In modern Japanese, the basic terms for 歯 "teeth" and 葉 "leaves" are both pronounced /ha/. When used in compounds, their voiced counterpart becomes /ba/, so that 下歯 "lower teeth" and 下葉 "lower leaves" are both pronounced /sitaba/. So you might wonder, are these homophones or is there a distinction between the two? The answer to this question is yes. In standard Japanese, the word 歯 "teeth" is accented, while 葉 "leaves" is unaccented.
The reason why the pitch accent is important is because it not only tells us that the words are different, but it also tells us that these differences may go back to very early forms of Japanese and that they may manifest themselves differently in other dialects. For example, in the Shuri-Naha dialect of Okinawan, the word "tooth" is pronounced はー /haa/, while "leaf" becomes ふぁー /ɸaː/.
Another consideration would be that emerging tonal contrasts in some languages tend to be the result of interacting sound changes. For example, in the South-East Asian language of Khmer, the elision of /r/ gave rise to a peaking tone in those words where it disappeared. So, it may be possible that Proto-Japanese-Ryukyuan (aka Proto-Japonic) had coda consonants in these words that later disappeared and gave rise to pitch differences.
Korean Relation: Word Parallelism
Although it is currently impossible to state whether Japanese and Korean are related, it is interesting to note that the words for "tooth" and "leaf" are very similar in Korean as well:
- 잎 ip "leaf" (historical variants include 닢 nip and 닙 nib)
- 이 i "tooth" (historically 니 ni, still used as a suffix).
Another curiosity is that the Wiktionary cites 이빨 ippal as being a word referring to the teeth of animals or beasts, but I cannot say whether the -pp- sequence is part of a following morpheme or if it stems from a historical form of the first (*nip?). At any rate, I find the similarity between the Korean words to be quite peculiar when you consider that a similar parallel exists in Japanese. But how exactly could Korean *ni(p) be related to modern Japanese *ha? From these words alone it is impossible to say, but it might be interesting to question a few things:
- Could the initial sound /n/ have been intrusive in Korean or have arose from an irregular sound change (p > m > n)?
- Could the Korean word for "leaf" actually be related to Japanese 菜葉 /nappa/ "greens; vegetable leaves"?
- Are there other similar words that could suggest a relation between the Korean vowel /i/ and Japanese /a/?
- Could the Japanese word for "leaf" have historically been */hap/?
For this last question, the reduplicated Japanese word 葉っぱ /happa/ as opposed to */haba/, as well as the Ainu word ハㇺ /ham/ or ハム /hamu/ could suggest that this might be possible. If so, I am still left to wonder if the Japanese and Korean words could have stemmed from the same source, and what would have motivated the following differences: *pap ~ nip and *pa ~ ni? Consider as well: Japanese 羽 /ha(ne)/ and Korean 날개 nalgae "wing". Otherwise, if no relation exists, it remains an interesting case of parallelism between the two languages.