The Last Speakers (5 page)

So how do you plumb the depths of a speaker's mind to retrieve the grammar of Tuvan, Igbo, Inuit, or Sora? And once you've “discovered” the grammar, what use is it to anyone?

In the field, we often start with body parts: ear, eye, nose, hand (oops, right hand). Things you can easily point to. But when you are living in a village, in a local environment, it very quickly becomes apparent why you can never figure out the whole grammar sitting in a classroom. Grammars are diffuse: they grow in gardens, flow along rivers, and float on air. One of the most fascinating sentences I ever collected in the nearly extinct Chulym language was “Worms have eaten our cabbage.” Though it was an entirely novel sentence, I understood it immediately. “Worms” was a totally new word to me, “eaten” was familiar, and “cabbage” was recognizable as a loanword from Russian. Never in a hundred hours of classroom work would I have asked for or heard such a sentence. It emerged spontaneously during a walk through the vegetable patch with last speaker Anna Baydasheva, as she thrust a worm-eaten cabbage under my nose for inspection.

Living with the Mongush family in Tuva, I collected such wonderful (and unasked-for) sentences as “The yaks pooped a lot yesterday—go and collect it,” or “The crooked-horned yak is licking salt.” In rural villages where I conduct much of my fieldwork, I always enjoy doing a photo-shoot walk. I may take a hundred images of local objects—cat, broom, canoe, locust, pebble—and play these on my laptop as a slide show. Each image elicits names for local, culturally relevant objects, as well as the stories behind them. A round pebble I picked up in the forest was not merely a pebble but an omen of good luck from the local spirits. A tiny purple flower was the sign for the sixth lunar month. Two-day-old dried yak poop had a different name than fresh poop.

I began to think of language as existing not only in the head, or perhaps not entirely in the heads of speakers, but in local landscapes, objects, and lifeways. Languages animate objects by giving them names, making them noticeable when we might not otherwise be aware of them. Tuvan has a word
iy
(pronounced like the letter
e
), which indicates the short side of a hill. I had never noticed that hills had a short side. But once I learned the word, I began to study the contours of hills, trying to identify the iy. It turns out that hills are asymmetrical, never perfectly conical, and indeed one of their sides tends to be steeper and shorter than the others. If you are riding a horse, carrying firewood, or herding goats on foot, this is a highly salient concept. You never want to mount a hill from the iy side, as it takes more energy to ascend, and an iy descent is more treacherous, as well. Once you know about the iy, you see it in every hill and identify it automatically, directing your horse, sheep, or footsteps accordingly. This is a perfect example of how language adapts to local environment, by packaging knowledge into ecologically relevant bits. Once you know that there
is
an iy, you don't really have to be told to notice it or to avoid it. You just do. The language has taught you useful information in a covert fashion, without explicit instruction.

Fieldwork is a constant string of “Oh my gosh, they actually have a word for that” moments. Standing on a high ridge in the Altai Mountains, my host Eres pointed down to the valley below, in which I could see numerous perfectly round circles etched into the brown landscape. UFO landings? Crop circles? No, these brown, round depressions, called
honash,
are the footprints of yurts that had been moved when families migrated. They may remain for several seasons or even years. Tuvans feel a great sentiment toward them, even extolling them in song. If they return to the same campsite, they never construct the yurt directly on an old honash, but impress a new one beside it.

So, language's proliferation doesn't stop with just having a word for something. Once in the lexicon, the “mental dictionary,” a named concept takes on a life of its own. It contributes to organizing thought and perception. We have no idea how deep this effect goes. If it goes deeper than we suspect, these unique words would render perfect communication among different languages impossible. Each language would remain a singularity of conceptual possibilities. At present, we have no idea how deeply (or shallowly) language may influence thought and perception. But we never will know if we allow most of the world's small tongues to pass into oblivion before they can be studied in their natural habitats.

HOW TO SAY “GO” IN TUVAN

One of the crucial parts of speech to learn in a language, after basic nouns, is a set of common verbs. I thought nothing could be simpler. I had learned to say “water” and “fetch water,” since I spent part of every day fetching water. But I also wanted to learn to say simply “go.” Yet every time I moved, pantomimed, or pointed to indicate “go,” I seemed to elicit a different word. It turned out that learning to say “go” in Tuvan is much more complex than I'd imagined. It requires not only an internal compass but also an acute awareness of the local landscape, even parts of it that may not be visible.

How does one acquire landscape awareness? Nomads are connoisseurs of geographic gossip, which they pass on in casual conversation, songs, stories, and their choice of “go” verbs. They talk frequently about where they have been, where the yaks are roaming, and where the neighbors are. I was amazed to find that my hosts always seemed to know the exact locations of migrating friends and relatives many miles distant.

With tools no more advanced than horses, binoculars, and gossip, the nomads in the community where I stayed managed to keep track of dozens of families, herd movements, and migration schedules. People would answer me with absolute confidence anytime I inquired as to the location or migration date of almost any member of the community. By daily observations through binoculars, talking about landscapes, and distributing bits of knowledge across domains (religious, aesthetic, acoustic), they kept track of a complex and dynamic system of multiple moving parts.

A young Tuvan in his familiar landscape, western Mongolia.

Tuvans live in a land where level spaces are unusual. Nearly every patch of ground slopes in one direction or another. This provides a framework for orientation—the directions of watersheds and river currents. Though Tuvan does have a general word for go, it is less often used. Most of the time, Tuvans use, as appropriate, verbs meaning “go upstream”
(còkta),
“go downstream”
(bàt),
or “go cross-stream”
(kes)
. You'd rarely hear, “I'm going to Mugur-Aksy” (the nearest town to the Mongush family camp), but rather, “I'm upstreaming [or downstreaming] to Mugur-Aksy.” Being a visitor rather than a lifelong resident, I was clueless as to what rivers were nearby and in which directions they flowed, so I could never confidently select the correct “go” verb. The Mongushes, on the other hand, could not explain to me the invisible orientation framework that was all around them and underfoot. No one ever said to me, “To say ‘go,' you must locate the nearest river, ascertain its direction of flow, then locate your path relative to the current.” They simply knew all this information without knowing that they did.

River-based systems are strictly local, leading to confusion. In one village, I got completely conflicting directions from two local ladies. One said (pointing due west), “Go upstream a bit more.” Farther along, another lady pointed west again, but told me to go downstream. I realized later that each was referring to a different river as her point of orientation. The Yenisei River, located several miles north of the village over a mountain ridge, was well known to all and served as a general axis of orientation. A much smaller stream, the Khüüls, flowed eastward just south of the village, and could also be used, depending on which way the speaker was facing. My ignorance lay in not knowing the local river systems and (even if I had known them) failing to pick up subtle cues about which river system was being referenced.

Locally embedded systems like Tuvan river-flow orientation may be found in many cultures, where the local landscape takes priority over the more abstract cardinal directions. Such languages force their speakers to specify at all times whether they are moving up, down, or across relative to a stream, mountain, wind, or some other frame of orientation. In order to speak correctly, one must attend to the lay of the land.

For the Mongush family, with their seasonal migrations among the mountain passes in unending pursuit of greener pastures, nothing was more pertinent to their survival than the local landscape. They were finely attuned to notice the tiniest details. In their world, but also in ours, landscapes permeate daily life. Landscapes are sculpted by humans, not just in the obvious physical ways—for example, by building roads or leveling hills—but also in cultural ways. What people pay attention to or name in the landscape may be deeply influenced by the language they speak. This holds especially true for small, indigenous cultures, well adapted by cultural habits to surviving in a particular place. Languages, too, adapt and equip their speakers with tools to describe, divide, and manage the local environment and its resources. Nor is this dynamic limited to small or indigenous cultures. If one Manhattanite says, “I'm cabbing it uptown,” another Manhattanite will understand perfectly, but outsiders may need a second to process the use of “cab” as a verb and to figure out what exactly “uptown” refers to.

Living with Tuvans, I learned that languages thus come to reflect local geography, not only in their vocabulary but also in more deeply structural ways, in their grammar. This knowledge is often accumulated over many centuries, and so geographic terms can represent an ancient layer of cultural knowledge encoded in language. Grammar can be “embedded” in the local landscape, and in fact cannot be understood or described in isolation from it. This finding, and similar ones by other scientists, has contributed to an emerging area of study known as “ethnosyntax.” This formulation goes against the traditional wisdom in linguistics, within the Chomskyan paradigm, by claiming that knowledge of grammar is not contained solely in mental structures—that is, rules in the mind—but also spills out to encompass the local landscape and cultural habits.

If this expansive view of grammar is true, then you cannot fully explore a language by simply sitting down with a speaker in a room and asking him questions, as is the current practice in linguistics courses. When I was co-authoring a Tuvan dictionary, I grappled with the problem of what information to include under the entry for “go.” To give an accurate and complete definition, I would have had to include not only the descriptions given above, but the very landscapes themselves. Similarly, as the following section describes, languages may lack what we think of as basic terms for colors, and instead have a complex system that describes the colors and patterns of animals. All these ways of encoding local knowledge in language are useful cultural adaptations, and they reveal a uniquely Tuvan worldview.

WHAT COLOR IS YOUR YAK?

One thing I learned about yaks is that they are incredibly skittish animals. They dislike strangers, noises, and camera flashes. The Mongush family's yaks did not allow anyone other than their master, Eres, to approach them. Yet once a day the entire herd, the cows and calves following their leader, an imposing shaggy bull, ambled downhill to the stockade. Here the lactating mothers were tied up and milked, while their mewling calves waited eagerly to suck the leftovers. Leaning on the split-rail fence of the stockade, I spent many hours observing and conversing about yaks. My take-home lesson? Every yak has its own color, and color (like other characteristics such as pattern, horn shape, or personality) provides a system of powerful descriptive detail.

The nomadic Tuvan yak herders have a complex hierarchal system for classifying yaks by the following traits in ascending order of importance: (1) fur color, (2) body pattern, (3) head marking, and (4) individual personality. They use different classifications for horses, goats, sheep, and cows. Mastering the system of yak-naming allows a herder to efficiently pick out or refer to a specific yak from a herd of hundreds. The color-and-pattern naming system is a strict hierarchy, determined by cultural preferences (which yak, horse, or cow colors and patterns Tuvans regard as more desirable, beautiful, or rare). If an animal possesses one or more special traits, you may omit mention of the less special ones, but if an animal has only a common trait, such as fur color (which all animals possess), you must mention it. If a horse or yak possesses one of several recognized body patterns—for example, star-spotted—then it will simply be called by that pattern, and its color need not be mentioned. A horse or yak possessing the highest trait, a spot on the forehead, will be named by that characteristic alone. If two animals each possess a foreheard spot, they will be differentiated by naming the spot plus the next trait down the hierarchy, for example “forehead spot brown stripe.” Learning extra labels for animals imposes a slight burden on memory, but as an information-packaging technology, it affords Tuvans great efficiency in breeding and herding.