Can I trust Google Translate? It's the wrong question!

Pandora's box or a can of worms?  I certainly opened up a larger question than I intended to when I decided to do an informal investigation into the accuracy of Google Translate (GT).  I'll state right up front that my experiments with using GT to convert the same ten English phrases into Spanish, Portuguese and French (see past three posts) convinced me that GT does provide accurate translations at the level of communication of travel phrases.  Neither of us was perfect, but GT and I scored about equally on the number of translations we could have improved on.

So my admonition to students for using GT to do their homework based on GT's lack of accuracy was (to use an idiom that GT might have trouble translating), barking up the wrong tree.  I really wasn't concerned about GT's accuracy.  I was convinced that overuse or misuse of the service would prevent students from acquiring another language.  And I am still convinced.  It is an issue that needs to be resolved in language education as GT is adding more features (seductive features, I might add) and becoming more accessible to more students.  It and other online translation services are not going away.

The bigger question is how to incorporate GT into what we currently believe about language acquisition. Let's assume that a person has made a commitment to acquire a second language. I'm ignoring the argument that  acquiring another language is so last century because GT can translate anything you want it to. That's a topic for another blog post. Maybe I'll tackle it in sometime.

Language acquisition has to happen in the learner's brain.  In simple terms, the brain is a language-making machine, taking information heard or read in the new language, making hypotheses about how the new language works, and testing the hypotheses by creating spoken or written language.  The road to accurate language use is strewn with errors.  And that's a good thing.   Errors are positive indications that the learner is cerebrally working out generalizations and exceptions to grammar rules .

If a learner immediately turns to GT for a quick translation, the goal of getting an accurate translation may be achieved, but, and this is crucial,  nothing has happened in the learner's brain. In other words, just informing someone how to say a phrase in another language is missing the brain activity necessary for language acquisition.  This argument may seem counter intuitive, but think about it.  Have you ever asked a native speaker of a language how to say something, maybe even repeated it, and a few minutes later, tried to recall it and couldn't?  No brain activity took place.  Memorizing language is not the same activity as creating language in the head.

Did you have to do any thinking?

Now I want to daydream about techniques that accept the reality that learners are going to use GT but encourage them to use it in a manner that will lead to language acquisition.  Has anyone come up with ideas for creative uses of GT? I'm all ears!  (How would Google translate that, I wonder?)