"Can" and "cam" has a different-sounding vowel from "cat" and "cap". Is this common? Do all North American dialects do this?
can and cat
(First of all, see http://en.wikipedia.org/wiki/IPA .)
Yes, it's common for the vowel phoneme /æ/ to have a different realization before the nasal consonants [n], [m]. This can range from a slight lengthening [æ:] or diphthongizaiton [æə] to a stronger diphthongization [ɛə].
But /æ/ before the nasal consonant is a different matter: for a lot of Americans, this vowel has turned into /e:/, as in "face": [eɪŋ].
Yes, it's common for the vowel phoneme /æ/ to have a different realization before the nasal consonants [n], [m]. This can range from a slight lengthening [æ:] or diphthongizaiton [æə] to a stronger diphthongization [ɛə].
But /æ/ before the nasal consonant is a different matter: for a lot of Americans, this vowel has turned into /e:/, as in "face": [eɪŋ].
How different do you think "can" and "cam" are from "cad" and "cab"? The vowel preceeding unvoiced consonants is slightly shorter than those preceeding voiced consonants.
That think that happens in front of other consonants too, but it depends on the dialect. I think I heard "gas" as something like /gɛəs/ several times, for example.
<<How different do you think "can" and "cam" are from "cad" and "cab"? The vowel preceeding unvoiced consonants is slightly shorter than those preceeding voiced consonants.>>
Well, /n/, /m/, /d/ and /b/ are all voiced consonants - so the comparison would be between "cad, cab" and "cat, cap". But in most North American dialects, this allophony wouldn't be very significant - that is, no more noticeable than similar allophony between "bed, bet", for example.
<<That think that happens in front of other consonants too, but it depends on the dialect. I think I heard "gas" as something like /gɛəs/ several times, for example.>>
Real phonemic splits have occurred in the New York-New Jersey and Philadelphia dialects, leading to a distinct /E@/ being used in many words ("gas" may be among them - I'm not sure). And aside from that, the Northern Cities Vowel Shift includes a general change of /{/ to a diphthong like [E{], [E@] or [I@].
Well, /n/, /m/, /d/ and /b/ are all voiced consonants - so the comparison would be between "cad, cab" and "cat, cap". But in most North American dialects, this allophony wouldn't be very significant - that is, no more noticeable than similar allophony between "bed, bet", for example.
<<That think that happens in front of other consonants too, but it depends on the dialect. I think I heard "gas" as something like /gɛəs/ several times, for example.>>
Real phonemic splits have occurred in the New York-New Jersey and Philadelphia dialects, leading to a distinct /E@/ being used in many words ("gas" may be among them - I'm not sure). And aside from that, the Northern Cities Vowel Shift includes a general change of /{/ to a diphthong like [E{], [E@] or [I@].
>>Well, /n/, /m/, /d/ and /b/ are all voiced consonants - so the comparison would be between "cad, cab" and "cat, cap". But in most North American dialects, this allophony wouldn't be very significant - that is, no more noticeable than similar allophony between "bed, bet", for example.<<
Hmm... to me at least, I've found the vowel length to be almost more important than the actual voicing in many cases. One example that stuck out to me recently was the pronunciation of "Jose", which canonically is [ˈhoːze̞ː] here, even though it "should" have [s] rather than [z]... However, for me at least, [ˈhose̞ː] just does not work, as the vowel to me is [oː] not [o] no matter what, with [o] just sounding wrong intuitively to me. The best pronunciation that "works" for me seems to actually be [ˈhoːz̥e̞ː], because vowel length seems to actually be more salient than actual voicing, with optional devoicing of obstruents being possible in all positions IMD, while actual fortisness/lenisness is more tied to vowel length except word-finally IMD (hence [ˈhoːz̥e̞ː] not [ˈhoːse̞ː])
Hmm... to me at least, I've found the vowel length to be almost more important than the actual voicing in many cases. One example that stuck out to me recently was the pronunciation of "Jose", which canonically is [ˈhoːze̞ː] here, even though it "should" have [s] rather than [z]... However, for me at least, [ˈhose̞ː] just does not work, as the vowel to me is [oː] not [o] no matter what, with [o] just sounding wrong intuitively to me. The best pronunciation that "works" for me seems to actually be [ˈhoːz̥e̞ː], because vowel length seems to actually be more salient than actual voicing, with optional devoicing of obstruents being possible in all positions IMD, while actual fortisness/lenisness is more tied to vowel length except word-finally IMD (hence [ˈhoːz̥e̞ː] not [ˈhoːse̞ː])