Here’s a quick experiment: make a pure yellow (#FFFF00) with ahex color tool. Looks yellow, right? Now drag the little crosshairs down chart to decrease the brightness. Looks…chartreuse…then olive…then greenish black. Not dark yellow.
I think, “what if it’s because the green cones are a tiny bit more sensitive than the red cones?” and decide to compensate by experimenting with a redder yellow (#FFDDOO). The darker shades still have a greenish cast, I think. Except that they also seem to have an orangish cast if I think of orange. Still not dark yellow. Turns out that if I make my starting yellow orangey enough, my dark yellow is perceived as brown.
Why won’t my brain allow me to imagine or see dark yellow without such strange problems? Is it something in visual processing that I’m not thinking of right now? A case of linguistic relativism?
If you can see and understand dark yellow, know more about color vision than I do, or speak a language that embraces this sketchy “dark yellow” character, please tell me about it.
Playing around is photoshop, the only dark yellow I could come up with was leaning brown. My WAG is that FFFF00 is dark yellow, and that your reference point of ‘yellow’ should be a lighter shade, say, FFFF50.
I picked FFFF00 as a pure yellow, not a light yellow. For the other pure hues, such as green (00FF00) you can slide down in brightness quite far before you start perceiving it as anything other than “just green”.
Green cones are more sensitive than other colors. 16-bit colors are traditionally broken up using 5-6-5 bits for R-G-B since we have more ability to distinguish between greens than other colors. Probably this is due to our having evolved in a natural setting where green is the predominant color.
But whether that is the cause of this effect, I don’t know. Another item to be pointed out is that additive colors don’t work the way you expect them to. For instance, if you use an art program to fill with a fade between yellow and blue, it will go enroute of grey instead of green like you would expect. So it might be some sort of artifact of this.
(Not one of those multi-thousand dollar gems that come with their own calibration hardware. Consumer grade. The type whose brightness setting is an insult because it doesn’t set brightness but only allows you to adjust which parts of the image you’d like to drown out in darkness. The type whose contrast setting borders on fraud because it only works over half the range and simply cuts off the remaining values for the other half. The type where you can’t tell dark blue from black. The type where there actually isn’t any black. The type that simply can’t display proper reds or cyans, but is too embarrassed to adjust its greens accordingly because that’s the last feature to distinguish it from becoming a monochrome display. The panel type that sacrifices viewing angle and color stability for a more marketable millisecond value (and lags several frames behind your mouse movement because of its crappy overdrive). The type that would properly need a 12-bit lookup table, but only has an 8-bit controller. Which fakes it with a dithered 6-bit noise pattern on top of it all.)
Their color reproduction sucks yak. The response of a liquid crystal cell is crooked and nonlinear, particularly in the darker ranges. Display makers somehow have to compensate for it. Many often really don’t.
Perhaps it’s just because we have names for all the dark yellows so you associate them with a name rather than a version of yellow. Same as how we have pink instead of light red…
The actual colors are scientifically quantifiable, but color names are cultural. Anthropologists Berlin & Kay have done a lot with this. If you’re asking “why is it that when I darken yellow, I don’t call the result ‘yellow’ but ‘brown,’” the answer is in part because your culture tells you where the boundaries between colors are. http://www.putlearningfirst.com/language/research/colour_words.html puts this better than I do. (Their theory that there is some sort of evolution going on until languages have the same terms as English is, of course, silly, but the data is interesting.)
I’ve heard it explained that the reason paint manufacturers come up with such apparently random names for their colors (i.e. “Morning Mist”, “Spring Snow”, “California Sky at 4:13PM on a Cloudless Day in August” (why yes, I am making these up)) is that they want names that are proof against somebody saying, “No it isn’t.”
This helps avoid situations where somebody sues the paint manufacturer with a claim of, “I wanted to paint my house turquoise, so I bought 20 gallons of paint labeled “Turquoise”, but what I actually got was teal! I want $1 million in compensation!”
It’s been a while since I read about that research, but if I’m remembering it correctly, you’ve misinterpreted it. The research doesn’t show that color boundaries are culturally determined. The research is perfectly compatible with the possibility (actually, the fact) that people from different cultures are able to make the same color distinctions in matching tests, no matter how large or small their color vocabulary.
The research you cited is about vocabulary only and has no implications for perception of color or thinking about color.
It has no implications for the perception of color, you’re quite right. I disagree that it has nothing to do with thinking about color. Recognizing shade, hue, and tone is one thing, but categorizing it is another. Language shapes how we organize the colors, even though the equipment we use to collect the data is, barring colorblindness, the same in every human. But I’m not an expert in color theory, and I’m spectacularly bad at recognizing shade, hue, and tone, so I won’t insist on the point.
