Talk:Consonant mutation

Adding Korean?
Would Korean also qualify as having consonant mutation? It sounds like it has some examples similar to Japanese, but I'm not sure if they qualify. These do not occur between words and are not allophonic (but may merge or separate sounds): Unreleased stops become a released stop, affricate, or fricative when followed by a vowel, such as the stem /it̚/ turning into /i.s͈jo/ when adding a politeness suffix. Which one it turns into is virtually always the same for a given morpheme, which is reflecting in their writing system. (In the example above, 있요 sounds like 이쑈, but even when the stem is alone, one writes 있- and not 읻-) Similarly, an tenuis stop, affricate, or fricative can become fortis when preceded by an unreleased stop, unless it's a coronal stop → fricative sequence (in which case they turn into /s.s͈/) or ㅎ followed by an unaspirated stop or affricate (which makes them aspirated). Unreleased stops followed by a nasal turn into nasals themselves. For example, /ip̚.ni.ta/ becomes /im.ni.ta/.

I would assume that since allophones are typically non-phonemic (the only phonemically significant ones occur when /t̚/ is followed by a vowel in a suffix or another word, which is sandhi, not consonant mutation), they would not count, despite most of them only occurring word-internally.

Blanket P.I. (talk) 19:14, 17 February 2017 (UTC)

Archive 1
All discussions pre-dating 2017 have been archived at Talk:Consonant mutation/Archive 1 due to the newest post being from 2013 going back to 2005. – Dyolf87 (talk) 14:37, 2 August 2020 (UTC)