jcb wrote: ↑Thu Feb 26, 2026 1:18 pm
Travis B. wrote: ↑Thu Feb 26, 2026 7:56 am
jcb wrote: ↑Thu Feb 26, 2026 3:19 am
At least in American English, one can listen for the differences between the vowels. /t/ will have a short vowel with a sharply rising tone precede it, and /d/ will have a long vowel with a gently falling tone precede it.
Also, /t/ when not preceded by a fricative is commonly glottalized, and the preceding nasal if not syllabic will be elided leaving the preceding vowel nasalized, whereas /d/ will not be glottalized and the preceding nasal will be preserved as-is.
I agree that /t/ is commonly glottalized at the end of a word, but I think that /d/ is also, at least in my dialect.
In my dialect all coda fortis plosives not preceded by a fricative are (pre)glottalized and result in a rising tone in the preceding vowel. Note that this is a good test for whether a fortis plosive falls in a coda or not, and results in some possibly-unexpected conclusions (e.g. the /p/ in
topsy(-turvy), the /t/ in
Nazi, and the /k/ in
boxy are not coda plosives because they are not (pre)glottalized and do not trigger rising tone in the preceding vowel).
jcb wrote: ↑Thu Feb 26, 2026 1:18 pm
What's hard to tell apart are word-final /n/ and /nd/. One can listen for the sandhi if the next word starts with a vowel, but otherwise it can be very hard.
There are quite a few word-final consonant clusters that get readily reduced in many NAE varieties (including my own), especially if not followed by a vowel in the next word, including /nd/, /st/, /ld/, /ft/, /kt/, /sk/, /skt/, and so on.