No translator's translation is perfect. As the story goes, St. Jerome, the patron saint of translators, made a notable translation interpretation that depicted Moses with horns in medieval and Renaissance art. Jerome translated the Hebrew word for "shining" or "emitting rays" as "horned." This is a good example of how interpretation affects human translation, and how a word's meaning in one language may not directly translate into another. Alternatively, no machine translation is perfect. Machine translation can not pick up on cultural nuances or interpret meaning as might be required. Translation isn't perfect, but rather interpretative, and this is how we are continuously working to make sure it is reflecting the original message.
Even the most experienced translators aren't perfect. There’s no such thing as a “perfect” translation. The Italians have a saying: “Traduttore, traditore.” It means “translator, traitor,” referencing the difficulty of being able to translate a work of literature from one language to another perfectly. Especially in literature and poetry, it’s often difficult to come up with a translation that exactly conveys all the nuances of the source text.
Some people assume translation is merely substituting one word for another, but it's not quite that simple. There's frequently more than one way to translate a word, a phrase or a sentence. As a result, people naturally have different opinions about which way is best.
Translators are human, and humans are imperfect. Even the most skilled translators may interpret something differently than a counterpart. Artificial intelligence is not free from translation errors either.
There’s no doubt that machine translation has improved at a rapid pace in recent years. However, it’s still far from perfect. For example, a 2017 study of the types of errors most common in machine translation found that “Out of 63 sentences, 60 sentences contained at least one error.”
how to improve translation quality
No translator is perfect, but strategies are implemented to avoid mistakes in the final product. Whether the translator is human or machine, here are some of the ways we work to be as close to perfect as possible.
Glossaries, style guides, and translation memories
Translation glossaries and style guides reduce ambiguity, helping your translators choose the right word every time. A glossary is a set of key terms, along with your preferred translation for them. A style guide helps translators understand the style and tone they should use when translating your documents. Glossaries and style guides improve consistency, ensuring that your translations stay on-brand.
Similarly, translation memories and computer-assisted translation software can save time and ensure that common phrases are translated the same way every time. After all, once you have a good translation, there’s no sense in re-translating it every time.
Human post-editing for machine translation
MT algorithms don’t understand all of the nuances of grammar or vocabulary. There’s no doubt that MT can save time and money, but it’s often best to have the output post-edited by a skilled translator to catch and correct any unintentional errors.
The translation quality control process
Because no translator can be perfect, quality assurance is one of the most critical parts of the translation process. After the translator is done with the initial translation, an editor fluent in the target language reviews it for errors. Often, the content is again examined by the desktop publishing team, and by the client's staff during the in-country review.
At United Language Group, our quality commitment informs everything we do. No translator can be perfect, but we're continually improving our procedures to serve our clients better. Take a look at the services we offer, the industries we serve, and contact us for your next translation project.
[et_bloom_inline optin_id="optin_3"] Google+ Facebook LinkedIn Twitter Instagram