A Quick Thought on Perception

 Posted by on 10 December 2013 at 10:00 am  Epistemology, Perception, Philosophy
Dec 102013

David Smith tweeted: “Mind = blown. These two blocks are exactly the same shade of grey. Hold your finger over the seam and check.”

I’d like to do some more thinking on perceptual illusions. I don’t think that the grays look different due to any conceptual inference. The grays look very different, until the seam is covered, and then they look the same.

Yet I don’t think that these are “perceptual errors.” Rather, this is exactly how our perceptual system is supposed to work, perhaps because such mechanisms enable us to properly judge shades and depth in the real world. However, particular with computer images, we can reveal these oddities and limits in our perceptual systems in a stark way.

Notably, these kinds of cases are very different from many traditional illusions like a stick bent in water, which are a function of the medium of perception (i.e. air versus water). Still, I don’t think that they reveal that our senses aren’t reliable or valid: they just reveal, in yet another stark way, that the diaphanous model of perception is wrong.


  • James

    As far as I’ve been able to determine, it’s a function of evolution. Our minds evolved in a particular context–light comes from overhead, shading indicates the direction of light, shadows make colors look darker, etc. Because these were true in almost every human experience until extremely recently, our brains evolved to take certain short-cuts, which these illusions exploit. These types of illusions play on that. In this case, the lighter portion of the objects (as an aside, these look like streak plates used in geology classes) are at the top. This causes our minds to assume that there is a light sorce above the objects, and this causes our brains to assume that the brighter section is the “true” color. In your every-day life, this is true; if you see two streak plates, one light and one dark, the parts where the light hits will show the true color. The presence of shaddows, blue “sky”, and a “ground” in the illusion enhance this perception. This is exactly how our perceptions are supposed to work. Again, in the real world this would provide useful data. And those shortcuts are actually extremely useful. While our senses pick up on the data, our brains don’t process it all; they select the important bits and ignore the rest. A great example of this is watching TV: you can lose your perception of the room around you if you get drawn into the program. In the Pleistocene, these mechanisms allowed us to filter out the environment around us and focus on the important data (food and predators). Modern illusions subvert those mechanisms in creative and fun ways. What I can’t say is whether we evolved these shortcuts directly, or evolved the capacity to develop them–it could be that we develop these so early in childhood that we don’t notice that it’s an internalized process. If you want to see more illusions that play with these shortcut mechanisms in our brains, particularly ones with very practical results, look at remote sensing images, such as photos taken from Mars orbitors. If the Sun is shining on the south of a crater, you will see a dome, for example. In paleontology one of these shortcuts is explicitely and conciously exploited–standard practice is to have the light sorce in the upper left when photographing fossils, which allows us to present 3D data in a 2D format consistently. As an aside, I’ve spent a LOT of time looking at how vision works in terms of how we interpret the data. I had an eye that didn’t work properly, and it took years of physical therapy to see in 3D. So I internalized different methods for determining distance. The end result is that I can’t play sports (I couldn’t tell how far away the ball was), but I always see topographic maps as 3D images. Very useful when you’re in a desert with a topo map and a compass!

  • Adam


    I would dispute your use of the term ‘true colour’. How bright does the light have to be to reveal the ‘true colour’? Does this mean that 99.9% of colour identification is invalidated? I think that there is no ‘true’ colour; the same exact ‘colour’ is perceived in a different form depending on how much it is illuminated, but there is no difference in veracity.

    In addition, I am skeptical of your appeal to evolutionary psychology. It is equally possible that the phenomenon you have described is the result of a learned interpretation, and not of some hard wired neural mechanism.

    It appears that a lot of three dimensional perception is learned, and may well be based on the interpretation of differences in illumination. This would account for the apparent difference in the colours on the CGI cube.

    • James

      I agree that the exact color seen will depend on the lighting. What I’m saying is that your brain assumes that things in shadows are darker than they should be, while things in the light are more or less the color they’re supposed to be. Think about going outside on a sunny day–the grass in the sunlight doesn’t seem washed out, the grass in the shade appears darker. This was stated in the specific context of the illusion: when presented with both light and dark patches on an object, and clear indications of shading, you are going to percieve the darker patches as shaded and therefore darker than they “should” be (with your brain taking a short-cut to deterrmine what the color “should” be).

      I agree with your skepticism about my arguments–in fact, I presented it first. :) It may be a learned interpretation; my own experience tends to suggest it is, though other data I’ve seen tend to suggest it’s more a physical issue. I was merely speculating on the origins of the feature. It’s fundamentally irrelevant, however. Our eyes and brain adjust the sensory input in a number of ways prior to us becoming conciously aware of that data. That processing is what causes this sort of optical illusion. As an aside, from a geological perspetive the color something like a soil “should” be is the color it reflects when fully saturated (but only just) and in direct sunlight. Then you break out the Munsell charts…..

  • William H. Stoddard

    I don’t think it makes sense to say that the axiomatic status of sensory perception requires every single perception to be perfectly accurate. For comparison, memory appears to be axiomatic, in that it identifies a necessary basis of all conceptual knowledge: If we could not remember the past we could not integrate the past and the present, which we do every time we form a concept, nor could we have an evidential basis for any proposition that goes beyond the limits of perception. And yet I can forget things, or remember things incorrectly.

    The axiomatic character of perception lies in the fact that if we have misperceived something, we perceive it correctly, and identify and correct our error, by looking more closely—that is, by relying on perception. There is no higher standard to appeal to. And only perception can give us evidence of error in perception.

    Psychological experiments in illusion cut short that process, requiring people to make judgments under conditions where they cannot look more closely. That is, they interrupt the process of error correction over time.

  • http://twitter.com/Radian_Angle Tjitze de Boer

    Covering the seam didn’t work for a while but looking at it intently without doing so did make me see the similarity. I’ve just been listening to some lectures on epistemology, its a good show on how we see precepts instead of unrelated sensations.

Suffusion theme by Sayontan Sinha