Linguistic Society of America Language

Linguistic Society of America
Moribund Dialects and the Endangerment Canon: The Case of the Ocracoke Brogue
Author(s): Walt Wolfram and Natalie Schilling-Estes
Source: Language, Vol. 71, No. 4 (Dec., 1995), pp. 696-721
Published by: Linguistic Society of America
Stable URL: http://www.jstor.org/stable/415741
Accessed: 11-11-2015 23:22 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
Linguistic Society of America is collaborating with JSTOR to digitize, preserve and extend access to Language.
http://www.jstor.org
This content downloaded from 130.179.16.201 on Wed, 11 Nov 2015 23:22:02 UTC
All use subject to JSTOR Terms and Conditions
MORIBUND DIALECTS AND THE ENDANGERMENT CANON:
THE CASE OF THE OCRACOKE BROGUE
WALT WOLFRAM NATALIE SCHILLING-ESTES
North Carolina State University University of North Carolina at
Chapel Hill and North Carolina
State University
Moribund dialects threatened by the encroachment of healthy varieties of the same
language have been overlooked in establishing the language endangerment canon. Endan-
gered varieties of languages as safe as even English exhibit structures not found in main-
stream language varieties and so are an invaluable resource to scholars of language
variation-and, indeed, of language patterning in general. Further, the insights into lan-
guage variation and change that we gain from studying moribund dialects inform our
study of the types of changes that characterize endangered and dying languages. Our
arguments are based on the examination of Ocracoke English, a dialect of American
English which is spoken on Ocracoke Island, located off the coast of North Carolina
and inhabited by about 600 year-round residents. This dialect developed in relative isola-
tion from mainstream varieties of American English but is now threatened by encroach-
ment from mainland dialects as the island becomes more accessible to the outside world.
Using the case of the Ocracoke production of the /ay/ diphthong as [Ak'], we present
linguistic and sociolinguistic evidence that Ocracoke English is indeed an endangered
dialect. We also describe the development of a community-based preservation program
that parallels the type of proactive programs that have been implemented thus far only
for endangered language situations.*
1. WHY STUDY ENDANGEREDDIALECTS?Recent symposia (for example, the

Linguistic Society of America's annual meetings in 1991 and 1995, the American
Association for the Advancement of Science annual meeting 1995) and collec-
tions of essays (Dorian 1989, Hale et al. 1992, Robins and Uhlenbeck 1991)
speak poignantly to the language-endangerment epidemic. Although linguists
show an understandable urgency about the moribund status of the majority
of the world's languages, they rarely mention endangered dialects of a 'safe'
language-that is, particular varieties of a language whose unique status is
threatened by other, encroaching varieties of the same language. The literature
on language death includes examples of endangered varieties of safe languages
(e.g. King 1989, Tsitsipis 1989), but these cases invariably involve encroach-
ment by different languages-not by different varieties of the same language.
Language death and language endangerment for dialects seem to be defined by
a bilingual contact situation. Only in the field of traditional dialectology has
there been widespread concern for documenting forms found in dying dialects
surrounded by thriving varieties of the same language, but there has been little
*
The research reported here was supported by National Science Foundation Grant SBR-93-
19577 and by the William C. Friday Endowment at North Carolina State University. Thanks to
Nancy Dorian, Kirk Hazen, Brian Joseph, and Dennis Preston for their helpful comments on an
earlier draft. Special acknowledgement is due to Yancey R. Hall for his assistance in extracting
some of the /ay/ data and especially to Dennis Preston for conducting the trinomial VARBRUL
analysis.
696
MORIBUND DIALECTS AND THE ENDANGERMENT CANON 697
connection between this area of study and the study of the language death
process itself.
There are a number of reasons for this neglect of dying dialects in the midst
of the escalating concern for languages on the brink of extinction. Linguists
may feel that there is little point in rushing into a dying dialect area if the core
language as a whole shows no signs of vanishing. At first glance, this seems
like a reasonable basis for omitting dying dialects from the canon of language
endangerment. On closer inspection, however, this exclusion is based on the
questionable assumption that interlanguage variation is of more value than intra-
language diversity, which in turn, is based on the assumption that the systemic
boundaries between language and dialect are obvious and discrete. Further-
more, it is inconsistent to admit endangered dialects of safe languages as worthy
objects of study only when they are situated in bilingual settings-not when
they are surrounded by safe varieties of the same language (e.g. King 1989,
Tsitsipis 1989).
The omission of dying dialects from the language-endangerment canon may
also be based on the descriptive problems encountered in sorting out the subtle
ways in which dialects often differ from each other. It can be difficult to deter-
mine which language features are characteristic of the receding variety, which
belong to the expanding dialect, and which features were already present in
both varieties before one began to encroach on the other. As Hoenigswald
(1989:348) observed:
So long as the differences between the two languages are gross, the distinction [between
healthy and dying languages] seems feasible and meaningful. But what about dialects'? Which
do you believe or expect: that dialect death is an unspectacular, endemic, everyday occurrence,
taking place pervasively and beneath the threshold of awareness; or contrariwise, that there
can be no such thing as dialect death by definition'? Or does it matter'?It is probably no accident
that none of the papers assembled here deals with such a situation however remotely.
Given the moribund condition of so many of the world's languages, it may

seem quite reasonable to treat dialect death in healthy languages, by compari-
son, as inconsequential and superfluous. However, our study of Ocracoke En-
glish, a moribund American English dialect spoken on Ocracoke Island in the
Outer Banks of North Carolina (Schilling-Estes & Wolfram 1994, Wolfram et
al. 1995, Wolfram & Schilling-Estes 1995; Wolfram & Schilling-Estes 1996)
shows that the meaningful study of endangered language varieties is, indeed,
both possible and necessary. Dying dialects of even languages as alive as En-
glish exhibit features not found in more mainstream varieties, and these struc-
tures need to be documented in order to provide a full portrait of diversity
within language. Further, some obsolescing forms and unique configurations
of forms in moribund dialect areas may be vital to our understanding of funda-
mental issues of language change and variation, as well as the formulation of
models describing and explaining such mutation.
Conversely, we can apply insights from the broad-based study of language
variation to the investigation of language death in bilingual and bidialectal situa-
tions with productive results. For example, conclusions reached in our study
698 LANGUAGE, VOLUME 71, NUMBER 4 (1995)
of Ocracoke English lead us to question the 'widely shared intuition' (Hill 1989:
149) that the processes of change in language death differ from ordinary change
processes primarily in the rapidity with which they occur. A number of investi-
gations of the patterning of linguistic variables with respect to internal linguistic
and external social factors in the course of language change indicate that vari-
ables display quite different-and often less linguistically and sociolinguistically
'natural'-patterning at the beginning and end points of a change than at other
points in its progression (Fasold 1973, Wolfram 1988, Labov 1994). Such un-
usual patterning may well characterize language changes we find in dying lan-
guages and language varieties, since, presumably, a language on the verge of
extinction will be characterized by a heavy concentration of incipient changes
to exterior linguistic norms and terminating changes in native language struc-
tures. A portion of the current analysis, for example, involves the detailed study
of the replacement of an obsolescing Ocracoke English vowel production-that
is, the pronunciation of the layl diphthong as [Ai']-with the Southern Ameri-
can English pronunciation of this diphthong as [a:]. Our investigation shows
that this replacement process does not parallel ordinary, natural change as
neatly as we might assume.
The data from intralanguage variation may even prove valuable for examining
templates of universal language, given the interplay between universality and
diversity in language. Trudgill and Chambers (1991:294), citing an empirical
case of nonstandard syntax variation in Wessex English that challenges the
universal IP template (lhalainen 1991), observe that familiarity with dialect
variation 'could potentially sharpen theoretical arguments as much as do im-
ported "exotica" from Kinyarwanda or Ponapean.' Although the in-depth in-
vestigation of the /ay! variable we present in this analysis does not directly
challenge current assumptions about the nature of universal grammar, the po-
tential significance of intralanguage variation for universal grammar certainly
must be admitted.
Arguments for the inclusion of dialects in the endangerment canon are not
limited to linguistic considerations per se, but extend to the sociocultural role
of language as well (Hale, in Hale et al. 1992, Woodbury 1993). Hale (1992:1)
points out that 'It [language loss] is part of a much larger process of LOSS OF
CULTURAL AND INTELLECTUAL DIVERSITY in which politically dominant lan-
guages and cultures simply overwhelm indigenous local languages and cultures,
placing them in a condition which can only be described as embattled.'
While the link between dialect and (sub)culture may be more subtle than
the link between language and culture, it is certainly no less real. In fact, the
subtlety of the dialect-subculture connection probably heightens the level of
the threat to a dialect's maintenance. Dialects are typically disregarded when
considering nonmainstream subcultures. Even worse, they are often dispar-
aged, since they are typically held to be unworthy approximations of a standard
language variety (Labov 1969).
Our investigation of Ocracoke English reveals that nonmainstream dialects,
despite their negative valuation by those who esteem standard English, indeed
play a large role in the shaping of cultural identity. Further, the threat to cultural
MORIBUNDDIALECTSAND THE ENDANGERMENTCANON 699
identity reflected in the loss of these dialects is just as real as the threat of
cultural loss symbolized by language death. For over two and a half centuries,
Ocracoke Island, located twenty miles off the coast of North Carolina and
accessible to this day only by ferry, existed in relative isolation from the main-
land U.S. The island is fourteen miles long and a little less than a mile wide at
its widest point. The village of Ocracoke is located within the southernmost
mile; the remainder of the island is now a National Seashore protected from
residential development. During an extended period of insulation, Ocracokers
developed not only a unique dialect but also a unique cultural identity which
became inextricably bound to their language variety. Shortly after World War
II, with the advent of a state-run ferry service and the construction of a paved
road on the island, Ocracoke began to host ever-increasing numbers of tourists
and new residents from the mainland. The unique culture of the island is now
threatened by mainland ways of life, and the traditional dialect is correspond-
ingly fading. During tourist season, newcomers to the island now outnumber
ANCESTRAL ISLANDERS-residents who can trace their island heritage at least
several generations back. Even more strikingly, a minority of the current gener-
ation of Ocracoke schoolchildren are ancestral islanders. Our conversations
with Ocracokers reveal that their traditional dialect has for many years played
a strong symbolic role in projecting island identity; correspondingly, a true
sense of cultural loss is felt as the dialect vanishes. As one of our key informants
in Ocracoke, a fifty-year-old male, told us:
(1) I got a little kid, see, he's four weeks old, [by the] time he gets grown,
his accent will be what they call 'dingbattish' [i.e. non-island]. But
I would like for him to keep the same accent and heritage that we've
had for years and years, but all this is gone now. The only way we
can preserve it is for you fellers to put it on tape.
The link between dialect and culture is even felt by younger members of the
Ocracoke community, as indicated in a thirteen-year-old girl's response to a
question about her feelings toward the disappearance of the traditional Ocra-
coke dialect.
(2) It's [i.e. the Ocracoke dialect] sacred, really, the way we talk; it's
something the island is special for.
Despite the cultural value embodied in their language varieties, speakers of
endangered dialects often relinquish their native speech in the face of pressure
from safer, encroaching varieties and from those who condemn vernacular vari-
eties in general, whether dying or thriving. Through the systematic and rigorous
study of endangered dialects-and, crucially, the dissemination of the results
to speakers of the dialects under study-linguists can show these speakers that
no language variety, no matter what negative social valuation it has acquired,
is linguistically inferior. The study of endangered dialects is thus important,
not only from an intellectual and scientific point of view but from a sociocultural
and humanistic perspective.
The threatened status of a number of small postinsular dialect communities
should be of great concern to dialectologists and language variationists who
value the full spectrum of diversity within a language. In fact, the history of
dialectology shows a fastidious concern for documenting the forms found in
nonmainstream communities, and one of the motivations for initiating the Lin-
guistic Atlas of the United States and Canada was the concern for capturing
some of the relic forms found in such communities before they disappeared
(Kurath 1939). Thus, our study of obsolescing forms in Ocracokefits well within
the focus of traditional dialect study. At the same time, we extend our concern
beyond the mere documentation of such forms so that we examine the linguistic
and sociolinguistic dynamics of dialect erosion. We further relate the status of
moribund dialects to the current concern with language endangerment. In the
process, we argue that the social responsibility of linguists toward dying dialects
is every bit as great as the responsibility of linguists toward dying languages
and their speakers (Hale et al. 1992).
In this presentation, we first consider evidence for the designation of the
Ocracoke dialect as endangered, with special focus on the patterning of the
/ayl variable mentioned above. Following the presentation of our evidence, we
discuss the social responsibility that linguists and dialectologists should assume
with respect to documenting and disseminating information about the state of
moribund dialects. We conclude by describing a community-based, collabora-
tive model that involves the development of materials and programs designed
to foster knowledge of and pride in the local dialect. Although such programs
exist for selected endangered languages (e.g. Watahomigie and Yamamoto,
Craig, in Hale et al. 1992), we are not aware of similar programs focused on
endangered dialects of safe languages. We hope that the description of our
program will encourage other students of language variation to become involved
with communities in a way that benefits the local research communities
as well as the linguists who use these communities for field-initiated
research.
2. OCRACOKEENGLISHAS AN ENDANGEREDDIALECT. In this section, we pre-

sent the sociohistorical, linguistic, and sociolinguistic evidence for maintaining
that Ocracoke is an endangered dialect. We consider the overall status of the
variety and examine in detail a receding phonological variable that has tradition-
ally marked this variety: the production of the /ay! diphthong with a raised,
backed nucleus-that is, [AF1]. The detailed discussion of a single dialect vari-
able will also show how the replacement of obsolescing forms with forms from
encroaching dialects can inform our understanding of language variation and
change.
2.1. THE SOCIOHISTORICAL

BASIS. The first Euro-American settlers on Ocra-
coke were English ship pilots who settled in the early 1700s on land claimed
by the English throne. Various land ownership acts in the mid-1700s apparently
brought in upperclass English settlers from Southeast and Southwest England,
as well as settlers of British descent who migrated from their original settlement
sites in the Maryland and Virginia colonies. Several island families trace their
lineage to these early settlers of English origin, though one of Ocracoke's oldest
and largest families, the O'Neals (the oldest living O'Neal woman had eleven
sons), claims Irish ancestry. The historically isolated islanders made their living
chiefly through such maritime-related trades as fishing and guiding ships
through the Outer Banks barrier islands. Ocracokers were isolated not only
geographically and economically from the mainland South but also socially.
Islanders tended to limit their social ties to the island, or at least to the Outer
Banks chain of islands, and marriages were confined almost exclusively to
island families, as evidenced in the relatively small number of surnames we
find on the island to this day. Many of these surnames can be traced back to
the island's original settlers.
Shortly after World War II, the longstanding isolation of Ocracoke Island
came to an end. A coastal highway was constructed along the Outer Banks
islands, and a state-run ferry service was instituted. These changes allowed far
greater access to Ocracoke than had been available in the pre-War years, when
boat transportation was sporadic and land travel was often over bare sand.
Ocracoke now began to host ever-increasing numbers of tourists and new resi-
dents from the mainland. Today, Ocracoke is home to only about 200-250
ancestral islanders. This represents less than half the year-round population
and less than one-tenth of the population on a typical day during the summer
tourist season. In addition, the island economy has been converted from one
largely independent of mainlanders to one almost wholly dependent on out-
siders. As with other islands along the Eastern seaboard, a vibrant tourist indus-
try has developed, and most ancestral islanders are employed in the tourist
service industry in one way or another. With the breakdown of economic and
geographic barriers separating Ocracoke from the mainland has come a corre-
sponding breakdown in social barriers. Marriage outside the island community
is becoming increasingly common, and the youngest islanders now interact
on a daily basis with classmates and teachers whose families are transplanted
mainlanders.
We summarize the social, cultural, and historical facts supporting our conten-
tion that the Ocracoke dialect is now endangered as follows:
1. Two and a half centuries of geographic isolation are brought to a sudden
end with the completion of highways and the implementation of a state-run
ferry service.
2. Ancestral islanders become a minority population on the island, as main-
landers establish permanent and vacation residences on the island.
3. The economic base shifts from a relatively self-sufficient marine-related
economy to one heavily dependent on the tourist trade.
4. Social networks extend beyond the confines of the island as Ocracokers
come into more contact with outsiders; marriage with mainlanders becomes
more commonplace, as do working and other social relationships.
The sociohistorical situation contextualizing the Ocracoke dialect neatly par-
allels what is found in a number of endangered languages, including East Suther-
land Gaelic, the focus of Dorian's classic study (1981) of the dying language of
a small Scottish fishing village, and the Newfoundland French spoken in Port-
au-Port, a Canadian fishing village community on which the outside world is
rapidly encroaching (King 1989).
2.2. THE LINGUISTICBASIS. The constellation of structures that defines the

traditional Ocracoke dialect, or BROGUE, as it is sometimes referred to by island-
ers, certainly sets this dialect apart from mainland Southern varieties. While
sharing some features with contiguous southern coastal varieties, the Ocracoke
brogue has many syntactic and phonological features in common with isolated
highland varieties such as Appalachian English and Ozark English (Wolfram
and Christian 1976, Christian et al. 1988). These include a-prefixing (e.g. Rex
uwenta-fishin'), absence of the plural -s marker with nouns of weights and
measures (e.g. four mile), singular verb concord with collective noun phrases
(e.g. People likes to fish) and conjoined noun phrases (e.g. Candy and Melinda
usually takes them), [e] lowering preceding r (e.g. [6ar] 'there'), -ire syllable
restructuring (e.g. [tar] 'tire'), schwa raising in final unstressed syllables (e.g.
[ekstri] 'extra'), and intrusive r in unstressed final syllables (e.g. [fela] 'fellow')
(Wolfram et al., 1996).
At the same time, Ocracoke English manifests a few peculiar characteristics
that reveal some relics of its historical roots and insular development. One of
the most salient features found in Outer Banks English is the raising and backing
of the nucleus of the /ay/ to [A'-], a diphthong near the phonetic quality of/oy/,
so that a word like tide sounds something like [to'd]. In fact, Ocracokers are
so well known for this diphthong, especially as contrasted with the neighboring
Southern American glide-shortened or unglided lay! (as in [ta:d] 'tide'), that
they are often called 'hoi toiders' for 'high tiders.' Unique morphological fea-
tures that occur in Ocracoke speech include iere/n't regularization (e.g. I/you!
(s)helwelthey weren't there), a morphological realignment rare among Ameri-
can English dialects, in which only the were form is used with negatives while
was is the preferred form in affirmative sentences, as in llyoul(s)helwelthey
was there (Schilling-Estes and Wolfram 1994).
The dialect spoken in Ocracoke naturally shares the vast majority of its lexical
items with other varieties of American English, although it combines words
typically considered Northern with Southern lexical items, as well as items
associated with both coastal and inland dialects. Ocracoke English also includes
a set of lexical items we have not found elsewhere. These Ocracoke-specific
words include terms for games and activities, such as meehonkey and whoop
and holler for local versions of hide-and-seek; call the mail over for 'mail deliv-
ery to the island post office'; and scud for 'car ride'. While there are few lexical
items or words that are unique to Ocracoke, they cannot be ignored, as some
of them are used fairly frequently in everyday speech and are sociolinguistically
quite salient. Another set of lexical items is common to Ocracoke and other
Outer Banks communities, including some of the communities along the coastal
mainland areas immediately adjacent to the Outer Banks. This set includes
items like mommuck 'harass', slick cam 'very smooth' (with reference to
water), dingbatter 'nonislander', quamish 'upset (stomach)', and so forth.
Since many of the dialect features which once defined the Ocracoke brogue
are now fading rapidly among younger speakers, the current state of the Ocra-
coke variety is of particular interest for the study of dialect erosion. To examine
the nature of this process, however, it is necessary to examine the detailed,
variable dimensions of particular receding forms as they are replaced by alter-

nates from encroaching dialects. An in-depth examination of particular forms
also shows how the study of obsolescing dialect forms can inform our under-
standing of language change. For example, it may provide an empirical basis
for challenging the assumption that change in language death is defined only
by its rapidity (Hill 1989). As an illustrative case, we examine in some detail the
fading use of the distinct /ay/ production in Ocracoke speech and its alternative
replacements.
2.2.1. THE CASEOF /AY/. In many respects, the examination of /ay/ in Ocra-
coke brings into focus the linguistic complexity of dialect death, as well as the
social embedding of phonological change and variation in dialect encroachment.
Although the nucleus and glide of /ayl show considerable phonetic range, there
seem to be three competing, socially meaningful options for /ay/ on the Outer
Banks of North Carolina. These are:
1. a unique Outer Bank production with a backed, raised nucleus with a front
upglide. We represent this as L[oy],but the phonetic quality is typically more
like [A-']. This is the classic hoi toider variant that has become an icon of the
so-called Outer Banks brogue.
2. a socially unmarked, low central variant with a front upglide-that is, [al']
(referred to henceforth simply as [ay]). This is the variant associated with non-
Southern varieties and with standard English.
3. an unglided or glide-shortened variant ([a:]), typically occurring with a
more fronted nucleus, the classic production associated with Southern varieties
of English spoken in the contiguous mainland.
The [3y] variant, which is the older, relic form in Ocracoke (Wolfram and
Schilling-Estes 1995), is at this point a highly variable production that is quite
sensitive to internal linguistic and external social constraints. As a central part
of our investigation of [3y] in Ocracoke, we conducted a quantitative study of
the incidence of [oy] and the alternative variants [ay] and [a:]. Although there
is considerable variation in the phonetic production of the /ay! nucleus in Ocra-
coke, we chose to organize our tabulations in terms of the three variant categor-
ies outlined above. In Table 1, the relative frequency of each of the variants
is given for a subgroup of 22 cross-generational speakers, taken from our more
exhaustive interview sample of over 70 speakers. Three different generations
of speakers are included in our sample in order to represent an apparent time
dimension: older (aged 59-81), middle-aged (aged 35-50), and younger (aged
10-22). This framework should allow us to observe the ways in which the
traditional, relic form is receding in the face of external pressures. Both men
and women are included for each group, although in our statistical treatment
we distinguish by sex only for middle-aged speakers. For middle-aged men, we
distinguish two different groups, referred to here as 'poker players' and 'non-
poker players.' The designation of 'poker players' is a convenient label for
members of an exclusive all-male islander group that meets several times a week
to play poker. More importantly, this group shows relatively dense, multiplex
networks (Milroy 1980) and shared cultural values. These men are typically
involved in fishing or other marine-related industries (such as working on the

ferry), and they socialize with each other on a regular basis. This group is well
known on the island and includes some of the indigenous leaders. It consists
of middle-aged men, but younger initiates sometimes cluster around the fringes
of the group. The members of the 'poker game network' are united by their
strong belief in the positive value of being true islanders, especially as con-
trasted with the tourists and new residents who increasingly threaten their tradi-
tional way of life. This belief is projected in a number of ways, including, as
we shall see, the use of the [3y] dialect feature.
The relative incidence of the three socially meaningful variants of /ay/ was
tabulated on the basis of four different kinds of following environments, using
traditional procedures for tabulating linguistic variables in variation studies
(Wolfram 1993a). These include (I) a following voiceless obstruent, as in nice
or light; (2) a following voiced obstruent, as in tide or dive; (3) a following
nasal, as in time or nine; and (4) a following lateral, as in island or mile.' In
Table 1, the raw numbers and percentages of the unique island variant [oy] and
Southern variant [a:] are given for each environment, separated on the basis
of different generational/social groups-that is, older speakers, middle-aged
poker game network members, middle-aged non-poker-playing men, middle-
aged women, and younger speakers. In Table 2, we give results of a VARBRUL
analysis for the raw data presented in Table 1. VARBRUL is a probabilistic-
based multivariate statistical procedure that shows the relative contributions
of different factors to the overall variability of fluctuating forms (Cedergren
and Sankoff 1974, Sankoff and Labov 1979, Guy 1993). Factor groups may
consist of independent linguistic constraints or external social ones. The
weighting values range from 0 to 1. For a binomial application, a factor with a
value greater than .5 favors the application of the process or rule whereas a
value of less than .5 disfavors its application. The input probability, which is
interpreted in the same way, is the likelihood of the rule application apart from
the influence of the environmental factors (Fasold 1990:280, Guy 1993). Since
this analysis involved three primary variants, [oy], [a:], and [ay], a trinomial
analysis was undertaken. In a trinomial analysis, a probability value greater
than .333 favors the incidence of the variant and a value of less than .333
disfavors it. This analysis permits us to examine the patterns of retention for
the traditional Ocracoke variant, the encroaching Southern norm, and the non-
Southern [ay] variant.
The overall figures clearly show that the traditional relic form [oy] is receding
rapidly among the younger generation of speakers. However, several important
patterns emerge from an examination of the figures in Table 1 and Table 2
indicating that this recession is not taking place unilaterally.
' Although we include following laterals in our tabulation, we shall not discuss them in detail
here since /1/, like its liquid counterpart Irl, affects preceding vowels in ways that bear on syllabicity
rather than sonorancy. For example, /ay! in items such as fire and island in Ocracoke may be
categorically monophthongized to [far] and [aland], respectively, in order to avoid syllabifying a
consonantal liquid.
SUBJECT VI. Obstr. Vd. Obstr. Nasal Lateral

[3y] [a:] Tot [3y] [a:l Tot [3y] [a:] Tot [oy] [a:] Tot
Older (3 M, N 117 20 318 139 8 191 138 14 236 35 17 106
3 F) % 36.8 6.3 72.8 4.2 58.5 5.9 33.0 16.0
Middle-Aged N 71 12 185 97 1 126 84 28 166 37 9 75
Poker Player %o 38.4 6.5 77.0 0.8 51.0 16.9 49.3 12.0
(3 M)
Middle-Aged N 29 1 155 63 4 109 82 14 159 15 16 70
Non-Poker 18.7 0.7 57.8 3.7 51.6 8.8 21.4 22.9
(3 M)
Middle- N 18 2 86 27 2 59 23 5 73 2 11 19
Aged % 20.9 2.3 45.8 3.4 31.5 6.8 10.5 57.9
Women
(3 F)
Younger N 25 14 246 54 1 163 61 37 196 6 16 55
(4 M, 3 F) % 10.2 5.7 33.1 0.6 31.1 18.9 10.9 29.1
Total N 135 25 555 203 5 376 202 74 518 53 40 203
% 24.3 4.5 54.0 1.3 39.0 14.3 26.1 19.7
TABLE 1. Variable Constraints on [3y] and [a:]
VARBRUL Probabilities
[3y] [a:) [ay]
Social Factors
Older .474 .267 .259

Middle-Poker .457 .319 .224
Middle-Non-Poker .360 .271 .369
Middle-Women .251 .355 .394
Younger .176 .417 .407
Linguistic Factors
VL Obstruents .242 .240 .518

VD Obstruents .600 .117 .283
Nasals .344 .428 .229
Laterals .150 .625 .225
Input Probability .374 .082 .544

TABLE 2. Trinomial VARBRUL Analysis for [3y], [a:], and [ay]
First, consider the constraint ordering for the variable incidence of the [oy]
variant with respect to the following phonological environment, as indicated
by the VARBRUL analysis: voiced obstruent > nasal > lateral > voiceless
obstruent. That is, a following voiced obstruent is the most favoring environ-
ment for the occurrence of [oy] as opposed to lay] or [a:], and a following
voiceless obstruent is the least favoring, with nasals and laterals in between.
Interestingly, the ranking of voiced over voiceless obstruents with respect
to the raising and backing of the /ayl nucleus in Ocracoke is the opposite of
the constraint ordering associated with the more centralized raising of layl in
other dialects of American English, including the 'standard' variety. In these

varieties, the raised variant[al'] is favored when the following segment is voice-
less (i.e. voiceless obstruent > voiced obstruent). That is, the nucleus of write
([ra't]) will be more raised (and shorter in duration) than the nucleus of ride.
[ra:ld]. This is the pattern Labov found for the raising of the nuclei of /ay/ and
/aw/ in Martha's Vineyard (Labov 1963) and the oft-cited pattern discussed in
early versions of generative phonology (Chomsky & Halle 1968).
Conversely, the ordering of constraints on the backing and raising of the
nucleus of [oy] in Ocracoke is the same as that on the ungliding of [a'] to [a:]
in the mainland South (e.g. Bernstein & Gregory 1994, Feagin 1994, Thomas
& Bailey 1994), where we find that ungliding is more likely to occur before
voiced segments than voiceless ones (e.g. side is more likely to unglide than
sight). The typical constraint pattern for Southern ungliding is: laterals > nasals
> voiced obstruents > voiceless obstruents.
How can we explain the fact that the constraints affecting [oy] in Ocracoke
with respect to the voicing of following obstruents are ranked in the opposite
order of those affecting /ay/ raising in standard English varieties and in the
Martha's Vineyard Dialect, while the constraints affecting Ocracoke [oy] and
Southern American English [a:] are parallel? As stated several times above,
the nucleus of the traditional Ocracoke [3y] is not simply raised but backed.
In other words, it is located along the outer edge, or periphery, of the phonetic
space of the vowel grid rather than in the non-peripheral space occupied by the
raised/centralized/s1/ we find before voiceless obstruents in standard English
varieties (Labov 1994). Similarly, Southern ungliding also represents the move-
ment of /a/ toward peripheral status. This parallel peripheralization is exactly
what we would expect, given that Southern [a:] and Ocracoke [3y] represent
more advanced stages in the diphthongal rotation system than Standard English
[ay], as outlined by Labov (1991, 1994) in his principles of vowel system and
subsystem rotation.
The diachronic path followed by the nucleus of the /ay! diphthong in previous
stages of English, as well as its future directionality as predicted by Labov's
principles, is outlined in Figure 1.
During much of the course of its downward movement from /iy/ to /ay/ as
a component of the Great English Vowel Shift, the nucleus of the /ay! phoneme
was nonperipheral. In its downward trajectory, this nucleus gradually became
more and more peripheral, until it entered the peripheral track on the bottom
of the vowel grid, as Ial. According to Labov, once the nucleus of/ay/ bottoms
out and becomes peripheral, it may follow one of two paths. One alternative
for /ay/ is to monophthongize to [a], thus abandoning the diphthongal rotation
system in favor of the movement pattern of tense monophthongs. This is the
path followed by /ay/ in much of the mainland South, where unglided [a:] has
been shown to be a more recent pronunciation than diphthongal /ay! (Thomas
& Bailey 1994). The monophthongization of /ay/ has taken place at least once
before in the history of English and was indeed one of the main triggers of the
Great Vowel Shift. A Middle English /ay/ diphthong, which is unrelated to
current /ay! and evidenced in spellings such as day and maid, became long
[a:], a peripheral vowel. This [a:] merged with the long /a:/ which had just been
[iy] [u']
[E "] /oy/
[a'] [: ]
\. /
/ayl
(adapted from Labov 1991:9)
FI;GURE i.
created by the lengthening of vowels in words with original short /a/; and the
new, highly peripheral phoneme began to front and raise, proceeding from /a/
to /ae/ to /e/, as we would expect under Labov's Principle I.2
The other alternative for /ayl as the nucleus bottoms out is to remain diph-
thongal and began to rise along a BACK peripheral path. Seemingly, this is the
path followed by Outer Banks [oy]. The different, yet equally peripheral loca-
tions of the nucleus of Outer Banks [oy] and the Southern [a:] vowel are illus-
trated in Figure 2, along with that of the nucleus of standard English unraised
[ay], which lies in peripheral position, and that of the nucleus of standard raised/
centralized [a'], which lies in nonperipheral space.
If the constraining effect of the following segment on /ay/ raising is ordered
in one direction along the voicing dimension in peripheral space and in the
opposite direction in nonperipheral space, we can readily explain the opposing
constraint orderings affecting Ocracoke [oy] vs standard English [ay], as well
as the parallel constraint rankings affecting the Ocracoke and Southern English
variants.
2.2.2. THE ENCROACHMENT
OF SOUTHERN[a:]. We need to examine the South-
ern unglided [a:] not merely as an external dialect norm but as a socially mean-
ingful dialect feature which is making inroads into the Ocracoke variety, as
islanders come into increasing contact with mainland Southerners.3 It is not
2 Labov's
principles for vowel system and subsystem movement are as follows:
Principle I: In chain shifts, tense nuclei rise along a peripheral track
Principle II: In chain shifts, lax nuclei fall along a nonperipheral track
Principle III: In chain shifts, back vowels move to the front
Principle III': In chain shifts, tense vowels move to the front along peripheral paths, and lax
vowels move to the back along nonperipheral paths
3 Note that at prior periods in the island's history, Ocracokers came into more contact with non-
Southerners, chiefly coastal residents, than with Southerners. For example, prior to World War
II, many Ocracoke men left the island for extended periods to work in such Northern port cities
as Philadelphia and Wilmington, Delaware. Today, islanders are much more likely to leave the
island for work in North Carolina and Virginia. Further, much of the sustained, service-related
contact Ocracokers now maintain with outsiders takes place with Southern mainlanders.
FIGURE 2.
surprising that Southern ungliding in Ocracoke is starting to make inroads, but

the patterning of this encroaching norm is not what we might predict based on
our knowledge of the structured variability of Southern ungliding. Tables 1 and
2 show a somewhat surprising constraint ordering for incipient ungliding in this
variety. We would expect that incipient ungliding is starting to make inroads
before nasal and lateral segments, since these are a favorable environment for
ungliding. It is surprising, however, that ungliding is more highly favored before
voiceless obstruents than voiced obstruents. The ordering of constraints on the
ungliding of [a:] in Ocracoke is: lateral > nasal > voiceless obstruents > voiced
obstruents.
While Southern ungliding in Ocracoke is promoted before laterals and nasals,
it is clearly being resisted where the [oy] associated with the traditional Ocra-
coke brogue is most strongly entrenched--before voiced obstruents. There is
virtually no ungliding for any of our speakers before voiced obstruents. This
constraint ordering-that is, the voiceless environment favoring ungliding over
the voiced environment-does not make sense in terms of a natural sonorancy
hierarchy and does not conform to our expectations about constraints affecting
peripheral vowels like the unglided [a:].
The extent of the shift to an external, Southern-based norm for /ay/ in Ocra-
coke can be seen clearly in Figure 3. Here, the incidence of the traditional
Ocracoke [cy] and the incidence of the Southern unglided variant are shown
graphically for the representative speakers in the different generational and
social groups into which we have divided our sample, based upon the VAR-
BRUL probabilities.
We might expect that the incidence of the traditional Ocracoke [3y] will
steadily decline with decreasing age, as islanders come into more frequent con-
tact with surrounding language varieties. However, the middle-aged poker-play-
MORIBUND DIALECTS AND THE ENDANGERMENT CANON 70(9
1.000 -
.750 -
. .500 -
L:9--- 474 ,
t ~~~~~~~.417
!L
.250 -??
=C-0.176
.000 I
Older Mid-Poker Mid-Non-Poker Mid-Women Younger
AgelSocial Group
FIGURE 3.
ing men and older speakers in our sample show no significant difference in the
use of [oy]. Does this indicate a reversal of, or at least resistance to, a change
in progress on the island toward reduced [3y] usage? A look at the incidence
of [oy] among middle-aged male speakers who are not members of the poker
game network, middle-aged women, and the youngest speakers shows that the
maintained incidence of the [3y] variant by the poker players is not a true
dialect reversal. The non-poker-playing age cohorts and younger speakers show
reduced frequency levels for the traditional hoi toider variant compared with
both older speakers and middle-aged poker players. In fact, if we restricted
our tabulations to speakers aged 16 and under, we would find only vestiges of
the [oy] production (less than 10% of all /ay! tokens are realized with [3y]),
showing how rapidly this form is fading.
The pattern of [Oy] production by the middle-aged poker players suggests
that the projection of the traditional vowel is a temporary maintenance of the
traditional island variant, apparently taking place before giving way to external
norms. This temporary rise in a declining variant is highly reminiscent of the
situation Labov observed for centralized /ay! and /aw! in his classic study of
Martha's Vineyard (1963). Labov, however, reported that the dramatic rise in
receding forms he observed among ancestral islanders who strongly identified
with traditional island ways of life represented a permanent reversal of a change
in progress. In the Ocracoke case, the maintenance of the traditional variant
represents a sort of dialectal last gasp before the [Oy] variant dies.
It is noteworthy that the traditional Outer Banks [oy] is maintained at a high
level in the speech of Ocracoke poker players even as the Southern variant is
added. The middle-aged poker players rank higher on the Southern [a:] index
than both older speakers and their middle-aged non-poker-playing male cohorts,
although middle-aged females and younger speakers outrank them in their use
of La:]. In a somewhat ironic way, these men are projecting island identity by
mixing traditional vernacular variants with encroaching Southern vernacular
variants. The selective, additive pattern uncovered with /ay/ seems to be con-
firmed by other emerging patterns we have observed in Ocracoke speech which
show that speakers may add vernacular features from surrounding dialects to
their speech; at the same time they maintain a select group of some distinct
island traits, in order to maximize the projection of island identity via vernacu-
larity in speech (Schilling-Estes & Wolfram 1994). The adoption of selected
features of external varieties is commonplace in threatened languages and is,
according to Dorian (1994), one of the more viable strategies for preserving
endangered languages, albeit in an altered form, in the face of external language
encroachment.
The pattern of [oy] recession and the selective, additive pattern of [a:] expan-
sion in Ocracoke suggest that obsolescing dialect forms do not simply represent
'ordinary' language changes that proceed at an accelerated pace. We would
expect linguistic variables associated with such ordinary changes to display a
more orderly correlation to external social factors such as age and internal
linguistic factors such as following phonetic environment than the unexpected
patterning we observe in the obsolescing Ocracoke [3y] and the incipient South-
ern American [a:]. Not only do the patterns we observe for hOy]and [a:] in
Ocracoke inform our understanding of obsolescing forms in language-death
situations, but they also lead us to refine our notions about the structured nature
of variability in language change. In the next section, we reformulate conditions
on the orderly progression of change, based on insights garnered from Ocracoke
[3y] and [a:].
2.2.3. EXPLAINING PHONETICALLY IMPLAUSIBLE CONSTRAINT HIERARCHIES. In
Table 3, we summarize the constraint orderings for /ay! variability with respect
to the following phonological environment. Boldface numbers indicate phoneti-
cally implausible constraint orderings in terms of the following phonological
environment-that is, constraint effect rankings for the variable phonetic var-
iants of /ay! that do not match the natural, expected tendencies established
empirically in other studies (e.g. Labov 1963, Bernstein & Gregory 1994, Feagin
1994).
How do we explain the apparent phonetic implausibility of the constraint
ordering for Southern unglided [a:] as it begins to compete with the traditional
VARIETY/ Lateral Nasal Voiced Voiceless

VARIANT Obstruent Obstruent
Ocracoke [oy] 4 2 1 3
Southern [a:] 1 2 3 4
Ocracoke [a:] 1 2 4 3
(1 = most favoring environment; 4 = least favoring environment)
TABLE3. Constraint Orders for Ocracoke [oy] and Southern la:]
variant? We suggest that there are three bases for this phonetic implausibility.
One is a shift in the peripherality of the nucleus of lay/. When Southern unglid-
ing enters the Ocracoke dialect, it affects /ay/ in just those contexts in which
the diphthong is most likely to have been realized in traditional speech with a
nonperipheral, more centralized nucleus-that is, before voiceless consonants.
Before voiced consonants, a favored environment for [3y], the /ay/ diphthong
is almost never unglided to [a:], perhaps because the nucleus is too far along
on its rotational path backward and upward to be readily unglided. Ungliding
affects diphthongs when they are near low central position where they are still
vulnerable to a shift to peripherality-not in the mid back area where the shift to
peripherality has already taken place.4 As we have seen, the socially unmarked
standard English raising of [a'] to [a'], a nonperipheral vowel, shows a mirror
image constraint order with that for the traditional Ocracoke and Southern
peripheral nuclei. We hypothesize that the transition between peripheral and
nonperipheral status, with their mirror image constraint hierarchies, may result
in a temporary nonnatural ordering of phonological constraints. We presume
that over time, as the new change is carried to completion, the natural constraint
order will override this unnatural one, but not necessarily in the initial stages
of the change.
Our second explanation relates to the nature of variation during change. Most
variationists now assume that variation takes place in a way conveniently de-
scribed as an S-slope curve (Weinreich et al. 1968, Bailey 1973, Labov 1994).
At the beginning and end points of change, the change slope is relatively grad-
ual, and one variant predominates. At the middle stages of the change, however,
the curve is sharp and relatively rapid, as variation among competing variants
is maximized. We hypothesize that the beginning and end points of a change
are more likely to display upheaval in the natural ordering of constraint effects
than the midpoint of a change. During the midpoints, we expect the most natural
phonological constraint orderings between alternating variants. At the end
points, however, the constraint orders may be less phonologically natural, par-
ticularly in cases of dialect mixture. At the end points, lexical constraints and
rote processing of variants are likely to be maximized, a situation conducive
for overruling natural phonetic tendencies.5 In other words, the environment
in which a given change occurs first may not necessarily be the most natural
phonological environment for this change; similarly, as a change reaches its
4 Thomas (1995) offers a quite different interpretation geared more toward explanatory percep-
tually based principles of phonetic change.
5 There are also implications about the orderly progression of change and the role of the lexicon
in change that seem related to the progression slope. For example, the role of the lexicon in
phonological change is more prominent at the incipient and cessation stages of a change than at
its midpoint. Part of the resolution of the ongoing controversy over regularity in phonological
change (Labov 1981) may be related to the trajectory slope of the change. lrregularity and lexical
diffusion are maximized at the beginning and the end of the slope and phonological regularity is
maximized during the rapid expansion in the application of the rule change during the midcourse
of change.
endpoint, it may begin to fade first in environments that ordinarily encourage

the occurrence of the changed segment. Acceleration and stagnation in change
have been recognized for some time in variation studies (cf. Bailey 1973, Fasold
1973); this unevenness in the progress of a change over time often is at its
highest at the beginning and end of the change. And there are other cases in
the literature (Fasold 1973, Wolfram 1988) where the end points of change show
patterns that are more marked than the more natural, unmarked constraint
hierarchies found at the mid-points of the S-curve change slope. Clearly, /ay/
ungliding to La:]in Ocracoke qualifies as an incipient change; it is characterized
by low frequency usage levels (less than 10%of all potential unglided /ay/ tokens
are actually unglided) and lexical constraints on its variability (for example, a
major item for ungliding is the lexical item Carolina [katrla:na]). In fact, there
is some evidence that the unglided variant may not even endure in Ocracoke,
if enough new islanders from non-Southern mainland areas (now estimated to
be over half of new islanders) take up residence on the island.
At the end points of change-for [a:] as an incipient change and for [3y]
as an obsolescing one-the natural phonological conditioning of variation is
apparently quite vulnerable. The resistance to the encroaching [a:] in Ocracoke
before voiced segments vis-a-vis other environments is somewhat phonologi-
cally unnatural. While the favored retention of [hy] before voiced segments
is not itself unnatural, the relative proportion of [oy] retention before voiced
segments as opposed to the level of retention in other environments is certainly
unnatural, as it indicates that the old variant is not simply fading in a linear
regression (see Fig. 3). With respect to language variability, then, obsolescing
forms may show some essential differences in patterning vis-a-vis healthy,
changing forms not situated at the end point of a change. We cannot simply
assume that obsolescing forms will recede in an orderly, linear fashion as has
sometimes been assumed in the literature on language death. Processes of
change in language death may very well differ from ordinary change processes
in ways other than sheer rapidity of progress (Dorian 1981, Schmidt 1985).
Finally, we appeal to a social basis for the peculiar constraint ordering on
the unglided [a:] variant in the Ocracoke dialect. The environment of a following
voiced obstruent, which favors the ungliding of [a:] in mainland Southern vari-
eties, is, in Ocracoke, the phonetic context most strongly associated with the
stereotypical brogue. Words like [soyd] side and [toyd] tide are an integral
part of the phrases Ocracokers-and outsiders-use to highlight the speech
differences between islanders and mainland Southerners. Not only are Ocracok-
ers known throughout North Carolina as 'hoi toiders,' but they often display
their dialect features to tourists with sayings such as 'It's hoi toide on the sound
soide.' In such 'performance' phrases, islanders tend to exaggerate the raising
and backing of [oy] (Schilling-Estes 1995). Such productions lead speakers to
expect the [3y] variant before voiced obstruents, and so they may resist using
the [a:] variant in this environment, even though they have begun using it before
voiceless obstruents. Strange shifts happen when people switch to a more ver-
nacular form of their dialect in the face of a standard or prestigious language
variety, and this may be one manifestation of this so-called SUBORDINATEDIA-

LECTSHIFT(Labov 1972).
We summarize the causes for the apparent implausibility of the ordering of
constraints affecting /ay/ in Ocracoke with a proposed principle of PHONETIC
which encompasses
IMPLAUSIBILITY, the three conditions discussed above that
can give rise to phonologically unnatural orderings of constraints on the pho-
netic realization of a particular variant.
(3) PRINCIPLEOF PHONETICIMPLAUSIBILITY:
Constraints on variability may
be altered in ways that violate the apparent phonetic naturalness of
constraint hierarchies if:
a. there is a significant categorial shift in the status of the unit, such as
a shift between peripheral and nonperipheral status for vowels
b. the change is taking place at a beginning or end point in the S-slope
change curve, thus involving incipient or obsolescing forms
c. the change involves competing socially significant norms at a con-
scious level
While none of the different explanations may, in itself, fully account for the
apparent phonetic implausibility in the ordering of constraints that affect the
replacement of the obsolescing traditional variant with Southern ungliding in
the Ocracoke community, certainly the convergence of the kinds of conditions
we discuss here could bring about a reordering of constraints on variability-a
reordering that runs counter to phonetic naturalness. If nothing else, an in-
depth examination of Ocracoke /ay! shows how simplistic models of language
change and dialect mixture are inadequate if we hope to achieve a full under-
standing of how and why languages change. It also shows how the examination
of change at its end points-both beginning and terminating-should inform
our understanding of this dynamic. Moribund dialects and dying forms thus
provide an essential proving ground for developing and refining empirically
based models of language change.
2.2.4. THE SOCIOLINGUISTIC SETTING: THE SEMI-SPEAKER AND STYLISTIC

SHRINKAGE.The consideration of the broader social context of language use
and particular performance phenomena alluded to in the previous section pro-
vide further evidence for the moribund status of the Ocracoke brogue. Among
the present generation of ancestral islanders, we find an increasing core of
islanders who qualify as 'semi-speakers' in Dorian's (1981) classic defini-
tion-speakers who persist in using the dying-language variety instead of the
encroaching language despite the fact that they possess only selective and in-
complete mastery of the receding language. Such speakers may, in fact, be
more proficient in the encroaching-language variety. Because their creative
ability in the dying language is diminished, semi-speakers may resort to formu-
laic utterances or confine their use of the dying language to ritualistic, perfor-
mance-oriented speech events. This narrowing of contexts in which the dying
variety can be used, referred to as STYLISTICSHRINKAGE(e.g. Campbell & Munt-
zel 1989:195) has been documented in dying-language areas throughout the
714 LANGUAGE, VOLUME 71, NUMBER 4(1995)
world. The speech of semi-speakers is further characterized by its metalinguis-

tic focus; semi-speakers tend to talk about their dying language far more than
do fluent speakers of healthy varieties.
In Ocracoke, 'dialect semi-speakers' are concentrated in a close-knit group
of middle-aged speakers and some younger speakers. These middle-aged speak-
ers, who grew up during the first wave of increased outside influence on the
island, express overt pride in the traditional Ocracoke brogue, and they lament
its passing, often remarking that the brogue is fading among younger speakers.
They are also quite aware of their speech as an object of curiosity and intrigue
to outsiders, and they tend to exaggerate salient features of the brogue, espe-
cially the Loy] variant discussed above, when in the presence of outsiders.
Exaggeration also occurs in stock performance phrases, such as that in the
following passage. In this exchange, a forty-year-old fisherman and carpenter
is explaining his initial meeting with Walt Wolfram to a fieldworker who had
become his friend.
(4) Subject: I got him [Walt Wolfram] going with that 'hoi toide
on the sound soide.
Fieldworker: What did he say to that? Did he get all excited?
Subject: Oh my God, yeah. Came out there, said, 'I'm studying
speech.' And I said, 'Well, it's hoi toide on the
sound soide. Last night the water fire; tonight the
moon shine. No fish. What do you suppose the mat-
ter, Uncle Woods?' Well, he got a laugh out of that.
He did.
This performance phrase is replete with the most marked phonetic features
of the brogue, including the unique [Oy]production discussed at length above,
the pre-palatal raising of [I] in fish flI's], the syllable restructuring of fire as
[far], and the fronting and raising of both the nucleus and glide of /aw/ in sound
[send]. Further, the phonetic distinctiveness of [Oy] is noticeably exaggerated
in performance phrases such as the above (Schilling-Estes 1995). The contrast
between the production of [oy] in nonperformance speech and its production
in performance contexts can be seen in Figure 4, in which the height and back-
ness of the nucleus of /ay! in the performance speech of the speaker quoted
above is compared with the height and backness of the /ay! nucleus in nonper-
formance speech. Height is indicated along the y-axis, by the frequency of the
first formant of each vowel token; backness is indicated along the x-axis by F2
minus Fl values.
Not only do Ocracoke semi-speakers maintain salient features of the tradi-
tional brogue, but they also skew the natural patterning of their speech variety
by adding vernacular forms from neighboring dialect areas. For example, as
discussed above, the middle-aged poker players in our sample use more un-
glided /ay/ than other middle-aged speakers, upsetting the orderly generational
progression toward steadily increasing [a:] usage we otherwise find. In addition,
the poker players are the only speakers in our sample who use nonstandard
ii'asn't (e.g. youlwely'all/they wasn't) rather than using weren't throughout the
F2-F1
1900 1700 1500 1300 1100 900 700 500 300
I I t _
I I --t----+- -l------- --T300
- 400
f - 500
Mean Values for Unraised, . 0 * Conversationallay/(Unraised) |

Raised and Performance layl . 0o * Conversationallay (Raised)
VARIANT F2-FI Fl A Performance/ay/
Unraised 502.4 630.9 700

Raised 464.6 594.5
Performance 439.2 565.2 800
(RO, 39-year-old male)

1 900
FIGURE 4. Performance and Nonperformance Productions of I|y1
negative past BE paradigm (e.g. l/you/(s)he/elely'calllthey weren't), or using a

mixture of generalized weren 't and a few tokens of standard *wasn't (e.g. I/(s)he
wasn't) (Schilling-Estes & Wolfram 1994). Even the youngest speakers do not
use nonstandard 'asn't. The exaggerated use of [a:] and the use of nonstandard
wsasn'tare suggestive of the performative speech of semi-speakers rather than
reflective of changes taking place in nonperformance speech in Ocracoke,
which is characterized by a lesser degree of La:] usage and no nonstandard
rasn't use.
Further, some middle-aged and younger speakers now use the traditional
Ocracoke dialect chiefly for performance, whether for telling humorous stories
about fishermen or for displaying the oddities of the language itself. The tradi-
tional Ocracoke dialect has almost become too 'quaint,' too much of an object
of curiosity-or an OBJECTLANGUAGE(Tsitsipis 1989:121)-to be used in many
situations in which the focus is not on language itself. The sociolinguistic situa-
tion in Ocracoke, along with the sociohistorical background and the linguistic
change observed within the Ocracoke community, clearly support our conten-
tion that the Ocracoke brogue in its traditional form is receding rapidly and
therefore should be classified as an endangered language variety.
3. SOCIALRESPONSIBILITY
AND ENDANGEREDDIALECTS.The relatively short
history of sociolinguistics has shown that it is quite possible to combine a com-
mitment to the objective description of sociolinguistic data and a concern for
social issues. At the same time that sociolinguistic researchers have contributed
substantively to our understanding of language variation, they have become
involved in several important social and socioeducational issues related to dia-
lect diversity over the past several decades.
According to Labov (1982), there are two primary principles that may moti-
vate (socio)linguists to take social action, namely, the PRINCIPLE OF ERROR COR-
RECTION and the PRINCIPLE OF DEBT INCURRED. These are articulated as follows:
(5) PRINCIPLE OF ERROR CORRECTION: A scientist who becomes aware of
a widespread idea or social practice with important consequences
that is invalidated by his own data is obligated to bring this error
to the attention of the widest possible audience (Labov 1982: 172).
(6) PRINCIPLE OF DEBT INCURRED: An investigator who has obtained lin-
guistic data from members of a speech community has an obligation
to use the knowledge based on that data for the benefit of the com-
munity, when it has need of it (Labov 1982: 173).
There are several outstanding instances in the history of social dialectology
where these principles have been applied. In the 1960s, sociolinguists took a
prominent prodifference stance in the so-called deficit-difference controversy
that was taking place within education and within speech and language pathol-
ogy (Baratz 1968, Labov 1969, Wolfram 1970). Consonant with the principle
of error correction, sociolinguists took a united stand against the classification
and treatment of normal, natural dialect differences as language deficits or disor-
ders. There is little doubt that sociolinguists played a major role in pushing the
definition of linguistic normalcy toward a dialect-sensitive one, although the
practical consequences of this definition are still being worked out in many
clinical and educational settings.
In keeping with the principle of debt incurred, social dialectologists also rose
to the occasion in the celebrated Ann Arbor Decision (1979). Linguistic testi-
mony was critical to the judge's ruling in favor of a group of African-American
children whose parents brought suit against the Ann Arbor, Michigan, Board
of Education for not taking their dialect into account in reading instruction
(Farr-Whiteman 1980, Smitherman 1981). In effect, the judge ruled that the
defendants had failed to take appropriate action to overcome language barriers,
in violation of Title 20 of the U.S. Code, Section 1703 (f).6 In compliance with
the judge's ruling, a series of workshops was conducted to upgrade awareness
and to apply sociolinguistic expertise in reading instruction.
These relatively well known cases of social involvement are not the only
ones in which the principles of social action articulated by Labov have been
applied. There are many other instances where linguists have been involved in
individual or corporate cases related to language equity in the workplace, the
educational system, and society at large. For the most part, however, the social
role assumed by linguists in these cases has been that of REACTIVE ADVOCACY,
where the linguist responds to a social inequity by providing sociolinguistic
evidence. In the typical scenario, a linguist who has conducted research in a
6
The relevant portions of this Code are as follows:
No state shall deny equal educational opportunity to an individual on account of his or her
race, color, sex, or national origin by. . .
(f) the failure by an educational agency to take appropriate action to overcome language
barriers that impede equal participation by its students in its instruction programs.
particular community is called upon or feels obligated to respond to some erro-

neous sociolinguistic assumption or conclusion about the language of the com-
munity, as the language is threatened sociopolitically or socioeducationally.
We believe, however, that investigators should make a stronger social com-
mitment to the language communities that have provided them data and that
this role is particularly appropriate with respect to endangered dialects. This
level of commitment is more proactive, in that it involves the active pursuit
of ways to return linguistic favors to the community. Thus, the PRINCIPLEOF
LINGUISTIC GRATUITY was proposed (Wolfram 1993b: 227).
(7) PRINCIPLE OF LINGUISTIC GRATUITY: Investigators who have obtained
linguistic data from members of a speech community should actively
pursue positive ways in which they can return linguistic favors to
the community.
For endangered dialects, there appear to be several ways we can take a
proactive role that is consonant with the principle of linguistic gratuity. Al-
though we cannot necessarily revitalize a moribund dialect, we can work with
community members (1) to ensure that the dialect is documented in a valid and
reliable way; (2) to raise the level of consciousness within and outside the
community about the traditional form of the dialect and its changing state; and
(3) to engage representative community agents and agencies in an effort to
understand the historic and current role of dialect in community life. In these
respects we follow the lead of those who have been engaged in preserving
endangered languages (e.g. Watahomigie and Yamamoto, Craig, in Hale et al.
1992).
3.1. A PROACTIVE DIALECT PRESERVATION PROGRAM. Our involvement with
the community in Ocracoke attempts to serve positive humanistic, scientific,
and sociohistorical purposes and involves institutions, agents, and curricular
programs. The local institutions involved in the dialect awareness programs
implemented in Ocracoke include the Ocracoke Preservation Society, the Ocra-
coke School, and various North Carolina museums such as the Outer Banks
museum in Manteo. Key community members have also served as active partic-
ipants in various phases of our program, including the president of the Ocracoke
Preservation Society, community leaders, and teachers in the school.
A set of products has been developed in support of our documentation and
dissemination goals. One of the products we have compiled for the community
is an audio tape of representative speech samples from our cross-generational
speakers, who range in age from ten to eighty-two. This archival tape includes
almost forty different speakers, which is about 20% of all ancestral islanders.
It will be deposited in several different places, including the Preservation Soci-
ety, the local school, and historical museums in North Carolina, so that the
dialect will be documented and preserved along with other physical and cultural
artifacts of Outer Banks culture.
We have also written a book about the brogue aimed at residents of and
visitors to Ocracoke (Wolfram & Schilling-Estes 1996). The book will be made
available in gift shops, museums, and other trade stores along the coast of
North Carolina,as well as on the mainland.We attemptto provide in readable

but nontechnicalformat historical and descriptive informationabout the dia-
lect.7 We also confront stereotypes about the dialect which range from the
mainstreamopinionthatthe brogue,as a vernacularvariety, is simplyan unwor-
thy corruptionof standardEnglishto the romanticnotion that it is 'Elizabethan
English' preserved intact on Ocracoke since the 1700s. In addition, we hope
to raise dialect awareness by capitalizingon the value that islanders place on
their historicalheritage.Islandersare quite knowledgeableabout theirgenealo-
gies and are conscious of some unique relic lexical items that have become
symbolictokens of islandquaintness.A separate,self-containeddialect lexicon
that we have compiled with considerableassistance from a variety of islanders
has helpedconcretize some of this historicallexical heritage.We even designed
a T-shirtwith a numberof dialect lexical items displayed on the back and the
inscription'Save the Brogue' on the front to highlightthe variety's endangered
state. We initially distributedthe shirt among participantsin our study and
schoolchildren in the dialect awareness curriculumwe implemented in the
school. The T-shirt has proven to be so popularamong islanders and tourists
that it is now sold by the Ocracoke HistoricalPreservationSociety, with pro-
ceeds going to the society.
Withthe cooperationof the educationalsystem, we produceda dialect aware-
ness curriculumfor the OcracokeSchool. The curriculumengages students, as
a part of their regular study of North Carolina history, in the study of the
structureand history of their dialect (Wolframet al. 1995).The curriculumhas
been implementedin the eighth grade social studies programwhich is focused
on state history. The programinvolves studentsnot simplyas passive observers
of languagevariationbut as collectors, collators, and analysts of the Ocracoke
brogue. For example, in some exercises, students are asked to formulatehy-
potheses about the structuraldetails of particulardialect forms. In another,
they are asked, as studentethnographers,to take partin the collection of dialect
lexical items, as well as in the analysis of the generationaldistributionof these
words.
Finally, we have produced a twenty-five minute documentaryvideo titled
'The Ocracoke Brogue' (Blanton & Waters 1994) so that local residents and
tourists can view for themselves an informativeportrayalof the state of the
dialect. This video documentaryis designedfor public, popularviewing as well
as for educationalpurposes.
The venues we use to disseminate informationinclude both traditionaland
nontraditionalagencies. For example, we have instituted experimental pro-
gramsin the OcracokeSchool and made presentationsto the OcracokeHistori-
cal PreservationSociety. In fact, our dialect awareness materialsare endorsed
by the society, and we will share royalties from book sales with them in the
hopes that our programsand materialsare considered as recognized preserva-
7 The book is aimed at the visitor whose curiosity about the dialect is piqued by a visit to the
island-the typical tourist-as well as islanders who desire to know something about their dialect
legacy. Writing popular, informative accounts for such an audience is not an easy task, but the
authors have profited greatly from Schilling-Estes's experience as a journalist before her career
in linguistics.
tion efforts. Along with these institutional efforts, we have shown our documen-
tary several times at the local bar and grill, Howard's Pub, where both local
residents and tourists typically congregate for informal socializing. These show-
ings resulted in animated, positive discussions about the dialect by both ances-
tral islanders and tourists. The threatened status of the Ocracoke brogue has
also been the subject of several local and regional television and radio news
programs, and there were at least seven major feature articles in local and
regional newspapers from 1993 to 1995 that focused on the state of the dialect
and the threats to its survival. Several of these stories were accompanied by
sound bites; readers were invited to call an advertised telephone number and
listen to a recorded sample of the brogue firsthand. The Virginian Pilot news-
paper reported that more readers called in to hear the brogue than any other
sound bite they had yet made available to their readership.
We have attempted to engage students, teachers, and community members in
collecting and documenting dialect data. Several community members routinely
make observations and take notes for us about lexical items; and one school-
teacher, following one of our lessons, returned the following day with over two
pages of lexical items and phrases elicited from her elderly relatives. On another
occasion, she wrote a poem composed of many unique Outer Banks lexical
items in celebration of the dialect. The level of community support and coopera-
tion we have observed throughout the course of this effort, on an institutional
and personal level, is probably unprecedented in the history of American En-
glish dialect study.
The fate of the Ocracoke brogue is still undecided, and circumstances beyond
our and even the community's control probably will determine its ultimate
destiny. But at least the community is becoming informed as never before about
its dialect heritage and the linguistic-cultural stakes involved in its demise. And
if nothing else, we have documented for the historical record the status of the
once-vibrant brogue. We hope that our being there has made a difference and
that we have given something in return for our seemingly inane and intrusive
sociolinguistic probing. In the process of conducting our sociolinguistic study
for scientific purposes, we have become partners with the community in raising
awareness about the dialect in its traditional form and also in its current, mori-
bund state.
REFERENCES
BAILEY,CHARLES-JAMES N. 1973. Variation and linguistic theory. Washington, DC.:
Center for Applied Linguistics.
BARATZ,JOAN. 1968. Language and the economically deprived child: A perspective.
ASHA 10.143-45.
BERNSTEIN,CYNTHIA,and ELIZABETH GREGORY.1994. The social distribution of glide
shortened /ai/ in LAGS. Paper presented at SECOL 50, Memphis, TN.
BLANTON,PHYLLIS,and KARENWATERS(producers). 1994. The Ocracoke brogue: A
video documentary. Raleigh, NC: The North Carolina Language and Life Project.
CAMPBELL,LYLE, and MARTHAC. MUNTZEL.1989. The structural consequences of
languge death. Investigating obsolescence: Studies in language contraction and ob-
solescence, ed. by Nancy C. Dorian, 181-96. Cambridge: Cambridge University
Press.
SANKOFF. 1974
HENRIETTA, and DAVID SANKOFF
CEDERGREN, HENRIETTA
CEDERGREN 1974. Var ab e ru
Variable rules:
es Per
Performance
ormance as
a s ca re
a sstatistical ec on oof compe
reflection ence Language 50
competence. 50.333-55.
333 55
CHOMSKY MORRIS HALLE. 1968
NOAM, and MORRISHALLE
CHOMSKY, NOAM 1968. The sound pa ern oof Eng
pattern sh New York
English. York:
Harper and Row
Row.
WOLFRAM; and NANJO DUBE
DONNA; WALT WOLFRAM
CHRISTIAN, DONNA
CHRISTIAN DUBE. 1988
1988. Var
Variation
a on and change
in
n geograph
geographically so a ed commun
ca y isolated communities:
es ach an Eng
Appalachian
Appa sh and Ozark Eng
English English
sh
Pub ca on oof the
(Publication he Amer can D
American Dialect
a ec Soc 74). Tusca
e y 74
Society oosa AL
Tuscaloosa, AL: Un vers y oof
University
abama Press
Alabama
A Press.
DORIAN,NANCYC
DORIAN NANCYC. 1981 1981. Language dea death: cyclee oof a Sco
h The lifee cyc Scottish
sh Gae
Gaelicc d
dialect.
a ec
ade ph a PA
Philadelphia,
Ph PA: Un vers y oof Pennsyv
University an a Press
Pennsyvlania Press.
--. ed 1989
(ed.) 1989. Inves ga ng obso
Investigating escence S
obsolescence: Studies
ud es in
n language
anguage con rac on and obso
contraction obso-
lescence.
escence Cambr dge Cambr
Cambridge: Cambridgedge Un vers y Press
University Press.
1994. Pur
--. 1994 Purism
sm vs
vs. comprom
compromise se in
n language
anguage rev a za on and language
revitalization anguage rev
revival.
va LanLan-
guage in n Soc e y 23
Society 23.479-94.
479 94
FARR-WHITEMAN,
FARR WHITEMAN MARCIA
M ARCIAed 1980. Reac
(ed.) 1980 ons too Ann Arbor
Reactions Arbor: Vernacu
Vernacularar B ack En
Black En-
sh and educa
gglish education.
on Wash ng on DC
Washington, DC: Cen
Centerer for
or App
Applieded LLinguistics.
ngu s cs
FASOLD,RALPHW
FASOLDRALPH concept oof ear
1973. The concep
W. 1973 er a er More or less
earlier-later: ess correc
correct. New ways
oof ana yz ng var
analyzing a on in
variation n Eng sh ed
English, ed. by Char es James N
Charles-James N. Ba ey and Roger W
Bailey W.
Shuy, 183
Shuy 183-97.
97 Wash ng on D
Washington, D.C.:
C George
Georgetown
own Un vers y Press
University Press.
--. 1990
1990. The soc o ngu s cs oof language.
sociolinguistics anguage Cambr dge MA
Cambridge, MA: BasBasil B
Blackwell.
ackwe
FEAGIN,CRAWFORD
FEAGIN CRAWFORD. Long i' as a m
1994. 'Long
1994 crocosm oof Sou
microcosm Southern a es speech
hern sstates speech. Paper
ed aat NWAV 23
presented
presen 23, S an ord CA
Stanford, CA.
GUY, GREGORYR
GUY R. 1993
1993. The quan a ve ana
quantitative analysis ngu s c var
ys s oof linguistic variation.
a on Amer
Americancan
dialect
d research, ed
a ec research Denniss R
ed. by Denn R. Pres on 223
Preston, 49 Ph
223-49. ade ph a PA
Philadelphia, PA: John BenBen-
jamins.
am ns
KEN; MICHAEL KRAUSS
HALE, KEN
HALE YAMAMOTO;COLETTE
WATAHOMIGIE;AKIRA YAMAMOTO
KRAUSS; LUCILLE WATAHOMIGIE
ENGLAND. 1992
JEANNE; and NORA ENGLAND
CRAIG; LAVERNE MASAYESVA JEANNE
CRAIG 1992. Endangered
anguages Language 68
languages. 68.1-42.
1 42
HILL, JANE H
HILL H. 1989
1989. The soc unc ons oof re
a functions
social a v za on in
relativization n obso escen and non
obsolescent non-
obsolescent
obso escen languages.
anguages In Dor
Dorian,
an 149-66.
149 66
HOENIGSWALD,
HOENIGSWALD HENRY
H ENRYMM. 1989
1989. Obso
Obsolescence anguage h
escence and language s ory Ma
history: ers oof linear-
Matters near
eve ng loss,
y leveling,
ity, ke In Dor
he like.
oss and the an 347
Dorian, 347-54.
54
IHALAINEN, OssI
IHALAINEN 1991. Per
OssI. 1991 Periphrastic
phras c n aaffirmative
do in rma ve sentences
sen ences in he d
n the a ec oof Eas
dialect East
Somerset. D
Somerse a ec s oof Eng
Dialects sh S
English: ud es in
Studies n gramma ca var
grammatical a on ed
variation, ed. by Pe
Peter
er Trud
Trud-
ggill and JJ. K
K. Chambers
Chambers, 148 160 New York
148-160. York: Longman
Longman.
RUTH. 1989
KING, RUTH
KING 1989. On thehe soc a mean
social meaning ngu s c var
ng oof linguistic ab y in
variability anguage dea
n language death
h
ssituations:
ua ons VarVariation
a on in n New
Newfoundland French. In Dor
ound and French an 139
Dorian, 139-48.
48
KURATH, HANS
KURATH HANS. 1939
1939. Handbook oof the ngu s c geography oof New Eng
he linguistic and Prov
England. Provi-
dence, RI
dence RI: Brown Un University.
vers y
LABOV, WILLIAM
LABOV 1963. The soc
WILLIAM. 1963 a mo
social motivation change. Word 19
va on oof sound change 19.273-307.
273 307
1969. The logic
. 1969 og c oof nons andard Eng
nonstandard sh George
English. Georgetown anguage and
own monographs on language
ngu s cs 22
linguistics ed. by James A
22, ed 44 Wash
a s 11-44.
Alatis, Washington, DC: George
ng on DC own Un
Georgetown University
vers y
Press.
Press
. 1972
1972. Some pr
principles ngu s c me
nc p es oof linguistic hodo ogy Language in
methodology. n Soc e y 11.97-154.
Society 97 154
. 1981
1981. Reso v ng the
Resolving he Neogrammar
Neogrammarian an con roversy Language 57
controversy. 57.267-309.
267 309
. 1982
1982. Ob ec v y and comm
Objectivity men in
commitment ngu s c sc
n linguistic ence Language in
science. n Soc
Society
ey
11.165-201.
11 165 201
1991. The three
. 1991 hree d a ec s oof Eng
dialects sh Quan
English. a ve ana
Quantitative yses oof sound change in
analyses n
progress,
progress ed. by Pene
ed Penelope
ope Eckert,
Ecker York: Academ
44 New York
11-44. Academicc Press.
Press
1994. Pr
. 1994 nc p es oof linguistic
Principles change: In
ngu s c change Internal ac ors Cambr
erna factors. Cambridge, MA: Bas
dge MA Basil
Blackwell.
B ackwe
M1LROY,LESLEY.1980
M1LROYLESLEY 1980. Language and soc a ne
social works Ox
networks. Basil B
ord Bas
Oxford: Blackwell.
ackwe
ROBINS, ROBERTH
ROBINS M. UHLENBECK (eds.)
ROBERTH., and EUGENIUS M 1991. Endangered
eds 1991 languages.
anguages
Oxford:
Ox ord Berg
Berg.
DAVID, and WILLIAM LABOV
SANKOFF, DAVID
SANKOFF he uses oof var
1979. On the
LABOV. 1979 ab e ru
variable es Language
rules.
in
n Soc e y 88.189-222.
Society 189 222
Th s con en down oaded om 130 179 16 201 on Wed 11 Nov 2015 23 22 02 UTC
A use sub ec o JSTOR Te ms and Cond ons
SCHILLING-ESTES, NATALIE. 1995. Production, perception, and patterning: 'Perfor-

mance' speech in an endangered dialect variety. Penn Review of Linguistics 18, to
appear.
--, and WALTWOLFRAM.1994. Convergent explanation and alternative regularization
patterns: Werel/weren't leveling in a vernacular English variety. Language Variation
and Change 6.273-302.
SCHMIDT,ALBERT.1985. Young people's Dyirbal: An example of language death from
Australia. Cambridge: Cambridge University Press.
SMITHERMAN, GENEVA(ed.) 1981. Black English and the education of black children and
youth: Proceedings of the National Invitational Symposium on the King decision.
Detroit, MI: Wayne State University Press.
THOMAS,ERIK. 1995. Phonetic factors and perceptual reanalysis in sound change. Aus-
tin, TX: University of Texas dissertation.
, and GUY BAILEY. 1994. The origins of monophthongal /ai/ in Southern speech.
Paper presented at SECOL 50, Memphis, TN.
TRUDGILL,PETER,and J. K. CHAMBERS(eds.) 1991. Dialects of English: Studies in
grammatical variation. New York: Longman.
TSITSIPIS, LUKASD. 1989. Skewed performance and full performance in language obso-
lescence: The case of an Albanian variety. In Dorian, 139-148.
WEINREICH,URIEL;WILLIAMLABOV;and MARVINHERZOG.1968. Empirical founda-
tions for a theory of sound change. Directions for historical linguistics, ed. by
Winfred P. Lehmann and Yakov Malkiel, 95-198. Austin, TX: University of Texas
Press.
WOLFRAM,WALT. 1970. Sociolinguistic premises and nature of nonstandard dialects.
The Speech Teacher (September). 176-86.
. 1988. Reconsidering the semantics of a- prefixing. American Speech 63.247-54.
. 1993a. Identifying and interpreting variables. American dialect research, ed. by
Dennis R. Preston, 193-222. Philadelphia: John Benjamins.
. 1993b. Ethical considerations in language awareness programs. Issues in Applied
Linguistics 4.225-55.
, and DONNA CHRISTIAN. 1976. Appalachian speech. Washington, DC: Center for
Applied Linguistics.
, and NATALIE SCHILLING-ESTES. 1995. On the social basis of phonetic resistance.
The NWAV Proceedings. Stanford, CA: Center for the Study of Information and
Linguistics.
, and . 1996. Hoi Toide on the Outer Banks: The story of the brogue. Chapel
Hill: University of North Carolina Press, to appear.
-;--; and KIRKHAZEN. 1995. Dialects and the Ocracoke brogue. An eighth-grade
curriculum. Raleigh, NC: The North Carolina Language and Life Project.
--; --; --; and CHRISCRAIG.1996. The sociolinguistic complexity of quasi-iso-
lated southern coastal communities. Language variation in the South revisited, ed.
by Cynthia Bernstein, Tom Nunnally, and Robin Sabino. University, AL: Univer-
sity of Alabama Press, to appear.
WOODBURY, ANTHONY C. 1993. A defense of the proposition 'When a language dies,
a culture dies'. Texas Linguistic Forum 33: Proceedings of the first annual sympo-
sium on language in society, Austin, TX.
Walt Wolfram [Received 5 April, 1995;
Department of English accepted 20 June, 1995.]
North Carolina State University
Raleigh, NC 27695-8105
Natalie Schilling-Estes
Department of Linguistics
University of North Carolina at Chapel Hill
Chapel Hill, NC 27599

Linguistic Society of America Language

Uploaded by

Copyright:

Available Formats

Linguistic Society of America Language

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Linguistic Society of America Language

Uploaded by

Copyright:

Available Formats

Linguistic Society of America

1. WHY STUDY ENDANGEREDDIALECTS?Recent symposia (for example, the

Given the moribund condition of so many of the world's languages, it may

2. OCRACOKEENGLISHAS AN ENDANGEREDDIALECT. In this section, we pre-

2.1. THE SOCIOHISTORICAL

2.2. THE LINGUISTICBASIS. The constellation of structures that defines the

variable dimensions of particular receding forms as they are replaced by alter-

involved in fishing or other marine-related industries (such as working on the

SUBJECT VI. Obstr. Vd. Obstr. Nasal Lateral

Older .474 .267 .259

VL Obstruents .242 .240 .518

Input Probability .374 .082 .544

other dialects of American English, including the 'standard' variety. In these

surprising that Southern ungliding in Ocracoke is starting to make inroads, but

VARIETY/ Lateral Nasal Voiced Voiceless

endpoint, it may begin to fade first in environments that ordinarily encourage

variety, and this may be one manifestation of this so-called SUBORDINATEDIA-

2.2.4. THE SOCIOLINGUISTIC SETTING: THE SEMI-SPEAKER AND STYLISTIC

world. The speech of semi-speakers is further characterized by its metalinguis-

Mean Values for Unraised, . 0 * Conversationallay/(Unraised) |

VARIANT F2-FI Fl A Performance/ay/

Unraised 502.4 630.9 700

(RO, 39-year-old male)

FIGURE 4. Performance and Nonperformance Productions of I|y1

negative past BE paradigm (e.g. l/you/(s)he/elely'calllthey weren't), or using a

particular community is called upon or feels obligated to respond to some erro-

North Carolina,as well as on the mainland.We attemptto provide in readable

SCHILLING-ESTES, NATALIE. 1995. Production, perception, and patterning: 'Perfor-

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.