13. The short- a and short- o conÞgurations
[Pages:16]13. The short-a and short-o configurations
Chapter 11 established a typology of North American dialects that is based on the phonological status of the two low short vowels /o/ and /?/. The chain shifts in progress are to a large extent governed, or even triggered, by the merger and/or split of /o/ and /oh/, /?/ and /?h/. Chapter 9 traced the ongoing merger of /o/ and /oh/. This chapter will present a brief account of the phonetic realizations of short-o and then proceed to the more complex topic of short-a configurations across the continent.
13.1. Short-o configurations
The historical short-o class of words is preserved in spelling as got, God, hop, cob, hock, hod, etc. To this must be added the class of words that are spelled with a and preceded by /w/ or /wh/. If these words end with velar consonants, they are realized as /?/ along with other short-a words: whack, whacky, wack, wag, waggle, wagon, wax, waxy, etc. If the spelling shows that the syllable was closed historically with a liquid, the vowel is realized as /oh/: wall, tall, call, talk, balk, caulk, stalk, wart, warm, wharf, warp. In the case of /r/, the coda can be simply /r/, as in war. Otherwise, words spelled historically with a and preceded by /w/ join the /o/ class: wasp, wander, Watson, wad, waddle, wadding, wan, and (variably) want, watch, wallet. Some of these, like watch and wallet, appear variably with /oh/ in dialects that maintain the /o/ ~ /oh distinction. Water commonly occurs with /oh/, but the alternation with /o/ is frequent and a matter of much social comment.
The formation of the /oh/ class and the opposition of /o/ and /oh/ was discussed in Chapter 9, along with the migration of a large number of /o/ words into the /oh/ class in North America. For dialects where /o/ merged with /oh/, there is no further history of /o/ to be discussed, since it then forms a part of a long and ingliding vowel that occurs in final as well as checked position. For those that do not merge /o/ with /oh/, the most common sequence is for /o/ to merge with /ah/. The /ah/ class has a small nucleus of native words: father, pa, ma, to which are added a large and variable class of loan words, the "foreign a" class (Boberg 1997): pasta, llama, macho, taco, Mazda, teriyaki, pajamas, etc. When /o/ is unrounded it normally merges with the /ah/ class, and becomes a low central peripheral vowel. Among the dialects where /o/ and /oh/ are distinguished, it is only in the South that this combined class remains a low back rounded short vowel in opposition to the long and ingliding vowel /oh/. As Chapter 18 will show, the differentiation of these two vowels is developed further as /oh/ shifts from a long and ingliding vowel to the back upgliding subsystem as /aw/.
The realization of the short-o class
Map 13.1 displays the realization of /o/ in both phonetic and phonemic terms. As 13.1 in Chapter 11, the green-oriented isogloss defines the region in which /o/ and /oh/ are merged. The red isogloss outlines the region in which /?/ is raised as a whole, with a mean value for non-nasal environments of less than 700 Hz. These red
symbols denote speakers for whom the /o/ ~ /oh/ distinction is preserved by the fronting of /o/, with an F2 greater than 1400 Hz. The red symbols fill the /?/-raising area, but extend beyond it westward to an area of eastern South Dakota and western Iowa, and also are heavily concentrated in western New England. Very few red symbols are found outside of these areas.
The blue symbols are speakers with back (and usually rounded) /o/, with F2 values of less than 1200 Hz, while the yellow symbols are the residual, or unmarked case of F2 greater than 1200 Hz but less than 1400 Hz. It is the distribution of the blue symbols that is noteworthy here. There are none in the Inland North. The heaviest concentrations are to be found in the low back merger region, especially in western Pennsylvania and Canada, where the merger is realized in low back rounded position. Table 1 shows the extent to which the back position of /o/ is associated either with the low back merger, or the diphthongization of /oh/. Only six blue tokens are found in other regions.
Table 13.1. Distribution of low back rounded forms of /o/
Area of low back merger Area of back upglide chain shift /oh/ g /aw/ g Other
No. speakers rounded vowel
136 62 242
Low back Percent
37
27%
12
19%
6
2%
In the initial position of Chapter 2, the /ah/ class of father, spa, pa, garage, Ba- 13.2 hamas, etc. is presented as distinct from /o/, although for the great majority of speakers, these two classes are firmly merged. Map 13.2 identifies those areas in which the two classes are distinct. The data here are somewhat sparse; /ah/ is not frequent in spontaneous speech, and no minimal pairs were elicited in the Telsur interview. Such pairs are rare and marginal (e.g. bomb ~ balm for the diminishing number who do not pronounce the /l/ in balm). We therefore have data on /ah/ and its relation to /o/ for only 310 of the 439 Telsur speakers who were analyzed acoustically.1
For the great majority of the speakers ? shown as yellow symbols on Map 13.2 ? there is no evidence of a phonemic difference between /o/ and /ah/. For those within the green-oriented isogloss, /ah/ forms part of the low back merger, so that the vowels of pa, paw, pot, pod, pawed, taught, form a single phoneme. For those outside that isogloss, where /o/ is generally distinct from /oh/, /o/ is unrounded as a rule and merges with /ah/, so that pa, pot, pod is opposed to paw, pawed, taught. The distinction between /ah/ and /o/ is maintained in only two narrowly circumscribed areas on Map 13.2: the area of Eastern New England centered on Boston, and some speakers in New York City. Though these areas are small geographically, they include large populations.
1 Map 32 of Kurath and McDavid (1961) gives phonetic realizations of the word father. Though there is a considerable variety of forms, they correspond quite closely to the vowel of /o/ except for Eastern New England.
168 The short-a and short-o configurations
Anchorage
Vancouver
Edmonton Calgary
Saskatoon
St. John's
Seattle
Spokane
Portland
Walla Walla
Eugene
Missoulas
Sacramento
Reno
San Francisco Fresno
Boise
Idaho Falls
Ogden Salt Lake City
Provo
Orem
Bakersfield
Las Vegas
Los Angeles
Flagstaff
San Diego
Phoenix Tucson
F2(o)
> 1400 Hz < 1400, > 1200 Hz < 1200 Hz F1(?) < 700 /o/ = /oh/ /oh/ back upglides
Regina
Winnipeg
St. John
Great Falls
Thunder Bay
Brockway
Minot
Chisholm
Montreal
Bangor
Billings
Bismark
Bemidji
Duluth
SSMarie
Burlington
Marquette
Ottawa
Rutland
Portland
Casper
Colorado Springs
Santa Fe
Lemmon
Aberdeen
Rapid City
Redfield Sioux Falls
Scotts Bluff No.Platte
Norfolk Omaha
Denver
Lincoln
GardenCity
Topeka Hays
Wichita
Eau Claire
Toronto
Syracuse Albany
Manchester Boston
Minn/ St, Paul St.James
Rochester
Sioux City
Mason City
Madison
Green Bay
London
Milwaukee
Grand
Flint
Rapids
De-
Windsor
Chi-
troit
Cleveland
cago
Rochester Buffalo
Scranton Erie
Pitts-
Hartford
Providence New York City
Cedar Rapids
Gary
Toledo
burgh Harrisburg
Trenton Philadelphia
Des Moines St. Joseph
Peoria
Indianapolis
Columbus Dayton
Canton
Cincin-
Clarksburg
Washington
Wilmington. Baltmore
Kansas City
Sprngfield
Terre Haute
nati Lexington
Richmond
Kansas City
Columbia
Louisville
Roanoke
Norfolk
Springfield
St.Louis
Evansville
Win-Salem
Durham
Toronto
Greenville
Tulsa
Nashville
Knoxville
Charlotte
Raleigh
Amarillo
Memphis
Chattanooga
Ashville
Fayettevlle
Wilmington
Albuquerque
Lubbock
Oklahoma City
Little Rock
New Albany
Birmingham
Greenvllle
Atlanta
Columbia
Charleston
Abilene
Dallas
Shreveport
Jackson
Montgomery
Macon
Savannah
El Paso
Odessa
San Antonio
Austin
Houston
Baton Rouge New Orleans
Mobile
Tallahassee
Jacksonville
Orlando
Corpus Christi
Tampa Miami
Sydney Halifax
Map 13.1. Front?back position and phonemic status of /o/ in North American dialects
The position of the word-class /o/ on the front?back dimension is closely connected with the various chain shifts that differentiate North American dialects. The red circles are the speakers with mean values of /o/ that are clearly in central (or even front) position. Red circles are dominant inside the red isogloss, which
marks the general raising of (non-nasal) /?/; this shows the close connection between these two sound changes. The fronting of /o/ extends to the east and west of the red isogloss, but not north or south of it. It is well dissociated from the area where /o/ and /oh/ merge, within the green oriented isogloss.
Short-o configurations 169
Winnipeg
F2(ah) - F2(o) > 200 Hz, F1(o) - F1(ah) > 200 Hz,
F2(o) - F2(ah) > 200 Hz /ah/ = /o/ The North /o/ = /oh/ E.N.E. area NYC
St. John
Minot
Chisholm
Thunder Bay
Montreal
Bangor
Bismark
Bemidji
Duluth
Marquette
SSMarie
Arnprior
Ottawa
Burlington
Waterville Portland
Rutland
Concord
Aberdeen
Wilmar
Minn/ St, Paul
Eau Claire
Oconto
Toronto
Stevens Pt
Syracuse
Albany
.Platte
Redfield
Watertown Sioux Falls
St.James
Rochester
Norfolk Omaha
Sioux City Denison
Mason City
Cedar Rapids
Des Moines
Lincoln Falls City
St. Joseph
Minneapolis
Appleton
Reedsburg Madison
Green Bay Milwaukee
Saginaw
Grand Flint Rapids
Flint
London
Rochester Buffalo
Binghamton
Hartford
Monroe
Kenosha
Detroit Kalamazoo
Windsor Erie
Scranton
Scranton
Quad Cities
Lena
Rockford Elgin
Sterling
Joliet
Chicago Gary
So. Bend
Ft. Wayne
Peoria
Muncie Indianapolis
Cleveland
AA Toledo
Youngstown Akron
Wmsport State College
Harrisburg
Mansfield Columbus
Canton Massillon
Pittsburgh
Dayton
Washington
Allentown
Trenton
Reading
Philadelphia
Wilmington. Baltmore
Sprngfield
Champaign
DC
Hamilton
Clarksburg
Hays
Topeka
Kansas City
St.Louis
Decatur
Terre Ashland Haute
Cincinnati
Huntington Lexington
Charleston
Richmond
Evansville
City
Columbia
Louisville
Kansas City
St.Louis
Roanoke
Norfolk
Manchester Boston
-&.+&/
Providence
New York City
!"# %&'( )*+,
Wichita
Toronto
Springfield
Knoxville
Win-Salem
Durham
Greenville
Raleigh
Map 13.2. Relations of /o/ and /ah/ in North America
In one region of North America, /o/ in bother and /ah/ in father are kept apart by the relatively front position of /ah/: in Eastern New England. In New York City, many speakers also retain the distinction between these two word classes, but here father is further back and higher than bother. In the rest of the country, /o/ of
bother, got, rock, Don, etc. is not clearly distinct from /ah/ of father, pa, spa, etc. This merged phoneme is best noted as /o/, to retain comparability of word classes across dialects.
170 The short-a and short-o configurations
The criterion here is a simple acoustic one, based on F2 measurements. It follows that some speakers, /ah/ is front of /o/. It has been found in a number of studies that a minimum of 200 Hz is needed to maintain a stable distinction on the F2 axis (LYS; Labov 1994: Ch. 12). For /o/ and /ah/ to be phonemically distinct in Map 13.2, the difference in the mean values of F2 must be greater than 200 Hz. This condition holds for only seven of the 310 speakers. Five are concentrated in the Boston area, where only one of the Telsur speakers fails to meet the criterion. The sixth is an isolated person in Lexington, Kentucky. The seventh (not shown in Map 13.2) is in Corpus Christi. Discounting these isolated points, it appears clear that /ah/ is maintained front of /o/ only in Eastern New England.
Figure 13.1 shows the distinction between the /ah/ and /o/ classes in the vowel system of Denise L., 21, of Boston. Figure 13.1a displays the clear separation of /ah/ and /o/. The four /ah/ tokens are well to the front of the 18 /o/ tokens, with
F2
2600 400
2400
2200
2000
1800
1600
1400
1200
1000
800
500
/o/ /ah/
m1 m2
F1 814 846
F2 1182 1486
N
18
4
m1-m2 t -32 -2.36
-304 -5.45 d.f.20
P ................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.