The Sunday Times - Published: 20 June 2010. Sam Leith

Through The Language Glass by Guy Deutscher

Even our most basic assumptions are called into question in this rich and provocative look at how language affects the way we see the world

This book seeks to answer a question that has preoccupied evolutionists, anthropologists and linguists for well over a century: to what extent is the way we see the world affected by the language we speak? Are the categories into which we organise our experience dictated by our mother tongues, or vice versa? Which came first, in other words: the chicken or “the chicken”? The obvious answer to this would be that perception is a matter for nature, and language one for culture — but this fabulously interesting book describes an area of intellectual history replete with brilliant leaps of intuition and crazy dead-ends.

Guy Deutscher, who combines enthusiasm with scholarly pugnacity — “How could such piffle be spouted by sober scientists?” he asks of one orthodoxy — is a vigorous and engaging guide to it.

Popular in the first half of the last century was the Sapir-Whorf hypothesis. It argued that our view of the universe is wholly determined by the language we speak, and that in any given language certain concepts are literally unthinkable. To bolster his argument, the hypothesis’s co-creator, anthropologist Benjamin Whorf, offered a compelling example: “after long and careful study and analysis”, the language of Arizona’s Hopi Indians “is seen to contain no words, grammatical forms, constructions, or expressions that refer directly to what we call ‘time’ ”.

Unfortunately, “long and careful study” hadn’t taken Whorf as far as Arizona. In 1983, the linguist Ekkehart Malotki — after extensive actual fieldwork — published a book called Hopi Time in which the first sentence quoted was “pu’ antsa pay qavongvaqw pay su’its talavay kuyvansat, pàasatham pu’ pam piw maanat taatayna”. Translation: “Then indeed, the following day, quite early in the morning at the hour when people pray to the sun, around that time then, he woke up the girl again.”

Sapir-Whorf was so discredited that even now, says Deutscher, to ask whether language can shape perception “carries with it a baggage of intellectual history which is so disgraceful that the mere suspicion of association with it can immediately brand anyone a fraud”.

Whorf’s charlatanism has done intellectual history a disservice. For, says Deutscher, the anti-Whorf backlash went too far. Deutscher’s is a relatively modest claim — that one’s native tongue does have a marginal effect on one’s perceptions — but it is a profoundly interesting one. And it cleverly stands the reasoning behind Sapir-Whorf on its head.

There’s no evidence that any language fundamentally restricts what can be expressed in it, he says. Rather, the question is: what does a given language force its users to express, and thereby force them to pay attention to?

In English, for instance, you can say you visited your neighbour without specifying whether that neighbour was male or female. In French, the choice between voisin and voisine is inescapable — giving your trip next door to borrow a cup of sugar a Gallic frisson of sexual awareness.

Evidence suggests the gendering of inanimate objects does have an effect on their mental associations. In German, the word for “bridge” is feminine — and in tests German speakers have tended to describe bridges as “beautiful, ­elegant, fragile, peaceful, pretty and slender”. In Spanish, where “bridge” is masculine, testers heard “big, dangerous, long, strong, sturdy, towering”.

In some cases these associations may be less straightforward. Deutscher explains that for linguists the term “gender” doesn’t just apply to male and female: “The African language Supyire from Mali has five genders: humans, big things, small things, collectives and liquids…and the Australian language Ngan’git-yemerri is said to have 15 different genders, which include, among others, masculine human, feminine human, canines, non-canine animals, vegetables, drinks, and two different genders for spears (depending on size and material).”

Bizarre and fascinating, too, is the question of how directions are described. In English, we can navigate by “egocentric co-ordinates”, based on the speaker’s position (left, right, forwards and backwards), and also by “geographical co-ordinates” such as north, south, east and west.

The Australian aboriginal language of Guugu Yimithirr is one of a handful of languages that do not use “egocentric co-ordinates” at all, and navigates simply by north, south, east and west. A Guugu Yimithirr speaker, for instance, might report that Long John Silver was missing his northwesterly leg. Depending, that is, on the orientation of the television on which he watched Treasure Island.

The speaker of Guugu Yimithirr has to know at all times which way is north — otherwise expressing himself becomes more or less impossible. (Tests have found that if you put him in a darkened room and spin him round until he is sick, he will still be able to orientate himself.) Here, then, is a linguistic accident that surely has a direct effect on the way the speaker perceives the world. Then there’s the question of colour. The sky is blue, right? You’d be surprised. What seems to you or me natural, fundamental and obvious turns out to be far from any of those things. In ancient texts, for instance, across a wide range of cultures — including the Old Testament, Homer and the Indian Vedas — the daytime sky seems to be no colour at all. Neither Homeric Greek nor biblical Hebrew even contains a word for blue.

Indeed, the oddities of Homer’s colour descriptions - the sea is “wine-dark” and sheep purple, but grass is never green nor the sky blue - led William Gladstone to argue that the ancients were effectively colour blind, discerning brightness but not hue. Ingenious, but wrong. Anthropologists have encountered tribes who - when pressed – will identify the sky as “black” or “white”. Yet these tribesmen can pick a pebble of blue lapis out of a bowl of red coral without trouble.

And Deutscher’s own daughter, who by 18 months could identify and name blue objects with ease, still thought the sky was “white”.

For a physicist, the visible spectrum is an even continuum of different wavelengths. But the way we divide in into “colours” and the extent to which our languages regard some distinctions in hue as worth noticing - is not simple. It belongs, it turns out, to both nature and culture.

Developing languages appear to acquire colour vocabularies in order of importance - red, the colour of blood and baboon sex, always comes first; blue, rate in nature, comes last. Their speakers, although equipped to perceive the whole visual spectrum, tend to make only those distinctions they have learnt to name.

Through the Language Glass is full of just such blue-sky thinking. It’s a remarkable rich, provocative and intelligent work of pop science.

Through the Language Glass

Author: Guy Deutscher

Publisher: William Heinemann

Length 320 pages

Price £18.00

Hint of a tint

Deutscher offers a striking example of how the vocabulary of colour in different languages can cause differences in the actual perception of colour. English speakers use ‘blue’ to cover everything from navy to cyan; Russians identify dark blue as siniy, and light blue as goluboy. Experiments done on the two language groups show that Russian speakers find it easier to distinguish shades across the siniy/goluboy divide than English speakers.