Talk:Color blindness/Describing color blindness

Dispute moved from Talk:Color blindness

Basically your premise is wrong: no form of red-green color blindness will cause red or green to appear gray. It simply makes them harder to distinguish from each other. For example, someone missing the red cone will have trouble detecting the addition of pure red to any other hue. Thus blue and purple often look the same, and green, yellow, orange, and red often look the same. A similar argument applies to people without the blue cone; adding pure blue to any other color tends to be undetectable. In the case of green the situation is more complicated, because it is not possible to add a color that would only stimulate the green cone in a normal retina; all possible greens stimulate a lot of red and at least a little blue. This can all be clarified by considering copunctal points (something I/we should probably cover either here or in its own article ASAP). Some diagrams from the web: [1] [http://webvision.med. for a deuteranope is an imaginary color far off the axis of the CIE color space. --Chinasaur 10:29, 1 Oct 2004 (UTC)

Your comment seems to indicate you don't understand how colour vision actually works nor what colourblindness is really about. Firstly, you seem to identify the so-called "red cone" with the perception of colour red, which is a complete misunderstanding of the functioning of colour vision. When one experiences colour white or colour yellow, there's nothing reddish in it; yet the "red cone" is being stimulated. Secondly, for a dichromat, what happens is not that red and green are "hard to distinguish from each other" as if they could see red and green but had some kind of cognitive problem identifying them; it's that the lights you see as red and green don't look ANYTHING like red and green to a dichromat. A dichromat doesn't know what red and green actually look like; those colours (the perceptions that a trichromat calls that way) are not part of a dichromat's colour experience. A dichromat has learned to use the words "red" and "green" as alternative words to call shades of what (s)he actually sees as yellow, beige, brown or grey. The following page: http://colorfilter.wickline.org/?j=1;t=p, uses a colourblind filter to simulate how a webpage appears to dichromat eyes. Enter "www.google.com" (or simply follow this link: http://colorfilter.wickline.org/?a=1;r=;l=0;j=1;u=www.google.com;t=p ) and see what happens. Can you now tell which letter is red and which letter is green? No, it's impossible to tell, and not because it's harder to distinguish red from green, but because now the letters appear only in shades of blue and yellow, there being nothing red or green to see. But a dichromat doesn't know—cannot actually know—that the colour (s)he is seeing in letter L is not actually the real colour green nor that the colour (s)he is seeing in letter E is not actually the real colour red. All (s)he can tell is that people seem to arbitrarily call what to his/her eyes appears as those yellowish-greyish insaturated colours you see in the simulation, sometimes "brown", sometimes "red", sometimes "green", sometimes "grey", without apparently any good reason even though other people seem to have no problem deciding when to use each colour name. Uaxuctum 22:41, 2 Oct 2004 (UTC)

No, my understanding of color vision is pretty good, at least at the photoreceptor level. We seem to be talking past each other.

When I say "someone missing the red cone will have trouble detecting the addition of pure red to any other hue", of course I don't mean the addition of some higher perceptual "reddishness"; that would make no sense, except perhaps through electrode/drug stimulation of higher brain areas, obviously not what I'm talking about. I'm talking about addition of power in the light spectrum that would normally stimulate only the red cone and not the green cone, i.e. pure, deep red, i.e. long wavelength light. A lot of your points seem to agree with my main point: that "no form of red-green color blindness will cause red or green to appear gray" (as long as we understand red-green color blindness to be a congenital, photoreceptor variety, which of course we do). Rather, lacking the red cone will cause reddish-blue (purple) to appear the same as blue, reddish-green (a higher perception non-entity but at the photoreceptor level perceptually yellow or orange) to appear the same as green, etc.. Yes, this does extend to reddish-gray appearing the same as gray, but reddish-gray is the only hue/saturation for which lacking the red cone will lead to confusion with gray. Copunctal points, confusion lines, and isochromatic elipses on the CIE color diagram really are the best way to understand color blindness at the photoreceptor level. Are you familiar with them?

You make points relevant to higher perception a lot. This seems mostly interesting to philosophers. If you are sure of your understanding of opponent processes in color blind subjects then I accept you could get some insight into higher perception. But in terms of just noticeable differences and basic perception, the relevant argument is at the photoreceptor level. When you start talking about how a color blind person "sees" the world I am skeptical. I agree that "a dichromat doesn't know what red and green actually look like" to a trichromat, but what makes you think you really know what red and green look like to a dichromat? Other than saying that protanopia makes it difficult to distinguish reddish-x from x at the lowest levels of processing, I don't see much other discussion as biologically useful or scientifically well founded. --Chinasaur 07:39, 4 Oct 2004 (UTC)

What you dismiss as "mostly interesting to philosophers" is actually the most useful and clarifying kind of information, while the discussion of what happens at the photoreceptor level is largely of little use for lay people who are just interested in understanding why colourblind people "confuse" two such different colours as red and green or what it is like to see the world through their eyes. If you only tell them about photoreceptors (which, by the way, are just the beginning of the story of the neurophysiology of colour vision), they will end up with a deep misunderstanding, let alone if you talk in the terms you've just used (identifying light spectra with the perceptions they most often prompt in a trichromat's mind, e.g. identifying long wavelength = deep red —> WRONG, the exact same light spectrum may be experienced as different colours depending on factors such as what colours surround it, and of course on whether the person is colourblind or not; a light of long wavelength is black for a deuteranope, not at all deep red). Just like what happens when you explain colour vision in similar terms instead of in terms of the perceptual colour space: people end up believing things like that yellow is "a mix of red and green", which is an unfortunately widespread misunderstanding. We know what kind of colour experience a dichromat eye produces at the mind level, because there are people who are dichromat only in one eye and have told us what difference there is between seeing the world in trichromacy and seeing it in dichromacy. And even if there weren't such kind of people, we could have inferred that information by asking dichromats how many colours they see as completely pure and cannot describe at all in terms of others, or by investigating the neuronal wiring of their retinocortex and comparing it with that of a trichromat's retinocortex. Having information from three kinds of cones enables the visual system, through a complex process of comparison involving ganglion cells, bipolar cells, cortical blobs, etc., to produce an endpoint colour signal with three layers of colour information: white-to-black, yellow-vs-blue and red-vs-green, and this is the kind of colour information that the mind of a trichromat experiences (and not at all the raw RGB-type of information that had started the processing path at the cones); while having information from only two kinds of cones produces, after all the processing, an endpoint signal with only two layers of colour information: white-to-black and yellow-vs-blue, so the red-vs-green layer is not experienced by a dichromat's mind. Now, unless we explain clearly that what happens to dichromat people is that they simply cannot experience the psychological perceptions of red and green, people won't truly understand why dichromat colourblinds cannot by the life of them "distinguish" two colours that to a trichromat appear so unmistakably different. OTOH, you're wrong, there are light spectra that a trichromat experiences as reddish or greenish but that a dichromat experiences as grey or greyish, and in fact, for deuteranopes some of them look plain black. Uaxuctum 23:14, 4 Oct 2004 (UTC)

I had wondered about the perceptual reports of the unilateral color blind people. I would be very interested in these case studies if you have references for them. Again, I don't see why you are getting so burned up, hot and bothered, high and mighty, or whatever you want to call it. We are not disagreeing about anything substantive except the relative importance of processing at the photoreceptor or opponent process levels. Please check out the summary of our combined answers that I tried to make above. Hopefully I did you reasonable justice, but change when you want; it's certainly not yet as simple or clear as I'd like. From what I wrote there, you should also be able to see why I think understanding copunctal points is fundamental to understanding color blindness; would you agree that reducing photoreceptor color blindness to a loss of red-green opponent channels is not quite a complete description of the deficiency?

Your arguments about perceived colors and opponent processes are basically an attempt to explain to a trichromat what a dichromat "sees" at a high peceptual level, an endeavor I am fundamentally skeptical about. I've always believed that trichromats should focus on understanding which colors dichromats will confuse and why, not on trying to imagine what they see. But if the unilateral color blindness case studies agree with your intuitive opponent process arguments then I can accept there is merit to your approach. Meanwhile my approach is not damaged in the least; hopefully you can see why it frustrates me that you have attacked all the nit-picks of my statements without apparently considering any of the merits of my underlying argument.

In terms of what is more clear to layman trichromats, I don't see how saying "dichromats just don't have a psychological understanding of red and green" is helpful. This leads to questions like "what do they see", which is the kind of vagueness that spawns mistakes like the original question above assuming that dichromats perceive gray in the place of both red and green, which is clearly wrong.

I don't really want to get into "red and green make yellow" with you; we both know that in human vision there's a simple and important truth to this statement, but if you're determined to find fault then you're bound to make a needless mess of the discussion. I agree with you that it's important to keep our terminology straight and not conflate a color with a wavelength, etc. Outside of the formal article though, I'm not going to qualify every statement I make; you can easily tell what I'm talking about, and if you didn't assume from the outset that I'm an ignoramus, you wouldn't have to waste your time "explaining" all my "mistakes" to me. If I work on the article, I'll be sure to be rigorous, and I will appreciate you rigorously checking me. I predict this may be a rare case in wikipedia history where a heated argument and general failure to communicate need not lead to an edit war, so in that sense alls well that ends well. But in terms of further discussion, your attitude makes it a pretty unpleasant prospect. --Chinasaur 01:09, 5 Oct 2004 (UTC)

What do they see? It's simple: red/green-colourblind dichromats see the world in terms of white, black, yellow and blue, with no red or green. The rainbow looks to them as a yellowish stripe below the neutral point, a thin whitish stripe around the neutral point, and a bluish stripe above the neutral point, with no red, orange, green or purple stripes. The array of light spectra that are metamers of a certain colour (so that a dichromat appears to "confuse" the colours those light spectra prompt in a trichromats' mind) varies for each case of colour blindness, but basically where a trichromat perceives a reddish or greenish colour they perceive insaturation of a usually yellowish shade, which may range from cream to brownish or greyish (some may appear black at the "red" end of the spectrum, white at the neutral point, and light bluish above the neutral point). If one presents the information about metamers in dichromacy by means of lines on the CIE chart, a lot of people won't understand anything, while they will easily when one tells them in plain terms which of the colours experienced by a trichromat look the same to a dichromat, like that a deuteranope sees deep red as black, pink as sky blue and forest green as brown, or if one offers a picture of the colours in the CIE chart as seen by a dichromat. People make a lot of fuss when they are explained colour blindness in terms of retinal cones, and usually end up with a lot of misunderstanding and confusion, while they understand the basics very easily once they compare a colourful image as seen by a trichromat and as seen by a dichromat. This is far more useful and clarifying for most people than a long discussion about photoreceptors and metamers and abstract lines on a chart which a lot of people are likely to find vague and arcane and are likely to prompt misunderstandings if a more down-to-earth explanation is not available. I'm not saying one should omit that technical side of the issue, but that an easy-to-follow explanation of what the world looks like to colour blinds in terms of what you like to dismiss as the "high perceptual level" (which is actually the intuitive, down-to-earth, easy-to-understand level because it's the level our minds work with) will be more useful for most people and should be offered first, before all the technical details underlying the phenomenon of colour blindness. Also, there is no simple and important truth in saying "red and green make yellow"; on the contrary, that statement is wrong and is the product of a widespread misunderstanding caused by explanation of colour vision in terms of retinal cones and using colour names improperly to refer to cones and to the wavelengths at their peak of sensitivity. You keep making this fundamental mistake of identifying wavelengths and cones with colours. Red and green do not make yellow at all, they make the redgreen novel binary hue of the Crane and Piantanida experiment. Yellow is an elementary percept completely independent of and impossible to define in terms of red and green, which belong to a different opponent channel. The perception of yellow is prompted mostly by an array of light spectra which stimulate both the M and S cones in an amount that surpasses the amount of stimulus of the L cone, but the presence of blue in a colour field can also prompt the perception of an amount of yellow in the adjacent colour field. Saying that "red and green make yellow" is both wrong and misleading; a more correct though incomplete statement is that "the stimulation of both the S and M cones prompts the perception of yellow". Uaxuctum 16:07, 5 Oct 2004 (UTC)

The percept yellow is not a mixture of pure red and pure green, it is a pure colour itself, there being nothing reddish nor greenish in yellow at all; yellow is just yellow, just like red is just red, green is just green, blue is just blue, white is just white, and black is just black. These six colours—white, black, red, yellow, green and blue—are the pure, basic, elementary percepts of human trichromatic vision, while all the other colours are perceptually mixtures of 2, 3 or 4 of these elementary percepts: orange is yellow+red, purple is red+blue, turquoise is blue+green, grey is white+black, etc. Red/green-colourblind dichromats have only two operative kinds of cones in their retinas, which allow them to make two kinds of distinctions: one dimension of lightness (white-to-black scale) determined by the intensity registered in their only short-wave receptor, and one dimension of hue (yellow-vs-blue opponency) determined by comparing the intensities registered in the long- and short-wave receptors. This means that, for them, the elementary colours are only four instead of six: white, black, yellow and blue; that's to say, the see the world in shades of white, black, grey, yellow, unsaturated yellow (beige, khaki, ochre, yellowbrown), blue and unsaturated blue (azure, navy, bluegrey). All the colours that trichromats perceive as involving the percepts red and green (red, green, purple, orange, pink, maroon, lilac, olive, chartreuse, etc.) are unknown to dichromats and their only real knowledge of those colours is limited to the words that designate them, because they lack the actual visual experience of what those words actually refer to, just like born-blind people lack the experience of seeing light even though they are aware of the word that refers to it. Trichromats, on the other hand, have two slightly different short-wave receptors, which allow them to make one further distinction by comparing their relative intensities, giving rise to an additional dimension of hue (red-vs-green opponency), which splits the yellowish band of the dichromatic spectrum into the greenish, yellowish and reddish bands of the trichromatic spectrum, enabling trichromats to experience a tridimensional colour space (with the white/black, yellow/blue and red/green axes, as shown in the NCS model). Since red/green-colourblind dichromats cannot make this distinction, they lack the experience of that additional dimension of hue and their colour space is bidimensional (with only the white/black and yellow/blue axes). However, as the article points out, dichromats are a minority within a minority, since many colourblind people are actually anomalous trichromats who do have the experience of the percepts red and green, only that many colours (especially unsaturated ones) that to a normal trichromat appear with a component of red or green do not look so to an anomalous trichromat, who requires more stimulation of the anomalous cone to experience an equivalent colour perception. Then, there is the possibility of tetrachromat women possessing two slightly different cones for what trichromats have only one, which should allow them to make a further chromatic distinction by comparing their relative intensities, expectably giving rise to an additional dimension with a pair of opponent hues unknown to trichromats (let's name them "septarine" and "octarine" after the unknown colour of magic and the fact that they are the seventh and eighth basic percepts after white, black, yellow, blue, red and green), allowing these women to experience a tetradimensional colour space with dozens of colours (like "reddish septarine", "octarinish purple", "dark greenish septarine", "yellowish octarinebrown", "blueseptarinish grey", "yellowgreenish octarine", etc.) for which there simply are no words (well, there weren't until now that I've come up with names for them ;-D) since such colours are unknown to the trichromat majority (just like red, green, light orange, reddish purple, bluish turquoise, etc. are unknown to the dichromat minority). From the point of view of a tetrachromat woman, trichromats are all colourblind since they cannot perceive any difference between colours that to her look as unmistakably different as red and green look to a trichromat (though not so to a dichromat); e.g. between a spectral yellow and a composite of spectral red and spectral green, both of which look yellow to a trichromat even though physically they are very different lights. Uaxuctum 21:17, 30 Sep 2004 (UTC)

You've made things a little confusing by first saying that yellow is just yellow (true of our perception of yellow in higher areas due to opponent processes), and then acknowledging that yellow can be a mixture of red and green at a physical or retinal perception level. Since we're talking about retinal color blindness, the higher perceptual arguments are not really relevant, only the retina matters. --Chinasaur 10:10, 1 Oct 2004 (UTC)

Prove that [all] yellow is not made from red and green. Do extra cones come with extra brain wiring to understand new hues, or does the wiring just stretch the hues to make new borders? lysdexia 14:48, 22 Nov 2004 (UTC)

Talk pages are where people discuss how to make content on Wikipedia the best that it can be. You can use this page to start a discussion with others about how to improve the "Color blindness/Describing color blindness" page.

Start a discussion