Skip

"In Japan, people often refer to traffic lights as being blue in color."
June 17, 2012 11:08 AM   Subscribe

The crayola-fication of the world: How we gave colors names, and it messed with our brains -- A two-part essay about (part I) how perception of colors affects our naming of colors which (part II) affects our perception of colors. Guest celebrities include Darwin's children, xkcd, a mantis shrimp, and Benjamin Whorf.
posted by ardgedee (44 comments total) 70 users marked this as a favorite

 
It is mentioned in passing in the article, but I wanted to highlight that Radiolab had a fascinating episode on this subject just a few weeks ago. You can listen to it and/or download it here.
posted by bjrn at 11:20 AM on June 17, 2012 [8 favorites]


Ugh. People have actually done experiments, and proven that all people have the same 'platonic' ideal for a few colors.

What they did is, they gave people hundreds of paint chips and asked them to come up with the names for 'basic' colors asked them to come up with the most 'pure' representation of that color

Basically, everyone chose the same chips for their color words. One interesting thing is that the order in which the colors appear in language comes in the same order. Starting with dark and light, then adding red, then adding green or yellow, then blue, then brown, and finally purple pink orange or grey. So it may simply be that Japanese skipped over the stage where blue and green separate.

(Also, the ancient Greeks didn't have a word for blue, in the Iliad the oceans are described as being the "color of wine")

They also found that people of other languages were most easily able to learn these platonic color terms.

This experiment was actually one of the major disproofs of he sapir whorf hypothesis, by the way. The idea that the Japanese, for example, would 'see' green and blue as being the same thing due to their language was precisely the conceit that was disproven.
posted by delmoi at 11:25 AM on June 17, 2012 [8 favorites]


Guys. The person who decided the color should be spelled FUCHSIA was not right in the head. One of my favorite colors but ugh, that pesky spelling.
posted by fluffy battle kitten at 11:33 AM on June 17, 2012 [1 favorite]


They renamed "flesh", which approximated a white person's skin tone, to "peach" after recognizing that human skin has many different colors. Here's a neat timeline about Crayola's colors and name changes.
posted by Renoroc at 11:37 AM on June 17, 2012 [2 favorites]


oh, delmoi...
posted by iamkimiam at 11:38 AM on June 17, 2012 [6 favorites]


So it may simply be that Japanese skipped over the stage where blue and green separate.

There is a Japanese word for what is considered to be "green" in European culture - midori.

Aoi doesn't just represent a colour, it represents the state of youth, vigour, vitality, and fecundity (and when it is time to proceed at a stoplight).
posted by KokuRyu at 11:41 AM on June 17, 2012 [9 favorites]


> Ugh. People have actually done experiments, and proven that all people have the same 'platonic' ideal for a few colors.

Man, people did a lot of experiments about a lot of things. If you followed your own link you'd have read that those people's research has had major criticism in recent years and even the researchers have had to loosen their own strict universal platonic truth of color, as it has shown to be less universal and platonic at as previously thought.

Not to mention the loaded concepts of classifying cultural aspects such as language on a spectrum from basic to more advanced, which is dripping with ethnocentrism.
posted by mrzarquon at 11:43 AM on June 17, 2012 [5 favorites]


I liked the very end of part II about child language acquisition of color terms...got me thinking about Genie, the feral child who was locked in a basement for her first 13 years. When you think of everything that she was deprived of—not being able to have that same 'click' into place of color and term, and all the parallels like that on other cognitive dimensions, big and small—it's no wonder she would never be able to do the most basic of tasks. And it's a miracle that she was able to do so much.
posted by iamkimiam at 11:44 AM on June 17, 2012 [1 favorite]


Also, the ancient Greeks didn't have a word for blue

Sure they did. κύανος, the root of 'cyan', was used by pre-Homerics. Homer actually uses the word too, but he uses it oddly; for instance, he uses it to describe the color of Hector's hair. The likely conclusion to draw here is that Homer was color-blind and apt to confuse colors, not that the Greeks didn't have a word for blue.
posted by painquale at 11:45 AM on June 17, 2012 [9 favorites]


"Guys. The person who decided the color should be spelled FUCHSIA was not right in the head. One of my favorite colors but ugh, that pesky spelling."

You can blame Mr. Fuchs for that.*

*Er, rather Mr. Plumier, who decided that fuchsia was just the thing, wacky spelling and all.
posted by iamkimiam at 11:52 AM on June 17, 2012 [2 favorites]


This thread is turning into a wikipedia flurry. Sorry!
posted by iamkimiam at 11:53 AM on June 17, 2012


They renamed "flesh", which approximated a white person's skin tone, to "peach" after recognizing that human skin has many different colors. Here's a neat timeline about Crayola's colors and name changes.
"Flesh" sounds kind of gross
posted by delmoi at 11:56 AM on June 17, 2012


Holy cats, I hadn't heard of that "Whorf was half right" paper. That is phenomenal!

(BTW, note that it's coauthored by Paul Kay, who is one of the principal researchers in the universal color category studies you mention, delmoi. It appears he has changed his mind. Half of his mind, anyway.)
posted by painquale at 11:57 AM on June 17, 2012 [4 favorites]


J.F. Ptak had a wonderful blog post recently about the medieval equivalent of the Crayola colour wheel, the urine wheel. Ah, the colours of medieval urine .. like a poem!

Albus color ut aqua fontis (White as wellwater)

Glaucus color ut cornu lucidum (Light blue/green/grey as lucid horn)

Lacteus color ut serum lactis (Milky as whey of milk)

Caropos color ut vellus cameli (Bluish-grey as camel skin)

Kyanos color ut vinum bene nigrum (Deep blue as very dark wine)

Viridis color ut caulis viridis (Green as green cabbage)

Lividus color ut plumbum (Livid as lead)

Niger ut incaustum (Black as ink)

Niger ut cornu bene nigrum (Black as very dark horn)

posted by verstegan at 12:07 PM on June 17, 2012 [7 favorites]


Man, people did a lot of experiments about a lot of things. If you followed your own link you'd have read that those people's research has had major criticism in recent years
Looking at the wikipedia article about the debate, it seems like most of it is about the order of emergence, rather then the basic idea of a handful of colors. There's apparently one language that contravenes these rules somewhat, but with anything done by humans there always going to be some exceptions.
Not to mention the loaded concepts of classifying cultural aspects such as language on a spectrum from basic to more advanced, which is dripping with ethnocentrism.
Oh please. If the color terms do follow some natural path, it makes sense to label them as basic to advanced.
posted by delmoi at 12:13 PM on June 17, 2012


In Japan, people often refer to traffic lights as being blue in color. And this is a bit odd, because the traffic signal indicating ‘go’ in Japan is just as green as it is anywhere else in the world. So why is the color getting lost in translation?

Because traffic lights are cyan, which really is just as much blue as it is green?
posted by Jpfed at 12:13 PM on June 17, 2012 [2 favorites]


J.F. Ptak had a wonderful blog post recently about the medieval equivalent of the Crayola colour wheel, the urine wheel. Ah, the colours of medieval urine .. like a poem!
What's interesting is that the sort of iconic image of a scientist holding a beaker and looking at it actually comes from medieval 'doctors'
Because traffic lights are cyan, which really is just as much blue as it is green?
Hmm, is that true? this light actually does seem much bluer then what I'm used to seeing in the US. this one seems greener, but you could still call it cyan. And you never know how well the cameras are capturing the actual color.
posted by delmoi at 12:22 PM on June 17, 2012 [1 favorite]


From Renoroc's link - Crayola renamed "Indian Red" to chestnut, despite it having nothing to do with American Indians, but was fine with "Fuzzy Wuzzy Brown"?
posted by sodium lights the horizon at 12:50 PM on June 17, 2012


I always kind of assumed "flesh" was the color of what's under the skin, not the skin itself. I never really think of flesh as skin. More like, the meaty stuff.
posted by Malice at 1:10 PM on June 17, 2012


"Looking at the wikipedia article about the debate, it seems like most of it is about the order of emergence, rather then the basic idea of a handful of colors. There's apparently one language that contravenes these rules somewhat, but with anything done by humans there always going to be some exceptions."

From your comments, I gotta assume that you read neither the linked articles nor the wikipedia pages you link to.

The debate isn't only over the order of emergence — which also contravenes the second part of your comment, since there doesn't seem to be a universal development process — but rather: 1) the validity of the question (whether color is a universally meaningful semantic space), 2) that color order acquisition is not correlated at all with the linguistic categories of a language, 3) that specific fieldwork counters the underpinning assumptions of universal color development in two ways: that some colors can be fully conceptualized through compound modifiers, and that some colors lack universal recognition as separated concepts, and 4) that the psychophysics of color and color development are not isomorphic, and that causality may be muddled between color discrimination and color naming, and that thereby individuals and cultures can take multiple paths to similar results.

Further, the linked article talks specifically about how Whorf was half-right, in that linguistic cross-category brain processing aids color discrimination, but only on half of the field of vision.

It's just frustrating to see you being immediately dismissive of something that you haven't bothered to engage with, and doing so in a way that's at least moderately discussed in the initial blog posts. It's like you wandered in after half the class was over and started blundering out opinions that the prof had already gone through instead of asking questions where you didn't understand.
posted by klangklangston at 2:05 PM on June 17, 2012 [8 favorites]


Is it true that we see three colours because we have three basic colour receptors and birds and insects have more than three?

The Japanese traffic lights that Delmoi posted are both not as green as the traffic lights where I live. To me the first one is definitely as much blue as green.

Isn't it true that knowing the name of anything makes it easier to recognize? Starting from when you are young reading a book that includes a lot of words you don't know so you just get a general idea from context but tend not to observe the words at all until you actually learn them so the passage gets a sharp meaning.

Until you learn the various names of the weeds in the tangled thicket you are apt not to see them as different from each other. Purple flowers become vetch and michaelmas daisies and monkshood. Giving things a specific name makes it easier to see them and to differentiate them. And it's not just a matter of genetic difference. Ornithologist had to reduce the number of different bird species after they started doing genetic studies because they found out that certain previously believed to be completely different species were actually the same type of bird either with a different plumage, or age, or habitat.

It seems to me that there are a great many words with meanings that have a fuzzy border. Take the word "chair" . Exactly when does a chair cease to be a chair and become a bar stool? Exactly when does a chair cease to be a stone throne and become rock that look a bit like a chair? When does an arm chair become a couch?

Wait... doesn't this bring us to the first verse of the Tao-te-ching, something or the other about names?

Flesh made perfect sense as the name of a colour to me because I went straight from eight packs of Crayola kindergarten crayons to a sixty pack of prismacolours which had a great deal more of various browns to choose from than the actual three shades of "flesh" which were "flesh" "dark flesh" and "blush". There was also a "peach" but it was an orange with pink tones. But at the same time I was also pitched into the Time Life series of books on the Great Masters which was painfully Euro-centric and also showed dozens of voluptous nudes. No doubt my understanding was also painfully Euro-centric. I remember making Indian paperdolls, coloured with different brown pencils, (except for Krishna who was not brown!) who wore saris made from winding strips of paper around them... It wasn't so much that I thought skin was "flesh" coloured but that only a certain pale pink brown bunch of people came in such a limited range of colours that one shade could describe it. I also remember deep and troubled discussions on how to colour skin in my colouring books when I was a moppet. If you coloured really, really lightly with the red they came out rosy pink like someone with severe burns. If you coloured them with the orange they came out looking like a pumpkin. If you blended the orange and the red they came out in a sort of impossible plaid because it was impossible to colour lightly enough and mix the colours. And if you left them blank with skin as white as the pape,r some officious grup was bound to tell you that you hadn't finished colouring the picture... The occassional chance to colour someone with brown skin was a relief.
posted by Jane the Brown at 2:07 PM on June 17, 2012 [1 favorite]


"Isn't it true that knowing the name of anything makes it easier to recognize? Starting from when you are young reading a book that includes a lot of words you don't know so you just get a general idea from context but tend not to observe the words at all until you actually learn them so the passage gets a sharp meaning."

The result of the experiment was: Yes, but only in the side of your brain that processes language, otherwise no, it makes you slower.
posted by klangklangston at 2:09 PM on June 17, 2012


Aoi doesn't just represent a colour, it represents the state of youth, vigour, vitality, and fecundity (and when it is time to proceed at a stoplight).

"In my aoi days when I was midori in judgment." Have we had a "Shakespeare was really Izumo no Okuni" thread?
posted by yoink at 4:25 PM on June 17, 2012 [1 favorite]


It's just frustrating to see you being immediately dismissive of something that you haven't bothered to engage with, and doing so in a way that's at least moderately discussed in the initial blog posts. It's like you wandered in after half the class was over and started blundering out opinions that the prof had already gone through instead of asking questions where you didn't understand.
Well, given this was something that I learned in a college class (in a fairly recent textbook), obviously I'm going to assume it's more accurate then a bunch of message board comments. The Wikipedia articles (which I did read) indicate there is a 'controversy' indicates that a handful of people disagree with the conclusions, and that a single language, Yélî Dnye, spoken on an island on Papua New Guinea doesn't fit the pattern, which the original authors explain by claiming that's an example of a language that hasn't 'partitioned' light and dark yet, and so has flexible rules on color terms.
Further, the linked article talks specifically about how Whorf was half-right, in that linguistic cross-category brain processing aids color discrimination, but only on half of the field of vision.
While that's interesting, it hardly seems to me that being able to perform a mental task slightly faster is really what people meant when they talked about the sapir-worf hypothesis back in the 60s. the idea was that language determines how we think, not that it slightly impacts in a barely measurable way. It isn't that people without the word weren't able to do it, but rather they were able to do it, but rather, it took them longer to do it. On the order of 200 milliseconds or so. I would hardly call them half right, especially since in the real world you use both sides of your brain to think about things.
Until you learn the various names of the weeds in the tangled thicket you are apt not to see them as different from each other. Purple flowers become vetch and michaelmas daisies and monkshood. Giving things a specific name makes it easier to see them and to differentiate them.
The thing is though, if one of those gave you a sting, like poison ivy, you probably wouldn't need a word to tell it a apart. Having different words for things give you a reason to learn to tell them apart, but it's not the only reason. On the other hand, even if you have a word for some weed, if you never see it or think about it you're more apt to forget.

An obvious counter example is people you just met. You can have a brief discussion with someone at a party, you can remember who they are without ever having learned their name.
posted by delmoi at 4:33 PM on June 17, 2012 [2 favorites]


Bloom County on crayons.
posted by hot_monster at 4:39 PM on June 17, 2012


Also, the ancient Greeks didn't have a word for blue, in the Iliad the oceans are described as being the "color of wine"

The sea isn't actually "blue" except in children's drawings. Its color depends on the weather, the depth, the seabed and the local flora. I just drove over a causeway that traverses a lot of grayish-yellow sea, but if my kids drew it it would be blue.

If Homer called the sea wine-dark it was probably because of the impression of depth and darkness given by wine in an opaque clay dish - remember, this was before glass! I think it would be very unsafe to assume that he didn't actually realise that the sea often looks like the sky, or sapphire, or lapis lazuli. I've been reading up on ancient dyes, and some of them are clearly named after things with similar colors - there's one that's furnace-like, one that's turquoise-like and so forth. The dyers obviously recognised that certain colors resembled other ones, and I don't think that the names they use are less real than ours.
posted by Joe in Australia at 5:49 PM on June 17, 2012 [2 favorites]


Because traffic lights are cyan, which really is just as much blue as it is green?

Really? I've never seen a cyan-colored traffic light. I've always seen green-colored traffic lights. But maybe that's confirmation bias on my part. But please show me a pic or video of a cyan-colored traffic light...I see green.

And I live in Japan and when the recorded voice at crosswalks says "Aoi ni narimashita..." ([The light] has turned blue) I'll look up and check and sure enough it's green. Not blue, green. My wife is Japanese and I asked her "when you look at that light, do you see green or blue?" She says "Green".

"So why do Japanese people call that light blue?"

She just smiles. "Because that's what it's called." Duh! She knows it's "wrong" to call the green light blue, but it's just a little quirk that is probably below the radar of most people's perception.
posted by zardoz at 6:25 PM on June 17, 2012 [1 favorite]




I've been reading up on ancient dyes, and some of them are clearly named after things with similar colors - there's one that's furnace-like, one that's turquoise-like and so forth. The dyers obviously recognised that certain colors resembled other ones, and I don't think that the names they use are less real than ours.
no one is saying the names are 'less real' but the problem is that in the past you couldn't really generate any color with the limited selection of materials and dyes. Tyrian Purple was considered a "royal" color because the dyes were so expensive that only royalty could afford it. So the colors they used for dyes were the colors that it was actually possible to acquire or create.
posted by delmoi at 6:58 PM on June 17, 2012


The sea isn't actually "blue" except in children's drawings.

I have seen the sea not be blue, sure. But I have also seen it--on many occasions and in many different parts of the world--be as pure a blue as lapis lazuli or as the Madonna's robe or as the sky. So, no, it is decidedly untrue to say that the sea is not blue.
posted by yoink at 7:45 PM on June 17, 2012 [1 favorite]


Pretty damn blue
posted by KokuRyu at 7:55 PM on June 17, 2012


Hmm, is that true? this light yt actually does seem much bluer then what I'm used to seeing in the US. this one yt seems greener, but you could still call it cyan. And you never know how well the cameras are capturing the actual color.

In the first linked video, the sky looks much closer to cyan than the light does. But I guess my point of reference is the cyan that CMYK printers use, so it might not be a true cyan.

Pretty damn blue

The sky is blue, the sea toward the horizon is emerald (or a dark green), and I can't describe what the color of the water nearer the shore is. But it's not something I'd call blue. Seagreen maybe? It's seafoam green without the muddy component.
posted by gjc at 8:11 PM on June 17, 2012


I've never seen a cyan-colored traffic light.

Forgive me if I seem like I'm going to excessive lengths here, but this is a serious pet peeve of mine. Traffic lights are, for the most part, not actually green. Everyone says they are green (except the Japanese), but really, they're not*. This ridiculous falsehood starts in elementary school, where they are apparently afraid to teach the word "cyan" (despite its enormous importance as a primary pigment (the C in CMYK)).

Let us say that cyan is centered at hue = 180 degrees and that green is centered at hue = 120 degrees. Splitting the difference (180-120 = 60) equally, we'll give each color word a tolerance of 30 degrees in either direction (cyan will then be hue between 210 and 150; green will then be hue between 150 and 90).

Consider a google image search for "traffic light". Further, we will consider only actual photographs rather than artistic depictions of traffic lights.

For reasons of time I will consider just the first two pages of results, but anyone can replicate and extend this work if they so choose, and from a glance it's obvious that the trend continues in deeper pages of results.

The photographs in which the "go" light is not lit:
1 2 3 4 5 6

The photographs depicting cyan traffic lights:
1 2 3 4 5 6 7 8 9 10 11 12 13 (a duplicate of 9)

The photographs depicting green traffic lights:
1 2

*I'll grant that it is possible that color words do not depict simply connected swaths through any particular color space. Taking a descriptivist mindset, perhaps by virtue of the way "green" is used, "green" might mean something close to "a region of HSV color space bounded by such and such hue values, such and such saturations, etc... except in the context of traffic lights, where the hue range is expanded". The programmer in me finds those kinds of special cases very distasteful, but maybe that's just how things are going to be.
posted by Jpfed at 8:16 PM on June 17, 2012 [5 favorites]


Re "pretty damn blue" : The sky pictured is blue (I consider blue to be centered at hue = 240 degrees; the sky has hue values around 220). The sea pictured is cyan, ranging from pure cyan in the shallows (hue values around 180 degrees) to a cyan very near blue in the deep areas (hue values around 205).

If I had a drum to beat, it would be colored cyan.
posted by Jpfed at 8:22 PM on June 17, 2012


I've only ever seen a few actually green traffic lights. Generally pretty old ones. Almost all the ones I've seen in the US, and absolutely all of the ones I've seen in Japan, are blue-green (cyan, as Jpfed says, I suppose, but I'm not good at color words).

As for "Japanese call the traffic lights blue", that's a great way to start an article with an interesting anecdote, but as the rest of the essay points out, it isn't true. They call it "ao", and the word "ao" includes both the English words "blue" and "green".

Saying "they call it blue" is about as misleading as saying "Americans say that hot water is cold", because Americans use the word "water" to refer to all water, while in Japanese you have the word "mizu" (cold water) and "oyu" (hot water). Since a dictionary says "water" = "mizu", then by this strain of logic, English speakers would put their hand in a pot of boiling water (oyu) and call it water (mizu), thereby indicating they think it's cold.

The essay goes on to explain well the history of the ao/midori situation, but really does a disservice with that lead, since people here are managing to discuss "why Japanese call traffic lights blue", despite the article explaining "they don't".

(I will note, however, that Japanese, when studying English, will often make the mistake of calling traffic lights 'blue' in English, but this isn't because they see them as 'blue', but just because they're grabbing the first word in English that comes up in their dictionary when they look up "ao". And, looking at, for example, the Progressive English Dictionary, it's defined as 1) blue, azure, sky-blue, 2) green. If the order were reversed, and students learned "green" before "blue" when learning English, we might all be discussing how Japanese describe the color of the sky as "green")
posted by Bugbread at 8:55 PM on June 17, 2012 [5 favorites]




Thanks for the info all re:Japan and green/blue. Looks like I've gotta pay attention if lights are really green or actually cyan.
posted by zardoz at 5:30 AM on June 18, 2012


The ancient Greeks did not, in fact, think the sea was literally wine colored.
posted by the man of twists and turns at 10:00 AM on June 18 [+] [!]

Eπιστερικός!
posted by Joe in Australia at 5:38 AM on June 18, 2012 [4 favorites]


Forgive me if I seem like I'm going to excessive lengths here, but this is a serious pet peeve of mine. Traffic lights are, for the most part, not actually green. Everyone says they are green (except the Japanese), but really, they're not*. This ridiculous falsehood starts in elementary school, where they are apparently afraid to teach the word "cyan" (despite its enormous importance as a primary pigment (the C in CMYK)).

In the US, traffic lights can be bluer than they used to be.

The colour of US traffic lights is specified by the publication Equipment and Material Standards of the Institute of Transportation Engineers: Vehicle Traffic Control Signal Heads.

I haven't actually read the 2005 update to the standard (it costs ~$30) but it significantly shifted the specified range for a green signal towards blue, and made some slight changes to red and orange. The old standard for red and green permitted colours that were not readily detectable by colour vision deficient drivers, as well as greens that could blend into the sky. Some changes were also made for reasons related to LED colour ranges, and to bring things closer to the international standard.

The following PDFs reference the colour change in terms of the CIE 1931 colour space.

This PDF of a presentation on the new standard is pretty excellent, if summaries of traffic control standards is your thing. The chromaticity stuff is at the end, after intensity and luminance requirements, testing, and electrical requirements.

This comparison of international transportation chromaticity standards (pdf) is also pretty great.
posted by zamboni at 8:16 AM on June 18, 2012


zamboni: "The old standard for red and green permitted colours that were not readily detectable by colour vision deficient drivers"

Thank you, thank you, thank you! I'd heard that traffic lights were bluer in order to be easier to see for red/green color blind folks (which would explain why the really green traffic lights I saw in Houston were all really old ones), but I didn't want to mention it because I wasn't sure if it was just an urban legend. Good to know that something I heard from a FOAF was actually true.
posted by Bugbread at 12:00 PM on June 18, 2012


They're also bluer because the blue spectrum is distinguishable from a longer distance (shorter wavelengths=less distortion).

But what I'll say is 1) most of those lights still looked green to me (and I do enough color corrections to know CMYK fairly well), 2) Color perception is so heavily tied to confounding factors that making clear distinctions on edge cases is pretty futile. Like, context has a massive effect on color perception, both from proximate colors and semantic context. Anyone who's ever had to make color illusions knows this stuff.
posted by klangklangston at 2:10 PM on June 18, 2012


most of those lights still looked green to me (and I do enough color corrections to know CMYK fairly well)

You can easily do what I did- grab the image and use an eyedropper tool*. Since I'm on Win7, I grabbed the relevant portion of each image with the snipping tool, then used MS Paint to eyedropper a lit pixel, then renormalized the hue values from 240 unique values to 360.

Color perception is so heavily tied to confounding factors that making clear distinctions on edge cases is pretty futile.

Meh, there were only 2 lights that ended up in the cyan category that were anywhere close to the boundary I'd set (the boundary was at hue = 150; a couple lights were around 155). There were a surprising number that were right at hue = 180. They really weren't edge cases. I only busted out the eyedropper tool to make sure that I was sufficiently generous to the "green" side, only to have my visual guess verified by the eyedropper every time*.

Anyway, one must make a distinction to translate from the percept to a word. To the extent that we are interested in using words to describe colors (and maybe you're not; that's cool), we kind of have to have a mapping of some kind. I argue that the word "cyan" is a better match than "green".**

Like, context has a massive effect on color perception, both from proximate colors

This is a more likely source of variation in this set of cases. The blue sky that the lights are often framed against may reduce the perceived blueness of a traffic light.

and semantic context.

A very plausible influence. Check out the image search results. The artwork depicting stoplights almost always shows them as quite green. It's what we expect to see.

*It's possible that my vision is weird in some way. I used to work for a visual perception lab, and so I've seen a lot of illusions. Judging from the differences in how I perceived those illusions from how others in the lab did, my vision does seem to be oddly "local"- not as influenced by lateral inhibition as others. This makes it easy for me to forget that an eyedropper tool is not an accurate measure of what color people will actually see when the pixel in question is in context.

**Another weakness of my argument is the supposition that 150 degrees is the appropriate place to divide green from cyan. It may be that others would consider the color word "cyan" to subtend a smaller volume of color space than "green"- perhaps much smaller. Experience must play a part in shaping this. I grew up playing CGA games that used 4 colors (black, white, cyan, red), so maybe the word "cyan" subtends a greater portion of the space of my perceptual possibilities than it does for others.
posted by Jpfed at 2:52 PM on June 18, 2012 [1 favorite]


Jpfed: "Another weakness of my argument is the supposition that 150 degrees is the appropriate place to divide green from cyan. It may be that others would consider the color word "cyan" to subtend a smaller volume of color space than "green"- perhaps much smaller."

Though not a rigorous, peer-reviewed survey by any means, a good starting point may be the results of the survey by Randall Munroe, the xkcd guy. After all, it is a map of what words people actually use (as opposed to what a photographer or illustrator or the like might use) to describe colors independent of context.

I looked at the first 10 photos with lit green lights when searching for "traffic light" in Google Image Search, eyedropped the traffic light using a 50px by 50px average setting or a 31x by 31x average setting (depending on the image size), and checked where it fell on the xkcd image.

The results were, for me at least, startling.

Image 1: Teal
Image 2: Teal
Image 3: Blue (Note: for this one, I had to highlight the glow around the light, because the light itself was too close to white to show up anywhere on the xkcd image)
Image 4: Green
Image 5: Cyan
Image 6: Teal
Image 7: Teal
Image 8: Teal
Image 9: Teal
Image 10: Teal

We started out debating whether traffic lights, independent of the traffic light context, are "blue" or "green". Then an alternative, "cyan" was proposed. But the word that most people would use to describe the color of a traffic light, when removed from the traffic light context, is "teal".
posted by Bugbread at 5:02 PM on June 18, 2012 [2 favorites]


This thread got me thinking about the terms for blue and green in Korean. The native Korean word "푸른 / 푸르다" can be used to denote either blue or green, the usage context clarifying what is meant. So if 푸른 is used with sky, it is inferred that the color meant is blue, while if it is used with forest, then 푸른 is interpreted as denoting green. But, 푸른 is used strictly as a descriptor, and is not an actual color name.

The color name for blue is 파랑 and the color name for green is 초록. You use these terms when discussing colors specifically, and very rarely as descriptors. The interesting thing is that 파랑 is a native Korean term, while 초록 is actually from hanja, 草綠, and is literally "grass green." But wait, then there's 靑 (청), which hanja dictionaries will tell you means 푸른, and also happens to be the Japanese ao / aoi (kanji: 青). Oh, and midori from Japanese? That's the 록(綠) hanja from 초록 (草綠), or 緑 as the Japanese write it. So it's not that Koreans or Japanese think that traffic lights are blue, but that a descriptive term is used that can mean blue or green depending on the context of its usage.

The article linked to in the post mentions that 靑 can be blue or green in Chinese also. So, in the end, it's probably all the fault of the Chinese.
posted by needled at 6:28 PM on June 18, 2012 [2 favorites]


« Older Aone in the Arctic   |   My Old Man - The Five Spirits... Newer »


This thread has been archived and is closed to new comments



Post