Diacritics to revisit

[bambooforest]

Some combining diacritics don't seem to appear in Phoneme, but do in Allophones:

velarized/pharyngealized (combining tilde overlay) phoible %>% filter(grepl("̴", Phoneme)) %>% select(Phoneme) %>% distinct()

Doesn't appear anywhere afaict:

denasalized (combining not tilde above) phoible %>% filter(grepl("͊", Phoneme)) %>% select(Phoneme) %>% distinct()

nasal emission (combining homothetic) phoible %>% filter(grepl("͋", Phoneme)) %>% select(Phoneme) %>% distinct()

derhoticized (combining breve below) phoible %>% filter(grepl("̮", Phoneme)) %>% select(Phoneme) %>% distinct()

This appears in one phoneme type in one language (638 NENETS):

stiff (combining caron below) phoible %>% filter(grepl("̬", Phoneme)) %>% select(Phoneme) %>% distinct()

This appears in two languages phoible %>% filter(grepl("̚", Phoneme)) %>% select(InventoryID, LanguageName, Phoneme)

InventoryID LanguageName Phoneme 1 1411 Ngomba q̚ 2 1567 Tangale d̪̚

[drammock]

Yes, I noticed this too. Things showing up in allophones but not phonemes is not all that surprising, so I didn't worry about it much. Here are some comments:

• I'd prefer to get rid of combining tilde overlay if we can, since we have dedicated, separate modifiers for "velarized" and "pharyngealized". But knowing which one to pick requires going back to the sources.
• I already removed (by commenting out) derhotacized here) and nasal emission here... I must have missed denasalized in my testing.
• I also noticed the rarity of stiff and would like to dig deeper into it
• unreleased is weird on a phoneme, it basically means it can only ever occur utterance-finally? Probably a mistake, should go back to the source on that one. (EDIT: technically it's no audible release which doesn't restrict it phonotactically, so maybe it's OK. still worth checking the source doc though)