New England phonology*
Naomi Nagy and Julie Roberts
1.
Introduction
The six states that make up New England (NE) are Vermont (VT), New Hampshire
(NH), Maine (ME), Massachusetts (MA), Connecticut (CT), and Rhode Island
(RI). Cases where speakers in these states exhibit differences from other American
speakers and from each other will be discussed in this chapter. The major sources
of phonological information regarding NE dialects are the Linguistic Atlas of New
England ((LANE)
LANE) (Kurath 1939-43), and Kurath (1961), representing speech patterns from the first half of the 20th century; and Labov, Ash and Boberg, (fc);
Boberg (2001); Nagy, Roberts and Boberg (2000); Cassidy (1985) and Thomas
(2001) describing more recent stages of the dialects.
There is a split between eastern and western NE, and a north-south split within
eastern NE. Eastern New England (ENE) comprises Maine (ME), New Hampshire (NH), eastern Massachusetts (MA), eastern Connecticut (CT) and Rhode Island (RI). Western New England (WNE) is made up of Vermont, and western MA
and CT. The lines of division are illustrated in figure 1. Two major New England
shibboleths are the “dropping” of post-vocalic r (as in [ka:] car and [ba:n] barn)
and the low central vowel [a] in the BATH class, words like aunt and glass (Carver
1987: 21). It is not surprising that these two features are among the most famous
dialect phenomena in the region, as both are characteristic of the “Boston accent,”
and Boston, as we discuss below, is the major urban center of the area. However,
neither pattern is found across all of New England, nor are they all there is to the
well-known dialect group. We present a brief description of the settlement of the
region as a whole and give examples of past and current pronunciation patterns
to illustrate both how New England differs from the rest of the country and what
region-internal differences exist. The material is rather thin in some areas, due to
a dearth of recent research on New England English. Nevertheless, the resulting
pattern is one that reflects the richness and diversity of the region itself.
2.
European settlement of New England
Our story begins with the European settlement of a region that was previously
populated by a variety of indigenous peoples. There has been no systematic study
of the possible influences of the indigenous languages on English, but we can see
256
Naomi Nagy and Julie Roberts
Figure 1 Eastern and Western New England according to Carver (1987: 31). Reprinted
with permission from the University of Michigan Press.
their influence in local toponyms, for example the Piscataqua River in NH, the
Kennebec River in ME, Lake Memphremagog in VT, and Contacook, a town in
Rhode Island, as well as the word Massachusetts.
European settlers in Eastern New England came primarily from Boston, on the
Massachusetts Bay, and were of English stock. This coastal area, originally home
to indigenous groups, was settled by English immigrants in the early 1600’s and
became one of the country’s cultural hearths. In search of better farm land, some
of these original European settlers moved west from the coast and settled the
Lower Connecticut River Valley in central CT. They were joined soon after by
new immigrants from eastern and southern England, and later from Italy, Scotland
and Ireland, among other places. Settlement spread, generally along river valleys,
into NH, VT, ME, and RI (Carver 1987: 7).
New England phonology
257
WNE was settled by migration from central MA and central and western CT,
including Hartford, Springfield, and New Haven, towns originally settled in the
1630’s (Boberg 2001: 4). Following this movement, Eastern and Western NE remained isolated from each other until the early 18th century (Rosenberry 1962:
facing 70; Kurath 1972: 42, cited in Boberg 2001: 4). Western VT was settled
speakers in the late 18th century by English-speaking migrants from western CT
and MA (Kurath 1939-43: 104, cited in Boberg 2001: 5) and from NY (Rosenberry 1962: 136, cited in Boberg 2001: 5), as well as some settlers from east of
the Green Mountains (NH, ME, and RI) (Kurath 1939-43: 103-4, cited in Boberg
2001: 5). WNE, in turn, was “the staging ground for the initial English-speaking
settlement of the Inland North” (Boberg 2001: 9).
WNE also “received a considerable admixture of Scotch-Irish in the half century preceding the Revolution [early 18th century]” (Kurath 1928: 391, cited in
Boberg 2001: 9), though they did not form a sizeable percentage of the population
at any time. Also present in NE are Franco-Americans who moved south from
French-speaking parts of Canada, and large Irish and Italian groups. Upper ME
(north of Penobscot Bay) is quite distinct from the rest of the region, due to ties
with New Brunswick, Canada (Carver 1987: 31).
Boston, the largest New England city, is still known as the hub, hearkening
back to its position as the center from which settlements radiated in New England.
Much of the rest of NE, however, is more rural, with many farms, forests, and undeveloped areas surrounding small towns and cities. Like many rural communities,
NE is undergoing changes including increased highways, in-migration from other
dialect areas, and change from small family farms to agribusiness (Frazer 1983;
Labov 1994). The rural, regional dialects appear threatened with obsolescence due
to the decrease in agriculture and increase in in-migration by speakers from other
states. This loss evokes mixed reactions within the communities, where it may be
seen as a sign of progress and increasing sophistication as well as a loss of cultural
identity (Ring 1997).
3.
New England dialect regions
The Linguistic Atlas of New England (Kurath 1939-43) divides the area into
Eastern (ENE) and Western (WNE) (divided by the Green Mountains of VT in
the north, the Berkshires in the middle, and the Connecticut River in the south),
with seven subregions dictated by settlement patterns (Carver 1987). However,
today there is little in the way of linguistic markers of these sub-regions, aside
from some distinctive characteristics of ENE. A Word Geography of the Eastern
United States (Kurath 1949) divides New England into only three regions (Northeastern, Southeastern, and Southwestern), better representing current linguistic
differences.
258
Naomi Nagy and Julie Roberts
As table 1 demonstrates, the English of NE is in many ways similar to that heard
in many other regions of the United States. In the following section, we will discuss the ways in which NE English may be different from other regions.
4.
Vowels
Table 1. New England vowels — summary
KIT
FACE
START
() ~ ()
DRESS
PALM
~
NORTH
() >
TRAP
>
THOUGHT
~
FORCE
()
LOT
~
GOAT
>
CURE
()
STRUT
GOOSE
happY
i
FOOT
PRICE
>
lettEr
()
BATH
> >
CHOICE
horsEs
>
Table 1. New England vowels — summary
CLOTH
MOUTH
>
commA
()
NURSE
()
NEAR
()
kittEn
~
FLEECE
SQUARE
()
aunt
In discussing the vowel patterns, we begin with the elements considered essential
as points of departure for the phonological analysis of North American English dialects, according to Labov (1991: 21). The lack of a merger between low, back, unrounded // (LOT) and mid, back, rounded, lengthened // (THOUGHT) and the behavior
of low front // (TRAP/BATH) as a unified phoneme (rather than split into tense and lax
classes) are seen as essential conditions for the Northern Cities Chain Shift (NCCS),
a major ongoing change in American phonology. The presence of these two phonemic patterns are necessary for the onset of the NCCS: TRAP/BATH raises, leaving a
space for LOT to move forward and maintain its distinction from THOUGHT (Boberg
2001: 11; Labov 1994: 184; Gordon, this volume), thus initiating a chain shift.
4.1.
TRAP, BATH, HAPPY
and DANCE
At the time of the Linguistic Atlas of New England (LANE) fieldwork, both BATH
and TRAP comprised a unified low front vowel across New England (Kurath 193943: Maps 150 sack, 344 pantry, and 371 dad, cited in Boberg 2001: 13). Laferri-
New England phonology
259
ere’s (1977: 102-3) findings from urban Boston show a less uniform picture. She
reported for BATH a non-productive backing: lexicalized and categorical before
many /f/ and // words and in some /n/ words (e.g., half, rather, aunt) and lexicalized but variable before /s/ and in other /n/ words (e.g., last, dance). Supporting
evidence comes from Calais, ME, where a majority of speakers report saying [ant]
for aunt. Some speakers report [nt], but none report [nt]. This differs from
much of the US, where [nt] is used (Miller 1989: 124). Our NH speakers use []
for all of these word classes except aunt, which is [].
Laferriere (1977) also reports a productive, phonological process raising TRAP
and BATH to [], demonstrated by her younger speakers. As this process was
found to affect both TRAP and BATH vowels, it thus encroaches on the lexical BATH
class that had been subjected to backing.
A more recent study of WNE found raising of the nucleus in TRAP and BATH in
all environments and tensing (as well as raising) before nasals (DANCE) (Boberg
2001: 17-19). A small sample of telephone survey data (Labov, Ash and Boberg
fc.) showed this to be the case across WNE with exception of the very northern
city of Burlington, Vermont. Words like bad and stack are pronounced with [e],
and words like stand and can are pronounced [].
Labov (1991: 12) suggests that unified raising of TRAP/BATH/DANCE is a pivot
condition for the NCCS (Northern Cities Chain Shift). Boberg (2001: 11) further argues that the NCCS may thus have had its beginnings in northwestern NE.
The existence of this raising pattern is surprising if one accepts the reported lack
of BATH-raising in the LANE data (Kurath 1939-43), especially given that Labov,
Ash and Boberg (fc.) does not show this to be an incipient vigorous change: older
speakers show more raising than younger speakers in Hartford, CT, Springfield,
MA, and Rutland, VT (Boberg 2001: 19).
4.2.
LOT, CLOTH
and THOUGHT
There was a major split within New England as early as the 1930’s at which point
ENE did not have a distinction between LOT and THOUGHT, while WNE had two
distinct phonemes, (Kurath 1939-43, discussed in Boberg (2001: 13). ENE pronounced both LOT- and THOUGHT-type words with [], while virtually all of WNE
used [] and [:] respectively, resembling NYC.
One modern exception to this pattern is Providence, RI, where the two vowels
are distinct (Labov 2000: Map 1). Another may be Calais, ME, where no speakers
reported a merger in Miller (1989: 101). More recent data (Labov, Ash and Boberg
fc.) presents a strikingly different picture for the LOT/THOUGHT merger. While all
western CT speakers keep the two values clearly distinct, resembling the Inland
North pattern, seven of eight VT speakers have completely merged the two vowels. One older northern VT woman did not merge these vowels, suggesting that the
merger is more recent in VT than CT (Boberg 2001: 20). This trend is supported
260
Naomi Nagy and Julie Roberts
by unpublished data from the McGill-Vermont-New Hampshire Survey (Nagy,
Roberts and Boberg 2002) which shows most New England speakers report merging these two vowels. Our two recorded NH speakers produced LOT, CLOTH and
THOUGHT with []. One of them also produced PALM with this vowel.
Boberg (2001: 22) attributes the presence of the merger in VT to lack of contact
with the Inland North (due to the barrier of Lake Champlain) combined with contact over the Green Mountains with the merged speakers of NH. In contrast, CT
speakers have more contact with NY and thus retain the distinction. Geographically located between CT and VT, western MA speakers exhibit an intermediary
variable pattern. In our data, however, MA has the highest rate of merger. Interestingly, Burlington, VT speakers show a tendency to merge LOT and THOUGHT in low
back position, similar to the ENE merger (and to the Canadian merger just north
of them), whereas the two Rutland speakers, 67 miles south, show a merger in
low-central position (like that of southwestern NE) (Boberg 2001: 24), providing
a gradual transition between the northern and southern WNE patterns.
To summarize, with respect to the LOT/THOUGHT merger and BATH/TRAP/DANCE
raising, ENE has full merger of LOT/THOUGHT (except RI) and no BATH/TRAP/DANCE
raising, except for that reported in Boston by Laferriere (1977). WNE is more
complex:
The CT portion of the lower Connecticut Valley (the Hartford area) is a pure Northern
[NCCS] system, with raised [bath/trap] and centralized [lot], distinct from mid-back
[thought]. Northwestern VT (Burlington) is a pure “third dialect” system, not unlike
the Canadian systems to the north of it [with no bath raising and a lot/thought merger].
Between Burlington and the lower Connecticut Valley are two transitional types.
Springfield, and perhaps western MA in general, is basically Northern [NCSS] but shows
a reduction of contrast between the low-back vowels, which may be tending toward
merger among the youngest speakers in that area. Southwestern VT (Rutland) shows a
solid merger of the low-back vowels but in the phonetic position characteristic of [lot] in
western MA and CT (Boberg 2001:25-6).
4.3.
FACE
and FLEECE
In general, there is nothing remarkable about these tense front vowels. However, Duckert (1986: 141) reports diphthongs in words like [ma
[maan] machine
and [dreijan] drain as a feature of rural New England dialects. Laferriere (1979:
431) lists the variable pronunciation of FACE as [i] or [e] as a marker of Boston
speech.
4.4.
GOAT
Avis (1961) described a complex pattern involving GOAT in ENE. Reporting on the
data from LANE, Avis argues that there are, in fact, two phonemes: an upgliding
New England phonology
261
phoneme that appears word-finally, and another phoneme in which alternation can
be found between monophthongal [o] and one with a fronted inglide [o]. Avis
(1961: 552) also notes that the monophthongal vowel is more likely to be found in
“dialectal” speech than in words “learned in school”. Avis does not report on this
vowel in WNE. Roberts (1997) indicates that GOAT is produced as a lowered, lax
vowel with either no glide or a shortened upglide in VT. All older and younger adult
speakers produce low, lax GOAT, overlapping with their productions of FORCE.
Laferriere (1977: 431) reports GOAT as [] as a feature of Boston English.
4.5.
GOOSE
Kurath (1939-43) found that both a tense ([u]) (as in too) and a lax ([]) (as in
took) production of GOOSE occurred in NE, but we hear only [u] today.
took
4.6.
PRICE
and MOUTH
Miller (1989: 110) reports Canadian raising (the production of PRICE and MOUTH before voiceless vowels as []and [] respectively) in Calais, ME —not surprising
as this town is on the border of Canada. Raising was reported in Calais in LANE
(Map 354, vol. II, Part 1 ; Map 481, vol. II, Part 2 ; Map 53, vol. I, Part 1, cited in
Miller 1989: 110), but not in neighboring towns. Kurath and McDavid (1961: 10910, cited in Miller1989: 112) cited patterns similar to Canadian raising for coastal
ME and southern NH. However, Canadian raising has not been reported elsewhere
in NE. Our NH speakers do not produce raised nuclei in these diphthongs.
A pattern that may be seen as similar to Canadian raising, however, has been
reported in Vermont for some time. Kurath (1939-43) reported a fronted, raised
nucleus of MOUTH was being overtaken by a fronted, but low production in VT.
He also found that change in progress was occurring with PRICE, in that the raised
nucleus was receding in favor of a lowered, more “standard” pronunciation. Work
by Amblo and Roberts (1997) notes the continuation of this trend in VT in that
women and younger speakers are pronouncing these vowels in a more standardsounding way than older rural men.
4.7.
START
Some variation between the central and back variants is seen for this vowel in NH.
Our older male western NH speaker produced START with the central [a], while the
younger female eastern NH speaker produced it with []. The vowel /// before ///
appears as [] even along the ME/New Brunswick border, in spite of the contact
with Canadian [r] pronunciations (Miller 1989: 88). Examples include tomorrow,
sorry and borrow. This pattern was also reported in LANE (Kurath 1939-43: Map
72, vol. I, Part 1 and Map 564-5, vol. III, Part 1). However, all of Miller’s sixteen
262
Naomi Nagy and Julie Roberts
speakers report [] for orange (ibid
ibid 89), while LANE (Map 273, vol. II, Part
1) reported [ ] for this area.
4.8.
NORTH/FORCE
ENEers traditionally made a distinction between pairs like for and four, or horse and
hoarse, which is not heard in most of the rest of the U.S. As a result of this distinction, combined with r-dropping, a Boston pronunciation of short rhymes with shot;
north rhymes with moth. This distinction may be disappearing among young people
(Labov, Ash and Boberg fc.). Our NH speakers have merged these two vowels.
Laferriere (1979: 428) defines the vowel in short and forty (NORTH) as [], in
contrast to the standard [o()]. The words which have this vowel in standard
American English are divided (apparently arbitrarily, cf. McCarthy 1999) into two
classes in the Boston dialect, some of which allow this alternation and some which
use only [o] (Laferriere 1979: 429).
4.9.
BOTHER
and FATHER
Bostonians and Northern New Hampshirites generally maintain a distinction between the vowels in the first syllables of bother [] and father [a], while many
residents of VT and southern NH, especially younger people, have merged those
vowels (Nagy 2001). Miller’s respondents (Miller 1989: 124) report that father
and bother do not rhyme in Calais, ME.
4.10.
MARY, MERRY
and MARRY
Many speakers in eastern MA and northern NH have three distinct pre-rhotic front
vowels, differentiated in the triplet Mary [e:] ~ merry [] ~ marry [], while those
in VT and southern NH pronounce the three words alike (Nagy 2001; Nagy and
Roberts 1998). Miller (1989: 99) reports that most speakers in Calais, ME, have a
two-way merger: for 80% of the speakers, Mary and marry are [meri] and merry
is [mri]. 13% of the speakers surveyed have merged all three. (7% have slightly
different two-way mergers.) This indicates a marked change from LANE, where a
three-way distinction was maintained across NE (Miller 1989: 100).
4.11. Mergers before L
Pre-lateral mergers that occur in other parts of the U.S. are documented as not occurring in NE in Labov, Ash and Boberg fc. These include the following tense and
lax vowel pairs before /l/: /i/ and /i// ((pill
pill and peel), /u/ and /u// ((pull
pull and pool),and
/e/ and /e/ (well
well and wail).
New England phonology
5.
5.1.
263
Consonants
T, D
Several types of substitutions involving the alveolar stops /t/ and /d/ appear in
the New England area. These include both substitutions of spirantized variants
for alveolar stops as well as alveolar stops substituting for interdental fricatives.
Glottal stop replacement of /t/ (e.g., [mn] mitten, [vm] Vermont, [rn]
Right on!) in VT appears to be a robust dialect phenomenon. Although considered
to be a traditional rural phenomenon most common to older male speakers, these
glottal forms are found in speakers of all ages in VT. Children produced at least as
many glottal stop forms as their parents, with girls producing more // than boys
(Roberts 2001). These findings demonstrate that dialect obsolescence, common
in rural areas, does not necessarily mean a change toward “Standard English.” In
this case, girls appear to be leading a change toward a resurgence of glottal stop
replacement. Similar findings have been reported in the United Kingdom where
research on the glottal stop has been going on for years (cf. Milroy et al. 1994;
Foulkes, Docherty and Watt 1999).
Nagy and Ryback-Soucy (2000) indicates the frequent use of alveolar stops /t/
and /d/ in place of interdental fricatives // and // among speakers who self-identify as members of the Franco-American community of Manchester, NH.
Finally, Miller (1989: 104) reports categorical flapping in butter for the speakers he surveyed in ME. LANE also reports flapping for most of NE (Map 496, vol.
III, Part 1, cited in Miller 1989: 105). This is in keeping with the general pattern of
northern AmE: categorical post-tonic flapping for all speakers (Strassell 1997).
5.2.
Word-initial H
The Franco-American speakers studied in Manchester, NH, who substitute [t,d]
for /, /, also variably omit word-initial H and insert an initial H in underlyingly
vowel-initial words (e.g., [oli h nd l a ] Holy Angel High). Interestingly, several of
these speakers are monolingual Anglophones, so this is not a case of mother tongue
interference in a second language, but rather a marking of cultural identity.
5.3.
W/HW
distinction
The distinction between word initial <wh> and <w> words, as in which and witch,
is retained to some extent in parts of NH, VT, and MA (Labov 2000). This pattern
was reported in LANE (Map 163, vol. I, Part 2, and Map 179, vol. I, Part 2, cited
in Miller 1989: 108). However, the distinction was not maintained by Miller’s ME
speakers. Kurath and McDavid (1961: 178) mention this merger as occurring “in a
narrow coastal strip of NE extending from Boston to the Kennebec in Maine.”
264
Naomi Nagy and Julie Roberts
5.4.
JU (jod-dropping)
Our survey data (Nagy and Roberts 1998) show the continuing presence, mostly
among older speakers, of a palatal glide or jod between alveolar consonants and
[u] in words such as new [n(j)u] and Tuesday [t(j)uzde]. This was also noted by
Duckert (1986: 141) as a feature of rural NE speakers. Interestingly, LANE shows
a preference for the jod-less pronunciation even among the oldest speakers (Kurath 1939-43: Map 4, vol. I, Part 1). Sixteen speakers from Calais, ME, surveyed
in the late 1980’s showed no use of the jod in either relevant survey question (the
pronunciation of during and reduce) (Miller 1989:86).
5.5.
R
vocalization and intrusive R
Finally, a frequently noted feature of ENE, also exhibited by speakers in the Virginia and North Carolina hearth areas, is the vocalization (popularly referred to as
“dropping”) of // in post-vocalic position. People talk about “New Hampsha” and
“Woosta” for New Hampshire and Worcester. Similarly, Laferriere (1979: 431)
indicates that the R-less production of START with [a:] is a marker of Boston speech.
Linking R is produced: if the following word begins with a vowel, the R is rhotic
(hearr it). A related NE pattern is the appearance of inter-vocalic // where the standard spelling does not indicate it, referred to as intrusive R, as in [sa: t] saw it.
According to Labov (1966), “the vocalization of /// is eroding under the influence of the post World War II convention that constricted /// is the appropriate standard for careful speech.” However, all three Boston speakers included in
Labov (2000) show some vocalization of //, and one Bostonian shows 50%. In
contrast, most of WNE shows consistent [].
Our recorded NH speakers vocalize // in reading the word list, in words such as
CURE, LETTER, FORCE, NORTH, START, SQUARE, and NEAR. Variable vocalization is also
evident in the recorded and transcribed narratives.
6.
Compound word stress
Duckert (1986: 141) reports a tendency for stress to appear on the second element
of compound words such as maple TREE, band CONCERT
CONCERT, polar BEAR, and battle FIELD in rural NE speech. We are not sure if this pattern is constrained to NE.
7.
Summary
As we have shown, NE presents a complex linguistic profile. There are a number of both consonantal and vowel patterns that preserve the distinction between
New England phonology
265
NEEnglish and other varieties present in the U.S. Some of these features are uniformly distributed across NE, while others illustrate the maintenance of distinct
dialect subregions. It appears that, as people more frequently move into the area
from all over the country, New Englanders increasingly sound like other AmE
speakers. However, some local features remain. Many New Englanders still “drop
their r’s,” though no longer as consistently or in as many words as they used to.
Others substitute glottal stop for T, and many retain a variety of fairly subtle vowel
differences. Thus, much as found by the scholars who documented the linguistic
patterns of this region in the early 20th century, both the NE dialect and its regional
subdialects operate as relevant markers of NE identity today.
*
This chapter is an extended version of a paper written by Nagy, Roberts and Boberg for
American Language Review (2000). We are very grateful to Charles Boberg for sharing
his large bank of knowledge about American dialects with us. We are also grateful for
the assistance of Joleen Hansen and Denis Jobin who recorded and transcribed the two
New Hampshire speakers.
Selected References
Please consult the General References for titles mentioned in the text but not included in the references below. For a full bibliography see the accompanying CDROM or online version.
Amblo, Rebecca and Julie Roberts
1997
Change and obsolescence in rural Vermont: /aw/, /ay/, and /uw/ in younger
and older speakers. Paper presented at NWAV (New Ways of Analyzing
Variation) Conference. Laval, Université Laval.
Avis, Walter
1961
The New England short o: A recessive phoneme. Language 37: 544-558.
Boberg, Charles
2001
The phonological status of Western New England. American Speech 76: 129.
Duckert, Audrey A.
1986
The speech of rural New England. In: Allen and Linn (eds.), 136-141.
Foulkes, Paul, Gerry Docherty, and Dominic Watt
1999
Tracking the emergence of structured variation. Leeds Working Papers in
Linguistics and Phonetics. Leeds, University of Leeds: 1-25.
Frazer, Timothy C.
1983
Sound change and social structure in a rural community. Language in Society
12: 313-328.
Kurath, Hans
1928
The origin of the dialectal differences in Spoken American English Modern
Philology 25: 385-95.
266
Naomi Nagy and Julie Roberts
Labov, William
1991
The three dialects of English. In: Eckert (ed.), 1-44.
Laferriere, Martha
1977
Boston short a: Social variation as historical residue. In: Fasold and Shuy
(eds.), 100-107.
1979
Ethnicity in phonological variation in change. Language 55: 603-617.
McCarthy, John
1999
The dialects of Eastern New England. Linguistics 402 course handout. http://
www-unix.oit.umass.edu/~jjmccart/ling402f01/11-Boston%20Vowels.pdf
Miller, Corey
1989
The United States-Canadian border as a linguistic boundary: The English
language in Calais, Maine and St. Stephen, New Brunswick. Undergraduate
thesis, Linguistics Department. Cambridge, MA, Harvard University.
Milroy, Lesley, James Milroy, Sue Hartley, and David Walshaw
1994
Glottal stops and Tyneside variation: Competing patterns of variation and
change in British English. Language Variation and Change 6: 327-357.
Nagy, Naomi, Julie Roberts and Charles Boberg
2000
Yakking with the Yankees. American Language Review 5: 40-43.
2001
‘Live free or die’ as a linguistic principle. American Speech 76: 30-41.
Nagy, Naomi, Julie Roberts and Charles Boberg
2002
McGill-VT-NH Dialect Survey. Unpublished research instrument.
Nagy, Naomi and Julie Roberts
1998
Yankee doodles in dialectography: Updating New England. Paper presented
at NWAV (New Ways of Analyzing Variation) Conference University of
Georgia.
Nagy, Naomi and Wendy Ryback-Soucy
2000
Exploring the dialect of the Franco-Americans of Manchester, New Hampshire.
Journal of English Linguistics 28: 249-264.
Ring, Wilson
1997
Time erodes all including traditional Vermont accent. The Caledonian Record:
1A, 12A.
Roberts, Julie
1997
/ow/ movement and chain shift: An example from rural Vermont speech.
Paper presented at NWAV (New Ways of Analyzing Variation) Conference,
Laval, Canada.
2001
An American variable? A continuing study of glottal stop in Vermont. Paper
presented at NWAV (New Ways of Analyzing Variation) Conference, North
Carolina State University, Durham, NC.
Rosenberry, Lois Kimball Mathews
1962
The Expansion of New England. New York: Russell and Russell.
Strassell, S.
1997
Variation in American English flapping. In Claude Paradis, Diane Vincent,
Denis Deshaies and Marty Laforest (eds.), Papers in Sociolinguistics NWAVE-26 à l’Université Laval, 125-35. Quebec: Nota bene.
View publication stats