Sunday, November 22, 2020

Non-native bias? (in translation)


[Bearded dragons*]

At any get-together of translators, one topic that surely heats up the conversation is the virtue or lack thereof of translators translating into their non-native language. In its most extreme form, one side declares that non-native speakers, regardless of their language level, are unable to express themselves in writing like native sons and daughters, who have had a lifelong education in that language. On the other hand, those excluded translators retort by claiming that this generalization is, at best, a form of unjustified elitism and, at worst, an attempt to limit competition, noting that many native speakers, even translators, are unfamiliar with the grammar rules of their own language. As usual, the truth is more complex with the full expectation that many will disagree with me.

To clarify, it is clear that interpreters, as compared to translators, can and often should be natives in the source language, not target language, since their task often involves almost instantaneous understanding of the speech of people from all levels of society and a need to understand the subtext. For example, an APTI conference in Valencia, a professional interpreter recounted how difficult and important it was to understand the testimony of uneducated women in Yugoslavia during the War Crimes Tribunal because they were talking about rape, a taboo topic. The interpreter had to understand the code language of these people while the judges could cope with less-than-perfect English. So, the arguments below do not apply to interpreters.

To help non-translators understand the problem, it is first necessary to realize that the demands of written communication are different from those of spoken communication in terms of learning process and flexibility. Formal education is not required in order to speak a language. Numerous people worldwide have studied a foreign language, even for several years and are barely able to get a sentence out while others, with no education but merely the opportunity and necessity to use the language, not only express themselves clearly but live their daily lives in that language. By contrast, people attain the ability to express ideas in a clear, acceptable manner in writing through many years of formal schooling. To one degree or another, written language is a dialect that is only taught in schools although reading and speaking contribute to its acquisition. Furthermore, speaking is an instantaneous act that does not allow for editing and thus accepts individual differences in style and even grammar. When we judge spoken language, the essential issue is whether the listener understands with accuracy a secondary factor. It is true that people may note grammar and vocabulary errors, especially teachers and translators, but these mistakes are generally forgiven. On the other hand, written language, especially English, a hodgepodge of various roots, is a polished product, like a diamond. Since writers (and translators) have the time to edit, readers expect a perfect result in terms of grammar, syntax and style. The requirements of those elements may evolve but do so quite slowly. The “accepted” manner of writing, with small variations, is de rigueur. Any writer failing to comply with those rules is harshly judged as the sharp reactions to grammatical errors in comments in social media shows.  The scope of acceptable written communication is rather limited.

For this reason, native speakers generally categorically reject translation by non-natives in their language. Unless the foreigners were educated in that language from childhood, it is stated that they simply cannot write like a native but instead write in a hybrid style combining their native and second languages. Teachers call this language interference, which can also happen to natives after sufficient years living in a foreign country. Examples include Hebrish, where commas and preposition use is rather whimsical and Russian Engish, famous for its curious use of articles.  Consequently, translations by non-natives may be accurate in terms of content but will sound “translated”, not seamless, the legendary goal of all translation. Specifially,  Ideally, a proper translation should sound like it was an original work. Clearly, the vast majority of non-natives, Samual Beckett aside, are not capable of achieving that goal. Thus, in terms of attaining seamlessness, the nativists are correct.

Yet, supply and demand create a strong niche for non-native translators. First, even in the common language combinations such as Spanish-English, excellent non-native writing may be good enough for the customer or the customer may lack sufficient knowledge to detect the errors. On a larger scale, many languages used on one country with low population suffer from a lack of non-natives that have learned the language proficiently. For example, few Americans and Brits have learned Czech or Hungarian, to name a few. Thus, law students in the Czech Republics are also trained as English translators as there are insufficient numbers of native English translators in these combinations. In addition, the translation rates in a country may be too low to attract foreign-based translators, effectively giving local, non-native translators a virtual monopoly. The Russian Federation is the most striking example where non-native, local translation is the norm due to the price structure for the most part. Thus, in practice, non-native translation is rather common and acceptable in some markets.

As in most issues, the question whether it is acceptable or not for translators to translate into their non-native language is not black or white. Ideally, translators should only translate into their native language since they have the proper ear for that language. On the other hand, market conditions create a situation requiring translation by non-native speakers. Reality is often a shade of grey.

* Caption pictures to allow the blind access. Picture via Pixabay


  1. you make no mention of people with several native languages

  2. What does that mean several native languages, spoken at home or learned and used at school?

  3. Mr Rifkin, thank you for an interesting and rather argumentative article on this perennial topic.

    Well, me as a native Russian interpreter and translator have had those language interferences since my childhood.

    Imagine, a grammar school in Russian starting back from brinkmanship of the Soviets in 1990, the Bachelor of Arts in Translation at the university English and Russian being both languages of instruction. Meanwhile raised by Russian and Azerbaijani speaking parents of then labour class.

    Woof, I guess, mine is beyond that interference twilight zone.

    Thanks again for being generous enough to consistently write on important translation related twisters.

    1. Thank you.
      BTW, I truly love the Russian language but had the misfotrune of being young and stubborn and having to take 9 courses from a Soviet teacher rather spoiled my chances of mastering the language. I wish you a successful year.

  4. There are no such things as “source” and “target” languages per se in interpreting. It is incorrect to insist that a less than optimal rendition into the target language is acceptable in interpreting but not in translation (I would say it is unacceptable in both). Another assumption made here is that any mistakes made would always be grammatical and not propositional (that the judge, for example, could still receive “accurate” evidence albeit with some grammatical errors). This clearly is not the case.

    The economic side of things is true. But to say that it is the sole reason for why people who translate into their non-native language are still in demand is to gloss over the issue. A lot of times that I have had to proof and edit work, without the source the translation may read very fluent like an original piece of writing. But that is highly deceptive because beneath that lies something along the spectrum of outright misunderstanding of terms/phrases/sentences to misperception of nuances in the non-native language which are then transferred to the resultant translation work.

    On the other side of the coin, translation from one’s native language into non-native presents its own dangers of varying degrees of seriousness. Possible errors could range from simple grammatical errors which don’t affect so much of the transfer of meaning to major distortions. But is that more problematic than the serious or major errors that could arise from the lack of understanding of a non-native language when translating into one’s native language? I don’t think so.

    In the ideal world a job should probably be translated by a native tongue and reviewed by someone whose native language is the target language. Of course the world isnt ideal when economic considerations usually come first.

  5. Thank you for your comments. My experience (as well as my wife's) is that we can generally "smell" a translation not because of grammar errors but syntatical features typical of the interfering language. Your experience may be different.

  6. One point besides price is that native speakers of the source language are often stronger on the accuracy side of the accuracy-fluency spectrum.
    I've taught Germans and non-Germans bidirectional translation (DE/EN), and the Germans often understood many nuances in German texts that foreigners did not. Their English is also good enough to produce a text without formal errors.
    On the other side, foreigners often understood better which nuances their readers might care to know.
    I found that both natives and non-natives make style errors like redundancy or insufficient paraphrasing, because they haven't properly grasped the purpose, which then appear as "interferences" to the reader.
    I know that English is special in this regard, I doubt there are many other languages which are mastered at such a high level by second-language speakers.

  7. Good point. I hadn't considered that there are many times of nuances.

  8. Please proofread yourself before writing an article, I am not native in English, but I spotted mistakes, one is a very basic grammar rule (plural adjective) [sic]

  9. Alas, everybody has a weakness. It is hard to see one's own mistakes, even one day after I write the text. It just shows how the standards of written text is higher than oral ones.