Low Central Merger

Milton   Fri Sep 05, 2008 12:34 pm GMT
cot [kät]
caught [kät]

(IPA [ä] - low central unrounded vowel)


I think Cot-Caught merger is a better name since the merged vowel can vary a lot (you can have low central [unrounded] merger, low back [unrounded/rounded] merger or raised back [rounded] merger )

'' The relative fronting and backing of /oh/ in caught, cause, law, etc.
where /o/ and /oh/ are merged, like Tucson, Arizona, and St. Johnʼs, Newfoundland, we find red symbols, indicating that the merger takes place at a relatively front position. ''

Chapter 10
Map 10.32 (page 107 in PDF, page 109 in the printed form)
ANAE by prof. Labov
Milton   Fri Sep 05, 2008 12:36 pm GMT
Interesting reading:

STABILITY AND CHANGE ALONG A DIALECT BOUNDARY:
THE LOW VOWELS OF SOUTHEASTERN NEW ENGLAND

http://www.ling.upenn.edu/~johnson4/dej2007b.pdf
Guest   Sat Sep 06, 2008 11:15 pm GMT
Nice to know.
Lazar   Sat Sep 06, 2008 11:46 pm GMT
I disagree; I think "low-back merger" is a better name for the merger commonly found in North America. The reason is that "cot-caught" merger, like "father-bother" merger, merely implies a merging of two lexical classes: just as General American is father-bother merged while remaining cot-caught (i.e. low back) merged, Eastern New England (my own native dialect) is cot-caught merged while remaining father-bother (i.e. low back) merged:

GA

father ["fAD@`]
bother ["bAD@`]

cot ["kAt]
caught ["kQt]

ENE

father ["fAD@`]
bother ["bQD@`]

cot ["kQt]
caught ["kQt]

The low-back merger found in Pennsylvania, the Western US and Canada involves both of these mergers - a full merging of the historical /A:/, /Q/ and /O:/ phonemes. So I think using "cot-caught merger" as a proxy name for the low back merger - especially considering that there are speakers in North America, and elsewhere in the Anglophone world (e.g. Scotland), who are cot-caught merged but father-bother unmerged - doesn't make sense. Often, when people refer to the cot-caught merger, they'll lump Eastern New England together with the Western US, just because they both happen not to distinguish between the LOT and THOUGHT lexical sets, but I think that Eastern New England has more in common with the unmerged parts of the US, because they both maintain two phonemes /A/ (as [A] or [a]) and /O/ (as [O] or [Q]) that are merged in the West.
Lazar   Sat Sep 06, 2008 11:50 pm GMT
Sorry, typos:

<<just as General American is father-bother merged while remaining cot-caught (i.e. low back) unmerged, Eastern New England (my own native dialect) is cot-caught merged while remaining father-bother (i.e. low back) unmerged:>>

And I would add, about the name, I think "low-back merger" is fine because the three phonemes involved, historical /A:/, /Q/, /O:/, are in the low back region. It's really a matter of phonemes, so I think the exact phonetic realizations don't matter, but in any case I think it's more common to have a near-back realization (and sometimes a rounded realization too) than to have a fully central [a].
Guest   Sun Sep 07, 2008 12:20 am GMT
there's Mary ~ merry ~ marry merger
and father ~ bother ~ daughter merger
it's better to call it like that
Guest   Sun Sep 07, 2008 12:35 am GMT
-cot ["kAt]
caught ["kQt]-

Most dictionaries that transcribe General American using IPA don't agree on the caught vowel; they use [kOt] .

(O=Open-mid back rounded vowel
http://en.wikipedia.org/wiki/Open-mid_back_rounded_vowel )
Guest   Sun Sep 07, 2008 4:37 am GMT
>> Most dictionaries that transcribe General American using IPA don't agree on the caught vowel; they use [kOt] . <<

I've never seen a dictionary that used [ ]. Dictionaries use phonemic transcription, not phonetic transcription, so the exact vowel is not important.
Lazar   Sun Sep 07, 2008 4:58 am GMT
The last Guest is right; but as for the phonetics, General American (if it's low-back distinguishing) usually uses a rather open realization between [Q] and [O]. If you compare GA "caught" to RP "caught" or to GA "court", you'll find that there's a significant qualitative difference.

(In fact, I think you can even make a case for transcribing the General American vowel, phonemically, as /Q:/, because it's possible for AmEng speakers to make contrasts like "Laura" ["lQ:r\@] versus "Torah" ["t_hO@`@]. You could also solve this problem by positing a distinct rhotic phoneme /O@`/.)

For what it's worth, Longman uses /Q:/ for General American ( http://www.ldoceonline.com/howtouse.html ); I have seen some other sources use [Q] as well. If you listen to the recordings at m-w.com ( http://www.merriam-webster.com/dictionary/caught ), you'll find that they use quite open realizations for /O:/ (or /Q:/).
Guest   Sun Sep 07, 2008 4:59 am GMT
Phonetic transcription is important for language learners. When you learn French, you need to know the exact sound value, unless you want to sound accented/foreign. The same thing in English.