Evolution does not select for veridical perception

The title is from a blog post by Brian Leiter. Leiter links to an article in the LA Review of books: Imitation and Extinction: The Case Against Reality. The article is written by Donald Hoffman.

We have discussed the general topic before, in several threads. So maybe this is a good time to revisit the topic.

Hoffman asks: “I see a green pear. Does the shape and color that I experience match the true shape and color of the real pear?”

My take is that there is no such thing as the “true shape and color of the pear.”

It is a common presumption, that there is an external standard of truth. Here, I mean “external to humans”. Truth is presumed to come from somewhere else. And our perceptual systems evolved to present us with what is true.

As I see it, this is backwards. Yes, our perceptions are mostly true. But this is not because perception is based on truth. Rather, it is because our human ideas of truth are based on what we perceive.

Open for discussion.

424 thoughts on “Evolution does not select for veridical perception

  1. BruceS: The philosopher Michael Tye

    For those who might care, Tye has published at least two papers on orgasms.

    And, IIRC, the acronym for his theory of mind is PANIC

  2. BruceS: Have you asked the current publisher to allow you to distribute limited copies of the book electronically? I’m assuming the current publisher owns the copyright.

    I have the copyright, but, well, you know.

    If you want a break from your onerous reading schedule, does this short paper do justice to the issue you raised in your previous note on ideal languages etc:
    Wittgenstein On Russell’s Theory Of Types
    https://projecteuclid.org/download/pdf_1/euclid.ndjfl/1093891616

    I don’t know that article (or Davant at all). But one of the classic papers on this matter that I DO know is this one:

    “Objects, Properties, and Relations in the Tractatus”
    Author(s): Irving M. Copi
    Source: Mind, Vol. 67, No. 266 (Apr., 1958), pp. 145-165

    I am pretty sure you would have seen the following, but just in case:
    https://bostonreview.net/philosophy-religion/katrina-forrester-future-political-philosophy

    No, I hadn’t seen that. I’ll take a look. Thanks!

  3. Neil Rickert: I don’t see any, either.

    Light gray ghost images in the white spaces at the corners of the squares. The disappear if you try to focus on them.

  4. Kantian Naturalist:
    This is slightly off-topic from whether selection would favor veridical perceptual processes, but relevant to the more general question of how we might think about consciousness in naturalistic terms: The Consciousness Illusion. Frankish makes a compelling argument for illusionism: our introspective awareness of phenomenal properties does not correspond to any objective features of the world or our brains.

    That’s better than Dennett’s brand of illusionism.

    He is too entranced by computer ideas, and is making things far too complex.

  5. Neil Rickert,

    I have a distorted macula, and the image in one eye is larger than in the other, because the cells are stretched a bit.

    Interestingly, my brain is able to fuse the images 99 percent of the time. I sometimes see double images at night, when the context is removed from text signs, or points of light.

    Back to the point, I have trouble seeing 3D illusions. I require context to fuse 3D images.

    I’m never going to agree with keiths about calling implied structures illusions. The little grey cells are illusions. As are Benham colors.

  6. BruceS: The philosopher Michael Tye claims bees are conscious but caterpillars are not.

    Which doesn’t make a lot of sense, since both are insects. He bases this on the fact that learning behaviour has been demonstrated in bees but not in caterpillars, I see. A little googling rapidly shows that the claim that caterpillars are incapable of learning is false, at least for the larvae of hawk moths. The learned behaviour even seems to carry over to the adult stage.

  7. Corneel,

    Hmm:
    Training caterpillars

    Using some mild electric shocks, Douglas Blackiston from Georgetown University trained hookworm caterpillars (Manduca sexta) to avoid the scent of a simple organic chemical – ethyl acetate. The larvae were then placed in the bottom end of a Y-shaped tube, with the scent of ethyl acetate wafting down one arm and fresh air coming down the other. Sure enough, 78% of the trained caterpillars inched down the odour-free arm.

  8. Neil Rickert: Even for us, I suspect that what we see as separate from the environment has a significant social component. That is to say, there is peer pressure for us to see things roughly the same ways as others in our society.

    Perhaps. What I opposed to was your suggestion that humans conventions are all there is and that there are no true structures in the world.

  9. phoodoo: I think the definition of consciousness is you have to be able to laugh at something funny.

    I think that disqualifies ants.

    and cats.

  10. Corneel: The learned behaviour even seems to carry over to the adult stage.

    Utterly sceptical about this. Where do insects go to learn?

  11. Alan Fox: I mean if you had the choice of fresh air or noxious chemical, which would you choose?

    You’d be surprised how many people choose the latter.

  12. Corneel: What I opposed to was your suggestion that humans conventions are all there is and that there are no true structures in the world.

    You missed the subtlety.

    I fully agree that there is a reality independent of us. And I agree that there are things that we call physical structures that exist independently of us. However, the very notion of structure comes from us.

    To say that human conventions are all there is to structure, is not to deny that there is a reality that appears to us as structured. But remove our abstract ideas about structure, and then that reality just exists. The concept of structure is what comes from us. The parts of the world that we consider to be structure may not come from us (except for the ones that we constructed). But the concept of structure comes from us.

  13. Alan Fox: Utterly sceptical about this. Where do insects go to learn?

    Those are standard tests using Y-shaped (or T–shaped) decision mazes. They are routinely used in (among other things) Drosophila research, and yes, you can train insects to associate smells with unpleasant consequences.

    Here is a nice video explaining the procedure
    https://youtu.be/-dPfZE5adYg

  14. Neil Rickert,

    But can animals share some of those “conventions”?

    Whether that speaks of a real underlying structure to the world is not something we are going to resolve, I fear.

  15. Neil is making some progress.

    Earlier he was denying the existence of structure:

    Structure is abstract. That should already tell you that it is not real (unless you are a platonist).

    Now he acknowledges its reality:

    I fully agree that there is a reality independent of us. And I agree that there are things that we call physical structures that exist independently of us.

    Next step: acknowledging that there are external truths about those independent features of reality we call ‘structures’.

  16. petrushka,

    I’m never going to agree with keiths about calling implied structures illusions. The little grey cells are illusions. As are Benham colors.

    Suit yourself, but the standard term remains “illusory contours”. And for good reason. They are illusory in the same sense that the little gray balls are illusory — they are absent from the stimulus but present in the percept.

  17. Corneel: Those are standard tests using Y-shaped (or T–shaped) decision mazes. They are routinely used in (among other things) Drosophilaresearch, and yes, you can train insects to associate smells with unpleasant consequences.

    Here is a nice video explaining the procedure
    https://youtu.be/-dPfZE5adYg

    Convincing. Though I did not notice any mention of the control, where the flies response to chemical odour in the absence of shocks was different.

  18. Alan Fox: Convincing. Though I did not notice any mention of the control, where the flies response to chemical odour in the absence of shocks was different.

    Or the reciprocal experiment where they swapped the odors, or the other negative controls where you don’t supply any odors but do give the stimulus, or where you supply neither.

    That is for the sake of simplicity in the instructional video, is my guess. Any researcher worth her salt will perform all the appropriate control experiments.

  19. Corneel: Which doesn’t make a lot of sense, since both are insects. He bases this on the fact that learning behaviour has been demonstrated in bees but not in caterpillars, I see.

    I don’t take that arbitrary boundary that seriously.

    You’d have to read more of his stuff to understand the details of his reasoning. He respects the science, but I gave up on reading him because he still is too much of an armchair reasoner for my tastes, at least in the stuff of his that I have read.

    ETA:
    Referring to our previous exchange on agents, this video just came up in my YT (YouTube) subscription to FXQi. It starts with an attempt to define an agent which is probably closer to Neils than mine; then it moves on to drawing some conclusions based on simulations using a simplified mathematical model. So it’s like Hoffman that way. But the presenter is not going to write any popularizations of her ideas, I suspect.

    She uses IIT (Integrated Information Theory) as part of her model. I don’t like IIT as a theory of consciousness because it is panpsychic but it seems she it not using that aspect of it.
    https://www.youtube.com/watch?v=f9wwcLBjrV0

  20. walto:
    Frankish, on twitter today:

    s=17″>https://twitter.com/keithfrankish/status/1177297339419111424?s=17

    Thanks, that’s helpful.

    The illusion in illusionism is that our introspected experience is constituted by a misrepresentation of brain states; specifically it represents them as having (pseudo-)phenomenal properties.

    So if perception is itself a representation built through sensorimotor processes, then our introspected experience is a (mis)representation of that representation. The “mis” being because of the (mis)attribution of phenomenal properties.

    Or so the idea seems to me.

    The are other essays in the JCS volume Frankish edited which try to deal with what such things could be and why brains would evolve to produce these types of misrepresentations.

    ETA: expanded for “clarity”

  21. Neil Rickert: That’s better than Dennett’s brand of illusionism.

    He is too entranced by computer ideas, and is making things far too complex

    Frankish accepts Dennett’s ideas, so they are trying to explain the same thing.

    ETA: It is true that, in the JCS essay at least, Frankish looks at variations of Dennett’s core idea.

    I agree Frankish does a better job of explaining illusionism than Dennett.

    Dennett comments on Frankish in the JCS issue on illusionism to agree that Frankish has captured Dennett’s concept correctly.
    https://ase.tufts.edu/cogstud/dennett/papers/illusionism.pdf

    I find that in general about Dennett; for me, his acolytes do a better job of explaining Dennett’s ideas than he does.

    ETA: clarity

  22. keiths: It’s interesting that you and Alan can’t see the black balls.

    Is that true for anyone else?

    I see them. But of course it is a double illusion. At least, according to Dennett.

  23. walto: I don’t know that article (or Davant at all). But one of the classic papers on this matter that I DO know is this one:

    “Objects, Properties, and Relations in the Tractatus”
    Author(s): Irving M. Copi

    Has roman numerals in its DOI, which was new to me:
    10.1093/mind/LXVII.266.145
    But scihub has it. The opening sentences are fun:
    [start quote]
    “Few recent works in philosophy compare with Wittgenstein’s Tractatus either in obscurity of style, range of influence, or depth of insight. Considering its difficulty, scandalously few studies of it have been published.”
    [end quote]

    I skimmed the rest; it seems much of it is intended as a reply to “Mrs. Daitz”.

    There also seems to be a lot of speculation about how we understand language; it struck me that the paper was written before Chomsky, although I don’t know if Copi would have considered his ideas relevant to the philosophy.

  24. petrushka to Neil Rickert,

    I have a distorted macula, and the image in one eye is larger than in the other, because the cells are stretched a bit.

    Interestingly, my brain is able to fuse the images 99 percent of the time. I sometimes see double images at night, when the context is removed from text signs, or points of light.

    Back to the point, I have trouble seeing 3D illusions. I require context to fuse 3D images.

    I’m never going to agree with keiths about calling implied structures illusions. The little grey cells are illusions. As are Benham colors.

    I am reluctant to call these illusions because they are natural effects of the way the eye deals with light. Goethe goes into this in detail in his Theory of Colours

    In the case of the black grid Goethe would say that the area of the retina on which the black falls is in a state of relaxation while the area of white is stimulated. The areas that have been stimulated take some time to recover in the same way that if you look at a bright light and then look at a blank wall the area where the light struck your retina will appear as a dark patch (I’m sure you all know not to look directly at the sun or you could do permanent damage to your eyes.) We see grey areas because the retina hasn’t fully recovered in these areas.

    The colours in the spinning Benham disc can also be explained in a Goethean sense. Goethe argued that colours were the effect of the interplay of light and darkness. Light moving in front of darkness gives an effect at the blue end of the spectrum and darkness moving in front of light gives an effect at the red end of the spectrum. So it is to be expected that the movement of black and white areas on the disc will produce coloured effects.

    The blue sky and red sunsets are also explained in this way.

  25. BruceS,

    I really like that paper and have often wished I could write as clearly/simply as Copi, who studied with Russell. One of his excellent and at-one-time widely used textbooks was used in the first logic class I took. (I wonder if they’re still used a lot: maybe KN knows.)

    His first few papers were published under the name “Irving Copilowish,” so he must have Anglicized it as an adult. He made some important contributions to logic, I understand, though I have no idea what they were. IIRC he was an unrepentant Platonist regarding mathematical objects, but maybe I’m wrong about that.

  26. CharlieM: The blue sky and red sunsets are also explained in this way.

    Complete rubbish. You can photograph sunsets. You cannot photograph Benham colors, because the Benham top does not “objectively” alter the colors.

    The Benham effect works regardless of the colors of the top. It actually works best with monochromatic light. The illusion is induced by alternation of light and dark. The misinterpretation happens in processing before the signal reaches the brain.

  27. petrushka: Complete rubbish. You can photograph sunsets. You cannot photograph Benham colors, because the Benham top does not “objectively” alter the colors.

    Goethe is attempting to describe what he sees without speculating as to any causes lying outside what he observes. The red/yellow and blue/violet that we see in the vastness of the heavens is also seen when we shine light through a semi-opaque liquid. The direction of the light source will determine whether the liquid appears yellowish or blueish. Try it with a glass of milky water and a torch (flashlight).
    Here is an excerpt from ‘The Theory of Colours”:

    150.

    The highest degree of light, such as that of the sun, of phosphorus burning in oxygen, is dazzling and colourless: so the light of the fixed stars is for the most part colourless. This light, however, seen through a medium but very slightly thickened, appears to us yellow. If the density of such a medium be increased, or if its volume become greater, we shall see the light gradually assume a yellow-red hue, which at last deepens to a ruby-colour.—Note L.

    [Pg 62]

    151.

    If on the other hand darkness is seen through a semi-transparent medium, which is itself illumined by a light striking on it, a blue colour appears: this becomes lighter and paler as the density of the medium is increased, but on the contrary appears darker and deeper the more transparent the medium becomes: in the least degree of dimness short of absolute transparence, always supposing a perfectly colourless medium, this deep blue approaches the most beautiful violet.

    152.

    If this effect takes place in the eye as here described, and may thus be pronounced to be subjective, it remains further to convince ourselves of this by objective phenomena. For a light thus mitigated and subdued illumines all objects in like manner with a yellow, yellow-red, or red hue; and, although the effect of darkness through the non-transparent medium does not exhibit itself so powerfully, yet the blue sky displays itself in the camera obscura very distinctly on white paper, as well as every other material colour.

    153.

    In examining the cases in which this important[Pg 63] leading phenomenon appears, we naturally mention the atmospheric colours first: most of these may be here introduced in order.

    154.

    The sun seen through a certain degree of vapour appears with a yellow disk; the centre is often dazzlingly yellow when the edges are already red. The orb seen through a thick yellow mist appears ruby-red (as was the case in 1794, even in the north); the same appearance is still more decided, owing to the state of the atmosphere, when the scirocco prevails in southern climates: the clouds generally surrounding the sun in the latter case are of the same colour, which is reflected again on all objects.

    The red hues of morning and evening are owing to the same cause. The sun is announced by a red light, in shining through a greater mass of vapours. The higher he rises, the yellower and brighter the light becomes.

    155.

    If the darkness of infinite space is seen through atmospheric vapours illumined by the day-light, the blue colour appears. On high mountains the sky appears by day intensely blue, owing to the few thin vapours that float before the endless dark space: as soon as we descend in the[Pg 64] valleys, the blue becomes lighter; till at last, in certain regions, and in consequence of increasing vapours, it altogether changes to a very pale blue.

    This is worth reading in its fuller context. I don’t have time just now to pick any other examples from the book

    petrushka: You cannot photograph Benham colors, because the Benham top does not “objectively” alter the colors.

    The Benham effect works regardless of the colors of the top. It actually works best with monochromatic light. The illusion is induced by alternation of light and dark. The misinterpretation happens in processing before the signal reaches the brain.

    Of course it cannot be photographed because it relies on the movement of the light and dark areas. There is no overlapping of light and darkness in a disc which is static relative to the observer.

    It doesn’t matter if we are observing an apple, a rainbow, a sunset or a spinning disc, in all cases four things are necessary, light, the entity under observation, the observing subject, and the visual apparatus of the subject.

    In each case the difficult question is: “Where does the colour reside?

  28. BruceS: I find that in general about Dennett; for me, his acolytes do a better job of explaining Dennett’s ideas than he does.

    That’s probably because Dennett’s version comes first. The others then have a chance to see how it was misunderstood.

  29. Neil Rickert: That’s probably because Dennett’s version comes first.The others then have a chance to see how it was misunderstood.

    Easier to improve something than to originate it.

    Or so I’ve heard.

  30. keiths: they are absent from the stimulus but present in the percept.

    Not sure what you mean. I do not see lines that are not there, but I see an implied figure. Just as I might see a figure hiding in the bushes. It’s not seeing something that is not there. It’s forming adaptive conclusions.

  31. I think I’m going to insist that the term “veridical perception” is equivocal or misleading. The issue is not about perception being veridical, but about it being more or less detailed. Evolution would select for detailed-enough. That perception is not too detailed, say, for us to “know” that atoms are mostly “empty” space, or even that there’s such things as atoms, doesn’t mean that our perceptions are “false.” It just means they’re not that detailed.

    There’s also a problem distinguishing between perception and interpretation. Dividing the issue this way helps a lot. If we perceive half a cat, because the other half is behind a wall, there’s nothing “unveridical” about the perception, and we still interpret the cat as a complete one.

    Have a great evening.

  32. The line between organisms that have consciousness and those that don’t is a bit like the line between bird ancestors and reptile ancestors. You have to draw the line before you can see it.

    Same with perception and interpretation. The bottom line is, senses are functional. They support actions necessary for survival. Success is statistical.

  33. petrushka,

    Not sure what you mean. I do not see lines that are not there, but I see an implied figure.

    Perhaps you aren’t seeing the Kanisza illusion as most of us do. (I know that some people don’t).

    I see a solid white triangle, with three full edges, superimposed on the black triangle and the black dots. Is that what you see?

  34. Entropy,

    I think I’m going to insist that the term “veridical perception” is equivocal or misleading. The issue is not about perception being veridical, but about it being more or less detailed. Evolution would select for detailed-enough. That perception is not too detailed, say, for us to “know” that atoms are mostly “empty” space, or even that there’s such things as atoms, doesn’t mean that our perceptions are “false.” It just means they’re not that detailed.

    I think detailed/coarse is a separate dimension from veridical/non-veridical. It’s easy to imagine a perception that is full of detail, most of it inaccurate.

  35. Science writer Kate Baggaley tells of a curious finding:

    Our more distant avian relatives can be fooled by the Ebbinghaus-Titchener illusion too. But, strangely, the size estimation bias is reversed. Chickens judge a disk surrounded by large disks to be bigger than it is in reality. Researchers found the same bias in pigeons and concluded that, “Pigeons may actually experience a visual world too different for us to imagine.”

    The range of binocular vision that a pigeon has is very narrow compared to its total field of vision which is very wide. Humans and owls have narrower fields but the proportion of binocular vision is much higher. Maybe this has something to do with how pigeons experience these illusions, I don’t know.

    In this piece Baggaley flips between what she says is actually being deceived by these illusions. Sometimes it is the individual creatures that are, “tricked by optical illusions”, sometimes it is “our eyeballs” which make a judgement on the length of lines, or it could be that, “brain is fooled” by the illusion.

    I would prefer that she showed a bit of consistency here and stuck to the one story that it is the individual that is fooled. Our eyes don’t expect there to be certain visual relationships and neither do our brains. It is we as sense aware, thinking individuals who have these expectations.

  36. I see the implied triangles, but not the full edges. I can assume there are edges, but I don’t see them. Before you started using the word triangles, what I “saw” was a six pointed star.

    But it’s not an illusion like Benham colors or the ghost gray thingies. I recall Joni Mitchel using the phrase cloud’s illusions, but I don’t think of them as illusions.

    When I see Benham colors, I really believe I am seeing colors. When I see figures in clouds or in incomplete drawings, I really believe I am seeing an implied figure.

    It’s just a fussy word distinction, but one I think is important.

    I think the two classes of phenomena occur at different levels of perception.

  37. Entropy: That perception is not too detailed, say, for us to “know” that atoms are mostly “empty” space, or even that there’s such things as atoms, doesn’t mean that our perceptions are “false.” It just means they’re not that detailed.

    If you believe that atoms are mostly empty space, what substance do you believe occupies the remainder of the atom? What would you say that the subatomic “particles” are made of?

  38. petrushka,

    I see the implied triangles, but not the full edges. I can assume there are edges, but I don’t see them.

    I see full edges on the white triangle, but only partial edges on the black one. The white triangle appears to be in front of the black triangle and the black dots, and it’s opaque.

  39. If that’s what you see, I would agree it’s an illusion. But I see no boundary for the implied white triangle. I’m viewing it on an iPad with retina display. If I squint the right way, I can imagine the white triangle is a fraction of a shade darker than the background, but I have to work to see that. If that’s what I’m supposed to see, it’s very weak.

    Benham colors, viewed under monochromatic light, are strong, dramatic, almost iridescent.

  40. BruceS,

    I finally sat down and read Frankish’s article. That, combined with the Aeon piece, have me very confused. Frankish claims that he is an anti-realist about “phenomenal properties” or “qualia.” I think I just don’t understand what the hell phenomenal properties are supposed to be. I think there’s a lot of bullying and posturing that goes on here (“you don’t believe in qualia! then you must be a zombie!”) but the more I read this stuff, the more I suspect that the belief in phenomenal properties relies on a lot of intuitions that I don’t share.

    As best I can tell, the idea of phenomenal properties is the idea that we discover through introspection a whole bunch of sensory states that we also think are states of perceptual objects, e.g. we think that the brown that we’re introspectively aware of as a visual sensation is also a state of the wood that we’re looking at.

    Is that right?

    I ask because that does not make any sense to me at all.

    Maybe it’s because my intuitions have been corrupted by too many years of phenomenology and cognitive neuroscience, but it’s pretty clear to me that there aren’t any such things as phenomenal properties.

    I mean, I do experience perceptual objects as having determinate sensory qualities but I experience those qualities as co-constituted by my own embodied relation to the world and I understand that co-constitution in causal terms as the looping dynamics in and across brain, body, and environment.

    So I have no idea if that means I experience phenomenal properties or not.

  41. For the record, I posted to Twitter this bit:

    “I do experience perceptual objects as having determinate sensory qualities but I experience those qualities as co-constituted by my own embodied relation to the world and I understand that co-constitution in causal terms as the looping dynamics in and across brain, body, and environment.”

    and asked Frankish if that meant I experience phenomenal properties. His response: “It sounds as if you don’t need to mention them! (And I like the picture you sketch.)”.

    I’m inclined to think that illusionism, as Dennett and Frankish call it, is a theory of consciousness only insofar as it tells us to avoid a whole bunch of philosophical mistakes — mistakes that Nagel and Chalmers became famous for making. It will seem counter-intuitive to anyone who is convinced that there are phenomenal properties and that they experience them. I don’t experience phenomenal properties (it seems) — not in the sense that makes Chalmers think that physicalism cannot be true, anyway.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.