Colour, Resolution, and David Hume.
October 31, 2008 5:08 PM Subscribe
To what extent can we discern different colours? What of this power is merely relational (a light blue, a lighter blue)? Have studies been done to determine if our "resolution of discernment" (a particular distance in nanometers of which we can differentiate two visual wavelengths with unaided vision)? What's the difference between how we speak of colour, how we conjure it up in its particularity in our imagination, and how we differentiate it through experience?
Many questions and an extended explanation, I know, I know. I considered narrowing this question down but I realize what I want is a broad answer, with specific issues addressed. So perhaps the most helpful answer would be a specific field of inquiry, not a specific journal article.
I imagine both philosophical and scientific approaches will agree that 'it depends on the perceiver' - a colour blind man may have a much lower resolution for differentiating some colours over others. Or are there contemporary arguments suggesting that colour is entirely subjective, any distinguishing factors are relational constructs out of scientific experiments that confuse qualitative experience with quantitative results?
The reason I'm looking for both philosophical and psychological approaches is that the question was stimulated from the beginning of Hume's Essay. There he uses a thought experiment as counterexample to his maxim that our simple ideas are composed out of equivalent simple impressions. He imagines a simple idea produced without a corresponding simple impression: a painter who has knowledge of all the shades of blue but one 'blue' in particular goes ahead and lays out the gradation of blue in front of him as a series of swatches, all except the the shade he has no knowledge of. Hume inquires as to whether the painter will see a gap in this array, for even if he has no knowledge of this particular blue, surely he'll see the relational problem of one blue 'leaping' to the other. He then speculates as to whether or not the man can produce a unique idea of the missing blue in his mind. He assumes (rather quickly) that he must be able to do, then discounts his experiment as too rare and trivial to change his initial maxim. Today, we may answer his problem by saying the painter has a relative idea of what blue should be there (lighter, more greenly) or perhaps we would say that the thought experiment itself is silly and filled with problems. That's not the concern I'm wrestling with.
Instead, I'd like to know more about the subjectivity of colour as is perceived in the mind, and if there is quantitative information on our abilities to do such a thing. (To give a completely arbitrary and baseless example, perhaps the majority of subjects can discern two different shades of red at a resolution of 50nm but two shades of violet at a distance of 10nm).
It is marveling to me that we can distinguish ten million different colours. Under what conditions can we do this? Can I put Colour #9342132 beside Colour #9342133 and distinguish them immediately?
From the initial question I posed in the heading as to whether the power is merely relational, I mean so from the act of reflection. We may be able to discern two colours side by side in immediate experience, but how about the powers of conjuring up different colours in our memory? I understand this is an entirely different field, but I want to approach colour from all angles here.
Many questions and an extended explanation, I know, I know. I considered narrowing this question down but I realize what I want is a broad answer, with specific issues addressed. So perhaps the most helpful answer would be a specific field of inquiry, not a specific journal article.
I imagine both philosophical and scientific approaches will agree that 'it depends on the perceiver' - a colour blind man may have a much lower resolution for differentiating some colours over others. Or are there contemporary arguments suggesting that colour is entirely subjective, any distinguishing factors are relational constructs out of scientific experiments that confuse qualitative experience with quantitative results?
The reason I'm looking for both philosophical and psychological approaches is that the question was stimulated from the beginning of Hume's Essay. There he uses a thought experiment as counterexample to his maxim that our simple ideas are composed out of equivalent simple impressions. He imagines a simple idea produced without a corresponding simple impression: a painter who has knowledge of all the shades of blue but one 'blue' in particular goes ahead and lays out the gradation of blue in front of him as a series of swatches, all except the the shade he has no knowledge of. Hume inquires as to whether the painter will see a gap in this array, for even if he has no knowledge of this particular blue, surely he'll see the relational problem of one blue 'leaping' to the other. He then speculates as to whether or not the man can produce a unique idea of the missing blue in his mind. He assumes (rather quickly) that he must be able to do, then discounts his experiment as too rare and trivial to change his initial maxim. Today, we may answer his problem by saying the painter has a relative idea of what blue should be there (lighter, more greenly) or perhaps we would say that the thought experiment itself is silly and filled with problems. That's not the concern I'm wrestling with.
Instead, I'd like to know more about the subjectivity of colour as is perceived in the mind, and if there is quantitative information on our abilities to do such a thing. (To give a completely arbitrary and baseless example, perhaps the majority of subjects can discern two different shades of red at a resolution of 50nm but two shades of violet at a distance of 10nm).
It is marveling to me that we can distinguish ten million different colours. Under what conditions can we do this? Can I put Colour #9342132 beside Colour #9342133 and distinguish them immediately?
From the initial question I posed in the heading as to whether the power is merely relational, I mean so from the act of reflection. We may be able to discern two colours side by side in immediate experience, but how about the powers of conjuring up different colours in our memory? I understand this is an entirely different field, but I want to approach colour from all angles here.
Best answer: How The Retina Works. (pdf) Many computer graphics texts go into a lot of detail on the physiology of vision as a precursor to being able to effectively simulate objects on a screen.
posted by GuyZero at 5:19 PM on October 31, 2008
posted by GuyZero at 5:19 PM on October 31, 2008
I have no idea what your actual question is, so I'll just post a long and meandering answer. :D
For resolution: some people can do perfectly on this test but lots of people cannot.
Also it's misleading to talk about colors as having a wavelength. There is no wavelength that is Magenta. There's about a million ways to make your eye see yellow (one pure wavelength around 560nm, or equal parts of green (510nm) and red (660nm) light. Or a bunch of various other combinations in between.
Unless you use special purpose equipment* every test you can think of to test this is testing altering various mixes of RGB. This is different from shifting the color to a 50nm longer wavelength. In other words all existing color reproduction technologies we have are based on the human three-color physiology.
* You could probably do something with a bright white light and a prism and an opaque card with a slot.
posted by aubilenon at 5:25 PM on October 31, 2008
For resolution: some people can do perfectly on this test but lots of people cannot.
Also it's misleading to talk about colors as having a wavelength. There is no wavelength that is Magenta. There's about a million ways to make your eye see yellow (one pure wavelength around 560nm, or equal parts of green (510nm) and red (660nm) light. Or a bunch of various other combinations in between.
Unless you use special purpose equipment* every test you can think of to test this is testing altering various mixes of RGB. This is different from shifting the color to a 50nm longer wavelength. In other words all existing color reproduction technologies we have are based on the human three-color physiology.
* You could probably do something with a bright white light and a prism and an opaque card with a slot.
posted by aubilenon at 5:25 PM on October 31, 2008
Best answer: Color is a complete illusion. Theres no discrete wavelength between colors. Colors are what we perceive when our eyes absorb a certain type of radiation. There's really an infinite amount of colors. You cant just say "anything between this many nanometers is purple." That's going backwards. We say "we perceive things this wide as purple."
The entry on the math of color perception at wikipedia is worth reading.
A lot of this is pretty theoretical and our absorbtion of color and perception of color is biological and has evolutionary roots. Our preceptions have a lot to do with detecting predators, detecting enemies, detecting mates, and detecting safe food. I dont think it falls into a neat graph like you seem to be asking for.
posted by damn dirty ape at 5:55 PM on October 31, 2008
The entry on the math of color perception at wikipedia is worth reading.
A lot of this is pretty theoretical and our absorbtion of color and perception of color is biological and has evolutionary roots. Our preceptions have a lot to do with detecting predators, detecting enemies, detecting mates, and detecting safe food. I dont think it falls into a neat graph like you seem to be asking for.
posted by damn dirty ape at 5:55 PM on October 31, 2008
You should check out "Basic Color Terms: Their Universality and Evolution" by Berlin and Kay. Unfortunately I can't find a free online copy, but it's a study of how languages evolve words for colors--turns out that there's a certain order that they usually follow. (IIRC, "black"/"white" come first, then "red", then a word for all colors except red, then that word splits into "yellow" and "green", where "green" also includes blue, etc.)
I assume somebody's probably written a better study in the 40 years since, but I don't really know (I didn't study psychology; I read Berlin and Kay in AI class). These guys look like they might have some helpful stuff.
posted by equalpants at 6:01 PM on October 31, 2008 [1 favorite]
I assume somebody's probably written a better study in the 40 years since, but I don't really know (I didn't study psychology; I read Berlin and Kay in AI class). These guys look like they might have some helpful stuff.
posted by equalpants at 6:01 PM on October 31, 2008 [1 favorite]
We may be able to discern two colours side by side in immediate experience, but how about the powers of conjuring up different colours in our memory?
You've hit the nub. What is the touch "rough"? What is the physical feeling "cold"? What, precisely, do you feel when a "cutting" remark cuts? Words describe generalities, not specifics.
posted by Mblue at 6:07 PM on October 31, 2008
You've hit the nub. What is the touch "rough"? What is the physical feeling "cold"? What, precisely, do you feel when a "cutting" remark cuts? Words describe generalities, not specifics.
posted by Mblue at 6:07 PM on October 31, 2008
Much of our modern concept of complementary colors comes from Goethe, writing at about the same time as Hume. For a more technical/scientific approach, Real World Color Management delves into a lot of the theory and science behind how we see colors and is a perennial favorite among photographers.
That hue test that aubilenon linked is cool!
posted by TedW at 7:19 PM on October 31, 2008
That hue test that aubilenon linked is cool!
posted by TedW at 7:19 PM on October 31, 2008
You might want to think about, if you haven't already, qualia, specifically, the inverted spectrum problem.
posted by Rock Steady at 8:27 PM on October 31, 2008 [1 favorite]
posted by Rock Steady at 8:27 PM on October 31, 2008 [1 favorite]
Best answer: Also psychophysics, and the idea of "resolution of discernment" tends to be called Least Noticable Difference.
posted by so_necessary at 12:11 AM on November 1, 2008
posted by so_necessary at 12:11 AM on November 1, 2008
Instead, I'd like to know more about the subjectivity of colour as is perceived in the mind, and if there is quantitative information on our abilities to do such a thing.
I believe some tetrochromats reported that their color vision was unusual long before there was scientific proof for why that was. Because they couldn't quantify it, the whole issue was ignored until genetic testing proved they had four active cones instead of three.
posted by Mouse Army at 5:33 AM on November 1, 2008
I believe some tetrochromats reported that their color vision was unusual long before there was scientific proof for why that was. Because they couldn't quantify it, the whole issue was ignored until genetic testing proved they had four active cones instead of three.
posted by Mouse Army at 5:33 AM on November 1, 2008
The Feynman Lectures on Physics, Vol 1, Ch 35 has a very good introduction to color perception, from, unsurprisingly, a mostly physics perspective. Glancing at it now, you probably wouldn't need much of any background to understand it beyond some basic science.
As others have said, color does not map one-to-one with light wavelength. In vision, light will stimulate three receptors that are sensitive to broad, overlapping spectrums, and it is the combination of their response that is our initial color perception. The eye itself can not necessarily distinguish between a 'pure' wavelength of light and a more complex set of lights that will stimulate the color receptors in the same way. A certain yellow can usually be created by mixing any three colors (on a monitor or tv, usually red green and blue) in varying quantities.
As for conjuring up different colors in your mind, you may be interested in V.S. Ramachandran's (others as well?) work on synesthesia. I may be remembering this wrong, but he apparently found a color blind man with grapheme-color synesthesia who was able to perceive colors that his eyes never saw. Called them martian colors.
posted by jacobbarssbailey at 8:01 AM on November 1, 2008
As others have said, color does not map one-to-one with light wavelength. In vision, light will stimulate three receptors that are sensitive to broad, overlapping spectrums, and it is the combination of their response that is our initial color perception. The eye itself can not necessarily distinguish between a 'pure' wavelength of light and a more complex set of lights that will stimulate the color receptors in the same way. A certain yellow can usually be created by mixing any three colors (on a monitor or tv, usually red green and blue) in varying quantities.
As for conjuring up different colors in your mind, you may be interested in V.S. Ramachandran's (others as well?) work on synesthesia. I may be remembering this wrong, but he apparently found a color blind man with grapheme-color synesthesia who was able to perceive colors that his eyes never saw. Called them martian colors.
posted by jacobbarssbailey at 8:01 AM on November 1, 2008
This thread is closed to new comments.
posted by dsword at 5:14 PM on October 31, 2008