Studies Have Shown - The Language Hoax: Why the World Looks the Same in Any Language - John McWhorter

The Language Hoax: Why the World Looks the Same in Any Language - John McWhorter (2014)

Chapter 1. Studies Have Shown

MY GOAL IN THIS manifesto is straightforward. I wish to show the flaws in, and even dangers of, the more sensational implications bandied about in our intellectual culture over whether and how language shapes thought. However, in this first chapter I need to ward off a possible misinterpretation, to the extent that this is possible.

I may be taken as dismissing the work of Neo-Whorfians, but I mean no such thing. I seek out the articles in question and read them with great joy. As far as I can assess, they are composed with great care, enviable imagination, and thorough training. In my teaching, I regularly note that new Whorfian work has shown some modest effects that one might want to know about.

What I take issue with is the tendency to interpret this work as suggesting something about the human condition that I think it does not. To be sure, this interpretation is one more talked about after the fact—by some of the authors and certainly by onlookers—than is actually engaged within the experiments themselves. Yet this interpretation, ripe for cocktail party chat, media-friendly, and beckoningly interdisciplinary, has much greater impact than the minutiae of the experiments. It requires, then, engagement and critique—even with full respect for the work itself.

Ultimately, almost all books settle into the public consciousness in shorthand. I assume that this book, if cited, will often be classed as simply dismissing Neo-Whorfianism. Regardless, I would be remiss without making my actual position clear for those engaging the text.

Hitting a Wall after a Long Night

One of my favorite Neo-Whorfian experiments is one the public doesn’t hear much about, perhaps because it doesn’t involve concepts quite as immediately enticing as colors and genders. Yet this experiment is flawlessly constructed, easy to understand, and exemplifies perfectly what good Neo-Whorfianism is—and isn’t—about.

It hinges on a difference between languages that one would be unlikely to consider important in the daily scheme of things. In English we say a long time. In Spanish, one says mucho tiempo, a lot of time. If you put it as “a long time,” un tiempo largo, no one will throw you off a bus, but it’s ungainly, not true español. In English, time is a distance. In Spanish, it’s an amount or a size.

Greek is the same way: you don’t have a long night in Athens, you have a big one, a “lot of” night. We might be tempted to read the Greek expression metaphorically—we have “big” nights in English, too, but Greeks don’t mean that the way we do. For example, in Greek you also have a “big” relationship rather than a long one, and what they mean by that is that the relationship lasted a long time. As in Spanish, time is stuff, something there can be a lot of, rather than a stretching out of something. “Long” night in Greek is weird Greek.

But then, in Indonesian it’s as in English: long times, long nights. These things vary from language to language: French is like English and Indonesian, while Italian is like Spanish and Greek.

One might suppose that a difference like this would be a mere matter of “feel” for the language in question and of no import beyond that. It’s what I would once have assumed. However, show an English speaker—who says a “long” time—a line slowly lengthening toward an end point on a screen, and then a square slowly filling up from bottom to top, and she’s better at guessing how long it will take the line to hit the end than for the square to be full. Yet a Spanish speaker is better with the square filling up than the line reaching its end! Plus, with their “a lot of night,” Greeks pattern with Spanish speakers and Indonesians with their “long” nights pattern like English speakers.

Among the reasons one might come up with for this difference, clearly the most plausible one is language: the metaphor for time in people’s language determined their performance on the test. Try fashioning an idea that Spanish, Greek, and Italian pattern together because of something about Mediterranean culture, and notice how hard it is to come up with how the beauties of the water and the splendiferousness of the seafood would make people better at predicting how long it will take before something is full. Then, good luck figuring out what cultural trait they have in common that would lessen people’s knack at the same task among people in Paris, Leeds, and Jakarta!

This guessing experiment was constructed by Daniel Casasanto, a psychology professor at the University of Chicago. He persuasively argues that a case like this, in which people are not asked about language during the experiment and thus were not primed to use their language’s expressions to help them make decisions, shows that language can shape thought. However, he makes no claims beyond this. After all, imagine what the claim would be. Speaking Greek creates a distinct mental world in which, well, you’re a little better at predicting how quickly a space will fill up with liquid, while speaking Indonesian makes you a little better at the always handy skill of predicting just when something’s going to hit a wall? How do those skills extend to life as it is lived—that is, to What It Means to Be Human? The Spanish speaker with his mucho tiempo walks about on a Saturday afternoon seeing his environment differently from me with my long time in that he … what?

Yet while writing The Stuff of Thought, Steven Pinker had to stop telling people he was writing a book about language and thought because regularly people assumed it must be about language as a lens—that is, about the structure of your language making you see the world “diversely” from other people. The cachet of this notion is not founded on findings of the kind Casasanto so elegantly identified, but on a tacit notion that such things are just preludes to something grander. We are to assume that, to adapt Al Jolson’s old catchphrase, we ain’t seen nothin’ yet, and that the payoff will be a confirmation that languages lend us worlds of different colors.

Kind of Blue

Yet the top-class Neo-Whorfian work on color, marvelous as it is in many ways, does not lend itself any more gracefully to the juicier, humanist angle of interpretation. For example, I forget why I know that the Russian word for “gay” is goluboj, but as it happens the word’s basic meaning is “light blue.” Not just blue, because there is another Russian word for the darker, navy, Prussian version of blue, siniy. There is no word that means just blue: in Russian, the sky and a blueberry are different colors.

A neat Neo-Whorfian experiment presented Russian speakers with various tableaus of three squares on a computer screen: one on top, the other two right below it. The squares were various shades of what English speakers call blue, occurring in twenty gradations stretching from dark to light blue. In each tableau, one of the bottom squares was the same shade as the top square, while the other bottom one was a different shade. The Russians were given a task: to hit a button when they identified which bottom square was the same shade as the top one.

It must have been pretty dull to take this little test, but the researchers were trying to get at something: whether having different terms for dark blue and light blue has any effect on perception—that is, can language shape thought? And they found that it did. For example, if the top square was dark blue and the stray, different-colored bottom square was a shade or three into the light-blue range, then the Russians hit the button in a flash, while if the stray square was just a different shade of dark blue the average time before hitting the button was longer. Things were the same the other way around: if the matching squares were light blue, then Russians hit the button without hesitation if the stray one was in the dark realm, but lingered otherwise.

Yet English speakers had the same response time wherever the stray square happened to fall in the blueness spectrum: a stray square’s lightness didn’t quicken them up when the matching squares were dark, and a stray square’s darkness didn’t quicken them up when the matching squares were light. This shows, in a really ingenious way, that having different terms for light blue and dark blue makes people differentiate those colors more quickly than people whose language has a single term for blue—and even when no one asks them about the words in question or even uses them.

Just in case anyone tried to find, say, some cultural reason why Russians would be more sensitive to the difference between dark blue and light blue than Americans, the researchers did another version of the experiment, to show that language really is what drives the Russians’ difference. The second experiment had the subjects not only distinguish the stray square, but at the same time recite a random string of numbers they had just been asked to memorize. The mental candlepower required of doing that puts a temporary block on the processing of language, and in this version of the experiment, suddenly whether the stray square was of the other kind of blue made no difference in the response times. So, without language, Russians were no more attuned to the difference between dark blue and light blue than a guy from Atlanta.

But. A current fashion advertises this kind of test as showing that what your language is like makes you see the world in a particular way. The Anglophone, intrigued, will strain gamely to imagine what the world must look like through the eyes of someone to whom light blue and dark blue are “more different” than they are to them. The attempt may be reminiscent of trying to picture a fourth dimension.

But there’s a problem. It’s not that this experiment by Jonathan Winawer, Nathan Witthoft, Michael Frank, Lisa Wu, Alex Wade, and Lera Boroditsky isn’t extremely clever, nor is it that it doesn’t show that language affects thought. Rather, we hit a snag when we try to go beyond the experiment and embrace the notion that it is telling us something about worldviews, being human, and the like. Namely, when I described the difference in reaction times, I used vague terms such as in a flash and linger. However, in actuality, to seriously evaluate what this experiment means beyond the world of academic psychology, it must be clear what the mean difference in reaction time was, depending on which color direction the stray square leaned toward. It was—wait for it—124 milliseconds.

124 milliseconds! When the matching squares were darker, if the stray square was also in the dark realm, then Russians hit the button just one tenth of a second more quickly than if the stray square was in the light realm. They didn’t linger for half a minute, or even a whole second, or even a half second. Really, we can’t even call a tenth of a second a linger at all.

Now, that there was an effect at all is still something—in itself. Think: among English speakers, just because of a difference in the language, there was no lag at all. But: upon what grounds are we to take a 124-millisecond difference in reaction time as signaling something about the way Russians experience life? Language affects thought? Apparently so, but as with so much in life, the issue is degree. At the current state of our knowledge, it would seem that goluboj is relevant to a Russian’s soul more vividly in terms of sexual preference than color!

Intuition corresponds with the 124-millisecond figure in suggesting that we are not dealing with anything like different glasses. Upon learning that Russian has separate terms for dark and light blue, it would seem that some are inclined to wonder whether it means that Russians see a robin’s egg and a preppy blazer as more distinct in color than English speakers do. However, to just as many English speakers, or, I highly suspect, more, the reaction is a certain bemusement that a language would make such a distinction. “Why would a language need to do that?” we might ask. “We certainly know that the color behind the stars on the American flag is starkly different from baby blue—but we don’t need different words for it!” That’s certainly how I felt when I first encountered Russian.

In that light, there are plenty of languages that do not make color distinctions an English speaker considers fundamental, in which case, to them, English looks as needlessly obsessive as Russian does to us. The Herero people of Namibia in Africa speak a language in which one term refers to both green and blue. Finding out that other languages have separate words for green and blue, the Herero were not given to wondering whether Westerners saw a different world than they. Rather, they were quite aware of the difference between the color of a leaf and the color of the sky—living on the land as they do would seem to have made it rather difficult to avoid noticing it at least now and then. They just found the idea of a language having separate words for those colors, when they learned such languages existed, faintly silly.

Some might still be open to an idea that, on some level, there is a scale of sensitivity to color upon which Russians are high up, English speakers are middling, and the Herero are down on the bottom. That ranking will feel distasteful to most of us—and we will see how often Whorfianism’s implications end up confronting us with similarly icky propositions when it’s not us that the studies depict as fascinatingly dim. It seems hardly irrelevant that the Herero, in terms of clothing and decoration, give all indication of reveling in color—including distinct greens and blues—just as much as Westerners. Despite all this, it may well be that an experiment could show that the Herero language wires the brain in some way that leaves its speakers a few milliseconds slower to distinguish a blue-green Crayola crayon from a green-blue one than the typical person on the street in Chicago or Stuttgart (German has grün and blau). But in this, we have departed from any meaningful discussion of differences in souls.

Yet souls are what we think of in response to statements like “As strange as it may sound, our experience of a Chagall painting actually depends to some extent on whether our language has a word for blue.” That was one of the most resonant phrases in the editorial based on Deutscher’s book and elicits almost 5,000 hits on Google at the moment I am writing this. As I have long experienced, the media (including publishers) tend to encourage academics to put things in that kind of way, in an endless quest for “eyes” (web hits). There are so many books out there; one must ballyhoo a bit. Editorials—and jacket copy—advertising the book will always have a certain rhapsodic quality that almost no actual text could embody.

However, phrases like the one about Chagall have more influence than the book itself, especially given the inherent frisson of the Whorfian idea, and it implies something the studies simply do not. Would lacking a word for blue really impact one’s experience of a Chagall more than education, experience, or even mere variation between individuals’ receptivity to art? The editorial did say only “to some extent,” but let’s face it, a hedge like that gets lost amid the sexy pull of the basic statement. The real question is to what “extent”? 124 milliseconds?

Tribe without Paper or Pencils Mysteriously Weak at Portraiture

There have been some claims about language affecting thought and culture, which, if valid, would indicate much more dramatic effects than infinitesimal differences in mental processing. However, what they demonstrate is cultural traits that language reflects, like Thai words for you, not linguistic traits magically shaping the culture.

You wouldn’t have known it in the summer of 2004. That summer is defined in my memory by three things. One is the melody my cell phone played when a text came through, as that was the summer I started texting. The second is a beautiful house plant, of a kind fashionable at the time in New York City, that proliferated its light green leaves all over my study’s windowsill and down to the floor. The third was endless media reports of the people who can’t do math because their language has no numbers.

This sounded off to me, like a song played with an off chord, or ice cream that has been in the freezer next to leftover linguini and clams, such that into the initial glow of strawberry or chocolate drifts a stray hint of garlic. The coverage was sparked by Columbia University psychologist Peter Gordon’s work on the language of a tiny Amazonian tribe called the Pirahã, and the result was that today an obscure language of the Brazilian rain forest has been discussed in various books written for the general public, and was especially publicized by Daniel Everett. It is always good to see a language so unlike Western ones getting so much attention. Nevertheless, it was still perplexing to see one publication after another exclaiming how counterintuitive it was that a group of people who don’t have numbers, don’t count things, and aren’t good at it if you try to make them do it. “Tribe without names for numbers cannot count” (Nature, August 19, 2004). “Experts agree that the startling result provides the strongest support yet for the controversial hypothesis that the language available to humans defines our thoughts” (New Scientist, same day).

It’s not that the Pirahã of the Amazon have been misportrayed. They really do not count and are all but hopeless at learning math. A Pirahã woman genuinely cannot tell you how many children she has, because the language has no words for numbers. There has been some controversy over just how utterly innumerate the Pirahã are. The evidence leaves me, for one, skeptical that they really have no concept of one and two, although it would appear that for them, “one” means what we would mean by “that there,” and “two” is more a matter of “a pair and optionally one more or so.” However, if someone lived with the Pirahã for several years as Everett did, then even if there is an extent to which people see what they want to see, we can take his word for it that the Pirahã don’t talk about the numbers 5 or 42. If the Pirahã do by chance have counting games that they hid from Everett (“No, no, not in front of him!”) then if all they have to work with is “that there” and “two and a bit” then we can assume that the game barely qualifies as what we think of as counting (“Here’s one banana, Junior, and now, heeeeere’s something like twobananas! Yaaay!!!!!”).

The problem is the announcement, “Tribe without numbers in their language cannot do math,” with breathless speculations about how the language shapes their existence. We have to imagine equivalent claims. “Tribe without letters cannot write”: notice how unlikely such a headline seems. Not having letters would seem to be the very essence of not writing. When we encounter a group without writing, we speculate as to the historical or cultural reasons that explain why they have not adopted it. What would we think of someone who was instead mesmerized with the fact that the group have no conception of letters, seeing it as a valuable insight that this ignorance of letters is what prevents the people from writing anything down or being much good at trying to do so if asked? “Illiteracy prevents writing,” the headlines announce—and we wonder whether we have had a small stroke.

Certainly not having numbers in your language will make learning math difficult. However, the fact that the language lacks numbers is not an independent variable in the way that having different words for dark blue and light blue or saying big night instead of long night are. Pirahã lacks numbers for a reason: an isolated hunter-gatherer culture has no need for a word for 116, or to do long division, or to speculate about the nature of zero.

If, nevertheless, Pirahã were the only language in the world to lack numbers, then there would be a case for treating it as a fluky matter with fluky consequences. That is, we might suppose that there are tribes who have no number words but still count to 7 or 54 silently with their fingers or by lining up little buds on the ground. However, as we would expect, small hunter-gatherer groups quite often have no numbers beyond two or so. That doesn’t get around much; only because the groups are small ones unknown beyond where they reside. Many of them, in fact, live in the Amazon. Hence it’s not that there is a mysterious lack of numbers in the language of one group that makes them bad at math. Rather, the lesson is that counting, as humanity goes, is an accessory, despite how fundamental it seems to us. Indigenous hunter-gatherers don’t need to count, and thus often their languages have no word for the number 307.

An interesting thing to know, but building a case for language “shaping thought” is out the window. “Tribe without cars doesn’t drive” sounds like something out of Monty Python, as does, really, the idea of marveling that people without numbers don’t take to math. For example, cultures differ in the degree to which they happen to elaborate their music, art, or food. All people have and cherish these things to an extent, but, for example, some groups take cuisine to a more prolific and universally captivating level than others. Take Italy versus Romania, perhaps. Yes, I know Romanian food has its moments. New York diner menus even feature something I’ve never quite got around to eating called “Roumanian Steak.” But still.

Suppose we encountered a tribe whose approach to food was relatively utilitarian, and found that in their language there was a single word that covered meat, vegetables, starches, and fruit. The person who came away saying that the reason these people weren’t gourmands was that they didn’t have words for different kinds of food would likely be a clever child, whom we would correct while chuckling warmly. Obviously, the cultural trait created the linguistic one.

Upon which we return to the likes of “Tribe’s not playing music is traced to their lack of musical instruments.” It is the warm attraction so many have to the idea of language shaping thought that leads people to treat this kind of reasoning as normal when it comes to language. Steven Pinker gets it just right: “The idea that Eskimos pay more attention to varieties of snow because they have more words for it is so topsy-turvy (can you think of any other reason why Eskimos might pay attention to snow?) that it’s hard to believe it would be taken seriously were it not for the feeling of cleverness it affords at having transcended common sense.”

It Depends on Where You Stand

It’s hard to avoid the same verdict on a case that was often advertised as the one for skeptics to beat when the Neo-Whorfian work started getting attention beyond academic psychologists in the late 1990s. As always, the literature starts with something you wouldn’t want to go through life not knowing, but then veers off into garlic ice cream.

There are groups in Australia who don’t think of things being in front of, behind, to the left of, or to the right of them. Rather, they think of north, south, west, and east. Always. Not just when they turn north, and not just when a reason comes up to explicitly figure out which way is up. To a group like the Guugu Yimithirr (the name, in their language, roughly meaning“talking like this”), if a tree is in front of them and to the north, then they say it’s north of them, and even when they turn around, they do not say it’s behind them—they say it’s north, which it still is. In front of them is now south, and they would describe a wall they might now be facing as “south.” This is how they describe where things are inside, outside, in the dark, in a room they’ve never been in: they can always instantly discern wherever they end up as north, south, west, or east.

It makes perfect sense; it’s just not what we would do. Here is a fascinating example of human diversity indeed. However, the scholars who have publicized this aspect of the Guugu Yimithirr call it stunning evidence for Whorfianism. Namely, they think of this not as something interesting about the Guugu Yimithirr as people but as something interesting about their language. To them it’s not that the Guugu Yimithirr process direction differently than others do—it’s that their language forces them to.

That is, “Tribe with no words for clothing do not wear clothes.” Imagine: according to Scientific American, “Previously elusive evidence that language shapes thought has been discovered in Papua New Guinea, where the Stnapon tribe, who habitually wear no clothes, have been found to exhibit this trait because their language has no words for clothing.” Unlikely—we assume that not wearing clothes came first, and that unremarkably the language developed no words for clothing.

In the same way, a Guugu Yimithirr man processes direction the way he does because his environment forces him to. The language part is just a result. Of course this is hardly a language that would encourage someone to think about behind and beside. But just as Eskimos have a reason to focus on snow, the Guugu Yimithirr have a reason to rely heavily on geographical coordinates: they live on flat land in the bush. In fact, this kind of reckoning is common in Australian Aboriginal languages.

I am hardly the first person to see it that way, but defenders insist that the language must be the driving force because there are similar cultures that do not rely on geographical coordinates. They posit that this means that it can’t be culture that creates this orientation, therefore leaving language as where it all starts. This reasoning, however, would not stand up in court.

No one has ever claimed that a given cultural trait always expresses itself in a group’s language. If it did, then every language spoken by a group with a strong sense of social hierarchy would have seven ways of saying you—even feudal European languages. Yet not one European language is ever recorded as doing so.

All evidence shows that people like the Guugu Yimithirr process the world as they do because of their environment, not their language. It is not even, as some might wonder, a chicken and egg case in which both sides are right. Exhibit A: There is no language like Guugu Yimithirr spoken in, for example, a rain forest or a town. People only rely on geographical coordinates to this extent in environments that would naturally make it urgent. No peoples surrounded by structures and roads in front of and behind them mysteriously insist on looking beyond them and saying “north” and “south.”

Exhibit B: It is documented that among generations of Guugu Yimithirr who grow up outside of the indigenous environment, the geographical orientation quickly falls apart—this seems to have happened with countless Aboriginal groups. Again, what drives this way of speaking is where its speakers are, not the language.

But can’t language play a part? Possibly, but the evidence suggests that it doesn’t in any significant way. For example, languages index aspects of environment in other ways. In the Mayan language Tzeltal in Mexico, one refers to “uphill,” “downhill,” “across,” and to place names rather than “in front of,” “behind,” and so forth. The Whorfian impulse starts with “What a fascinating language that channels its speakers into thinking that way!” However, more intuitively, we are also interested to know that the Tzeltal live on the side of a mountain!

Now, while some might try to save the Whorfian analysis by finding a group of people who live on the side of a mountain somewhere and yet speak and think in terms of left and right—“the language determines the thought pattern!”—there is another group that pretty much closes the case in favor of the prosecution. Upon which: Exhibit C: Next door to the Tzeltal live the Tzotzil, in the same kind of mountainside environment. As you might guess from the similarity of the names (one must guiltily admit they sound like two groups created by Dr. Seuss!), Tzeltal and Tzotzil are essentially variations on the same language: one, two, three is hun, cheb, oxeb in Tzeltal and jun, chib, oxib in Tzotzil. Yet the Tzotzil differ from the Tzeltal in that they do speak in left-right/front-back terms linguistically—yet if you submit them to a psychological experiment, they still reveal themselves to conceptually process direction in terms of geographical coordinates like the Tzeltal and the Guugu Yimithirr.

If a Tzotzil is presented with three objects laid out in a row on a table and is then asked to turn around to a table in back of them and arrange the objects “the same way,” they will place them in a way that we would consider backward, as if the order of the objects on the first table were mirrored. For them, when they move, the world doesn’t change—just like with the Tzeltal in the same experiment. What the Tzeltal and the Tzotzil have in common here is culture, not what their language—practically the same one—makes them do.

The cool insight is about the world, not what one’s language makes you see in it. Processing direction geographically is something about culture, which can occur whether it penetrates language or not. Calling it language shaping thought looks plausible from the Tzeltal, but falls apart when we pull the camera back and bring in the Tzotzil. Calling it language shaping thought looks plausible from the Guugu Yimithirr point of view, but falls apart when we pull the camera back and bring in a hypothetical issue of the Onion with the headline “Legless Tribe Incapable of Walking Because They Have No Word for Walk.”

Mommy, the Park Is Covered with Squirrel! Can I Go Feed Some of It?

And so it goes. I am unaware of a Neo-Whorfian study in which neither of these things are true: (1) it’s hard to say what it has to do with what it is to be human, or (2) the whole claim is like saying a tribe’s lack of a word for calf is why they don’t raise cattle. The studies themselves are always intriguing, but if they are showing anything like different lenses on life, then the difference between the lenses is like the one between the two lenses that your optometrist shows you during an exam for glasses or contacts when you have to have her alternate between them several times to decide whether you see better through one or the other, because really, the chart looks the same through both. “Better? Or better? Better? Or better?” she says. “Well, uh …,” one ventures. E, T X P R E, G J N B C … “Better? Or better?”—but actually you would experience life the same way in a pair of glasses fitted with either of them.

My praise of these studies in themselves is not a backhanded compliment. For example, there is work on Japanese that gets less attention than it should because it came along before the media happened to pick up on Neo-Whorfianism. It perfectly illustrates how Neo-Whorfianism can be great work despite offering little or nothing to those of a mystical bent.

In Japanese, when you talk about a number of something, the number has to come with a little suffix. That suffix is different according to what kind of thing or material something is. Two is ni, dog is inu. However, two dogs is not ni inu, but ni-hiki no inu. Hiki is used when you are talking about small animals and using a number. But if you say “two beers,” ni biru is incomplete, and ni-hiki no biru would make the beer into a small animal. One neither pats, feeds, nor swats at a beer. You say ni-hon no biru, because hon is used for long, thin things, like bottles.

In Japanese this translates into saying “two little critternesses of dog,” “two skinninesses of bottle.” Dog and bottle are treated as substances, just as in English we say two ounces of water but three pounds of meat, except that in Japanese you have to do this with all nouns when accompanied by a number. In English only some nouns are substances: three pounds of meat, but we say I have two desks in my office, not I have two woodnesses of desk in my office, and There are a lot of acorns over there, not Behold, there are many seednesses of acorn! But whenever there’s a number, woodnesses and seednesses are the lay of the land in Japanese.

There are dozens of these suffixes in Japanese. They are about the hardest things in the language, after getting used to the different word order, for English speakers to master because knowing which suffix to use for which noun gets a little arbitrary. Are bottles really long and thin in the way that pencils are? And when you find out hon also has to be used with phone calls and movies you just have to suck it up.

In any case, the Whorfian seeks to see if this grammatical trait, where everything is marked as stuff instead of as an object, has any reflection beyond. In fact, it does. In what is definitely the best-smelling Whorfian experiment yet, Mutsumi Imai and Dedre Gentner laid out for their subjects triads of objects: say, a C-shaped mass of Nivea (have you ever smelled Nivea? Truly heaven, I’ve always thought), a C-shaped mass of Dippity-Do (a hair gel more popular in the old days, which smells pretty good too, although currently they push an unscented kind, anyone’s preference of which reminds me of people who poo-poo mackerel and sardines as “tasting like fish” as if that’s a minus), and scattered little dapples of Nivea. Or a porcelain lemon juicer, a wooden lemon juicer, and then some pieces of porcelain (that part was just plain nice to look at).

Yes, all of this did apply to Whorfianism. Asked which two things go together out of the three, Japanese children were more likely to group the mass of Nivea with the little clumps of it, while American kids were more likely to group the similarly shaped masses of Nivea and Dippity-Do. The Japanese kids thought of the porcelain lemon juicer as forming a pair with the pieces of porcelain, while American kids grouped the two juicers and left those crummy shards of porcelain to the side. Americans group by shape, Japanese by material.

This is all the more fun because if you are American, you almost surely feel the American choices as more natural, even if you can see the basic sense in the Japanese kids’ choices. Nivea with Nivea, well, of course! But to an American, somehow the fact that the two masses of Nivea and Dippity-Do are shaped alike “pops” more.

And wouldn’t you know, when they hear about experiments like these, people who speak languages whose numbers work the Japanese way tend to find grouping by material more intuitive. That is even scientifically confirmed in experiments with other languages that treat all things as substances. Near Tzeltal and Tzotzil in Mexico is their relative language Yucatec, and its speakers have number suffixes like Japanese. Eight out of 10 of them given a paper tape cassette box (it was in the 1980s) group it with a small piece of cardboard, while 12 out of 13 English speakers grouped it with a plastic box. Yucatec speakers went by material, English speakers by shape.

One cannot assess Whorfianism without awareness of studies like these. Yet we must return to the big picture. Clearly, the Japanese and Yucatec experiments show that language can shape thought. The question is what is meant by thought. Many seek to read experiments like these as shedding light on larger issues: real life, the human condition. But what could that really mean from data of this sort? A difference in thought must be of a certain magnitude before it qualifies realistically as a distinct “worldview.”

Is there anything a Japanese person has ever done in the 1,800 years since chopsticks have been used in that country, anything that any of the 125,000,000 Japanese do with chopsticks now, or anything that any Japanese-to-be will ever do with or even think about chopsticks, that seems even remotely traceable to them thinking of chopsticks as a substance rather than as a thing? That is, what effect of any kind has this mental trait ever had on a Japanese person’s behavior, outlook, health, argumentational skill, artistic sensitivity, sexuality, or anything at all? “Goodness, this room is fairly bedecked with chopstick!”

At what shall we aim our subsequent experiments to find out how these Whorfian ripplets affect people and life as we know them? One can’t help noting how few such experiments seem to actually occur. And if somewhere, somehow, Japanese people suggest that they think of chopsticks as a substance like water or sex in some stupendously minimal, ambiguous way—of the kind even scholars would likely have trouble even agreeing on anyway—why, really, should it occupy our attention long-term?

Any Whorfian study that suggests any effect on “worldview” less evanescent than this still meets trouble. For example, in Mandarin Chinese next month is “the month below” and last month was “the month above.” Does that mean Chinese people think of time as stretching vertically rather than horizontally? Now, there would be a worldview—and for a while a paper by Stanford University psychologist Lera Boroditsky (last encountered heading that study of blueness in Russian) taught us that Chinese people do sense time as up and down, and the study comes up often in conversations about Whorfianism’s plausibility.

In Boroditsky’s experiment, Mandarin speakers were faster to answer a question like “August comes earlier than October” when they had just been shown pictures of objects oriented vertically (a ball over another one, for instance) rather than horizontally (worms following each other, for instance). However, I often noticed that the Chinese people I asked about this often said they didn’t sense time as going up and down, and as it happens, various researchers have not been able to replicate Boroditsky’s findings. Most indicatively, in one study English speakers were about even in terms of how well they did with sentences like “August comes earlier than October” depending on whether they had just seen vertically arranged balls or horizontally creeping worms.

To give up on a hypothesis upon the first volleys of criticism would be unscientific, and thus as expected, Boroditsky has refined her experimentation. In the latest rendition, subjects are asked to hit early/late buttons arranged vertically as well as horizontally in response to pictures (such as of a young and an old Woody Allen). Mandarin speakers are quicker when the buttons are vertical, paralleling the time expressions in their language.

But, we get back to the question as to what quick-er is and what it means. Mandarin speakers are 170 milliseconds faster at nailing “up” as previous. That is certainly a result. But then, English speakers are almost 300 milliseconds quicker at nailing what their language marks as previous, “left” over “right.” And, even Mandarin speakers get “left” as earlier about 230 milliseconds faster—which we would expect since left-right time orientation exists alongside the up-down one in Mandarin. So: English speakers register their language’s way faster than Mandarin speakers register their language’s way—and who knows why?—but even Mandarin speakers register the English way faster than their main one!!

So we do not know that Mandarin speakers “experience time vertically,” and it is even precarious to leave it that they even experience time “more vertically,” in that they demonstrably do at all, because let’s face it, mental habit encourages letting that qualified assessment drift into the easier “Mandarin speakers feel time as up and down,” period. What the studies show is that Mandarin speakers sense time primarily as horizontal, with a background openness to a sense of it as vertical that you can tease out from very, very careful experimentation.

Is that layered, subordinated twinge a “worldview”? In deciding whether it is, we must ask: where are the Mandarin speakers who say, “Oh, that isn’t going to be ready for years and years!” pointing animatedly to the ground?

Language Is about All of Us

Yet no one would deny that human cultures are quite diverse, nor would anyone deny that the diversity means that humans of different groups experience life differently. However, language structure is not what creates this difference in experience. Culture certainly percolates into language here and there. Why would it not, since people with cultures speak language? However, language reflects culture—in terms of terminology, naturally, and also things like honorific levels of pronouns and geographical ways of situating oneself. But pronouns and topographical terms are, themselves, terminology in their way. They come for free from what life is like for a language’s speakers.

What language does not do is shape thought by itself, in terms of meaningless gender divisions of the kind that in German makes forks female, spoons male, and knives something in between (die Gabel, der Löffel, das Messer), or in terms of how people see the world’s colors, or in terms of whether we think of a cat as a clump of cuteness in the same way as we see a glorious-smelling white glob as a clump of Nivea. All attempts to find otherwise splutter. Even if you can, as it were, trick someone into revealing some queer little bias in a very clever and studiously artificial experiment, that weensy bias has nothing to do with anything any psychologist, anthropologist, or political scientist could show us about how the people in question manage existence.

Make no mistake: languages, like cultures, differ massively, and far beyond the terminological features that drift into them from the cultures. I have written about how vastly languages differ, and we will see it in the next chapters. The degree of divergence is awesome indeed: languages with only a handful of verbs (many Australian languages), languages with no regular verbs (Navajo), languages where a word’s meaning differs according to nine different tones you utter it on (Cantonese), languages with only ten sounds (Pirahã again), languages with whole sentences that you need only one word to utter (Eskimo), languages with dozens of click sounds (did you ever read about the “Kung” “bushmen” in an anthropology class? “Kung” is actually lazy Westerner for!Xũũ where the! is a click), languages with no tense at all (Maybrat in New Guinea), languages with two hundred genders (Nasioi, again in New Guinea), languages where the only ending in the present tense is the third-person singular one (English).

But the wonder is how in all of their diversity, these languages convey the same basic humanity. The cultural aspects qualify as scattered decoration. That will sound naïve to many—until they consider what it is to learn a language, upon which it becomes clear how ancillary the cultural aspect of a language is. How much of the Spanish or Russian or Chinese you hacked your way through was “cultural”?

If you want to learn about how humans differ, study cultures. However, if you want insight as to what makes all humans worldwide the same, beyond genetics, there are few better places to start than how language works. We will see why in subsequent chapters.

In this light, we must revisit some deeply seductive questions in Guy Deutscher’s editorial based on his book: Did the opposite genders of “bridge” in German and Spanish, for example, have an effect on the design of bridges in Spain and Germany? Do the emotional maps imposed by a gender system have higher-level behavioral consequences for our everyday life? Do they shape tastes, fashions, habits, and preferences in the societies concerned? At the current state of our knowledge about the brain, this is not something that can be easily measured in a psychology lab. But it would be surprising if they didn’t.

But what if they don’t?