Through the Language Glass: How Words Colour Your World

Does language mirror the objects and structures inscribed in reality, or is it rather that reality is reified, structured and even invented by language? Is what we assume to be naturally and commonsensically out there merely a matter of linguistic convention? And if so, given the variety of languages and forms of language, why should we ("we" here being the European "we") presume that it is Indo-European languages that pick out the necessary saliences and paint the only correct world picture?

Guy Deutscher's light-hearted survey of these linguistic and anthropological debates begins in 1858 with the question posed by William Gladstone, the British prime minister, in his monumental work on Homer. Why, asked Gladstone, does Homer speak of "the wine-dark sea", describe honey as green, and sheep and iron as violet? Why are his imaginary skies never blue, despite the real blueness that must have dazzled him? This deficient discrimination cannot be blamed on Homer being blind or colour-blind, since other ancient Greek writers seemed to share it, along with - as scholars soon pointed out - the authors of the Indian Vedas, the Bible and early Chinese texts.

Gladstone surmised that there was a universal anatomical deficiency in the ancient world, and that colour discrimination evolves. Initial awareness of black and white graduates to red, and then to other shades, culminating in blue. This sort of answer was eventually rejected, however: evolutionary change was found to be far more gradual than originally supposed, and unaffected by changes over a few generations. Humans must have had the same degree of colour vision for millennia.

But if colour distinctions are not determined by anatomy, are they formed by language? Perhaps, speculated anthropologists, "primitive" languages poor in colour categories produce speakers with correspondingly limited colour discrimination. Perhaps it is culture, rather than vision, that evolves, and linguistic form that controls how we perceive.

The politically correct flip side to this idea emerged in the 1950s in the theory of linguistic relativity proposed by American linguist Benjamin Lee Whorf. One of his claims was that Europeans are so blinkered by their languages' noun/verb dichotomy that they artificially split reality into things and events; whereas Hopi Indians, with their noun-lacking language, are not forced to read fictitious agents into nature. They can conceive reality as processes, and thus are more receptive to Einstein's theory of relativity.

Nowadays, however, linguistic relativity is largely rejected. Equally, anthropological studies have shown that speakers of colour-rich and colour-challenged languages discriminate more or less the same colours. Deutscher agrees with the current consensus, but argues that even though Whorf is wrong - and that our perception is indeed constrained by nature - languages do, in fact, variously provide some room for manoeuvre within these constraints. Plus, of course, they impose constraints of their own.

He describes experiments that suggest that speakers tend to make male or female associations with nouns depending on the male or female genders these nouns have in the speaker's native language, while linguistic studies on Australian Aborigines subvert the age-old assumption that all humans naturally orient themselves in subjective relation to their bodies ("it's on the right-hand side of the table at the back"), and only secondarily via the objective points of the compass. Speakers of Guugu Yimithirr, it turns out, use only the more objective, sophisticated geographical coordinates ("it is on the southern edge of the northern table"), not the primitive egocentric ones that we use.

What follows from this? These findings are, like the book as a whole, immediately intriguing but ultimately less than illuminating. Deutscher thinks that writing for the uninitiated obliges him to be skittish and facetious, which he is - at the expense of stringent theorising.

Although it is all about discrimination in perception and language, Through the Language Glass is disappointingly short on the fine philosophical distinctions these questions demand.