Automated Translation: Why Not?

I'm a bit of a technophile, so I think these are very attractive gadgets:

"Translate, and two more new gadgets for your dynamic blogs"
Blogger Buzz (May, 2012)

"If you're like many Blogger users, you write in one language, but may have readers from around the world. Not all of them speak your language. Today, we've made it easier to bring content from your blog to everyone, with the release of a new Google Translate gadget...."

I'm not going to use them, though. Not any time soon.

It's not because I don't like new technology: like I said, I'm a bit of a technophile. I like gadgets.

I could use Blogger's new gadgets, since I use Blogger's hosting service for my blogs: and I'm pretty sure that a fair number of folks who see these posts speak a language other than English at home.

I put this post's title, "Automated Translation: Why Not?" into Google Translate, got "自動翻譯：為什麼不呢？" as the translation into traditional Chinese: which comes back as "Automatic translation: Why not?" That's pretty close to the original.

So, why wouldn't I jump at this opportunity to provide posts in any of the Google Translate's selection of languages?

Links, Languages, and Going Global

The possibility that someone who isn't entirely familiar with English may read these posts is one reason that I link some words to online definitions. I also love language, so those links are also an expression of that interest.

If I spoke Hmong, I would probably decide to learn another language for blogging. Of the 7,000,000,000 or so folks living today, about 2,600,000 speak Hmong 'when they're home.' ("Hmong language," Wikipedia) That's a lot more folks than live in the small town I call home, and roughly as many as live in Minnesota.

But globally? That's a fairly small fraction. Let's see: 2,600,000/7,000,000,000 is very roughly 1/2692, if I did the figures right. For someone wanting to write for as many folks as possible, in as many places as possible, Hmong probably isn't a very good choice.

Mandarin, English, and a Lemming

Mandarin is spoken by a great many folks: but most of them live in China. Again, probably not a very good choice for someone wanting to write for folks in as many different places as possible.

Besides, I don't know Mandarin. My wife and one of my daughters are learning Korean, and that's another language. And another topic.

The only language I use well enough for easy communication is English. Happily, that's one of those languages that, although not the 'cradle language' of all that many folks, is learned later in life by folks living in a remarkably large number of places. I've been over this in another blog:

• "Why isn't there More Mandarin on the Web?"
  Apathetic Lemming of the North (April 4, 2008)

"The Slippery are Very Crafty"

It would be nice, if some automated gadget could transform the ideas I express in English into whatever language folks who reading this blog prefer. But I know enough about languages, translation, and communication theory to appreciate how difficult that task is.

Today's software does a pretty good job, translating simple ideas like "I am hungry," or "this is the way out." On the other hand, when I put "this is the way out" into Google's translation service (translate.google.com), I got "這是出路," which was translated back into English as "This is the way." Not quite the same idea.

I'm impressed with today's software, and what it can do. But I also appreciate how complicated language is, even 'simple' language. Impressive as translation services are, I do not want some well-designed piece of software taking the words in these posts, and translating them into another language.

It's not because I don't like technology. Remember, I'm something of a technophile. I use Google's translation service sometimes, and it does a pretty good job. Usually.

But, particularly since I write idiomatic English, I'm not entirely confident about what software might do. Particularly when human beings, translating one language into another, make mistakes like these:

"Beijing stamps out poor English"
BBC News (October 15, 2006)

"China has launched a fresh drive to clamp down on bad English in the run-up to the 2008 Beijing Olympics.

"Previous attempts to wipe out Chinglish - the mistranslated phrases often seen on Chinese street signs and product labels - have met with little success.

"Emergency exits at Beijing airport read 'No entry on peacetime' and the Ethnic Minorities Park is named 'Racist Park'.

"Beijing city authorities will issue new translation guides by the end of the year, Xinhua news agency said.

"Running joke

"The booklets would be handed out to hotels and shopping malls, on public transport and at tourist attractions.

"Chinglish has become a running joke among many foreigners in China, and several websites have been set up listing humorous examples of mistranslation.

"A road sign on Beijing's Avenue of Eternal Peace warns of a dangerous pavement with the words: 'To Take Notice of Safe; The Slippery are Very Crafty'.

"Menus frequently list items such as 'Corrugated iron beef', 'Government abuse chicken' and 'Chop the strange fish'.

"The mistranslations arise because many Chinese words express concepts obliquely and can be interpreted in multiple ways, making translation a minefield for non-English speakers...."

I suppose one lesson to be learned from China's experience is that - if foreigners start publishing collections of hilariously inappropriate translations, there may be a reason. Other than the all-too-common 'foreigners are dense' assumptions.

That phrase, "foreigners are dense," by the way, came through Google's translation into traditional Chinese as "外國人密集," which got translated back as "Foreigners intensive."

And that is why I won't use automated translation services. Not until they get a whole lot better.

By the way, the title of that BBC article, "Beijing stamps out poor English," came out of Google's translation software as "北京郵票了英文不好," which came back into English as "Beijing Stamp English is not good."

I'm definitely not letting translation software rewrite these posts.

Somewhat-related posts:

• " 'Visible and Invisible' "
  (February 8, 2012)
• "CREDO! New Roman Missal Coming"
  (November 6, 2011)
• " 'EF,' 'OF,' Jargon, and Making Sense"
  (April 21, 2011)
• "Radio Vaticana: What? Only 38 Languages?!"
  (March 7, 2011)
• "When 'Secret' Doesn't Mean 'Secret:' The Vatican Secret Archives"
  (November 12, 2010)

caution: blind crossing