I have a question about numbers.

[Totakeke423]

Why is it that the word for 300 is さんびゃく, but the word for 400 is よんひゃく? I understand about rendaku and all that, but さん and よん both end in "ん." So why are they different? Is it just an irregularity that needs to be learned? Thanks.

[becki_kanou]

Perhaps someone else can enlighten you as to the why, but I think you are better off just remembering it. This rule goes for almost all counters from the はひふへほ row, as far as I know.

いっぱい、にはい、さんばい、よんはい
いっぴき、にひき、さんびき、よんひき
いっぽん、にほん、さんぼん、よんほん

one exception is:

いっぷん、にふん、さんぷん、よんふん

[keatonatron]

I don't know of it's correct or not, but when I pronounce よん and さん, the ん is often a lot weaker on よん. When I say さんびゃっく, my tongue completely touches the roof of my mouth on the ん, but in よんひゃく it only closes halfway (the ん sound perhaps comes more from my throat). I think this is because さ ends with my mouth stretched wide, and よ ends with my mouth in a small O shape (which is harder to change to ん quickly).

You can see this in some words as well: 四時 becomes よじ (the ん disappears completely!).

[richvh]

The reason is fundamentally that both さん and ひゃく are originally Chinese words containing consonants that don't map well to Japanese, while よん is pure Japanese. The combination of さん plus ひゃく to form さんびゃく has to do with the interaction of the final consonant of さん and the initial consonant of ひゃく, and how they sounded to the Japanese ear when they were introduced.

[Yudan Taiteki]

There is no reason.

There has been a lot of scholarly research into rendaku for over a century (it's one of the oldest topics in the western study of Japanese linguistics) and nobody has ever been able to find a pattern for when it occurs or why. Lyman's Law gives a rule for one case in which it doesn't occur, but that's as far as it goes.