Hanunoo Color Categories
In an important paper published in Southwestern Journal of Anthropology in 1955, Harold Conklin presented a description of the color categorization system of Hanunoo, a Philippine language spoken in the southern part of the Mindoro island (see map below).
According to this description, Hanunoo has only four basic or “Level I” color terms (here and below, I am ingoring some of the diacritic marks in Hanunoo words as well as in the name of the language itself): (ma)biru (BLACK), (ma)lagti? (WHITE), (ma)rara? (RED) and (ma)latuy (GREEN). Additional color distinctions are made by using color terms from the “Level II” list, consisting of hundreds of “specific color categories, many of which overlap and interdigitate”. He further noted that these Level I terms refer not only to a kind of hue, but also to other perceptual properties of objects, such as wetness or dryness, freshness etc. For example, “a whiny, wet, brown-colored section of newly-cut bamboo is malatuy (not marara?)”. The resulting intracultural schema is shown below:
This description fits nicely with the later seminal work by Berlin and Kay (1969) who proposed that there is a certain order in which color categories are added to the system: according to them, languages with two basic color terms have BLACK and WHITE, with three — also RED, and with four — YELLOW or GREEN. A language with five color terms with have both YELLOW and GREEN. The sixth color to be added is BLUE, then BROWN, then PINK, ORANGE, GREY and PURPLE (in any order).
While a given language is more likely to add new color terms rather than disuse the old ones, the latter is by no means impossible. However, Berlin and Kay’s proposal has often been misunderstood to imply a strict order of historical/evolutionary development from simpler, more primitive to more sophisticated systems. For example, Charles O. Frake in his commentary on Conklin’s paper says:
“To say that there is an implicational order (and, according to Berlin and Kay, an evolutionary order) in the number of basic terms in human languages from two… to eleven or so… is not a count of the color words in a language but a measure of the inclusiveness, the generality, of the most inclusive terms. The order of implication/evolution here is not a progression from a primatively few color words to a civilized many, but a regression from generality to specificity — abstractness to concreteness — in talking (and thinking? about color. In color, it is the civilized Westerners who come out to сиу concrete-minded clods unable to recognize, for example, the similarities that make “grue” [green and blue] a labeled category for so many of the less evolved folk (including, in this case, the ancient Greeks.”
In other words, Frake and his fellow anthropologists try to save the so-called “technologically primitive peoples” (itself a questionable concept for sure) from the stigma of being also linguistically primitive.
Yet, it is thoroughly incorrect to ascribe to linguists such as Berlin and Kay the claim of linguistic primitiveness of peoples like Hanunoo. Earlier research on color terminology, which predates Berlin and Kay’s study, seems to indicate that (certain) ancient peoples made fewer color distinctions (cf. for example Gladstone’s 19th century work on Homeric Greek). However, historical development is not what Berlin and Kay are trying to describe and explain. All that they’ve shown is that only certain combinations of basic color terms are possible. For example, as mentioned above, a language with three color terms will have BLACK, WHITE and RED, and not YELLOW, RED and PURPLE, or BLACK, WHITE and BLUE, or any other logically possible combination. The same is true of languages with any other number of basic color terms. For example, a language with six basic color categories will not have BLACK, WHITE, PINK, PURPLE, ORANGE and BROWN.
In this respect, Berlin and Kay’s work fits very nicely with later work on language variation in a number of linguistic frameworks, which all share an understanding that while languages differ widely in a great number of lexical and grammatical respects, this cross-linguistic variation is not boundless. Thus, the primary goal of a contemporary linguist interested in cross-linguistic variation (and most linguists are!) is to describes the bounds of variation, or in other words, what is a possible as a human language and what is impossible. These parameters of cross-linguistic variation then need to be explained, but even establishing them is an achievement in and of itself. And this is exactly what Berlin and Kay did for color terminology.