Sod the source text!

When we translate marketing material, or indeed any text with a commercial interest, there is one thing that is infinitely more important than the source text: the poor bugger who is going to read the translation.

The relation between the source and target texts has been discussed to death, usually presented as variants of literal versus free translation. In the early 19^th century, the German philosopher Friedrich Schleiermacher stated that a translator has two options: to leave the author alone and focus on the reader, or to leave the reader alone and focus on the author.

Put bluntly, sod the source text or sod the audience.

Yep that’s it, pretty much. Image from themetapicture.com.

Schleiermacher himself was a strong advocate for the latter option, to focus on the author. More recently, his views on the matter have been picked up by the eminent American-Italian translator and researcher Lawrence Venuti. Both Schleiermacher and Venuti are concerned with literary translation, and more specifically with translation to English from other languages. For Venuti, translation is a political act that should resist the common practice of Americanising vernacular works of art.

In the field of commercial translation however, there is little to gain from this method. The aim of promoting a product is hardly to render the style of the author behind the text – a copywriter is often just as anonymous as a translator. The aim is rather to engage the reader, inform them of their options and ultimately, to sell the product. As for the language-political viewpoint, translating from English into smaller languages (which constitutes the most common language direction in Europe) is, if anything, the opposite to the situation Venuti describes. Hence translation of marketing texts calls for a method that entails localising, editing, removing, adding and paraphrasing.

Not always the biggest and the best. Image from panix.com.

I don’t mean that we should write about polar bears if the text is about lions. By all means, stick to the facts and respect the client’s predetermined terminology. I’m saying that we should present the lions in a way that is appealing and makes sense to the target audience. Why do we even want a translation to be identical to the original? It is not going to be read by the same people. Also, the source text is not necessarily bigger and better by default – even the most well-written copy may not work in the target culture. If somebody really is crazy enough to compare a translation with the original, segment by segment, surely the interesting thing is not how much the translation resembles the original, but how the translator has chosen to negotiate the inevitable differences between the source and target cultures.

The term transcreation is sometimes used to address what I’m talking about. There are different definitions of this concept – I’d say it refers to translations that have been adapted to work as texts in their own right and not just in relation to the original. Personally I am reluctant to use this term, perhaps because I cannot differentiate between transcreation and translation in general. I prefer to think of it as bilingual writing. Translation without an element of transcreation is a pointless activity that we hopefully can leave to Google Translate in the near future.

In my experience, human translators resort to literal translations for one of the following reasons:

• They are under pressure to produce an insane number of words and do not have the time to think about presentation;
• they did not quite understand the meaning of the original text (this is not always the translator’s fault);
• they are scared to deviate from the structure of the original (in case somebody questions their interpretation, they can defend themselves by saying “but it is what the source text says”);
• they do not have sufficient writing skills to express themselves differently.

As for machine translation programs, it is obviously futile for human translators to compete in terms of quantity – surely the only way forward is to provide value in terms of quality. After all, Google is getting pretty good at mass-producing uninspiring, ugly lions.

Couldn’t agree more. Image from panix.com.

Read more

Friedrich Schleiermacher on Wikipedia
Lawrence Venuti on Wikipedia

On intralingual translation and how it can shed light on “translation proper”

In the influential essay ”On Linguistic Aspects of Translation”, Roman Jakobson divided translation into three types: intralingual translation (monolingual translation), interlingual translation (bilingual translation) and intersemiotic translation (translation of words into symbols). After establishing that interlingual translation is ”translation proper”, the rest of the paper focuses on translation issues specific for bilingual translation.

That was 1959. Naturally, the field of translation has gone through huge changes since then, and Jakobson’s categories have been explored and expanded by a number of different researches. One thing that seems to linger, however, is the notion that ”translation proper” is the act of translating from one language to another. This is a shame, as monolingual translation can shed light on the translation process in general.

three pints of beer and a couple of non-teetotallers

Let’s revisit Jakobson’s proposals and see how he described them:

1) Intralingual translation or rewording is an interpretation of verbal signs by means of other signs of the same language.
2) Interlingual translation or translation proper is an interpretation of verbal signs by means of some other language.
3) Intersemiotic translation or transmutation is an interpretation of verbal signs by means of signs of nonverbal sign systems.

Note how only intralingual translation is described as a “rewording”. This is unfortunate, as it somehow implies that interlingual translation is not a rewording, which in turn reinforces the old-fashioned idea that there are exact equivalents that can replace each other in bilingual translation – i.e. as long as we manage to identify the “correct” equivalent in the target language, the translation process does not constitute a rewording. This may not be Jakobson’s intended meaning, but it is nevertheless a possible interpretation.

So what do we mean by intralingual translation and when do we use it? If somebody asks me what a teetotaller is, I will try to explain the meaning, i.e. I will translate from English to English. How I do that depends on the situation. If a child asks me, I might describe teetotallers as “people who choose not to drink alcohol”. If the question comes up at a wine tasting conference, a more formal response, like “those who abstain from intoxicating beverages”, may be more appropriate. If I want to try to be funny in a group of friends, I could claim that teetotal is the same as “unsociable” or “boring”. Etc.

And if I am asked to translate “teetotaller” into another language? Again, the answer depends: Who has requested the translation? What do we know about the target audience? Where will it be published? The actual translation does not take place on the level of individual words but in relation to the context. Hence the strategies described above are just as valid for bilingual translators, and the question is not so much whether a translation is “correct” but rather if it communicates the intended message in a suitable manner. If that does not constitute a “rewording”, I don’t know what does.

Final words: In Polish, the word for translation – tłumaczyć – also means “to explain”. I say cheers to that.

Reference
Roman Jakobson’s On Linguistic Aspects of Translation

Disputes within the translation industry

The translation industry is not exactly a harmonious, happy family where everyone silently agrees on best practice. Below are few topics that are constantly discussed among translators, reviewers, agencies and clients.

Translators should stick as closely to the original as possible.

This view, in my experience a common attitude at many translation agencies, is problematic for many reasons. It is a well-known fact that translations that follow the original to closely sound a bit rubbish in the target language. Moreover, following the original doesn’t necessarily mean that it is more faithful than a translation that doesn’t. And finally, it gets much more interesting for everyone if we allow translators to use their judgement to find the most suitable way to express the meaning of the original text for the intended target audience.

Translation qualifications are unnecessary.

Quite on the contrary, good translation courses prepare the students for the translation industry, give them valuable practice with possible specialisation, allow them to get familiar with CAT tools and machine translation, teach them translation theory and about the history of translation and/or encourage them to think analytically about the translation process.

Having said that, it is clearly fully possible to be a decent translator without formal qualifications – and a translation qualification does not guarantee decent translations. Just like it is fully possible to be a decent say, website designer without a computing degree, and a computing degree does not guarantee that a person is a decent webdesigner.

Translation theory is pointless.

Translation theory may not be everyone’s cup of tea, but it is certainly not pointless for those who appreciate it. Personally I find it not just hugely interesting, but also of practical use in my daily work as a translator.

Translators don’t need CAT tools.

I am always surprised when I meet professional translators who think they would not benefit from using a CAT tool in their work. For me, the most amazing thing with CAT tools is that they are excellent at remembering everything I don’t. Every translation assignment will inevitably involve some degree of research, using dictionaries and thesauri, checking terminology and thinking about the most useful expressions for the project. Next time I get a new assignment from the same client, the CAT tool allows me to access my previous translations in a second and use the relevant translation memory for reference. Without a CAT tool, I would have to start from scratch every time I translate a new text, relying only on what I may remember from previous tasks.