>>-But as far as I know "collar" is shifted to [A] not [a].-
Not true. COLLAR has [A], already, and it shifts to [a] and then to [æ]...
caller [Q]-->[A]-->[a]
collar [A]-->[a]-->[æ]
I guess you can hear all these in Chicago, black speakers don't seem to have any shift at all ( caller [Q], collar [A]), young white males have the 1st shift (caller [A], collar [a]), while young and middle aged middle class females have the most progresssive shift (caller [a], collar [æ])). So, misunderstandings can be a problem, since callar and collar pronunciations overlap in 67% of cases, you have to guess them from the context... So, to a foreign ear, there is a phonetic merger in the area, when all population is discussed although there is no merger on the single user level.<<
The dialect spoken in most of the Milwaukee area, AAVE aside, has a very clear NCVS, but is actually rather conservative in this regard. For instance, historical /ɔː/, that is, modern GA /ɒ(ː)/ remains rounded as [ɒ], as in "caller" [ˈkʰɒːʊ̯ʁ̩ː], while GA /ɑ(ː)/ is shifted to [a]. However, though, there are many phonetic environments in which historical /ɑː/ is not fronted much or is even backed relative to its position in most NAE dialects. For instance, I have "collar" [ˈkʰäːɯ̯ʁ̩ː], and even moreso I have a very clear back vowel in "car" [ˈkʰɑːʁ], which is actually further back than the realization of historical /ɑː/ in most NAE dialects without the NCVS, and which is actually often rounded to [ˈkʰɒːʁ] in many parts of the Milwaukee are, even though the dialect of the exact area where I grew up lacks such rounding.
Not true. COLLAR has [A], already, and it shifts to [a] and then to [æ]...
caller [Q]-->[A]-->[a]
collar [A]-->[a]-->[æ]
I guess you can hear all these in Chicago, black speakers don't seem to have any shift at all ( caller [Q], collar [A]), young white males have the 1st shift (caller [A], collar [a]), while young and middle aged middle class females have the most progresssive shift (caller [a], collar [æ])). So, misunderstandings can be a problem, since callar and collar pronunciations overlap in 67% of cases, you have to guess them from the context... So, to a foreign ear, there is a phonetic merger in the area, when all population is discussed although there is no merger on the single user level.<<
The dialect spoken in most of the Milwaukee area, AAVE aside, has a very clear NCVS, but is actually rather conservative in this regard. For instance, historical /ɔː/, that is, modern GA /ɒ(ː)/ remains rounded as [ɒ], as in "caller" [ˈkʰɒːʊ̯ʁ̩ː], while GA /ɑ(ː)/ is shifted to [a]. However, though, there are many phonetic environments in which historical /ɑː/ is not fronted much or is even backed relative to its position in most NAE dialects. For instance, I have "collar" [ˈkʰäːɯ̯ʁ̩ː], and even moreso I have a very clear back vowel in "car" [ˈkʰɑːʁ], which is actually further back than the realization of historical /ɑː/ in most NAE dialects without the NCVS, and which is actually often rounded to [ˈkʰɒːʁ] in many parts of the Milwaukee are, even though the dialect of the exact area where I grew up lacks such rounding.