Chorusing

[Bakunin]

Chorusing is a technique to improve your accent. The basic idea is to repeat short speech segments many, many times together with the recording to internalize pronunciation and prosody. In its current form, chorusing was suggested by Olle Kjellin. He recommends to use sentences or sentence fragments of only a few seconds length and loop them using Audacity or some other audio software.

Catherine, the author of the very comprehensive Thai learning blog WLT, has put up a blog post detailing the technique and linking to a few of Kjellin’s papers. For those who would like to download these papers her post is a convenient source. You can find the post here.

I personally don’t have a lot of experience with this technique and also no definitive opinion on it, but his main point, that prosody is a key element in native speech totally underrated in most language learning approaches, is convincing.

[Jar-Ptitsa]

I agree: we learn the pronunciation of the vowels, consonants and words, but not so much the rhythm and almost not at all the prosody. The sentence pronunciation I suppose, and the pitch. I know that languages like Enlgish haven't got pitch like the tonal ones, but nevertheless it has its own pitch patterns I think.

I can imagine that chorusing would be very helpful. It seems a bit like the scales and arpeggios on a musical intrsument.

[neofight78]

I integrate audio into everything I study. I always make sure have the text and audio for everything. A big part of my study routine is flashcards, and all my flashcards have audio. I always answer out loud and include correct pronunciation as part of getting an answer right. I do this not only for words but sentence and grammar cards too. The sentence cards of course give me the prosody aspect. If I sound off, I mark the card as failed and listen to and repeat the audio copying it as closely as I can. It's not exactly the same approach as described but I suspect the effects are similar.

Pronunciation and listening are my two strongest skills, which I put down to my approach rather than any particular talent. In my B1 listening exam I scored 100% when the examiner said this usually the part people find the hardest. So I think there is something to be said for increasing the amount of audio used, not just for output but also input.

[coler1]

For what its worth, I've found that chorusing the L2 versions of my flashcards out loud has dramatically increased my recall ability when learning vocabulary.

[garyb]

I agree on the importance of prosody. It's sadly given no attention in most teaching. I think it's part of a bigger problem, commonly held beliefs that accent is something that you need to have a natural talent for, or that you'll pick it up naturally as you learn the rest of the language; for most learners, neither are the case.

I experimented with Olle's chorusing method around a year ago. I think it helped take my prosody from bad to okay, but not from okay to good. It helped me identify and fix a few bad habits. It made me very good at pronouncing the example sentences I used, but the improvement didn't generalise as much to other phrases or spontaneous speech. However I just did it alone, so it was limited by my own ability to give myself feedback based on an imperfect mental model. Doing it with feedback prompts from a trained tutor, as described in Olle's paper, would probably take me much further.

Take everything I write with a pinch of salt because my own accents are still not very good, despite a lot of work that is still ongoing. I'm just throwing out ideas, not giving advice. To me it seems that there are two important parts in improving pronunciation (both prosody and individual sounds): building an internal model, and applying that model. But like most learning, it's an iterative process: improve model, apply changes, get feedback, improve model more, etc. I suppose Olle's idea (complete with the external feedback) tries to do all of this in a loop, which makes it pretty smart.

Some people have a talent for accents and seem to be able to construct and apply quite an accurate model just from listening. Immersion seems to work wonders in some cases as it gives lots of listening and lots of practice - I've met quite a few people who've lived in France for years and have great accents, even though their level in the language otherwise often isn't very impressive - although others still have a strong accent, so as usual it's not a magic bullet for all. For these people with a "good ear", methods based on listening and repeating (chrousing, shadowing, Pimsleur, etc.) work great. For the rest of us, it seems like we need to study the prosody and understand how it works (rhythm, intonation, stress) before we can effectively practise and apply it. The hardest part is often eliminating bad habits from our native language and then stopping them from returning. A lot of the work I've done has given temporary improvements that have gone away again once I stopped paying attention.

[chobbs]

I have seen great results from Olle's chorusing method, for my own French, for my son's German, and from a variety of fellow language learners who have posted their results for a number of different languages. It doesn't take much time at all to really see some improvements. I plan to work through the various languages on my target list well before I start learning them just so that prosody is fully ingrained. I am curious to see how well it helps me with a language I start from scratch as I was already well into my French studies when I utilized his approach.

I think the mirror neuron aspect is a critical piece of the method. I do wonder though how much it varies person to person and how much volume it needs to solidify in each of those cases. On a semi related note, I know that even after a few days of having some of our UK coworkers around I start falling into their accents and not too many of my other coworkers seem to experience that affect. I have the same thing happen for me when visiting other parts of the US as well so it isn't something particular to a specific accent. Not sure if that type of malleability is a big factor in success here or not.

[neofight78]

For what its worth, I've found that chorusing the L2 versions of my flashcards out loud has dramatically increased my recall ability when learning vocabulary.

I've read several times the criticism that if you learn flashcards it's a waste of time because you won't be able to remember the word when you need to use it. I've never found this to be the case, it seems to me that saying something out loud cements it in a more accessible way, I wonder if that is the difference in some cases. You often hear the claim that writing things out helps people, so there is definitely something about the physical that boosts the memory in a practical way.

[neofight78]

To me it seems that there are two important parts in improving pronunciation (both prosody and individual sounds): building an internal model, and applying that model.

That's an interesting description. I just listen and imitate, I'm not building any kind of model, at least not on a conscious level. In fact if I start to think about things too much that's when it goes wrong. But perhaps you are not describing a conscious process?

[Aozu]

I personally don’t have a lot of experience with this technique and also no definitive opinion on it, but his main point, that prosody is a key element in native speech totally underrated in most language learning approaches, is convincing.

I agree that it's underrated, and I personally find it useful. I even sort of talked a little about this in the dumb video I made earlier this week on my log where I do the Glossika steps, since Glossika does a lot of this, so that's kinda funny timing.

I think part of the reasons it's underrated might be historical and technological - until fairly recently it would have been an incredible pain to try and do chorusing, especially if you want to hear yourself. You'd have to have some kind of goofy rigged-together tape system and be rewinding all the time, guessing how far back to go, and then have some way to record yourself and play back after and during this whole endeavor.

Whereas now you just fire up Audacity, lean into the mic, and go. Waaaay easier.

[samfrances]

The article by Olle Kjellin talks mostly about using chorusing as an absolute beginner. Will this technique help intermediate learners improve their accents too, do you think, or is it too late by then?