Do animals see optical illusions?

February 22, 2015 • 2:00 pm

This gif strongly suggests that at least cats do:

uE9ihTO

This page will show you a larger version of the “wheel” illusion that the cat is inspecting (and attacking). I think it’s very clever to present such images to animals who can show us directly (as hunting cats can) whether their senses are fooled. This moggie was clearly duped!

That’s not all that surprising, of course, as we naturally think that other mammals’ eyes work the same way as ours. But this shows more than that—it shows that other mammals’ brains work the same way as ours (at least insofar as the illusion works), for what is taken in by the eye is interpreted by the brain. This cat sees a nonmoving image and, like us, interprets it as being in motion.

Optical illusions are a byproduct of the evolved interaction between vision and its neural interpretation, and many such illusions—like the “checker-shadow” illusion, perhaps the best of all optical illusions— reflect our evolved ability to compensate for natural phenomena. Greg Mayer has written about how countershading in animals takes advantage of predator’s evolved tendency to be fooled about the hue of an object. Likewise the checker-shadow illusion fools us because our brains evolved to imagine things in shadow to be lighter than they really are.

Of course many of the selective pressures that molded our own interpretation of what we see must be similar in animals. Predators and humans are both fooled by countershading, and I bet a mammalian predator’s brain would also be fooled by the checkershadow illusion. (Actually, that could be tested, at least in birds, by training them to peck at squares of a certain hue and then giving them the illusion.) Or perhaps we’ve simply inherited the eye-brain connection that is fooled by illusions from our ancestors in which that visual interpretation was critical for survival, and we’re the victims of “evolutionary inertia.” Regardless, the cat above shows that we’re not the only species fooled by these two-dimensional tricks.

h/t: John S.

31 thoughts on “Do animals see optical illusions?”

  1. The kitty does seem to be reacting to the illusion of movement that would be presented by this pattern.
    I wonder if animals like cats, dogs, and other primates react to pareidolia? It seems to me that they should.

  3. This may sound a bit southpark-ish, but you know, I think I’ve learned something today. I can watch this gif all night (it’s half past 9 here). It’s very interesting to think about! I wonder if animals can suffer from mental illness too.

    1. Yes. Traumatic stress, ADHD, obsessive compulsive disorder, anxieties, phobias, depression, addiction…. quite a lot of disorders.

      1. I don’t have pets, but I’ve always thought it was possible. How are these illnesses diagnosed? A dog can’t fill in a form.

        1. Birds pull out all their feathers, dogs may act aggressively for no apparent reason or mope for no apparent reason. There are a few different ways pets react. If you live with pets (especially cats and dogs), you understand that they express happiness and sadness in ways easily understood by humans.

    2. Pinning down what constitutes “Mental illness” is tough enough in people & of course social norms change…

      That said

      According to the NYT at least 5% of US military dogs develop PTSD symptoms

      To this layman I’ve come across what seems like an almost permanent anxiety in some pet dogs. It feels to me as if dogs have as rich an emotional life as people & I would assume they can develop hangups as much as people. Cats I can’t comment on – can’t figure ’em out…

      1. Cats are not pack animals. Perhaps that’s why they’re more resistant to loneliness and emotion.

        1. Loneliness they may not feel as much, but being the descendants of wide-ranging territorial animals, I’d expect they are particularly susceptible to confinement-related problems.

  4. I would like to see chimps tested for optical illusions. Since they are much closer to us than cats, I would expect the effect to be pronounced. It does not seem to me it would be difficult to do such tests.

    1. I think I saw somewhere footage of a magician doing some sort of sleight of hand trick to a chimpanzee. I do not recall the detail but it had something to do with making the animal expect to find a treat in a certain spot and it was in a different spot. They were definitely puzzled as if they knew what to to expect to happen and it did not happen that way.
      This is of course not at all surprising.

  5. Jerry,

    “Likewise the checker-shadow illusion fools us because our brains evolved to imagine things in shadow to be lighter than they really are”

    I don’t think that is the correct explanation.
    Square B looks lighter than it actually is because it is surrounded by dark squares, and square A looks darker than it really is because it is surrounded by light squares.
    The purpose of the “shadow” in that picture is to ensure that square A and square B are actually the same.

  8. Funny … I went to the larger image and didn’t see any movement at all.

    Then again, my cats always did say they were superior to me.

  9. I can never find the hidden object or name the species.
    But I do know that these are not rotating wheels but moving snakes. You can even see the forked tongues.

  10. I don’t see why they shouldn’t (no pun intended). Their optical engines in the brain probably work just about the same way ours does.

  11. Could someone let me know what the wheels are supposed to be doing? I looked at the large view of the illusion, and they don’t seem to move at all.

    Are they supposed to be spinning? How quickly? Uh-oh, is there something wrong with my evolved interaction between vision and its neural interpretation?

    1. You may want to click on the link provided. I don’t see the effect in the kitten video, either, though I see it in other presentations of the same pattern.

      1. I did click on the link, but the larger image looks completely still, too.

        What does the effect look like?

  12. Likewise the checker-shadow illusion fools us because our brains evolved to imagine things in shadow to be lighter than they really are.

    I still think it’s a mischaracterization of what’s going on in our brains to say that we’re being “fooled” by the checker-shadow “illusion”.

    Our eyes are not meant to be accurate spectrometers or photometers. There’s little or no adaptive value in knowing the precise wavelengths and intensity of light striking our retinas. What we need is information about the physical objects in our visual field. We therefore care more about an object’s intrinsic color (determined by its reflective properties) than about the wavelength of light striking it.

    On that view, being able to distinguish dark squares from light despite differences in illumination is not a case of being fooled by an illusion; it’s a demonstration of our ability to correctly extract accurate information about object colors under variable lighting conditions.

    So the only sense in which the checker-shadow effect is an illusion is that it represents a 3D scene (with shadows) as a 2D array of pixels. Readers’ wildlife photos also fit that description, but nobody calls them illusions.

    1. There is so much wrong with your post that it is hard to know where to start. Perhaps I’ll just start at the end…

      That the checker shadow illusion is a 2D array of pixels that represents a 3D scene is irrelevant. The demonstration would work equally as well with a white disc superimposed on a larger black disc next to which you have a black disc superimposed on a larger white disc. You can manipulate the shades of he various components so that the two smaller discs are actually the same shade but look entirely different. That’s the purpose of the “shadow” in the checker shade illusion.

      1. If the two smaller disks are identical, then it’s incorrect to describe them as white-on-black and black-on-white. What you actually have is gray-on-black and gray-on-white.

        I grant that in this case the brain has difficulty perceiving the colors of the inner disks accurately, and I’d be comfortable calling that an illusion, because in this abstract 2D composition there’s no deeper layer of information to be extracted, and it’s an error for the brain to try to extract information that isn’t there.

        But in the checker-shadow scene there is a deeper layer of 3D shapes and colors beyond the 2D pixel values, and it therefore seems perverse to call it an illusion when our brain successfully extracts accurate information from that deeper layer. That is, again, what our visual system is evolved to do. It may be the same brain mechanism as in the disk illusion, but in this case it’s working properly on the sort of input it’s optimized for. The fact that it says “light” on some gray pixels and “dark” on others isn’t a bug; it’s a feature.

        1. GK: “If the two smaller disks are identical, then it’s incorrect to describe them as white-on-black and black-on-white. What you actually have is gray-on-black and gray-on-white.”

          You must have missed this bit:

          BJ: “You can manipulate the shades of he various components so that the two smaller discs are actually the same shade but look entirely different”

          GK: “But in the checker-shadow scene there is a deeper layer of 3D shapes and colors beyond the 2D pixel values, and it therefore seems perverse to call it an illusion when our brain successfully extracts accurate information from that deeper layer”

          In fact, the “checker shadow scene” does not have a deeper layer of 3D shapes. It is simply a 2D array of pixels. In any case, your brain tells you those squares in that 2D array of pixels are different. But they are, in fact, identical. This is the very definition of an illusion.

          1. But there are no squares in the 2D array of pixels. The gridlines intersect at oblique angles to form rhombuses. The only reason to call them squares is because you recognize that it’s not simply a 2D array of pixels; it’s a projection of a (possibly imaginary) 3D scene in which the gridlines do form squares. The 2D rhombuses may be identical, but the 3D squares are different, and that’s what my brain is telling me.

            You can’t have it both ways. Either the squares are different (even if their projections are the same), or there are no squares at all, just meaningless pixels whose colors don’t matter.

          2. You are being obtuse now.
            Of course they,re not squares. They’re representations of squares. The 2D array of pixels represents a 3D object. The point is that it is not an actual 3D object. It’s just a 2D array of pixels.
            But you have lost sight of the argument…that this really is an optical illusion (in fact it is the best example of an optical illusion ever produced in my opinion) This is what you challenge and, so far, have failed to justify.
            The rest Of your original post is also full of errors but we’re obviously never going to get around to exposing those at this rate.

          4. There is a video of a 3D version of the checkerboard illusion which is equally impressive and which clearly demonstrates my points about the scene having to be manipulated to produce the effect.
            In other words, the 2D array of pixels does not represent an actual checkerboard and shadow. In other words, you cannot set up a real checkerboard and shadow to produce the effect, you have to use a checkerboard on which the colours have been altered.
            So, not only have you been fooled into thinking the “squares” are different in “colour”, you’ve also been fooled into thinking this is an actual checkerboard!

Comments are closed.