Translation doesn’t teach language

Recently, I wrote about what makes learning a language hard and pointed out that using data on word difficulty from the language learning application Duolingo is not particularly credible. Duolingo relies on gamification and translation to “teach” learners decontextualized words and sometimes useless phrases such as, and I am not making this up, “the butterfly drinks the milk.” Outside of some Alice-in-Wonderland-esque scenario, I’m hard-pressed to think of a circumstance under which one would ever be asked to translate the English phrase “the butterfly drinks the milk” into Portuguese.

Setting aside the silly instructional content, though, many people have asked me why translation is such an ineffective approach. Often, when people think about language learning, they think of translation, assuming that the best way to start learning a new language is to work from your first language and then figure out how to say basic words and phrases in the language you are learning. That might sound intuitive, but in practice, it’s a terrible way to learn. Here’s why:

First, it simply takes too much time. When you translate, you are spending your time in both languages, trying to go from one to the other and not immersing yourself in the language you are learning. If you spend your time trying to figure out how to go from English to Spanish, then you are losing half of your language practice time to your mother tongue. Learning a language is learning a skill, and learning a skill requires practice–way more practice than most people realize, actually–so cutting that time in half by constantly thinking in your first language is inefficient.

Second, and perhaps more important, languages don’t work that way. Think about the first Spanish class you took, when you learned to introduce yourself by saying ‘Yo me llamo Katie.” Translated literally into English, that statement is actually “I myself I call Katie,” NOT “My name is Katie.” The first lesson you learn in Spanish 101 is that you cannot actually translate sentences word-for-word.

Finally, translation is, itself, an art. People go to school for years to be professional translators. In order to be successful, they need to not only be bilingual, but also bicultural. Without thinking, native speakers use metaphor, word order, tone, and slang to convey meaning. Someone translating, for example, a movie scene must not only understand that when a new boss says “let’s hit this baby out of the park” at the beginning of a brainstorming session, she’s not talking about physically smacking a human baby out of a playground (or baseball stadium!) but instead motivating her team to hit a metaphorical homerun. And then the translator must also, after understanding what the new boss is saying, find a way to express the same sentiment in another language, which is almost certainly going to involve a different metaphor.

When you think about it, it seems almost miraculous that we can take conversations, poetry, and speeches in one language and express them in another without more things getting lost in, well, translation. Translation is an impressive skill for well-trained, fluent speakers, but not an effective approach for learning a language.

This post was originally published on the Voxy blog in January 2018. 

17 thoughts on “Translation doesn’t teach language”

  1. Dear Katie,
    Let’s see. Your criticisms are quoted below:
    1. “These particular translations are silly.” Bravo! Let’s have more silliness and play inside language. Silliness breaks boundaries and shakes people out of preconceived notions of how language should work. Moreover, why color all translations with this particular criticism? Are you so upset with Duo lingo that you want to indict everyone using that method?
    2. “It takes time.” Well I have yet to see a chart of what kinds of exercises are the most fruitful (over time). If you know how to best spend a student’s time could you share that with me? Language learning is a holistic exercise. Lots of things are important. I would argue that a little dabbling in translation can be quite fruitful. Any activity done to the extreme is not helpful. Over 30 years in many different countries I have been surrounded by people that have been immersed in the language and culture of others. I am forever astounded by how many people can’t/don’t pick up the language of the larger culture. If learning a language was simply a function of using language for meaningful communicative tasks then this problem should not exist. Translation focuses the mind. This focusing is a positive thing which forces people to engage in a much different way than when they are casting about for words to communicate in their broken “whatever.” I would argue that for many, the focused effort of translation gives learners a new perspective on language learning.
    3. “Language learning doesn’t work this way (by translating).” But, how do you know this until you start translating? Translation is a helpful tool in teaching us exactly this. It helps us see barriers and similarities between languages. At its best it can be consciousness raising!
    4. “It is an Art.” Wow! Thank you. If translation is an art, then we need more it. Can you imagine? A student starts to learn language for a practical reason and ends up learning an art form. Ah, the subversiveness of it all.

    I can see you are not happy with Duo lingo but you don’t need to broad brush those of us that teach translation in balanced and meaningful ways. The problem you have is with Duo lingo not translation.

    • Some criticism of Michael’s criticisms:

      1. You seem to be implying that we should use translation to show how we shouldn’t be translating since it can lead to nonsensical outcomes. If that’s a valid interpretation of your criticism, then I totally agree.
      2. Katie’s comment that it takes too much time is spot on. I’d argue that most students have little chance to use English outside of class, so time is better spent creating opportunities for production. You could probably make the argument that translation could be done at home and checked in class. But if I told any of my paying students to use valuable (and expensive) class time to translate, I wouldn’t have any more students.
      Rather ironically, the only time I’ll use translations in class is to save time. I’d rather not waste 10 minutes of valuable time trying to teach the concept of ‘sarcasm’ to a class that struggles with basic communication. A simple translation suffices and we can move on to time better spent.
      3. The larger problem for me is that students who rely on translating never learn to think in English. It’s always L1 to L2, and it’s so easy to see students who are doing this. Their conversations are invariably unnatural and stilted.

      I also think you happen to be right. How much translating is okay? ‘Dabbling’ is probably the best word for it.

      • Dearest Mean Teacher,
        (Nice name!)

        1. I don’t want to drawn into a defense of nonsensical sentences except to say this. All my translations are drawn from speech that is excerpted from paragraphs. I think that is a wise guideline. I would also note that to some, humor is nonsensical.
        2. I agree in part. Personally, I assign it as homework and correct it in class. I spend maybe 5-10% of class time going over the translations (depending on errors). I don’t do it in every class but in those situations where I sense students are experiencing problems with the huts and bolts of the language. I find it works best with beginners.
        3. I only work from L2 to L1. I think this form of translation is the most productive, quickest, and the easiest. I never “do” L1 to L2. I find it tends to create the results you describe. But, I try to stay away from the word never in language learning.
        4. Expect excellence. Require excellence. And do it in small amounts. I agree.

  2. Michael Butler points out that translation is an important skill, much as Ms. Nielsen does. Unfortunately, Mr. Butler seems to be defending translation in general when Ms. Nielsen is only pointing out the shortcomings of translation as a language learning technique. Takes me back to the early days of CLL (that’s Communicative Language Learning) in the 1970’s when it was overcoming the common practice of teaching language with grammar and translation. Yawn.

  3. Thanks, Kevin Ryan. Yawn is right! I was certainly not meaning to re-start the old Grammar Translation versus Communicative Language Teaching debate. I wrote the article to explain the drawbacks of translation as a pedagogic approach for the many laypeople who assume that translation is a reasonable way to teach someone a new language–which, generally speaking, it isn’t.

    Michael Butler–it seems you may have missed my point. Translation is an art, and a valuable art that serves society well. My point is just that it is not the right way to learn a language. If you’d like to have an actual conversation about how long it takes to learn a language, what approaches have been empirically established to work well, and/or the individual difference factors that interact with learning environment to hasten or impede ultimate attainment, feel free to drop me a note. I just ask that you leave any preconceived notions behind.

  4. Mr. Ryan, I don’t know if my argument rises to the level of a general defense of translation. I certainly did not intend to rise that far. Let’s just say that I refuse to accept a blanket repudiation of translation. I believe there are specific instances where translation can be used to good effect.

  5. As a committed communicative EFL teacher with Celta under my belt and Delta well under way, I was surprised to discover I was able to learn two languages (Esperanto and Welsh) via Duolingo, at least to a point where I had enough of a basis to start practicing (CEFR A2 maybe). One thing I would say is not all the courses there are equally silly in terms of content. In particular the Welsh course seems grounded in daily life, perhaps because the course writers are used to teaching it as a second rather than foreign language.

  6. Hi Katie, in all honesty I have never met anyone who could leave preconceived ideas behind. When people say this they usually mean, “ideas that are not like mine.”

    So to be clear, I acknowledge that I come to this forum with preconceived notions and I am glad for it. And to be specific, one preconceived notion is that translation, in small doses at the right times can be useful, effective and worthwhile.

    In the context of your business or of Duo lingo perhaps translation does not make sense. I get that. But, because it does not make sense in your context does it therefore also mean that translation is ineffective for everyone, everywhere, at all times. I think not. Will you grant me this?

    Teaching is like archery. It is nice to have arrows capable of doing many things in your quiver. When I was young and starting out I thought everyone in authority was an “authority.” Now I realize that theories of language learning ebb and flow. I would encourage anyone interested in translation to try it out and see for themselves what works and what doesn’t. And if anyone needs a reason to try it out, I will say in my small voice that for me translation works.

    Using translation at times has made it possible for me to DISCERN problems my students have. Translation helps me better understand how students SEE the target language. Translation helps me see what students are MISSING. Translation helps me see how students are or are not accessing hidden MEANINGS. I love translation when it helps me DISCERN problems, SEE what students are MISSING and get insight into how students understand the MEANING of what they are reading or listening to.

    You will notice that I conceive of all this in terms of how translation can help me become a better teacher. To be even more clear let me say that by becoming a better teacher I feel it directly helps students become better learners. Large organizations with fixed curricula on tight schedules may miss this link. I think it is important.

    Expressed as a statement it goes something like this: I believe by getting better insight into how an individual student sees the target language it makes me a better teacher – and as a better teacher I am better prepared to make each student a better learner. That makes translation useful, effective and worthwhile in my book.

    There, it is said. I hope now I have said my peace. And for all of you YAWNING out there I beg your indulgence

  7. Thanks for the article.
    As a language learner, I find that translation is a handy tool in the toolbox. It enables me to identify grammar areas or structures where there are parallels between my native language and the target language. For example, I think the construction and use of the present continuous overlaps nicely between English and Portuguese. For other areas though, it’s possible to identify that there is no easy bridge/shortcut between the two languages and that another approach is needed.
    This technique is probably most relevant when the L1 and L2 are relatively close.

    • Hi Michael Bain–

      Thanks for your note. One of the things I thought about including in my article was the fact that *all* language learners use translation as a tool (even if we tell them not to do it!), which is why our approach to teaching shouldn’t use it. They’re going to do it anyway. And, you’re right, it can help learners identify parallels and points of convergence.

  8. Rather than just offering unsubstantiated opinion, reference to research findings would be helpful. And there’s a lot of it out there! Your article takes an extremely narrow definition of what ‘translation’ is and goes no further. Rubbishing Duolingo’s silly sentences is easy (and fun, I have to admit), but translation has never really been about word-for-word translation (which seems to be the core of your argument), and I think you should know better.

    • It’s been interesting to watch the small but revealing sub-sample of aggressive, condescending (male) ELT bullies congregate around this post. Why on earth should Katie have to refer to research findings? To pander to your delicate ego because you have some kind of degree in teaching English, your mother tongue, which involved reading books written by similar geniuses to yourself? Where are YOUR references to research findings, or could you not be bothered to enlighten us? I would say her successful career and expertise fully gives her the right to an “unsubstantiated” opinion, but according to the gatekeeper Philip Kerr (whoever he might be) she “should know better”. Goodness me…

  9. Dear Larry,

    As you suggest Dr. Nielson is a leader. In fact, she is a teacher, a scholar, and a business leader. Given these positions, I feel she has a responsibility to us that attends these accomplishments. She should be alert to the fact that when she attacks a competitor and their methods that such “attacks” should not be done lightly. Specifically, I feel her criticisms were over-general, unsubstantiated by third parties, and largely appealed to her implicit authority.

    And so yes, I wanted to question the notion that her authority was sufficient to solve this question. I attempted to show that my experience leads me in a different direction. Of course you may ask the question, is personal experience the best way to solve such an argument?

    No, but it is a start.

    I countered Dr. Nielson’s account with a first person explanation of how translation has been received by my students. Based on my experience, I am fairly sure that translation has a very useful, very effective place to play in one aspect of the language learning process. I countered her truth with my own truth.

    Now, if you look at the scholarly research you will find that translation has enjoyed a resurgence of sorts in the last 15 years. You would find that her proposition, that translation is ineffectual, is debatable. However, after reading Dr. Nielson’s article one would think that the issue had been settled in her favor long ago.

    It has not. And moreover personal interest is at stake in this debate. The issue of whether translation is important pits people who can and do translate in a language class (and with algorithms) against people who can’t or don’t.

    Larry, I don’t have a dog in this race. I don’t use translation enough (2-3% of my week) to be wholeheartedly committed to it. In point of fact, I am interested in listening to professionals who have well-thought out positions in this matter. Dr. Neilson’s account fails to reach this level. And by criticizing Duo-lingual (which I too have criticized) she puts her self-interest into question (she works at Voxy).

    Finally Larry, while trying to honor your desire to come to Dr. Nielson’s aid, I do believe that your arguments are not helping her case.

    • Dear Michael,

      I am not interested in helping Katie’s case one way or another. I too have no dog in this game. What I really took exception to were the patronising “wows” and “bravos” that accompanied your initial post responding to her article. That sort of thing immediately puts people’s backs up and is not conducive to civil discussion.



Leave a Comment

Get our newsletter

Learning Included is our newsletter for learning professionals who want to tap into the latest in learning design innovation, research and best practice.

Evidence-informed, inclusive learning design ideas every two weeks.

Not sure? Check it out first

Learning Included Newsletter