KaiTheHomoSapien wrote: ↑10 Jun 2021 17:54
Anyone else notice that in American English /ɪ/ is often an allophone of /ə/ and syllabic consonants?
Yes, a great many people have noticed this. It's called "the weak vowel merger". It's also sometimes called the "Roses/Rosa's merger" or the like, after its most famous minimal pair; another clear minimal pair (that is merged in these dialects) consists of the famous 20th century names, "Lenin" vs "Lennon".
Most speakers have this merger. In many dialects, it's not universal (particularly among older speakers); however, it's only systematically resisted in England, in RP-influenced 'colonial' English dialects (Caribbean, African, Indian Englishes), and in Southern US English. Contrariwise, it's the norm in all other US dialects, in Australia, New Zealand, South Africa, Ireland, and to a lesser extent Scotland.
However, there is one complication: in the US, the merger is usually in favour of /I/, but in the rest of the world it is usually in favour of /@/.
Wikipedia says that it's also increasingly common in SSBE. I'm not sure about that in my experience; for me (an SSBE speaker, albeit of a rather 'posh' or 'old-fashioned' sociolect) the distinction is very clear, and while I wouldn't be surprised by someone being confused by a minimal pair in allegro speech, I would expect them to understand the difference, and articulate which one they intended, in careful speech.
What IS definitely a thing in SSBE is the replacement of unstressed /I/ with /@/ in many morphemes and contexts. Wikipedia lists -ace, -ate, -less, -let, -ily, -ity, -ible, be-, de-, re- and e-. I mostly agree with this list, and indeed didn't remember that -less and -let were 'meant' to have /I/, although some are less clearcut than others - I think I often have /I/ in -ily and -ity, and in re- in some contexts (eg /rIspQnd/, but /r@pi:t/). Hence, although I don't merge Lennon/Lenin, I do merge edition/addition except in careful speech (and then I'd hypercorrect to /E/, not /I/). I suspect in practice I may also merge elicit/illicit, although I feel that I "shouldn't". On the other hand unstressed /I/ within the root morpheme itself is usually maintained, as it is in inflectional morphemes like -ing and -ed (so I don't merged battered/batted).
I noticed a difference between the way some people and I say the word "button". For me, it comes out as /ˈbʌʔn̩/, but others seem to say /ˈbʌʔɪn/, turning that syllabic /n/ into /ɪn/
Conceptually, it's the other way around: you eliminate the schwa because it's a schwa, whereas they don't eliminate the schwa because for them it's not a schwa. The elimination of schwa before sonorants is an ongoing 'optional' rule in English.
. Likewise, when I say "important", it's /ɪmˈpɔɹʔn̩t/ but I hear from some others /ɪmˈpɔɹʔɪnt/. I'm not great at narrow transcription, but I notice also that for me the glottal stop occurs while my tongue is in the alveolar position
There's a continuum in how people pronounce this (if they don't just have /t@n/). It can be an unreleased alveolar stop; or, it can be an unreleased glottalised alveolar stop; or it can be a glottal stop. Personally, I mostly just have an unreleased alveolar, sometimes with a bit of glottalisation.
*Also sometimes I just can't tell the difference between syllabic /n/ and schwa followed by /n/. I seem to use syllabic /n/ in words like "important" and "button" but others use /ən/ or /ɪn/.
There is no phonemic difference: /tn/ and /t@n/ are generally interchangeable (with the former more common in allegro speech, but the latter used in more 'careful' speech). The difference can be even greater for some Americans, who flap /t/ before schwa but not before a syllabic nasal, but there are no minimal pairs.