Oops! Mistranslated French

No Word-for-Word

The sad part in human translation services is that exact translations are not always the BEST translations; sometime trans-creation is needed to achieve best quality translations. Mostly it is proverbs and sayings, which are so colorful and fanciful hence result in literal translations that would startle and cause apprehension. It is always fun to learn some French proverbs, it is an essential ingredient to become fluent in the language. Using some appropriate proverbs into your speech help you in acculturation and appear more local.

Oops image

Fun with Miss Translation

It is best to speak the new language you are learning; and even better if you practice some irritating and pretentious proverbs. It is very amusing to give very literal translations.

The new trick:

“on n’apprend pas aux vieux singes à faire des grimaces”

Literal meaning: ‘you can’t teach old monkeys to make faces!’

Fair English translation: ‘you can’t teach an old dog new tricks’

but it’s much more fun to use the literal version.

The one about the youth:

“si jeunesse savait, si vieillesse pouvait”

Literal meaning: ‘Youth is wasted on the young,’

Fair English Translation: ‘If youth only knew, if old age only could.’

It captures a sadness that comes with experience that I find very affecting, and lacks the English version’s cynicism.

“un malheur ne vient jamais seul”

Literal meaning: ‘misfortune never arrives alone’ 

Fair English Translation: ‘When it rains, it pours’ but the literal translation

Of course, translation professionals have an odd sense of humor as well as sense of language. Which is why the profession is joined with so much solitude!


                                Will Machine Translation Ever Beat Human Translation?


It is so temptation when we think of a world where communicating with others in different languages is effortlessly easy, all hail to machine translation. Will the rise in machine translation mean that people won’t have to learn languages anymore?

It is evident that evolution of machine translation is spreading fast, but the fact of the matter is that we haven’t yet reached the ideal vision. Will it ever happen?

Machine translation dates back since around 1950s, where researchers began the simple use of software to translate text or speech into different languages. The first commercial machine translation system appeared in 1991, with the first web applications following along a few years later.



In the tech-age today, machine translation technology varies from extensively used, free online translation services such as Google Translate, to cheap, on-the-go mobile phone apps such as Apple’s iTranslate, to rather more expensive, customizable, professional software packages.

The pros of machine translation are clear:

It’s cheap and sometimes even free depending on the kind of service you use.

It’s swift for on-the-spot translation needs, or time-critical web content.

It’s innovating almost every other day as researchers strive to make machine translation technology better.

But we can’t ignore the drawbacks which are as clear as the benefits:

It lacks localization as it produces straightforward translation. There is little or no localization in reproducing the content into a suitable cultural context.

Tone in the content is nearly nonexistent as machine translation focuses to translate more restrained aspects of language such as humor and metaphor.

Machine translation often reads very awkwardly breaking the momentum of content, which results in poorly structured sentences which are difficult to read.

If you’ve invested a good time to create content; then you definitely need human creative translation to do justice to the original content and make certain that all client objectives are met.



Probably the most prominent machine translation failure incident is about the Chinese café which wanted to provide a sign in both English and Chinese. The Chinese sign (in Chinese characters) said ‘Dining Hall’, while the shiny new English sign said ‘Translate Server Error.’

The incident attracted global publicity but it did not prove harmful for the café. Although in the serious world of business machine translation such errors can cause some significant complications. Sometimes the organization can end up facing legal action.

Therefore we conclude, that human translation is here to stay. There is unquestionably a place for machine translation, especially in those circumstances where urgency is more important than precision, but in a professional perspective nothing will ever beat the impact of creative, precise, localized, human translation.



The San Zi Jing (Three Character Classic), written in the 13th century and attributed to Wang Yinglin (1223-1296), is not one of the traditional six Confucian classics, but rather is a distillation of the essentials of Confucian thought expressed in a way suitable for teaching young children. Until the latter part of this century, it served as a child’s first bit of formal education at home. It is written in couplets of three characters (syllables) for easy memorization. One might call it a Confucian catechism.

George Yeo, then Singapore’s Minister for Trade and Industry, in an address at the Chinese University of Hong Kong, described the work thus: “For centuries, Chinese children, before they could read or write, were taught to recite the San Zi Jing through which the Confucianist idea of society being one big happy family is programmed into young minds. The three-character phrases are like strands of cultural DNA which are passed on from generation to generation.” (Cf. Sanzijing at raptor.depauw.edu)


Notes & illustration:

2004-07-20: Note from Name/pseudo (Commenter on Sanzijing 0.)
The San Zi Jing as appeared in the link you provided
is actually a copy from

The person who put up the trimetric.htm page copied the text from Giles book by hand back in the late 1960s. He made several mistakes in his transcription, which he could not decipher when he put his Web page a few years ago. The “side-notes” of the undecipherable text showed up as “distinct markers” whenever someone else reprints
the material from his site.
So far, all the reprints of the Giles version on the Web has been traced back to the same source.

The Three-Characters Classic --- San Zi Jing

Do we need to localize keyboard shortcuts?

When I’m talking regarding keyboard shortcuts to things like Ctrl-S or maybe one key shortcuts like ‘A’ (used to Archive an email on Gmail). On the opposite aspect the method, called hotkeys or the underlined letters in menus or different GUI controls are clearly one thing that ought to be localized and that I not going to argue regarding this.

I have did some analysis to check however, some different firms are handling the shortcuts once people are using different languages and different keyboard layouts. Here are a brief list of my results:

  • If doable, the shortcut isn’t modified
  • If a key not available on a particular layout we must always attempt to use constant physical position
  • There needs to be no link between practical application language and keyboard patterns. It’s a mistake to consider that application language might match the keyboard format. I’ll reveal just a good example: in Romania over 95% of the computer systems are using US keyboard layout and the other 5% are using among the 4-5 various Romanian keyboard layouts.
  • Some individuals are using several keyboard layouts – generally, they are in fact power users and we can’t ignore them
  • Keyboard patterns can be converted anytime
  • Characters are secure to be used only when they are Latin

Physical position rules

I think that the actual physical position of the key is a lot more important than the letter written on it. Below are my arguments:

  • Human brain is understanding the shortcuts by hand movements (notice Pavlov’s dog experiment)
  • It is appropriate for the usage of multiple keyboard models

However, it’s not always so easy: we can’t assume that the French users would switch the usage of Ctrl-Q with Ctrl-A (Quit vs Select All) simply because they are using different layouts. Therefore, a fundamental rule would be to stick with the format for letters and numbers. However, attempt to use the actual physical position for other keys.

Anyways high, likely those individuals using multiple layouts would select alike layouts (derived from the similar root).

One-to-one key matching

Since the number of keys is similar we need to find a way of remapping the keys so that all shortcuts working on US to be workable on different layouts. When I’m talking about keys, I’m referring to the hardware keys.

Do we need to localize keyboard shortcuts?

Question: How do I create OS X tell me that the key pressed by the user was ‘q’ while current format is Arabic?


  • We are not allowed to consider it’s a ‘q’ because the key code is similar, probably it’s a Arabic DVORAK keyboard!
  • The shortcuts should not be hardcoded (user can edit them)

On Windows it does work 98% using virtual key codes and you can use them to obtain this info, but on OS X the key codes (actually Apple call them sometimes virtual key codes) they are some kind of scan codes. For the rest of 2% you need to do some small hacks with the OEMs.

Do we need to localize keyboard shortcuts?