Saturday, September 11, 2010

Words Cannot Express

New York Times Book Review

Is language first and foremost an artifact of culture? Or is it largely determined by human biology? This issue has been argued back and forth for a couple of centuries with no clear resolution in sight. Guy Deutscher’s 2005 book “The Unfolding of Language” placed him firmly in the pro-culture camp. Now, in his new book, “Through the Language Glass,” he examines some idiosyncratic aspects of particular languages that, in his opinion, cast further doubt on biologically based theories of language.

Deutscher starts with the puzzling fact that many languages lack words for what (to English speakers) seem to be basic colors. For anyone interested in the development of ideas, Deutscher’s first four chapters make fascinating reading. Did you know that the British statesman William Gladstone was also an accomplished Greek scholar who, noting among other things the surprising absence of any term for “blue” in classical Greek texts, theorized that full-color vision had not yet developed in humans when those texts were composed? Or that a little-known 19th-century philologist named Lazarus Geiger made profound and surprising discoveries about how languages in general divide up the color spectrum, only to have his discoveries ignored and forgotten and then rediscovered a century later? Did you know that Siegfried Sassoon’s World War I psychiatrist, William Rivers, carried out the earliest psychological experiments to test the precise relationship between the colors people could name and the colors they actually saw?

Deutscher does not merely weave little-known facts into an absorbing story. He also takes account of the vast changes in our perceptions of other races and cultures over the past two centuries. Although the strange sequence in which color terms appear in the world’s languages over time — first black and white, then red, then either green or yellow, with blue appearing only after the first five are in place — still has no full explanation, Deutscher’s suggestion that the development of dyes and other forms of artificial coloring may be involved is as convincing as any other, making color terms the likeliest candidate for a culture-induced linguistic phenomenon.

But then Deutscher switches to another issue entirely, that of linguistic complexity. He brings off a superb “emperor has no clothes” moment by demonstrating that the “fact” (attested in countless linguistic texts) that all languages are equally complex has no empirical basis whatsoever. Moreover, as he points out, such a claim could not be made even in principle, since there are no objective, nonarbitrary criteria for measuring linguistic complexity across entire languages.

Deutscher then goes on to addresses the relationship between language and thought. Do speakers of all languages think in similar ways, or do different languages give their speakers quite different pictures of the world (a view sometimes referred to as “linguistic relativity”)? Deutscher rejects linguistic relativity in its strong form, pouring scorn on its most vehement defender, the early-20th-­century linguist Benjamin Whorf, and again firmly locating his account in the cultural-historical background. His skepticism extends even to promising cases like that of the Amazonian language Matses, whose arsenal of verb forms obliges you not only to explicitly indicate the kind of evidence — personal experience, inference, conjecture or hearsay — on which every statement you make is based, but also to distinguish recent inferences from older ones and say whether the interval between inference and event was long or short. If you choose the wrong verb form, you are treated as a liar. But the distinctions that must be expressed by verbal inflections in Matses, Deutscher argues, can all be easily understood by English speakers and easily expressed in English by means of circumlocutions.

Deutscher does find three areas where a weaker version of linguistic relativity might hold — color terms, spatial relations and grammatical gender. Ever since Mark Twain mocked the pronoun confusions of “the awful German language” — a young girl is an “it” while a turnip is a “she” — most people, including linguists, have treated gender assignment as largely arbitrary and idiosyncratic, devoid of any cognitive content. But recent experiments have shown that speakers do indeed, on a subconscious level, form associations between nonliving (“neuter”) objects and masculine or feminine properties. As for spatial relationships, we English speakers relate the positions of objects or other people to ourselves (“in front of,” “behind,” “beside”) or to each other, but some languages use compass references (“east of,” “southwest of”) for identical relationships. Deutscher argues that repeated use of such expressions forces speakers of these languages to develop an internal cognitive compass, so that regardless of where they are and what they are facing, they automatically register the location of the cardinal points.

Deutscher presents his material in a chatty and accessible (if sometimes verbose) style, and if he had left things at that, he would have written just the kind of language book most readers love — heavy on quirky detail, light on technicalities and theory. But he also burdens his findings with more theoretical weight than they can bear.

First, the facets of language he deals with do not involve “fundamental aspects of our thought,” as he claims, but relatively minor ones. Things like location, color and grammatical gender hardly condition our thinking even in the day-to-day management of our lives, let alone when we address issues of politics, science or philosophy. Moreover, with the possible exception of color terms, cultural factors seldom correlate with linguistic phenomena, and even when they seem to, the correlation is not causal. For instance, languages of small tribes tend to have words with multiple inflections, while those of complex industrial or post­industrial societies do not. However, this phenomenon is not directly caused by differing degrees of social complexity. Rather, complex societies tend to have much larger and more ethnically diverse populations, hence they experience far more interactions between native speakers of different languages and dialects. It is this factor that encourages simplification and erodes word endings.

Take a hypothetical correlation that ­really might have cultural causes. Suppose relative clauses appeared only when a society entered the market economy. Any such finding would revolutionize our understanding of the interface between language and culture. But not only has no such relationship ever been demonstrated, nothing remotely like it has ever been found.

Explaining why he rejects biologically based explanations of language, Deutscher states that “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.” Actually, it’s quite easy. Simply suppose that biology provides not a complete grammar, but rather the building blocks out of which such a grammar can be made. That is, in fact, all biology could be expected to do. With physical organs, biology can mandate — two legs instead of four, five fingers instead of six. But when it comes to behavior, biology cannot mandate. It can only facilitate, offering a range of possibilities from which culture (or more likely, sheer chance) can choose.

Fortunately, relatively little of “Through the Language Glass” is devoted to these issues. Readers can ignore Deutscher’s broader claims, and enjoy the little-­trodden linguistic bypaths along which he so knowledgeably leads them.


Bickerton, Derek. 2010. "Words Cannot Express". New York Times. Posted: September 3, 2010. Available online:

No comments: