McWhorter: Language Doesn't Affect How You Think

More Edge questions about what ideas popular in science should be kicked to the curb: linguist John McWhorter attacks a theory on the Nurture side of the Nature-Nurture divide.

John McWhorter

Professor of Linguistics and Western Civilization, Columbia University; Cultural Commentator; Author, Doing Our Own Thing

Languages Conditioning Worldviews

Since the 1930s when Benjamin Lee Whorf was mesmerizing audiences with the idea that the Hopi people's language channeled them into a cyclical sense of time, the media and university classrooms have been often abuzz with the idea that the way your language works gives you a particular worldview.

There are two closely related ideas here:

1. Thought is affected by language form, such as differences in grammar

2. Thought is affected by language content, most notably differences in vocabulary

Content differences include Franz Boas's famous contention that the Eskimos have a gazillion words for snow. Wikipedia says about Eskimos and snow:

The claim that Eskimo languages have an unusually large number of words for snow is a widespread idea first voiced by Franz Boas and often used as a cliché when writing about how language may keep us more or less alert to the differences of the natural world. In fact, the Eskimo–Aleut languages have about the same number of distinct word roots referring to snow as English does, but the structure of these languages tends to allow more variety as to how those roots can be modified in forming a single word.[1][2]

But English has a large number of words for snow, too, as you'll note during the upcoming Winter Olympics coverage. So it's silly to be surprised that the size of vocabulary of a tiny illiterate culture isn't larger than that of the huge literate culture that produced the Oxford English Dictionary. It would be a more apples to apples comparison to contrast the number of words for snow in an Eskimo language to the number of words for the white stuff on top of Kilimanjaro in the language of a small African tribe.

Most languages appear to be fairly elaborate in form, but languages differ wildly in quantity of content. The Oxford English Dictionaryfor instance, features 600,000 words.

Orwell depicted Newspeak in 1984 as an attempt to reduce vocabulary to conceptually impoverish the subjects of the tyranny: e.g., the Declaration of Independence translates into simply "crimespeak."

For example, in the Edge essays, numerous scientists on the social engineer side of the spectrum rail against the nature - nurture conceptual framework devised by Shakespeare and Galton as something that should be permanently retired. Why? They offer lots of reasons, but a basic reason is that "nature-nurture" makes it easier for citizens to think skeptically about social engineering plans like Obama-Blasio's notion of fighting income inequality with universal pre-K, so junking the phrase "nature and nurture" would help intellectually disarm taxpayers.

McWhorter goes on:

You just want this to be true, but it isn't—at least in a way that anyone would be interested in it outside of a psychology laboratory (or academic journal).

Or as McWhorter later points out, not in ways we're supposed to be interested in, which would seem like very different things, but is increasingly the same thing to modern people as crimestop -- the predilection for feeling bored by potentially subversive trains of thought -- becomes more beaten into contemporary skulls.

It's high time thinking people let go of the idea, ever heralded as a possibility but never actually demonstrated, that different languages represent different ways of experiencing life.

Different cultures represent different ways of experiencing life, to be sure. And part of a culture is having words and expressions to express it, to be sure. Cell phone. Inshallah. Feng shui. But this isn't what Whorfianism, as it is often called, is on to. The idea is that quiet things in a language's very structural architecture—how its grammar works, how its vocabulary happens to cut up space—channel how the speaker experiences life.

And in fact, psychologists have indeed shown that such things do influence thought—in tiny ways elicitable via fascinatingly peculiar experiments. So, Russian has different words for dark and light blue and no one word that just means blue, and it has been shown that Russians are, indeed, 124 milliseconds faster at identifying grades of dark blue to other ones and grades of light blue to other ones. Or, it has been shown that people whose languages divide nouns into masculine and feminine categories are more likely, if asked, to imagine those things talking in the appropriately sexed voice if they were cartoon characters, or to associate them with gendered traits.

This kind of thing is neat—but the question is whether the quiet background flutterings of awareness they document can be treated as a worldview. The temptation is endless to suppose that it does. Plus we are always reminded that no one has said that language prevents a speaker from thinking anything, but rather that it makes it more likely that the speaker will.

... In the early eighties, psychologist Alfred Bloom, following the Whorfian line, did an experiment suggesting that Chinese makes its speakers somewhat less adept at processing hypothetical scenarios than English speakers. 

After all, nobody ever noticed that the Chinese tend to be pretty concrete in their thinking compared to, say, the Hindus or the Ancient Greeks. Oh, except that tends to be everybody's impression. (Whether it stems from language or not is another question, and which aspects of language would be involved is a third ...)

Whoops—nobody wanted to hear that.

Kind of a general problem with the human sciences these days: there are lots of things nobody wants to hear.

There was long train of rebuttals, ending in an exhausted draw.

In other words, Bloom's argument apparently wasn't disproved despite strongly motivated attempts to do so. (This doesn't mean it was proven, just that it was still standing after the spirit of the age took its best shots at it.)

But there are all kinds of experiments one could do that would lead to the same kind of place. Lots of languages in New Guinea have only one word for eating, drinking, and smoking. Does that make them slightly less sensitive to the culinary than other people?

Are you implying that Papuan cuisine isn't quite as sophisticated as Italian or Cantonese? That a French sommelier may come equipped with a more sophisticated vocabulary for thinking about wine than a New Guinea highlander?

Nobody wants to hear that!

Or, Swedish doesn't have a word for wipe—you have to erase, take off, etc. But who's ready to tell the Swedes they don't wipe?

In cases like this our natural inclination is to say that such things are just accidents, and that whatever wisp of thought difference an experimenter could elicit on their basis hardly has anything to do with what the language's speakers are like—or what their worldview is. But then, we have to admit the same thing about the wisps that happen to tickle our fancies.

No, there is an obvious difference between the two examples.

What creates a worldview is culture—i.e., a worldview. And no, it won't work to say that culture and language create a worldview together holistically. 

How do we know? For example, consider ancient Greece's transformation from barely literate in 700 BC to philosophically sophisticated in 350 BC. The Greek vocabulary developed tremendously during this period as you might imagine. Richard Nisbett argued in The Geography of Thought that ancient Greek was peculiarly well-adapted to coining new conceptual words, a role that it continues to play today in the coining of new scientific and technological terms.

Is Nisbett right? This stuff's over my head. But I wouldn't rule it out. If McWhorter's upcoming book demolishes Nisbett's arguments that Japanese speakers seem to be better at seeing the context and English speakers seem to be better at "object oriented" thinking, with Japanese raised in the U.S. in-between, then good for him. But, so far, I'd consider Nisbett's argument viable if unproven.

Remember, that would mean that Chinese speakers are—holistically—a little dim when it comes to thinking beyond reality.

Who wants to go there?

Phrased conversely, the Chinese tend to be particularly sharp at thinking about current reality, that they seem to devote a higher proportion of their mental horsepower to the palpable here-and-now.

Especially when even starting to, decade after decade, leads us down blind alleys? Hopi, it turned out, has plenty of markers of good old-fashioned European-style time. ...

A lot of anthropological examples turn out not to be very good since it's so hard to check up on something about some small tribe. Plus, there's the simple brute fact that a lot of languages of small illiterate tribes tend to be conceptually impoverished because the tribesmen don't think abstractly very often, and their brightest intellects who do come up with abstractions can't write them down to communicate them over time to future very bright fellows who would be on their wavelengths. So, the brightest illiterate sages end up playing a game of Telephone with their abstractions, with generally depressing results.

But anthropologists frequently feel the need to gloss over this with highbrow explanations of the tribe's alternative abstractions. (I'm not saying this is the case for the Hopi-Whorf tale in particular. Benjamin Whorf, by the way, was an interesting guy: an MIT chemical engineer who was the top chemical factory fire prevention inspector for a big insurance company, who took occasional breaks to go to Mexico to study indigenous languages. It's hard not to imagine that his death from cancer at age 44 was a real loss to the human sciences because who knows what he would have come up with if he'd lived long enough to turn to linguistics full time.)

What it comes down to is this. Let's ask how English makes a worldview. Our answer requires that the worldview be one shared by Betty White, William McKinley, Amy Winehouse, Jerry Seinfeld, Kanye West, Elizabeth Cady Stanton, Gary Coleman, Virginia Woolf and and Bono.

Let's face it, what worldview would that be?

I don't know, but I suspect a lot of Frenchmen, much less Japanese, would have an opinion on the subject.

Evelyn Waugh's theory was that English had evolved not to be precise like French (with its limited vocabulary that seems to encourage the French to ring dazzling changes on a fairly distinct set of ways of the Pleasures of Being French -- French radical philosophers tend to be dazzling but oddly conservative: underneath all the novelty is a core conviction that the highest form of life is to live in France, and the highest form of being French is to sit in a sidewalk cafe and philosophize) or thorough like German (with its ability to turn every sentence into giant words, which may encourage German pedantry and profundity), but to sound good, to be a language for poets to weave their spells.

Sure, a lab test could likely tease out some infinitesimal squeak of a perceptive predilection shared by all of those people. But none of us would even begin to think of it as a way of perceiving the world or reflecting a culture. Or, if anyone would, then we are on to an entirely new academic paradigm indeed.

Perhaps we can broaden Waugh's notion to include playwrights, politicians, salesmen, celebrities, rock stars, actors, rappers, comedians, and the like. Which now that I think of it, pretty much covers "Betty White, William McKinley, Amy Winehouse, Jerry Seinfeld, Kanye West, Elizabeth Cady Stanton, Gary Coleman, Virginia Woolf and and Bono."

To take a more cynical view than Waugh, perhaps English turns out to be the finest language in practice for salesmen and other BS artists to use to infiltrate their ideas into the minds of others, that English is the ideal language for world-domination?