Is That a Fish in Your Ear?

by David Bellos

  However, the rapid exploration of the diversity of human languages in the nineteenth century also led people to wonder in what ways the languages of less developed peoples were different from “civilized” tongues. Greek had “produced” a Plato, but Hopi had not. Was this because so-called primitive languages were not suited to higher thought? Or was it the lack of civilization itself that had kept primitive languages in their irrational and alien states? Von Humboldt’s hypothesis of an indissoluble bond between language and mentality could be used to argue either way around. Were there any general features of the languages of “natives” that marked them off as a class from those few languages that were spoken by the civilized nations of the world? And if so, what were they?

  Explorer-linguists observed quite correctly that the languages of peoples living in what were for them exotic locales had lots of words for exotic things, and supplied subtle distinctions among many different kinds of animals, plants, tools, and ritual objects. The evidence piled up at a disproportionate rate simply because the explorers wanted to know first of all what all these strange objects in their new environment were called. Accounts of so-called primitive languages generally consisted of word lists elicited from interpreters or from sessions of pointing and asking for names.8 But the languages of these remote cultures seemed deficient in words for “time,” “past,” “future,” “language,” “law,” “state,” “government,” “navy,” or “God.” Trique, a language spoken in Mexico, has no word for “miracle,” for example, only specific words for “heal the sick,” “part the waters,” and so forth.9 Consequently, it was difficult to translate into such languages most of the things that colonial administrators and missionaries needed to say. How could these strange folk be granted the benefits of civilization if the languages they spoke did not allow for the expression of civilized things? More particularly, the difficulty of expressing “abstract thought” of the Western kind in many Native American and African languages suggested that the capacity for abstraction was the key to the progress of the human mind.

  Savages will have twenty independent words each expressing the act of cutting some particular thing, without having any name for cutting in general; they will have as many to describe birds, fish and trees of different kinds, but no general equivalents for the terms “bird,” “fish” or “tree.”10

  The “concrete languages” of the non-Western world were not just the reflection of the lower degree of civilization of the peoples who spoke them but the root cause of their backward state. By the dawn of the twentieth century, “too many concrete nouns” and “not enough abstractions” became the conventional qualities of “primitive” tongues.

  That’s what people actually mean when they repeat the story about Eskimo words for snow. The multiplicity of concrete terms “in Eskimo” displays its speakers’ lack of the key feature of the civilized mind—the capacity to see things not as unique items but as tokens of a more general class. We can see that all kinds of snow—soft snow, wet snow, dry snow, poudreuse, melting snow, molten snow, slush, sleet, dirty gray snow, brown muddy snow, banks of snow heaped up by wind, snowbanks made by human hand, avalanches, and ski runs, to name but fourteen—are all instances of the same phenomenon, which we call “snow”; “Eskimos” see the varieties, not the class. (This isn’t true of real Inuit people, only of the Eskimos who figure in the Great Eskimo Vocabulary Hoax.)

  Translation between “civilized” and “primitive” languages distinguished in this way was clearly impossible. The solution was to teach colonial subjects a form of language that would enable them to acquire civilization, and the obvious tool to carry out the mission civilisatrice was the language of the imperial administrators themselves. In some cases, as in the Spanish conquest of the Americas, the impoverished resources of native languages were seen as such a threat to the spread of civilization that the languages and their written records had to be eradicated. But the destruction of the Maya codices wasn’t solely an expression of naked power, religious fervor, and racism.11 The suppression of lesser tongues was not a policy reserved by the Spanish for other continents—it was already the European norm. France had already begun its long campaign to stop people speaking anything that was not French within its own borders. Breton, Basque, Provençal, Alsatian, Picard, Gascon, and many other rural patois were almost hounded out of existence by laws and institutions over a period of several hundred years. The long pan-European drive toward “standard languages” was powered not only by political will, economic integration, urbanization, and other forces at play in the real world. It also expressed a deeply held belief that only some languages were suited to civilized thought.

  What, then, can it mean to “think in Hopi”? If it means anything, can it be called “thought”? The linguist Edward Sapir came up with a revolutionary answer in the early part of the last century. Breaking with millennial practice and prejudice, he declared that all languages were equal. There is no hierarchy of tongues. Every variety of human language constitutes a system that is complete and entire, fully adequate to performing all the tasks that its users wish to make of it.

  Sapir didn’t argue this case out of political correctness. He made his claims on the basis of long study of languages of many different kinds. The evidence itself brought him to see that any attempt to match the grammar of a language with the culture of its speakers or their ethnic origins was completely impossible. “Language,” “culture,” and “race” were independent variables. He turned the main part of von Humboldt’s legacy—European linguistic nationalism—upside down.

  Sapir showed that there is nothing “simple” about the languages of “simple” societies—and nothing especially “complex” about the languages of economically advanced ones. In his writings on language he showed like no one before him just how immensely varied the forms of language are and how their distribution among societies of very different kinds corresponds to no overarching pattern. But he did not reject every part of the inheritance of von Humboldt’s study of Basque. Different languages, because they are structured in different ways, make their speakers pay attention to different aspects of the world. Having to mark presence or absence in languages that have evidentials (see here and here), or being obliged to mark time in languages of the Western European type, lays down what he called mind grooves—habitual patterns of thought. The question for translation (and for anthropology) is this: Can we jump the grooves and move more or less satisfactorily from one “habitual pattern” to another?

  The view that you can’t ever really do this has become known as the Sapir-Whorf hypothesis, despite the fact that Edward Sapir never subscribed to the idea. The trouble with the simple form of this misnamed prejudice—that translation is impossible between any two languages because each language constructs a radically different mental world—is that if it were true you would not be able to know it. The parable of the NASA captain’s report of an alien language, given here, is one way to show how flawed the Sapir-Whorf hypothesis really is. A more sophisticated version of the same line of thinking runs up against equally powerful blocks. If we grant that different languages provide different kinds of tools for thinking but allow for substantial overlaps—without which there could be no translation—we are left with the idea that there are just some things in, let us say, French that can never be expressed in English, and vice versa. There would then be an area of “thinking in French” that was “ineffable” in any other tongue. That contradicts the axiom of effability, which, as we argued in chapter 13, is the sine qua non for translation to exist. It makes no difference to the argument against it whether the ineffable is held to be an attribute of God or of poetry, or a property of French.

  Sapir actually had much more interesting things to tell us about languages and the way they relate to social and especially intellectual life. Greek and Latin have served as the vehicles of sophisticated thinking that deals in abstract entities. Both have grammatical features that make it easy to create abstract nouns
from verbs, adjectives, and other nouns. A strong trace of the grammatical facility for creating abstract entities in classical languages can be seen in those large parts of the English vocabulary that have Latin and Greek roots: humanhumanity, justjustice, civilcivility, translatetranslation, calculus calculatecalculation, and so forth. Sapir’s point was that instead of saying that Latin and Greek are well suited to abstract thought, we should say, rather, that abstract thought is well suited to Greek and Latin and view the particular kinds of philosophical discourse that were developed by their speakers as a consequence of the grammar they used. The mind grooves laid down by the forms of a language are not prison walls but the hills and valleys of a mental landscape where some paths are easier to follow than others. If Plato had had Hopi to think with, he would not have come up with Platonic philosophy, that’s for sure—and that’s probably not a merely retrospective illusion based on the observable fact that there’s no Hopi speaker who thinks he is Plato. Hopi thinkers think something else. That does not make Hopi a primitive language unsuited to true thought. It means that speakers of what Sapir called “Average West European” are poorly equipped to engage in Hopi thought. To expand our minds and to become more fully civilized members of the human race, we should learn as many different languages as we can. The diversity of tongues is a treasure and a resource for thinking new thoughts.

  If you go into a Starbucks and ask for “coffee,” the barista most likely will give you a blank stare. To him the word means absolutely nothing. There are at least thirty-seven words for coffee in my local dialect of Coffeeshop Talk (or tok-kofi, as it would be called if I lived in Papua New Guinea). Unless you use one of these individuated terms, your utterance will seem baffling or produce an unwanted result. You should point this out next time anyone tells you that Eskimo has a hundred words for snow. If a Martian explorer should visit your local bar and deduce from the lingo that Average West Europeans lack a single word to designate the type that covers all tokens of small quantities of a hot or cold black or brown liquid in a disposable cup, and consequently pour scorn on your language as inappropriate to higher forms of interplanetary thought—well, now you can tell him where to get off.

  FIFTEEN

  Bibles and Bananas: The Vertical Axis of Translation Relations

  Let’s start with the math. For any three languages there are 3 × 2 = 6 different translation relations: FrenchRussian, Russian French; FrenchGerman, GermanFrench; RussianGerman and GermanRussian. Among any four there are 4 × 3 = 12; for n languages there are n × (n–1) directions of translation possible. So, since there are approximately seven thousand known languages in the world, there are 24,496,500 pairs of languages between which translation could in principle take place in either direction, giving rise to nearly 49 million potentially separate translation practices, each with its own tools and conventions. Translation is a universal capacity of human societies, and a level playing field of that size cannot be ruled out on purely theoretical grounds. In reality, however, the number of language pairs with established practices of translation is infinitesimal compared with all those that could exist.

  Translation does not happen every which way nowadays and never has. But in which ways does it happen? The fundamental answer, though a very broad one, is that it happens either UP or DOWN. As these are technical terms of my own invention, I’ve put them in small capitals.

  Every human language serves as a full means of communication for some community, and in that sense there is no hierarchy among them. But acts of translation, which are rarely isolated events, typically exploit and support an asymmetrical relationship between source and target tongues.

  Translation UP is toward a language of greater prestige than the source. The prestige may be the fruit of ancient tradition—as it was when Akkadian was translated into Sumerian in the Assyrian era, for example, or when translation into Latin was used to spread news of Marco Polo’s adventures far and wide (see here). At other times UP may be toward a language with a larger readership—typically, when then the target tongue is used, like French in nineteenth-century Russia, as a vehicle of intercultural communication. It may also simply be the language of the conquerors, or of a people with greater economic power, such as Russian in the Central Asian lands in the period of the U.S.S.R. Prestige can be located in a language also because it is the preferred vehicle of religious truths. Arabic, Latin, and Sanskrit, among others, have played this role at different times.

  Translation DOWN is toward a vernacular with a smaller audience than the source, or toward one with less cultural, economic, or religious prestige, or one not used as a vehicular tongue. Translation from German into Hungarian during the dual kingdom of the Austro-Hungarian Empire, for example, was DOWN, as is translation from English nowadays into any other tongue.

  The rank order of languages when seen as pairs is extremely hard for any individual act of translation to shift, but it is not stable over long periods of time. Sumerian, Greek, Syriac, Latin, English, and French, to take obvious examples, have seen their places in the pecking order change dramatically over the centuries. In addition, the ranking is often not all-encompassing. In specific fields, the relationship can be reversed or substantially modified. The standing of German as the language of a prestigious philosophical tradition means that shifting Kant, Hegel, or Heidegger into English (or French) is usually handled by translators as if they were translating DOWN; the translations of French novels into English in the nineteenth century exhibited the most obvious signs of that same direction of travel.

  What distinguishes translating UP from translating DOWN is this: translations toward the more general and more prestigious tongue are characteristically highly adaptive, erasing most of the traces of the text’s foreign origin; whereas translations DOWN tend to leave a visible residue of the source, because in those circumstances foreignness itself carries prestige. When Marcel Duhamel launched the Série Noire crime-fiction imprint in Paris just after the Second World War, for example, he ensured that the translations of the American novels he aimed to make popular in France used plenty of Americanisms in French. He went further: he insisted that his French-language authors (who provided more than half the texts) adopt American-sounding pseudonyms to deceive readers into thinking they were getting the real thing.

  However, the complexity and contradictions of language hierarchies are most richly illustrated by the history of Bible translation—in the West, to begin with, but subsequently worldwide.

  Bible translation got off to a slow start. The first foreign-language version of the Jewish Torah was the Septuagint, written in koiné Greek around 236 B.C.E. (see here). Other Greek-language versions followed, but it was not until shortly before the start of the Christian era that it came into Latin, around the same time that the Jews themselves began writing down the oral translations they had long practiced to make their holy texts accessible in Aramaic. Five centuries later there were still only eleven languages possessing versions of the Old and New Testaments (Greek, Latin, Aramaic, Syriac, Coptic, Armenian, Georgian, Old Gothic, Ge’ez, and Persian); and five more centuries were needed for the total to grow to nineteen, around the end of the first millennium. By the time printing was invented in the late fifteenth century, there were maybe fifty; by 1600 there were sixty-one, by 1700 there were seventy-four, and by 1800 there were eighty-one. A remarkable number, admittedly, but small change compared to what happened thereafter. In the course of the nineteenth century, more than five new languages were added every year, bringing the total to 620 by the turn of the twentieth century. Then things really began to shift. On average, one new Bible translation was completed every month between 1900 and 1999, and so, by the year 2000, the number of languages possessing all or part of the Old and New Testaments in translation shot up to 2,403.1

  Despite its roots in ancient and medieval times, in quantitative terms Bible translation is a preponderantly twentieth-century affair. Throughout many decades of that era, much of it was overseen by one
man, Eugene Nida, who has long been the most respected authority on Bible translation in the world.

  Nida never translated the Bible himself. He worked as linguistic consultant to the United Bible Societies, helping to exercise quality control over a great number of Bible translation projects that arose after the Second World War. In that capacity, he lectured all over the world and sought to explain in layman’s terms some of the contentious issues of language and culture that have been tackled from a different perspective in chapters of this book.

  Nida made a distinction between two kinds of equivalence in translation: formal equivalence, where the order of words and their standard or common meanings correspond closely to the syntax and vocabulary of the source; and dynamic equivalence (later renamed functional equivalence), where the translator substitutes for source-text expressions other ways of saying things with roughly the same force in the culture of the receiving society. He was an unashamed proponent of the view that, as far as the Bible was concerned, only dynamic equivalence would do. In that sense he was renewing the translator’s defense of the right to be free and not “literal.” Nida’s overriding concern, which is also that of the United Bible Societies, is that the holy scriptures be brought to all people—and that what is brought to them be the scriptures, as nearly as can be managed. A Bible that makes no immediate sense in the target language, or Bibles that can be read or understood only by trained theologians or priests, are not well suited to missionaries’ aims. Nida’s preference for dynamic equivalence was in the first place an encouragement to translators to sacrifice whatever was necessary to “get the message across.” As he titled one of the chapters of the handbook he co-authored with Jan de Waard: “Translating Means Translating Meaning.”2