There& #39;s far more to this glottalization scheme than I realized! After analyzing a series of songs, dialogues, and other miscellaneous recordings, I& #39;ve been able to generalize the rule I formulated in my previous thread and get a rough idea of the environment to which it applies. https://twitter.com/adam_farris1/status/1243015371751211011">https://twitter.com/adam_farr...

From the data I observed, it looks like the extent of applicability varies greatly throughout the colloquial register - for purposes herein, I propose three different subregisters, formal/honorific-colloquial (FHC), conversational-colloquial (CC), and vulgar-colloquial (VC).

Let& #39;s start with VC (certain songs and comedic sketches).

Rule 1. [-Vt# -Vk# -Vp#] > [-ṼʔN̩#] - Word-final, unvoiced stops (denoted by P from now on) are glottalized (preceding -V is often nasalized) and followed by homorganic nasal phonation- in all environments, EXCEPT ...

Rule 1a. [-VP1 #___ #P1] > [-Vʔ #___ #P1] - When gemination occurs across word boundaries, P1 is glottalized although nasal phonation is never sounded and the preceding vowel is never nasalized.

e.g. ⟨rahasak kīwwa⟩ >rahasaʔ kīwwa & #39;secret-IND tell-PST& #39;

Rule 2. [-VP1 P1-] > [-Ṽʔᴺ̩¹ P1-] - For geminates occurring within words, the first stop gets glottalized although the nasal release is much weaker. The vowel preceding the first member is nasalized.

e.g. ⟨duppat⟩ > dũʔᵐpãʔn̩ & #39;poor& #39; (notice Rule 1 also applies here)

Exceptional VC processes requiring further study:
1. The reappearance of glottalized "ghost" consonants in certain environments:

e. g. ⟨kaⁿduḷut ekka⟩ > kaⁿduḷũʔn̩-(t)-ẽʔᵑka? & #39;tear-PL along with& #39; (Rule 2 also applies)

2. Initial syllable glottalization causes medial w elision and ghost stop:

⟨ekwuṇu⟩ > ẽʔŋ̍-(k)-uṇu?

Thoughts?

I& #39;ll cover the rest (FHC and CC) on my blog. This thread is already getting way too long!