Top Ten Translation Myths

In today’s digital society, information, both true and false, spreads like wildfire, so it can be difficult to get correct, factual content. The same rings true for the localization and translation industry, which is largely misunderstood by anyone not involved with it. So, we thought it would be helpful to debunk the top ten translation myths we’ve heard in the course of doing business and from family and friends.

Anyone who knows more than one language can translate; translators only need to be native speakers.

It depends on the quality level. Sure, if you need a quick translation for personal use or need someone to glance over a document to validate the language used, you certainly don't need to hire an LSP. Most international companies have native speakers who can help with non-critical tasks.

But for customer-facing content or documentation that can inform critical business decisions, you will need to engage with professional translators who are experienced native speakers, bilingual and preferably living in the country where the target language is spoken. They have a linguistics education or a degree in translation and need to pass rigorous tests to be accepted to work for an LSP.

Translation and interpretation are the same thing.

Interpreters are essentially verbal translators and must choose their words with split-second reactions. Skilled in their craft, they are hired for specific events or engagements, and as a result, spend time examining meeting briefings and extending their knowledge of the terminology they’ll need prior to stepping into the live situation.

Translators, as you know, solely work with the written word. Of course, they also have deadlines and use tools and skills to maintain consistency and terminology compliance, but they have more time to deliberate word selection and research topics.

As you can see, their industries require different qualifications and skill sets.

Translation is just looking up the right words in a dictionary.

Who on earth came up with that idea? Even early machine translation was way more sophisticated than that. Let's just look at the basic elements of a language: words and rules.

If translation was just about replacing one word for another, there would be no localization services industry in the first place. In addition to translating words and language syntax, there are the additional aspects of emotion and intent. If you need more convincing, watch the famous movie "Lost in Translation."

Translation is a dying art. Everyone will be replaced by machine translation anyway!

Not so fast...although the author is a big believer in AI and a future in which tasks like translation will be handled completely by machines, it is still too early to even make predictions by when this may become reality.

Content contains a varying amount of emotions and semantics based on what its intent is. A product user guide or knowledge base article should be very technical and free of emotion, which means very little linguistic creativity is required to write it. (A mature, well-trained machine can handle this.) A marketing brochure or a product campaign's webpage will require more creative minds, and if successful, will evoke emotions and the desire to buy the product. (Humans will likely always handle this.)

There is still a long way to go before machines can understand emotions, let alone translate them from one language and culture into another. Only when we have created singularity will the translator's profession become obsolete—but when that happens, the future of translators may not be our biggest concern.

There's just one correct way to translate a sentence. Two experienced translators will always produce the same translation.

Wouldn't that be nice? A simple example will refute that idea: "I saw a man on a hill with a telescope." In English, this simple sentence can have five different meanings—and thus, five different translations.**

When recycling from a translation memory, it is crucial to look at the context of the segments as well. Imagine having to correctly translate the sentence above without knowing the story surrounding it.

Translators may still want to translate the same sentence in different ways, such as with different grammatical structures and synonyms, but that’s where linguistic assets can help—style guides and terminology glossaries, to be exact, to ensure linguists conform to client requirements. Another way to reduce the potential for different translations with the same meaning and improve recycling is to use simplified source language and provide standardized and modular content.

**Wonder what those 5 meanings are?

• There’s a man on a hill, and I’m watching him with my telescope.
• There’s a man on a hill, who I’m seeing, and he has a telescope.
• There’s a man, and he’s on a hill that also has a telescope on it.
• I’m on a hill, and I saw a man using a telescope.
• I'm on a hill, and I’m seeing a man with a telescope.

High-quality translation can be achieved by assigning experienced translators and telling them 'I want the best quality'.

"I want the best quality" is a statement that needs a reference in order to make sense. In translation, there is no "best" quality. Linguistic quality is difficult, if not impossible, to measure and define by itself. You always need to consider factors such as target audience, topic, content type and purpose.

Of course, linguistic quality is important when translating a help file, but a technical manual or a legal document requires subject matter expertise, and for marketing content, a creative copywriter will be the best fit.

Any translator can perform linguistic quality assurance; being a native speaker qualifies to review a translation.

To be fair, translators often review each other’s translations. But certain reviews require special knowledge, background information or just an expert familiar with the product. Therefore, some translation work includes an additional client review step—by someone who really knows the product or service and can verify accuracy. Marketing content, for example, benefits from a client reviewer who not only understands branding style and tone, but also has a deep understanding of the product or service and target market.

MT will replace TM recycling.

Yes, it will. The boundaries between traditional recycling from translation memories and machine translation (MT) are slowly vanishing. Many translation management systems (TMSs) already allow for a combination of TM recycling and MT. Eventually, the source of a pre-translated string will not matter, because MT engines and TMs will be combined in the training step: recycling quality levels will become completely transparent, meaning MT engines and TMs will merge into a single source.

What will remain important, though, is the subject of the source content so that the correct domain can be selected for the pre-translation process.

Every translator can do MT post-editing (even without prior training).

From a linguistic perspective, yes. Translators who meet our linguistic quality standards for translation will be capable of performing post-editing. But there's another aspect to this service: the productivity/quality paradigm. A good post-editor needs to be able to balance the time needed to review and update the machine-translated text against the required level of quality. They also have to make quick decisions on whether the translation is editable or to discard it and start from scratch, which may be more efficient.

The soft skill a good post-editor should have is flexible thinking, resulting in adjusting the work style to the project's objective. Also, ideally a post-editor understands the type of errors a particular engine can produce, which will help them look for and fix them.

Software localization is just translating strings.

If you don't care about the product’s functionality, user interface or user experience, then yes.

Translating applications and operating systems is challenging and complex. The risk of breaking a working application by translating the wrong strings still poses a significant risk, even with modern software development tools. Factors adding to the complexity are different operating systems and display settings on devices like phones, tablets and appliances. Now more than ever, it’s imperative to build into software localization projects sufficient time for functional, localization, linguistic and user experience testing to ensure your products are device-ready and flawless.

Many thanks to Solution Architect Jan Grodecki for this post! Have you heard other translation myths? Share them here!