Monthly Archives: March 2012

How to Safely Cut NHS Translation and Interpretation Costs

By | Interpretation, Translations | No Comments

We’ve talked before about how blindly cutting back on translation and interpretation costs can yield false economies at best and obstructions of justice at worst.  Well as the United Kingdom continues to go through austerity measures in all public services, the NHS has come under fire for spending on translation and language services.

A health policy think tank, 2020health, has revealed that the NHS spends over £59 000 a day on language services.  This number can be staggering at first, but when one considers that the NHS has a legislated responsibility to provide proper care to non-English speakers.

2020health has also provided some recommendations of ways the NHS can cut costs related to language services.  While the suggestion that the NHS use free only translation software; i.e. Google Translate, is incredibly dangerous when it comes to matters of health, others are not as absurd.  The use of simpler language is a great one, as specialized translators and interpreters are more expensive.  In fact simpler language should really be used for all patients.

This video from provides some great recommendations for how to decrease expenses without leaving non-English speakers in the cold.  These suggestions are useful for any company looking to cut costs without falling prey to poor translation related gaffes.



Computer Aided Translation is not Google Translate

By | Translations | No Comments

At a recent non-translation related trade show I was occasionally asked the shocking question, “So, how often do you use Google Translate?”  The answer is of course, never, but it gets to a matter we’ve discussed a lot on the blog, the lack of understanding of how professional translation is actually done.

At AZ World we use Computer Aided Translation (CAT) software, now when I say “translation software” people immediately think Google Translate or machine translation.  That is simply wrong.  CAT software can have tools like glossaries, spell checkers, or translation memories.  They in no way translate for the translator, but just provide additional tools, like any modern word processor does.  The translation memory is the most powerful of the CAT tools, and the one that causes the most confusion.  It’s best explained with an example.

Say your company, Initech, releases weekly Human Resources bulletins that must be translated into Spanish so that employees in both New York and Madrid can read them.  Assuming Initech, uses a standard format for each of the bulletins, there are probably words repeated each week.  These can be things like company slogans or mottos.  For example, “Is This Good for the COMPANY?”  Instead of paying full rate for translating “Is This Good for the COMPANY?” every week, Initech would only have to pay full rate once, then that phrase would enter into Initech’s TM and every subsequent translation would charge that phrase at a significantly lower rate.

This system rewards customer loyalty by providing our clients with a lower rate the more they use AZ World, but it has another benefit for clients as well.  The TM is the property of the client and will be provided to them if they request it.  The client then still has a database of translated words they can take to a new translator should they choose to leave AZ World.

The way the TM actually works depends on the specifics of the translation software used, but it can be useful to think of it as a sort of find and replace function.  And this is where people become confused about translation memories and CAT and machine translation (Google Translate).

This is where it is important to point out the difference between using translation memories, TM’s, and using machine translation, MT.  MT uses a general database of words and simply replaces all entered words with their translated equivalent.  While MT software is getting better and better at detecting context and finding equivalent verb tenses, it is still nowhere near as good as professional translation.  While MT uses a general database, TM’s are client specific databases that contain the words that best fit that client’s needs and environment.  The TM also isn’t used for every single word in a project.  Large bodies of text cannot be properly translated with a TM, but technical and specialized words within a body of text can be.

Nowadays everyone from teachers to cabinet makers utilize specialized computer software to improve their work, translators are no difference.  These tools improve the quality and quantity of work, and in no way will replace the irreplaceable human element

Malaysian Ministry of Defense Uses Google Translate, with Expected Consequences

By | Translations | No Comments

A looser civet

It appears that the Malaysian Ministry of Defense used Google Translate to translate their website into English.  The internet was tipped off when the inevitably humorous translation began to pop up on Facebook and Twitter.  Since then the Ministry has removed the translations, there is no longer an English version available and admitted to using Google Translate.  The Ministry, which handles a $3.5 billion yearly budget, decided to go for the free translation service, and got exactly what it paid for.  They have since said that the translations in future will be done manually.

Here are some great examples of the inevitable shortcomings of machine translation:

“After the withdrawal of British army, the Malaysian Government take drastic measures to increase the level of any national security threat”

Women should avoid: “clothes that poke eye”

Men are permitted to wear a “tight Malay civet”

Will Mandarin become the new “Lingua Franca?”

By | Interpretation, Translations | No Comments

Last week we talked about what a “Lingua Franca” is and about the history of the term.  Today we will discuss if there is much truth to notion that Chinese, specifically Mandarin, will replace English as the lingua franca for international business.

My general answer to this question is a simple, “no.”  It’s a no for two reasons.  First, Mandarin is a tonal language.  This means that words have different meanings depending on the tone one uses.  This isn’t like in English where one’s tone can indicate whether the word is to be taken seriously, sarcastically, or as a question; in Mandarin different tones indicate drastically different meanings.  This page has some great examples.

While speakers of other tonal languages have little trouble dealing with this when they learn Mandarin, speakers of non-tonal languages do.  This is a serious obstacle for speakers of non-tonal languages, and increases the metaphorical cost of becoming fluent in Mandarin.

Another aspect of the language that will impede its widespread adoption is the script.  Chinese calligraphy takes years to learn, and learning how to read can take even longer.  This presents another roadblock for Mandarin as it makes it even less likely for Mandarin to be adopted if the cost of writing up business documents is so great.

The second reason why Mandarin will not become the new lingua franca I see every time I get on an airplane.  “Five times more people are learning English in China, than there are people in England,” is what I see every time I board a plane.  This factoid comes on HSBC posters found on the passenger walkways around YVR.  It indicates the most interesting reason why Mandarin won’t replace English; Mandarin speakers are learning English.  This means that the Chinese are willing to conform to the current lingua franca.  So long as this is the case, English will remain the lingua franca of international business.