Few fields of study suffer from a more complete public misunderstanding than linguistics. It isn’t uncommon for a linguist to be asked, on meeting a non-linguist, how many languages he or she speaks, or to hear the exclamation, “Oh dear, I must watch my grammar!” Linguists study languages and their structures, but speaking many of them isn’t a job requirement, nor is being a professional grammar scold. A slightly rarer misimpression is usually held by those with just enough knowledge to be dangerous. These people think they flatter a linguist when they say how important linguistics is, “because what we think depends on the words we use to think it.”
This last belief is the bugbear that’s been eating John McWhorter’s trash, and that he hopes to kill off once and for all with his latest book, The Language Hoax: Why the World Looks the Same in Any Language. McWhorter’s writing appears frequently in the liberal New Republic and the conservative City Journal, often on the subject of race and politics. (McWhorter subscribes to a number of political heterodoxies.) But before he went into punditry, McWhorter trained as a linguist and contributed to the study of creolization, the process by which two or more languages coalesce into a full-featured third language.
The belief in question—that the languages we speak shape the thoughts we think—is known in linguistics as the Sapir-Whorf hypothesis, and among the linguistic establishment, Whorfianism has fallen on very bad times indeed. The hypothesis’ namesakes, Edward Sapir and Benjamin Whorf, have been dead for 70 years, and in my own linguistics classes I rarely heard them invoked except to be ridiculed, like biologists of yore who thought maggots grew spontaneously from rotting meat, or historians who thought the world began 6,000 years ago. What Whorfianism claims, in its strongest form, is that our thoughts are limited and shaped by the specific words and grammar we use. Mayans don’t just speak Mayan; they think Mayan, and therefore they think differently from English speakers. According to Sapir-Whorf, a person’s view of the world is refracted through her language, like a pair of spectacles (not necessarily well-prescribed) superglued to his face.
The Herero people use the same word for green and blue, but they have no difficulty distinguishing the color of a leaf and the color of the sky.
Whorf came up with his version of the hypothesis through his study of the language of the Hopi Indians. Hopi, he believed, lacks tense markers, like the “-ed” in “I walked to the store,” or words meaning “before” and “after.” In English we can’t say a sentence about walking to the store without saying when the walking happened. Whorf turned out to be wrong about Hopi time-words and tense-markers, McWhorter notes: Hopi has them. But Whorf viewed Hopi’s supposed lack of them as a sign that the Hopi see the world with less reference to time than we do, and that they are a culturally “timeless” people, living in communion with eternity while we English speakers are slaves to tense markers and clocks.
Perhaps the most famous invocation of Sapir-Whorf is the claim that because Eskimos have dozens of words for snow, they have a mental apparatus that equips them differently—and, one assumes, better—than, say, Arabs, to perceive snow. (I once watched the wintry film Fargo with an Egyptian who called everything from snowflakes to windshield-ice talg—the same word she used for the ice cube in her drink.) To get a hint of why nearly all modern linguists might reject this claim, consider the panoply of snow-words in English (sleet, slush, flurry, whiteout, drift, etc.), and the commonsense question of why we would ever think to attribute Eskimos’ sophisticated and nuanced understanding of snow to their language, rather than the other way around. (“Can you think of any other reason why Eskimos might pay attention to snow?” Harvard’s Steven Pinker once asked.)
McWhorter’s first order of business in his book is to show that the Sapir-Whorf hypothesis is testable, and that the cognitive differences between speakers of different languages turn out to be infinitesimal. A Sapir-Whorfian might expect a speaker of a language with a sophisticated inventory of words for colors—not just Newton’s seven basic hues, but hundreds of names for the gradations in between—to be better at differentiating colors. Unlike English speakers, for example, Russians have distinct words for dark and light blue. But if you show a Russian the two shades of blue, his speed at differentiating them is just 124 milliseconds faster—not even the blink of an eye—than an English speaker’s. Some languages are particularly forlorn in their inventory of color words; the Herero people of southwest Africa use the same word for green and blue. But they have no difficulty distinguishing “the color of a leaf and the color of the sky,” McWhorter notes. “Living on the land as they do would seem to have made it rather difficult to avoid noticing it at least now and then.”
What’s amazing about Sapir-Whorf, given the fairly negligible differences that linguists have thus far detected between language speakers, is that educated people from many other disciplines—philosophy, cultural studies, and literature—claim that their own thoughts are shaped by language in exactly the way Sapir-Whorf proposed, and that linguists now consider risible.
If you look for folk-Whorfianism you’ll see it all over. McWhorter rightly calls its appeal “almost narcotic.” George Orwell’s essay “Politics and the English Language” (1946), which generations of writers have accepted as a model of clarity and common sense, takes as its premise something awfully like Whorfianism: “the slovenliness of our language makes it easier for us to have foolish thoughts.” The novelist Anthony Burgess, who wrote two delightful books on linguistics, lapsed into Whorfianism when explaining why he wanted the protagonist of A Clockwork Orange to borrow the Russian ruka (“rooker,” in the book) as his slang for “hand.” “Russian makes no distinction,” Burgess wrote, “between ... hand and arm, which are alike ruka. This limitation would turn my horrible young narrator into a clockwork toy with inarticulate limbs.” A character in a novel by John Crowley rhapsodizes about how his Puerto Rican girlfriend’s Spanish enhances her sexiness by surrounding her with gendered nouns and making the world “a constant congress of male and female, boy and girl.” (What must life be like for the Nasioi speakers of New Guinea, whose language has over 200 genders? Like living in the Castro District of San Francisco, one imagines.)
So The Language Hoax is the work not of John McWhorter, pundit, but of John McWhorter, zombie killer: slayer of an undead theory. He is a talented guide to linguistics, one of a number of gifted lay and professional linguist popularizers, including Burgess (stronger on philology and phonetics than on language and mind), Pinker, Geoffrey Pullum, Robert Lane Greene, and William Safire. McWhorter’s writing is just a bit less graceful than these others, in part because of his chatty habit of including bafflingly irrelevant detail (we learn that one of his teachers looked like a sad Tom Petty, and that McWhorter thinks Nivea cream smells heavenly).
But on the substance, McWhorter is exhaustive, fair-minded, and convincing. He is withering toward most of Sapir-Whorf ’s lay acolytes, yet also generous to the academic linguists and psychologists who have led a minor neo-Whorfian revival. These researchers, such as Lera Boroditsky of the University of California-San Diego, have found minuscule ways—far smaller than anything proposed by the original Whorfians—in which language does affect thought. McWhorter praises Boroditsky and others for elegant and sound experimental design, but says they have resurrected the hypothesis in such a diluted form that their work only serves to show that language barely affects thought at all, as measured cognitively.
Even academic neo-Whorfians haven’t found evidence for a further claim of Whorfians, that languages express “cultural needs”—that the intricacies of grammar reflect cultural facts about the groups who speak it. Whorf’s original hypothesis about the Hopi—that they are “timeless” because their language doesn’t mark tense—is but one example. McWhorter offers another, this time of a language that marks more, not less, than we do in English. Tuyuca, an Amazonian language, marks sentences with endings that show the provenance of the information the speaker is giving. So just as in English we have no choice but to include a marker on our verbs to say when something happened (“he walked to the store”), Tuyuca speakers are forced to add a marker that says how they know the information: I hear (-gí), I see (-í), they say (-yigï). Having these “evidential markers” would, in a Whorfian world, make Tuyuca speakers more attuned to the accuracy and provenance of information—perhaps making them more skeptical, or better on the witness stand.
Again, McWhorter shows that evidential markers just don’t correlate with these broader cultural traits. Korean has them, and researchers have found that Korean children are no better at attributing things than English-speaking children who speak with no evidential markers. Of European languages, Bulgarian is perhaps the only major one that does, and no one seriously thinks Bulgarians are the supreme skeptics of Europe. And plenty of cultures (the ancient Greeks, the French philosophes) seem to have been plenty skeptical without the assistance of Tuyuca-like suffixes. Indeed, this Whorfian attempt at a compliment to the Tuyucans for their skepticism comes across as a slander against the speakers of languages that have no evidential markers. Are those languages less skeptical? Do languages that mark very few things make their speakers ignorant? Chinese syntax is particularly bare in this regard, and its speakers regularly say sentences of the form “he go store,” without explicit marking of time, evidence, or much else. Yet no one considers Chinese speakers especially lacking in discernment. McWhorter encourages us to see linguistic structures as nothing more than products of “chance,” and to accept that we are all “mentally alike.”
McWhorter’s book should convince any doubters that strong-form Whorfianism is a bad idea whose time will never stop coming. So why does this bad idea, apparently without evidence, keep getting re-discovered and confidently touted? It is, after all, not the sort of thing one would expect people to want to believe. It implies that we are prisoners of our dictionaries, and that our words distort our worlds in ways we cannot escape. The belief in a strong version of Sapir-Whorf means that other people are in a sense unknowable, and that some have virtues and skills (skepticism, color perception) that are unattainable to us. Call it a belief in the inevitable inequality of language speakers.
One reason for its persistence, McWhorter proposes, is that we are humble. Social and cognitive science has repeatedly shown that our minds are particularly bad at knowing their limits. A Whorfian linguistic constraint would be just one more such distortion, like the visual blind-spot that is on the retina of every human who has ever lived, but that most of us never learn about unless we take a neuroscience class. It also appeals to our sense that human diversity is greater than we once thought. McWhorter’s anti-Whorfian sentence “all humans are mentally alike” is, for me anyway, not an immediately appealing one. Whorfianism encourages the belief that every language is a beautiful and unique snowflake—or some other snow-entity, thinkable only by Eskimos. But it’s telling, McWhorter notes, that Whorfians tend only to celebrate languages for their differences (making Tuyuca speakers out to be skeptics, say), and never vilify them (making Chinese out to be credulous).
I would propose another attraction: Sapir-Whorf lets us off the hook. The theory suggests that some thoughts or ways of seeing the world are simply not possible for us—and that can be comforting, particularly if it means we’re limited in the same way that thousands or millions of other speakers of our language are limited. Less comforting is the post-Whorfian reality: that our potential thoughts are not limited by constraints of language, but by our own deeper inabilities to imagine, perceive, and feel. For those who thought they could blame words, that is an unsettling thought in any language.