OK so here’s a new thing I’m trying. As I’ve mentioned, my biggest problem is speech. I can understand most of what I hear, and I can read the things I need/want to be able to for the most part (and I know how I’ll be improving that over the next ~11 months). My writing is OK, and I’ll be working on that even more with a tutor starting next term. But my speaking is just not there yet.

I have no problem getting my meaning across in most any situation, but the problem is that I never feel sure if I’m saying things right. Or maybe I know that what I’m saying is correct, but people still look at me funny and say things like “Oh, you can just say X.” Sometimes they just don’t know what I mean, or I think I’m saying one thing but what they understand is something else. This boils down to the fact that 1) I’m not comfortable enough with putting fluid, correct sentences out yet, and 2) I need to learn to say things the way native speakers would, rather than expressing things the way an American would, but in Chinese. The biggest problem though is that I just can’t spit out Chinese. I have to think about it, I make a lot of unnatural pauses, say “怎麼講” a lot, etc. I can’t, as they say in Chinese, just 脫口而出.

So to that end, I’m trying a method that Mike Campbell (Glossika on YouTube and Facebook) has talked about in some of his videos. His Chinese is unbelievable, and he says he developed this method while learning Chinese and uses it to learn other languages and to teach people English at his school here in Taipei. From what I understand, his students pay a serious amount of money to study at his school, so I think these factors all add up to someone we should pay attention to.

What I understand of his method is this. First he buys books containing thousands of sentences in parallel format. For our purposes, the best thing that’s out there are for learning English, not Chinese, but as long as the MP3 recordings contain both languages, it’s fine. Second, he goes through and records himself reading these sentences. 500 per day. I currently shoot for somewhere in between 100 and 200, but I’m trying to figure out ways to make it go more efficiently so I can record more. Then over the next several days he reviews them, repeating after the tape when he can. After he’s gone through the entire book, he goes through and does it all again, a total of 2 or 3 times.

He claims that a good base level can be achieved after 10,000 sentences (sounds like a familiar number), but that after 20,000 sentences, your <insert language here> should be really good. I take that to mean 20,000 different sentences, rather than, for example, 5 repetitions each of 4000 sentences, because when he talks about numbers of repetitions he tends to add a zero. I’ve collected several of these books now, and I do have about 20,000 sentences ready and waiting. I have a book of 10,000 sentences besides this, but probably won’t use it because the MP3 recordings don’t contain Chinese (more on this in a moment).

Mike says that the most important thing is that it gets your tongue moving, and I agree. But another big thing for me is that listening to myself speak really brings focus to the errors I make. I listen to it every day for a week, and every time I hear X error, it makes me mad, and I end up saying that sentence a few times so I get it right. So double bonus there.

I feel like this method is already helping a bit, but it’s too soon to know for sure. I plan on sticking with it for at least a few weeks to see how it goes, because I think this really has promise. There’s something about it, and I think part of it is that you’re involving more senses while doing the recordings, and then invoking that by reviewing. You could just listen to the MP3s that come with the book, but you have no connection to that. When you’ve recorded the MP3s yourself, you remember sitting there recording it. You remember when you flubbed that sentence three times before you got it right, you remember that the thud you just heard on the recording was when you dropped the book, etc. All those associations come back, which I think helps to make stronger connections with the material. And while you’re recording, you hear the native speaker say the sentence, you hear and feel yourself repeating the sentence, and you see the sentence written on the page. Using more senses helps to reinforce the neural connections, which helps you to learn better (score for having an awesome teacher for a wife, who talks about these things).

He talks a lot about muscle memory, and I think that’s a huge point. You say something correctly enough times, and it’s going to feel wrong to say it incorrectly. There’s other stuff happening here too though. By looking at the English, you’re turning the sentence into comprehensible input, even if you don’t know one of the words. So you’re essentially feeding your brain a ton of comprehensible sentences in a very focused way, and then repeating them, which involves all those processes I mentioned above. Then when you review, you can do it all over again, only the fact that you don’t have English on your own recordings forces you to recall the meaning, enhancing the connections even further.

As an aside, I don’t think this is the most suitable method if there tend to be a few words you don’t know in most sentences, because it’s too much for your brain to pick up in such a short period of time. I could be wrong, but Mike himself says the method is for people who already know the target language well, and that you should use another method to get yourself to that point. I definitely see the logic there and agree, though I feel like some of this method can be successfully adapted to other materials to get you to this point. I haven’t thought through that yet, but if this goes well I may do something like that for Japanese.

Now, here’s what I believe I’m doing differently than Mike. I think he just reads the sentences aloud himself, without the aid of a recording of a native speaker. I may be wrong about this, but that’s what I gather from the videos I’ve seen. I’m using Audacity to add silences to the recordings after the native speaker reads each sentence (or more accurately, I’m muting the English sentences because I don’t need them, which leaves a space of silence after the Chinese sentences). Then I put the track onto my phone. While listening to the MP3 through earphones so the recording I’m making doesn’t pick it up, I repeat the sentences into Audacity. I can do this more or less on the fly, stopping only to re-record when I make a mistake. Then using Audacity I edit the silences out of my recording, leaving only enough space to repeat after it while reviewing later. There’s a plug-in that comes with Audacity that does this. Doing all this takes more time, but I think it’s worth it because then I’m modeling a native speaker rather than starting from my own conception of how a sentence should sound.

Things you have to be careful about: One, some of these books have direct translations of the English rather than an equivalent Chinese phrase. You don’t want that, because it’s just explaining the English rather than telling you how a Chinese-speaker would express the same thing. Two, some of them don’t have Chinese on the recordings, because again, they’re books for learning English, not for learning Chinese. That last bit is a good thing because you know the Chinese will be correct since it’s written by a native speaker (some of these books written solely by Taiwanese people have absolutely horrendous English), but on the other hand, sometimes they don’t see the need to include Chinese, so you have to look out. You won’t always know until you get the thing home either, so it’s a bit of trial and error.

So we’ll see how this goes. I think there’s a lot of potential here, and I’m excited to see if it works as well as I think it might. I’ll report back here.