American Scientist

The Power of Babel: A Natural History of Language. John McWhorter. vi + 327 pp. Times Books, 2001. $26.

The Atoms of Language: The Mind's Hidden Rules of Grammar. Mark C. Baker. xii + 276 pp. Basic Books, 2001. $28.

There are a couple of facts about languages that everyone should know. The first is that they are extraordinarily diverse. Obviously, they sound different—anyone can tell Japanese from Spanish from Finnish—and every child knows that different languages call things by different names. But this is just the beginning. Every language contains its own metaphysics—tacit commitments as to what parts of reality are worth talking about and how these parts should be packaged. John McWhorter's deeply entertaining The Power of Babel discusses several examples of this. When we say in English, "He is chopping trees," we are forced to be clear about when the chopping takes place (is—right now) and whether there is one tree or more than one (trees—more than one). In other languages, tense and number are not typically marked, and in Tuyuca, spoken in the Amazon rain forest, a sentence must encode how the speaker came to know the information that the sentence expresses: Did he or she see the tree-chopping, infer that it happened or hear about it from someone else? (Tuyuca is a trial lawyer's dream.) In English, you might say "I sank in mud up to my ankle." The German word-for-word translation is subtly different ("I am until the ankles in the mud sunk"), the Japanese translation a bit more unusual ("mud of within at ankle until soaked put-away"), and the two-word Ojibwe equivalent is downright bizarre (first word: "knob-bone-at"; second word: "I-extending-to-mud-moved").

How did the 6,000 or so human languages get this way? The approach McWhorter takes (following Darwin's lead) is to understand the evolution of language as being akin to the evolution of species, complete with winnowing, competition, adaptation and extinction. This analysis leads to some surprises. For one thing, although all languages are highly complex, some are more complex than others. (There is no simple metric with which to judge this, but a crude measure is how hard the language is to learn.) The most complicated languages are those you probably have never heard of, spoken by small hunter-gatherer groups. These are isolated for long periods of time, and thereby come to acquire, as McWhorter puts it, "accreted layers of decorative gunk." (At the other extreme are creoles, languages built up in a single generation, which are crisp, clean and relatively simple.) Languages like English have their own problems—they have writing systems. Literacy has obvious advantages, but it also instills stilted and unnatural patterns into people's speech and helps freeze languages in time, creating grammatical and lexical fossils.

The Power of Babel is sprawling, gossipy and fascinating. McWhorter has a skill at answering questions you have never thought to ask: When Ginger Rogers said that Fred Astaire "made love to me" in a 1935 movie, what precisely did she mean? Why is Charlie Brown bald? Do Germans fully appreciate the translated version of the American TV sitcom Married . . . with Children? Why does Mickey Mouse wear white gloves? Surprisingly, the answers bear in interesting ways on the nature of language and linguistic diversity.

The second fact you should know about languages is that they are basically the same. Noam Chomsky revolutionized linguistics—and the social sciences more generally—by maintaining that all languages share the same basic design features because language arises from an innate "language organ" that all neurologically normal humans possess. From this perspective, English, Tuyuca and Ojibwe are nothing more than minor variations on a theme.

This is the framework adopted by Mark Baker in The Atoms of Language. He pursues an analogy between chemistry and linguistics, suggesting that diversity is best explained in terms of a set of building blocks that interact in complicated and constrained ways. In chemistry you have atoms; in linguistics, you have parameters—basic choice-points in how universals can be expressed. Hence Chomsky is "the Democritus of modern linguistics."

In much of his book, Baker explores this approach by looking at the very hardest cases. He compares English with the language that is perhaps most different—Mohawk, spoken by a few thousand Native Americans in Quebec, Ontario and New York. Mohawk is a polysynthetic language. Whereas English speakers communicate complex ideas by putting words into sentences, Mohawk speakers just construct bigger words. Mohawk contains no stand-alone words such as dog and eat; if you were taking Mohawk 101, the first word you would learn might be katerihwaiénstha': "I habitually cause myself to have ideas," which means, roughly, "I am a student."

Baker makes two arguments about Mohawk and English. First, they really are very similar—he lists eight basic grammatical properties that Mohawk and English share, such as "Agents are subjects, and undergoers are objects." Second, the contrast between Mohawk and English is due to just a few subtle differences, most of them captured in terms of a single "polysynthesis parameter." The terminology here might sound forbidding, but Baker does an excellent job of conveying technical ideas in a way that is accessible and yet not patronizing. This is one of the best introductions to the formal approach to language I have ever read.

There is little overlap in the specific phenomena that these two books discuss, but their theoretical approaches are entirely consistent. McWhorter does not reject the notion of a language organ, and he is sympathetic toward the intellectual enterprise of studying the formal universals of language (although somewhat ambivalent about the status that this research program has in the field of linguistics, which he summarizes as "Chomsky-smart/other-also-ran"). Baker is plainly interested in the nature of linguistic diversity, and he grants the need for a theory of the forces that flip the parameters in various ways across linguistic history. Both authors are intensely concerned with the fate of marginalized and dying languages.

They also agree on certain things that might surprise nonlinguists. They are both skeptical about the commonsense notion of "language." There is a temptation to treat categories such as "English" and "Chinese" as bona fide real entities, to be clearly distinguished from dialects and from "mixes" such as Spanglish. But in fact, what gets counted as a language is a matter of social and political considerations and has little to do with the sort of linguistic facts that linguists are interested in. Danish and Swedish are counted as separate languages even though they are far more similar than Cantonese and Mandarin, which are commonly viewed as dialects of the same language, Chinese. As the linguist Uriel Weinreich put it, "a language is a dialect with an army and a navy."

McWhorter and Baker are also adamant that culture plays a limited role in explaining linguistic diversity. Plainly, some aspects of language connect with the beliefs, values and technology of the language's speakers. This explains why English has modem, whereas Tuyuca does not. This explains what the psychologist Steven Pinker has dubbed the "euphemism treadmill," which is what happens when a word becomes tainted by negative connotations, and so a new word is coined, but that new word gets tainted, and so on, as in slum giving way to ghetto and then inner city.

But for the most part, to use McWhorter's analogy, languages do not reflect culture any more than a pattern of spilled milk reveals anything about the shape of the bottle it came from. If you classify languages in terms of their grammar, you get mush from a cultural viewpoint—German puts verbs at the end of sentences, just like Hindi, Japanese, Mongolian and Nama, in sharp contrast to languages spoken by more culturally kindred groups such as the English and the Portuguese. In French you take a beer (prendre une bière), in English you just drink one; in English you take a nap, and in French, you make one (faire un petit somme). It is unlikely that this reflects different cultural attitudes toward drinking and napping. More generally, these authors come to the same conclusion from very different directions: Language—both as a mental capacity and as a social entity shaped by historical forces—really is special.