The Awkward Issue of L3

Translation Studies deals typically with relations between entities called 1 and 2 (language, culture, text)—in an older jargon, ‘source’ and ‘target.’ These long-standing binaries seem solid enough. But what if a third language, L3, is involved? Asking how to deal with Latin tags in French texts translated into English, for instance, or Yiddish words in American English for translation into German, may seem like a diversion or an irrelevance. But however quirky and curious specific L3 problems may be, the question of what to do about language in T1 that is not L1 raises issues that are central to the phenomena that Translation Studies seeks to study. And it raises them in a way that makes the standard approaches to translation practice and theory quite problematic.

L3 is usually seen as lying on the wiggly boundary of translation and translatability. Famous, unendingly rehashed examples include the French-language passages of Tolstoy’s War and Peace. Many people have asked: what is the right, or the best, or the least awful way of dealing with French-language utterances in a text to be translated from Russian to English? Practically, there are only two things you can do: translate the French into English, or leave it in French. But both solutions to the L3 problem fall foul of the very idea of ‘translation’ as it is commonly construed.

Leaving the passages in French is certainly not ‘a translation’ of the text, since nothing gets changed by the hand of a translator. Yet everything is changed by the shift in the linguistic environment of the L3 passage. The cultural meaning of the speaking of French by nineteenth-century Russians is particular, local, and specific; it contributes to many of the higher-level meanings of the novel (individual characterization, social description, the politics of language and nation…). But the cultural meaning of French spoken by English-language standins for nineteenth-century Russians bears no relation. For obvious reasons, it is not possible to represent in English the social meaning attached to speaking French in Russian: no language possesses grammatical, lexical or stylistic resources to mark metalinguistic meanings attaching to another language and its linguistic environment. You can footnote it, you can explain it, you can understand it (up to a point), but introductions, footnotes, glosses and essays do not count as translation.

Conversely, translating the passage into English cannot represent what is meant by the L3 in L1, since it reduces the polyglot to the monoglot, and elides a significant part of the utterance-meaning of the original. Again, you can indicate— through italics, footnotes, glosses, translator’s introduction and so on—what is going on and what it might mean, but translating the French into English does not fit with the general notion of translation as an attempt to create an ‘equivalent effect.’ The effect isn’t remotely equivalent. It can’t be. It’s not that you can’t quite meet the demand for equivalent effect when L3 is involved: you can’t even try.

War and Peace can be sidelined as a unique case that should not distract us from the central concerns of translation studies. But that would be doing an injustice to our own language use. In an issue of the New York Times in fall 2008, for example a solution to the “financial tsunami” then rocking the nation was described as the “big megillah.”¹ So, in an article indubitably written in English we have an event in the U.S. economy named in Japanese and swiftly solved in Hebrew. Conventions laid down by generations of copy-editors declare words printed in roman face to be words of the English language, and words in italics to be words of a foreign language ‘used as foreign words’ in English. The distinction is just about as useful as any of the rules invented by printers. By this rule, ‘megillah’ is now a word of English. But of course it isn’t. It’s the Hebrew word for ‘scroll,’ and the supplementary meanings it has when used in English are taken over from Yiddish. Tsunami is no less a word of Japanese when printed in roman than when it is said in Japanese. So what is English? One obvious answer is that to all practical purposes, English isn’t. Or rather, texts written in any L are also characteristically written in other Ls as well. That is one of the things that make translation so interesting, and so poorly explained by the ‘doctrine of equivalent effect.’

L3 is handled at micro-level, the level of individual words, and at macro-level, the level of whole chunks of text, in ways that vary in accordance with ill-analyzed and still barely recognized assumptions and aspirations that are cultural, social, historical, and sometimes pedagogical. For example, the Latin tag nec plus ultra when said in French is ne plus ultra. Why? No reason, other than the anecdotal and probably now irrecoverable histories of Latin quotations in France and England. But what of a French text about Occupied Paris that refers to the Kommandatur? A British copy-editor did not blink at German in English, and what’s more, put it in roman, not italic, as a proper name (of a building, as a metonym for the functions fulfilled by the people who worked there); but the American copy-editor did blink, and wanted a footnote or a glossary entry. Is that the sign of a cultural difference between two contemporary receiving audiences? Or is it evidence that from one professional point of view, there is no difference in English between German and French, such that ‘Kommandatur’ can be treated as a remnant of L1 in L2?

Examples of bizarre practices in the handling of L3 are legion. From languages like French (broadly speaking, the Romance languages), Latin quotations are most often translated into English, on the assumption that Latin is less well-known in the English-speaking world. But Latin is not really well known in France. Latin words are often a little more motivated because of the historical roots of the Romance languages, but the function of a Latin tag in a French text is much more cultural than it is semantic. Translating Latin when translating French into English also says something like, “Anglo-Saxons are better at plain speaking than pretentious continentals.” Both in L1 and L2 there’s a game going on that has very little to do with meaning, and a great deal to do with ideology, in the specific form of mental constructs that I would like to call Language Fictions.

Language Fictions include the unformulated assumptions we make about the meaning of the use of foreign languages in our own and in other tongues. In French it is an established custom to indicate in a footnote any occurrence of L3, especially if it is also, as it often is, L2: en français dans le texte is so common an indication in all kinds of books published in Paris that it serves as a figure of speech in many contexts having nothing to do with translation. There is no English equivalent. Why do the French need to know that characters in many foreign novels—not just by Tolstoy, but by Dostoevsky (l’homme de la nature et de la vérité, in Notes from the Underground), Evelyn Waugh, Thomas Mann and Paul Auster—come out more or less often with phrases and sentences in French? No reason, if we restrict ourselves to issues of translation, questions of meaning, and ‘equivalent effect.’ But the French would not footnote megillah, yarmolke, kippah, schul, or drek as ‘en juif dans le texte’: they translate them, from Hebrew or Yiddish as the case may be, without signaling that in the original they are in something like L3. They do not hesitate to transliterate Italian chiaroscuro as clair-obscur; but never once have I seen Inch-allah footnoted as ‘en arabe dans le texte.’ En français dans le texte adds precisely nothing to the meaning of the phrase not translated, it contributes only in the vaguest way to the force of the expression not translated, and it doesn’t even affect the translator’s rate of pay. But it does say quite a lot about France’s view of its place in the world, and of the universal aspiration of its language. En français dans le texte is a not so subtle reminder of a national-cultural ideology, insisting that French is still a global language. Vive la France!.

These few examples might seem too slight for constructing a hypothesis, let alone for demolishing a temple. But the degree to which the issue of L3 is so little understood or analyzed can be underscored by the following examples of non-translation which turn out to be mistranslation of the most obvious kind. They arise in the specific loop when L3 is L2—when you are translating a text into a language (L2) and find that T1 contains material that is in L2 already. Now, given the printers’ rules referred to above, in a well-edited English text any French sentence is going to be printed in italics, signaling that it is in a foreign tongue. The rule is the same in French: in Senkovksi’s Bolshoy Vykhod U Satany the Devil speaks all languages, and when he says Gut! in German, the French translator and editor put it, unchanged, in German, in italics. And similarly for his brief outbursts in Polish, Italian, etc. But the French translator of Anna Seghers Das wirkliche Blau does something different in the following instance:

Sur la table, il y avait une montagne de brioches et aussi de la tequila destinée aux hommes quand ils viendraient.*

*Note du traducteur: Eau-de-vie mexicaine que l’on fabrique en distillant le suc d’une variété d’agave appelée maguey.²

She puts the word tequila — a word rather well-known in English, if not in bars world-wide—in italics, as a word of “Mexican,” perhaps. She adds a footnote explicating this dubious foreignism as “Mexican brandy distilled from the sap of a variety of agave called maguey.” Agave, a word now lexicalized in French, is printed in roman, but maguey is put in italics. What is going on here? The footnote is evidence of the fact that the translator does not expect a French-language reader to have a clue about non-French alcoholic beverages. In a country famous for its wine and brandy, and with the highest per capita alcohol consumption in the developed world, it is not very likely that the average French (L2) reader has never heard of tequila. What is more likely is that the translator, whose presumed expertise is in German (L1) culture, wasn’t sure what it was. However, what we learn most obviously from this curious addition has nothing to do with translation per se. What it shows to my mind is that irrespective of L1, whether you are starting from German or from Swahili, you end up with explicating the foreign with the same or another word that is foreign when written in L2—in this case, maguey. Just a simple bottle of hooch on a German table shows that you can’t really write much in French without also writing some other language as well. I suspect that goes for all other natural languages too.

Just as crazy is the fate of a sentence in French in a novel by Gloria Emerson. In italics in English, to signal its status as a foreign body, it goes back into French with its italics still on.³ The translator has meticulously preserved the form of the original (Nida would call it “formal equivalence”) but non-translation of these formal marks utterly transforms the force of the sentence. Because it is in French in French, being in italics signals it as significant, important or emphatic in a way entirely absent from the (French in English) original. Not translating offers no way out of the fruitful paradox of translinguistic writing.

Let me suggest some lines of inquiry that might begin to sort all this out:

First case

When L3 is neither L1 or L2 (Tolstoy into English, for example) then the choices made between the two options available map the force and cultural status of L3 in L2, alongside various ancillary phenomena such as the education system, wars won and lost, current political sensitivities, and so on (in Tolstoy’s case, of the meanings that can be attached to English people speaking French). If sufficient material could be assembled, it might be possible to write an interesting sociolinguistic history of English in its relation to other languages through the decisions made by translators when confronted with L3. It would map the Empire’s rise and fall as well as the history of English vocabulary, adding detailed substance to both. What those choices do not map, represent or enact, however, are meanings related to the L1 culture. No attempt can possibly be made to represent the significance of the L3 for the reader of the original. The ‘doctrine of equivalent effect’ is just not applicable. None of the conventional dichotomies of translation studies—foreignizing versus domesticating, literal versus free, formal versus dynamic, etc.—are pertinent to the issue of L3, which is why this awkward issue ought to prompt some serious thought.

Second case

When L3 is L2, then there is no choice. But there is a huge shift. Hélène Berr, a Jewish woman whose diary of life in wartime Paris was recently published in English,⁴ was a brilliant student of English literature at the Sorbonne. In her harrowing chronicle of a descent into hell, she has recourse to the literary treasury she had accumulated, and quotes Keats, Shelley, Shakespeare and many other English writers; she also slips into English vocabulary for many of the small things she is fond of (marmalade, potatoes) and for sayings and expressions that have stuck in her mind. This makes her French just a little bit pretentious and pedantic, and often quite difficult to understand for those many French readers who do not know English as well as she (the French edition translates the English in parentheses after the foreign word or words, making it much longer than the manuscript.) But in English ‘translation,’ Hélène Berr is as clear as a bell: she reads more smoothly than in the original French. God forbid! Translators are not supposed to make texts better than they were. So shocking is this possibility that the rights-owners of the original French were horrified by my mentioning in print that the English version is less difficult for English readers than the original is for the French. It was only a storm in a demitasse, but the emotional energy of the original argument underscored how deep-seated are assumptions about the hierarchical relation between T1 and T2.

The awkward issue of L3 in these and hundreds of other cases ought to lead us towards a reconfiguration of translation studies with these basic and commonsensical principles in mind:

• • To know a language is also to know that other languages are different. This is only apparently trivial. Most translations and almost all discussion of them seem to be based on the idea that L2 readers have no clue what L1 is like. But in fact, we could not even speak English unless we were aware that it was different from any other tongue.

• • To know a language is to be able to use (elements of) other languages within it. If you speak English you can also say sayonara, amigo and adopt as time goes by whatever foreign expression you care or need to. To know a language is to have a translingual capacity, exercised to a degree that is determined by social, historical, economic, and cultural factors, but not by properly linguistic criteria at all.

• • No language exists as an autonomous linguistic universe. Overlaps between languages—words and expressions that float across the so-called language barrier—are precisely what is natural about natural languages.

• • For these reasons, the act of translation cannot be interestingly described as an attempt to produce in the reader/hearer of T2 an effect equivalent to the one created by T1 on the L1 hearer/reader. The issue of L3 can be used as a loose thread that allows us to pick apart this particular piece of intellectual embroidery.

• • When L3 is L2, it is not at all impertinent to suggest that a translation may have qualities — including readability, even elegance — that are not present in L1. The hierarchy of source and target is not as simple as our romantic inheritance tells us. Some books in L1 are just screaming out to be put into L2, where they can find a more powerful, economic or poetic voice. Every translator knows that. We should dare to say so.

Envoi

On holiday on the West Coast of Ireland, sauntering along a beach in springtime where local fishermen were clubbing a squawking host of baby seals to death, Giulio Lepschy, a noted linguist, got into conversation with a local. The latter was delighted to discover that his interlocutor was not an Englishman. Pointing to the marine mammals messing up the beach, the Irishman said: “Well now, you Italians have t’e right word for t’ese pests. Bloody focas is what t’ey are!”⁵

Lepschy’s story shows that the device of translinguistic remotivation is not just a ‘problem’ for students of James Joyce or a curio in the discourse of the learned, the witty and the pretentious. It’s a basic feature of ordinary language use, and it serves to remind us that the primitive model of language relations implied by the use of ‘L1’ and ‘L2’ in contemporary Translation Studies needs to be replaced with something a little more complicated.