''bath'' and ''Mary''

Nigel   Mon May 29, 2006 8:59 am GMT
>>Long a is not universally used in Australia. Take the word DANCE as an example. In many regions it is pronounced like in the USA and not like in London.<<

Not merely in "many regions"; it is almost universal in Australia. You will hear "dahnce" only in South Australia, and in a few circles in which affected speech is used.

My point was that when the broad A <IS> used in Australia, it is the same broad A used in RP.

Jim will undoubtedly disagree, as usual, but in a pleasant cockroachy way.
Kirk   Mon May 29, 2006 9:09 am GMT
<<My point was that when the broad A <IS> used in Australia, it is the same broad A used in RP. >>

Well strictly speaking that's still not the case because the vowel quality is different. For instance, RP "bath" is [bA:T] while in General Australian it's [b6:T], I believe. If I'm not mistaken then for Australians with 'broad A' that means that "dance" and "dunce" are differentiated only by vowel length. They'd be:

"dance" [d6:n(t)s]
"dunce" [d6n(t)s]

Compare to RP where the vowel qualities are different as well as the length. It's a fundamentally different distinction:

"dance" [dA:n(t)s]
"dunce" [dVn(t)s]
Adam   Mon May 29, 2006 9:25 am GMT
Terms like "vowel quality" are just another excuse for poor English speakers to justify their sloppiness.
Aussie bleeder   Mon May 29, 2006 9:47 am GMT
Nigel,

I agree with Kirk. In comparing "ah" or "ar" and "uh" in AusE, contrastive vowel lengths are of primary importance. Vowel quality in this case is far less significant.

In some RP accents, the [A] in "bath" can be extreme to the point where it seems to approach the [O] vowel of "hot" in AusE. You might think you're hearing "bohth" for the word "bath" (think of Tony Greig, the South African cricket commentator).
IMO, in Aussie English, the vowel in "bath" can range from [6] to [a]. The [6] vowel sound seems to be more prevalent in broader Aussie accents, while the [a] sound can give the impression of the [A] in RP. Maybe that's how you're perceiving it.
Ed   Thu Jun 01, 2006 1:15 am GMT
To me, the first syllable of "mary" sounds the same as "air", "mary" rhymes with "airy" and "fairy".
Travis   Mon Jun 05, 2006 10:47 pm GMT
The problem is that mjd is an absentee moderator who isn't around here to delete this goddamn shit that has been spammed to no end onto this forum.
Kirk   Tue Jun 06, 2006 3:42 am GMT
<<Actually, Travis, I'm here everyday and delete them everyday.

Why bother paying attention to these people. The posts are only going to be gone when I get to them each evening.>>

Thanks for deleting those. But I do have to wonder about this particular thread as it's been saddled with those multilink messages for a long time. Does that spammer just keep on posting the same messages every night after you delete them?

<<To me, the first syllable of "mary" sounds the same as "air", "mary" rhymes with "airy" and "fairy".>>

I think it's that way for many people regardless of how else they're merged. For instance I'm "Mary-merry-marry" merged so of course "airy" and "fairy" also have the same vowel as "Mary." I assume that for you it's something like ["mE@r\I] (which would be the RP pronunciation)--is that right? For me "Mary" is ["mEr\i], identical to the vowel I pronounce in "airy" "fairy," "carry" and "Kerry."
Lazar   Tue Jun 06, 2006 3:56 am GMT
<<I think it's that way for many people regardless of how else they're merged.>>

Yeah, I think what Ed wrote would apply to all English dialects. For me,

Mary - ["mE@`i]
air - ["E@`]
airy - ["E@`i]
fairy - ["fE@`i]

But since I'm 3M-unmerged,

carry - ["k_h{r\i]
Kerry - ["k_hEr\i]
Kirk   Tue Jun 06, 2006 4:01 am GMT
<<Yeah, I think what Ed wrote would apply to all English dialects. For me,

Mary - ["mE@`i]
air - ["E@`]
airy - ["E@`i]
fairy - ["fE@`i]

But since I'm 3M-unmerged,

carry - ["k_h{r\i]
Kerry - ["k_hEr\i]>>

Hehe, I think I'm gonna hafta start using the term "3M" as long as it's clear in context (wouldn't wanna dip any further into linguistic jargon!). It goes along well with other online ling forum terms such as "c-c" merger, "IMD" or "fauxnetic."
Lazar   Tue Jun 06, 2006 4:05 am GMT
I perceive the fundamental difference between pre-rhotic-merging and pre-rhotic unmerging accents to be that unmerging accents have sequences of V+/r\/ that can't occur word-finally. Compare,

hurry - ["hV.r\i]
furry - ["f3`.i]

Sirius - ["SI.r\i.@s]
serious - ["SI@`.i.@s]

merry - ["mE.r\i]
marry - ["m{.r\i]
Mary - ["mE@`.i]

For me, the first syllables of "furry", "serious", and "Mary" can occur as words (eg, "fur", "sear", "mare"), but the first syllables of "hurry", "Sirius", "merry", or "marry" can't.

On a phonemic level, you could interpret my pre-rhotic distinctions as a matter of syllabification.

hurry - /"hV.ri/
furry - /"fVr.i/

Sirius - /"SI.ri.@s/
serious - /"SIr.i.@s/

merry - /"mE.ri/
marry - /"m{.ri/
Mary - /"mEr.i/
Father Jeff   Tue Jun 06, 2006 7:36 am GMT
Two nuns in a bath, one says to the other "Where's the soap?" the other replies "Yes it does doesn't it".

Joke doesn't quite work if "where's" and "wears" aren't homophones when you say them!
Kirk   Tue Jun 06, 2006 8:17 am GMT
<<On a phonemic level, you could interpret my pre-rhotic distinctions as a matter of syllabification. >>

Yes. Your examples are interesting since they're a lasting effect of a historical New England nonrhotacism (since it's especially nonrhotic accents that make all those distinctions) which I would argue is a substrate influence upon your speech even considering your rhotic speech (also, take for example, your occasional 'intrusive r' you've mentioned). You're the first fully rhotic generation in your family, right?
Lazar   Tue Jun 06, 2006 11:00 pm GMT
<<Yes. Your examples are interesting since they're a lasting effect of a historical New England nonrhotacism (since it's especially nonrhotic accents that make all those distinctions) which I would argue is a substrate influence upon your speech even considering your rhotic speech (also, take for example, your occasional 'intrusive r' you've mentioned).>>

Yep, many of the distinctive features of my speech are things strongly associated with non-rhotacism, even though the actual non-rhotacism itself has "worn out" in my case.

It's often occurred to me that my own accent is, as it were, "a rhotic accent that was built as if it should be non-rhotic". For example, the impetus for having the intrusive R was the neutralization of phonemic distinctions such as /@r/ ~ /@/ at the end of words. Take the example of "fetter" vs "feta". If both of these words are pronounced the same in isolation, and if "fetter" receives a "linking R" when it's followed by a vowel, then why shouldn't "feta" receive such an R as well? It was an instance of simplification; if two words are pronounced identically in one context, then there is a pressure for them to be pronounced identically in all contexts, with identical rules of allophony. But since I'm rhotic, I of course pronounce "fetter" and "feta" differently. The original impetus for having the intrusive R has disappeared from my speech, yet I have the intrusive R nonetheless.

There's another element of my dialect which I think might reflect my "built as if it were non-rhotic" observation above, but I'm less sure on this one. The element in question is my CC-merged, FB-unmerged low vowel system. I'm not at all certain on this, and it's really just speculation on my part, but I think that non-rhotacism might have played a part in the development of this particular vowel system. I will present my thoughts on this:

You see, in traditional CC-unmerged American dialects, you have a system where historical /A/ and most cases of historical /Q/ become /A/, while historical /O/ and some cases of historical /Q/ become /O/. This results in a relatively "even" distribution of the phonemes /A/ and /O/, with each phoneme having a large inventory of words that use it.

But in my idiolect, in which the distinction between /A/~/Ar/ is maintained, and in which *all* cases of historical /Q/ and /O/ have merged into /Q/, there aren't really that many instances of "plain /A/" (ie, /A/ unaccompanied by /r/). The "plain /A/" words, for me, are pretty much "father", "aunt", the "-alm" words, and some foreign names and loanwords. On the other hand, the "/Q/ words" in my dialect are extremely numerous, including everything in the "cot", "cloth", and "caught" classes. In my idiolect you have:

- a fair number of /Ar/ words
- a few sporadic instances of plain /A/
- a huge number of /Q/ words

Most of the occurrences of /A/ in my idiolect are in the form of /Ar/ ("park", "car", "are", you name it). There is very little contrastive information riding on the preservation of an /A/ phoneme that is distinct from the sequence /Ar/. Thus, my ratio of "ah" words to "aw" words is extremely unbalanced. You could say that "ah" (ie, "plain /A/") isn't really put to much use in my dialect.

But in a non-rhotic Massachusetts accent, in which /A/ has merged with /Ar/, and in which the "bath" words have in many cases shifted to /A/, then you have a much more even distribution of /A/ vs /Q/. Thus I speculate that my low vowel system may be "designed" for a non-rhotic accent. When you take the non-rhotic New England /A/ and apply rhotacism to it, then most instances of it become /Ar/, leaving a few sporadic instances of plain old /A/.

Maybe I could solidify my speculation above and write a thesis paper in which I argue that GA rhotacism and NE non-rhotacism have influenced the development of their respective low vowel systems. ;-)

<<You're the first fully rhotic generation in your family, right?>>

Yep. My mother is variably non-rhotic, but she has told me that as a child she was completely non-rhotic. My father is almost completely rhotic now, but I'm pretty sure that he was either partially or completely non-rhotic as a child.
Kirk   Wed Jun 07, 2006 12:57 am GMT
<<Yep. My mother is variably non-rhotic, but she has told me that as a child she was completely non-rhotic. My father is almost completely rhotic now, but I'm pretty sure that he was either partially or completely non-rhotic as a child.>>

Interesting. I wonder when or where they acquired their rhotacism. Was it encouraged at school (since nonrhotic accents there are by this point generally perceived as working-class and less-educated) or has everyone just changed with the times? Or was it due to the move out to the closer-to-General-American suburbs (assuming they grew up in the actual city of Boston)? That kind of a drastic change in an accent in an individual (especially one that never left the region or anything) is fascinating to me. I'm positive nothing as drastic as that has occurred in the speech of my parents compared to a few decades ago when they were younger.

<<Yep, many of the distinctive features of my speech are things strongly associated with non-rhotacism, even though the actual non-rhotacism itself has "worn out" in my case.

It's often occurred to me that my own accent is, as it were, "a rhotic accent that was built as if it should be non-rhotic".>>

Yeah, I'd say there's a clear nonrhotic substratum framework upon which rhotacism has just recently been laid.

<<For example, the impetus for having the intrusive R was the neutralization of phonemic distinctions such as /@r/ ~ /@/ at the end of words. Take the example of "fetter" vs "feta". If both of these words are pronounced the same in isolation, and if "fetter" receives a "linking R" when it's followed by a vowel, then why shouldn't "feta" receive such an R as well? It was an instance of simplification; if two words are pronounced identically in one context, then there is a pressure for them to be pronounced identically in all contexts, with identical rules of allophony. But since I'm rhotic, I of course pronounce "fetter" and "feta" differently. The original impetus for having the intrusive R has disappeared from my speech, yet I have the intrusive R nonetheless. >>

Yeah, that makes sense. On another note, do you really have [E] in "feta?" I've always had [e(I)] for "feta" as with my "beta" "ate a" or "theta."

<<There's another element of my dialect which I think might reflect my "built as if it were non-rhotic" observation above, but I'm less sure on this one. The element in question is my CC-merged, FB-unmerged low vowel system. I'm not at all certain on this, and it's really just speculation on my part, but I think that non-rhotacism might have played a part in the development of this particular vowel system. I will present my thoughts on this:

You see, in traditional CC-unmerged American dialects, you have a system where historical /A/ and most cases of historical /Q/ become /A/, while historical /O/ and some cases of historical /Q/ become /O/. This results in a relatively "even" distribution of the phonemes /A/ and /O/, with each phoneme having a large inventory of words that use it.

But in my idiolect, in which the distinction between /A/~/Ar/ is maintained, and in which *all* cases of historical /Q/ and /O/ have merged into /Q/, there aren't really that many instances of "plain /A/" (ie, /A/ unaccompanied by /r/). The "plain /A/" words, for me, are pretty much "father", "aunt", the "-alm" words, and some foreign names and loanwords. On the other hand, the "/Q/ words" in my dialect are extremely numerous, including everything in the "cot", "cloth", and "caught" classes. In my idiolect you have:

- a fair number of /Ar/ words
- a few sporadic instances of plain /A/
- a huge number of /Q/ words

Most of the occurrences of /A/ in my idiolect are in the form of /Ar/ ("park", "car", "are", you name it). There is very little contrastive information riding on the preservation of an /A/ phoneme that is distinct from the sequence /Ar/. Thus, my ratio of "ah" words to "aw" words is extremely unbalanced. You could say that "ah" (ie, "plain /A/") isn't really put to much use in my dialect.

But in a non-rhotic Massachusetts accent, in which /A/ has merged with /Ar/, and in which the "bath" words have in many cases shifted to /A/, then you have a much more even distribution of /A/ vs /Q/. Thus I speculate that my low vowel system may be "designed" for a non-rhotic accent. When you take the non-rhotic New England /A/ and apply rhotacism to it, then most instances of it become /Ar/, leaving a few sporadic instances of plain old /A/.>>

Yeah your vowels are still built upon a nonrhotic system's vowels. I wonder what this asymmetry and thus the (relative) instability of the vowel system you and others like you there have will mean for the future of the system. I wouldn't be surprised if things got shifted around a bit in a generation or two to balance things out. For instance, looking at other accents even the absence or asymmetry of one vowel in a system may prompt a full-scale vowel shift. Many linguists studying vowel shifts have theorized that ultimatey the "c-c" merger sparked such shifts as the California Vowel Shift and the similar (and nascent) Canadian Vowel Shift due to the sudden absence of a vowel which other vowels have been shifting around towards to balance the space. This doesn't mean such shifts are inevitable but mergers such as the "c-c" one may mean subsequent shifts become more likely.

<<Maybe I could solidify my speculation above and write a thesis paper in which I argue that GA rhotacism and NE non-rhotacism have influenced the development of their respective low vowel systems. ;-) >>

That'd be awesome.
Lazar   Wed Jun 07, 2006 1:48 am GMT
<<Interesting. I wonder when or where they acquired their rhotacism. Was it encouraged at school (since nonrhotic accents there are by this point generally perceived as working-class and less-educated) or has everyone just changed with the times? Or was it due to the move out to the closer-to-General-American suburbs (assuming they grew up in the actual city of Boston)? That kind of a drastic change in an accent in an individual (especially one that never left the region or anything) is fascinating to me. I'm positive nothing as drastic as that has occurred in the speech of my parents compared to a few decades ago when they were younger.>>

Well remember, the city was Worcester in this case. ;-) They did both grow up within the city of Worcester. I think it's more of a changing-with-the-times thing, as rhotacism made more and more inroads through national media and interstate migration. From what my mother has told me, I don't think non-rhotacism was really discouraged when she went to school. I get the impression that when she grew up, nearly everyone in an urban school (even including the teachers) would have probably been non-rhotic. She's told me that she became more rhotic during her professional life, in her 20s and 30s. (I haven't had as many linguistic discussions with my father, but I get the impression that his story is pretty much the same.) Another thing that has changed in my mother's idiolect is her maintenance of the trap-bath split. She's told me that when she was a child, she clearly remembers saying things like [baT] and [p_haT] (for "bath" and "path") when talking with her friends. Now she would usually say [b{T] and [p_h{T].

And an additional factor is that some years ago they moved to the suburbs, which are noticeably more rhotic than the urban areas.