Disney’s Let It Go from Frozen, According to Google Translate

Usual Disclaimer:  I didn’t make it, just found it.

I really love this video, but for more than one reason. I like Malinda’s voice – and she also performs the song well, especially considering it’s mostly nonsense lyrics, which I would presume is much harder to do. The result of the Google Translate experiment is also pretty funny, especially considering it manages to actually reverse the meaning of the song’s chorus. And, the new lyrics tend to stick in one’s head.

But there is also a much more serious side to the Google Translate series of which most people are probably not aware.  I have an academic background in anthropological linguistics, having studied it in both under-grad and grad school, as well as reading various books on linguistics from several disciplines and even watching video courses on the subject.  I wouldn’t say I’m an expert, not by a long shot, but I know more about the study of language than your average person.  This series of music videos illustrates, precisely and in an easy-to-grasp and humorous way, the problem with machine translation.  Machine (that is by computer) translation is often very inaccurate, especially with figurative language or language that contains a lot of slang and argot.  Machine language translates, literally, by single words.  Sometimes it might use a phrase, but it’s mostly by words – and it should be obvious that words can often have different meanings.  In conversation, native speakers will automatically pick-up on the meaning of a word by context and pronunciation/intonation.  A native speaker knows the difference between the sentences, “Attention, class, read chapter two for tomorrow,” and “Yes, I have read Oliver Twist.”  However, someone who is learning English might be confused by those two sentences, especially initially.  And in machine translation, the machine can’t tell the difference, and might choose the wrong tense when translating into another language.  This is compounded when several layers of translation are used – because each layer builds on assumptions of the previous layer.

However, machine, computerized, or automatic translation aren’t the only cases where translations are messed-up.  People also make mistakes in translation. Sometimes because they are working with written records that are very difficult to read. Sometimes because the meanings of words have changed. Sometimes because political views make one translation more desirable than another one. Sometimes out of laziness. And sometimes simply by mistake – no one is perfect, not now, and not in the past.

And, as anyone who has ever studied a foreign language knows – it can be literally impossible to translate certain words from language A to language B if language B doesn’t have that word.  A good translator will actually both use the word from, say, Japanese, or Russian, then describe it’s meaning in a series of words in English.  All languages world-wide are unique, and almost always contain words that are unique to that language.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s