9

It’s common to hear in translation theory that when you translate a work you change it fundamentally, for example because words have multiple meanings and very specific connotations depending on their context of use, so every translation is in some way a rewriting.

I feel somewhat skeptical of this point of view. While there are doubtless things in some language that are untranslatable, because they make use of very language-specific features, often there are very strong correspondences between languages - the word “cat” means something very similar when translated across languages - and the same is true for a number of expressions and idioms, and certain sentence structures or grammatical structures.

So has anyone formalised a technique in which the translator tries to ensure a standard translation, always seeking the word, expression or grammatical structure that corresponds most closely?

7
  • 1
    I don't think that standard is really the right word here; rather, I think what you mean is "close translation".
    – ruakh
    Mar 6, 2022 at 0:33
  • @ruakh Translations are said to be accurate or idiomatic. That's for all translation by the way.
    – Lambie
    Jan 20, 2023 at 19:02
  • The word "cat" can be translated into Russian as кот, кошка, котёнок, котик, кошечка, котёночек, киса, киска, etc.
    – Anixx
    Jan 21, 2023 at 7:38
  • And cat in formal English can refer to big cats, or in informal English to a man. That said, I think the premise of the question is correct, that there are sentences that are relatively straightforward to translate generically, at least between certain language pairs. But there may be less than we assume, because even things like pronoun or tense, which are hard to avoid, can be fundamentally not 1:1. Jan 21, 2023 at 13:36
  • 1
    There is no such thing as a standard translation for texts.
    – Lambie
    May 8, 2023 at 23:39

3 Answers 3

12

Language is a form of communication made up of many tools: e.g., phonetics, lexis, morphology, syntax, semantics, and pragmatics. To be efficient, the number of tools must be limited, e.g., no language has words that can uniquely describe the shape of every leaf, every tree, or every head of hair on the planet. Even different dialects of the same language often have a different range of tools to express similar meanings, so there is almost always a mismatch in translations that requires interpretation, guesswork, or accommodation.

When a sentence is ambiguous in some way, different languages will use different tools to resolve the ambiguity. On the other hand, sometimes the tool in a particular language is more specialized than the actual communication needs require. Translating between languages involves deciding on where to draw the line between under- and over-specification because of the mismatch between the tools depending on communication goals, style, and real-world knowledge.

An example of the mismatch at the level of phonetics is that even such a simple English sentence as "the dog is chasing the rat" requires a choice of sentence stress when spoken in order to express the pragmatics of what information you are trying to update. If I translate this sentence into Latin, I must make a choice of word order that will correspond to different pragmatic meanings. Having the verb first would tell what is happening. Having the word for "dog" first would tell what the dog was doing. Having "the rat" first would tell what was happening to the rat. In a translation, you would often have to guess at the intended pragmatics of the English, since the pragmatic tools of the languages don't line up perfectly.

An example at the level of lexis is that some languages routinely divide the terms for siblings by gender, others by age. In translating between such languages, you must have real world information to give a natural translation or use some awkward construction. Without a very large context, it can be unclear whether it is important or not to convey the extra information about gender or relative age that the language does not routinely convey.

At the level of morphology, I could say that some languages routinely force the speaker to convey information they may be indifferent to. In the English royal anthem "God Save the King/Queen," you must specify the gender of the current monarch. Such distinctions are made in the morphology of some languages, but I use an English lexical example for simplicity. In translating this phrase into some languages, it would be highly marked to specify the gender and so this information would not be routinely conveyed. Translators will differ on what to do in translating between such languages.

The same morphological mismatch exists for the category of number between many languages. If I translate "the dogs chase the rats" into Classical Arabic, I must choose between a form that specifies two and a form that specifies three or more. If I translate that same sentence into Japanese, I could not easily differentiate the number of either noun unless it were an essential part of the communication.

At the level of syntax, languages differ widely on what they require you to specify about tense, aspect, and mood. A good example is the profound mismatch between English and Mandarin. The same sentence "The dogs are chasing the cats" would have a "standard" default translation in a teaching grammar, but the Mandarin equivalent would normally not disambiguate between past and present reference. The English is also ambiguous from a Mandarin perspective, because there is a mandatory morpheme in Mandarin used to specify if the sentence is referring to a new situation relevant to the conversational context. Similarly, a simple phrase like "How old are you" would differ in syntax depending on whether it was said by a passport inspector or a relative, but there would be cases in between where a speaker would have free choice. Translators will differ in how or whether to communicate such nuances.

At the level of semantics, consider this example using a word mentioned in the question. In English, you can say: "A lion is a big cat," but you can't say: "A wolf is big dog." We can assume that these semantics will differ widely among the worlds languages, just as some languages do not distinguish between "rats" and "mice" or might have more fine-grained distinctions for ordinary reference. Even such a seemingly fundamental word as "mouth" covers different facial geography between languages, requiring a translator to estimate what information is necessary to convey.

At the level of pragmatics, consider what often happens if you meet someone new and want to chat. A common question in English aimed at an older adult who you suspect is married is: "Do you have any kids?" If talking to a child, you might ask, "How many brothers and sisters do you have?" In Mandarin, in the same situation, you might ask, "How many people are in your household?" These questions are just means of filling out the familial picture of the person you have met, but will yield different information at different rates. They usually have the same pragmatic goal, but with different cultural assumptions; however, they also can be asked for other purposes. In translating such questions between the languages, you usually have to translate them into an unnatural form in order for the responder to give an answer along the lines expected by the questioner.

If you put all these mismatches together, it is easy to see that there can be nothing like a "standard" translation beyond what is often conventional in teaching grammars.

Even if one has perfect knowledge of what a speaker means by an utterance, if a language over-specifies something, the speaker can change course mid conversation and smoothly continue using the information already supplied for a new pragmatic purpose. If the language underspecifies something, the speaker will have to introduce the new information first before being able to change course.

Lastly, there is an irreducible issue of opinion and style. Take the first words of the Hebrew Bible, Genesis 1:1: "In the beginning, God created the heavens and the earth." Hebrew has several words and several forms that are the equivalent of "God" and so you have to decide whether to convey that information if possible. The Hebrew word translated as "heavens" above could equally be translated as "Heaven," "sky," and "skies." The choice has theological implications. If you translate these words into Chinese, you confront yet different choices about theology and emphasis, since many terms for "God" and "heavens" are loaded with different cultural connotations specific to Chinese traditional culture.

Let's take a simple example of translation between two dialects of American English. If you are familiar with some dialects of African American speech, how would translate into standard American English: "You'll stay away from him, because he be acting crazy"? This dialect often requires a differentiation between second person singular and plural in commands. Do you force that difference into standard English, which cannot make this distinction? Do you translate the command simply as "stay away" or do you rephrase it with "you/you all/you guys/youse need to stay away" or with something else. I have heard all these variations in pronouns depending on region.

The verbal aspect of "he be acting crazy" is explicitly habitual and is different in syntax and meaning from "he acting crazy" in this dialect. Do you translate this difference into the standard English version in some way, even though it lacks such a form? If you do, what word should you add to translate it? Do you translate it as "he is often/always/usually crazy" or do you take a pragmatic approach and translate it as something like: "Stay away, because he seems to be acting crazy all the time"?

If it is hard to have a standard translation even between two dialects of English, it is easier to see why translators often differ in their translations between different languages.

7
  • 1
    Well, I think you raise the good point that undecidable questions of how to translate something are inescapable. But we might reframe the issue as that there is a high degree of consensus between people of what the most standard translation of some sentences is, and not of some other sentences; even if we can’t explicate what defines “standard” in terms of rules, we can still agree in our subjective judgment. So what I’m saying is some sentences do have strong correspondences between languages and some do not and even though the notion cannot be formalised that doesn’t mean it doesn’t exist. Mar 4, 2022 at 23:38
  • @PeterElbert: If all you're suggesting is that it sometimes happens that a certain sentence is pretty clearly the best translation of a given sentence into a given language, then -- sure, yes, that can happen (at least if the contexts are known, or assumed to contain no surprises). But that's not what the question is about; the question speaks of "translat[ing] a work", not just a sentence, and of "a technique [...] to ensure" a result, not just whether the result might occasionally be possible.
    – ruakh
    Mar 6, 2022 at 8:38
  • That’s true, perhaps my question was exploring a number of connected thoughts and now I should ask a new question to get to the heart of the matter as I see it. The current question could be rephrased to ask whether in professional / industrial translation contexts there are guidelines or even vetting procedures to settle on - given a highly specific context - what the best translation is. But then I’d be pretty curious to see how such a system plays out on literary fiction - not living in a dichotomy of “adventurous reinterpretation” vs “robotically literal translation” but believing in Mar 21, 2022 at 13:59
  • something beyond either: the most appropriate and accurate translation taking all subjective factors into account. For example in the example above someone points out how they don’t really say “How do you do?” in Chinese or something, they say “Have you eaten?” or something. So given the cultural context that is the most accurate translation overall. The point is if it could be shown empirically that human judgment produces consistent consensus then you could ask an individual to produce a translation according to that benchmark - so you could ask of a person for a “standard” translation. Mar 21, 2022 at 14:05
  • I really don't see how this answer really addresses the question at all.
    – Lambie
    Jan 21, 2023 at 17:49
6

There is a clear no answer to this question. There are different styles of doing translations, giving different weights to naturalness in the target language and fidelity to the original language. Those styles are most evident in the translation of literary or religious works (compare different translation of the Bible or the Quran for this as illustrations) and are probably smaller for translations of technical or legal texts. The article on translation in the Wikipedia gives a longer account on that.

4
  • Well, it does seem like there are some good theoretical concepts to discover there (en.m.wikipedia.org/wiki/Dynamic_and_formal_equivalence). But the subjective freedom and creativity of those theories makes me think an argument can be made for what is “standard”, and there is probably someone who has written about it. Actually, I think “fidelity” is basically what I meant, but not awkward word for word translation, sort of like fidelity with a small degree of naturalness, at least if what is “standard” is to be framed by those terms. Mar 4, 2022 at 18:14
  • @PeterElbert: Well, it depends on who wants that translation and what they intend to use it for. If it is a translation of a work of fiction, naturalness is probably more important than fidelity, possibly to the point of completely rewriting some parts to be better understood by the intended audience. OTOH if you're translating a legal document, then that's probably a terrible idea.
    – Kevin
    Mar 5, 2022 at 23:52
  • I’ve been thinking about this and I think one way forward is to show empirically that people have convergence in terms of when you ask them in some neutral way what the “best” translation for a certain phrase is (given a real world context) and then basically considering a theory or explanation for why there’s so much convergence in people’s opinion. Mar 21, 2022 at 13:55
  • This is not a matter of style. Some meanings can have standard translations and others are a matter of opinion. In response to the OP, the answer is no, and I explain why in my answer.
    – Lambie
    Jan 15 at 19:27
5

Has anyone formalised a technique in which the translator tries to ensure a standard translation, always seeking the word, expression or grammatical structure that corresponds most closely?

First of all, I am a professional translator with 35 years of experience and have worked as a literary, technical and business translator. Also, I have done a lot of academic work (translation of articles in sociology, economics, history, etc.).

Translation is broken down into types of translation: commercial (business), law and patents, economics, advertising/marketing, scientific, etc. There is overlap in many of these areas.

In the field of translation, the idea of a standard translation can be applied to terms like standard phrasing or to someone's work (of art or academic work) but overall, it would not be used to describe a translation. Non-literary translations are said to be idiomatic or unidiomatic, among other quality parameters. Translators are required to have a near-native grasp of an L2 to competently translate. I won't go into machine translation here.

Whatever the multiple meanings are of words, in context, they do not have multiple meanings in practice unless one is dealing with poetry, which is a specialized type of translation and certainly does not include anything like "standard translation".

Generally speaking, in non-literary translation translators aim to capture meaning. So, translating into L1 requires in-depth knowledge of the L2 and how the L2 works in various fields or subject matter areas from which one is translating.

The problem is not whether un chat is a cat. The issue is what is idiomatic in L1 and whether it properly conveys the meaning in the L2 text. So, the French phrase "des bottes en cuir" are leather boots. However, in a literary context, this might be a context where one would translate it as: the boots made of leather, bearing in mind that in French terms like en plastique or en cuir are how materials are referred to where in English one usually uses a simple adjective like leather or plastic (leather boots, plastic boots). This is, all told, a simple example.

Knowledge of the L2 is crucial for translation because of the mistakes or non-idiomatic translations that may arise due to a lack of in-depth knowledge of an L2.

Translators who know their L2 well and know the field in which they are translating often will look for standard translations for certain technical-type terms but otherwise the term is not used to characterize a text which is translated.

Translation involves moving meaning from L2 to L1. So what is sought in non-literary translations (and that does not including advertising/marketing) is to properly capture and convey meaning. Often, there can be more than one way to translate something even when the translation is for, say, business. Beginning translators often miss this point: translation is really about good writing and conveying the accurate meaning of the L2 text in L1 language.

For example, in English "Arriving early is important." means the same thing as: "It is important to arrive early". The point is which one should be used? That depends on who is sending this message to whom and over what channel. Both those utterances can be used to translate a Spanish utterance such as: "Es importante llegar temprano". No hand will come down from on high to say one is standard and the other is not.

In the end, translation is an art, not a science.

And George Steiner, one of the 20th century theorists of translation (though somewhat dated), especially literary translation thought this:

George Steiner’s monograph After Babel is a living plea for translation. For him, translation as meaning transfer is not an isolated phenomenon, but rather a basic anthropological activity and ‘formally and pragmatically implicit in every act of communication’ (Steiner 1975/1992: xii; cf. also 49).

After Babel: Aspects of Language and Translation

This answer only touches the tip of the iceberg but at least it has the merit of being properly focused on the subject.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.