Such a wide variety of phonetic possibilities...I'm mostly interested in American accents that have a phonetic distinction between clear and dark /l/. I've heard some that do (in the South maybe?) but the vast majority don't seem to.
Realisation of /l/
I'm not familiar if the distinction really occurs... I have a few close friends from Mississippi and they don't distinguish... I definitely don't (unless I'm making fun of the English...)
My dialect does have a covert distinction between clear and dark /l/ even though it does not correspond with the usual clear and dark /l/ realizations of [l] and [5]. Rather, clear /l/ corresponds to a significant likelihood of being realized as [L\] or, if vocalized, generally being more closed as [M\], whereas dark /l/ corresponds to generally being vocalized as [M] or [U] (depending on the rounding of the preceding vowel). Note that this follows the same distribution of [l] and [5] that generally appears in dialects that have clear and dark realizations of /l/, including with respect to morphology (such as with "really" having a "dark" realization in my dialect but "freely" having a "light" realization in my dialect), even though the realizations themselves are different.
I have a few different allophones of /l/, but I don't really want to call them clear and dark, because a) they're both relatively dark, and b) they are distributed the same way as in RP, or in Travis's accent, for example. I have
[5] in most positions, but I also have a darker linking /l/ when a final l precedes a vowel (eg. fill up). I'm not sure exactly how to transcribe the second /l/, so I'll use [L]. Then of course there is the vocalized /l/, which is somewhere between [u] and [o] depending on the surrounding vowels. Unlike your 'dark l' (or the equivalent thereof), mine only occurs between words, so 'really' and 'freely' have the same allophone.
filling [fI5IN]
fill [fIo]
fill up [fIoLVp]
[5] in most positions, but I also have a darker linking /l/ when a final l precedes a vowel (eg. fill up). I'm not sure exactly how to transcribe the second /l/, so I'll use [L]. Then of course there is the vocalized /l/, which is somewhere between [u] and [o] depending on the surrounding vowels. Unlike your 'dark l' (or the equivalent thereof), mine only occurs between words, so 'really' and 'freely' have the same allophone.
filling [fI5IN]
fill [fIo]
fill up [fIoLVp]
>>fill up [fIoLVp]<<
I do have a phenomenon like this for syllabic /l/ between vowels, as while I normally realize syllabic /l/ as [M], I will usually realize it between vowels as [ML\] or at least [MM\]. However, in careful speech I may also realize it in other positions as [ML\] as well.
I do have a phenomenon like this for syllabic /l/ between vowels, as while I normally realize syllabic /l/ as [M], I will usually realize it between vowels as [ML\] or at least [MM\]. However, in careful speech I may also realize it in other positions as [ML\] as well.
I find that my normal /l/ isn't really as dark as most Americans'; I wrote it as [5] because it was a handy symbol. Maybe I should use [l_G] to show velarisation. Is there a symbol to show partial velarisation or the amount of velarisation?
>>I find that my normal /l/ isn't really as dark as most Americans'; I wrote it as [5] because it was a handy symbol. Maybe I should use [l_G] to show velarisation. Is there a symbol to show partial velarisation or the amount of velarisation?<<
Not that I know of, but considering that [5] is equivalent to [l_e], with _e referring to velarization *or* pharyngealization (at least according to Wikipedia, that is), using [l_G] could be a useful ad hoc method of showing such a difference.
Not that I know of, but considering that [5] is equivalent to [l_e], with _e referring to velarization *or* pharyngealization (at least according to Wikipedia, that is), using [l_G] could be a useful ad hoc method of showing such a difference.
Makes sense. Maybe I'll start doing that. I don't usually tend to use really narrow transcriptions though, except for the actual phonetic element in question. So I'll use [l] for both, since there's no phonemic distinction, unless we're specifically discussing realisations of /l/. I'm still not sure if my 'linking l' is really [L\] or not. (I meant [L\] above, not [L]. Though I do sometimes have [L] for intervocalic /lj/ sequences, like familiar [f@mIL@`].
The easy way to know whether it is [L\] or not is that if it is [L\] it is purely velar, being articulated solely with the dorsal region of the tongue and with no contact or even close proximity of the front portion of the tongue with the top of the mouth.
There's definitely tongue contact, so I suppose it's not [L\]. It seems strange to have two velarized allophones of /l/ though.
>>There's definitely tongue contact, so I suppose it's not [L\]. It seems strange to have two velarized allophones of /l/ though.<<
It is not really any more weird than my dialect having two allophones of /s/ which differ solely in their degree of laminalness, one having the usual level of laminalness for English [s] and the other having the level of laminalness typical for English [S] but being articulated at an alveolar rather than postalveolar location.
It is not really any more weird than my dialect having two allophones of /s/ which differ solely in their degree of laminalness, one having the usual level of laminalness for English [s] and the other having the level of laminalness typical for English [S] but being articulated at an alveolar rather than postalveolar location.
What's the distribution of your two /s/ allophones?
Anyway, I was wondering if you consistently make a distinction between /o/ and /ol/. I've found that while I may produce such a distinction, I can't really hear it in my own speech. I suppose with your unrounded l-vocalisation, it would be easier, but for me, 'code' and 'cold' are essentially homophones.
Anyway, I was wondering if you consistently make a distinction between /o/ and /ol/. I've found that while I may produce such a distinction, I can't really hear it in my own speech. I suppose with your unrounded l-vocalisation, it would be easier, but for me, 'code' and 'cold' are essentially homophones.
>>What's the distribution of your two /s/ allophones?<<
Here I will denote the less laminal (and more normal for English) allophone [s_m] and the more laminal (and [S]-like) allophone [s_m_m]. I have [s_m_m] for /s/ before /p/, /t/, /r/, /l/, /w/, /m/, and /n/ but not before /k/ or /tS/ (but usually /stS/ becomes [StS] or even [S:] in my dialect anyways unless I'm really carefully pronouncing it). In all other positions, aside from cases where voicing would occur, including before /k/, I have [s_m].
>>Anyway, I was wondering if you consistently make a distinction between /o/ and /ol/. I've found that while I may produce such a distinction, I can't really hear it in my own speech. I suppose with your unrounded l-vocalisation, it would be easier, but for me, 'code' and 'cold' are essentially homophones.<<
I make a consistent distinction in speech between /o/ and /ol/, which is largely aided by my /o/ normally being monophthongal in nature as [o] and because /ol/ does involve a (usually slight) degree of loss of rounding even though it is still rounded as [oU]. The matter is that my l-vocalization is unrounded by default but undergoes rounding assimilation to any preceding vowels - but often, especially in more careful speech, such rounding assimilation is not quite complete. Even when I do realize /o/ as a diphthong, it still normally has much shorter of a glide than than in [oU] from /ol/.
At the same time, in listening /o/ and /ol/ in the dialect here can be rather close to each other but generally are still distinguishable (as /ol/ commonly comes off as sounding as if it were /ow/ with a quite marked glide), especially in more careful/formal speech as in such there is more derounding than in more informal speech, where little to no derounding may occur in /ol/ relative to /o/. At the same time, if one makes the mistake of realizing /o/ as a diphthong with too strong of an offglide (akin to /o/ in some North American English dialects) it may be confused with /ol/; for instance, I remember one of my parents once mistaking "Boeing" for "bowling" because I placed too much of an epenthic [w] between the two vowels in it.
Here I will denote the less laminal (and more normal for English) allophone [s_m] and the more laminal (and [S]-like) allophone [s_m_m]. I have [s_m_m] for /s/ before /p/, /t/, /r/, /l/, /w/, /m/, and /n/ but not before /k/ or /tS/ (but usually /stS/ becomes [StS] or even [S:] in my dialect anyways unless I'm really carefully pronouncing it). In all other positions, aside from cases where voicing would occur, including before /k/, I have [s_m].
>>Anyway, I was wondering if you consistently make a distinction between /o/ and /ol/. I've found that while I may produce such a distinction, I can't really hear it in my own speech. I suppose with your unrounded l-vocalisation, it would be easier, but for me, 'code' and 'cold' are essentially homophones.<<
I make a consistent distinction in speech between /o/ and /ol/, which is largely aided by my /o/ normally being monophthongal in nature as [o] and because /ol/ does involve a (usually slight) degree of loss of rounding even though it is still rounded as [oU]. The matter is that my l-vocalization is unrounded by default but undergoes rounding assimilation to any preceding vowels - but often, especially in more careful speech, such rounding assimilation is not quite complete. Even when I do realize /o/ as a diphthong, it still normally has much shorter of a glide than than in [oU] from /ol/.
At the same time, in listening /o/ and /ol/ in the dialect here can be rather close to each other but generally are still distinguishable (as /ol/ commonly comes off as sounding as if it were /ow/ with a quite marked glide), especially in more careful/formal speech as in such there is more derounding than in more informal speech, where little to no derounding may occur in /ol/ relative to /o/. At the same time, if one makes the mistake of realizing /o/ as a diphthong with too strong of an offglide (akin to /o/ in some North American English dialects) it may be confused with /ol/; for instance, I remember one of my parents once mistaking "Boeing" for "bowling" because I placed too much of an epenthic [w] between the two vowels in it.
That's pretty much the same situation as me. I do produce a difference, but if you were to record me saying 'cold' and 'code', I might have a hard time telling them apart out of context.
However, IMD /U/ and /Ul/ are much harder to distinguish than /o/ and /ol/ in non-prevocalic positions as they really are only distinguished by /U/ being realized as [U_"] and /Ul/ being realized as the falling diphthong [U_"U_^]. Also, even though in my own idiolect /u/ and /ul/ are distinguished as [u] versus [uU_^], it seems that in my girlfriend's idiolect the two are practically merged in non-prevocalic positions as [u].