The Conquering Tongue If languages do not preserve world views, what makes them worth saving?

The siren song

You can count on it to leave a classroom quiet enough to hear a pin drop: the idea that one’s language shapes how one views the world. The idea was first popularized by amateur linguist Benjamin Lee Whorf in the 1930s, based on his study of the language of the Hopi. Whorf claimed that Hopi has no markers of tense, and that this was why the Hopi have a cyclical rather than linear sense of time.

Almost ninety years later the very idea of such a thing is catnip to roughly anyone. Yet Whorf’s depiction of Hopi turned out to be inaccurate, and his early death deprived us of knowing how he would have adjusted his framework as a result. But he sparked a school of thought that has lived on since, in which linguists, anthropologists, and psychologists—today, especially the latter—seek to show that language influences thought to some degree.

No worker today has the near-mystical take on the matter of Whorf, who took the idea as far as supposing that European languages predestined their speakers to create modern science, an endeavor he thought to be marred by neglect of the more holistic bent of languages less bound by notions of past, present and future. However, a commitment continues to showing that to some extent, hardly absolute but potent enough to merit discussion, your language shapes your thought patterns. By extension, we are to suppose that how a language’s vocabulary divides up things and actions (e.g., whether like Hopi it has different words for water in nature and water that you are drinking) and how its grammar works (e.g., whether it has tense, whether it distinguishes the from a) corresponds to the essence of a culture.

The purpose

That idea is not only inviting but, to some, useful: Whorfianism can serve as a resonant tool for consciousness-raising about a looming linguistic catastrophe, the disappearance of as much as 90 percent of the world’s 6000 languages over the course of this century. Already, 96 percent of the world’s people speak one or more of just twenty “big” languages alongside their local ones. Many of those languages are the linguistic equivalent of kudzu: English, Mandarin, and Indonesian, for example, tend to edge out and exterminate smaller languages. The preface of a grammatical description of the obscure language of a small group quite often includes an observation that the language is no longer spoken by children, which almost guarantees its extinction as a living language.

A scan of the world’s languages will reveal as many awkward implications about world view as laudatory ones.

To ordinary people, the disappearance of an obscure language can seem inconsequential. One may suppose that life progresses, that as modernity brings groups together it is inevitable that smaller languages will no longer be useful, and that in the grand scheme of things, the welfare of a people is more important than whether or not they happen to still be using their elders’ language.

But views like that horrify those dedicating their lives to documenting and helping keep alive tiny languages few will ever hear of. They argue that the disappearance of a language is the disappearance of a particular way of seeing the world, which the rest of us are poorer for having missed, in the same way that threatened plants or animals are often described as possible sources of chemicals or survival strategies.

The argument is powerful, handy, and heartfelt—which is what makes it all the more awkward that it doesn’t hold up. It leaves a question: If languages are not world views, then is there value in saving them regardless?

The inconvenient truth about “world views”

On the world view notion, the data are in. Does language influence thought? Yes, a wee bit, but in no sense significantly enough to be called a world view. Certainly, a culture is encoded partly in words for things and beliefs, something easily perceived, utterly undeniable, and, therefore, of little true interest. A culture based on strict social hierarchy will have elaborate proliferations of different terms for you. In the language of a culture living on a mountain, the words for up and down will be more central to the grammar than in the language of a people living on the tundra.

But Whorfianism claims something more than this: that to know the basic vocabulary and grammar of, say, Chinese is to have your thoughts urged strongly into a “Chinese” worldview distinct from a Portuguese or Indonesian one. But scholarly inquiry has revealed no such thing. Language differences do create fascinating peeps of cognitive difference, to be sure. Russian has different words for dark and light blue, and Russians turn out, in experimentation by psychologist Lera Boroditzky, to be a tad faster at distinguishing shades of blue straddling the line between dark and light than English speakers. Just a tad—124 milliseconds on the average. That is indeed fascinating in itself, and results of that kind have been shown in countless ways for various language pairs at this point. Believe it or not, people whose language assigns a gender to inanimate objects (Spanish’s el sombrero for hat versus la luna for moon) are more likely to assign those things traits traditionally assigned to males and females under the conditions of a psychological experiment.

The question, though, is whether the experiments demonstrate a “world view.” That is, does blue “pop” more at an art museum if you speak Russian? Do Spanish speakers actually think of hats as “guys” and moons as “ladies”? One wants the answer to be yes—until learning of other kinds of “world views” Whorfianism can suggest in perfectly innocent people.

For example, in Mandarin, one can leave aspects of the counterfactual and hypothetical to context much more than we are used to in a language like English. For example, to say If you had seen my sister, you would have known she was pregnant, a typical way to say it in Mandarin would be If you see my sister, you know she is pregnant, which is also the way you could say If you saw my sister, you knew she was pregnant. Now, according to the Whorfian view, we would expect that Chinese makes a person somewhat less sensitive than we are to the hypothetical—what English conveys with obligatory had’s and would’s and would have’s. And this is exactly what work by Alfred Bloom in the early eighties found: Mandarin speakers were somewhat less adept at answering questions requiring reference to the counterfactual.

However, because of the unsavory implications of a conclusion that suggests that Chinese speakers are less quick on the uptake than Westerners, Bloom’s work elicited a long series of attempts at refutation. The result was essentially a draw—no one was capable of entirely discrediting Bloom’s conclusions. Here’s the rub: faced with a proposal that to be Chinese is to be a touch dim, most of us will decide that whatever was shown in the artificial conditions of a psychological experiment can hardly be treated as indicating a world view. That is: Chinese grammar is somewhat uninterested in hypotheticality, but Chinese people are just like anyone else. Language may influence thought a tiny bit, but not enough to, basically, matter outside of a psychology lab.

Note, however: if this is the intelligent response to Whorfian results that are unflattering, then consistency requires that it be the same when the result suggests a higher sensitivity to blue, interpersonal relations, interest in what material things are made of, or any number of other “cool” traits that have been attributed to speakers of languages based on how their vocabularies and grammars work. To wit: the fun bits involving blue and gender make us ask “Who’s to say what a world view is?” And the answer is that these experiments are showing us a world view to exactly the extent to which we are ready to call the Bloom experiment’s results evidence of a world view. In other words, no world view at all.

Overall, today’s affection for Whorfianism is ironic given that assorted thinkers in earlier times were given to similar claims in ways that are no more attractive to us today than the implications of Bloom’s work can seem. Prussian thinker Heinrich von Treitschke opined in the nineteenth century that “differences of language inevitably imply differing outlooks on the world,” within a typology of estimation under which Third Worlders (and Lithuanians) were “barbarians.” Of course today’s Whorfian only intends to identify “differing outlooks” in an approving or perhaps neutral way. But that brings us back to cases like Chinese: a scan of the world’s languages will reveal as many awkward implications about world view as laudatory ones.

Linguistically assimilated groups can retain cultural distinctiveness. … However, it would be just as difficult to deny that loss of language is typically accompanied by a significant degree of assimilation to the dominant culture, and is, in fact, a symptom of it.

Then why save languages?

Why save dying languages, then? The question is especially urgent given two challenges rarely addressed squarely by people devoted to language preservation.

Challenge One: If languages are valuable in reflecting a world view, then wouldn’t we be doing our job in simply recording the language’s words and structure for posterity? With the world view duly preserved for us, wouldn’t the languages’ continuing to be actually spoken become irrelevant?

Challenge Two: Language preservation skeptics are sometimes heard to say that the world would be better off with just one, or maybe just a few, languages, because translation is expensive and imperfect, and language diversity is overall an impediment to communication and understanding. In other words, there’s a reason the Biblical tale of the Tower of Babel is written as a tragedy. There are those who read this observation as a callous thumbs-up for English eating up the world’s languages, a language laden with sins of empire centuries deep. However, the observation need not be taken as such: one might imagine the sole remaining language as Hungarian or Xhosa. The question can be posed as a schematic one, hardly unworthy of posing: would a monolingual world necessarily be a lesser one? Especially when the argument that each language represents a world view is not as impregnable as often implied?

However, even without the flawed science of treating these languages as “world views,” and in the face of these challenges, I believe that there is an argument that 1) at least a goodly number of threatened languages should be kept alive as living systems, and 2) that a monolingual world would indeed be poorer in contrast. That argument is threefold.

Totemic value

Language preservationists often say that when a language dies, the culture dies with it. From them, the statement is founded in the Whorfian perspective, but even if languages are not world views in the way Whorfianism seeks to show, there is value in the idea that the very fact of having a separate language is a keystone to constituting a culture.

Certainly, linguistically assimilated groups can retain cultural distinctiveness. Few would claim that Reform American Jewishness is no longer a “culture” because its members no longer use Yiddish in the home. One would be on even thinner ice to claim that Native Americans in the United States who no longer speak their ancestral language are therefore no longer meaningfully “Indian.”

However, it would be just as difficult to deny that loss of language is typically accompanied by a significant degree of assimilation to the dominant culture, and is, in fact, a symptom of it. It is hardly accidental, for example, that Jews who do continue to use Yiddish in the home are also those who observe the directives of the Torah much more closely than others. In general, the very ability to communicate in a language incomprehensible to outsiders occasions, and even embodies, an in-group cohesion less self-generating if no such code exists.

Importantly, this is true without a language corresponding to a cultural world view: the issue is not what the language is like, but its sheer existence as a language different from others. Under this analysis, a language is like a tartan. The very fact of a Scottish clan’s association with a particular pattern is a vibrant token of its distinctiveness, despite that no one would say that there is any inherent association between that tartan’s colors and patterns and what the clan is like culturally.

Thus if there is a commitment to cultural preservation—and note that no one could deny that cultures, as opposed to languages, differ in terms of world views—then language preservation can be seen as a component of that quite independently of the Whorfian perspective and its pitfalls.

Cognitive value

Recent studies have shown that bilingualism enhances cognitive function. Again, the point is the fact of bilingualism itself, not which languages are spoken and which “world views” they purportedly lend. The analogy is with exercise or playing a musical instrument: The sheer act is the key, not the particularities of swimming over the treadmill or the violin over the flute.

To control two languages has been shown in children to enhance the brain’s executive function, which directs problem-solving, concentration, and multitasking. There is even evidence that the higher the degree of bilingual proficiency, the later the onset of dementia is in elderly people. As such, another argument for keeping languages alive is that speaking more than one language is, in itself, good for mental functioning and by extension, the world as a whole.

Aesthetic value

The layman often supposes that languages differ simply in having different words for things, partly because of the prescriptive tradition under which grammar is taught mostly as a collection of sentence patterns one is to avoid. However, languages differ far more than we tend to be aware, such that there is a value in language preservation simply because of the magnificence of variation among the world’s 6,000 different ways of speaking. The analogy is with, again, flora and fauna. Languages bedazzle in their diversity to a degree that most, if aware of it, would understand the value in preserving—and even as living languages rather than as data on pages and in recordings. Languages differ much more than in which word they use for water and how one says goodbye.

Three randomly chosen examples can illustrate the point:

Flemish: There are dialects of Flemish in which how you say yes depends on person and number just as verb endings do. If I say yes I say jok; if we say yes we say jom; if someone asks if your friend did something and you say “Yes,” you say “Jot, he did it.”

Hmong: In this language of Southeast Asia, all syllables have different meanings according to seven different tones. So, to say po with a swoop down means “female.” With a swoop up it means “throw.” Say it on a high pitch and it means “ball-shaped.” On a low pitch it means “thorn.” On a pitch somewhere in the middle it means “pancreas.” Say it with a creaky tone and it means “see,” and say it with a breathy tone that sounds kind of like how Marilyn Monroe talked, and it means “paternal grandmother.” This is the case with every syllable, not just some—it’s how the whole language works. There are some who might admire the very speaking of such a language in the same way that they admire the flight of an albatross or the acrobatics of an Olympic gymnast.

Bemba: In this African language of Zambia, there are eight shades of past tense: things that started happening today, things that finished happening today, things that finished happening today and are truly done with, the “regular” past, the past further back (Elvis left the building), the perfective form of that past (as in Elvis has left the building where repercussions of the event persist), the more distant past, and things that happened continuously in that more distant past. Clearly, English with its mere past, perfect and pluperfect is less interesting in comparison.

Any language harbors traits of this kind, fascinatingly different from what Whorf termed “Standard Average European.” Crucially, traits like these emerge only over centuries and even millennia of daily usage by communities, within which new habits drift into the system and take hold, becoming tomorrow’s grammar. Once a language is dead, this kind of efflorescence is, of course, forever stanched.

Which languages?

One remaining question is whether we can expect to save all of the world’s dying languages, or even most of them. The answer to that question requires reference to an aspect of the matter insufficiently attended to: that acquiring fluency in a language is difficult as an adult, and the tragedy is that the smaller a language, the more complex it tends to be. When history has thrown a lot of non-native speakers at a language—which is usually because it has been a language of empire or jostled around between some of them—that language has tended to become somewhat streamlined grammatically. Hence a language such as English, where the Viking invasions of the eighth and ninth centuries left it with no gender for inanimate objects, barely any case marking suffixes, only a single ending, –s on verbs in the present tense, etc.

Speakers of English—and a few other Western European languages with a similar degree of imperfect learning in their pasts—often have no way to spontaneously imagine how fearsomely more complex most languages in the world are. That means that language preservation asks a massive learning task, much more than memorizing some words and expressions and a few lists of endings. To a person raised in English, Spanish is hard enough—but in a Native American language like Navajo, all verbs are irregular! It is one thing to acquire a halting command of a language like this—and quite another to imagine speaking it 24/7 to your child, when this is precisely what truly reanimating a language would entail.

For that reason, as I have opined elsewhere, in the case of languages whose fluent speakers are now all elderly we cannot expect them to be reborn as languages lived in by whole communities. Sadly, the learning task is perhaps greater than many have any reason to be aware, such that a reasonable expectation is that descendants of speakers will know words and expressions as a cultural hallmark, taught down the generations certainly with the help of scholars who have recorded what the language was like when it was alive. Certainly there will be uniquely dedicated individuals who manage stronger command of the ancestral language than this, but the eternal question will always be whether there are enough such people to constitute a community passing the language down to new generations.

However, there are languages that are still spoken by enough people today that the trend towards extinction could possibly be reversed, especially with the help of modern tools such as online pedagogical reinforcement and texting. Navajo is one of them: as recently as 1981 85 percent of Navajo children spoke it as a first language and it was often cited as a Native American language success story, but more recently that number is down to 25 percent. Maori in New Zealand is another example: by the 1980s, only one in five Maori spoke the language with true fluency, and a revival movement flowered into the 1990s but has waned considerably since. However, both languages have had rather ample media footprints where they are spoken, and both have been well-studied and therefore have robust representation on the page, to assist those who seek to study or teach them.

In cases like these—languages that are “on the ropes” but realistically revivable—their reanimation is

worth the difficulty of acquiring languages as an adult.

The real deal

However, because of the nature of that task, more linguists and educators should be working to address how to teach adults to actually speak languages quite unlike English and its close Romance and Germanic relatives, for the very purpose of assisting the survival of threatened languages. That effort, of course, must happen in conjunction with language preservationists as well. All concerned must attend closely to what is actually required in reviving a language as a spoken vehicle of actual communication. Experimentation will be healthy, and results ought to be shared and used rigorously as the basis for further efforts. Two issues could be seen as among those foremost to consider.

Speaking over writing. Teachers must be fully aware that speaking is quite different from writing and, in reference to keeping a language alive, much more important. Too often, for example, students of Chinese and Arabic are taught to focus so much on mastering the writing system of the languages that they learn little of how to actually express themselves, which makes boredom and frustration too likely after a year or two. The analogous mistake should not be made when imparting languages on the verge of disappearance.

Throughout the history of humankind, almost all languages have been passed on by illiterates; even after the invention of writing 5,500 years ago, only over the past few centuries has literacy been common in certain privileged societies. Even today, only about a hundred languages are written ones, while the vast majority are passed on orally. Someone able to pass a language on to their child despite not knowing how to write in it is more valuable than someone who can teach their child how to write some words but not how to express themselves in spontaneous sentences.

What is a word? Many linguists are arriving at a conception of the “word” that includes what are often considered “idioms” traditionally. To speak a language is to manipulate a continuum of “chunks” of which isolated words are just one kind, while “idioms” like might as well or expressions such as “You have the wrong X”—a concept expressed variously in languages, hardly only with the word for wrong—are others. Beyond a certain point—and an earlier one than often thought—learners should be given the full range of “chunks” that are needed to communicate in a basic fashion, beyond the “My uncle is a lawyer but my aunt has a spoon” school of language-textbook vocabulary. To wit,  learners who can convey anyway, upside-down, right in the middle, I am good at this, They are the same size, It does not fit, and He will get over it are well on their way to actually speaking a language even if they have no idea of what the words are for things traditionally presented to learners as basic such as forks, cousins, horses, and shoes.



Among language preservationists, the claim that languages do not express world views is not popular. An article such as this one, because of its first sections alone, will occasion a degree of irritation among those who have so often argued that a language is a cultural perspective. However, in the end, this article has been intended as a positive one. Even if languages do not afford their speaker a particular pair of glasses through which they see the world, there is value in preserving as many of them as we can.

I have sometimes surmised that many language preservationists stress the “world view” argument out of a kind of benevolent pragmatism, seeing it as the easiest sell to a lay audience amidst a situation that demands immediate address. I propose, however, that the public may be open to arguments of a different kind, that avoid the pitfalls of overselling academic results, of leaving groups open to unsavory descriptions as well as pleasant ones, and even of a kind of unintended exoticization.

On the latter, for example, who among us is ready to say that English embodies a “world view” that all of its speakers share, one more robust than the obvious cultural differences between English speakers in Massachusetts, Melbourne, Norwich, Kenya, Johannesburg, and New Delhi? Other languages can be said to express “world views” to exactly the extent that English can: to so minimal a degree that few would consider the issue worth extended engagement as a matter of sociological import.

Languages are valuable quite beyond this, and the news must be spread as far and wide as Whorfianism has been.