Manny Rayner's book reviews

We've just started a project where I'll need to know some Chinese, so I thought I'd dig out this book, which had sitting on my shelf for the last year and a half. The project looks like it will be fun; we're using our speech recognition and speech translation technology to build a kind of automatic conversation partner for students who are learning a new language. The computer shows you a picture indicating what you're supposed to say, the student tries to say it, and then they get rated on how well they did. So far, our examples are about ordering food in restaurants. So the machine might show you a picture of the client smiling and saying something to the waiter, plus the figure 2 and a picture of a hamburger. The student is supposed to say something like "Could I have two hamburgers please?", and then they get a score depending on how close they were.

The initial version of the system will be for English, French and Japanese, languages where we're reasonably confident that we know what we're doing. But we'd like to understand better from our own experience what it feels like to be a beginning student using the system, so we also want to build a Chinese version; none of us know any Chinese worth mentioning. Hence my interest in this language. I've been spending maybe an hour or so every day with Po-Ching and Rimmington for the last two or three weeks, and I'm quite surprised how much I've picked up. I haven't really been making any attempt to learn how to speak; my goal was to learn the basic grammar and pick up some vocabulary, and that's succeeded better than I'd dared hope. This is clearly a good introductory text.

I'm trying to figure out what makes it so effective. I think it's something to do with the way they introduce new material. Some texts are absolutely slavish about starting from the beginning, and only ever using words and constructions that they've already explained. The problem is that, for most of the book, you can hardly say anything at all. These authors have a freer style. They're happy to use things they haven't explained yet, but they do it in moderation, and in contexts where you can half guess how it works. Then when you get to the formal explanation, you already have some feel for it, so it immediately makes sense. It's a bit of a tightrope act, but it certainly seems to work. I got to the end of the book easily, and was able to follow almost everything. I'm not sure how much passive vocabulary I've picked up, but it feels like quite a lot.

One organizational point is worth mentioning: they are completely consistent about giving everything in both Chinese characters and Romanized form, so that you always know both how it's written and how it's pronounced. It sounds obvious, and maybe it's standard for Chinese; but it certainly isn't in Japanese texts, where often you only get one form and have to guess the other. It's extremely distracting, and I hate it. Full marks to these people for taking the trouble to get it right.

As several people had told me, Chinese is quite fascinating. I think it's the most logical language I've ever come across... very, very clean and elegant. I'm looking forward to learning more about it.