Contractions in english
>>I haven't really looked into vowel length in my dialect that much, but I don't think that a hypothetical word *wone would be the same as "won't" before vowels. It would probably be like yours: *wone [wo~:n] as opposed to "won't" [wo~n] (before a vowel). <<
I strongly suspect that many if not most NAE dialects are really not all that different from my own with respect to vowel length except that the distinction in my dialect between short and long vowels may be stronger than in most NAE dialects (for example, I have heard some individuals have what sounded to me like long vowels in places where I would expect short vowels) and some people may have extra rules with respect to how vowel length allophony operates (such as with respect to syllable closedness). The main thing, though, is that most dialects do not have rules like final devoicing which really make vowel length allophony strongly distinctive, and hence it tends to be overlooked a lot or just mentioned in passing, even if there are limited cases present where it is potentially distinctive.
Yeah, I think I would likewise maintain a slight distinction between prevocalic "won't" and prevocalic "wone". In a similar way I can perceive a slight difference in my pronunciation of the words "batter" and "badder", although I don't think I can detect any distinction in my speech between "better" and "bedder".
I don't generally mark allophonic vowel length when I transcribe my speech because it seems rather variable for me, and I don't find the alternations to be very noticeable in my speech. (And I think that my allophonically lengthened /E/ may not be quite as long as, say, the [E:] of Estuary "care" - so I maybe I'd have to resort to half-length marks.)
I've considered the idea of systematically using length marks to mark free monophthongs: I think that even accounting for allophonic length variations, my free monophthongs always tend to be a bit longer than my checked vowels in any given context. This convention would help me make clear the slight but audible difference between my [Q:] and RP [Q], for example. (And interlinguistically, it would let me show the difference between my [i:] and, say, Spanish [i], without resorting to the "super narrow" transcription [Ii] or [I_+i].)
>>Yeah, I think I would likewise maintain a slight distinction between prevocalic "won't" and prevocalic "wone". In a similar way I can perceive a slight difference in my pronunciation of the words "batter" and "badder", although I don't think I can detect any distinction in my speech between "better" and "bedder".<<
I can detect differences in all these words, with the strongest being between "won't" and "wone" and the weakest being between "better" and "bedder" (even though this is still definitely detectable), with "batter" and "badder" being in between (but still quite salient).
>>I don't generally mark allophonic vowel length when I transcribe my speech because it seems rather variable for me, and I don't find the alternations to be very noticeable in my speech. (And I think that my allophonically lengthened /E/ may not be quite as long as, say, the [E:] of Estuary "care" - so I maybe I'd have to resort to half-length marks.)<<
The thing is that for me, such alternations are well-defined (between short, long, and overlong vowels), and are not optional or interchangeable, especially but not solely in final syllables. They are particularly noticable in final syllables, due to such being more distinctive than other positions, but they are present in all positions. Also, using the wrong length in a place where a different length would be expected may be confusing, even in medial positions, due to things such as such likely resulting in the wrong postvocalic obstruent phonemes being perceived, amongst other things. Furthermore, elided consonants generally do not lose their effect on vowels IMD, and consequently vowel length enables listeners to "hear" elided fortis obstruents, and consequently realizing vowels as long rather than short would sound wrong, as it would sound as if such obstruents weren't there to begin with.
(Note that the usual vowel allophony does not produce overlong vowels - rather, they are produced through lengthening vowels that would otherwise be long through other phonological phenomena, usually assimilation of a following vowel phoneme after the elision of an intervening consonant.)
>>I've considered the idea of systematically using length marks to mark free monophthongs: I think that even accounting for allophonic length variations, my free monophthongs always tend to be a bit longer than my checked vowels in any given context. This convention would help me make clear the slight but audible difference between my [Q:] and RP [Q], for example. (And interlinguistically, it would let me show the difference between my [i:] and, say, Spanish [i], without resorting to the "super narrow" transcription [Ii] or [I_+i].)<<
That is the thing: my dialect lacks such a differentiation between free and checked vowels, and rather the differentiation between such in my dialect is solely with respect to phonotactics (and does not correlate to tenseness versus laxness, as /Q/ is free in my dialect despite being lax).
I've been wondering about phonemic length recently, because I perceive my tense vowels as being longer than my lax ones, but I don't have the equipment to check if that's accurate. NAE is generally said to have lost the phonemic length distinction from Middle English, but doesn't the tendency toward diphthongisation of tense vowels make them all longer anyway?
>>I've been wondering about phonemic length recently, because I perceive my tense vowels as being longer than my lax ones, but I don't have the equipment to check if that's accurate. NAE is generally said to have lost the phonemic length distinction from Middle English, but doesn't the tendency toward diphthongisation of tense vowels make them all longer anyway?<<
Of course, though, another trend in NAE has been the loss of diphthongization of tense vowels; while this is definitely not the case across the whole of NAE, large sections of such either have monophthongized tense vowels or at least reduced their diphthongization to being rather marginal.
<<Of course, though, another trend in NAE has been the loss of diphthongization of tense vowels; while this is definitely not the case across the whole of NAE, large sections of such either have monophthongized tense vowels or at least reduced their diphthongization to being rather marginal.>>
This is the case in my accent: /i, e, A, u/ are all usually mononophthongs and /o/ is a very narrow diphthong. I still perceive them as longer than the corresponding lax vowels though. I was wondering if there are any studies that confirm or refute my perception of longer tense vowels.
"won't" and *"wone" both have short vowels for me. /wont/ and /won/.
>>I've been wondering about phonemic length recently, because I perceive my tense vowels as being longer than my lax ones, but I don't have the equipment to check if that's accurate. NAE is generally said to have lost the phonemic length distinction from Middle English, but doesn't the tendency toward diphthongisation of tense vowels make them all longer anyway?<<
When I speak individual words carefully, there *may* be a *very* slight difference between checked and free vowels, but in actual speech there really is no difference for me; and even the aforementioned difference between checked and free vowels in extremely careful speech is far smaller than the typical difference between short and long vowels due to vowel length allophony IMD, which does not need to be listened for carefully at all.
Note though that such is separate from diphthongization, even though I do actually have a slight tendency towards diphthongization of tense vowels other than /o/ - but this diphthongization seems to not be linked to historical Late New English diphthongization. It is the sporadic *lowering* of the starting points of tense vowels /e i u/ when long (as opposed to the raising of the ending points of /e/ and /o/ present in historical Late New English), not affecting short tense vowels, which remain monophthongs consistently. The thing that really sticks out in my mind about this, even though it is generally rather slight and sporadic, is that it is a form of asymmetry between short and long vowels beyond just vowel length, and hints at short and long vowels not just being superficially different phonologically despite being similar in origin phonemically.
I should modify what I said above: it is really only particularly marked diphthongization of long /e/, /i/, and /u/ which has stuck out, and such only occurs only sporadically, but from looking into it more carefully, it seems that /e/, /i/, and /u/ are in general slightly diphthongized across the board, with slightly lowered starting points relative to the positions of short /e/, /i/, and /u/ combined with ending points equal to said positions.