Why translation mistakes can be dangerous
I work as a freelance translator. In fact, everyone is a translator. But since everyone does it, why do most translations suck so much?
According to research in translation studies, there are three main modes of translation: intralanguage, interlanguage, and intersemiotic. Let’s take Japanese as an example to see how these three modes work:
- intralingual: translation, or ‘rewording’: ‘an interpretation of verbal signs by means of other signs of the same language’;
- interlingual: translation, or ‘translation proper’: ‘an interpretation of verbal signs by means of some other language’; and
- intersemiotic: a translation, or ‘transmutation’: ‘an interpretation of verbal signs by means of non-verbal sign systems’ (Munday 2001: 5).
If you rephrase a sentence (e.g. a more or less elaborate language register or slang or dialect to say the same things), you have executed an intralingual trnaslation. The same applies if you reformulate a sentence written in Japanese Katakana by using Hiragana, Kanji, or Romanji. The translation of Roman numerals or numbers (e.g. VI) into Arabic (i.e. 6) is an intralingual translation. If you start speaking dialect (e.g. Bavarian) instead of Standard High German with your relatives, you are translating as well. The prefix intra- is Latin for “within”. Although some meaning may get lost as you switch registers or dialects or the cultural context, you will most likely be able to avoid misunderstandings, since you are still speaking within the framework of the same language. Talking about the weather in England serves the same purpose and encompasses the same terms, regardless whether you are the Queen or a member of the working class.
If you translate a sentence from Japanese to Chinese (regardless whether it is written in Hiragana, Katakana, Kanji, or Romanji), it is called an interlanguage translation: the Latin prefix inter- means “between”. This is where it gets tricky and where scholars argue at length about the question: Why do so many translations suck?
To say it with Lawrence Venuti and George Steiner: sometimes, the translation turns out to be “stronger” than the original text; which means that the primary (original) text fades in light of the secondary (translated) text. A text may be more beautiful in its translation than it was in the original — but the other phenomenon holds true as well. No translation of Hemingway or Harry Potter is as good as the original Hemingway; not only because English is the “natural” language of this text, but also because the (British) English cultural context and mindset of the author are vital parts of the text as an overall artifact. Semiotics and semiology (as a part of semiotics or as a discipline in its own right) discuss these issues within the text and of the text as an ensemble within cultural or societal contexts. Walter Benjamin even wrote an article about the aesthetics of punctuation marks — which can in my opinion only fully be understood by speakers of the German language, because even punctuation marks have different meanings in different languages and their cultural contexts. Some cultures use ! more or less often; and the English long dash “—” indicating a rather long or significant break is not so common in German. In essence, there is a huge debate going on between those who want the translation to be “invisible” and those who want it to be “visible”. An invisible translation would be a translation that allows the reader to read the translated text as if it was written in the target language. “It’s raining cats and dogs” would be translated into the German equivalent, which in English would mean: “It’s raining like someone is pouring water out of buckets”. The reader is thus lured into the illusion of reading an original text; there is no barrier of appropriation, other than it may be odd to read about the English traditions explained in Harry Potter while the text is in German — but that is something to be discussed in the context of intercultural translations; early 20th century fiction has indeed been translated interculturally, which means that an event taking place in London would take place in Berlin instead. Certain literary genres (e.g. cheap Romance dime novels in which the location doesn’t matter or even “otherworldly” Fantasy / Science Fiction) may even today be modified culturally — and we should keep in mind that I only know about translation in Western Europe and America; and that I have no idea how other cultures handle those questions.
A “visible” translation however allows the original text to “shine through” the translation. For example, this will occur if a sentence order that is typical in English is kept in the German translation; or when metaphors and word plays are translated literally for lack of a more suitable equivalent in the target language. This is actually the main reason why so many translations on the Internet suck, why machine translations still suck so much, and why I wrote this piece: the translations have been “done” (you cannot say “created” for something like that) too fast and with no feeling whatsoever for both the original and the target language or cultures. The reason why I wrote this post in the first place was such a fucked up (“pardon my French” — see, that’s another one you can’t translate into another language “just like that”) translation. The original English headline must have been something like: “Child marriages are expected to increase drastically”. In an interlanguage translation, you might reformulate the sentence to say: “We think that child marriages are going to drastically increase”. The interlanguage translation at Refinery29 could be interpreted as: “The number of child brides is supposed to increase drastically” — as if it was a good thing to be supported. Just because they had to make the headline short to fit in the newsletter or to be suitable for social media sharing, they completely changed its meaning. Interestingly, the URL says “www.refinery29.de/zahl-der-kinderbraeute-steigt-an” — which is a simple and correct statement: “the number of child brides is rising”. But it’s not a good headline for social media, that’s why it was changed — for the worse.
According to George Steiner, translation is also a “hermeneutic motion”, which means that it is a form of appropriation, internalization, intepretation, modification, and externalization. Steiner even names it a “violent act” and he lists the following four movements within the hermeneutic motion:
- Trust (élancement): the translator has to trust that everything the author writes down is meaningful and has a specific meaning.
- Aggression (penetration): the translator has to invade the text and break the code that lies within the text. Of all the possible meanings, he has to select the one that seems most appropriate for him right at this place and moment in time. This is especially problematic when it comes to translating scientific literature. The German word “Wechselwirkung” (Georg Simmel) has been translated as “interaction”; however, “Interaktion” in German exists as well and means something different than “Wechselwirkung”, whereas “interaction” can be both “Wechselwirkung” and “Interaktion”.
- Incorporation (embodiment): the translator has to make the text his own. A translation is not just transferring words from one sentence to another, it is also an act of composing a full text that is harmonious and consistent; just like a melody is more than just lining up musical notes. The flipside of the incorporation coin is the fact that the text will also change the target language. Maybe a new word will be introduced here that hasn’t been used yet, thus the target language will be “infected” by what the translator writes. Examples for that are especially common in entrepreneurial and tech slang; the “lean” in lean project management hasn’t been translated yet (and maybe never will) and the same holds true for the term “hacker”.
- Enactment of reciprocity (restitution of balance): in the last step, the translator has to “fill in the blanks” of meaning that have been created in the rupture of translating from the primary into the secondary text. This is an act of harmonization not within the translated text or between the translated text and the target language, but between the original and the translated text. The translator has to make sure that the author’s voice can still be heard in the translation, but that the readers of the translation will understand it, even if they do not know or speak the original language:
“The original text gains from the orders of diverse relationship and distance established between itself and the translations. Genuine translation will, therefore, seek to equalize, though the mediating steps may be lengthy and oblique. Where it falls short of the original, the authentic translation makes the autonomous virtues of the original more precisely visible. Where it surpasses the original, the real translation infers that the source-text possesses potentialities, elemental reserves as yet unrealized by itself.”
The problem here is that everyone who speaks two or more languages thinks they can also translate between them. But just as Steiner puts it here, the “mediating steps” between the languages “may be lengthy and oblique”. There are certain objective rules and patterns to translate by, but translation is (and of course there is a debate going on about that as well) a question of subjective “gut feeling” as well. That’s why only trained translators (trained by socialization, by experience, or by formal education) should work as translators — or as interpreters, which is just as hard.
I should add that there is another very important problem of universals one has to consider in the context of translation studies: who or what determined the way how we named the things in this world? Have the words come to us from some sort of “universal conception of the world”, which means they have already been predetermined by the world and we only had to “find” them (e.g. according to Noam Chomsky); or does language “create” nature and “determine” our way of thinking (relativism, e.g. according to George Steiner). For instance, a tree and a shrub belong to the same “biological family” of plants with trunks and branches, but they are completely different in our language. So translation studies are also interested how the concepts of nature “translated” into language (or written signs) in the first place.
But for now, let’s wrap up that post with the third mode of translation:
Just like intercultural translation, this mode actually requires a long post of its own. In the French sémiologie and sémiotique studies, a lot attention is brought to this topic. One interesting thought of the sémiotique is to explore how and why some signs (pictograms or images) “translated” into icons, symbols, and icons representing other things. Charles Sanders Peirce explains the three terms as follows:
- The icon (icône): a sign resembling the object
- The indice (indice): a sign that is like a symptom or a hint of its object
- The symbol (symbole): a sign with an abstract signification with no direct resemblance whatsoever
The dove is a symbol for piece. Smoke is an incide of fire (and a picture of smoke is an icon for smoke and an indice of fire); and a man/woman pictogram is most likely an indice for a toilet. The “cultural icon” is not the same as the icon in semiotics; Che Guevara is a “cultural icon” of a certain revolutionary spirit, but a semiotic symbol for revolution.
Sémiotique is the science of signs and their signification for “meaning-making” (including linguistics); sémiologie is the study of both verbal and non-verbal signs; including but not limited to their meaning in a context.
If you translate code (e.g. <a href=”…”>link</a>) into a GUI, you are probably conducting an intersemiotic translation as well. Emoticons and emojis are the most recent forms of intersemiotic translations — and they are said to be understood universally. But they are not! I once argued with a friend about the meaning of some of the emojis and we had to check the universal emoji chart with the written explanations to determine who was right. In fact, we were both wrong.
Many emojis are however universally understood (although some are discriminating or bringing up religious issues); and they can also help people diagnosed with autism spectrum disorders (or who are socially awkward, like me) to better understand the feelings of others. They also allow for faster and more immediate communication, which is both a good and a bad thing. So even intersemiotic translations can suck or result in misunderstandings.
Translation mistakes can be dangerous: intralanguage, if you use the wrong language register in a given situation, you may get into trouble. Interlanguage, if the translation falsifies the meaning of what was said in the original text; intrasemiotic, if the translation into non-verbal signs results in mistunderstandings. All these errors can be traced back to intercultural differences based on nationality, stratification, race, gender, religion, and other social markers.
The points mentioned are however just some of the reasons why translations tend to suck so much if they are not conducted by professionals — and this is the most important takeaway from this post: hire trained translators to do the job, whatever it may be.