How the World’s Languages Name the Rainbow.

Claire Bowern writes in The Conversation about her research into color terms:

My colleague Hannah Haynie and I were interested in how color terms might change over time, and in particular, in how color terms might change as a system. That is, do the words change independently, or does change in one word trigger a change in others? In our research, recently published in the journal PNAS, we used a computer modeling technique more common in biology than linguistics to investigate typical patterns and rates of color term change. Contrary to previous assumptions, what we found suggests that color words aren’t unique in how they evolve in language. […]

That there’s something unique about the stability of color concepts is an assumption we wanted to investigate. We were also interested in patterns of color naming and where color terms come from. And we wanted to look at the rates of change – that is, if color terms are added, do speakers tend to add lots of them? Or are the additions more independent, with color terms added one at a time? […]

In order to answer these questions, we used techniques originally developed in biology. Phylogenetic methods use computers to study the remote past. In brief, we use probability theory, combined with a family tree of languages, to make a model of what the history of the color words might have been. […]

Our results supported some of the previous findings, but questioned others. In general, our findings backed up Berlin and Kay’s ideas about the sequential adding of terms, in the order they proposed. For the most part, our color data showed that Australian languages also show the patterns of color term naming that have been proposed elsewhere in the world; if there are three named colors, they will be black, white and red (not, for example, black, white and purple). But we show that it is most likely that Australian languages have lost color terms, as well as gained them. This contradicts 40 years of assumptions of how color terms change – and makes color words look a lot more like other words.

I freely admit I don’t understand the method works (“we estimate likely reconstructions, evaluate that model for how well it fits our hypotheses, tweak the model parameters a bit to produce a different set of results, score that model, and so on”), and I’m curious what my readers think of the method and the results.

Comments

  1. What I think is the most interesting fact about color terminology is that there are four common hues that every human encounters: red (from blood), yellow (from the sun), green (from chlorophyll), and blue (from the firmament). Everybody basically agrees what a generic red looks like, because everybody has had a cut. There are not the same kinds of universal reference prototypes for other hues like orange, purple, violent, etc. So I would expect to see the evolution of color names to be different depending on which kind of color one is talking about.

  2. Voltaire, in the 6th of his Lettres Philiosophiques on Newton’s optics, gives the colors of the rainbow (or rather of the prism) as: Le premier est couleur de feu; le second, citron ; le troisième, jaune ; le quatrième, vert ; le cinquième, bleu ; le sixième, indigo ; le septième, violet. “couleur de feu” instead of rouge seems particularly odd.

  3. The Sun isn’t anything like focal yellow: it’s white light with a little of the higher frequencies removed by scattering. So I doubt that’s it; more likely flowers such as daisies, or else leaves in the fall. Snow is focal white, but not every culture has experience of it.

  4. January First-of-May says:

    So I doubt that’s it; more likely flowers such as daisies, or else leaves in the fall.

    Or sand (it can have a lot of assorted shades, but most of the time it’s some variety of yellow).

  5. Everybody on this thread, just read Deutcher’s second book (Through the Language Glass, I think) and spare all of us much idle subjective back-and-forth…

  6. Chinese don’t distinguish between green and blue.

    So let’s take them out from list of basic colors.

  7. It seemed a little suspiciously convenient to me that a well-known Australianist just happened to conclude that the Pama-Nyungan languages were ideal for studying this question; but I guess it might be a convention of the genre. 🙂

    I haven’t read the PNAS article, but from the article that you link to, it sounds reasonable to me; but I’m not sure how one would cross-validate the claim that “Estimates that are very consistent […] are highly likely to be good reconstructions”. Intuitively it makes a lot of sense, but how would you know?

  8. @SFReader: “Basic colors” are a language-dependent concept. So Chinese may not have distinct basic color words for blue and green, but that doesn’t change the fact that many languages do.

  9. The sun is, famously, red in Japan. (And the sky is grue, except under Sinitic influence where it is sometimes Cerulean, Lapis Lazuli, and other basic poetic colors.)

    Re sand, amusingly that is mentioned in the first comment, which asserts that English has many more basic colors than CB claims, e.g. the beige of sand, which no English speaker would ever call yellow or brown. (To be fair, “pristine beige beach” does sound like the sort of phrase that would appear in a despairing couplet by Larkin.)

  10. Probably irrelevant is the curious fact that F blanc and E black are from the same PIE root. The meanings coincide in the ON doublet blakkr (of white horses and black things).

    But at least it’s possible to conceive of a language with only two color terms, which are homonyms…

  11. David Eddyshaw says:

    The question is not how many colour words a language has altogether, but how many words suffice to cover the entire spectrum. Speakers of three-colour languages typically have many more accepted and conventional words (or expressions) than three for colours, but any given colour can be correctly called by one of only three names. In English, it’s an actual mistake to call something “blue” if it’s green; not in a four-colour language. There’s also the question of whether the relevant colour terms have complex internal structure morphologically or semantically within the language in question.

    “Ash-coloured” and “grass-coloured” are both common and standard for “grey” and “green” across quite unrelated West African three-colour languages lacking simplex colour terms for those hues. Not at all surprisingly, when you think about it.

  12. David Eddyshaw says:

    The Seri language isolate of Mexico has six basic colour terms: white, black, red, yellow, green/blue, gray/brown, but there is no basic term for “colour”; according to Stephen Marlett’s grammar:

    There is no easy way to express the general word “color” in the language. If one wants to know the color of something, the opening question is ¿Zó hapácta ya? (literally, “How does it appear?” less literally “What is it like?”) which is understood as referring to color in the common situation. The fact that one is asking about color and not shape can be made clear by prompting the response with suggestions about the expected answers, such as ¿Cooxp ya? ¿Cheel ya? “Is it white?” “Is it red?”

  13. Hungarian has two words for “red”: piros and vörös. To non Hungarians, these look exactly the same, but to Hungarians, vörös is used for the Red Cross and red wine, piros is the colour of clothing and fruit, but that is just an approximation.

  14. Irish has “glas” for natural green (including some blue and grey) and “uaine” for artificial green.

    The green in the Irish flag is “uaine”, which interferes somewhat with the idea that it symbolises the verdant countryside. Of course, in the flag it symbolises the Catholic/Gaelic/Nationalist population, which is highly artifical; but nationalist adoption of green as a national colour is popularly ascribed to its natural prevalence.

    The flag’s “orange” is “flannbhuí” ‘flame yellow’ rather than “oráiste” ‘orange [fruit]’, despite the latter being used for political Orange

  15. Irish also has two words for red, “rua” and “dearg”.

    Color terms don’t depend just on hue, they also take saturation into account.

    “Glas” could correspond to English blue, green or grey, but with a similar low saturation.

    “Rua” is a low saturation red like red hair, but “dearg” is a high saturation red like blood.

    People who only know hue-based languages sometimes think this is universal, but it’s not.

    It takes three numbers to define a color, which can be chosen different ways, but hue, saturation and brightness is one commonly used approach. Some languages base color terms on just one axis, others might use two. I don’t know of any language that uses all three, but it’s possible.

  16. David Marjanović says:

    I’ve read the paper! 🙂

    I freely admit I don’t understand the method works (“we estimate likely reconstructions, evaluate that model for how well it fits our hypotheses, tweak the model parameters a bit to produce a different set of results, score that model, and so on”), and I’m curious what my readers think of the method and the results.

    Entirely unremarkable ancestral reconstruction using Bayesian inference (as opposed to maximum likelihood or simple parsimony). Using not just the topology of the trees, but also their branch lengths and a model of evolution, the program calculates probabilities for state changes along each branch ( = internode). The model contains probabilities for different kinds of state changes; these parameters of the model are themselves calculated by the program to fit the data as well as possible, but there is the option to set some of them as equal to each other or to 0, and that’s what the authors played around with. Notably, when they set the probability of losing color terms to 0, the program was unable to calculate the other parameters (“failed to converge [on a result]”).

    Or sand (it can have a lot of assorted shades, but most of the time it’s some variety of yellow).

    Latin had a word canis that covered things we’d call yellowish, like fields in summer, or gray. I’m not aware of a Latin word for “yellow”.

    It seemed a little suspiciously convenient to me that a well-known Australianist just happened to conclude that the Pama-Nyungan languages were ideal for studying this question; but I guess it might be a convention of the genre. 🙂

    The reason is spelled out in the paper: most IE and Austronesian languages are in stage VI, having words for brown, pink, purple and the like, while PN covers the whole range.

    Chinese don’t distinguish between green and blue.

    Well, 青 qīng does cover saturated greens, blues and blacks, but 绿 lǜ is just green and 蓝 lán is just blue, right?

    Apparently, 青 is what brown used to mean: dark and shiny.

    We can agree that Modern English brown, Modern German braun and Modern French brun (borrowed from Frankish) are “basic colour terms” and can be used to describe the colour o[f] a beaver’s coat. It doesn’t follow, however, that the same can be claimed of their Proto-Germanic ancestor, *βrūna-. In early Germanic languages the word meant ‘dark, swarthy, dusky’ (as well as ‘shiny, bright’, often with reference to forged metal or the sea), and while it could be used to modify virtually any hue for which there was a name, it was hardly a specific colour term itself. […] The “colour conspiracy” of the modern languages of Europe, which have developed identical or very similar basic colour systems, is a case of recent cultural convergence. As late as the seventeenth century, German braun could still refer to hues in the violet/purple range (e.g. the colour of the amethyst).

  17. Color coded names of Latin colors. Clever idea, but they unaccountable left out flavus (famously the color of the Tiber), and caeruleus and cyanus.

  18. @mollymooly: Another one of those counterintuitivities that I love – in the Irish conflict, the side that rejects orange has orange on their flag, while the side that embraces orange has none on theirs.

  19. No one’s going to discuss the methods?

    According to the anthropologist Harold Conklin, Hanunó’o has no basic color terms in the commonly understood, Berlin-Kay sense. What observers initially thought to be black, white, red and green turns out to be a two-dimensional continuum, with dark vs. bright on one axis and “dry” vs. “fresh” on the other, which, albeit prototypically “red” and “green”, may be realized as glossy vs. matte, or else as not purely visual, material considerations; “A shiny, wet, brown-colored section of newly-cut bamboo is malatuy [‘green/glossy’] (not marara [‘red/dry’])”. I wonder how many other “color systems” subtly differ in the very notion of “color”, without anyone having realized it yet.

    I’ve first heard of this in David Batchelor’s Chromophobia, which is a great little book.

  20. I’ve read the paper!

    Thanks for your dedication and very helpful commentary! (Which was, alas, held up in the moderation queue until I got up this morning.)

  21. David Marjanović says:

    Color coded names of Latin colors.

    Too bad it makes no attempt to separate the basic color words from the technical terms.

  22. per incuriam says:

    in the Irish conflict, the side that rejects orange has orange on their flag
    They don’t reject it, they embrace it, that’s the whole point of the flag (“cherishing all of the children of the nation equally”). Orangemen don’t much like this embrace of course.

    while the side that embraces orange has none on theirs
    That flag is presumably the Union Jack which is much displayed in NI to symbolize unity with Britain, just as the green, white and orange flag stands for Irish unity. There is symmetry insofar as the Union Jack contains an element representing Ireland.

    More striking surely is the absence of orange from the flag of the Netherlands. There can hardly be another country more strongly associated with a particular colour, certainly in the sporting arena where they even refer to their national teams as “Oranje”.

  23. Trond Engen says:

    It may seem even more striking than Italy’s failure to include azurro in their flag, but at least the flag of the Unified Nether Regions used to have an orange strip. Or maybe that was before the Regions were Unified. Anyway, at some point the standard was lowered and orange was replaced by red to improve visibility at sea. Soon red&white&blue became the colors to copy for every nation inspired by the freedoms of the Dutch republic. Back home the tricolor with orange maintained as a signifier of the royal family and by extension useful for occasions of special national or historical significance. And now it’s been hijacked by the darkest forces in the Dutch parliament.

  24. leoboiko: Berlin and Kay actually explored focal terms already. So the question they tried to pinpoint is not “What is the boundary between blue and green?” but “What is the best blue?” and “What is the best green?” In that sense, the Hanuno’o terms are color terms, even if they cover things that aren’t colored.

  25. Trond Engen says:

    Wikipedia once again can provide more details than me. I’d failed to notice the use of the flag by the WW2 collaborator movement, and even failed to remember Apartheid South Africa. So a considerable part of the hijacking took place before the current wave of darkening-rebranded-as-defence-of-light.

  26. David Eddyshaw says:

    On this subjects of flags, it has just occurred to me that there may be a Hatter who can put me out of my misery on an issue that has bugged me since I was about twelve:

    I was at a tiny travelling circus in northern Italy, and the crosstalk double-act team that filled in between animals and acrobats was telling jokes. One was about the Italian flag and the significance of the colours.
    “So: the red stands for the Communists, and the white for the Christian Democrats.”
    “OK! So what does the green stand for?”
    Unfortunately at this point my never-firm grasp of Italian gave out, and so I never got the punch line. It got a big laugh …
    I would imagine from the general venue and ambiance that it was probably a pretty corny old chestnut. Anyone know it?

    (This was of course long before there were such things as Green Parties.)

  27. gwenllian says:

    That flag is presumably the Union Jack

    My first thought was the Ulster Banner.

  28. per incuriam says:

    And now it’s been hijacked by the darkest forces in the Dutch parliament
    In the same way as the colour orange itself, which back in the benighted 17th century evoked the religious tolerance of the Dutch both at home and in their American colonies, was hijacked by the Orange Order, a protestant supremacist organization in Ulster, Scotland, Canada etc., directed at the exclusion and intimidation of catholics and others.

    @David Eddyshaw
    Perhaps the line played on “al verde” (“broke”), which often features in such jokes e.g. siamo al verde e passiamo le notti in bianco perché abbiamo i conti in rosso.

    The truth of course is that the flag is based on the pizza margherita (if not vice versa).

  29. The truth of course is that the flag is based on the pizza margherita

    Now that’s a flag I’d be proud to march under.

  30. ə de vivre says:

    Apparently, 青 is what brown used to mean: dark and shiny.

    Huh, “zagin” (𒍝𒆳 if you’ve got the right fonts installed), ‘lapis lazuli’, meant the same thing in Sumerian. Kings’ beards and Inana’s pubic hair (her “bearded mouth” as one incantation had it) in particular were apt to be described as “zagin”. Has anyone tried to connect Old Chinese to Sumerian yet? Maybe we could make some money off some overly credulous nationalists.

  31. @per incuriam, gwenllian: I meant both Unionist flags, and “reject” in the sense that prominently wearing or displaying orange – outside of the tricolor – would be an outgroup signifier for Nationalists.

  32. David Eddyshaw says:

    @per incuriam:

    Grazie mille. That looks really pretty plausible.

  33. David Marjanović says:

    Has anyone tried to connect Old Chinese to Sumerian yet?

    Yes: the polysynthetic grammar of Sumerian has a few intriguing similarities to those found in various Dené-Caucasian branches.

    Further comparison is, of course, hampered by two things: 1) Proto-Sino-Tibetan and its grammar have not been reconstructed. Progress can be expected now that it’s dawning on people that the Rgyalrong and Kiranti branches are generally more useful for that than Old Chinese, Old Tibetan and Old Burmese, but a lot of work remains to do, not least because the family is really large and many of its easily 250 languages are still poorly known to science. 2) The Sumerian writing system doesn’t say as much about the sound system as we’d like, so it’s not hard to read regular sound correspondences into the material…

  34. “Blue” for shiny black hair is used in English too, and while’s it’s old-fashioned, it’s not obsolete. I associate it with Robert E. Howard’s vaguely antisemitic character descriptions. I always figured that was the meaning in “Bluebeard” as well, although that was originally French.

Trackbacks

  1. […] Hat looks at how different languages name different colours over […]

Speak Your Mind

*