Games, Diversions & Perl Culture

Chapter 26. Learning Japanese with an HTML Filter

Tuomas J. Lukka

I like to learn new languages by plunging into a good book. For Italian it was Pinocchio, for English The Moon is a Harsh Mistress. I keep a dictionary handy, and I spend many hours rechecking words until I remember them. It’s tedious, but in the end it’s still a more interesting way to learn than the usual route of beginning with very simple phrases and vocabulary and building up slowly, reading childish stories about uninteresting subjects.

I tried this with a book on Go strategy written in Japanese, and quickly hit a wall. With many languages you can infer the meaning of words by what they look like and remember them by how they sound. But in everyday Japanese text, there is no way for a beginner to know how a given phrase is pronounced until he learns the two thousand characters in common use. Furthermore, each character can usually be pronounced in two or more ways depending on the context (see the sidebar on Japanese Characters).

It might still be possible to learn Japanese with this method, but the task is complicated further still by the fact that the character dictionaries are tough to use—Japanese has 2,000 characters, so you have to find the words graphically, which is much more time-consuming. You can’t leaf through the dictionary as you can with Western writing systems.

So I ended up auditing the Japanese course at the university where I work. Even though the teacher made the course as much fun as a language ...