Is That a Fish in Your Ear?

by David Bellos

  We can grant that emotional relationships to things, including poems and forms of language, may be ultimately incommunicable. However, beliefs about the uniqueness and ineffability of emotional attachments have no relevance to the question of whether poetry is translatable. That is a much less abstruse matter.

  Some people doubt that there are any affects or experiences that cannot be expressed, on the commonsensical grounds that we could say nothing about them and would therefore have no way of knowing if they existed for other people. The philosopher Ludwig Wittgenstein presumably meant to adopt an agnostic position on this issue in the famous last line of his Tractatus when he wrote, “What one cannot talk about must be left in silence.” 3 The infinite flexibility of language and our experience of shared emotion in reading novels and poems and at the movies must also cast doubt on whether there are any human experiences that cannot in principle be shared. On the other side of this thorny tangle is the intuitive knowledge that what we feel is unique to us and can never be fully identified with anything felt by anyone else. That inexpressible residue of the individual is ineffable—and the ineffable is precisely what cannot be translated.

  Should translation studies pay any attention to the ineffable, or to notions, intuitions, feelings, and relations that are held to be unspeakable? Oddly enough, anguished engagement with the problem of ineffable essences is not at all characteristic of Bible translation, where you might expect to find mystical and religious issues taken seriously. Instead, it has preoccupied secular scholars of the twentieth century, from Walter Benjamin to George Steiner and Antoine Berman. I would rather approach this boundary of translation from the opposite direction, for it seems to me more important to realize not that the ineffable is a problem for translation, but that translation is one big problem for the ineffable.

  Let’s imagine a crew returning from a space flight at some future point in time. They’ve visited a faraway Earth-like planet and are holding a press conference at NASA headquarters. They have something spectacular to announce. Yes, KRX291 is inhabited, they say, and, what’s more, the little green men that live on it have a language.

  “How do you know that?” a journalist asks.

  “Well, we learned to communicate with them,” the captain responds.

  “And what did they say?”

  “We can’t tell you that,” the captain answers coolly. “Their language is entirely untranslatable.”

  It’s not hard to predict how our descendants would treat the captain and his crew. They would have the astronauts treated for flight-induced insanity, and, if that proved to be unjustified, treat them as liars, or as laughingstocks. Why so? Because if the inhabitants of the distant planet did have a language, and if the space crew had learned it, then it must be possible for them to say what the aliens had said. Must, not should: radically untranslatable sounds do not make a language simply because we could not know it was a language unless we could translate it, even if only roughly.

  There are intermediate and problematic positions, of course. Not all utterances can be translated even when we are quite sure they are in a language. Egyptian hieroglyphs were indecipherable until two brilliant linguists, Thomas Young and Jean-François Champollion, worked out how to do it with the help of the Rosetta stone. More generally still, we can’t translate from languages we don’t know. But to claim that something is in a language is to posit that, with the appropriate knowledge, it can be translated.4

  Translation presupposes not the loss of the ineffable in any given act of interlingual mediation such as the translation of poetry but the irrelevance of the ineffable to acts of communication. Any thought a person can have, the philosopher Jerrold Katz argued, can be expressed by some sentence in any natural language; and anything that can be expressed in one language can also be expressed in another. What cannot be expressed in any human language (opinions vary as to whether such things are delusional or foundational) lies outside the boundaries of translation and, for Katz, outside the field of language, too. This is his axiom of effability. One of the truths of translation—one of the truths that translation teaches—is that everything is effable.

  Especially poetry. America and Britain are awash with poetry magazines, and every year small publishers put out hundreds of slim volumes containing poems in translation. Our present army of amateur poetry translators is keeping poetry alive. Poetry is not what is lost but what is gained from their work.

  An individual poem may have a quality that, for any one of us, is so personal and unique that it might as well be ineffable, but the issue of unspeakable ideas arises much more obviously in a quite different domain. It is in our interactions not with works of genius but with other species that the ineffable looms before us like a brick wall.

  On a short trip to South America, Romain Gary picked up a twenty-three-foot-long python, whom he called Pete the Strangler and then donated to a private zoo in California. When he was consul general in Los Angeles, Gary used to go and see Pete in his enclosure.

  We would stare at each other in absolute astonishment, often for hours, deeply intrigued and wondering, awed and yet incapable of giving each other any kind of explanation about what had happened to us, and how and why it had happened, unable to help each other with some small flash of understanding drawn from our respective experiences. To find yourself in the skin of a python or in that of a man is such a mysterious and astonishing adventure that the bewilderment we shared had become a kind of fraternity, a brotherhood beyond and above our respective species.5

  Maybe Gary was right to feel that a python can no more imagine what it is like to be one of us than we can imagine what the mental world of a reptile is like—and it’s typically generous of him to allow a fearful and pea-brained monster like Pete the Strangler a reciprocating intuition of the ineffability of human life. On the other hand, many nonhuman species—and perhaps all living things—do communicate with one another, and some most definitely communicate with us. Dog owners, to take the most obvious example, easily distinguish among the meanings of different kinds of bark. But the dog language we can access is a fairly limited thing. It consists of a small set of individual signals. Signals are generally treated as the isolated vehicles of specific pieces of information—“There’s an intruder in the house,” “Hello and welcome,” or “Take me for a walk.” They can’t be combined with one another to produce more complex meanings—as far as we know, dog language has no grammar. In addition, the set of signals possessed by domesticated dogs—like the signals used by monkeys or bees—is inherited and fixed. There’s no new word formation going on in dogs, just as the signaling system of traffic lights is incapable of producing more than “slow down,” “stop,” “get ready,” and “go.” (The green-and-orange “get ready” combination is used in the U.K., as a courtesy to drivers of ancient sports cars with gearshift sticks.) Those are the main criteria by which human language is distinguished from all other kinds of communication by most modern theorists of language. Monkeys can say only what they have to say, and nothing else; whereas human signaling systems are forever changing and always capable of adapting themselves to new circumstances and needs. These are fairly persuasive reasons for keeping animal language outside the field of “language proper” and far away from the concerns of translation. But we could try to be as generous and as imaginative as Romain Gary. From such a perspective, human language may well seem to a dog to be just as limited and inflexible a signaling system as linguists imperiously declare dog language to be.

  From infancy to the onset of puberty, children of every culture have always known that animals have things to say to them. There’s no folklore in the world that doesn’t similarly break the alleged barrier between human and other.6 But in our Western, script-based cultures, growing up (which is so heavily entwined with formal education that it might as well be treated as the same thing) involves unlearning the instinctive childhood assumption of communicative capacity in nonhuman species. No wonder our philosophers an
d priests have long insisted that language is the exclusive attribute of humans. That self-confirming axiom makes children not yet fully human and in real need of the education they are given.

  However, the traditional reasons for making a radical separation between “signaling” and “speaking” are not quite as hard-edged as they are often made to seem. Some animal signaling systems that have been studied (among ants and bees, for instance, where the channels are not by voice but by physical and chemical means) communicate what for us would be extremely elaborate geographic and social information. Whales emit long streams of haunting sounds when they gather in a school in waters off the coast of Canada. The tonal and rhythmic patterns of whale song are of such complexity as to make it quite impossible to believe that what we can hear (and pick up on instruments more sensitive than human ears) is just random noise. Even more striking is the recent behavior of a group of monkeys in a Colchester, England, zoo: they have added two new gesture signals to their prior repertoire of communicative behavior. Even if the “monkey sense” of these gestures is not absolutely certain, they are indisputably meaningful signs within the community, and indisputable inventions of the monkeys themselves.7

  But what makes the communicative behavior of ants, bees, whales, monkeys, dogs, and parrots mysterious to us, what takes cross-species communication into the realm of the ineffable, is the fact that, save for a very limited range of noises from a limited range of long-domesticated pets, nobody knows how to translate “animal signals” into human speech or vice versa. When and if we ever can translate nonhuman noises into human speech, species-related ineffabilities will evaporate like the morning haze.

  Translation is the enemy of the ineffable. One causes the other to cease to exist.

  FOURTEEN

  How Many Words Do We Have for Coffee?

  The number of New Yorkers who can say “good morning” in any of the languages spoken by the Inuit peoples of the Arctic can probably be counted on the fingers of one hand. But in any small crowd of folk in the city or elsewhere you will surely find someone to tell you, “Eskimo has one hundred words for snow.” The Great Eskimo Vocabulary Hoax was demolished many years ago,1 but its place in popular wisdom about language and translation remains untouched. What are interesting for the study of translation are not so much the reasons this blooper is wrong but why people cling to it nonetheless.2

  People who proffer the factoid seem to think it shows that the lexical resources of a language reflect the environment in which its native speakers live. As an observation about language in general, it’s a fair point to make—languages tend to have the words their users need and not to have words for things never used or encountered. But the Eskimo story actually says more than that. It tells us that a language and a culture are so closely bound together as to be one and the same thing. “Eskimo language” and “the [snowbound] world of the Eskimos” are mutually dependent things. That’s a very different proposition, and it lies at the heart of arguments about the translatability of different tongues.

  The discovery and understanding of what makes different languages different and also the same has a curious modern history. In a lecture on the culture of the Hindus given in London to the Asiatic Society in 1786, an English judge posted to Bengal made a claim that overturned long-held beliefs in the superiority of the languages of the “civilized” West and the lesser jargons of the rest of the world:

  The Sanscrit language, whatever be its antiquity, is of a wonderful structure; more perfect than the Greek, more copious than the Latin, and more exquisitely refined than either, yet bearing to both of them a stronger affinity, both in the roots of verbs and the forms of grammar, than could possibly have been produced by accident; so strong indeed, that no philologer could examine them all three, without believing them to have sprung from some common source, which, perhaps, no longer exists; there is a similar reason, though not quite so forcible, for supposing that both the Gothic and the Celtic, though blended with a very different idiom, had the same origin with the Sanscrit; and the old Persian might be added to the same family.3

  This is generally reckoned to be the starter’s gun in a fascinating race that lasted for much of the nineteenth century to map all the world’s languages and to work out how they were related to one another, in “family trees” each springing from a single progenitor. But even on the Old Continent some languages—Albanian, for example—didn’t seem to have any close relatives at all, and one of them stuck out like a sore thumb. Basque, spoken in parts of northern Spain and southwestern France, was just so different as to resist any kind of “family” treatment. Wilhelm von Humboldt, elder brother of the great explorer Alexander, learned this strange idiom and wrote a grammar of it,4 and in so doing developed the intellectual tools that in watered-down form ultimately led to the Great Eskimo Vocabulary Hoax.

  Von Humboldt was struck not so much by the list of words that Basque has for different things as by the radically different structure of the language. It seemed to him that the grammar of Basque was the core and also the mirror of Basque culture. The observation was generalized into a theory: insofar as the formal properties of different languages are different from one another, each of the world’s languages gives access to a different mental world.5 Basque cannot be “reduced” to French, German, or anything else. It is just itself—the embodiment and the root cause of “Basqueness.” Different languages, von Humboldt saw, were different worlds, and the great diversity of natural languages on the planet should be seen as a treasure house of tools for thinking in other ways.

  The observation “other people just don’t think the way we do” was made long before von Humboldt’s essays appeared, but for most of human history it was dealt with quite easily. In Greek eyes, “barbarians” who couldn’t speak Greek were obviously not capable of saying anything interesting. Similarly, for the grammarians of seventeenth-century France, other languages could barely allow their speakers to engage in approximations to real thought, which was truly possible only in Latin and French. It must have taken great courage to express von Humboldt’s insight in the colonial era, when the otherness of other languages was generally thought to confirm the intellectual inferiority of people less fortunate than the French (or the Greeks, or the Romans, and so forth). Like Sir William Jones, the Bengal judge, von Humboldt dared to assert that other languages offered speakers of “West European” a wonderful mental resource.

  Colonial expansion and conquest brought Europeans into contact with languages that were even more different than Basque. Some of them, dotted here and there around the globe in no obvious pattern, are very different indeed. Imagine a language in which there is no term for “left” or “right” but only expressions for laterality cast in terms of cardinal orientation. “There’s a fly on your southwest leg” might mean “left” or “right,” depending on which way the speaker and his interlocutor are facing. (This is less unfamiliar than it first sounds: in contemporary Manhattanese we use cardinal orientation whenever we say “go uptown from here.” To the dismay of many a lost tourist, that can’t be translated into tournez à gauche or à droite unless you also know which of the four cardinal points you are facing.) Speakers of Kuuk Thaayorre (Cape York, Australia), for example, lay out ordered sets (say, numbers from one to ten, or photographs of faces aged from babyhood to maturity) not from “left” to “right” or the other way around but starting from east—wherever east happens to be with respect to the table at which their anthropological linguist interrogator is seated.6

  But languages can be even weirder than that. In Nootka, a language spoken on the Pacific coast of Canada, speakers characteristically mark some physical feature of the person addressed or spoken of either by means of suffixes or by inserting meaningless consonants in the body of a word. You can get a very faint idea of how this works from vulgar infixes such as “fan-bloody-tastic” in colloquial English. In Nootka, however, the physical classes indicated by these methods are children, unusually fat or he
avy people, unusually short adults, those suffering from some defect of the eye, hunchbacks, those that are lame, left-handed persons, and circumcised males.7

  One example of the radical difference of human languages was made famous by the American linguist Benjamin Lee Whorf, who had learned and studied many Native American languages. In the language of the Hopi (but also in quite a few others, distributed with no obvious pattern around the globe), there is a grammatical category called evidentials. For each noun-phrase, the grammar of Hopi marks not so much the categories of definiteness or indefiniteness (“a farmer,” “the farmer”) but whether the thing or person referred to is within the field of vision of the speaker. “The farmer I can see” has a different form from “the farmer I saw yesterday,” which is different again from the form of “the farmer you told me about.” As a result, the English sentence “The farmer killed the duck” is quite untranslatable into Hopi without a heap of information the English sentence doesn’t give you—notably, whether the farmer in question is present to the speaker as he speaks and whether the duck is still lying around. If you speak Hopi, of course, and are speaking it to other Hopi speakers in an environment where the duck and the farmer are either with you or not, you know the answers to these questions and can express your meaning grammatically. What you can’t translate in a meaningful way is the sentence “The farmer killed the duck” out of context. But as we have seen in earlier chapters, this kind of untranslatability holds for any de-contextualized sentence in any language. The use of Hopi-type grammars as evidence of the untranslatability of tongues is really a red herring. Isolated, unsituated, written example sentences are often more hindrance than help when it comes to thinking about translation.