In light of all the evidence, it seems to me that the balance of power between culture and nature can be characterized most aptly by a simple maxim: culture enjoys freedom within constraints. Culture has a considerable degree of freedom in dissecting the spectrum, but still within loose constraints laid down by nature. While the precise anatomical basis of these constraints is still far from understood, it is clear that nature hardly lays down inviolable laws for how the color space must be divided.* Rather, nature suggests optimal prototypes: partitions that are sensible given the idiosyncrasies of the eye’s anatomy. The color systems that are common among the world’s languages orbit within reasonable distance of these optimal partitions, but languages do not have to follow the prototypes to the letter, so nature’s guidelines can be supplemented or perhaps even overridden by cultural choices.
The explanation for Geiger’s sequence should also be sought in a balance between natural constraints and cultural factors. There is undoubtedly something biologically special about our relation to red: like other Old World monkeys, humans seem to be designed to get excited by it. I once saw a sign in a zoo that warned people dressed in red not to venture too close to the cage of a gorilla. And experiments with humans have shown that exposure to red induces physiological effects such as increasing the electrical resistance of the skin, which is a measure of emotional arousal. There are sound evolutionary reasons for this, since red is a signal for many vital things, most importantly danger (blood) and sex (the female baboon’s big red bottom, for example, signals she is ready for breeding).
But cultural reasons also contribute to the special status of red, and these ultimately boil down to the fact that people find names for things they feel the need to talk about. The cultural importance of red is paramount in simple societies, above all as the color of blood.* Moreover, as Gladstone suggested in 1858, the interest in color as an abstract property is likely to develop hand in hand with the artificial manipulation of colors, when color comes to be seen as detachable from a particular object. Red dyes are the most common and least difficult to manufacture, and there are many cultures that use only black, white, and red as artificial colors. In short, both nature and culture give red prominence over other colors, and this agreement must be the reason why red is always the first prismatic color to receive a name.
After red, yellow and green are next in line, whereas blue comes only later. Both yellow and green appear brighter to us than blue, with yellow by far the brightest. (As explained in the appendix, the mutation in the primate line that brought about the special sensitivity to yellow increased our ancestors’ ability to spot ripe yellowish fruit against a background of green foliage.) But if it was simply brightness that determined the interest in naming colors, then surely yellow, rather than red, would have been the first color to be given a separate name. As this is not the case, we should seek the explanation for the precedence of yellow and green over blue in the cultural significance of these two colors. Yellow and green are the colors of vegetation, and the difference between them (for example with ripe and unripe fruit) has practical consequences that one might want to talk about. Yellow dyes also happen to be relatively easy to make. The cultural significance of blue, on the other hand, is very limited. As noted earlier, blue is extremely rare as a color of materials in nature, and blue dyes are exceedingly difficult to produce. People in simple cultures might spend a lifetime without seeing objects that are truly blue. Of course, blue is the color of the sky (and, for some of us, the sea). But in the absence of blue materials with any practical significance, the need to find a special name for this great stretch of nothingness is particularly non-pressing.
A lot of water has flowed down the Scamander since a great Homericist, who occasionally dabbled in prime ministry, set off on an odyssey across the wine-dark sea in pursuit of mankind’s sense of color. The expedition that he launched in 1858 has since circled the globe several times over, been swept hither and thither by powerful ideological currents, and got sucked into the most tempestuous scientific controversies of the day. But how much real progress has actually been made?
It is a sobering thought that, on one level, we are hardly further advanced today than Gladstone’s original analysis of 1858. So sobering, in fact, that you would be hard-pressed to find contemporary accounts owning up to it. If you look up the subject in linguistic discussions, you will be lucky to find Gladstone mentioned at all. If he does make an entrance, he will be relegated to a perfunctory “pioneering efforts” footnote, reserved for those whom one feels one ought to mention but whom one cannot be bothered to read. And yet Gladstone’s account of Homer’s “crude conceptions of colour derived from the elements” was so sharp and farsighted that much of what he wrote a century and a half ago can hardly be bettered today, not just as an analysis of Homeric Greek but also as a description of the situation in many contemporary societies: “Colours were for Homer not facts but images: his words describing them are figurative words, borrowed from natural objects. There was no fixed terminology of colour; and it lay with the genius of each true poet to choose a vocabulary for himself.” In one oft-quoted passage, for example, the anthropologist Harold Conklin explained why the Hanunoo in the Philippines call a shiny, brown-colored section of newly cut bamboo “green”—essentially, because it is “fresh,” which is the main meaning of the “green” word. Conklin probably never set eyes on Gladstone’s explanation for why Homer used chlôros for brownish fresh twigs. But anyone comparing their analyses might be forgiven for thinking that Conklin simply lifted his passage wholesale from Studies on Homer and the Homeric Age.
What is more, Gladstone’s fundamental insight that the opposition between bright and dark was the primary basis for the Homeric color system could also stand virtually unimproved at the cutting edge of current thinking on the development of color vocabulary. Not that anyone would admit nowadays that the insight is Gladstone’s, mind you. In modern accounts, the idea that languages shift the emphasis from a brightness-based system toward hue is presented as a shiny new and ultramodern theory. But while this modern theory is far more impressive than the old one in the complexity of its terminology, in actual content it offers little that cannot be found in Gladstone’s original analysis.
But perhaps the greatest irony in the whole story is that even the seemingly infantile evolutionary model that Gladstone invoked at the very beginning of the color debate was actually spot-on. The Lamarckian evolution-through-stretching mechanism is a perfect way to explain the changes between Homer’s time and ours—if only we overlook one little detail, namely that Gladstone thought he was describing biological developments. For while the Lamarckian model, whereby the acquired aptitudes of one generation may become the inherited and inborn aptitudes of another, is a ridiculous way to explain anatomical changes, it is a perfectly sensible way to understand cultural evolution. In biology, characteristics acquired within the lifetime of an individual are not passed on to the offspring, so even if exercising the eye could improve one’s own sensitivity to colors, the improvement would not be genetically transmitted to the next generation. But the Lamarckian model does fit perfectly with the reality of cultural developments. If one generation exerts its tongue and “stretches” the language to create a new conventional name for a color, then the children will indeed “inherit” this feature when they learn the language of their parents.
So Gladstone’s assertion that the developments in the vocabulary of color involved the “progressive education” of mankind is in actual fact entirely correct, and so is his belief that “Homer’s organ” still needed to be trained in the discrimination of colors. It is only that Gladstone did not realize which human faculty underwent this progressive education and which organ it was that needed to be trained. And it is exactly in clarifying this troublesome question, in telling apart the eye from the tongue, education from anatomy, culture from nature, that substantial headway has been made in the century and a half–long debate. It is here that our vie
w has sharpened since the culture blindness of Gladstone in 1858, of Geiger in 1869, of Magnus in 1878, and of Rivers in 1903, but also since the nature blindness of Leonard Bloomfield in 1933 (languages mark off color boundaries “quite arbitrarily”) and of Verne Ray in 1953 (“there is no such thing as a ‘natural’ division of the spectrum”), and even since the culture myopia of Berlin and Kay in 1969.
BEYOND COLOR
The fighting over the rainbow may have been fiercer and more prolonged than over any other concepts, but the insights that have emerged from the debate can be applied with equal benefit elsewhere in language. The framework of freedom within constraints, which I suggested above, provides the best way to grasp culture’s role in shaping the concepts of language more generally, and even its grammatical system.
Different cultures certainly are not at liberty to carve up the world entirely at whim, as they are bound by the constraints set by nature—both the nature of the human brain and the nature of the world outside. The more decisive nature has been in staking out its boundaries, the less leeway there is for culture. With cats and dogs and birds and roses, for instance, culture hardly has any freedom of expression at all. We can be quite certain that in any society where there are birds and roses, there will be words that correspond to our “bird” and “rose,” and there will not be words that correspond to the Ziftish “rird” and “bose.” Even if one tried to construct an artificial language brimful of unnatural Ziftish concepts, it is not clear that children would be able learn these. For obvious humanitarian reasons, the experiment has not been conducted, but if ever anyone is cruel enough to raise young children on a monolingual diet of rirds and boses, dats and cogs, steaves and lones, the result will probably be that the hapless children will fail to learn these concepts “correctly” and instead impose an “incorrect” interpretation with more sensible and more natural meanings, which will correspond to our birds and roses, cats and dogs, leaves and stones.
On the other hand, when nature has shown even the slightest dithering or fuzziness in marking its boundaries, different cultures have far more sway over the division of concepts than anyone exposed only to the conventions of one society would imagine. Of course, concepts must be based on some sensible logic and internal coherence if they are to be both useful and learnable. But within these limits, there are still many ways of dissecting the world that are perfectly sensible, perfectly learnable by children, perfectly suitable for the communicative needs of the speakers—and yet totally different from what we are used to.
The field of color made it glaringly obvious that the unfamiliar may not always be unnatural. A language in which yellow, light green, and light blue are treated as shades of one color, for instance, may seem to us almost incomprehensibly alien, but this division makes perfect sense within a system whose primary emphasis is on brightness rather than hue and where the main prismatic color to be set apart is red, so that all bright hues that have no tinge of redness naturally belong to the same concept.
But there are many other examples of the discrepancy between what is unnatural and what is merely unfamiliar. We will encounter one striking but little-known case in a later chapter: the concepts used to describe space and spatial relations. A more famous example is kinship terms. The language of the Yanomamö Indians in Brazil, for instance, appears to us incomprehensibly hazy, because it lumps together relatives of entirely different kinds under one concept. Using one and the same term, šoriw∂, for both cousins and brothers-in-law may already seem rather peculiar. But this is nothing compared with the unification of brothers and certain cousins: the Yanomamö term εiw∂ makes no distinctions between one’s own brothers and the sons of a paternal uncle or of a maternal aunt! On the other hand, the Yanomamö would consider English unbearably vague in having just one term, “cousin,” which lumps together no less than four distinct type of relatives: amiw∂ (daughter of a paternal uncle or of a maternal aunt), εiw∂ (son of paternal uncle or of maternal aunt), suw∂biy∂ (daughter of maternal uncle or of paternal aunt), and šoriw∂ (son of maternal uncle or of paternal aunt). There are even weirder systems of kinship terms, such as the one that anthropologists call the Crow system, in which the same concept is used for one’s own father and for some of one’s cousins (the sons of a paternal aunt). All these ways of dividing up one’s relatives have their own internal logic and coherence, but they nevertheless diverge radically from the categories that we find natural.
The freedom of culture is even more pronounced in the realm of grammar, since grammatical structures are by nature more abstract and, as we have seen, nature’s hold loosens considerably in the realms of abstraction. One striking aspect of the grammatical system that varies even among mainstream languages is the order of words. Japanese and Turkish, for instance, arrange words and grammatical elements in a way that seems to us perversely back to front. In The Unfolding of Language, I discussed examples such as the Turkish sentence Padiah vezir-ini ordular-ι-nιn ba-ι-na getirdi, where a literal translation of each element—“Sultan vizier his troops his of head their to brought”—is almost as unenlightening to an English speaker as the Turkish itself. But for a Turkish speaker encountering English for the first time, the English arrangement—“the Sultan brought his vizier to the head of the troops”—would appear just as peculiar.
While the extent of variation among different grammars is not contested, there have been vociferous arguments about how to interpret it. The divergence between grammatical systems poses a particular challenge to the nativist idea of an innate universal grammar, because if the rules of grammar are meant to be coded in the genes, then one could expect the grammar of all languages to be the same, and it is then difficult to explain why grammars should ever vary in any fundamental aspects. One influential nativist response to this challenge has been the theory of “parametric variations” within universal grammar. According to this idea, the genetically coded grammar contains a few “parameters,” that is, a small set of preprogrammed options that can be thought of as “on-off” switches. Children who acquire their mother tongue, so the argument runs, do not need to learn its grammatical rules—their brains simply set the preprogrammed parameters according to the language they happen to be exposed to. Nativists have claimed that different settings of these few switches must account for the whole variation in grammatical structures across the world’s languages. The only freedom that different cultures are accorded is thus to decide on how to set each of the parameters: press a few switches one way and you’ll get the grammar of English, set a few switches the other way and you’ll get the grammar of Italian, and flip a few more and you’ll get the grammar of Japanese.
The theory of parameters has met with much criticism and some ridicule among non-nativists, who maintain that the scope of variation among the world’s languages is far too wide to be covered by a few parameters, and that from an evolutionary perspective it is exceedingly unlikely that a genetically determined grammar would emerge with such a set of switches (whatever for?). But the main argument against the theory of parameters is that it is merely a convoluted way to account for grammatical variation that can be explained far more simply and far more easily if one does not insist on believing that specific grammatical rules are innate.
In short, the adamant claims of nativists about the innateness of grammar have met with equally resolute opposition from culturalists. The controversy over grammar has thus produced a most impressive pile of paper over the last decades, and many a library shelf across the globe quietly groans under its burden. This book will not add much weight to the debate, because it concentrates on the concepts of language rather than on grammar. But there is one aspect of the grammatical system that nonetheless cries out for attention, precisely because it has—wholly unjustifiably—escaped the controversy almost entirely: the complexity of the grammatical system. On this subject, an eerie consensus prevails among linguists of all creeds and persuasions, who unite in severely underestimating the influence of cultur
e.
* In 2007, three researchers, Terry Regier, Naveen Khetarpal, and Paul Kay (same one), made a tentative suggestion for explaining the nature of these anatomical constraints. They started from the idea that a concept is “natural” if it groups together things that appear similar to us, and they argued that a natural division of the color space is one in which the shades within each color category are as similar to one another as they can be and as dissimilar as possible from shades in other categories. Or put more accurately, a natural division maximizes the perceived similarity between shades inside each concept and minimizes the similarity between shades that belong to different concepts. One might have imagined that any division of the spectrum into continuous segments would be equally natural in this respect, because neighboring shades always appear similar. But in practice, the accidents of our anatomy make our color space asymmetric, because our sensitivity to light is greater in certain wavelengths than in others. (More details can be found in the appendix.) Because of such non-uniformities, some divisions of the color space are better than others in increasing the similarity within concepts and decreasing it across concepts.
* In many languages the name of the color red actually derives from the word “blood.” And as it happens, this linguistic connection has exercised the minds of generations of biblical exegetes, because it bears on the name of none other than the father of mankind. According to the biblical etymology, Adam owes his name to the red tilled soil, adamah, from which he was made. But adamah derives from the Semitic word for “red,” adam, which itself comes from the word dam, “blood.”
Through the Language Glass: Why the World Looks Different in Other Languages Page 11