Posting up links to Google Translate pages - you must be joking. Google Translate is notoriously poor when it comes to Japanese-English translation. I would not agree that it has gotten better each year. Rather, it would be more accurate to say that it has a very long way to go yet.



I was really talking about how those characters for sumitsuke would be rendered in hiragana, since, if you look up kanji in a Japanese dictionary, the pronunciation for the character is rendered in hiragana (or katakana as the case may be). Romanizations of Japanese can only derive from the hiragana/katakana.

The issue here seems to be whether the character which stands alone as 'tsuke' is read as '~tsuke' or '~dzuke' as a suffix, when appended to 'sumi'. Correct? I realize that this will seem like some insanely minor and obscure detail to many reading this, but it is not so if you are a student of the Japanese language, so please bear with me.

I'll try to illustrate the issue firstly with a relevant example - and apologies to those for whom this is old hat. Take a word beginning with the 'su' sound, such as 'sushi'. When the word 'sushi' forms a suffix, the 'su' sound is voiced as 'zu', as in maki-zushi (rolled sushi). The sound 'su' has a hiragana character representing it, and the sound 'zu' has the same hiragana character with a couple of tick marks added above, indicating that it is to be voiced as 'zu'.

The same pairing occurs with the phoneme 'tsu', which as the initial consonant of a suffix may, at times, be voiced as 'dzu'.

Romanizations attempt to convey these voice shifts by their spelling.

This phenomenon of these voice shifts is called rendaku in Japanese. This area of Japanese is a bit of a minefield - as the wiki link mentions, "it's unpredictable".

Such is the case for sumitsuke. I might add it is the same situation as well for another relevant pair, namely 'sumi' + 'tsubo' - a 'sumitsubo' or Japanese inkline. It is never written as 'sumidzubo' - and again, I'm talking about how the word is rendered in hiragana.

Looking through the link on rendaku provided above, I do believe the reason sumitsuke is not rendered as sumidzuke is 'Lyman's Law'.

By the standard logic of how rendaku is manifest, both sumitsuke and sumitsubo might be voiced as, and written as 'sumidzuke' and sumidzubo', but they are not, at least not according to Japanese dictionaries. That's why I asked you to provide a reference to a Japanese language source for that, not 'Google Translate', and not "as was explained to me". Japanese people are, it would be worth noting, no more expert on their own language than English speakers are experts on their own. That is to say, expertise varies widely. Many English speakers miss-spell and mispronounce English. We all know that. Imagine if a Japanese person sought guidance as to how to spell or pronounce an English word from an English speaker who was not all that savvy about their own language?

And hearing a thing secondhand does start a process of telephone as we all know.

Again, I said simply, show me a link to a Japanese dictionary which shows a hiragana spelling of the kanji for sumitsuke as sumidzuke. The Japanese dictionary entry, just like our English dictionaries, will indicate the preferred pronunciation for any word listed. If you don't have that, the rest is, I'm afraid, waffling.