This is something I learned a long time ago, but when I mentioned it to a friend who’s been studying Japanese for years, she apparently didn’t know this trick. So this might be more obscure trivia than I thought, and it might help somebody to post it here.
Let’s look at the Japanese kun’yomi number words from one to ten:
| Arabic numerals | Kun’yomi |
|---|---|
| 1 | ひと hito |
| 2 | ふた futa (huta) |
| 3 | み mi |
| 4 | よ yo, よん yon |
| 5 | いつ itsu (itu) |
| 6 | む mu |
| 7 | なな nana |
| 8 | や ya |
| 9 | ここの kokono |
| 10 | とお too, と to |
The pattern is that halves have the vowels /i/ and /o/, and doubles have the vowels /u/ and /a/ in the same places. So what’s half of huta? Hito. What’s the double of mi and yo? Mu and ya. Although yo is more often read as yon nowadays due to influence from the on’yomi of san.
There’s no easy trick to memorize nana and kokono, though, and itsu/too doesn’t really follow the pattern, either. The best you can do is squint and say “well they both have a T in them, close enough.”
The /i/ → /u/ sound change is also used in a number of compound words that exhibit apophony (転音 ten’on), but not /o/ → /a/, as I understand it. In ten’on, it’s rather /e/ that turns into /a/. Less often both /e/ and /i/ can turn into /o/ as a result of ten’on.