The Case for Non-Native Translators?

While the idea of a translator working out of their native tongue is the stuff of nightmares for many established professionals, I recently took part in a round table discussion that briefly centred around questioning whether or not this method could actually be beneficial to translation.

Though the topic only came up in passing and there was little real support for the notion at the time, the discussion stuck with me and I thought that I would dedicate today’s post to entertaining this idea.

In the real world of translation, professional associations such as the ITI state that:

members shall translate only into a language that is either (i) their mother tongue or language of habitual use, or (ii) one in which they have satisfied the Institute that they have equal competence

[I find this definition of a “native language” as a “language of habitual use” to be a much more balanced distinction than “mother tongue” and, as such, here we’re using “non-native” to designate a language that is not that of habitual use]

However, while these guidelines forbid non-native translation, Antoine Berman’s conception of the aim of translation as receiving the “foreign” as “foreign” – where this encounter with the “foreign” enables the receiving culture to grow and develop – hints at the potential benefits of such a method (after all, who better to transmit this “foreignness” or “otherness” than somebody with extensive experience of that very other?).

So, aside from the obvious situation involving a rare language pair, where the use of a non-native translator is inevitable (though not necessarily preferable), what else can non-native translation offer us?


A seemingly obvious benefit is that native speakers should have a comprehensive command of the source language and culture, allowing them to unpick the most convoluted jargon with greater ease than a non-native – undoubtedly a valuable attribute in a translator.

However, simply being a native speaker doesn’t guarantee comprehension. There are plenty of topics in English that are beyond my understanding and, as a result, you won’t be able to get to the heart of a text without hiring a specialist.

Furthermore, unless this command of the source language is mirrored by the translator’s target language ability, that understanding won’t be reflected in the translation. It’s all well and good to perfectly understand a source text but if you can’t transfer that understanding into the target text then it is all for nothing.


Taking Berman’s lead once again, it can be argued that the non-standard use of a target language (influenced by an increased proximity to the source language) may result in innovation that can help the receiving language to develop.

One interesting anecdote I came across recently was the story behind the name of famous video game character Donkey Kong (the tie-wearing gorilla at the top of this post). Allegedly, creator Shigeru Miyamoto believed “donkey” could mean “stupid/stubborn” (depending on which sources you read) in English and assumed that the name Donkey Kong would convey the sense of “stupid/stubborn ape” to an American audience.

When he suggested this name to Nintendo in America he was initially ridiculed but the name stuck and, subsequently, some of those intended associations will have surely attached themselves to the English term given the character’s iconic status. While this usage stems from poor target language skills rather than his proximity to the source language, it neatly demonstrates the potential value of non-native language use.

In reality, though, how often will this be the case? The other side of the coin is that this improper usage can simply see us disregarding necessary grammar rules and misrepresenting a language by failing to adequately capture potentially key elements of meaning. For every Donkey Kong there will be a thousand translations like the one below (and much worse, no doubt) that negate any potential positive effects.


With concerns remaining over whether or not translators get suitable recognition for their work, perhaps the implementation of non-native translation and these “innovative” usages offers us a way to differentiate translators’ work as an independent form of writing and increase the translator’s visibility. If all translations were produced by non-natives, we’d have a much better idea of when we are reading a text in translation.

However, when confronted with today’s standards of translation, which require perfectly readable, error-free target texts (a standard that professional guidelines – including the stipulation outlined above – are seemingly designed to uphold), these new translations are likely to be rejected outright and, if they did get released into the world, they would perhaps just serve to make translation synonymous with unreadable garbage.

Sure, we’d know when a text has been translated but we’d also be likely to quickly decide to steer clear of any translations in the future… Not quite the positive result we had in mind.

Clearly, concrete benefits of non-native over native translation are fairly hard to find and, while collaboration between native and non-native translators could prove to be a beneficial course of action in some situations, this solution is often impractical.

Ultimately, however, I find this entire discussion to be overly general and prescriptive. The basic label of native or non-native is not enough to assess what really matters: translation competence.

The fact that a translator is stronger in the source language than the target language does not necessarily dictate that their translations will be poor as a range of other, supporting skills also play an important role.

Both the Donkey Kong and the “don’t touch yourself” examples demonstrate a clear lack of linguistic/cultural knowledge and research skills and these are a vital part of any translator’s skill set.

However, while we cannot routinely assume that non-native translations will always be inferior, perhaps the most important element of translation competence in contemporary professional practice is the ability to produce error-free target language texts and, in my experience at least, that ability is much more likely to be found in a qualified “native” speaker.

– If you’re looking for a different take on the discussion, check out these interesting posts: ‘The Importance Of Being A Native‘ & ‘Native or Non-native… This Is the Question‘. Enjoy!

Managing mistakes: Getting ‘wrong’ right

While the topic for today’s post is something that I touched upon in my recent guide to translating medical texts, I feel that it is something so engrained in the translation act that it is worth a more detailed exploration.

In that post (under the sub-heading of ‘Poorly written texts’) I explained that technical texts are very frequently marred by the presence of many mistakes (typographical, stylistic, grammatical, formatting…) and here I want to further explore this feature of texts along with an array of weird and wonderful examples from a variety of sources in order to highlight the issues that they pose for the translator’s work.

Firstly, just to show that I’m not exaggerating the importance of this issue, why not have a quick glance at this text from the African Journal of Neurological Sciences – an interesting article and a quite typical medical journal entry, and one that is littered with examples of each of the aforementioned errors:

Examples of formatting and stylistic errors include the incorrectly placed space following the phrase ‘une alternative intéressant .’ (line 93), the unnecessary usage of a capital for ‘Plus de 77%…’ (line 61) as well as the inconsistent or unnecessary usage of full stops and commas in decimals, capitals for references to tables, and last but not least between using numerals or their full, written versions e.g. ‘Quatre enfants’ and ‘2 cas’ (lines 38 and 29 respectively).

In addition to these examples, instances of typographical and grammatical errors include the misspellings of ‘haemophilus’ as ‘heamophilus’ (23), ‘aiguë’ as ‘aigue’ (64), ‘tétraventriculaires’ as ‘tétraventiculaires’ (80) and ‘hypertension’ as ‘hyperetension’ (figure 3) or the lack of agreement for ‘séries européenne(s)’ (55). Finally, the worst examples of errors must be the reference to table 6 on line 33 when there are only five tables and the usage of the non-existent ‘dérivation standard’ (table 4) – presumably a guess at the French version of the English term ‘standard deviation’ (which is actually ‘écart type’).

This may seem like an extreme example yet the truth is that mistakes can and will happen in almost any kind of text (even in big budget Hollywood movies, as we will see) and the importance of taking the time and effort to proofread work cannot be overstated.

Films obviously offer different challenges to written texts in the way that there are aesthetic and audio elements serving to complicate matters. For example, how could the translator go about dealing with this example from Aladdin? The animators clearly tried to make all the writing in the film look Arabic yet when we see the faces of Jafar and the Sultan reading a scroll, their eyes move from left to right (Arabic is read right to left). In this instance – when the translator has no control over the error – they must shift their focus to ensuring that they don’t introduce any new errors in their work, which can see the translation process turn into something resembling a game of chinese whispers as is the case with the error below.

In this scene from French film Amélie (below, 0:28 onwards) where the death of Princess Diana is announced on the news, the narrator declares that the death occurred ‘dans la nuit du 30 août 1997′ [on the night of 30th August 1997] when it was in fact widely acknowledged to be the 31st August – a small detail, perhaps, but one which causes the translator/subtitler a bit of a problem.

When a factual error such as this is found in a ST, the translator faces the dilemma of whether to remain faithful to their source and knowingly reproduce the error with the risk of it later being attributed to the translator themselves or whether to stray from the text and correct the error. In his excellent 2001 Revising and Editing for Translators, Brian Mossop suggests that ‘factual errors should be corrected if they seem to be inadvertent but not if they are important as author’s ignorance of the facts’ and, as the former appears to be the case here, the translator should presumably correct the error.

However, in the English subtitle the translator not only neglects to change the date (perhaps it was not spotted or they deemed it unimportant) but also renders the French simply as ‘on August 30th 1997′ which, while seemingly only a small change, actually serves to compound the error (the night of the 30th and early hours of the 31st could perhaps be seen as nearly interchangeable, whereas this subtitle loses that margin for error).

Facts and figures are an obvious place to find errors that can easily be overlooked as it is easy to assume that something is correct, yet it is always worth double checking.

While it is quite obvious in this previous example that the error is not intentional, there are other times when the author’s intentions are must more difficult to ascertain:

In his 2012 film Django Unchained – which toys with the traditional spaghetti western, known for playing fast and loose with history – Quentin Tarantino treats us to all kinds of historical inaccuracies and anachronisms (how about the use of dynamite years before its invention…). In the introduction we’re greeted with the subtitle ‘1858: Two Years Before The Civil War.’ when the war actually began in 1861.

Yet here, despite the obvious errors, the translator’s task is more difficult than ever. Given Tarantino’s previous work (think Inglorious Basterds) it is actually quite likely that the mistakes are intentional and – rather counter-intuitively – the correct decision for the translator to take is to maintain the errors even when it is in their control to change them, as with the opening subtitle.

Ultimately, it is important that the translator fulfils their extended role as proofreader and editor in both ensuring that mistakes present in the source text are spotted and that they don’t introduce more mistakes in their own work. This task is made slightly more manageable when working with written tasks due to the fact that the translator can often contact clients or authors to seek clarification on certain issues while the single-layered medium of writing (i.e. no contradictory visual or audio issues to contend with as in the examples above) reduces some of the burden. However, despite its importance, it remains an under-appreciated element of the translators task.