Through the Language Glass: Why the World Looks Different in Other Languages (27 page)

Since there is no evidence that any language forbids its speakers from thinking anything, as Humboldt himself recognized two hundred years ago, the effects of the mother tongue cannot be sought in what different languages
allow
their speakers to think. But where then? Humboldt went on to say, in somewhat mystical terms, that languages nevertheless differ in what they “encourage and stimulate to do from their own inner force.” He seems to have had the right sort of intuition, but he was clearly struggling to pin it down and never managed to get beyond the metaphors. Can we turn his hazy imagery into something more transparent?

I believe we can. But to do so, we need to abandon the so-called Sapir-Whorf hypothesis, the assumption that languages limit their speakers’ ability to express or understand concepts, and turn instead to a fundamental insight that can be dubbed the Boas-Jakobson principle.

We have already encountered the anthropologist Franz Boas as the person who introduced Edward Sapir to the study of Native American languages. In 1938, Boas made an acute observation about the role of grammar in language. He wrote that, in addition to determining the relationship between the words in a sentence, “grammar performs another important function. It determines those aspects of each experience that
must
be expressed.” And he went on to explain that such obligatory aspects vary greatly between languages. Boas’s observation was rather inconspicuously placed in a little section about “grammar” within a chapter entitled “Language” within an introduction to
General Anthropology
, and its significance does not seem to have been fully appreciated until two decades later, when the Russian-American linguist Roman Jakobson encapsulated Boas’s insight into a pithy maxim: “Languages differ essentially in what they
must
convey and not in what they
may
convey.” The crucial differences between languages, in other words, are not in what each language allows its speakers to express—for in theory any language could express anything—but in what information each language obliges it speakers to express.

Franz Boas, 1858–1942

Roman Jakobson, 1896–1982

Jakobson gives the following example. If I say in English, “I spent yesterday evening with a neighbor,” you may well wonder whether my companion was male or female, but I have the right to tell you politely that it’s none of your business. But if we are speaking French or German or Russian, I don’t have the privilege to equivocate, because I am obliged by the language to choose between
voisin
or
voisine, Nachbar
or
Nachbarin, sosed
or
sosedka
. So French, German, and Russian would compel
me to inform you about the sex of my companion whether or not I felt it was your business. This does not mean, of course, that English speakers are oblivious to the differences between evenings spent with male or female neighbors. Nor does it mean that English speakers cannot express the distinction should they want to. It only means that English speakers are not obliged to specify the sex each time the neighbor is mentioned, while speakers of some languages are.

On the other hand, English does oblige you to specify certain bits of information that can be left to the context in some other languages. If I want to tell you in English about a dinner with my neighbor, I may not have to tell you the neighbor’s sex, but I do have to tell you something about the timing of the event: I have to decide whether we
dined, have been dining, are dining, will be dining
, and so on. Chinese, on the other hand, does not oblige its speakers to specify the exact time of the action each time they use a verb, because the same verbal form can be used for past or present or future actions. Again, this does not mean that Chinese speakers are unable to express the time of the action if they think it is particularly relevant. But as opposed to English speakers, they are not obliged to do so every time.

Neither Boas nor Jakobson was highlighting such grammatical differences in relation to the influence of language on the mind. Boas was concerned primarily with the role that grammar plays in language, and Jakobson was dealing with the challenges that such differences pose for translation. Nevertheless, it seems to me that the Boas-Jakobson principle is the key to unlocking the actual effects of a particular language on thought. If different languages influence their speakers’ minds in varying ways, this is not because of what each language allows people to think but rather because of the kinds of information each language habitually obliges people to think
about.
When a language forces its speakers to pay attention to certain aspects of the world each time they open their mouths or prick up their ears, such habits of speech can eventually settle into habits of mind with consequences for memory, or perception, or associations, or even practical skills.

If this all still sounds a little too abstract, then the contrast between the Sapir-Whorf hypothesis and the Boas-Jakobson principle can be
brought into focus with another example. Chinese may seem to us rather lax in allowing its speakers to equivocate about the time of the action, but just try to imagine what a speaker of Matses from Peru might feel upon hearing about the incredibly crude and careless tense distinctions of English.

The Matses are a 2,500-strong tribe, and they live in the tropical rain forest along the Javari River, a tributary of the Amazon. Their language, which was recently described by the linguist David Fleck, compels them to make distinctions of mind-blowing subtlety whenever they report events. To start with, there are three degrees of pastness in Matses: you cannot just say that someone “passed by there”; you have to specify with different verbal endings whether this action took place in the recent past (roughly up to a month), distant past (roughly from a month to fifty years), or remote past (more than fifty years ago). In addition, the verb has a system of distinctions that linguists call “evidentiality,” and as it happens, the Matses system of evidentiality is the most elaborate that has ever been reported for any language. Whenever Matses speakers use a verb, they are obliged to specify—like the finickiest of lawyers—exactly how they came to know about the facts they are reporting. The Matses, in other words, have to be master epistemologists. There are separate verbal forms depending on whether you are reporting direct experience (you saw someone passing by with your own eyes), something inferred from evidence (you saw footprints on the sand), conjecture (people always pass by at that time of day), or hearsay (your neighbor told you he had seen someone passing by). If a statement is reported with the incorrect evidentiality form, it is considered a lie. So if, for instance, you ask a Matses man how many wives he has, unless he can actually see his wives at that very moment, he would answer in the past tense and would say something like
daëd iko
h
: “two there were [directly experienced recently].” In effect, what he would be saying is, “There were two last time I checked.” After all, given that the wives are not present, he cannot be absolutely certain that one of them hasn’t died or run off with another man since he last saw them, even if this was only five minutes ago. So he cannot report it as a certain fact in the present tense.

But finding the right verbal form for directly experienced events is
child’s play compared with the hairsplitting precision required when you report an event that has only been inferred. Here, Matses obliges you to specify not just how long ago you assume the event occurred but also how long ago you made the inference. Suppose you saw wild pigs’ footprints on the ground somewhere outside the village, and you want to tell your friends that the animals passed by at that place. In English, saying “wild pigs passed by there” is exactly as much information as you have to specify. But in Matses, you have to reveal both how long ago you found out about the event (that is, how long ago you saw the footprints) and how long before that you think the event itself (pigs passing by) actually occurred. For example, if a short time ago you discovered tracks that were still fresh, you assume that the wild pigs passed by only shortly before you saw the tracks, so you would have to say:

kuen–ak–o–
h

passed by–
HAPPENED SHORTLY BEFORE EXPERIENCED–EXPERIENCED RECENTLY
–they

“they passed by” (I found out a short time ago, it had happened shortly before that)

If a short time ago you discovered tracks that were already old, you would have to say:

kuen–nëdak–o–
h

passed by–
HAPPENED LONG BEFORE EXPERIENCED–EXPERIENCED RECENTLY
–they

“they passed by” (I found out a short time ago, it had happened long before that)

If a long time ago you discovered tracks that were still fresh, you would have to say:

kuen–ak–onda–
h