Code-switching as a Teaching Method.

Lameen Souag recently posted at Jabal al-Lughat about an intriguing teaching method:

I haven’t done much language teaching in my life, but as a person who likes learning new languages, I’ve seen a fair range of different teaching methods applied, from only speaking the target language to saying almost everything in English. But the approach used in Simon Bird‘s “#LilMoshom” series of Cree-teaching videos was new to me, and very interesting. Take a moment to watch some of them before reading further [links omitted click through to Lameen’s post for them]

There are a lot of strong points one could comment on – the CGI, the subtitles, and the humour, for instance – but what particularly draws my attention is the way he combines the two languages. To introduce the words he’s teaching, he usually speaks in English – but he doesn’t just gloss, much less lecture (contrast, say, the more conventional approach used in this Ojibwe video series). When speaking in English, he throws in Cree discourse particles and sometimes even content words, gives the sentence a distinctly non-mainstream English intonation pattern which I assume reflects Cree, and even pronounces the English with a Cree accent. In different contexts, the maker of these videos speaks English like any other Canadian academic, so this appears to be a deliberate teaching strategy. The beauty of this is that, before the learner can even formulate a full sentence, they’re already getting a chance to acquire some aspects of language – discourse structure and intonation – that are super-important for actually making yourself understood, yet play a minor role or get left out entirely in many traditional curriculums and textbooks (not to mention grammars!).

Have you ever encountered such a teaching method? If so, did you find it effective?

I’ve never heard of such a thing, but it seems like it could work, and I’m curious to know my readers’ thoughts.


  1. This is more of a technique than a method. The closest I can see this resembling is the Community Approach ( but it also has elements of the Silent Way ( and the Natural Approach (

    Code switching is something most communities of learners engage it. And I can see it as being much more effective than ‘x means y’ approach to teaching vocabulary. I’ve always been puzzled by the fact that some learners (often but not always older) seem incapable of assimilating hardly any foreign language vocabulary even in an immersion situation. But they often have no problem with different names of friends or coworkers or location names. I suspect that’s because they learn these in a kind of code-switching context.

    The problem for some learners is that they are much slower at letting the new language establish its own context based on which they can figure out new meanings. Some years ago, I was proof reading a part of Czech English dictionary and I was able to make corrections on words I was sure no one ever taught me. Yet, I was as sure about my intuitions in one language as the other. I’ve assimilated them as I let existing words form sufficient context (and allowed the formation of mental representations). Yet, when I work with some students they seem to be completely unable to let go of their need to know ‘the dictionary meaning’. So for these, I can see this as being a great technique for habituating new vocabulary. I’m not sure how far it would go. I’ve experimented with the techniques of the Silent Way and Community Learning but never felt confident enough to make them the main approach to my teaching.

    I’m just coincidentally rereading Davidson’s famous paper ‘A Nice Derangement of Epitaphs’ which for some reason reminds me of this. Unusually, for a philosopher of language he talks about how easy it is to understand ‘erroneous’ speech – and things like malapropisms.

    He says “There is no word or construction that cannot be converted to a new use by an ingenious or ignorant speaker.”, I think our ability to deal with this in our first language, can be harnessed in new languages through techniques such as code switching.

  2. “There is no word or construction that cannot be converted to a new use by an ingenious or ignorant speaker.”

    Great quote!

  3. A slightly related hypothesis: actors learn languages more readily than others; method actors, doubly so. Maybe the first step in teaching, say, French to anglophones, should be to get everyone in the class to speak English with a French accent.

  4. Trond Engen says:

    I knew someone who was taught Russian while enlisted in the Norwegian military. One of the first things they were told to do was to imitate the Russian accent and nonsense Russian of the late great Norwegian comedian Harald Heide-Steen.

    Edit: The clip is from about 1980. The context is rumors of Soviet submarines in Norwegian fjords.

    Edit 2: Oh, and there are English texts if you find the right button, Not that the translation is important.

  5. I like mollymooly’s idea – sounds like a fun class! The link does make sense: language is so closely tied to identity, after all…

  6. I was taught Russian exactly as Trune’s friend was at Case Western Reserve University: two weeks of intensive learning to speak English with a Russian accent. I have long since forgotten everything I learned in that one semester, but my Russian-accented English is still pretty good.

    Dominik Lukes: My mother, Marianne Cowan, used CLL to teach both German and English as a Second Language quite successfully in alternation with more conventional techniques (as Curran himself suggested). I wish it were better known. I know she studied SW as well, but didn’t find it as useful.

  7. Back in my young days, ore-sama and my nakama* certainly used a lot of henna gaijin nihongo mid-English, with the accompanying intonation da yo ne~~

    *Translation note: nakama means pals

  8. By the way, thanks Hat for reposting this, and thanks everyone for the comments!

    leoboiko: Funny, Nihongo wo benkyou shinagara sonna koto wo zenzen shinakatta kedo, good idea-mitai ne…

  9. Yes, I think this would be a great way to learn a language, if the teacher is careful and deliberate about her code-switching.

    My Hindi teachers were usually unaware of when they’d switched from Hindi to English (“Hinglish” being their native tongue). When they started on a topic they thought we really needed to understand, like, say, Holi, they’d go on and on about it in English. It was so frustrating.

  10. GT will not touch romanized Japanese (although the language detection algorithm knows it is Japanese). Arrgh.

  11. That’s annoying; surely it can’t be that difficult to do.

  12. So you would think. If you paste “これが私たちが英語で書く方法です” into GT, you get “This is the way we write in English”, and Google politely supplies “Kore ga watashitachi ga eigo de kaku hōhōdesu” at the bottom of the page. But if you copy and paste the romaji back into the Japanese field, Google will not translate it into English.

    But if you use an online transliterator to convert the above romaji to the hiragana-only equivalent “これ が わたしたち が えいご で かく ほうほうです” (you have to adjust ō to ou) and paste that into the Japanese field, it comes out “This is the way I write in English” (almost right). What happens if you then tell it that both the source and target languages are Japanese (which for most languages causes it to do nothing)? Then it becomes “これ が ワタシたち が 英語 で 書く ほうほうです”!

    AI, or Artificial Stupidity.

  13. If Google wanted to do the world a service, it would use accented Rōmáji in both input and output. Kōiūfú ni. Nihongo no Ákusento wo Múshi-suru no wa Iyá da.

  14. I approve of this proposed service. It’s especially hard to find accents for proper nouns.

Speak Your Mind