The bad translation tag features articles with examples of bad translations and mistranslations taken from around the world — often with amusing consequences.


Eating carpet prohibited, by Justin Ross Lee

The hazards of machine translation

Eating carpet prohibited, by Justin Ross Lee, the hazards of machine translation
Eating carpet strictly prohibited – a sign of the hazards of machine translation / Image credit: Justin Ross Lee via BBC

The Hazards of Machine Translation and Beyond

The Jackie Chan bus stop, a restaurant called Translate Server Error, wife cake, children sandwiches, wide-boiled aircraft – they sound like comical lines at a stand-up show. But, in fact, they’re real-world examples of glaring mistakes and the hazards of machine translation.

For big firms, poorly translated text can have major consequences such as the risk of offending potential customers or losing business. Although we often hear promises of machine learning tech that will process language translation effortlessly and naturally, just as a real [human] translator would.

But when will such technology be available to businesses?

Skype and its real-time translation upset

Last January, Skype launched its real-time translation tool which instantaneous voice-to-voice translation in seven languages. However, it suffered heavy blows when users complained that it turned ordinary Mandarin words into obscenities. The glitch came to light during the shooting of a Skype commercial in China; apparently even the simple phrase ‘it’s nice to talk to you’ translated into offensive swear words.

Google Translate, the traditional approach

Translation tools like Google Translate have traditionally been built around phrase-based statistical machine translation. Machine translation works by analyzing a back catalogue of texts already translated, such as academic papers and glossaries.

The texts are analyzed in parallel – both original and target languages. Using statistical probabilities, it selects the most appropriate translation to the phrase submitted; the better the quality of the original language, the greater its effectiveness. But it’s prone to howlers, like the ones mentioned earlier. Often the translations sound mechanical and dull.

The End of Machine Translation?

Alan Parker, director of engineering language technology at Facebook, recently commented on statistical machine translation reaching “the end of its natural life”. It has been said that translation technology is on path towards artificial neural networks much like the neural pathways of the human brain.

These neural networks are structured similarly to the brain, using complex algorithms to select phrases appropriate to the translation. Astonishingly the sophisticated network can learn metaphors, idioms and the subtle meaning behind language. This will effectively transform language translation today – rather than direct literal translation, the neural network can translate the same meaning to a different culture avoiding any possible offense.

While Facebook and Google have reported that they will switch over to neural network translation this year, they have not publically announced specific dates.

Auto-translation, not perfect yet

Despite the two tech giants rolling out plans to use neural networks, there are still major hurdles to cross before we’re quite there yet. According to Professor Philipp Koehn: “there are very hard problems with semantics and knowledge representation that have to be solved first, and that we are not close to solving.” Professor Koehn hints at less explicit information in the source language such a gendered nouns and verbs in languages such as Portuguese, Italian and German. Prof. Koehn is a computer scientist and expert in translation technology at the University of Edinburgh.

‘Chinese doesn’t use plurals, verb tenses or pronouns as we do in English, which makes exact translation very difficult’, Prof. Koehn added.

The Hazards of Machine Translation Tech and the Future

Albeit, translation technology has come a long way and provides decent literal translations; there is the need for a tech that speaks the real language of the end user.

Machine translation technology is a handy tool, but don’t rely on it entirely.

The STAR Team

Source: BBC Business News

UFC Logo

Conor McGregor UFC Win, Lost in Translation

UFC Logo, Conor McGregor UFC Win, Lost in Translation

Conor McGregor UFC Win

Immediately after being knocked out within 13 seconds of the UFC 194’s main event, José Aldo’s comments post-fight were met with booing from the packed MGM Grand arena in Las Vegas.

Being a Brazilian national, an interpreter was brought in to translate Aldo’s remarks on the fight. The former champion was quoted

We need a rematch, it was not a fight.

The interviewer at the fight — a stand-up comedian — Joe Rogan took to Twitter to apologise for the confusion and the mistranslation of Aldo’s words. Many people did wonder however, how could Aldo’s long speech post-match [in Portuguese] be easily translated into only nine words!

The Real Translation

A Portuguese-speaking Reddit user, RandyLiddell translated what he heard José Aldo say…

Rogan: How much, if anything, of the fight can you remember?
Aldo: He threw a jab on my chest, I was already expecting that. When I went to attack him he hit me with a good cross and there is where he got me. I believe that after this fight we have to go for a rematch, is not done yet. He got me with a good shot and was able to finish the fight.

Rogan: How much did all the taunting affected you?
Aldo: It didn’t affect me in anything. Whatever he said it doesn’t matter, I don’t fall for provocations. My mind is always calm inside there, I try to just get in there and do my job. He was happy today, caught me with a good blow. I think we have to move on now and now I am waiting for a rematch, and God willing, next time I will be back much better trained and recover what is mine.

Perhaps the translator can be excused as it was a loud arena with little time to interpret everything Aldo had said.

Did you watch the fight?

The STAR Team


Graham O'Mahony, Blogger and Web Designer
Web Designer and Blogger
The STAR Team
Follow the conversation on Twitter logo @STARTranslation

Context is important in translation: row or row?

Why context matters in translation: row or row?

Why context is important in translation / STAR Translation Imaging

Is it row or row?

Row or row are known as homonyms (i.e. words with the same spelling or pronunciation). While they share the same spelling, they have different pronunciations.

We’re often asked why we cannot translate a single word or a collection of short phrases on the spot for someone. Professional translators will tell you that sometimes the smallest phrase can be the most difficult to translate. Sounds silly, but look at the illustration above …

In this example, a single word has two completely different meanings – often referred to as homonyms.

Is row a verb, referring to rowing a boat? Or a noun / verb, meaning to argue or an argument?

The only way you’d know is if you …

  • heard the word spoken
  • knew the context of the conversation
  • knew where the text came from and what it’ll be used after it’s translated

For those interested, this is the number one reason why free translation or automatic translation on the internet delivers such bad results.

Another example of this type of mistranslation is the word armed. Does this refer to an alarm on a phone, a watch being set or a person with weapons?

Try entering a few sample sentences into Google Translate to see what you get. Pick a target language that you have some knowledge of first to better understand how context helps.

The STAR Team

Funny Machine Translation Errors

Funny Machine Translation Errors

One of the things we find amusing is how machine translation can make some really silly mistakes.

A common misunderstanding is that translation is just about words however context is very important in understanding exactly what the words mean.

Machine translation engines provide free translation, but the quality is often very bad. They translate words and phrases but can never perform like a human in understanding the text.  Here are some fun examples we’ve found in our own research.

“A School of Fish was spotted in the sea”

A school of fish was spotted in the sea

The translation above literally means that a school (schools where kids go) full of fish was in the sea. In French the word école means a physical school. They don’t have the term “school of fish” in French. We also translated – Sharks swim in schools and got the same result. Sharks swim in colleges!

The translation is thus taken out of context since the right phrase to translate “a school of fish” would have been “un banc de poisson”. Google and all the other machine translation engines we tested made the same mistake.

French Machine Translation Errors

Poor French translation example from Google Translate

A interesting French idiom is Se faire la belle”, which means to run away and not to make beautiful!

Another similar idiom would be to to take french leave, which describes when a guest leaves a party without informing their hosts.

Google translates this as Pour prendre un congé françaises” which means to take French holidays and we won’t say anything else about the grammar mistake. However, the correct translation is Filer à l’anglaise.

An example of a bad translation from a status update on Facebook by the Bing Translator

Amanda at Spiderworking spotted this one for us using Facebook’s translation with Bing. The Japanese user was discussing Scotland but we’ve no idea what the translation is meant to say correctly.

Do you have any funny examples of mistranslation? We would love to see them. Share them with us on Facebook or Twitter.

The STAR Team

Lost in translation: Welsh sign translation error

Welsh sign translation error, Swansea council

In November, Swansea council made the ultimate translation faux pas with a Welsh sign.

The original sign read, “No Entry for heavy goods vehicles. Residential site only“. The sign was to be erected close to a supermarket near a residential area.

The text was sent for translation but unfortunately, the translator was not in the office. Instead his email sent an automated response.

The automated email stated:

Nid wyf un y swyddfa ar hyn o bryd. Anfonwch unrhyw wiath i’w gyfieithu.“, which in English translates to “I am not in the office at the moment. Please send any work to be translated.

As the original message contained two sentences, the receiver assumed this was the correct translation. Thus forwarding the translation to the sign makers to be printed. The sign was printed and duly erected.

Swansea council stated that the sign would be corrected as soon as possible – a new sign now stands in its place.

The moral of the story is always proofread before you print.

“People rely on email all too often for basic communication on projects. When a translation is only two lines, people tend to treat it with indifference”, Damian Scattergood commented. “We have found that some people translate two lines like the safety warning stickers you see on kitchen appliances, signs and posters. The use of free translation tools and expecting the results to be perfect do not mix.”

“For us, every translation whether it’s two or two million words should to be handled with the same due care and attention. A simple phone call in this case would have saved money and embarrassment for all”, added Damian.

Damian Scattergood is the Managing Director or STAR Translation.

The STAR Team