Too literal, or word for word, japanese translations are the defeat of good writing style

The problem with literal translations in/from Japanese

Word for word translation is the end of the good writing style.

Literal translations are often an attempt to hide a translator’s inexperience or lack of ability

There are two opposing schools of thought in the translation world: 1) translations must always be literal, or word for word, and 2) translations should be adapted and localized while "improving" upon the source text if necessary. In other words, they should be semantic, or free. In any case, translation from and into Japanese should never be done too literally.

What is literal or word for word translation?

Literal translation aims to find the best possible matches between individual words in the source and target languages. This is achieved by giving more attention to single words than to sentence-level meaning, a strategy which carries the risk of creating a confusing and syntactically jumbled target text that is difficult to understand. The meanings of words in literal translations can sometimes differ significantly from what was intended, despite the efforts of the translator to perfectly match each and every source word with its closest equivalent in the target language. In contrast, semantic or free translation seeks to match meaning at the sentence level instead, an approach which greatly benefits the natural flow of the text and makes it more immediately comprehensible.

Dorothy Parker

Pretty women are like good translations: very seldom they are faithful.

Dorothy Parker
Poet, short story writer, critic, and satirist

Literal translations can make sense in some circumstances and for very specific purposes, which however are not the norm. For example, and although only to a certain extent, it can be more justified in technical/scientific or specialized translations when the main requirement is the absolute fidelity to the content, while it should be avoided when working with the vast majority of other texts such as those within the literary and business domains, otherwise it would end up breaking the naturalness of the source language and - even if basically correct - would be much less effective to the ears of the reader. The result of an anomalous positioning of two or more words that can give rise to misunderstandings of meaning slows down the perception of the translated text and can produce an alienating effect on the reader.

Literal translation from Japanese or Italian/English

Compared to Indo-European, and in particular Romance or neo-Latin languages such as Italian or English, the Japanese language has a fundamentally different structure: the parts of a sentence are arranged in a different order; virtually all superfluous information is discarded; the subject of a sentence is omitted except in cases where its presence is absolutely required for understanding; there are more nuanced and ambiguous ways to express a thought; nouns, verbs, and adjectives are not modified according to gender, number, or grammatical person; there is no distinction between definite and indefinite articles; there is only a loose distinction between verbs and adjectives and, well, it's profoundly irrational, all peculiarities that on the whole contribute to make it in a linguistic world apart.

Written Japanese in particular features less interconnection between statements, relying more on redundancy and standard formulations for which there is no equivalent in Italian and English and therefore are difficult to translate. The use of synonyms is not as important in Japanese as it is in Italian or English, and since the language doesn’t contain pronominal particles or relative pronouns, the object of a sentence is often repeatedly identified. In the Italian language, in particular, this would negatively impact the fluidity and readability of a text.

For these reasons, as a general rule, literal translation from Japanese should be avoided as much as possible, let alone translating it too literally. Or better, the above said sintax differences are so marked that translating word for word from Japanese is not possible at all. But to be clear, we don’t recommend translating word for word from English to Italian either, despite the structural similarities between the two. The specific issue with applying this method to Japanese is that its syntactic structure will inevitably shape the target text in a way that sounds quite unnatural to an Italian reader.

For Italian → Japanese translators the situation is is relatively simpler, even if only up to a certain point. In fact, since the Italian language is much more precise and rich in semantic instruments than the Japanese language, if mastered there is not much way to make mistakes. Besides, among other things most Japanese translators have a very solid knowledge of written English, which should make it easier for them to better understand Italian too. Certainly, some types of texts as in the legal and literary domain are rather twisted and sometimes so marked by an artificial search for the effect to seriously test even the best Japanese translators, whom in such cases can only take refuge in a word for word translation. That is, greater adherence to the form than to the substance of the content, ending up producing something that even the Japanese user would find hard to fully understand.

In any case, since translatiing in a more natural way from Japanese requires more effort, time and even a certain determination - especially here in Japan it must always be assumed that translations are checked by the typical Japanese reviewer who as such is tendentially prone to literal translation - the temptation to go down the relatively safest road of this translation mode is not negligible.

Below is an extreme yet real-world example of excessive literal translation from Japanese—a common case of unnecessary repetition, something which is unfortunately not confined to the realm of technical translation. (A) is a sentence from a Japanese printer manual, which has been literally translated into English (B). Sentence (C) is a word-for-word translation of either the Japanese or English versions, done by an inexperienced translator who is a native speaker of Italian.

(A) Original Japanese text:

プリンターを使う前に、プリンターがアラームを出さないようにプリンターのカバーを閉じてください。そしてプリンターの電源ケーブルをコンセントに差し込んでください。

(B) Word-for-word English translation:

Before using the printer please close the main cover of the printer so that the printer will not output an alarm. Be also sure to plug in the printer’s power cable.

(C) Word-for-word Italian translation:

Prima di usare la stampante chiudere il coperchio principale della stampante in modo che la stampante stessa non emetta un allarme. Assicurarsi inoltre di collegare il cavo di alimentazione della stampante.

The word "printer" appears four times in (A). In the Japanese language, this is not uncommon and is often a necessity. But because the Japanese text has been translated literally, the word is also used four times in the English and Italian translations. Most native speakers of Italian will recognize (C) as a clear example of poorly written prose (in this case, a single instance of “printer” is plenty). However, this unnaturalness is not always the fault of the translator; it is sometimes the result of a Japanese translation company or end client demanding strict adherence to the original text in a misguided attempt to maintain uniformity with other languages.

Let’s look at another common feature of the Japanese language: redundancy:

(A) Original Japanese text:

監督者の明確な事前の許可がない限り、この部屋に入ってはいけません。許可なしで入室してはなりません。

(B) English literal translation:

Do not enter this room without first obtaining the authorization from the security staff. Make sure not to enter without authorization.

To an English reader, it is obvious that the second sentence is redundant and should be eliminated for clarity. However, cases like these are quite common in technical texts and can become a source of anxiety for Japanese translation companies or direct clients who are hesitant to accept what they deem to be considerable deviation from the original text, whether that source is Japanese or English. By their rationale, such modifications could indicate an inaccurate translation.

In Italian schools, it is (hopefully still) taught that repetitions and redundancies like those above should be avoided, and this is very easy to do given the wealth of grammatical tools available within the Italian language.

And it is hard to argue that even technical or specialized translations would not benefit from a careful balance of content and writing style that improves their readability and interest while still maintaining the repetition and terminological consistency required to prevent any confusion or misinterpretation.

It is unfortunate then that the typical Japanese customer has little knowledge or interest in foreign languages and believes that all Western languages should conform to the structure of English, because such an individual may view the elimination of redundancy as a hallmark of low quality.

A curious way of checking translation quality

All of this leads to a common query: "Why does the word "printer" appear only once in the Italian translation but four times in both the original Japanese text and its English translation?"

While a patient explanation is sometimes sufficient to dispel any doubts, in other cases the client might decide to hire an external editor to verify the accuracy of the translation.

The problem is that editors in Japan are rarely native speakers of the target language, which means that they often zealously deconstruct translations that were carefully crafted for Italian readers. By rewriting a translation to be perfectly “true” (i.e. literal) to the original at the client’s request, such an editor can quickly create a troublesome situation for an Italian translator. The same scenario can also occur when a Japanese text is translated by a non-native speaker of the target language then that result is in turn translated into another language.

This happens rather often in multilingual projects where the client requires that translations into the various languages be made not from the original Japanese text but from a translation of it (commonly an English version). These “intermediary” translations are usually done by a Japanese translator or even an employee of the client and are in most cases plagued by the repetitions and redundancies illustrated above.

And since they reflect so clearly the structure of the original Japanese document, subsequent translations into other languages can end up being difficult to read, confusing, and often fundamentally wrong if the translator is not careful.

Why are there so many literal translations then?

  • Translating literally is easier and takes less time because the translator doesn’t have to search for synonyms or rephrase the text to match the target audience. It makes the translation process less mentally taxing, which boosts productivity and allows translators to offer cheaper rates.
  • Literal translations are sometimes the result of a lack of context, in which case it’s hard to blame the translator. There may be no way to know from the source text how a particular statement will be used, what will precede or follow it, or what it is really referring to. Although good translators will submit questions to clients for clarification, there isn’t much they can do if responses aren’t forthcoming because the client is busy or has trouble understanding why a professional would need to ask such questions in the first place. What sort of professional are they? Don’t they know languages well?
  • Translation companies have been known to require literal translation because it facilitates what they perceive as an expedient form of quality control: an Excel file with each language contained neatly in its own column so that the project manager can quickly notice any deviation. This sentence was translated this way in French and German. Why not in Italian?
  • Sometimes it’s just a by-product of the widespread decline in linguistic ability. Cultural globalization tends to homogenize and simplify human thought and consequently written expression. Horrid examples of anglicism abound in both the Italian and Japanese languages, many of which are incorrect usages.
  • Increased use of machine translation (even as a pre-translation tool) has also contributed to an extent. Stylistic blunders like the examples provided above are surprisingly prevalent and can often be traced back to the use (or misuse) of machine translation that purports to "make everything easier and faster."

Localization goes beyond merely translating

Most professional translators would agree that localizing a source text—interpreting and translating it so that it is suited for the target audience—is the most critical component of translation quality and involves much more than simply translating words.

Translations from Japanese into Italian in particular need to be reexamined from a different perspective sometimes, perhaps to coax them more toward something that Italians are familiar with and can relate to. This might involve getting rid of unnecessary repetitions, enriching the text with choice synonyms, or substantially revising the text to arrive at something that sounds like it was written directly in Italian. These are all techniques which can be applied to a translation that is fundamentally correct without altering its message or meaning.

Failing to take this step when it is needed is doing a disservice to the translation world and shows a lack of respect for the beauty and efficiency of the Italian language. It is especially common among inexperienced translators who may wish to avoid getting into time-consuming debates with the client or significant overhauls by a bad editor.