-awareness is “the mirror test” – put dye on a creature in a place it can’t see but could see in a mirror, give it a mirror, see whether it investigates the dyed spot. Cats, in general, fail this.

But I’ve found a test they pass using a mirror. Well, that one cat passes.

My cat is consistently surprised if touched from behind. Touch the back of her head, and at the very least she twitches her ears.

Do it while she’s facing a mirror, and she is completely unconcerned.

I hypothesize that she can tell that the thing in the mirror is me-holding-her, and that this means that she knows I am about to touch her, thus, isn’t surprised. Further science is called for. I require a large number of cats, some mirrors, and an NSF grant.

Comments [archived]

From: J.T.
Date: 2013-11-20 13:52:22 -0600


I have to chime in on this one. We have a few cats, and two of them (brother and sister, interestingly enough), had some interesting results when we tried a quasi-mirror test: When they were looking in the mirror, we waved at them from behind, so the only way they could see us waving was in the mirror. Both of them, instead of thinking the movement was coming from the mirror, immediately turned around and stared us, recognizing that we were the ones waving, and not some mirror-based versions of us.

From: Charles Walker
Date: 2013-05-10 09:02:27 -0500

Perhaps the difference in outcome is because touching your cat is a movement, which she can see in the mirror and expect; while the spot on fur is background, since it doesn’t move. This would be consistent with the tendency of hunting animals to track motion against a largely ignored background.

It also implies that the mirror test is not a good test for self-awareness. It is too human-centric. For the test to be valid, an investigator needs to consider the different ways animal sensoria work. This is more or less the same problem with IQ tests – which are loaded in favor of one particular human worldview, the one most of the investigators are most familiar with.

Just imagine the effects of an action-oriented sense of sight on the self, rather than our object-oriented sense. The less vivid color sense of many predators would make sense – during WW2 it was noticed that many camouflage patterns didn’t conceal in black and white photographs but did in full color.

Now about the NSF grant, would you accept a secondary investigator? (I have three cats, and purely in the Interests of Science (ahem!), I could get more.)

From: northierthanthou
Date: 2013-06-27 19:22:54 -0500

I contend the cats just won’t give us the satisfaction!

From: Silver
Date: 2014-01-22 19:58:00 -0600

This is amazing! I never considered how human-centric in a way the mirror test is! Since we would be interested in investigating a dot on our body, we assume that an animal would be too.

This reminds me of a quote from the documentary The Cove: “We keep spending billions of dollars sending signals up into the sky when we have a species here that can conceivably be more intelligent than we are.

It sometimes amazes me that the only language which has been extensively taught to dolphins is a version of American sign language, in which people use their hands to give messages to dolphins. But dolphins don’t have hands, so this is inherently a one-way process. It’s this anthrepomorphic “we have something to teach them, or to control them. When perhaps we ought to be looking at what they can give to us.

I wonder if living in an environment with mirrors (or anything that can produce a good reflection) can make an animal more self-aware.