Computer keyboards derive from typewriter keyboards, of course. The basic QWERTY layout we now use was created originally for manual typewriters to minimize the number of key jams, by physically separating the letters of common bigrams -- or so the mythology goes. Actually, part of the goal was to slow the typist down.

Regardless, we're still using it even though the original function has long since ceased to matter. (When's the last time you saw someone using an unpowered manual typewriter? When's the last time you even saw such a beast?)

What with Americans inventing all of this, and with English having a relatively small number of characters, it wasn't all that tough to fight them onto a typewriter keyboard. It's more of a problem for some of the European languages, especially in the far north, who have a lot more characters than English.

But for the Japanese, it's really a mess. A manual, unpowered Japanese typewriter bears no resemblance at all to one of ours. It has a tray of type pieces. There's only one key, but it's mobile. What happens is that you move a pointer over the type you want to use and press down. The type is popped up out of the tray, grabbed by a hammer, which swings it up to strike the page. When you release the key, the type is returned to the tray.

The tray contains the whole hiragana set plus maybe 50 or so very common kanji, and if you need a kanji not in the tray, you go to a file cabinet and get the one you need and drop it into an empty spot in your tray. It's an amazing device but it's also run by "hunt and peck" and an experienced typist can maintain a rate of maybe two characters per three seconds, with occasional pauses to visit the cabinet.

None of that makes a lot of sense for a computer data entry device, and what they ended up doing was to adapt American computer keyboards by giving all the keys alternate readings. You can switch the keyboard from romaji mode to hiragana mode.

That strikes me as very complicated. There are 74 characters you need in order to write Hiragana, even ignoring punctuation or any kanji. Fitting all of those onto a keyboard is tough. (I don't know for sure how they do it, but I assume that they're putting two characters per key and using the shift key.)

If we were designing a Japanese keyboard from scratch, without any knowledge of computing equipment history, how would we do it? It occurs to me that what you'd do is to create a double-action keyboard, to take advantage of the regular design of the hiragana set. Your left hand has 12 keys. Your right hand has 14. What are they?

Left hand: あ い う え お ん ゃ ゅ ょ っ わ を

Right hand: あ か が さ ざ た だ な は ば ぱ ま や ら

Most characters are created by simultaneously pressing two keys, one with each hand. To get す su you press う u and さ sa simultaneously. To get え e you press え with the left hand and あ with the right.

In other words, the first five keys for the left hand are the leftmost members of the rows of the chart and the keys for the right hand are the top characters of the columns. The left hand also gets single-action keys representing the orphans ん ゃ ゅ ょ っ わ を. Those get pressed without any right-hand key.

That's only 26 keys, evenly divided between the hands, which is very manageable. It leaves room on the keyboard for digits, borrowed punctuation marks, and shift keys for getting katakana.

To get kanji, it's the same as now: you enter in hiragana and the word processor substitutes kanji for kana when it can. If it picks wrong, you can put your cursor on the wrong kanji and use function keys to scroll through alternate choices.

Anyway, that's how I would design it...

Posted by: Steven Den Beste in Japanese at 12:07 PM | Comments (12) | Add Comment
Post contains 686 words, total size 4 kb.