Why is [好]{す}き pronounced with a voiceless/almost silent U, but [隙]{すき} is not?
I've used 2 sound sample sources for both, Jisho and Rikai sama. They both pronounce the first (like) with a voiceless U. However, they both pronounce the second (gap) with a voiced U.
I thought that if you disregard external factors like mood/emphasis/politeness/dialect/whatever for a second, that devoicing occurs when U and I are sandwiched between two voiceless consonants, or if there's a voiceless consonant first and then nothing after.
But here 1 word takes that rule and the other doesn't. Does that mean that some specific words have a voiced U where it normally shouldn't be there?
/u/
sound in all the words starting with すき that I looked at -- 好き、隙、スキー, etc. Note that the NHK dictionary is based on Tokyo pronunciation, where the initial す has a low pitch accent, vaguely similar to an unstressed syllable in English. I have a friend from Kōbe who uses the Kansai pitch pattern, where the す in 好き has a high pitch accent, and the/u/
is not devoiced. Is it possible you're hearing a Kansai-Kantō difference?