Is That a Fish in Your Ear?

TWENTY A Question of Human Rights: Translation and the Spread of International Law

Translation studies as currently practiced in the academy concentrates heavily on the circulation of books, especially books of literary merit. But despite the six-figure numbers bandied about in our survey of global translation, literary works make up only a small part of translation in the world today.

Legal texts are translated in vaster quantities than books and in more varied directions. Dreary as it may seem to all but legal eagles, the translation of law is a prerequisite for the construction and maintenance of a global society. Without it, business and diplomacy would come to a stop. But there’s something quite important to learn from it. Law is the very model of an untranslatable text, because the language of law is self-enclosed and refers to nothing outside of itself. In practice, however, laws do get translated, because they must.

In France you can say impossible n’est pas français when you want to assert that something hard can nonetheless be done. “Impossible” doesn’t exist in other languages, either, when it comes to translation. Translation is a voluntary act.

Layfolk the world over know why the law is untranslatable. It’s written in a language of its own that is almost impossible to understand, and what can’t be understood can’t be translated. We pay our lawyers good money just to reassure us that they understand the small print on the contract we’re about to sign without reading it through to the end.

The words of law often look like words of the language you speak, but when they are legal terms, they are not. They don’t refer to anything outside of the social institution and intellectual system that the law constructs. You may have a pretty good idea what murder means when you use it in an English sentence, but what looks like murder to you may be first degree, second degree, manslaughter, homicide, or even collateral damage in a legal description of the event. The offense committed is determined by the legal system in force in the place where the killing occurred, and within that system it is determined only by the definitions of the offenses that the system distinguishes—by the words of the law as it has come to be written down.

In the first years of the twentieth century, a professor of linguistics at the University of Geneva gave a course of lectures about the nature of human languages. Ferdinand de Saussure never wrote the lectures down, but after his early death in 1913 students put their notes together and produced a Course in General Linguistics, which has served as a breviary for much of the thinking about language that has gone on since then. Whether or not Saussure’s teaching should be regarded as the last word on what language is overall, it’s an excellent tool for getting at the reasons why the language of law is such a tricky thing to translate.[135]

Saussure was already very learned in the history of languages, but in his lectures on general linguistics he sought to explain what a language is as a whole and systematic entity at any given point in time. His account was grounded in what was then a revolutionary new definition of the linguistic sign. A sign possesses both a material existence as a string of sounds or written marks, which he called the “signifier” (in French, signifiant); but it necessarily also has a power to mean—a “significandum,” or signifié. The sign is neither a signifier nor a significandum, but their combination, in a pairing so tight that the one can no more be separated from the other than the two sides of a single sheet. However, unlike a piece of paper, the two sides of the sign are attached to each other for no necessary reason—they just are attached that way. A sign in Saussure’s teaching has five special qualities. It has to be inherited, because the signs of a language can never be invented on the spot. It has to be shared, because signs mean what we agree they should mean, not just what some individual thinks they mean. It has to be unchangeable, because nobody can turn table into cable just for fun and still be using the same sign. It has to be free to be combined with other signs in an act of speech or writing. Finally, the inner relationship between signifier and signified that makes the two together coalesce as a sign has to be arbitrary: you can’t explain why the letters T-A-B-L-E as distinct from any others can be used to refer to a “table” except by saying they just do.

What, then, allows us to know that this sign is not that sign? That table and cable represent two different signs? Because they differ in respect of something that is a structural part not of “language” in any abstract sense but of the language called English. That is to say, the difference between the sounds represented by t and c is a basic element of the structure of the English language—and the entire structure that is the English language consists exclusively of sets of differences or oppositions of this fundamental kind. A language is then nothing other than a system of differences, because a sign in any language is exhaustively defined by all the things that it is not. What makes English not French or Chinese, for example, is the specific set of differences on which it is built. Rising and falling tones, for example, exist in any act of speech, but they are not parts of English. On the other hand, tones are signs in Chinese. Similarly, the difference between the sounds usually written as l and r is part of English but not of Japanese. To map the differences that are made use of in a language is to map the structure of the language itself.

Saussure’s approach to language makes each actual language sui generis, “of its own kind,” that is to say, an internally coherent system that can never be satisfactorily mapped onto any other. The automatic consequence is that no sign in any one language is fully identifiable with any sign in any other equally unique system of signs. Throughout the twentieth century, the Saussurean doctrine of the sign provided a reason for disregarding translation and ignoring the resources it gives for understanding how languages are used.

Saussure certainly didn’t have law in mind when he pursued this rich train of thought, but his doctrine of the sign is directly applicable to it. Law is a systematic use of language that relies for its coherence on the precise distinctions it makes between its own constituent terms. In any given legal language, “murder” is what the book of statutes and the records of cases judged have said it is—not what the ordinary language sign murder might be taken to mean among layfolk. Law is a system of signs.

Legal systems have different histories, different norms, different distinctions and ways of doing things. Even when the languages of different legal systems look the same—as in English and Scottish law, for example—the terms they use are not interchangeable. Each one is truly sui generis, constituted exclusively by the particular distinctions it makes. That’s the reason you can’t translate legal language—except that you must.

Defendants in many parts of the world are entitled to understand their own trial, and courts are obliged to find translators and interpreters for whatever languages are involved. They often have to scour far and wide. A request for an English–Hungarian interpreter for a murder trial in rural Scotland landed on my doorstep thirty years ago. The brave person who took on this awesome responsibility in the end had never seen a courtroom before and was barely more aware of the meaning of what was going on than the defendant herself. In the state of New Jersey today, the courts service employs many hundreds of mostly parttime interpreters, predominantly in Spanish, at low rates of pay and with little supervision. In New York City, where no fewer than 140 languages are represented, finding language intermediaries for court cases is a huge administrative task. In South Africa, too, where eleven languages now have official status, court interpreting is often a lamentable mess.[136] The language rights of linguistic minorities are important achievements, but their implementation often leaves a lot to be desired.

Court interpreting of this kind is internal to a single system of law: where the minority language does not have a strictly equivalent term—for prosecutor, attorney, or QC, for example—the source-language term is mostly used, as it is indeed the proper term for the individual or instance that matters at that point. But the interpreter may also have to add explanations or rephrase what is said in altogether different terms in order to make sure that what are understood are not just the words but the force and real-world consequences of the expression used. It is an extremely difficult and responsible job. It is rarely recognized as such.

Legal translation between the official languages of countries that have more than one—Canada, Belgium, or Finland, for example—is not exactly easy, but it is usually better rewarded and less stressful, partly because the translators often have legal training themselves. The issue of the incommensurability of legal systems does not really affect this kind of work, since it is the same language of law that is being expressed in both tongues. All the same, it is crucial that the two versions be construed in exactly the same way. Given the natural anisomorphism of languages, that is often very difficult to achieve. Law translation in such circumstances tends toward a homogenization of tongues—creating similar-sounding formal equivalents in the two versions of the law—to reduce the risk of a clever lawyer exploiting an apparent verbal discrepancy between two versions of the same text.

The trend toward making legal languages look the same when put into a different tongue seems to be driven on the one hand by a rather naïve idea of how languages work and on the other by an overriding concern that laws be seen as the same by all who fall under their sway. An illustration of the seemingly irresistible drift toward homogenized transnational legalese is provided by the history of the words used to express the broadest and least national juridical principle of all—the notion of fundamental human rights.

In 1789, the new revolutionary regime in France drew up its famous declaration of the rights of man and called it the Déclaration des droits de l’homme et du citoyen. Its purpose was to sweep away the religious and feudal underpinnings of the legal system inherited from the monarchy and to establish, under the authority of a Supreme Being who could not be called God lest that be seen as a sop to the Catholic Church, the basic rights of the citizen in his relationship to the new French state.

There was no question of these rights being accorded to any who were not fully emancipated citizens. As no one had yet thought of enfranchising women, the use of a masculine term, homme, was not just a convenience of language—it was what the declaration meant to say. It established and made explicit the rights of male subjects who were also citizens.

Unlike French or English, German has a noun for “human being” that covers men and women without distinction—ein Mensch is just any member of the human race. The other word for “man,” Mann, refers exclusively to a male, and in many contexts it also means “husband” or “married man.” That’s why Männerrechte can’t serve as a translation of droits de l’homme—it could too easily be taken to cover conjugal and domestic affairs, which “human rights” obviously do not. So French droits de l’homme was quite naturally represented by Menschenrechte in German translation. In fact, the declaration needed translating into German within a few years of its drafting, because large parts of what is now Germany were conquered by France and incorporated into the republic and then the empire, where they stayed until 1814.

Because Mensch cannot be translated directly into English without saying either more or less than the original, it became customary to refer to Menschenrechte in English as Human Rights, even though the phrase “Rights of Man” had been made famous by Thomas Paine’s pamphlet of 1791. The English formula of a generalizing adjective plus a plural noun (human + rights) is the third alternative form of a concept that began as a plural noun plus a singular noun phrase linked by a genitive (droits + de + l’homme) that had transited by way of a noun-plus-noun compound of which both parts are plural (Menschenrechte ). These changes in grammatical form engendered subtle shifts in implication that became apparent only in later times. “Human rights” was intended as a “translation” of les droits de l’homme et du citoyen, but it was something more, and something less. It went on to acquire a life—and a power—all its own.

From the inception of the United Nations Organization in 1945, Eleanor Roosevelt, the widow of the U.S. president, devoted her energies to promoting the declaration of a World Charter of Human Rights, which was duly adopted in 1948. In its official French version it is called Déclaration universelle des droits de l’homme. But one major thing had changed since the phrase droits de l’homme had first been monumentalized in 1789: in 1946, Frenchwomen became entitled to vote and were now citizens in the same sense (or almost the same sense) as men. The traditional use of the masculine homme to mean Mensch began to seem discriminatory. By the 1970s, French feminist campaigners were clamoring for a parallel declaration of les droits de la femme, even though such a thing, if it had ever been made, would most likely have excluded women from the provisions of the Universal Declaration of Human Rights—which would have been counterproductive, to say the least.

The German Menschenrechte would have solved the problem for everybody, but German is not an official UN language.[137] So it was the English adjectival formulation that was transported back into nearly all other European languages of the Germanic and Romance families—Italian diritti umani, Spanish derechos humanos, Swedish mänskliga rättigheter, and so on.

In French, however, the expression droits humains has a real problem: humain means, indistinctly, what we mean by human and what we mean by humane. Consequently, to call human rights droits humains in standard French puts them closer to humanitarian concerns, which are not the principal objects of laws on human rights.

These areas of ambiguity have led to the exclusion of the term human rights from many international instruments that deal with them: the International Covenant on Civil and Political Rights (1966), the International Covenant on Economic, Social and Cultural Rights (1966), the Convention on the Elimination of Discrimination Against Women (1979), and the Convention Against Torture (1984) all avoid the term, and even Europe, home of the original formulation, felt the need to complement it in the title of its European Convention for the Protection of Human Rights and Fundamental Freedoms (1953). With the passage of time and because of the spread of the ideas that it conveyed, human rights slowly ceased to be a term of law. As it percolated into general use, it found itself expelled from juridical language. Which is precisely what the systematic nature of legal language would require.

This has created an awkward issue for French. The historical priority of the revolutionary decree of 1789 has made France unwilling to dispense with what it still regards as the classical, transparent formulation of the idea.

The solution found was to change the language to make the old formulation still fit for use. The word Homme written with an uppercase letter now refers to men and women indistinctly and is the declared exact written equivalent of the German Mensch; whereas homme with a lowercase initial letter refers only to males. Although legally enforceable, the distinction is hard for people to remember. I’ve read many newspaper articles where upper- and lowercase homme are used in alternation, as substitutes for each other.

Russian, on the other hand, retains even in its current constitution a form of words copied directly and somewhat unnaturally from eighteenth-century French: права и свободы человека и гражданина, “rights and freedoms of man and citizen,” with the masculine serving for the general and the word for “human being” being used in place of the more plausible человечество, “humanity.”

Despite this French-style solution to the problematic status of a phrase that began life in French but has come back to plague it, the adjective-plus-noun version of “human rights” has continued to spread across all European languages—even in German, where it is hardly needed. In journalism and general usage, menschliche Rechte, patterned directly on human rights, is now used as an uncontentious alternative to Menschenrechte, which remains the proper term. Despite French language laws, moreover, droits humains is heard more and more often as the functional equivalent of droits de l’Homme. Rama Yade, French secretary of state “chargée des droits de l’Homme” from 2007 to 2009, was frequently called (and also called herself) the minister for droits humains.

It’s quite likely that this new use of humain in French will shunt its parallel sense of “humane” into the cognate word humanitaire and cause a minor reorganization of the lexical and semantic environment.

With the assistance of San Marino, the smallest member state of the UN, the Commission for Human Rights (UNCHR) encourages and disseminates translations of the Universal Declaration into all languages. The set currently exceeds three hundred, from Abkhaz to Zulu, and what is obvious from the effort so far is that, with only a few possible exceptions to add to Russian, the source language for translation is not French but English.

The intellectual, political, moral, and other consequences of the homogenization of languages into a single structure for the semantic field of “the human” is beyond the scope of this book. What we can say, however, is that the history and present state of the translation of human rights provides clear evidence that international law tends to create a language of its own. In this instance, which is undoubtedly typical, the language of international law—whatever language it seems to be in—is increasingly calibrated to English-language norms.

It could be seen as historical revenge, for England was under the thumb of Law French for many centuries. French was the language of law imposed by the Norman Conquest in 1066, but it was understood only by the ruling class. It continued to be used for centuries in the courts, in spite of or probably because of the fact that the majority of the population didn’t have a clue as to what was being said. But Law French underwent its own process of contamination from below over a period of six hundred years, adopting phrases, words, and grammatical structures from the actually dominant tongue. By the seventeenth century, the official language of English justice sounded like something out of the late Miles Kington’s comic column for The Times (London):

Quite different problems arise when a court of law seeks not only to prosecute defendants speaking a different tongue but to do so in a jurisdiction that has authority in a transnational sphere. The idea of there being an international law—universal norms of legitimate behavior not determined by any one sovereign state—is very recent. It first dawned in horrified reaction to the sufferings of troops in the Crimean War in 1857–58, then took its initial form in the various Geneva conventions about the rules of combat. The first major institution resting on an idea of international law was the League of Nations, set up in the aftermath of the First World War. But it was only the Second World War and awareness of the unspeakable persecutions carried out by the Nazi state that finally prompted sovereign nations to abandon their historical prerogatives and to establish a jurisdiction that sat above them all.

Translation was at the heart of the International Military Tribunal that opened in Nuremberg, Germany, in November 1945. What had to be established first was the overall legal procedure to be used, and that was no straightforward task. Two of the victorious Allies used a common-law system, and the other two, France and the U.S.S.R., like defeated Germany, had different but related versions of what is called civil law. In civil-law systems, defendants make opening and closing statements but do not participate in any other part of their own trial. They sit in a special place and cannot be subject to further examination, since that is supposed to have been conducted exhaustively by the examining magistrates, who brief the prosecuting team. In the common-law tradition, on the contrary, a defendant is held to be innocent until found guilty and is therefore treated formally as just another witness to the crime. That’s the reason American courtroom dramas are so much more exciting than French versions of the same genre. The Nuremberg court adopted a mixed system: it was not a jury trial, as it would have been had it been conducted entirely within British or American systems, but a tribunal judged by an international panel of judges. But it did impose cross-examination on the defendants, who were called to the witness box in German. However, in German, “witness” is Zeuge, and a Zeuge cannot give testimony at his own trial. The arguments about how to proceed at the Nuremberg Trials were not only about language but about the incommensurable differences among the languages, institutions, and customs of different languages of law. Law translation in international affairs always runs up against huge obstacles of that kind: law words do not mean the same thing when translated, and the institutions they serve are not the same.

Over the last sixty years, the scope and implementation of international law has expanded at a prodigious rate. The sought-after effect—sought after by political will but implemented by legal translation teams—is to bring the different meanings of words belonging to incommensurable systems of law into greater harmony, or, as critics of this process protest, to homogenize and standardize the idea of what the law is. Karen McAuliffe reports that lawyer-linguists working at the European Court of Justice are aware that European Union law is a legal system “built from approximations of law and language from different legal cultures and different legal languages, which come together to form a new supranational legal system with its own language.”[139]

This is exactly what the Saussurean theory of the sign would entail. What language scholars rarely take into account is that, given sufficient effort and political will, new systems can be made.