Languages with no word for blue

Short answer

Yes, even if you don't have a separate word for a color, you still see it: your eye and your visual system tell colors apart independently of language. But language has a small effect on how you group colors and how fast you tell two colors apart; when two colors fall on either side of a boundary in your language, you tell them apart a little faster, and this effect disappears once verbal thinking is kept busy. The debate isn't about whether language changes color perception, but about how deep and how lasting this effect is.

Languages split colors differently

The world's languages don't divide the spectrum in the same places. Some languages have a single word for green and blue; linguists call this "grue," a blend of green and blue. In other languages, the reverse happens: a region English names with a single word is named with two separate words. Russian is the most studied example of this: "goluboy" is used for light blues and "siniy" for dark blues, and for speakers the two aren't a color and its shade, the way an English speaker's "light blue" and "dark blue" are, but two separate colors.

In 1969, anthropologist Brent Berlin and linguist Paul Kay compared basic color terms across different languages and proposed that these terms don't appear at random but in a specific order: first light and dark, then red, then green or yellow, then blue, brown, and the rest. The theory's ladder, and a counter-example raised against it, is covered in RH-01. What matters here is that a separate word for blue comes relatively late in this ladder; that's why in many languages, blue shares a word with green or with dark colors.

Three Turkish words for blue

Turkish is an interesting example in this debate. In Old Turkish, the sky's color was called gök (sky); the word Turkish uses today, mavi (blue), entered later from Arabic. RH-02 and RM-03 already touched on this history. But today's Turkish color vocabulary has an even more surprising feature: lacivert, a distinct dark blue.

In 1998, Emre Özgen and Ian Davies ran experiments with Turkish speakers that measured consistency and salience in color-naming tasks. Their results suggested Turkish might have twelve basic color terms — one more than the eleven Berlin and Kay counted in languages like English. The candidate for the twelfth term was lacivert: a small region of color space was consistently named lacivert, and that region behaved like its own dark blue category with a focal point separate from mavi. The authors suggested that a new basic term might be forming at the intersection of blue and black.

This makes Turkish similar to Russian: both languages split the region English calls a single "blue" into two. In Turkish, the split falls between light and dark blue — between mavi and lacivert.

Does language change perception?

The real question is whether these word differences change how people see colors. This is called the linguistic relativity debate.

In 2007, Jonathan Winawer and his colleagues tested Russian and English speakers with a fast color discrimination task. Participants chose, as quickly as possible, which of several blue shades spread across the goluboy–siniy boundary matched a target color. Russian speakers discriminated two colors faster when they fell into different categories in their language than when they fell into the same category. This advantage disappeared when participants were simultaneously occupied with a verbal task — that is, when verbal interference was created — but it didn't disappear when they were occupied with a spatial task. The effect was stronger for colors that were very close to each other and hard to tell apart. No such boundary effect appeared in English speakers.

Another study, published a year earlier, looked at which half of the brain this effect was in. Aubrey Gilbert and colleagues found that color discrimination speed was affected by color names in the right half of the visual field — the half that mostly projects to the left hemisphere, where language processing largely happens — but not in the left half. This suggests the effect sits at a processing stage where language gets involved, not in the eye itself.

Berinmo, a language spoken in Papua New Guinea, offers an even more striking comparison. It has five basic color terms, and one of its boundaries doesn't sit where English draws its blue–green boundary, but somewhere else. In 2000, Debi Roberson and colleagues found categorical perception effects in both English and Berinmo speakers, but in each group only at the boundaries of their own language.

Opposing view

These findings don't support a strong claim like "language determines what you can see." No study shows that someone without a word for a color can't see that color, or can't tell it apart from its neighbor. The Russian speakers' advantage is on the order of milliseconds and disappears once verbal thinking is kept busy; this suggests the effect comes not from seeing itself, but from a verbal process that kicks in when sorting what's seen into categories.

On the other hand, defenders of the Berlin–Kay tradition point out that different languages' focal colors — the most typical example of each category — are strikingly similar across languages; this shows that color categories aren't entirely arbitrary, but draw on the structure of the visual system and the distribution of colors in the world. Today's overall picture isn't two-sided: language sits on top of a universal visual system and slightly tilts the information coming from it toward its own categories, especially in fast comparison and memory tasks.

Example — words for skin tones

An everyday example shows how color words have only a small effect on perception but a large effect on communication: the words people use for human skin.

Turkish uses words like buğday tenli (literally "wheat-skinned") or esmer ("olive" or "dark-skinned") to talk about skin tones; the cosmetics and fashion industries, on the other hand, have developed a much finer vocabulary to distinguish undertones and degrees of lightness. A makeup team or a photo studio needs this vocabulary to talk about the difference between "a slightly warmer, yellow-leaning medium tone" and "a pink-leaning light tone." Without the word, the difference is still seen; but describing that difference to someone else, building a consistent product range, and asking for a correction in a photo all become harder.

The vocabulary can also show who stands at the center. As TK-18 showed, calling a single light pink shade "skin color" for many years showed that this word was describing only one group. A vocabulary that can name different skin tones with equal precision also makes it easier to see and protect them with equal precision.

Common misconception

Summary

1. Languages divide the spectrum differently; in 1969, Berlin and Kay proposed that basic color terms emerge in a specific order, and in that order, a separate word for blue comes relatively late.
2. After gök and mavi, Turkish splits the blue region again with lacivert; Özgen and Davies showed that Turkish may have twelve basic color terms, much like Russian's goluboy–siniy split.
3. Experiments show that colors are discriminated faster when they fall on either side of a boundary in a language, but that this effect is small and disappears once verbal thinking is kept busy; even without a word for a color, you still see it.

Test yourself

1. (Easy) What are the two words Russian uses for light and dark blue?
2. (Easy) What is a basic color term?
3. (Medium) If you don't have a word for a color, do you still see it?
4. (Medium) Why is Özgen and Davies's finding about Turkish interesting from the standpoint of Berlin–Kay theory?
5. (Hard) What does it mean that the categorical advantage in Winawer and colleagues' experiment disappeared with a verbal task?
6. (Hard) How does having a fine-grained vocabulary for skin tones affect the communication of colors rather than their perception?

Answers

1. "Goluboy" for light blues, "siniy" for dark blues ("Languages split colors differently").
2. A color word in a language that isn't a subtype of another color word, is monolexemic, used by everyone, and isn't tied to a specific object ("Languages split colors differently").
3. Yes. The visual system tells colors apart independently of language; language's effect is limited to small, verbally dependent effects, such as colors being discriminated a little faster when they fall on either side of a boundary in a language ("Does language change perception?", "Opposing view").
4. It suggests one more than the eleven basic terms Berlin and Kay counted in languages like English — twelve terms with lacivert included — and that a new basic term might be forming at the intersection of blue and black ("Three Turkish words for blue").
5. That the effect comes not from the eye itself but from a verbal process that names and categorizes the colors seen; while verbal thinking is busy, language's categories can't contribute to the comparison ("Does language change perception?").
6. Differences are seen even without a word for them, but describing that difference to someone else, building a consistent product range, and asking for a correction all require a shared vocabulary; a vocabulary with equal precision also makes it easier to see and protect different skin tones with equal care ("Example — words for skin tones").

Sources

• Berlin, B., and Kay, P., Basic Color Terms: Their Universality and Evolution, University of California Press, 1969.
• Özgen, E., and Davies, I. R. L., "Turkish color terms: tests of Berlin and Kay's theory of color universals and linguistic relativity," Linguistics 36(5), 1998.
• Winawer, J., et al., "Russian blues reveal effects of language on color discrimination," PNAS 104(19): 7780–7785, 2007.
• Gilbert, A. L., Regier, T., Kay, P., and Ivry, R. B., "Whorf hypothesis is supported in the right visual field but not the left," PNAS 103(2): 489–494, 2006.
• Roberson, D., Davies, I., and Davidoff, J., "Color categories are not universal: replications and new evidence from a stone-age culture," Journal of Experimental Psychology: General 129(3): 369–398, 2000.

Next step

TK-14 moves from language to meaning: it discusses how the same color can carry meanings of mourning, marriage, luck, or danger in different cultures, how globalization has leveled out some of these meanings, and how to talk about color meanings without falling into stereotypes about cultures. PS-02, meanwhile, looks at the link between color and emotion: it examines what multi-country color–emotion studies have found, the distance between association and effect, and how lightness and saturation can be more decisive than hue.