2.}:
1.9 in...

23‘.

.1 ':

In."
. g p

A:
q. ..

 

 

     

 

   
    
   
  

        

       

 

    

     

      

 

 

     

   

 

    

 

fiaxfifififim .mfifl?

. . v Vr. ..

E... , a...» 35...... 23%.332.93.332.33; ..3

      
          

      

.“fimfimflfi: . .35

 

,1! I: ..

91.nov,.

5.1.
I):

.5131
1:?

g
Hummwu.
t . aha.
Mumqwfi.
....‘.r_.w..wnr.uuw....
3N .nnummxf
....b...r¢.. r1.
:ififlfx 2".

 

. 13 .1
$113.1: 9
21.1.1):
. .3.-
I»!
.50.;

l‘
l-n‘

. -, 1..
mun... . .1
{I

.3:
m

5
3!

 

.wsfldhfl‘\:
351....m3m}
qumnhv
(32‘. .. :1
Oils-20.061!

“In! :-

 

ug”: -
hll'tflh'i'l

”I”

 

:‘fvflggy’
unarm-

      

 

 

 

 

 
        

 
    
 

          

               

        

 

   
  

 

éégéiwmfi . mm. ..

. .. . «1.2.

 

 

émfifiifia . : .

   

3333.333...

:. l

 

 

 

 

NNER [TV

:3. n'i'mi‘iiiiiimuufiuuxiiiii

3 1293 01405 7438

 

 

 

This is to certify that the

thesis entitled

OREGONIAN PERCEPTIONS OF AMERICAN
REGIONAL SPEECH
presented by

Laura C. Hartley

has been accepted towards fulfillment
of the requirements for

M.A. degree in Linguistics

  
 
 

 

  
 

l
[119.4‘7 '4‘
.1434"

“K

   
 
 

  

Date 3,51%]95;

0-7639 MS U is an Afiirmaa've Action/Equal Opportunity Institution

 

 

LIBRARY

Michigan State
University

 

 

 

PLACE It RETURN BOX to roman this checkout from your rocord.
TO AVOID FINES rotum on or baton dot. duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/EM Opportunity lmtlmtion
nuns-9.1

 

; 7 -—-———————__—

OREGONIAN PERCEPTIONS OF
AMERICAN REGIONAL SPEECH

By

Laura C. Hartley

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Linguistics

1996

ABSTRACT

OREGIONAN PERCEPTIONS OF
AMERICAN REGIONAL SPEECH

By

Laura C. Hartley

Although much is known about the varieties of American speech
from a linguistic standpoint, less work has been done which examines
folklinguistic attitudes and beliefs about those varieties. "Perceptual
dialectology” studies have been done in several areas of the United States;
however, no research to date has examined the perceptions of west-coast
residents.

In this study, the attitudes of Oregon residents towards American
speech are examined through the use of hand-drawn maps and ratings of
”degree of difference,” "correctness," and ”pleasantness." The results of
the ratings data are analyzed primarily using Chi-Square Tests of
Independence, Multi-Dimensional Scaling, and K-Means Cluster Analysis.

Results of the study indicate that overall, residents of Oregon exhibit
a good deal of linguistic security. They tend to agree, in general, with the
perceptual dialectology of respondents from other geographical areas,

although there are a few noticeable differences in their ratings.

ACKNOWLEDMENTS

I am very fortunate to be able to study under and work with Dr.
Dennis Preston. His enthusiasm, interest and support in this project have
been invaluable, particularly as we have explored together new ways of
analyzing perceptual dialectology data. I am very aware of the fact that not
all graduate students are able to work with an advisor who is both a mentor
and a friend.

Much thanks goes to Dr. A. Sue Harris and Lynn Landweer and the
students of the Sociolinguistic Survey class at the University of Oregon
Summer 1995 session for their help in this project: Troy Bailey, Rich
Berger, Donna Berger, Naomi Dobbins, Deb Duncan, Marijane Elliott,
Tina Fiddelka, Aric Haley, Janna Hopkins, and Michele Miles.

Grateful acknowledgment is made to the following authors and
publishers for the rights to reprint figures from copyrighted material:
University of Michigan Press for Figure 8.1 (reproduced as Figure 1, p. 3)
from American Regional Dialects, by Craig M. Carver, 1989; Walt
Wolfram for Figure 4—15 (reproduced as Figure 2, p. 4) from Dialects and
American English, 1991.

Finally, I would like to thank my husband, Ben Hartley, for his
support and encouragement throughout the duration of this project and my
graduate school career. His confidence in me is a continual source of

strength as I work to fulfill the vocation to which I have been called.

iii

TABLE OF CONTENTS

LIST OF TABLES ........................................................................... vi

LIST OF FIGURES ......................................................................... vii

CHAPTER 1

INTRODUCTION ............................................................................. 1
1.1 Language Attitudes Research ................................................... 1
1.2 The Linguistic Situation in the United States .............................. 2
1.3 Language Attitudes and Folk Linguistics ................................... 5
1.4 Perceptual Dialectology Studies in the US. ............................... 6

CHAPT ER 2

METHODOLOGY ........................................................................... 1 1
2.1 Data Collection .................................................................... 11
2.2 Research Tools ..................................................................... 11
2.3 Quantitative and Qualitative Analysis ...................................... 13

CHAPTER 3

RESULTS ....................................................................................... 16
3.1 Hand-Drawn Maps ................................................................ 16
3.2 Degree of Difference ............................................................ 27
3.3 Correctness .......................................................................... 35
3.4 Pleasantness ......................................................................... 43

CHAPTER 4

DISCUSSION .................................................................................. 53
4.1 Perceptions of Western States ................................................ 53
4.2 Perceptions of Midwestern States ........................................... 59
4.3 Perceptions of Northeastern States ........................................ 63
4.4 Perceptions of New York City, New York and New Jersey ....... 65
4.5 Perceptions of Southern States ............................................... 67
4.6 Effects of Gender, Age, and Visit Status ................................. 70
4.7 Comments on Research Methodology ..................................... 75
4.8 Conclusion ........................................................................... 77

iv

APPENDIX A

RESEARCH TOOLS ........................................................................ 80
APPENDIX B
STATE NAME ABBREVIATIONS USED ........................................ 86
APPENDIX C
SAMPLE HAND-DRAWN MAPS ..................................................... 87
APPENDIX D
LABELS USED ON HAND-DRAWN MAPS ..................................... 92
APPENDIX E
RATINGS (RAW NUMBERS) .......................................................... 98
APPENDIX F
SELECT TRANSCRIPTS OF TAPED INTERVIEWS ...................... 103
LIST OF REFERENCES ................................................................ 114

LIST OF TABLES

Table 1. Frequency of Identification of Dialect Regions .................... 16
Table 2. Degree of Difference Ratings for Illinois by Age Group ....... 27
Table 3. Degree of Difference Ratings for Indiana by Age Group ...... 28
Table 4. Degree of Difference Ratings for Arizona by Visit .............. 29
Table 5. Degree of Difference Ratings for Kentucky by Visit ............ 29
Table 6. Degree of Difference Ratings for Massachusetts by Visit ...... 29
Table 7. Chi-Square Results for Degree of Difference Ratings ........... 31
Table 8. Correctness Ratings for North Carolina by Age Group ......... 35
Table 9. Correctness Ratings for South Carolina by Age Group ......... 35
Table 10. Correctness Ratings for South Dakota by Age Group ......... 36
Table 11. Correctness Ratings for Tennessee by Age Group .............. 36
Table 12. Correctness Ratings for Texas by Age Group .................... 36
Table 13. Correctness Ratings for Virginia by Age Group ................ 37
Table 14. Correctness Ratings for Maine by Visit ............................. 37
Table 15. Correctness Ratings for Nebraska by Visit ........................ 38
Table 16. Chi—Square Results for Correctness Ratings ....................... 39
Table 17. Pleasantness Ratings for Alabama by Gender ..................... 43
Table 18. Pleasantness Ratings for Kansas by Gender ........................ 43
Table 19. Pleasantness Ratings for Louisiana by Gender .................... 44
Table 20. Pleasantness Ratings for South Dakota by Gender .............. 44
Table 21. Pleasantness Ratings for Texas by Gender ......................... 45
Table 22. Pleasantness Ratings for Minnesota by Age Group ............. 45
Table 23. Pleasantness Ratings for New Jersey by Age Group ............ 46
Table 24. Pleasantness Ratings for Rhode Island by Age Group ......... 46
Table 25. Pleasantness Ratings for Alaska by Visit ............................ 47
Table 26. Pleasantness Ratings for Kansas by Visit ........................... 47
Table 27. Chi-Square Results for Pleasantness Ratings ...................... 48
Table 28. Degree of Difference Ratings (Raw Numbers) ................... 98
Table 29 Correcness Ratings (Raw Numbers) .................................. 99
Table 30. Pleasantness Ratings (Raw Numbers) .............................. 101

vi

LIST OF FIGURES

Figure 1. Major American Dialect Regions based on Word Geography

Data ............................................................................................. 3
Figure 2. Major American Dialect Regions based on Vowel Systems ...4
Figure 3. Results of Hand-Drawn Maps of a "South" ......................... 18
Figure 4. Results of Hand-Drawn Maps of a ”Northeast" ................... 19
Figure 5. Results of Hand-Drawn Maps of a "Midwest" ..................... 20
Figure 6. Results of Hand-Drawn Maps of a "West" .......................... 21
Figure 7. Results of Hand-Drawn Maps of a "Pacific Northwest” ....... 22
Figure 8. Results of Hand-Drawn Maps of a "California” .................. 23
Figure 9. MDS Analysis for Degree of Difference Ratings ................ 32
Figure 10. MDS Analysis of Degree of Difference Ratings with

K-Means Clusters ........................................................................ 34
Figure 11. MDS Analysis of Correctness Ratings with K-Means

Clusters ...................................................................................... 42
Figure 12. MDS Analysis of Pleasantness Ratings with K-Means

Clusters ...................................................................................... 52
Figure 13. Oregonian Perceptual Speech Regions Summarized .......... 78

vii

1 INTRODUCTION

1 . 1 Language Attitudes Research. For the past three
decades, research on "language attitudes," i.e., beliefs about and
perceptions of varieties of language and the people who speak those
varieties, has had a substantial place in the literature of a number of fields,
including Sociolinguistics, Social Psychology and Communications. The
methods used in this research have typically fallen in one of three
categories: content analysis of societal treatment, such as examining laws
and policies regarding language use; direct measurement through questions
regarding the desirability of language varieties and self-report of language
use; and indirect measurement, primarily through the use of techniques
such as matched guise (Ryan, Giles & Sebastian, 1982).

Language attitude research is an effective tool in uncovering
stereotypes and beliefs that one group of people hold about another.
People often disguise the attitudes that they hold towards those who are
different from them by making claims about the language varieties that
those others use. "Attitudes toward particular varieties are then taken to be
attitudes towards the speakers of those varieties" (Ryan, Giles & Sebastian,
1982, p, 2). Various kinds of language attitudes research have been
conducted to investigate in- and out-group feelings and beliefs about
language varieties along a number of social dimensions, including ethnicity,
socio-economic status, gender, and region.

From a linguistic standpoint, all language varieties are equally
" good" as linguistic systems (T rudgill, 1974). From a social standpoint,
however, this is not the case. Particularly in areas with widespread

literacy, the elevation of one variety to the category of "standard", with the

1

2
subsequent attribution of the category or label "non-standard" to other

varieties often results in a popular belief that the standard is the ”correct",
”proper", or "educated” way of speaking. This in turn may lead to an
evaluation of non-standard varieties as "incorrect”, "rough" or "ignorant”,
especially (but not exclusively) by those who are native speakers of a
standard variety. This is often the case with US. varieties of English,
thanks especially to prescriptive grammar books used in elementary and
secondary education.

An individual who grows up speaking the standard variety of a
language will likely have a great deal of "linguistic security", i.e. their
attitudes towards their language variety will be generally favorable (Labov,
1966). Those who grow up speaking a non-standard variety, however,
may experience a great deal of ”linguistic insecurity” as a result of being
told, either directly or indirectly, that the way they speak is wrong or

ignorant.

1.2 The Linguistic Situation in the United States. The
varieties of American English from a linguistic point of view have been
defined primarily in terms of either lexicon (Carver, 1987) or phonology
(Labov, 1991). The massive LAUSC (Linguistic Atlas of the United States
and Canada) project begun in the 19303, identifies language regions based
on data collected and compiled in regional linguistic atlases (cf. Reed,
1957; Atwood, 1962; Kurath, 1949; Bright, 1971; Allen, 1973-76;
Pederson et al, 1986). Further data was collected during fieldwork from
1965 to 1970 for the DARE (Dictionary of American Regional English)
project (Cassidy, 1985). The compilation and publication of this data is

still in progress. Although data was collected in some areas of the Pacific

3
Northwest (primarily Washington state) for the LAUS C and DARE

projects, there has not been a great deal of analysis of the data which has
been published, compared to other regions. The information which is
available, primarily in the form of word geographies, (Reed, 1956, 1957,
1961) identifies the Pacific Northwest, which includes Washington, all but
southernmost portion of Oregon, northern Idaho and the westernmost

portion of Montana, as a distinct dialect area (see F1 gure 1).

 

 

 

 

 

 

 

 

 

W
m
In.
m
. ..
. x
1 3; W — .. m
. out-our \ W"
M' flaunt-n
can-uh mu
M
~7-
lu-

 

 

 

Figure 1. Major American Dialect Regions based on Word Geography
Data

In contrast to Reed's primarily lexical data, Labov (1991) classifies
the major dialect areas of the US. based on a careful study of the overall
vowel systems used in different regions. In Labov's study, the Pacific
Northwest is included in a much larger region, called for convenience
simply ”West". This region is characterized by the relative stability of
most of the vowels (unlike the "North" and "South", which are

characterized by different systematic vowel rotations in progress),

4
combined with a complete merger of the low back vowel /3/ with /a/, so

that words such as cot and caught are homonyms (both pronounced as
/kat/). Labov claims that the merger is found in an area extending north
into eastern New England from Boston and westward from Pittsburgh,
through the traditional Midland area of Ohio, Indiana, and Illinois, then
extending northward and southward to encompass most of the traditional

West (see Figure 2).

 

 

 

o - Northern Cities Vowel Shift
33:? - low vowel merge

- Southem Vowel Shift

 

 

 

 

Figure 2. Major American Dialect Regions based on Vowel Systems

Despite popular notions, there is no single variety of American
English which could properly be labeled "the" standard variety (Preston,
1993b). It is true that when questions regarding usage arise, people will
often turn to grammar books as an authoritative guide to "correct" or
"proper" English. These texts establish the rules for a variety of English

which would be most accurately labeled Formal Standard English or

5
Prescriptive Standard English (Wolfram, 1991). Formal Standard English

is most often used, however, in written rather than spoken language.

When it comes to spoken English, it is more accurate to say that each
region of the US. supports its own standard, which can vary from other
regional standards in terms of phonology, lexicon and even grammar.
Nevertheless, the popular belief that there is a "correct" way of speaking

(in addition to writing) American English is pervasive.

1 .3 Language Attitudes and Folk Linguistics. Early
language attitudes studies which used actual speech samples to elicit hearer
responses generally made the assumption that what was being rated was the
speech itself. Preston (1993b), however, points out that later studies done
with monolingual speakers suggest that nonlinguistic features may be
equally involved in shaping the perceptions of speakers. Because of this, it
is important to investigate both the linguistic reality and the folk
perceptions of dialect areas.

This concern for understand folk linguistic reality is by no means a
new undertaking. Rensink (1955) reports briefly on data collected in 1939
in The Netherlands in which respondents were asked to point out areas
where people speak dialects that were different than their own. Grootaers
(1959) included questions related to perceptual dialectology in a survey he
conducted in one area of Japan. Based on the data he collected, Grootaers

concluded that

the dialect consciousness of the average speaker has no linguistic
fundament. It is based essentially on an elusive feeling, fostered by
community life, it is of essentially transient nature, because a change
in village administration, if lasting approximately for one human life
span, suffices to give it new shape (p. 384).

6
Building on the methods used by Rensink and Grootaers, Preston

examined the perceptual dialectology of some regions in the US. through a
series of studies with residents from Hawaii (1982), Indiana (1985, 1993b),
western New York and New York City (1986), and Michigan (1993b).
Other researchers have utilized methods similar to Preston's to study the
”perceptual dialectology” of other countries, including Japan (Inoue, 1986;
Long, 1996), Germany (Dailey-O'Cain, 1994, 1996), Wales (Coupland,
Williams & Garrett, 1994), Turkey (Demirci, 1996), France (Kuiper,
1996), and Spain (Moreno, 1996).

1.5 Perceptual Dialectology Studies in the US. Preston's
perceptual dialectology research reveals a general lack of differentiation
among the varieties of speech in western states by non-Western residents,
with the exception of California often being singled out as a separate dialect
area. This is the case in research which compares composites of hand-
drawn maps, in which respondents circle and label areas where they believe
"people talk alike,” done with residents of New York City, Hawaii,
southern Indiana, western New York, and southeastern Michigan (1986).

In this study, composite maps from all five areas differentiated California
as a separate dialect area. All respondents except those from Hawaii also
drew a ”West”, although the boundaries of this area varied somewhat.
Only respondents from western New York designated a "Northwest"
region, which included all the states from Oregon to Minnesota. In terms
of other regions, all five groups differentiated regions called
”South/Southern", "Northeast/New England", "Texas", "North/Northern",

”Midwest" and "New York City." There was also a tendency to single out

7
the respondents' home state or area as a distinct dialect region; in fact, all

respondents except those from Michigan did this.

In further research in southern Indiana and southeastern Michigan,
Preston (1993b) also measures respondent ratings of the 50 states plus New
York City and Washington DC. in terms of ”correctness”, ”pleasantness”,
and "degree of difference." He concludes from this research that residents
of Michigan exhibit a great deal of linguistic security, reflected in their
extremely high rating of Michigan in terms of ”correctness” as well as high
ratings in terms of "pleasantness", although several other states are rated
equally pleasant. Residents from southern Indiana, however, demonstrate
some linguistic insecurity by rating several other areas higher in terms of
”correctness" but Indiana as highest in terms of "pleasantness." According
to Preston, ”these results suggest, further, that the preference for local
norms along affective lines is stronger in areas where there is linguistic
insecurity” (p. 35).

This conclusion is well supported by past language attitudes research
which has claimed that a language or language variety of a minority group
can become highly symbolic in terms of creating feelings of solidarity.

Ryan, Giles and Sebastian ( 1982), for example, make the following claim:

The language or dialect of one's family life, intimate friendships and
informal interactions acquires vital social meanings and comes to
represent the social group with which one identifies. One's native
language typically elicits feelings of attraction, appreciation and
belongingness. In situation where a group's identity is threatened,
the variety with which it is associated can become a key symbol of
the group's culture and identity (p. 9).

Because Preston's Indiana respondents are part of a stigmatized group in
American culture (i.e. they have a somewhat "southern" accent), their

language variety has become a means of creating in-group solidarity.

8
Although Preston's work has provided much helpful information in

understanding American perceptions of US. regional speech, a
comprehensive picture of the language attitude situation is not possible until
data has been collected from each of the major geographical and dialectal
areas of the United States. To date, no research has been done which
examines the perceptions of west-coast residents toward varieties of speech
in the US. This study helps to fill in this gap in the literature by focusing
on the perceptual dialectology of residents from one west-coast state,
Oregon.

Based on standard claims within the sociolinguistic literature, results
from Preston's studies, and my own experience and intuition as a native of
the west coast (although not a native of Oregon), the following hypotheses

are tested in this study:

HYPOTHESIS 1: Oregonians will differentiate varieties of speech
in western states to a greater degree than non-west coast residents,
although they will still outline a single major Western dialect region.
As far as other regions are concerned, Oregonians will not as finely
distinguish among states in southern and northern regions (east of the
Mississippi River) as do residents in those areas.

This hypothesis is based primarily on the tendency to identify the local area
as a distinct dialect region which Preston (1986) found in respondents from
other geographical regions. The prediction that they will still generally
include themselves in a greater Western dialect region is based on his claim
in the same article "that the local identity is not strong unless the area
supports some linguistic or other cultural caricature" (p. 230). The second
part of the hypothesis, which deals with other regions, is simply the flip-
side of the self-identification fact, i.e. Oregonians will not as strongly

identify the local areas of respondents from other geographical regions.

9

HYPOTHESIS A Oregonians will not generally distinguish
themselves from Washington, but may do so slightly from
California, due especially to caricatures of southern California (e. g.
Hollywood, ”valley” talk, and a large proportion of native Spanish-
speaking residents).

Since there is a relatively weak cultural caricature of Oregon and
Washington together based on the history of the Oregon territory and the
unique climate (i.e. "it always rains there") but a strong caricature of
California, as already described above, the self-identification tendencies

will serve to differentiate these areas.

HYPOTHESIS 3: Oregonians will exhibit a strong degree of
linguistic security, reflected in high ratings for Oregon in terms of
both "correctness" and "pleasantness”, with equally high ratings for
other western states in the "pleasantness" category.

This hypothesis is based largely on my own native speaker intuitions about
the status of western language varieties, but also on the fact that western
speech varieties are not stigmatized and therefore should not create a
situation of linguistic insecurity.
HYPOTHESIS 4: New York City will be rated the lowest overall
for both "correctness" and "pleasantness". Southern states will be
rated low for "correctness” but slightly higher for "pleasantness".
Northern and New England states will be rated fairly high for both

"correctness" and "pleasantness", but not as high overall as Western
states.

These prediction are based on the specific cultural caricatures associated

with these regions, as well as on past studies such as Preston 1993b.

HYPOTHESIS 5: Exposure to language varieties in different states
may effect ratings slightly, but in general, stereotypes will be
stronger than contradictory experiential evidence, particularly in
highly caricatured areas such as New York City and the deep south.

10
This hypothesis is based on past sociolinguistic work such as Williams'

(1972) study with teachers in Texas in which ratings were found to be

much more a result of stereotypes than responses to actual linguistic details.

HYPOTHESIS 6: Gender may influence ratings slightly, but not
significantly overall. In terms of "pleasantness" and "correctness”,
women will have a tendency to rate more extremely overall. Age
may also have a small effect on ratings, but overall this effect will
not be significant.

This hypothesis is based on past language attitudes research which has
shown that in some cases, gender and age may have a significant effect on
ratings (e. g. Kramer, 1977; Newcombe & Amkoff, 1979; Paltridge &
Giles, 1984; Condon & Pittman, 1993).

In the following chapter, research tools used, respondent
demographics and methods of data collection are detailed. Chapter 3 gives
the results of the research. Finally, Chapter 4 provides an interpretation
and discussion of the results, discusses problems and issues related to this
study and perceptual dialectology research in general and offers

suggestions for further research.

2 METHODOLOGY

2. 1 Data Collection. The data were collected primarily from
residents living in the greater Eugene-Springfield area of Oregon. A small
number of residents of Portland also participated in the research.
Respondents were approached in a variety of public areas, such as outside
shopping areas, parks and on the University of Oregon campus. In
addition, permission to interview residents of a local senior citizens home
was obtained.

There were a total of 66 respondents, 32 males and 34 females. The
respondents ranged in age from 20 to 78, with the following number in
each of five age categories: 20-29, 8 respondents; 30-39, 13 respondents;
40-49, 18 respondents; 50-59, 14 respondents; 60+, 13 respondents. All of
the respondents were of European-American descent, with the exception of
one Native American. 27 respondents were lifetime residents of Oregon.
6 were born in other states, but have lived continuously in Oregon since
elementary school. 4 moved to Oregon during their high school years. 7
spent their school years in other west-coast states but have lived in Oregon
for most of their adult lives. Finally, 22 were raised and have lived in
Oregon for most of their lives, but have lived in other states or countries

for some part of their adult lives.

2.2 Research Tools. The following research tools were utilized
in the data collection (copies of each tool can be found in Appendix A).
These tools are basically the same as those utilized in Preston 1993b, with a

few exceptions, as noted below.

11

12
Maps - Respondents were first given a map of the US, including

state boundary lines but no state names. They were asked to draw
circles around areas ”where people talk the same" and to label those
areas. Respondents were encouraged to use their own descriptive
labels for this task.

Degree of Difference - Respondents were next given an alphabetical
listing of the 50 states plus New York City and Washington DC. and
asked to rate each area on a 4—point scale depending on how similar
the speech in that area sounds compared to the respondents' own
speech. For this task, respondents were given a US. map labeled
with state names to consult if they so chose.

Correctness - Respondents were given a similar alphabetical list and
asked to rate each state or city on a 7-point scale (Preston used a 10
point scale for this task) as to how "correct" the speech in that area
is. No definition or criteria for determining "correctness" were
provided by the researcher. If the respondents asked questions about
what was meant by "correct", they were told to use their own
judgments.

Pleasantness - Respondents were then asked to rate each state or city
on a separate sheet, on a 7-point scale as to how "pleasant” the speech
in that area is. (Again, Preston used a lO-point scale.) This task was
switched with the correctness task for half of the respondents in
order to eliminate overall any possible ordering effects for the two
tasks.

Exposure to other varieties - Respondents were asked to indicate
which states they have visited by placing a check mark next to the

state names on their last survey sheet. They were instructed to count

13
any state that they had ever been in, regardless of for how long they

had been there.

6. Interviews - For some respondents, data collected on the
questionnaires was explored in greater detail through short
interviews. In particular, these interviews were used to determine to
what extent individual experiences, general stereotypes, media

influence, etc. have influenced labeling and rating.

2.3 Quantitative and Qualitative Analysis. Data obtained
from the degree of difference, correctness, and pleasantness tasks were
first checked using a normal probability plot to determine if ratings were
normally distributed, an assumption of parametric statistical tests. It was
determined that, in fact, the data were not normally distributed, and
therefore needed to be subject to non-parametric statistical analysis.

To begin with, Chi-Square tests were run comparing the ratings of
the two gender groups and each age group for degree of difference,
correctness, and pleasantness for each state. When it was found that overall
these factors were insignificant (the exact results will be given in Chapter
3), the subsequent analysis ignored gender and age breakdowns.

Chi-square tests were then run comparing each state to every other
state in each of the rating tasks. Although these tests provided some
interesting results, they did not provide enough detail to allow strong
conclusions to be drawn. Because of this, the data were also subjected to
Multi-Dimensional Scaling (MDS) and K-Means Cluster Analysis.

MDS is "a procedure for fitting a set of points in a space such that
the distances between points correspond as closely as possible to a given set

of dissimilarities between a set of objects" (SYSTAT, 1992, p.94). The

14
output of an MDS run is a two-dimension plot with points scattered fairly

evenly throughout the space. The axes can be interpreted as principle
components or factor analysis. Clusters of objects or obvious patterns can
also be interpreted from the plot (SYSTAT, 1992).

Cluster Analysis in general is a multivariate procedure for finding
natural groupings in a data set. K-Means clusters are partitioned clusters,
rather than hierarchical, i.e. rather than detecting groups within groups
within groups, it splits the objects of analysis into separate, non-
overlapping clusters. It does this by calculating between- relative to
within-cluster variation until it has minimized the within-groups sum of
squares. It is important to note that K-Means clustering makes no
assumptions as to how many groups there are in the data set. The
researcher must specify the number of groups that should be calculated and
will likely need to try different numbers of clusters before the best analysis
is achieved (SYSTAT, 1992).

Although Preston and Howe (1987) developed a technique for
extracting computerized generalization from hand-drawn maps, Preston
(1996) questions whether the computerized results are really more
instructive than a qualitative analysis, particularly in the case of a relatively
small set of maps. Because of this, hand-drawn maps in this study were
analyzed in a more qualitative fashion, although some attempt was still
made to quantify the results. Regions that were circled reoccuringly were
identified. For each of these regions, the researcher determined what
percentage of respondents included various states within their boundaries
of the regions, producing a kind of overall layered perceptual map of each

region.

15
Finally, the tape recorded interviews were listened to and relevant

parts of the discussions were transcribed. The information obtained in this
way supplemented the survey data, although it was not extensive and was

not collected from every respondent.

3 RESULTS

3. 1 Hand-Drawn Maps. The following table shows general
regions differentiated by respondents in hand-drawn maps of areas "where
people talk alike." Regions are listed in order of frequency of
identification, with most frequently to least frequently identified. Areas
identified by fewer than 12% (8 respondents) are not included. Several
examples of hand-drawn maps are included in Appendix C.

Table 1. Frequency of Identification of Dialect Regions

 

-Region Identified # Respondents %4Respondents
identifying region identifying region

 

(n=65)
South 60 92.3
Northeast 49 75.4
Texas 36 55.4
Midwest 3 1 47.7
West 29 44.6
New England 24 36.9
California 17 26.2
Pacific Northwest 16 24.6
Hawaii 15 23. 1
Northern 14 2 1 .5
Southwest 12 18.5
Alaska 9 13.8
Plains & Mountains 9 13.8
West Coast 8 12.3
Louisiana/”Cajun” 8 12.3

 

Although Table 1 provides an overall frequency with which different

regions were identified in hand-drawn maps, it does not indicate which

16

17
states were actually included in each area. Since this varied some from

respondent to respondent, the best way to interpret overall perceptions of
each region was to tally which states were included in each region when
that region was identified by a respondent as a distinct dialect area. It is
then possible to differentiate ”core" states which make up an overall
perceptual region (i.e. those most commonly associated with that area)
from ”peripheral" states in a kind of layered map. Layered perceptual
maps are presented in the following pages for a number of the regions
given in Table 1.

Of the 60 respondents who circled a southern dialect region, all 60
included Alabama within this region, thus making Alabama the ”most
southern” state in the perceptual dialectology of the Oregonian respondents.
Between 56 and 59 respondents also included Mississippi, Georgia and
western South Carolina in their drawings of a "South." Thus, these four
states form what could be called the "Deep South" in the respondents'
perceptions (see Figure 3).

It is interesting to note the degree to which Texas and Florida are
included in a general "South" for the Oregon residents. Recall from Table
1 that Texas itself was identified as a distinct region by 36 respondents.
This map, however, reveals that almost as many respondents (29) simply
included Texas in the "South." Florida also gets incorporated into the
South rather than being singled out as a distinct dialect region, although its

degree of "southernness" decreases as you move from northern to southern
Florida.

l8

 

   
 
        
     

 

     

it.

.Efiaee

.......... iteeae

fifififi§§§§é

agateeett

.aeeeeee
$§§$§$$figg+

‘aaarma

'~.

 

Number of Respondents
Including State in Region

We?”
ot» t
3293‘ 21-30

  

 

 

 

Figure 3. Results of Hand-Drawn Maps of a "South"

For the Oregonian respondents, the "Northeast" clearly centers
around New York state, extending as far west as Wisconsin for some
respondents, although the majority draw the western boundary at Ohio (see
Figure 4).

Table 1 reveals that 24 respondents identified a "New England"
region in addition to or rather than a "Northeast". The "New England"
maps have as their focal area the states of Vermont, New Hampshire and
Maine. Those who did indicate a separate region generally referred to it as
"New England," although three respondents made reference to a Boston
accent. As Figure 4 shows, more respondents overall simply included the

New England states within a general "Northeast" rather than singling them

19
out in a separate dialect area. Also, New York City is incorporated in this

area, with very few respondents identifying it as a separate area, although

several respondents labeled the entire area as having a New York accent.

 

 

 

Number of Respondents
Including State in Region

 

  

11-20
1- 10

 

 

 

 

 

 

 

 

Figure 4. Results of Hand-Drawn Maps of a "Northeast"

20

 

- oo-
ooooooooooooooo
OOOOOOOOOOOOOOOO
oooooooooooooo
ooooooooooooooo
I

 

 

   

vvvvv
......
.......
IIIIII
OOOOOOO

 

 

 

 

 

 

3140 1:22:55: 11-20
§§§$§ 21-30 2:23 1-10

 

 

Figure 5. Results of Hand-Drawn Maps of a "Midwest"

Nearly half of the respondents (31) indicated that the "Midwest” was
a distinct dialect region. The center of this perceptual region is the state of
Iowa. For a few respondents, this area extended as far west as Idaho
and/or as far east as central Pennsylvania (see Figure 5).

As Table 1 shows, 14 respondents drew a "Northern" region in
addition to or instead of a Midwest. This northern region was centered on
Minnesota, and for a few respondents included only Minnesota. Also, 9
respondents designated a "Plains and Mountains" region, focused around
Wyoming, Montana, Colorado and sometimes the Dakotas. Finally, a small
number of respondents (6) also indicated a "Great Lakes" dialect region,

which included Wisconsin, Illinois, Indiana, Michigan and Ohio.

 

    

  
   
   

543‘
4' u.
g 3%

    

. .
r" 403:"
.

   
  
   

.. ’33.? ‘0
.‘ei’ei’oi’oro
39343153
gang

      
 

“=9- . .

  
    
 

F

  
    
 

33%:
333%
.6333?)

'3‘ ‘4
egg

‘8
...g? ‘

 
   
 

 

 
     
 
  

‘3’ o
§’#§#
+4. ..
«gorge.
M- ;
“or so

   

$0

4'
4o

"3:.
arc

     

3,.
1:.

   

      
         
     
   
 

1“}?
'33:?
its

  

Number of Respondents
Including State in Region

mm
9' 3.3.1.3: 21-30 1-10

11-20

 

 

 

   

 

 

 

 

 

Figure 6. Results of Hand-Drawn Maps of a "West"

A few of the respondents who did not draw a separate midwestem
region simply included many of the midwest states, and even some eastern
states in a large region, often labeled "Normal" or "sound the same as me"
(see Figure 6). Of the 29 respondents who identified the "West" as a
dialect region, most of them drew the boundary for this region between
Montana/Wyoming/ Colorado and the Dakotas/Nebraska/Kansas. In fact of
all the regions drawn, this area was the most uniform, i.e. the least
"layered." All of the states west of the boundary mentioned above were
included in this western dialect area by 25-29 respondents. In other words,
for those respondents who indicated a distinct "West" dialect region, there

was 86-100% agreement that all states west of the Dakotas should be

included in this region.

22

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

eeeeeee
eeeeeee

#2333: 11-16

eeeeeee

 

[215:3 1-10

 

 

 

 

 

 

 

 

 

fa

 

Figure 7. Results of Hand-Drawn Maps of a "Pacific Northwest"

Some respondents were more specific in their identification of
western dialect regions. As Figure 7 shows, 16 respondents drew a
”Pacific Northwest" area, which included primarily Oregon, Washington
and western Idaho. The splitting of California into two dialect regions is
reflected in both the perceptual boundaries of the Pacific Northwest, which
includes northern California for at least some respondents, and the dialect
region drawn for ”California” (see Figure 8).

Finally, a handful of respondents (8) drew a "West Coast" region,
rather than a general "West", a "Pacific Northwest" and/or a "California."
For these respondents, Washington, Oregon, California, Idaho, Utah and

western Montana and Arizona formed a separate dialect region.

 

 

 

 

 

 

 

 

   

 

 

Number of Respondents
Including State in Region

1147
'3'313. 1-10

 

 

 

 

 

 

 

 

Figure 8. Results of Hand-Drawn Maps of a "California"

In addition to the boundaries which were drawn, an analysis of the
types of labels used for the various regions sheds light on how each of the
areas is viewed in the perceptions of the Oregonian respondents. Building
on the analysis of labels from Preston's (1982) research with Hawaiian
respondents, the Oregonian labels can be divided into nine categories:
Geographical location, Variety descriptors, Evaluative terms, Other
languages/Countries/Ethnic groups, Sound/Tempo Qualities,
Topography/Nature, Cultural/Historical, Respellings, and Other. By far,
the majority of labels fell into one of the first four categories. A complete
listing of labels provided on the hand-drawn maps can be found in

Appendix D.

24
An comparison of the categories of labels used for each of the major

regions reveals some interesting generalizations. To begin with, the most
frequently employed category of labels was " geographical location". For
each of the four major regions (West, Midwest, Northeast, and South), this
category had more members than any other category. Thus it appears that
Oregonians divide up the US. primarily in terms of geography.

An analysis of the other categories provides more insight into how
these major geographical regions are perceived. The "evaluative terms"
category, which includes labels which suggest that some regions support a
"better" variety of English than others, was used almost exclusively to
describe western regions (including Alaska and Hawaii). The respondents
most often used the labels "Normal" or "No Accent" to describe the speech
in western areas. Other labels included "Plain Western", ”Commoners",
and "The same as me." Both the Midwest and Florida received only one
label each in this category ("less accent" and "very little accent"
respectively). Evaluative labels were completely absent for the
northeastern and southern regions.

With respect to this category, it is interesting to note the differences
between the Oregonian respondents and Preston's Hawaiian respondents.
Many of the Hawaiian respondents used the term "Standard" within their
labels. This term did not show up once in Oregonian labels. Preston

(1982) suggests that, in the Hawaiian respondents' perceptions,
while eastern and northeastern varieties may belong to some such

historically prescriptive variety known as "Standard,” they may not
belong to the category "normal," reserved for varieties farther west

(p. 39).
While it seems that Oregonians agree with Hawaiians that "normal" speech

belongs to western speakers, they do not seem to support the same notions

25
of a prescriptive standard variety which is the property of east coast

speakers.

The category of "variety descriptors" was the second most common
(after "geographical location”) for the Midwest, Northeast, New England,
South and Texas. While the Midwest and Northeast varieties were most
frequently called an ”accent", however, the varieties of speech in Texas and
the South were most often referred to as a "drawi." Other descriptors of
Northeast speech included "thick brogue", "mumbo jumbo" (used for
Washington DC.) and “speak w/ accent where they draw out vowels.” In
terms of southern speech, ”southem twang” and "Rebel slang" were each
used once.

Another category which was fairly frequently employed for all
regions by the Oregonian respondents was ”Other languages/Countries/
Ethnic groups." A closer examination of which languages/ethnic groups
are singled out, however, reveals a distinction between early immigration
in the settlement of the Northeast, Midwest and South and on- going
immigration in the West. The primary group singled out in the
(south)western areas is Spanish/Mexican, although one respondent also
referred to southeast Asian groups, using the label "Vietnamese/Laosian/
Hispanic" to describe southern California.

In contrast, the Midwest received labels such as "More Scandinavian
like", "Germanic lgs" and "Northern European." The Northeast and New
England were referred to as "More British like," "European", and "Older
English terms." The South received the labels "More Scottish like” and
"slower 'French' roll." Finally, Louisiana was primarily labeled with

terms which fell into this category, such as "French,” "Cajon" and
”English/French Creole."

26
Hawaii was also differentiated primarily through labels which fell

into the category of "Other languages/Countries/Ethnic groups." There
was a perception of influence on the speech in Hawaii by both the
indigenous island population (e. g. "Hawaiian natives - native accent") and
Asian ethnic groups (e. g. "Asian influence”, "Japanese Golfland”). Alaska
also had a few references to the native population ("Eskimo", "Alaskan
Indian influence"), although not as many as Hawaii.

Finally, Florida seems to be the other area where continuing
immigration seems to be a significant influence on the respondents'
perceptions of language varieties. The labels here were primarily
references to Cuban Spanish, although one respondent used the label "7-11
Turbins", suggesting a perception of a large number of middle eastern
immigrants holding lower level jobs (such as working in all night
convenience stores).

As far as the other categories are concerned, a few comments should
be made. "Sound/Tempo Qualities" were used most often to describe the
speech of the Northeast and New England. These included "harsh, talk
fast", "meaningless mumble", "Nasal" (used 3 times) and "clipped and
direct.” In contrast, three respondents referred to speech in the South as
"slow.” The Midwest received two labels in this category - "twangy,
brash" and ”Nasal." The speech in the Northwest, on the other hand, was
considered "Soft - TV like" by one respondent.

The category of "Cultural/Historical" labels tended to reflect either
stereotypes of certain areas or occupations. In the former group are labels
such as "Leisure living" and "Hip-Californian" for California, "Last

Frontier” for Alaska, "Hillbilly" for Tennessee and Kentucky, and "Back

27
Bay” for New England. Occupations show up in the Midwest "Central

Farmers" and the ”cowboys" of the Plains states and Alaska.

Finally, respelling are used only four times in this data, all for
highly stereotyped accents; twice for New York City ("New York, New
Yawk!" and "neuw yawk"), once for the Midwest ("Chicaco accent") and

once for the South ("suthron").

3.2 Degree of Difference. Chi-square tests were first run for
each state based on gender, age group, and whether or not the respondents
had ever visited the state. In the gender run, chi-square values were not
significant at the .05 level for any state. In terms of age, chi-square values
were significant at the .05 level only for the states of Illinois and Indiana.
Tables 2 and 3 show the ratings of these two states in terms of age group.
Figures are percentages of respondents within each age group who gave the

corresponding ratings. The rating scale is as follows:

1 = people there sound like me

2 = people there sound a little different from me

3 = people there sound very different from me

4 = people there sound so different from me I can't understand them

Table 2. Degree of Difference Ratings for Illinois by Age Group

 

 

 

 

Age Group 20-29 30-39 40-49 50-59 60+
% Giving Rating (n=8) (n=13) (n=18) (n=13) (n=13)
l 0.0 23.08 27.78 38.46 38.46
2 75.00 46.15 72.22 61.54 61.54
3 25.00 30.77 0.0 0.0 0.0

 

 

Chi-square = 17.1—73 with d.f. = 8

28
Table 3. Degree of Difference Ratings for Indiana by Age Group

 

 

Age Group 20—29 30-39 40-49 50-59 60+
% Giving Rating (n=8) (n=13) (n=18) (n=13) (n=13)
l 25.00 15.38 22.22 38.46 46.15
2 37.50 76.92 72.22 61.54 53.85
3 37.50 7.69 5.56 0.0 0.0

 

 

Chi-square = 15.848 with d.f. = 8

For both of these states, the significant chi-square values (17.173, with 8
d.f. and 15.848, with 8 d.f. respectively) are likely a result of the overall
harsher rating of younger respondents (in the 20-29 and 30-39 age
groups).

Related to whether or not respondents had visited each state, chi-
square values were significant only for the states of Arizona, Kentucky and
Massachusetts. Tables 4-6 show the distribution of ratings, in terms of
percentages of respondents within each group who gave the corresponding
ratings.

For both Arizona and Kentucky, significant chi-square values (6.697,
2 d.f. and 6.007, 2 d.f.) are a result of the more favorable ratings (in terms
of being similar to the respondents) given to the states by those who have
actually visited them. In the case of Massachusetts, however, (chi-square =
10.146, 3 d.f.) the opposite is the case. Those who have actually visited
this state rated it more different from their own speech than those who

have not visited it.

29

Table 4. Degree of Difference Ratings for Arizona by Visit

 

 

—Group Have Not Have Visited
% Giving Visited State State
Rating (n=16) (n=45)

1 56.25 84.44
2 37.50 15.56
3 0.0 0.0
4 6.25 0.0

 

—Chi-square = 6.69Wirh d.f. = _2

Table 5. Degree of Difference Ratings for Kentucky by Visit

 

 

-Group Have Not Have Visited
% Giving Visited State State
Rating (n=43) (n=18)

l 0.0 0.0

2 20.93 50.00
3 76.74 44.44
4 2.33 5.56

 

Chi-square = 6.007with d.f. = 2

Table 6. Degree of Difference Ratings for Massachusetts by Visit

 

 

-Group Have Not Have Visited
% Giving Visited State State
Rating (n=48) (n=12)

1 8.33 0.0

2 50.00 25.00
3 39.58 50.00
4 2.08 25.00

 

Chi-square = 10.146 with d.f. = 3

30
Since significance in terms of age groups and visit status was found

in only a very small number of states (and in different states for the two
factors), subsequent statistics were run on the entire population, without
regard for the factors of gender, age group and visit status.

Chi-square tests were run comparing the degree of difference ratings
for each state against every other state. Table 7 shows the result of this run
in terms of which states were npt significantly different from others at the
.05 level. (State names are abbreviated using standard postal abbreviations.
See Appendix B for list of abbreviations used.)

This chi-square run reveals sets of states for which the ratings were
basically identical, i.e. the states were not significantly different from
exactly the same other states. These sets are Alaska, Arizona, and
Montana; Califomia, Colorado, Nevada, and Utah; Connecticut, Maine and
West Virginia; Georgia and Mississippi; Illinois and Indiana; Maryland and
Rhode Island; Massachusetts, Missouri and New Hampshire; Michigan,
Minnesota and Ohio; Nebraska and North Dakota; Tennessee and Texas.
Furthermore, it provides a general picture of clusters of states. These
clusters become even more apparent through the use of Multi-Dimensional

Scaling (MDS) analysis, as shown in Figure 9.

31

Table 7. Chi-Square Results for Degree of Difference Ratings

 

State States which were not significantly different

 

DDDDDDDDDD

Emma mac

v w
w A

.V
mwv v VN
..w w mm
mm“ a mm
am. A... w. a.
mmm mwm m we
mm man A . as
M.Am ..flM fl w .W
WAHHC .Nmmum m Emma

MMWWWH .KKY NARA
.mmm-..mm.-- Rm.

MAAmmoGDDDNNGMM AMED A

wmmLLmanaaallaaa “mam

mammxmmmmmmm mmwmmwm

w v
A w v
N m V M m
a m a VW A
mA . m WQ .
.V m .S W
«A . A x Km .m
QR m v m. D“ RD
mm m m mm may an
“mam m a. se mmw my
LKmML M.mmcc ALm LH
.ޣ N WMKHNN INA NW
mmm . wwwmwmmmmm m
a. “mm as. .W.
. .J. JLVMM W.
WWN mm L NNtss mm
. . . .S.WS.S. . . . .W.”
Y .mwm SWMWW .mmmmwo
SWK .WmmK AKYMYY mWMJM.%
KmKAAm NnK AXKMflflflmnfi
MHMWWGMLMMNMM mmmmmNM

9 9 9 9

9 9 9 9

umanmmDamr .mmmamacmLA

wmmommmmmmmmmw

:mzz
mwmw

MW

 

32

 

IE

IFL IOK éR 13:1ch
ITNIC I “L

UI‘ IAK
NM IMO .533“ IMS'LA

‘—
d -_
-

I1.5

NVI MT ND I I IIN
MN IDE
NYC

-1 . 5 CA - 1 OH
.NE Im P: INID INH-va .NY
'RI 'CT IMA

- ._ IVT
0- 5 IWDC ”“5 INJ

 

 

-1.5 J.

 

 

Figure 9. MDS Analysis for Degree of Difference Ratings

33
MDS analysis provides a graphic picture of how the states cluster

together along two dimensions. Although the horizontal dimension is
fairly easy to interpret, something like similarity to Oregon speech, the
vertical dimension is less clear and is apparently working primarily to
differentiate Hawaii from all other states. It may be a "kind of accent" or
even a ”pleasantness of accent” dimension, since the southern states are
clearly differentiated from the northeastern states, with western and
midwestem states clustering around the zero point on this dimension.

Although the MDS analysis provides a better picture of state clusters,
it is still difficult to tell which clusters border states such as New Mexico
and Pennsylvania belong in. Calculating a K-Means Cluster clarifies the
situation. Figure 10 shows the MDS analysis with the K-Means Cluster
analysis superimposed in the form of circles.

K-Means cluster analysis provides both an idea of which states fall
together into different groups as well as which states belong on their own.
In this case, Hawaii, Florida, Missouri and New Jersey did not fit well into

the other ratings groups.

34

 

 

 

 

 

-1-..

 

 

-1.5 -L

 

 

Figure 10. MDS Analysis of Degree of Difference Ratings with K-Means
Clusters

3.3 Correctness.

35
In terms of correctness ratings, the same

statistical procedures were used as with the degree of difference data.

Preliminary chi-squares tests of the factors of gender revealed no

significant differences in correctness ratings for any of the states.

Within

the factor of age group, the states of North Carolina, South Carolina, South

Dakota, Tennessee, Texas and Virginia were rated significantly differently

by respondents of different ages, as Tables 8-13 reveal. The rating scale is

1=least COITCCt t0 7=most COITBCI.

Table 8. Correctness Ratings for North Carolina by Age Group

 

 

 

 

Age Group 2029 3039 4049 573—59 60+
% Giving Rating (n=8) (n=13) (n=16) (n=13) Q=12)
l 0.0 0.0 0.0 7.69 16.67
2 0.0 23.08 18.75 0.0 0.0
3 12.50 15.38 6.25 0.0 25.00
4 25.00 46.15 25.00 30.77 41.66
5 62.50 15.38 25.00 30.77 0.0
6 0.0 0.0 6.25 23.08 0.0
7 f 0.0 0.0 “3er 7.69 16.67
hT-square = 38.023 with d.f. = 24

Table 9. Correctness Ratings for South Carolina by Age Group

 

 

 

A—figeGroup 20-29 3039 40.49 5639 60+
% Giving Rating (n=8) (n=13) (n=16) (n=13) (n=12)
1 0.0 0.0 0.0 7.69 8.33
2 0.0 7.69 31.25 0.0 0.0
3 12.50 30.77 6.25 7.69 25.00
4 12.50 30.77 12.50 23.08 50.00
5 75.00 23.08 18.75 23.08 0.0
6 0.0 0.0 12.50 30.77 8.33
7 0.0 7.69 18.75 7.69 8.33

 

(Tm-square = 43.124 with d.f. = 24

36

Table 10. Correctness Ratings for South Dakota by Age Group

 

 

 

 

 

 

 

 

 

 

 

Age Group 2029 3039 4049 50—-59 60+
% Giving Rating (n=8) (n=13) (n=16) (n=13) (n=11)
l 0.0 0.0 0.0 0.0 0.0
2 0.0 0.0 0.0 0.0 9.09
3 12.50 0.0 12.50 15.38 0.0
4 25.00 46.15 18.75 0.0 27.27
5 50.00 53.85 31.25 15.38 45.45
6 0.0 0.0 12.50 46.15 9.09
7 _12.50 0.0 25.00 23.08 9.09
Chi-square = 33.437 with d.f. = 20
Table 11. Correctness Ratings for Tennessee by Age Group

T—georoug 20-29 30-39 4049 5039 60+
% Giving Rating (n=8) (n=13) (n=16) (n=13) (n=12)
l 0.0 0.0 0.0 0.0 0.0
2 0.0 23.08 25.00 15.38 0.0
3 0.0 23.08 25.00 0.0 33.33
4 75.00 38.46 12.50 15.38 50.00
5 25.00 15.38 25.00 23.08 8.33
6 0.0 0.0 0.0 30.77 0.0
7 _ 0.0 0.0 18.75 15.38 8.33

Chi-square = 39.323 with d.f. = 20
Table 12. Correctness Ratings for Texas by Age Group

'Z—georoug 20-29 30-39 4049 50-59 60+
% Giving Rating (n=8) (n=13) (n=16) (n=13) (n=12)
1 12.50 7.69 0.0 0.0 0.0
2 12.50 15.38 31.25 23.08 0.0
3 0.0 23.08 6.25 0.0 25.00
4 37.50 46.15 18.75 7.69 66.67
5 37.50 7.69 18.75 38.46 0.0
6 0.0 0.0 6.25 23.08 0.0
7 0.0 0.0 18.75 7.69 8.33

 

(SE-square = 41.581 with d.f. = 24

 

37
Table 13. Correctness Ratings for Virginia by Age Group

 

 

 

Age Group 2029 3039 4049 5'0‘59 60+
% Giving Rating (n=8) (n=13) (n=16) (n=13) (n=12)
1 0.0 0.0 0.0 0.0 0.0
2 0.0 7.69 0.0 7.69 0.0
3 12.50 0.0 37.50 0.0 8.33
4 25.00 53.85 12.50 23.08 33.33
5 25.00 38.46 25.00 23.08 50.00
6 25.00 0.0 6.25 30.77 0.0
7 12.50 0.0 18.75 15.38 8.33

 

Chi-square = 31.858 with d.f. = 20

In terms of visit status, significant differences in correctness ratings
were found between those who had visited the state and those who had not
only for the states of Maine and Nebraska. Tables 14 and 15 show the

distribution of the ratings for these two states.

Table 14. Correctness Ratings for Maine by Visit

 

 

Eroup Have Not Have Visited
% Giving Visited State State
Rating (n=47) (n=9)

1 0.0 11.1 1
2 6.38 22.22
3 6.38 0.0

4 21.28 1 1.1 l
5 25.53 22.22
6 27.66 0.0

7 12.77 33.33

Chi-square = 12.827Lwith d.f. = 6

 

38
Table 15. Correctness Ratings for Nebraska by Visit

 

 

-Group Have Not Have Visited
% Giving Visited State State
Rating (n=35) (n=22)

1 0.0 0.0

2 0.0 13.64
3 22.86 0.0

4 22.86 9.09
5 22.86 27.27
6 20.00 31.82
7 l 1.42 18.18

 

—Chi-square .-. 12.5'7'3 with d.f. T:

Given the fact that the ratings of so few states were affected by the
factors of age group and visit status, subsequent statistical analysis was
again done on the respondent ratings as a whole. Chi-square test
comparing each state with every other state are given in Table 16. Sets of
states which are not significantly different from each other and from
exactly the same other states are reported together.

Once again, these chi-square ratings provide a very general picture
of how states are grouped together in terms of correctness. These groups,
however, are even larger than with the degree of difference ratings, and
are therefore that much less informative. It is very difficult to tell, for
example, whether the West, Midwest, and Northeast should all be grouped
together. Even within the southern states, it is difficult to tell whether the
states of the ”outer south" belong in the same group as the "deep south."
MDS analysis combined with K-Means cluster analysis once again helps

clarify the picture.

, MT, NE, NV, NH,
, MN, MT, NE. NV, NH, NM,

39

Table 16. Chi-Square Results for Correctness Ratings

,IL, ME, MD, MA, MI, MN, MT, NE, NV, NH,

, WDC. WA, WY
, CA, CT, DE, ID, IL, IN, IA, ME, MD, MA, MI, MN, MT, NV, NH, ND,

OR, PA, RI, UT, VT, WA, WY

AL, FL, GA, HI, LA, MS, NJ, NYC, NY, NC, SC, TN, TX, WDC

States which were not significantly different
AR, GA, KY, LA, MS, NJ, NYC, NY, NC, TN, TX

AK

State(s)

CA:
CO:

 

 

 

 

 

 

 

 

. . . m . . Q .. . . . .
M m mm v0A w m m m w Mn mm m M m Mwmm
m m .. MMw . w . . R my .0 m m x m mm

. . A .W m .Wm m . v m Ms . . 0 .Mn.
m w 6. mm. m M.m .mm wme m WWW m wmmw

9V: 9 ll 99 99 99 9 999 9 9
wwmwm. mm. . mm» “mm “We mm mmmmw mwmm

.1. ..SH . H W .VW V.,. N. S A. h .hI. .AW“
wwmmm. m . . mw .wm rm. up ammw. mw ..

.A...fl. ..w W .mWYO.. C. m. KK. MWN.W +0.. AVn
mwmmm. m c. . m .wmmm s.s.A s.o s.A.s.mJ. mm .

... w K.s m .me.w. KKJ KR KVKWN mew
mmmA.c om. . mA ww.m bAm A6 ArArm mAn.

.WJVWS CS. WW.NDQEM. CJ. JD JVJV. .VEW
kat m NKM ..MWNWN.M fimm mN mmmmm thu

.v.vm. {3.mmm.. «va0. 00.0.0 .Mvss
s.A.NA Mlk AJmmAm.m NJm JN JDJDS 1.x.
K.WI.W .P. .N.O. A.WD .V .m . C.D D . DSDSM LW .x
A.Wumx @In 1.Jum.m.5 WAC Am J.J.. J.m.
mmneem Nee m ameniww Lem AM nwemmvnsme
memwue new 1.emmm.mM mam emwem.emwamem
memmm. mmm 6gmmmmmem mmmwmwwmmmenmmmmm
mflmbmm mmmm m bmmmpm m mMmmv mammmabmm

9H9 99 999 999 991999 999 9 9 9 99 99
3.6666 ma meme.memanwmm.wmwmewe.gwe.em
m.mmmwmenewame.e.46mmeamma..a.a.mmama
ewa.memnem.e.emm.m.A.Afl..fl.memewm em.”
mmmmamwenemnmemew mnmnmmmnmwnmnmnmmwnm

. 6
.. M. mew 0.. ..
mmnmmmmmmmnmmmwmmmm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A.

, DE, ID, IL, ME, MD, MA, MI. MN, MT, NE, NH, NM.

.VT.

 

 

,WV,\VI

AL, AR, FL, GA, KY, LA, MS, MO, NJ, NY, NC, OK, SC, TX, VA, WDC, WV
AL, AR, FL, GA, HI, KY, LA, MS, MO, NJ, NYC, NY, NC, OK, SC, TN, VA,

WDC, WV

S, MO, NE, NH, NJ. NM.
9 VA9

, MS, MO, NE, NH, NJ, NM, NY,
, WV, WI
, IA, KS, KY, ME,
. ND. OH. OK, PA,

.MD.

. ND. OK. PA, RI, SC, S
0, CT, DE, HI, ID, IL, IN, IA,
, ND, 0 , OR, PA, RI, SD,

, DE, FL, HI, ID, IL, IN, I
C

9
9

CA

9
9

, AR, FL, GA, HI, KY, LA, ME, MA, MS, MO, NJ, NM, NYC, NC, OK, SC,

TX, WDC, WV
. NM.

FL, GA, HI, IL, KS, KY, LA, ME

AK, AZ, CA, CO, CT, ID, MT, NV, OR, UT, WY

AL, AR, GA, KY, LA, MS, NJ, NY, NC, SC, TX
NC, ND, PA, RI, S , SD, TN, TX, VT, VA,

AK, AZ, CA, CO, ID, NV, UT, WA

CT, DE, FL, HI, IL, IN, IA, KS,

NM, N , ND, OH, OK, PA, RI,

FL, GA, HI, IL, KS, KY, LA,

NY, NC, ND, OH, OK, PA, RI, SC, SD,

AK, AZ,CT
NH N

AL
TN
AK. AZ
NE.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NYC:
NC:
OH:
OK:
OR:
SC:
VA:
WA:

NJ:

 

 

 

 

Table 16 (cont'd).

Aum

. HI. ID. IL, IN, I
, DE, FL, HI, ID, IL,
, NM, ND, OH, PA. RI. 3

, ND, OH, OK, PA, RI, SC, 8
0.

C

A

C

9
9

AZ, CA, CT, DE,

,AZ

9 9

AK
NE
AK
MT

 

 

41
The solid circles in Figure 11 represent the K-Means cluster analysis

for correctness if 10 clusters are used. When this is done, Oregon and
Washington together falls out as a separate cluster, as does Florida. If 9
clusters are chosen, Oregon and Washington are included in a larger
West/Midwest group. When only 8 clusters are used, Florida falls in with
the southern states.

The northeastern states form a cluster, with a few noticeable
exceptions. These include Michigan being part of this cluster, and
Massachusetts, Washington DC. and New Jersey forming their own
individual groups. Also, New York City and New York are grouped
together as a separate cluster. Finally, as in the degree of difference data,
Hawaii also is singled out in its own group. Unlike the degree of
difference data, however, the vertical dimension seems to be playing a
larger role than simply differentiating Hawaii from mainland states. It is

not obvious, however, as to what this dimension might actually be.

42

 

 

 

 

 

 

 

Figure 11. MDS Analysis of Correctness Ratings with K-Means Clusters

43
3.4 Pleasantness. Ratings of pleasantness were the only data

which showed any gender effects when chi-square tests were run for each
state. More specifically, the ratings of Alabama, Kansas, Louisiana, South
Dakota, and Texas were significantly different for the male and the female
respondents. Tables 17-21 show the chi-square results for these states.

Table 17. Pleasantness Ratings for Alabama by Gender

 

 

m Female Male
% Giving Rating (n=32) (n=30)
l 0.0 16.67
2 6.25 20.00
3 9.38 6.67
4 31.25 20.00
5 9.38 23.33
6 25.00 13.33
7 18.75 0.0

 

 

Chi-square = 17.087 with d.f. = 6

Table 18. Pleasantness Ratings for Kansas by Gender

 

 

Em FcEale Male
% Giving Rating (n=3 1) (n=3 1)
1 0.0 0.0
2 3.22 9.68
3 6.45 19.35
4 38.71 16.13
5 25.81 12.90
6 12.90 41.94
7 12.90 0.0

 

Chi-square = 15.980 with d.f. = 5

Table 19. Pleasantness Ratings for Louisiana by Gender

44

 

 

mu; anale Male
% Giving Rating (n=32) (n=32)
1 0.0 9.38
2 3.12 12.50
3 6.25 15.62
4 18.75 15.62
5 31.25 18.75
6 15.62 25.00
7 25.00 3.12

 

Chi-square = 13.313 with d.f. = 6

Table 20. Pleasantness Ratings for South Dakota by Gender

 

 

Gm fimale Male
% Giving Rating (n=32) (n=30)
1 0.0 0.0
2 0.0 3.33
3 9.38 6.67
4 34.38 23.33
5 28.12 26.67
6 12.50 40.00
7 15.62 0.0

 

Chi-square = 11.095 with d.f. = 5"—

45

Table 21. Pleasantness Ratings for Texas by Gender

 

 

imp Female Male
% Giving Rating (n=32) (n=31)
1 0.0 9.68
2 6.25 12.90
3 9.38 12.90
4 18.75 29.03
5 37.50 12.90
6 15.62 22.58
7 12.50 0.0

" "_ —Chi-square = 12.730 with d.f. = 6

In terms of the factor group of age, significant effects were only

found for the states of Minnesota, New Jersey, and Rhode Island, as shown

in Tables 22-24.

Table 22. Pleasantness Ratings for Minnesota by Age Group

 

 

 

 

 

Age Group 2029 3039 4049 5031 60+
% Giving Rating (n=8) (n=13) (n=18) (n=13) (n=10)
1 0.0 0.0 0.0 0.0 0.0
2 0.0 0.0 0.0 7.69 0.0
3 0.0 0.0 11.11 0.0 10.00
4 25.00 53.85 11.11 23.08 40.00
5 12.50 30.77 5.55 15.38 0.0
6 62.50 15.38 55.55 23.08 50.00
7 0.0 0.0 16.67 30.77 0.0

 

c'fhi-square = 31.655 with d.f. = 20

45
Table 23. Pleasantness Ratings for New Jersey by Age Group

 

 

 

 

—A_geGroup 2029 30"39 4049 5'0_59 60+
% Giving Rating (n=8) (n=13) (n=18) (n=13) (n=1 1)
1 0.0 15.38 16.67 0.0 0.0
2 0.0 30.77 5.55 7.69 45.45
3 25.00 0.0 38.89 15.38 0.0
4 37.50 30.77 27.78 15.38 18.18
5 25.00 15.38 11.11 15.38 9.09
6 12.50 7.69 0.0 30.77 27.27
7 0.0 0.0 0.0 15.38 0.0

 

Chi-square = 41.054 with d.f. = 24

Table 24. Pleasantness Ratings for Rhode Island by Age Group

 

 

 

Age Group 2029 3039 4049 5'T59 60+
Rating (n=8) (n=13) (n=18) (n=13) (n=10)
1 0.0 0.0 16.67 0.0 0.0
2 0.0 7.69 5.55 7.69 0.0
3 12.50 7.69 0.0 7.69 30.00
4 12.50 46.15 44.44 23.08 40.00
5 50.00 38.46 16.67 7.69 0.0
6 0.0 0.0 16.67 38.46 30.00
7 25.00 0.0 0.0 15.38 0.0

 

Chi-square = 42.190 with d.f. = 24

Finally, looking at the factor group of visit status, significant
differences in ratings between those who had visited the state and those who
had not occurred only for the states of Alaska and Kansas. These results
are presented in Tables 25 and 26.

47

Table 25. Pleasantness Ratings for Alaska by Visit

 

 

-Group Have Not Have Visited
% Giving Visited State State
Rating (n=44) (n=13)

1 0.0 0.0

2 0.0 0.0

3 0.0 15.38
4 20.45 23.08
5 13.64 23.08
6 20.45 30.77
7 45.45 7.69

 

Chi-square = 11.721 with d.f. = 4

Table 26. Pleasantness Ratings for Kansas by Visit

 

 

-Group Have Not Have Visited
% Giving Visited State State
Rating (n=36) (n=22)

1 0.0 0.0

2 0.0 18.18
3 19.44 4.54
4 25.00 27.27
5 27.78 9.09
6 25.00 36.36
7 2.78 4.54

 

Chi-square = 11.800 with d.f. = 5

Once again, because of the small number of states significantly

affected by the factors of gender, age and visit status, remaining statistical

procedures for pleasantness were computed on the entire data set, without

regard to these factors.

9

, DE, GA, HI, ID, KY, LA, MD, MN, MT, NE, NV, ND, OH,

9 C09
A. RI. UT. WA, WI, WY

P
, CT, DE, FL, GA, HI, IN, IA, KS, KY, LA,

NV, NH, NM, NC, ND, OH, OK, PA, RI, SD, TN, TX,
AR, CT, DE, FL, GA, HI, IL, IA,

9
9

AR, CA, CT, DE, FL, GA, KS, KY, LA, ME, NH), MA, MS, MO, NE, NH, NJ, NY,

OH, PA, RI, SC, TN, TX, VT, VA, WDC, WV, WI
AK, CA, CO, GA, HI, ID, IA, KY, MI, MN, MT, NE, NV, NM, ND, OH, OR, SD,

UT, WA, WY
AL, CT, DE, FL, GA, IL, IN, KS, KY, LA, ME, MD, MA, MS, MO, NE, NH, NJ,

NY, NC, OH, OK, PA, RI, SC, SD, TN, TX, VT, VA, WDC, WV, WI

States which were not significantly different

0

ID, WA: AZ, CA, HI, MT, NV, OR, UT, WY

State
AK, CO,
CA:

GA:

IL‘

 

 

 

 

 

 

 

 

 

 

 

 

Table 27. Chi-Square Results for Pleasantness Ratings

KY, LA, ME, MD, MI, MN, MO, MT, NE,
D

S.
,S

K
RI
, GA, HI, IL, IN, KS, KY, MD, MI, MN, MT, NE, NV, NH, NM,

, TN, TX, UT, VT, VA, WV, WI, WY

, PA, RI, SD, TN, UT, VA, WI, WY
FL, GA, IL, IN, IA, KY, LA, ME,

r

, NC, ND, OH, PA,

9NI-I9
FL
O

 

 

, MS, MO,

, ND, OH, OK, PA, RI, SC, SD, TN, T , UT, VT, VA, WDC,

KS:

 

A .

KY:

, WA, WI, WY
, FL, GA, HI, ID, IL, IN, IA, KY, LA, MD, MN, MT, NV,

, HI, ID, IL, IN, IA, KY, LA, MD, MI, NW, NE,
, WA, WI, WY

, GA, HI, IL, IN, KS, K
, OH, OK, PA, RI, SC.

ND
, DE, FL, GA, IL, IN, IA, KS, KY, LA, ME, MD, MN, MO, MT, NE, NH,

AL, AR, CT, DE, FL, GA, KS, KY, LA, ME, MD, MN, MS, MO, NH, NJ, NY, NC,

PA, RI, SC, SD, TN, TX, VT, VA, WDC, WV, WI

AZ.

AL, AR, CT, DE, FL, GA, KS, KY, LA, ME, MD, MA, MO, NH, NJ, NY, NC, OK,

NM, NC, ND, OH, PA, RI, SD, TN, TX, UT, VT, VA, WV, WI, WY
PA, RI, SC, SD, TN, TX, VT, VA, WDC, WV, WI

MA:
MI :

MN:
MO:
IVfT:

MS:

 

 

 

 

 

 

 

 

 

 

49

 

 

Table 27 (cont'd).

, GA, IL, IN. IA. KS.

ND
, CT, DE, FL, LA, IVE, MD, MA, MS, MO, NH, NYC, NY, PA, RI, TN,

me.

C

S, MO, NE,

A, KS

I
A, WV, WI

. OH. OK, PA, RI, S
AK, AZ, CA, CO, ID, MT, NV, WA, WY

UT, VT,

, FL, GA, IL, IA. KS, KY, LA, IVE, MD, MS, MO, NE, NH, NC, OH,

DE, FL, GA, HI, IL, IN, IA, KS, KY, MD, MI, MN, MT, NE, NV, NH, ND,

WDC,WV

9
9
9

A

V
.CT.DE.FL.GA,IL,I

AR
NH, ND, OH, OK, PA, RI,

9
9

OH, SD

AL, AR, MA, MS, NJ, NYC, PA, RI, TX, WDC

AZ, CA, CT, DE, FL, GA, HI, IL, IN,
NV, NH, NM, NC, OH, PA, RI, SC, S ,

AL
TX
NJ, NY

NYC:
NC:
OH:
OR:
PA:

NJ:

 

 

 

 

 

 

 

 

 

 

 

, MI, MS,

, NJ, NY, NC, ND, OH, RI, SC, SD, TN, TX, VT, VA, WDC,

, DE, FL, GA, IL, IN, IA, KS, KY, LA, IVE, MD,

NH

a

, MI, MS, MO, NE, NH,
, IA, KS, KY, LA, NE, MD, MI,

, MA, MS, MO, NE, NH, NJ, NY,
, UT, WA, WI

, IL, IN, IA, KS, KY, LA, IVE, MD, MA, MI,
M, NC, ND, OH, OK, PA, RI, SC, SD, TN, TX,

, OH, OK, PA,

O, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, MD, MI, W,
, ND, 0 , OK, PA, SD, TN, WA, WI, WY

NC, ND, OH, OK, PA, RI, SC, SD,

, DE, FL, GA, IL, IN, KS, KY, LA,
C

9

A.
OH, OK, PA, RI, SC, SD, TN, TX, VA, WDC, WV, WI

CT
NY

C

9
9
9

AZ
NE
, AR, CT, DE, FL, IL, IN, KS, KY, LA, ME, MD, MA, MI, MS, MO, NE, NH,

AR.
NJ,
ND

9

AR, CT , DE, FL, GA, KS, KY, LA, IVE, MD, MA, MS, MO, NH, NC, ND, OH,

PA, RI, SD, TN, TX, VT, VA, WDC, WV, WI, WY
AL, AR, CT, DE, FL, GA, IL, IN, IA, KS,
MO, NE, NH, NJ. NC.

AK
MT

NC

SC:
SD:
VA:

 

 

 

 

 

 

 

 

 

 

 

 

50
Chi-square test results on the overall pleasantness ratings of each

state compared to every other state are reported in Table 27. As with the
correctness data, states which are not significantly different from one
another and from exactly the same other states are reported together.

Like with the correctness data, the chi-square results for pleasantness
give a general idea of how to group states together. However, many states
are not significantly different from states in almost every other
geographical region. Thus the boundaries of these groups are unclear if
we rely on the chi-square data alone.

A combination of MDS analysis and K-Means Cluster analysis
provides a better understanding of how the states group together in terms
of pleasantness ratings, as shown in Figure 12.

Regarding the MDS analysis, it is apparent here that, like the
correctness data, the vertical dimension is more significant than it was in
the degree of difference data. Once again, however, it is difficult to say
precisely what this dimension is measuring. While the horizontal
dimension seems to represent general geography in this case (with a few
noticeable exceptions such as Alaska), it may be that once again the vertical
dimension designates ”type" of accent.

In terms of the K-Means cluster analysis, the solid circles in Figure
12 represent the output when 9 cluster groups are selected. If only 8
groups are used, Oregon joins the larger West/Midwest cluster. When 7
groups are chosen, Washington also is included in this larger cluster. If
only 6 cluster groups are selected, Florida falls in with the southern states.

The pleasantness clusters are similar to the correctness clusters in the
fact that the western and midwestem states are lumped together in a single

cluster group, with the exception of Michigan and Wisconsin being

51
grouped with the northeastern states. (Recall that Michigan, but not

Wisconsin, was grouped with the northeastern states for correctness.)
Also, California and Hawaii form their own individual groups, rather than
being included in the large West/Midwest group.

There is a clear northeastern cluster, although New York City, New
York, New Jersey, Massachusetts and Washington DC are not included in
this group but rather in a separate cluster. Virginia does get incorporated
into this northeastern group, but West Virginia is included in a cluster of

southern states.

52

 

qu-

1.54-

 

 

 

 

 

 

 

 

Figure 12. MDS Analysis of Pleasantness Ratings with K-Means Clusters

4 DISCUSSION

By comparing the evaluations of states with one another for each of
the three rating tasks (degree of difference, correctness, and pleasantness)
in combination with the regions outlined and the labels given on the hand-
drawn maps, it is possible to make some generalization about this

respondent group's perception of American regional dialects.

4.1 Perceptions of Western States. To begin with, the hand-
drawn maps and the degree of difference ratings show that Oregonians
consider the states of Oregon, Washington, Idaho, Colorado, Utah,
California, Nevada, Montana, Arizona, Alaska and Wyoming to constitute a
distinct dialect region. This is illustrated dramatically by the distribution
of the ratings for degree of difference (see Appendix E). Of the 66
respondents, between 47 and 63 rated each of the above states as a "1", i.e.
"people there sound like me."

The next highest number of "1" ratings is 32, given to New Mexico,
although almost an equal number of respondents (25) rated it a "2." It is
clear that New Mexico is a kind of border state between the perceptual
regions of West and Midwest. Although slightly more people rated it a "1"
than a "2" in terms of degree of difference, the K-Means cluster analysis
for degree of difference groups it with the midwestem states. The MDS
analysis places it midway between Wyoming (a western state) and South
Dakota (a midwestem state). Finally, on the hand-drawn maps, the state is
split, with the majority of respondents who drew a "West" including the

northwest corner of the state in the "West" (see Figure 6).

53

54
While these western states are clearly distinct for Oregonians in

terms of degree of difference, this distinction is less clear in the correctness
and pleasantness ratings. The majority of western states are lumped
together with the midwestem states in these categories according to chi-
square tests (Tables 16 and 27) and K-Means cluster analysis (Figures 11
and 12). In terms of correctness, Oregon and Washington are
distinguished from the rest of the western states as a separate cluster, and
each makes up its own cluster in the pleasantness K-Means cluster analysis.
These results reveal a desire on the part of the respondents to distinguish
the local area (in this case Oregon and Washington) from surrounding
states, even when there is no noticeable difference in terms of accent
(reflected in the degree of difference ratings). This result conforms with
the general tendency for self-identification found in respondents from
other geographical areas (Preston, 1986).

This desire to differentiate Oregon and Washington as a distinct area
is also illustrated by the 16 respondents who drew a separate "Pacific
Northwest" area on their hand-drawn maps (see Figure 7). In the follow-
up interviews (see Appendix D), several respondents also commented on
the fact that the speech in Oregon and Washington was very similar but
could be distinguished from that of surrounding states. For example, one
respondent made the comment, "I can be in Washington and not feel like
I'm not in Oregon but I can be in Idaho and I can tell a difference"
(respondent #122).

The desire to distinguish California somewhat from other western
states, and particularly from Oregon and Washington, is reflected most
strongly in the pleasantness ratings. While 18 respondents rated California

as a "7" in terms of pleasantness, almost as many (16) rated it a "4". It is

55
unclear how to account for this discrepancy, since there were no significant

effects on the ratings of California in terms of gender, age group, or visit
status. What is clear is that many of the respondents wanted to differentiate
California from other western states, but did not choose to do so in terms
of degree of difference or correctness. Instead they opted to use the
available affective dimension of "pleasantness.”

Past research has revealed that status and in-group solidarity are the
two primary evaluative dimensions along which language attitudes can be
measured (Ryan, Giles, and Sebastian, 1982). In this study, the category of
"correctness” was used to measure status distinctions, while "pleasantness"
was chosen to reflect degree of solidarity. In view of this, the ratings of
California reveal that while these Oregonian respondents recognized the
more or less equal social status/power of California residents, they feel a
lack of solidarity with them. This lack of solidarity is underscored by the
fact that California is singled out from other western and midwestem states
in terms of pleasantness K-Means clusters even when Oregon and
Washington are collapsed into this large cluster.

The following comments illustrate the somewhat negative view that
some respondents have towards California (interviewer comments and

questions appear in brackets):

...I lived in California for a short time when I was a teenager and
wanted to be away from home, to start my own life, and I hated it...I
was in L.A., well I lived in Wittier, which was outside L.A....I only
lived there 11 months and I was ready to come home. I remember
the first time it rained, I looked up in the sky and I cried and said,
"Thank you God"... (respondent #119)

56

[why the very different rating for CA?]

Well, I was thinking about it and I think there's just, I don't know
some speech patterns, it just - sounds different [in terms of
pleasantness-- not necessarily in terms of correctness?] Yeah, right.
(respondent #121)

people in California talk their own language which I don't
understand [what do you mean by ”talk their own language?"] their
whole terminology is a different - task it seems like their words are
bigger and a lot of it seems to be, at least in the areas I've been in,
technical terms and a lot it I guess I think of Silicon Valley in
regards to computers and all the technical variety (respondent #122)

The hand-drawn maps also reveal that quite a few of the respondents
differentiate northern and southern California (see Figures 7 and 8),
grouping northern California with Oregon and Washington. Support for
this dialect division comes from traditional word geography data (see
Figure 1) as well as from the greater influence of Spanish in southern
California.

While California is differentiated slightly from other western states
in the respondents' perceptions, Alaska is situated squarely within this
cluster of states on all three scales. In fact, its ratings in terms of degree of
difference and correctness are almost exactly the same as those of Arizona,
and it is rated fourth highest in terms of number of "7" ratings for
pleasantness. It is likely that this identification of Alaska with the west is a
result of the historic connection between Alaska and the port cities of
Seattle and San Francisco. The hand-drawn maps support this conclusion
as well. Although 9 respondents (13.8%) indicated Alaska as a separate
region, a greater number (12) included it in their "West" or "Pacific
Northwest.” The remaining respondents simply did not include it in any

specific region. Even when Alaska was circled as a separate speech area,

57
several respondents labeled it as "Frontier”, another indication that Alaska

is perceived more in terms of the settlers that migrated there, embarking
from the Northwest, rather than in terms of the native populations which
inhabited the territory long before the arrival of the European-Americans.

The inclusion of Alaska within the ”West” is also interesting in light
of the fact that Hawaii was not considered part of this region on any scale.
In fact, in the K-Means cluster analysis for all three rating categories,
Hawaii emerges as a distinct cluster. As mentioned in the results section,
on the MDS analysis of the degree of difference data (see Figure 10), it
appears that the vertical dimension is used primarily to differentiate Hawaii
(which received a value of +1.23) from all other states (which fall between
-0.55 and +0.44 on the same dimension). Furthermore, 15 respondents
(23.1%) indicated on their hand-drawn maps that Hawaii comprised a
distinct dialect region, while only 7 included it in another region (either
West or California). Several respondents labeled the speech in Hawaii as
”native-like", which seems to indicate that the influence of the indigenous
population of the islands on the speech in that state is significant in the
respondents' perceptions. Again, this makes Alaska's ratings even more
interesting because there does not seem to be an equally strong association
of indigenous languages effects on the speech of Alaska. Even more
substantial is the perception of Asian influence on the Hawaiian speech,
with one respondent even labeling it ”Japanese Golfland."

Overall, these results confirm the first half of Hypothesis 1, namely
that Oregonians will differentiate varieties of speech in western states to a
greater degree than non-west coast residents, although they will still outline
a single major Western dialect region. The most common boundary drawn

in relation to western states on the hand-drawn maps divided the West from

58
the rest of the country at the Montana/Wyoming/Colorado/Arizona state

lines. 24 respondents (37.9%), however, more finely differentiated the
western region in terms of speech areas. 16 (24.6%) of these respondents
included Oregon in a "Pacific Northwest" while 8 (12.3%) drew a "West
Coast.” When we compare these results to the respondents from other
regions in Preston (1986), we find that the respondents from New York
State were the only group to draw 3 "Northwest", although the eastern
boundary of this region extended all the way to Minnesota in their
perception.

Hypothesis 2, which states that Oregonians will not generally
distinguish themselves from Washington, but may do so slightly from
California, due especially to caricatures of southern California (e. g.
Hollywood, ”valley" talk, and a large proportion of native Spanish-
speaking residents), is also generally confirmed by these data. In
particular, the separation of California from other western states is clearly
borne out as discussed previously, although not as decisively as the results
of respondents from other regions (Preston 1986).

Regarding the relationship of Oregon and Washington, in the MDS
analysis for degree of difference, Oregon and Washington received
identical scores (-1.46, +0.07). Although the MDS scores are not exactly
the same for Oregon and Washington in terms of correctness, the K-Means
cluster analysis produced an Oregon/Washington cluster. When it comes to
pleasantness, however, Oregon and Washington emerged as distinct
clusters. This suggests that Oregonians want to maintain at least a small
claim of uniqueness, but the interpretation of this result is not entirely
straightforward. Preston (1993b) claims that "the preference for local

norms along affective lines is stronger in areas where there is linguistic

59
insecurity" (p. 35). Since Oregonians view themselves as entirely unique

only on the affective scale, this might indicate a slight degree of linguistic
insecurity. This claim is somewhat tenuous, however, in light of the fact
that Oregonians also rate themselves highest in terms of correctness, a fact
which points to a great deal of linguistic security. It may be that the
traditional categories of linguistic security/insecurity do not apply in the
same way in western states, where a multiplicity and therefore awareness
of distinctive dialects is not as prevalent as in eastern and southern states.
Thus Hypothesis 3, which states that Oregonians will exhibit a strong
degree of linguistic security, reflected in high ratings for Oregon in terms
of both "correctness" and "pleasantness”, with equally high ratings for
other western states in the "pleasantness" category, is tentatively confirmed

but remains somewhat in question.

4.2 Perceptions of Midwestern States. According to the
degree of difference MDS and K-Means cluster analysis, the Oregonian
conception of the Midwest consists of the states of New Mexico, North and
South Dakota, Minnesota, Nebraska, Kansas, Wisconsin, Iowa, Illinois and
Indiana, Michigan and Ohio. The hand-drawn maps reveal that Iowa is the
"heart" of this region (see Figure 5). The cluster of midwestem states is
the most closely related to the western states in several ways.

First, in the K-Means cluster analysis of the degree of difference
data, the midwestem and the western clusters of states collapse together if
only 7 (rather than 8) groups are chosen. The midwestem and western
states are also lumped together on the correctness and pleasantness tasks
with the western states (with the exception of Oregon, Washington and for

pleasantness California) in the K-Means cluster analysis, although they are

60
clearly two subgroups in terms of their MDS values (as seen in the MDS

plots, Figures 11 and 12).

Michigan appears to be a kind of border state between the Midwest
and the Northeast for these respondents. While it falls clearly within the
center of the Midwest cluster on the degree of difference ratings, it is part
of the K-Means cluster of northeastern states in terms of both correctness
and pleasantness. On the hand-drawn maps, 20 respondents included it
within their boundaries of a "Midwest," while only 6 incorporated it into a
”Northeast.”

It is possible that the ”split personality" of Michigan is a result of the
sensitivity of respondents to the North Cities Vowel Shiftl, a phonological
change in progress which began in major cities on the east coast and is
slowly spreading from urban center to urban center across the north (see
Figure 2). This shift is well underway in Michigan cities. If this were the
case, however, one would suspect that it would be in the degree of
difference ratings, which are based on the sounds of the language in each
state, that Michigan would be rated most like northeastern states. Perhaps
the traditional association of Michigan with other midwestem states,
particularly Great Lakes states such as Illinois and Wisconsin, is too great
to be overwhelmed by a phonological change in progress, while at the same
time there is some recognition that "something” (i.e. the NCVS) makes it
sound more like states in the northeast.

The most recent work within Michigan suggests that university
students rate their own speech as much more similar to east coast varieties
than respondents in studies done ten years ago did (Preston, 1996). Thus it

seems that the influence of the NCVS on Michigan speech is a recognizable

 

1 For a detailed description of the North Cities Vowel Shift, see Iabov (1991).

61
reality not only for professional linguists, but also within the general

population.

If this is in fact the correct interpretation of the Michigan ratings, it
might also explain why Wisconsin clusters with the midwest in terms of
degree of difference and correctness, but with Michigan and the
northeastern states on the pleasantness task. Since the NCVS began on the
east coast and is slowly moving westward, overall the shift is further along
in Michigan than Wisconsin. Because of this, Wisconsin may sound a little
bit "northeastern", but not as much so as Michigan.

Ohio appears also to be a kind of border state between the Midwest
and the Northeast, but in a different way than Michigan. Ohio clusters
with the midwestem states on all three scales, but the hand-drawn maps
reveal some disagreement among respondents as to which region it belongs
to. 22 respondents include the western half of the state in the midwest, but
only 11 incorporate the entire state in this region (see Figure 5). On the
"Northeast” map, 19 respondents include only the eastern half of the state,
while 11 respondents incorporate the whole state. Ohio thus appears to be
a border between the Midwest and the Northeast in the same way that New
Mexico is split between the West and the Midwest. It is important to note
here that the first large group of settlers in the Willamette Valley in
Oregon came from the Ohio Valley states and Tennessee. There were also
large numbers of settlers from Missouri, Illinois and Iowa (Carver 1987).
Thus while the geographical location of Ohio may have caused some
respondents to incorporate it into the northeast, particularly on the hand-
drawn maps, the historical connection between Ohio and other midwestem
states as the major migration origination points for Oregon settlement was

more influential in the rating tasks.

62
Hypothesis 2 predicted that Oregonians would not as finely

distinguish among states in northern regions as do residents in those areas.
Comparing the results from this study with results from previous studies in
northern areas shows that this hypothesis is also confirmed. In particular,
this is seen in the differentiation that respondents from Michigan and
Indiana (Preston 1993b) tend to make between Great Lakes states and other
Midwest states.

More specifically, respondents from Michigan separate the "North"
states of Minnesota, Wisconsin, Illinois, Indiana, Michigan and western
Ohio from the ”Midwest" states of Iowa, Missouri, Nebraska and Kansas in
hand-drawn maps. The Dakotas are grouped with Montana and Wyoming
in a ”Plains and Mountains” region. As for the Indiana respondents, their
”North” contained only the states of Wisconsin and Michigan, while the
”Midwest” included Kansas, Iowa, Illinois and northern Indiana.

It is more difficult to compare the ratings in terms of degree of
difference, correctness and pleasantness of respondents from Michigan and
Indiana with the ratings in this study since different statistical measures
were used in Preston 1993b. However, it is apparent that Michigan raters
see themselves as a unique region in terms of correctness with the states of
Minnesota, Wisconsin, Illinois, Ohio and Pennsylvania forming another
region. Indiana raters show the same tendency to favor local norms in
isolating Indiana as a unique region in terms of pleasantness. Again, no
differentiation of Great Lakes states from other midwestem states in made
by the Oregonian respondents in terms of correctness and pleasantness, and
only a very small number of respondents (6) drew a Great Lakes region on

their maps.

 

63
4.3 Perceptions of Northeastern States. The next group of

states which emerges from the MDS and K-Means cluster analyses of the
three tasks is a northeastern cluster. The primary states in this cluster are
Maine, Vermont, New Hampshire, Rhode Island, Connecticut,
Pennsylvania, Maryland, Delaware, with some variability among the three
tasks for Washington DC, Massachusetts, Virginia, West Virginia and
Michigan and Wisconsin (as discussed previously).

For the degree of difference task, Washington DC. is included in the
northeastern cluster. It is isolated as its own group in terms of correctness,
falling somewhere in the middle of the ratings for midwestem and
northeastern states (see Appendix E). For the pleasantness task, however,
it is grouped with Massachusetts, New Jersey, New York, and New York
City, having received the third highest number of ”1" ratings (after New
York City and New York, see Appendix E).

In terms of correctness, the fact that Washington DC. emerges as a
distinct cluster is likely a result of the specialized styles and jargon
associated with the discourse of government functions. That it is grouped
in terms of pleasantness with the cluster which includes New York City,
clearly the cluster with the least favorable ratings, is perhaps the most
interesting result. This may well be an indication of growing
dissatisfaction with the current political scene, particularly in light of
controversy surrounding Oregon senator Bob Packwood, which later led to
his resignation from Congress.

Massachusetts follows the same pattern as Washington DC, being
clustered with the northeastern states in terms of degree of difference,
emerging as a distinct cluster for correctness (also with ratings somewhere

in the middle of the northeastern and midwestem states), and grouping

64
with New York City, New York, New Jersey and Washington DC. on the

pleasantness task (receiving the fourth highest number of "1" ratings).
Since this grouping cannot be a result of the perception of government,
however, an alternative explanation must be found. The most likely reason
for the differentiation of Massachusetts from other northeastern states is
the stereotypical (and actual) Boston accents. On the hand-drawn maps, in
fact, several respondents used the label "Bostonian” to refer to a general
New England area.

The historic connection between Boston and the Pacific Northwest,

 

particularly in terms of the fur trade in the first half of the 19th century,
may also be important here. Speaking of the New Englanders who
controlled much of the Oregon coast during this time, Carver (1987, p.
242) says, "So prevalent was their influence and presence that the Indians
called all white men 'Bostons.'" If this historic connection was important
in the ratings, then this would suggest a negative perception of the role of
the early Bostonian merchants, since Massachusetts was rated so low in
terms of pleasantness. This interpretation is highly speculative and should
be investigated further.

Virginia and West Virginia appear to be transitional states between
the Northeast and the South for the respondents in this study. Although
grouped with the northeastern states in the K-Means Cluster for degree of
difference, they fall in with the southern states in terms of correctness.
They are split in the pleasantness ratings, with Virginia grouped with the
Northeast and West Virginia clustering with the South. The hand-drawn
maps also confirm these states as border states, since almost an equal
number of respondents included them in the Northeast region as in the

South (see Figures 3 and 4).

65

4.4 Perceptions of New York City, New York and New
Jersey. Figure 4 shows that New York is the center of the hand-drawn
maps of a Northeast, and that in general New York City and New Jersey
get incorporated into this perceptual region. It is interesting then that New
York City, New York and New Jersey are never included in the K-Means
clusters of the northeastern states. In fact, New York City and New York
both get included in the cluster of southern states on the degree of
difference scale. They form their own cluster in terms of correctness, and
as stated already combine with New Jersey, Massachusetts and Washington
DC. to form a cluster in terms of pleasantness. New Jersey emerges as a
distinct cluster for degree of difference and correctness.

The results for New York City are unsurprising and confirm the
first part of Hypothesis 4, which states that New York City will be rated
the lowest overall for both ”correctness" and "pleasantness”. In fact, it did
receive the lowest ratings by far in terms of pleasantness, with 32
respondents giving it either a "1" or "2" rating (see Appendix E). In terms
of correctness, New York City also received the greatest number of "1"
ratings (12, compared to Alabama's 7). These ratings are clearly based on
the stereotype of New York City inhabitants as fast-talking, cold, and rude.

Consider the following comments made in the follow-up interviews:

When we used to go to the flea markets in, well, Greenfield and
those we used to think New Yorkers were mad at each other--that's
just how they talk, they always sound like they're arguing and
that...yeah, and they're just, that's the way they speak, and we used to
kind of, they'd yell and scream and you'd look back thinking they
were fighting and they were just visiting (husband of respondent
#104).

66

People seem to talk very fast up there from what I've experienced...
(respondent #122)

While the results of the New York City ratings were expected, it is
particularly interesting that New York state gets "dragged down" with New
York City. Chi-square tests reveal no significant difference between the
ratings given to New York and New York City on any of the tasks.
Furthermore, these two regions are always in the same K-Means cluster.
Thus, there appears to be little difference in Oregonian perceptions
between the City and the rest of the state.

This result is strikingly different from the results of maps drawn by
residents of both New York City and western New York reported in
Preston 1986. Both sets of respondents indicate that New York City and
New York state are distinct dialect regions. The respondents in western
New York even further differentiated western New York as a separate
region. This contrast between the Oregonian respondents and the New
York respondents further confirms the portion of Hypothesis 1 which
claims that residents in northern states are more likely to distinguish
dialects regions in those areas than Oregonians are.

The ratings of New Jersey are also clearly influenced by the New
York City stigma, although not in precisely the same way as New York
state. For both degree of difference and correctness, New Jersey forms its
own cluster, as mentioned previously. This may be due in part to New
Jersey having its own stereotypical accent (i.e. ”New Joisey"). It may also
be the case that New Jersey is seen as a kind of ”transitional” or "buffer”
state between New York City and other northeastern states such as

Pennsylvania, Delaware and Maryland. In general, New Jersey does seem

67
to occupy a location in-between New York City and the northeastern states

on the MDS plots.

4.5 Perceptions of Southern States. On the hand-drawn
maps, the South is clearly the most salient dialect region, with 92.3% of
respondents indicating at least some portion of it as a distinct region. The
states unequivocally part of this region are Alabama, Louisiana,
Mississippi, Georgia, Arkansas, North and South Carolina, Kentucky,
Tennessee, Texas and Oklahoma. Missouri is included in K-Means clusters
for correctness and pleasantness, but forms its own cluster in terms of
degree of difference. West Virginia joins the group in correctness and
pleasantness, and Virginia is part of the cluster in terms of correctness, as
discussed previously. Florida forms a distinct cluster on all three scales,
although it joins the southern cluster on the correctness and pleasantness
scales if fewer clusters are selected.

In terms of actual ratings (see Appendix E), the southern states are
rated the most dissimilar to Oregon. Between 43 and 58 respondents rated
all the "core" southern states except Oklahoma either a "3" or a ”4" on this
scale. Of non-southem states, only New York City received such a large
number of low ratings. For correctness, the southern states also fall at the
bottom of the list in terms of ratings. Finally, the ratings for pleasantness
rise dramatically for all of the southern states, with Georgia and Louisiana
actually receiving the most "7" ratings after the western states.

Missouri's marginal status as a southern state (reflected in the fact
that it forms its own K-means cluster) is interesting for several reasons.
First, the largest number of settlers in Oregon in the latter half of the 19th

century came from Missouri (Carver 1987). Thus the recognition of

68
Missouri as not as different from Oregon as the other southern states (in

terms of degree of difference ratings) may be a result of this historic
connection. That these Oregonian respondents group Missouri with
southern states at all, however, is intriguing in and of itself, since the
respondents from the five areas that Preston (1986) examined tended to
place Missouri in the Midwest, rather than the South. It seems that in
Missouri we once again find a kind of transitional state, in this case
between the South and the Midwest. This is supported by the hand-drawn
maps in which more respondents include Missouri in the Midwest (Figure
5) than in the South (Figure 3), even though the respondents clearly
associate it with the South in the ratings tasks.

Texas is also an interesting state to examine in terms of the
discrepancy between hand-drawn maps and the ratings tasks. 55.4% of
respondents singled out Texas (sometimes including Oklahoma) as a distinct
dialect area on their hand-drawn maps. In none of the three ratings tasks,
however, did Texas appear as a distinct K-Means cluster. In fact it is only
slightly peripheral in terms of the MDS plots in the degree of difference
category (although it received almost exactly the same ratings as Tennessee
on this task), and it was not at all peripheral on the other two tasks. It
seems, then, that while there is some recognition of a distinct dialect in
Texas, this distinctness is not great enough to overcome the general
category "Southern” in the minds of the Oregonian respondents.

With Florida there appears to be precisely the opposite discrepancy
between the hand-drawn maps and the ratings tasks than occurred with
Texas. In this case, Florida was generally incorporated into the maps of

the South but emerged, at least potentially depending on the number of

69
cluster groups used, as a distinct K-Means cluster on all three of the ratings

tasks. There are two possible explanation for the singling out of Florida.

First of all, there may be the recognition of influences from
northern dialects due to (particularly) retiree migration from the North to
the milder climate of Florida. This appears to be the explanation, for
example, of Michigan raters' perceptions of Florida, since Florida falls
together with Michigan on factor analyses of both correctness and
pleasantness data (Preston 1993b). The second explanation has to do with
influence of (primarily Cuban) Spanish on the language in Florida. Given
the fact that Florida would not be a likely retirement spot for Oregonians,
in combination with several respondents who used labels such as ”Cuban"
on their hand-drawn maps, this second explanation is more plausible.

In terms of overall ratings for southern states, the portion of
Hypothesis 4 which states that Southern states will be rated low for
”correctness" but slightly higher for "pleasantness” is confirmed by this
study. As a group, the southern states received the lowest ratings in terms
of correctness, especially the four states of the ”deep south”, i.e. Georgia,
Mississippi, Louisiana, and Alabama. At the same time, the pleasantness
ratings are substantially higher; many southern states are in fact rated
higher than many midwestem and northeastern states. This conforms to
the stereotypes of ”Southem hospitality" and a slower pace of life.

Speaking about people from Kentucky, for example, one respondent says:

...they act like they've got all the time in the world, you know—-
"nothing wrong with me, I'm just taking my time," I can't do it, but
it sounds great... (respondent #119)

70
In general, then, Oregonians view the South as one large dialect

region. Although they discriminate somewhat a ”deep South" from an
”outer South" and ”Texas” on their hand-drawn maps, this distinction is not
great enough to create separate K-Means clusters in any of the ratings
categories. An examination of the MDS plots, however, does reveal that
the Southern cluster is the "loosest" of all the clusters, i.e. its points are the

most spread out.

4.6 Effects of Gender, Age and Visit Status. Of the three
demographic factors studied, gender had the least effect on the ratings. In
fact, it was only in pleasantness category that gender played a role in the
ratings at all. The states that were influenced by the factor of gender were
Alabama, Louisiana, Texas, Kansas and South Dakota (see Tables 17-21).
The female raters gave higher ratings than the male raters to the southern
states, but the opposite was true for the midwestem states. The higher
ratings of the southern states by the female respondents must again be a
result of the stereotype of "Southern hospitality", a stereotype which is
more likely to appeal to women than men, due to romanticizations made
popular by media portrayals of southern life such as "Gone With the
Wind. " On the other end of the spectrum, associations of hard-working
farmers, ranchers and cowboys are more likely to appeal to male raters,
resulting in higher scores given to these two midwest states.

It is clear that overall, the factor of gender played a very minor role
in the ratings of the respondents. It had no significant effect on the degree
of difference and correctness ratings, and only figured in the ratings of a
few prototypical southern and midwestem states in terms of pleasantness.

Thus, Hypothesis 6, which states that gender may influence ratings slightly,

71
but not significantly overall is confirmed by these results. Another

prediction was that in terms of "pleasantness” and "correctness”, women
will have a tendency to rate more extremely overall. Since gender was not
a significant factor in the majority of ratings, this part of Hypothesis 6 is
not borne out in this data. In the few cases were gender played a role in
ratings, female raters were less harsh to southern states but were harder on
midwestem states.

The factor of age group seemed to play the biggest role in affecting
respondents' ratings. In the degree of difference category, it influenced the
ratings of Illinois and Indiana. For correctness, North and South Carolina,
Tennessee, Texas, Virginia, and South Dakota were affected. In terms of
pleasantness, significant differences among age groups were found for the
states of Minnesota, New Jersey and Rhode Island. It is difficult to find
any obvious generalizations to make about how age group factored into the
ratings overall. On the degree of difference scale, the youngest group of
raters were least likely to associate the speech of Illinois and Indiana with
their own speech, while the older respondents (in the 50-59 and 60+)
categories were most likely to do so. This may be due to the older group
being more cognizant of the roots of Oregon settlement. Since the
settlement of Oregon and the West in general is rather recent, especially
compared with the South and the East, it is likely that older respondents
have a greater connection to other areas of the country. One older
respondent, for example, describes growing up in California and Oregon
by saying:

...all my friends in childhood had come from other states and there

was almost no one in Long Beach when I lived there that -- had

really been there for very long and the same was true in Oregon --
people came in and it wasn't settled, you know the oldest churches

72

and the oldest schools and things down in Grant's Pass are only 100
years old, so that's that's within - my parents didn't quite make 100
but they, they're gone now, but there's just just, changed radically
and it's in a constant flux... (respondent #57)

Thus it seems that the few effects that do show up in terms of degree of
difference are because of the difference in life experience between the
Oregon in the childhood of those who are now 50+ and the Oregon in
which younger respondents grew up.

What is most striking about the effects of age group on the
correctness ratings is that five of the six states which showed significant
effects were states of the ”outer South”. In general, the youngest and the
oldest age groups seem to be the most uniform in their ratings, with the
scores of the three middle age groups varying widely . Rather than being
the result of the perceptions of these states themselves, I believe this
variation in terms of correctness is a reflection of the greater reluctance of
those in the three middle age groups to participate in the correctness task.
I will discuss this more in-depth in section 4.7, but four respondents in
each of the three middle age groups gave all the states the same ratings for
correctness. The variation occurs because there was some difference as to
which rating they picked. Most chose ”4", but others used "5", "6", and
"7." Only one respondent in the youngest group and two respondents in
the oldest group did this. Since there was not a very large number of
respondents within each age group, these ”strategies of protest" against the
correctness task resulted in significant age group effects in several states.

The effects of age group on pleasantness ratings for Minnesota, New
Jersey and Rhode Island are even less straightforward. There again tends

to be a wider variation in ratings among the three middle age groups than

73
among the youngest and oldest. Since, in general, respondents did not

object to the pleasantness task as they did to the correctness task, the
explanation doesn't seem to lie in ”strategies of protest." Perhaps all that
can be said is that the small number of respondents in each age group may
have played a role in all the instances where age group appeared to be
significant. Furthermore, age group, unlike gender or visit status, is not a
discrete category. The divisions between age groups were thus somewhat
arbitrary and perhaps did not represent age effects realistically. Still, age
group did not play a large role in ratings in the majority of cases, and thus
it seems that the portion of Hypothesis 6 which states that age may also
have a small effect on ratings, but overall this effect will not be significant,
is confirmed.

Finally, whether or not respondents had visited a state also played a
very minor role in ratings. In terms of degree of difference, significant
differences surfaced for only Arizona, Kentucky and Massachusetts. In the
case of these first two states, respondents who had visited the states rated
them to be more similar to Oregon than respondents who had not visited.
For Massachusetts, the opposite was true. The Massachusetts results may
again be the result of the historic connection between Boston and the
Pacific Northwest. For those who had not actually visited Boston and
heard the distinct accent, the assumption may have been that since much of
the early influence in the Oregon territory was from Boston, people in
Boston must sound like people in Oregon.

For the correctness task, visit status was only significant for the
states of Maine and Nebraska. In the case of Maine, those who had not
visited the state generally rated it higher than those who had, although the

highest rating (7) was given by more respondents who had visited than

74
those who had not. As for Nebraska, in general it was rated more

favorably by those who had visited than those who had not, although the
lowest rating for those who had visited was a ”2", whereas the lowest
rating by those who hadn't visited was a ”3.” Because of these
discrepancies within the two demographic groups for each state, it is
difficult to draw any general conclusions as to how visit status influenced
people's ratings in this task.

Finally, in the pleasantness category, visit status was significant for
the states of Alaska and Kansas. For Kansas, the scores again seem to be
somewhat random, i.e. it is difficult to say precisely what effect having
visited the state had on ratings. With Alaska, the results are more
straightforward. Those respondents who had not visited the state rated it as
more pleasant than those who had. This is again likely due to the historic
connection between Alaska and the Northwest, with the resultant perception
that inhabitants in Alaska must talk like Oregonians (thus producing high
pleasantness ratings). It may also be due to the somewhat romantic
depiction of life in small-town Alaska in the popular television show
"Northern Exposure."

Although visit status did produce significant effects in a few
instances, it is clear that overall there was little difference in ratings
produced by actual experience in a state. Those few instances where visit
status did have some effect were not in highly caricatured areas such as the
South and New York City. Thus, Hypothesis 5, which states that exposure
to language varieties in different states may effect ratings slightly, but in
general, stereotypes will be stronger than contradictory experiential
evidence, particularly in highly caricatured areas such as New York City

and the deep south, is confirmed by these data.

75

4.7 Comments on Research Methodology. The most
interesting difficulty that emerged in the course of this research was the
reluctance on the parts of many respondents to provide correctness ratings.
14 respondents gave all states the same rating (anywhere from a ”4" to a
"7”), while an additional four respondents simply left this task completely
blank. As they were actually filling out the correctness questionnaire sheet
itself, many respondents either wrote or said comments such as ”People are
correct for wherever they are from", "I don't consider speech in terms of
correctness or incorrectness, but in terms of difference”, ”Grammatically--
as seen by a prescriptive linguist!", and "It's all so subjectivel”. In the
follow-up interviews, people also offered explanations of their objections

to this task. Consider the following comments:

Well, of course, I think each individual thinks that they speak the
proper way, and so somebody that differs real drastically, you're
thinking they really don't know how to speak, you know, or they
don't know how to express themselves, but I'm sure that somebody
who speaks drastically different than I do thinks their speech is
perfect too (respondent #54)

I honestly don't believe that I have ever consider-- I have never rated
things when I was listening to them as being correct or incorrect, it
was just that person's way of doing it -- living in the neighborhood
where I did, why, we accepted everybody or we didn't get along and
- I I was aware that, well you can't say that what was spoken in
English was correct and what we spoke was not and it's the same
thing with what was spoken in the east coast and I never thought that
what was spoken in Boston was any better than what I spoke out here
(respondent #57)

...the correctness issue, I mean, well what is correct language
anyways? you know, it's all relative to who's looking at it and the
person you're coming from...(respondent #122)

76

[based on the fact that you haven't traveled much, what were you
basing your answers on?)

my inherent philosophical belief that there is no correct, in terms of
speech... (respondent #123)

The fact that so many respondents objected to the correctness task is
a sharp contrast to Preston's experience in giving this same task to

respondents in Michigan and Indiana. He says of his raters:

It should be noted that very few respondents complained about this
task...Although they complained that they did not have information
about this or that state, the ranking of most areas for correctness was
for them a reasonable task and represented opinions overtly held
about the sites where better and worse English was spoken (Preston
1993b, p. 31).

I believe this difference between raters in Oregon and Michigan and
Indiana is largely a result of the cultural and linguistic heterogeneity which
has been a part of West Coast experience since early settlement days and
continues to be a dominant force in the experience of westerners.

Another possible explanation for the difference between Preston's
raters and the respondents in this study may also be the factor of time.
Since Preston's data was collected roughly 10 years ago, the reluctance of
the respondents in this study may be an indication that Americans are
becoming more sensitive to the issue of what constitutes "correct"
language.

Interestingly enough, while there was so much objection to the
correctness task, there was little complaint about any of the other tasks.
Thus, to rate someone's speech as less pleasant than one's own didn't appear
to be as big an offense as labeling them "incorrect." One respondent

describes the difference in the two tasks in this way:

77

[why do you rate everyone the same on correctness but differentiate
in terms of pleasantness?]

Because it seems like it's an aesthetic quality and something that
brings you know, it's kind of an artistic thing it seems like it's
something you can appreciate where correctness to me seems to me
like there's some scale and there's some right and wrong to it and I
don't see that as relevant in terms of speech (respondent #123).

While Oregonians may be less dogmatic about what constitutes ”correct"
speech, it would not be fair to say that they are true linguistic relativists.
Clearly the results of the pleasantness ratings show that they have some
definite ideas that not all regional speech is equally acceptable.

While the research methodology and analysis worked well overall
for this study, some improvements could be made. The analysis of the
hand-drawn maps was a difficult undertaking, since it was more qualitative
than the other tasks. Also, quite a few respondents circled regions on their
maps but did not provide any labels for these regions. Without these labels
and/or more in-depth interviews to determine why people indicated the
areas they did, it is difficult to draw conclusion about what the respondents
had in mind while they were doing this task, or the ratings tasks for that
matter.

It should be noted that the respondent group in this study represents
only a small portion of Oregonians. They were almost all European—
Americans and residents of two urban areas on the western side of the
state. To make more comprehensive generalization about Oregonian
language attitudes, it would be necessary to survey a wider range of both

geographical and ethnic groups.

4.8 Conclusion. The results of this study show that residents of

Oregon tend to view the United States in terms of five major dialect

78
regions: West, Midwest, Northeast, South and Hawaii. Within each of these

major regions, there are a few smaller sub-regions, such as the Pacific
Northwest, (southern) California, metropolitan New York City, and
Florida. Between each region, there seemed to be one or two ”transitional"
or ”border” states. These included New Mexico (between West and
Midwest), Michigan (between Midwest and Northeast), Virginia (between
Northeast and South) and Missouri (between South and Midwest). The

overall findings of this study are summarized graphically in Figure 13.

 

 

 

 

 

 

Pacific Northeast Ngw J
Northwes Englan

- " ew York

W jest Iidwest City
Californl

South
Texas
Florida

 

 

 

 

“I'- Major Dialect Region Boundary
M Dialect Subregion Boundary

 

 

 

 

Figure 13. Oregonian Perceptual Speech Regions Summarized

The states of Oregon and Washington were rated the highest in terms
of correctness and pleasantness, while New York City was rated the lowest

in both categories. In general, the other western states as well as the

79
midwestem states were rated similarly, fairly high on both correctness and

pleasantness scales. The South, which was the most salient dialect region
for the respondents in this study, was rated poorly for correctness but
fairly well for pleasantness. Both Hawaii and Florida tended to fall into
their own groups in terms of the ratings, primarily because of the influence
of non-native English speakers in those states.

This study is the first to examine the perceptual dialectology of any
west coast residents. In order to gain a more general understanding of the
language attitudes of western residents, it is necessary to replicate this
research throughout the western region. At a minimum, it would be good
to survey residents of Washington, northern California and southern
California. Of course, each state has a unique settlement history, and
therefore the most comprehensive picture of western language attitudes
would require research in every state. This study is thus merely a

beginning to what could become a much more extensive research program.

APPENDICES

APPENDIX A

Research Tools

 

 

Please provide the following information for classification purposes only. In no way will
this information be used to identify you individually in published results of this study.

 

 

 

 

 

Name: Age: ________
Address: Sex: M / F
Ethnicity:

1 European-American
Phonelt: 2 African-American

3 Hispanic-American
Highest grade level completed: 4 Asian-American

5 Other

Would you be willing to participate in a short interview as follow-up to this questionnaire?
Yes / No

Please answer with city and state for each of the following questions:

Where were you born?

 

Where did you spend your elementary school years?

 

Where did you spend your high school years?

 

Where have you spent your post-high school years (list locations and approximate dates,
up to current residence):

 

 

 

 

Where did your mother grow up?

 

Where did your father grow up?

 

 

80

 

81
APPENDIX A

 

 

Consent Form for Oregon Study

1, . have consented to participate in a linguistic study
involving the identification, labeling and rating of ways of speaking in regions of the US.
I understand that both written and tape-recorded data (in the form of an interview) may be
obtained from me in this project. I have been informed that all demographic information
about myself in addition to all data I provide is strictly for the purpose of linguistic
research and will be kept confidential. In no way will I be placed at risk in this study. I
also understand that my participation in this study is strictly voluntary and that I can
withdraw my participation at any time during the project, including mid-process. Any
questions regarding this project or inquiries regarding the results of the study may be
addressed to: Laura C. Hartley, Department of Linguistics, A614 Wells Hall, Michigan
State University, East Lansing MI 48824.

 

 

Signature Date

 

 

 

AAAAAAAAA

 

 

 

 

 

 

 

 

 

1000
H
1000

 

 

 

 

 

 

 

 

 

0
a
0

 

 

 

83
APPENDIX A

 

 

DEGREE OF DIFFERENCE

Please rate the 50 states, New York City, and Washington, DC. on a scale of 1 to 4 as

follows:

l=people theresoundlikeme

2 = people there sound a little different from me
3 = people there sound very different from me

4 = people there sound so different from me I ain't understand them

Alabama

Alaska

Arizona _
Arkansas __
California _
Colorado _

Connecticut

Delaware _
Florida _
Georgia __
Hawaii _
Idaho _
Illinois _
Indiana __
Iowa _
Kansas _

Kentucky

Louisiana

Maine _
Maryland _
Massachusetts
Michigan __
Minnesota _
Mississippi __
Missouri __
Montana _
Nebraska _

Nevada

New Hampshire

New Jersey _
New Mexico __
New York City
New York _
North Carolina _
North Dakota _

Ohio

 

Oklahoma

Oregon _
Pennsylvania

 

Rhode Island _
South Carolina _
South Dakota _
Tennessee _

Texas

Utah

Vermont

Virginia

 

Washington DC. __
Washington _

West Virginia _
Wisconsin __

Wyoming

 

 

84
APPENDIX A

 

 

Correctness

Please rate the speech of people in the 50 states, New York City, and Washington, DC. on a

scale of 1 to 7 as follows:

Delaware _
Florida
Georgia
Hawaii
Idaho __
Illinois __
Indiana

Iowa

Kentucky _
Louisiana _

leastcorrect l 234567mostcorrect

Maine __
Maryland _
Massachusetts
Michigan __
Minnesota _
Mississippi _
Missouri _
Montana _

Nebraska __
Nevada

New Hampshire

New Jersey

New Mexico

New York City

New York
North Carolina
North Dakota __

Ohio

Oklahoma _
Oregon __
Pennsylvania _
Rhode Island _
South Cm'olina __
South Dakota _
Tennessee _
Texas

Utah

Vermont __
Virginia _
Washington DC. __
Washington _
West Virginia _
Wisconsin __

Wyoming

 

 

AL:

AK:

AZ:
AR:

CO:

DE:
FL:

GA:

ID:
IL:
IN:

KS:

KY:

State Name Abbreviations Used

Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan

Minnesota

APPENDIX B

MS:
MO:
MT:
NE:
NV:
NH:
NJ:

NM:

NYC:

NY:
NC:
ND:
OH:
OK:
OR:
PA:
RI:

SC:

SD:
TN:
TX:
UT:
VT:

Mississippi VA:
Missouri

Montana WA:
Nebraska WV:
Nevada WI:

New Hampshire WY:

New Jersey
New Mexico
New York City
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas

Utah

Vermont

86

Virginia

WDC: Washington DC.

Washington
West Virginia
Wisconsin

Wyoming

APPENDIX C

 

 

 

 

 

 

 

Sample Hand-Drawn Maps

mmmmmmmmmm

QQQQQ

mmmmm

 

 

 

 

87

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

89
APPENDIX C

 

 

 

 

90
APPENDIX C

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

use.
a sec tame .. .ca
\ .3: L

m 03.3
the Q

 

.4 he

 

91
APPENDIX C

 

 

CIA—3

3 can

 

 

 

 

 

 

 

 

 

 

a
o.
mamhmx‘OJC.
E
009 o
Qua-S
r o
4<om
.\..~
.../.

 

 

APPENDIX D

Labels Used on Hand-Drawn Maps

WEST

Geographical location

Western [6]

West [5]

West Coast [2]

West/Midwest

Western English

Western states

N.W.

Pacific NW

Northwest

Oregon

Southwestern (UT,CO,AZ,NM)

Four corner language
(UT, CO, NM, AZ)

Variety Descriptors
West US. English

Mixed

Evaluative Terms
Normal [4]
Normal Accent
Normal to me
Commoners

No Accent [2]
Similar

The same as me
Plain Western
Western Drawl (MT, WY)
Western Twang

Other Lgs/Countries/Ethnic Grpg

Spanish (AZ, NM)
Substantial Mexican & Spanish
accent (soCA, AZ, NM)

Spanish (soCA, soNV, UT, AZ, NM)
Spanish speakers & English speakers

(soCA, AZ, NM, TX)

Sound/Temm Qualities
Twangey Western (MT ,ID,WY)
Flat

Dry

Toppgraphy/Nature
More Mountain like

(UT.CO.NM)
Rocky Mtn Country

Cultural/Historical

like cowboys
(ID,MT,WY,NV,UT)

Old Oregon territory

cowboy (ID-KS)

Quiet
"cold" (WA)

PACIFIC NORTHWEST
Geographical location
Pacific NW [2]

Northwest

Nor West

Variety Descriptors;
West Coast English
Western English

Evaluative Term_s_
Normal [2]
No Acc

Sound/Tempo Qualities
Soft - TV like

93

APPENDIX D

CALIFORNIA
Geographical location
So. Calif [2]

Variety Descriptors
Californian English
"Valley talk"

Silicone Valley Gargin

Evaluative Term;
distinctive

Other Lgs/Countries/Ethnic Gr_'ps
Chinese (SF)
Vietnamese/IaosianlI-Iispanic (so.)
Local - Mex (50.)

Hispanic (30.)

Topography/Nature
Peninsula (SF)

Cultural/Historical
Leisure living
Hip-Californian

Other
local

ALASKA
GQgraphical location
Western [2]

West [2]

Pacific Western
Alaskan

Variety Descriptors
Western English

Evaluative Terms
Plain western
Normal

ALASKA (cont'd)

Other Lgs/Countries/Ethnic Grp§
More Canadian like

Eskimo

Indian Accent/Native

Alaskan Indian Influence

Culfitural/Historica]
Last Frontier
cowboy

Frontier

Other
Helpful

HAWAII
Geographical location
Hawaii [2]

Hawaiian [2]

Pacific Western

Variety Descriptors
Western US. English

Evaluative Terms

Normal
No Acc

Other Lgs/Countries/Ethnic Grp_s_

Hawaiian natives - native accent

Island - Asian influence

Hawaii with a mix of Asian

influence

pidgeon English

Multi-lingual Hawaii

Mixed Continental & Pacific
Island

Japanese Golfland

Sound/Tempo Qualities
Lovely

 

94

 

 

APPENDIX D
MIDWEST
Geographical location Topography/Nature
Midwest [8] Mosquitoe slapping shout
Midwestern [3] Midwestern Plains (KS, OK)
Northem/Midwestem Plains
North
Oregon (WI, MI) Cultural/Historical
Dakotas Central Farmers
N.D.lS.D./Wisc.
Wisconsin Minnesota Resppllings
Ohio Valley Chicaco accent
Variety Descriptors Other
Midwestern English gentile
Midwest Accent English don't know
Midwest Accent unqualifiable
slight Midwest accent
Dakotas English NEW ENGLAND
Great Lakes English Geographical location
southern accent New England [3]
Midwest drawl N .E.
slight draw (ND,SD, NE, MN, IA) Northern N. England
slight Eastern draw Nor-wester

Talk w/ accent (Chicago)

Evaluative Term;
less accent

Other Lgs/Countries/Ethnic Grps
More Scandinavian like
Germanic lgs

Canadian

French Canadian Influence (MI)
Northern European

Sound/Tempo Qualities
Twangy, brash
Nasal

Boston (MA, CT, R1)
New England - Boston

Variety Descripgrg

New Englanders

New England - Bostonese
New England English
Maine Accents

Other Lgs/Countries/Ethnic Grps
European

Sound/T empo Qualities
clipped and direct (ME)

Cultural/Historical
"Back Bay"

95

APPENDIX D

NORTHEAST
Geographical location
Northeast
Northeastern

Nor East

New England [4]

East Coast [4]

Eastern [4]

East [4]

Northern

Mid Atlantic
Brooklin

New York/New Jersey

Variety Descriptors
Northern Accent

Eastern Accent English

Eastern Accents

New England English

New York Acc

New Yorkers

Maryland English
(WV, VA, MD, DE)

southern accent

thick brogue

mumbo jumbo (WDC)

Eastern Heavy Draw

speak w/ accent where they draw out
vowels

Accent

Boston Accent/New York Accent

Other Lgs/Countries/Ethnic Grp_s
Older English terms

sound Jewish

British mix

More British like

Sound/Temm Qualities

slower except in New York
harsh, talk fast

meaningless mumble

Nasal sound

Nasal [2]

Cultural/Historical

Old English/Colonial

Eastern upturned nose British
immigrant wannabees in love
with the Queen Mum

NEW YORK CITY
Geographical locations
Bronx

N .Y.

New Jersey/New York
Northeast

Variety Descriptors
NY English Accent

New Yorkers
Yankee English (cf. Cult/Hist)

Cultural/Historical
Yankee English

Resgllings
New York, New Yawk!

nuew yawk

APPENDIX D

SOUTH
Geographical location
Southern [17]

South [5]

South east

Deep South

VarietyJDescriptors
Southern Accents [3]
Southern English
Southeastern Am. Eng.
southern accent/drawl
Floridian drawl
Southern drawl [4]
drawl [3]

Heavy draw

talk w/ drawl

southern twang

Rebel slang

Accent

English with Southern Accent

Other Lgs/Countries/Ethnic Grps
More Scottish like
slower "French" roll

 

 

Sound/Tempo Qualities
slower "French" roll
slow [2]

Resmllings

suthron
Topography/Nature
Backwoods

Tennessee Hills

Cultural/Historical

Rebel slang

Hillbilly (TN, KY, WV, VA, NC)
hillbillish (KY, TN)

TEXAS
Geographical location
Southwestern [2]
Southwest

Texas [5]

Texas - Oklahoma
Oklahoma (OK)

Variety Descripprs
Texan Accent
western drawl

Heavy country twang
southern drawl [4]
Texas drawl [3]
Southern Acc

Texan [2]

Near Texan (OK)

Other Lgs/Countries/Ethnic Grps
Spanish

Cultural/Historical
Tex Mex

LOUISIANA

Geographical locations
Arkansas (AR, LA)

Variety Descriptor§
Louisiana English

Other Lgs/Countries/Ethnic Grps
More French like

French

"Cajon"

French/Cajun

En glish/French Creole

Topography/Nature
Bayou

97
APPENDIX D

FLORIDA
Gegraphical location
Miami

Variety Descriptors
Midwest drawl

Evalruative Terms
very little accent

Other Lgs/Countries/Ethnic Ggps
Cuban

French Creole/Spanish

Spanish

Hispanic

Cuban Spanish/English speakers

Cultural/Historical
7-1 1 Turbins

APPENDIX E

Ratings (Raw Numbers)

Table 28. Degree of Difference Ratings (Raw Numbers)

1

 

Ewe
State

1

 

Rating
State

 

 

18
24
19

35
27
36
31

11

OR

10

62

WA

CO

25
19
17
20
29
3O
29

3O
38
38
34
29

ll

55

10
I4
14
14
17
25
19
34
33

55
52

51

wmu

VA
NJ

47

25
20

42

3O
38
18
31

32

NH
NC

10

29

H1

30

OK

SD

29

31

32
42

27

NE

20
19
17
17

TN
TX
SC

22
22

41

39
38
43

22

36
41

NYC

AR

21

20
19
18
16
14
12

OH

20

ll

41

IN
IL

45

41

52

GA

KS

51

MS

14
10

34
42

PA

 

 

98

APPENDIX E

Table 29. Correcness Ratings (Raw Numbers)

7:

3

 

Rating

State
OR

 

l8
19
17
15

27

25
21

WA

12
11
l4
l4
14
12
10
16
13
11
16
17
14
11
13
16
12
13
17
12
17
18
15

21

CO
ID
CA

11

21

l7
17
14
15
14
14

22

19
18
18
15
15
15
13
13

11

l3
14
12

mwa

14
14
16
14
21

19
18
13
13
12
18
17
15
15
15
14
13
12
11
11

11

IL

11
11

l6
I3
18
17
14
15
10
2O
20
18

11

10
10
10
10
10
10
10
10
10
10
10
10

wmwm

IN

OH

15
22
14
13
13

10

12
I4
15
14

H1

17
15
15

NH
NE

 

100

APPENDIX E

Table 29 (cont'd)

 

SD

20
19
18
19
19
16
17
16

21

10
2O
20

14
10

KS

VA
NJ

10
13
10

12

18
15
15

11

10

SC
NC

11

21

TN

10

17
2O
17
18

21

16
16
13

15

15

11

12

TX

12

NYC

GA

10
12
ll

13

16
11
l4
12

14
17
I6
18

10
10

MS

 

101

APPENDIX E

Table 30. Pleasantness Ratings (Raw Numbers)

,7

 

‘Rating

State
OR

 

13
18

11

32
26
25

WA

13
12
19
14
16

10
12
15
10

ID

13
12
15
13
18
18

22
18
18
17
17

CO
NV

CA

12
l6
14

10
9

mm

13

17

ll
16
19
11
14
18

14
8

18
17
16
13
18
25

14
14

mm

16

11

NE

18
18
15
10
13
17
12

13

11

21

21

15
10
17

ll

15
15

15

OK
MS

10

20
16

OH
TN

14

10
5
6

10 15
10 16
2O 12

15
12
11

Rum

20
18
19
18

13
l4
l4
17

10
19
17
16

IN
IL
SD

 

102
APPENDIX E

Table 30 (cont'd)

 

18
l8
13
19
14
17
17
17
ll
15
22

13
I3
19
12
12
12
13
12
l4
l6
13
12
14
18
l7
l4
l6
18

16
16
15
14

11

22
17

17
14
12

ll

10

TX

12

16
10

10
22

AR

VA

2O

16
16

11

NC

13
14
ll
16
17

NH

11
11

16

11

NJ

11

10
17

15

NYC

 

APPENDIX F

Select Transcripts of Taped Interviews

Respondent #104 and husband

[Is Oregon your favorite place?]

(husband)

..to live, yes. to visit-~we like to go visit, but it's always nice to come
home, appreciate Oregon more every time we leave..Oregon has a lot of
diversified landscape-you can go to the coast and the high desert, eastern
Oregon--it's all different terrain, like most states back east you don't have
that, they have, well, most of your New England states are all wooded, they
have one type of landscape-here we have a large diversity...

[Do people sound different as you travel around the U.S.?]

Oh back east, yes. When we used to go to the flea markets in, well,
Greenfield and those we used to think New Yorkers were mad at each
other--that's just how they talk, they always sound like they're arguing and
that...yeah, and they're just, that's the way they speak, and we used to kind
of, they'd yell and scream and you'd look back thinking they were fighting
and they were just visiting

[Do New Yorkers come up to that flea market in Greenfield?]

Oh yeah, and they're elderly people, older people, and it's not like they're-
-most flea markets have fairly young people, but in Greenfield it's the
older, professional antique people out of New York who come-they're not
the only ones there but the ones from New York seem to be older people,
they're doing it professionally and they're good at it, they're good at it,
you don't beat those fellows. It's very competitive back there. It used to
be that you could go back east and buy things inexpensively, but now it's
very competitive, you can actually buy it here cheaper than you can back
there...We collected the more collectable things, back east they collected
the more fine, better antiques, so they'd throw out the oaks [the primitives]
yeah, there you go, the primitives, the oaks and the primitives and of
course we'd love to have 'em, but now they've changed their approach on
the primitives and of course, real old primitives are real expensive back
there...

...We take a lot of short trips in Oregon, and Montana and Idaho--there's
some beautiful country in Idaho...

103

104
APPENDIX F

( #1 04)

I have a friend who has lived in the United States for, oh about 25 years,
but she's Canadian, but she continues to, I always notice the "could" and the
"would"...and another friend too whose parents were Canadian but she's
always lived here, but you can tell it in those words ["about"?] "about" !,
yes, yes.."aboot"--it is, I like linguistics are interesting...you can detect so
many little things...

(husband)

...most beautiful state in the Union and she said "your coastline is
gorgeous"--we've lived here all our life and when we left that restaurant
we said, "my god, let's go home and go see our coast", and we did, and we
did, and now, when we go to the coast we thoroughly enjoy it, but we were
so accustomed to it that it meant nothing, until we go back east and the east
coast is not, not very pretty--it's flat, you don't get to see it [but you get to
swim there] yeah, we didn't...

(#104) that's an advantage, you're right

(husband) you mean it's warmer, the water's warmer? [yeah...they call NJ
Philadelphia's sandbox...) ...well we were up further north, up in Maine
and CT ...and that's about equal to our area here...well, thanks to that gal in
that restaurant we came home and we appreciate our coastline more and
more everyday

(#104) it's so breathtaking because you're high

...(husband) oh you've been over there? See, north of Florence is where
the pretty part starts...it's one of the most beautiful views on our coast...

...one of our favorite areas back there, of course, is D.C., it should be
mandatory that every teenager goes to Washington, D.C...it should be...

Respondent #119 and #120

( #1 19)

...We lived for a short time in Idaho when we were first married..but we
hated it...

...I lived in California for a short time when I was a teenager and wanted
to be away from home, to start my own life, and I hated it...I was in L.A.,
well I lived in Wittier, which was outside L.A....I only lived there 11
months and I was ready to come home. I remember the first time it
rained, I looked up in the sky and I cried and said, "Thank you God"...

105
APPENDIX F

{regarding people with a KY accent} ..they act like they've got all the time
in the world, you know--"nothing wrong with me, I'm just taking my
time," I can't do it, but it sounds great...

{comments made while drawing map}...I don't pay attention to their
accents, except when I go way over here to the east coast or way over in
Louisiana...this is kind of mixed a lot, and I suppose it's mixed over here,
but see I don't know the difference between-~this is the east coast—~but I
don't know where like Louisiana would be, where's Louisiana?mThis is
purely guesswork...my husband's family came from Iowa...

(#120 )

...when you're travelling you're usually talking to tourists, so that's why I
say this whole thing is ridiculous--you can quote me on that!...I wish I had
that quote from this man that said, he had three or four sentences, he'd
picked the sounds involved in the words in those 3 or 4 sentences and he
could identify where you were raised, not necessarily where you were
born--it's a cultural thing primarily [did he do that by ear or did he use a
machine?] well, I, you learn to speak by ear, I suppose, but you imitate the
people where you live, we imitate speech just like we imitate anything
else...

(#119)

{filling out pleasant}...l really do like everybody's language, but I think to
be "cutsey", certain ones sound cuter, you know, cause you like to listen,
kind of like you like to listen to an Aussie, you know someone from
Australia, because they're unique, so does that count as...?

{filling out correct} I bet there's a lot you can learn from this...prejudices,
for instance the rating how most pleasant, if you're not particularly fond of
those type of people or the people who live there...I think this is
fascinating...this is interesting what I'm doing-gradually as I go to the east
coast I'm raising the quality [what's that mean?] more wealthy, I don't
know, I'm just interested as to why I'm doing that...oh, not necessarily, oh
well

106
APPENDIX F

Respondent #54

I really love Western Oregon...

[I'm curious as to as you were rating these, what you were thinking of?]
...there are parts of the states that I've never been to, so maybe I've only
heard the people's accents or their conversations on a TV show or a radio
program, so my rating is purely based on that...

[any neighbors or people you've met here who are from different parts of
the states?]

Yes, um, I have a son-in-law who's from TX, and it was really hard to
understand him when he first came up, and I thoroughly enjoyed listening
to President Kennedy speak, I like his eastern seaboard accent very much,
people from the southeast are very hard for me to understand, their ways
of expressing themselves-J really find that all people's ways of expressing
themselves with words is very interesting, like our oldest daughter always
says she has a love affair with words and it's true, words are so expensive
and so picturesque that I really enjoy listening to people, no matter where
they're from...

{about filling out the map} it was really hard to know where to divide for
the different areas, the way people talk in this group of states or that group
of states, but, it really was an exercise in my mind to decide where the
dividing line would go....

[how did you come up with the different terms that you used?]

Um, mainly by the type of people that live there and the area of the United
States that they happen to be in, you know, central, southwestern, or...

[have you met people from most of these areas?]
I think pretty much so, yes, I've met a lot of people from a lot of different
parts of the United States and a lot of foreign countries too

[what about the correct sheet, how did you come up with some of those
determinations?]

Well, of course, I think each individual thinks that they speak the proper
way, and so somebody that differs real drastically, you're thinking they
really don’t know how to speak, you know, or they don't know how to
express themselves, but I'm sure that somebody who speaks drastically
different than I do thinks their speech is perfect too, so it really is hard to
check on that which you think, you know, are they speaking properly or

107
APPENDIX F

not? I'm sure for all of us we aren't speaking really proper English that
came from England, the original, so you know it's an American mixture

[anything else you can add?]

...I find people's speech, the way they express themselves, interesting and
really educational. I like to talk with people that are really quite well
educated, because I'm not well educated it kind of puts me in a corner, but
just the medium everyday person is, people that I feel more comfortable
with...

Respondent #71
I've been to practically every part of Oregon except the southeast...and we
. both love to go on the side roads and various town and places...

[did you find that people all around Oregon pretty much talk the same way
or did you find difference?]

Oh, I think that if they had lived in Oregon for any length of time they talk
the same way, don't you MaryLou?

{MaryLou: Uh huh}

I think they were very similar in speech

[I had one man tell me that over in the eastern part of the state there was a
little - different - but you, you didn't...)

Well, it could be, more the cowboy type of living -- but a lot of people
from the eastern part of the state come over here to live and, and play

[did you find a real distinct difference in the Canadian speech just on the
other side of the border?]

Canadian speech? Yes, somewhat. A lot of people speak like the English,
you know with an English accent and also with the way they form their
sentences and speak you know -- I've never been to England but I'm sure
that's the way they talk

[did you find any difference between northern and southern California in
the people or the way they talk?]

Not in the way they talk particularly, of course northern California is more
like Oregon in climate and what but, uh as far a talking is concerned I
couldn't tell any difference. Maybe the people that I talked to are just
friends and relatives and so on and I wouldn't ( )

108
APPENDIX F

Respondent #57

...it has been my impression all the time that I had lived on the west coast
that there was no, no pattern - my father had a, a pattern of speaking that
was different but (...) the constant influx of other people has changed the
language in exactly, considerable Spanish which was used as a child in the
area has changed and gone out and there isn't that much but in the ( ) that
we worked with when our children were young there were people from
every nation on the face of the earth. We had 19 girls and I don't think
there were more than three of them, maybe maybe 5, that could be called
anglo, anglo—American from the east coast but we had one that was directly
from England, now she did talk funny (laugh) she talked a different
language, we had some from Guatemala, Puerto Rico, Argentina, our
neighbors were completely mixed, we were fairly close to UCLA and the,
the married student housing there and so we had, they weren't really within
our area but they were really close and a lot of the apartments in our area
were student families or people who would come in relocating, we had a
lot of very well educated people, educated from other areas but hadn't
really established themselves in our area yet in this country and so they
were in a rather blue-collar neighborhood, even though they had much
more - things but I hadn't, I've only been once to Alabama for a couple of
weeks and I believe we went to the Grand Canyon once but I didn't talk to
anybody because I was camping -- and I really couldn't say anything about
the speech patterns because I've always had to accept all kinds of speech
patterns, all kinds of languages, we had the Japanese people that have lived
in California for quite a long time, which speak precisely but have certain
um, expression that they use but we have also down there the ones that have
come in much later and we have people that learn the new language
perfectly, people that don't -- learn the language, I had one set of neighbors
that, uh the man would do the speaking and the woman would listen, we
were building a brick wall between the two properties so we were
negotiating how we were going to split the cost, why uh, they both had to
come over a talk with us, one of them had worked where he had a Mexican
crew, they had been teachers in Cuban before they had fled and so he was
able to work, he had a supervisory job of some kind with a Spanish-
speaking crew from Mexico and he could speak the language very well.
She watched TV at home, she did not work outside the home and she could
understand it quite well and so it took both of them to sign the papers as to
who was going to pay for this fence (laugh) and I couldn't fill out those
papers thoroughly because I just never judged, if I could understand people
that was fine, we would, we would converse in any kind of sign language

109
APPENDIX F

and -- my daughter's Japanese friend's mother always smiled and said "Ah
so" like she understand every word I said to her and later my daughter says
"It's too bad Sachi's mother never learned English" -- I'd been talking to
her for a long time, she recognized me (laugh) a yeah, we knew each other
when we saw each other, I guess I had many friends that I couldn't
communicate with and my speech patterns, my children said they were
very strange -- so, I guess I never even bothered to change mine when the
language changed or didnn't watch TV as much as the others now TV tends
to level out the language and you don't keep many colloquialisms, things
from home after you have children that are raised with a lot of TV-- so I
didn't talk like TV mothers so my children thought I was (...)

I honestly don't believe that I have ever consider-- I have never rated
things when I was listening to them as being correct or incorrect, it was
just that person's way of doing it -- living in the neighborhood where I did,
why, we accepted everybody or we didn't get along and - I I was aware
that, well you can't say that what was spoken in English was correct and
what we spoke was not and it's the same thing with what was spoken in the
east coast and I never thought that what was spoken in Boston was any
better than what I spoke out here I mean, uh, my dad used to read some of
the books by the cowboy authors, that wrote a lot and I can't remember the
man's name, one of 'em used to write a lot about his adventures, you know,
with the cow camps and things like that and he'd like to read that because it
was the kind of language he had heard when he was a boy in California
which wasn't spoken any more and it wasn't even spoken in our home --
and certainly my kids didn't pick it up, they didn't pick up very much of
my vocabulary

[did you have any experiences with people from other states within the US
that spoke differently?]

well, yes, yes, as I say all my friends in childhood had come from other
states and there was almost no one in Long Beach when I lived there that --
had realy been there for very long and the same was true in Oregon people
came in and it wasn't settled, you know the oldest churches and the oldest
schools and things down in Grant's Pass are only 100 years old, so that's
that's within my parents didn't quite make 100 but they, they're gone now,
but there's just just, changed radically and it's in a constant flux -- I
couldn't say, how can you say one person's language is better than others
it's better if the person can understand you and this is the reason my
children don't know -- I was born before TV, I read books, my family had

110
APPENDIX F

a library and whenever they got a chance they would get books and we
would have books in the house and you played cards or played checkers or
read a book in the evening because that was what you could do -- and uh,
you don't -- I have a lot of books and I notice reading some of the books
that I read younger now, I thought where did these words come from? I
think, yeah, I know what these words are but I haven't heard them for
years. It was really, every is like advertising literature now and short
sentences and not, the beautiful language is not being used because if
anybody does write it everybody's in such a hurry to read it, you can't you
can't do that now

Respondent #121

[the first thing that I noticed is that you didn't do anything with these states
in the middle. why is that?]

I think it would be, it seems like it would be more of a mixture 1 mean, I
mean I know on the west coast there's a mixture of people with different
accents and stuff like in eastern Oregon and stuff speaking more of what I
guess what I would consider midwestem, but I didn't really know...

[in try to do this whole thing, what did you try to base your answers on?]
well, some of them I based on people I've met from those states, like
Michigan, I know several people from Michigan sol guess I have a
conception of that and how it sounds, um east coast you know I know some
people from Pennsylvania and certain states and probably New York City
is mostly from TV and movies and stuff...

[do you consider WA and OR to be pretty similar?]

Yeah - I mean it depends on where you're talking about in Washington and
Oregon but [for example, what?] well, I mean if you like, like I was saying
the eastern side of the state is more of a, sounds more twangy or
something, I'm not sure what, it sounds more country and I don't know
why that is

[why the very different rating for CA?]

Well, I was thinking about it and I think there's just, I don't know some
speech patterns, it just - sounds different [in terms of pleasantness-- not
necessarily in terms of correctness?] Yeah, right.

111
APPENDIX F

[how did you find this task?] It was hard. [why?] just because I don't feel
like I have a very good concept of a lot of areas, the way they speak
because I haven't been too many places

Respondent #122

[can you explain your labels to me? first of all you've got this one that
includes kind of the NW and it's "Western Twang"]

Because just cause the scenery the way it is I think the people understand a
certain level of outdoor terms that most people don't from the outside
[and then you have "Silicon Valley Jargon" which is California]

Um, yeah, people in California talk their own language which I don't
understand [what do you mean by "talk their own language?"] their whole
terminology is a different - task it seems like their words are bigger and a
lot of it seems to be, at least in the areas I've been in, technical terms and a
lot it I guess I think of Silicon Valley in regards to computers and all the
technical variety

["Rebel Slang"] It's just because I've spent a lot of time studying history of
that area and when I'm with friends I can pick it up really fast and it seems
very "rebelish" after visiting there and then going to a another part of the
states everybody told me that I sound like a rebel so I equate that with
["meaningless mumble"] people seem to talk very fast up there from what
I've experienced

I can be in Washington and not feel like I'm not in Oregon but I can be in
Idaho and I can tell a difference..it's sort of still the same area

[what did you base your ratings on?]

when we used to travel I used to love to get out of the car and listen to the
gas station attendants talk and that's where I got some of mine from, from
what I remember some of the states it's sound crazy but it's from what
some of the gas station attendants talk, cause when you're driving through a
state you don't get much of a chance except for listening to people on the
radio and how they talk

[why such a low correctness rating for TX?]

it's just seemed very different to me, the family I stayed with I couldn't
understand a lot of what they were saying and their general terms, I felt
like I was kind of in a different culture, cause their general terms are very
vague compared to what I was used to hearing

112
APPENDIX F

...the correctness issue, I mean, well what is correct language anyways?
you know, it's all relative to who's looking at it and the person you're
coming from

Respondent #123

[based on the fact that you haven't travelled much, what were you basing
your answers on?]

my inherent philosophical belief that there is no correct, in terms of speech
kind of, um so...

[why do you rate everyone the same on correctness but differentiate in
terms of pleasantness?]

because it seems like it's an aesthetic quality and something that brings you
know, it's kind of an artistic thing it seems like it's something you can
appreciate where correctness to me seems to me like there's some scale and
there's some right and wrong to it and I don't see that as relevant in terms
of speech

[why low ratings for California? I noticed on your map you divided
Southern California-~it has it's own]

it has it's own English and so I spent three years in California [in southern
califomia?] in southern California and um there's a variety of English
there that I adopted for a period of time and um [can you describe that
variety at all?] I think it's vocabulary and uh I don't think phonetically or
phonologically there's a whole lot going on that's different but it's more
terms and you know um prosody [but northern California seems to you to
go more with Oregon and Washington] yeah, it seems that speech forms
seem like they're more similar

I mainly thought of degree of difference in terms of sounds and how
closely they were to the way that I talk, and subjectively then I thought of
the map as more like the whole speech forms, maybe the words that they
have might be like the Germanic areas, in some of the migrations in the
Great Lakes areas that maybe the things that they say come from those
background languages and, so I think with this degree of difference I was
probably only thinking of sounds

113
APPENDIX F

[what about these areas that you left out?]

they just, they don't talk there (laugh) [this seems to be the biggest area
right here MT, ID, WY] They could, if you want me to put [no I don't
want you to change anything I'm just curious] they don't seem real
prototypical, if I was to make stereotypes, if I was to stereotype or
prototype any of these folks, those are kind of the areas that I would say
are most like each other

[what is for example "Dakota English"? The English of people in the
Dakotas is different than say the English in Montana or Wyoming?] I don't
have a whole lot of first hand experience but I've heard people say that,
that this, um that there are things here in this English that are a little
different but I don't have any experience up there, I've never been there
and I don't know.

LIST OF REFERENCES

LIST OF REFERENCES

Allen, H.B. (1973-76) The Linguistic Atlas of the Upper Midwest, Minneapolis:
University of Minnesota.

Allen, HE. (1977). Regional dialects, 1945-1974. American Speech, 52 (3-4),
163-261.

Atwood, EB. (1962) The regional vocabulary of Texas, Austin: University of
Texas.

Bright, 13.5. (1971) A word geography of California and Nevada. Berkeley:
University of California.

Cargile, A.C., H. Giles, E.B. Ryan, and 1.]. Bradac (1994) Language attitudes as a
social process: A conceptual model and new directions. language and
Communication, 14 (3), 211-36.

Carver, C. M. (1987) American Regional Dialects: A Word Geography, Ann
Arbor: University of Michigan.

Cassidy, F.G., ed. (1985-) Dictionary of American Regional English, Cambridge,
MA: Harvard University.

Condon, S. and Pittman, P. (1993) Language attitudes in southern Louisiana.
In Crochetiere, A., Boulanger, ].C., & Ouellon, C. (eds.), Endangered
Languages: Proceedings of the xvth International Congress of Linguists,
Quebec: PU Laval, pp. 311-14.

Coupland, N., Williams, A. and Garrett, P. (1994) The social meaning of
Welsh English: Teachers' stereotyped judgements. journal of
Multilingual and Multicultural Development, 6, 471—89.

Dailey-O'Cain, I. (1994) The wall in the mind: Language attitudes in Eastern

and Western Germany. In NWAV23 Abstracts, Stanford: Stanford
University.

114

115

Dailey-O'Cain, I. (1996) The perception of German regional speech. Paper
presented at AILA (11th World Congress of Applied Linguistics).
Jyvéiskyla, Finland, August 4-9.

Dernirci, M. (1996) The folk perceptions of the regional varieties of Turkey.
Paper presented at AILA (11th World Congress of Applied Linguistics).
vaaskyla, Finland, August 4-9.

Fasold, R. (1984) The Sociolinguistics of Society, Oxford: Blackwell.

Frazer, T.C. (1993) Settlement patterns and language variation: A model for
research. In A.W. Glowka & D.M. Lance (eds.), language Variation in
North American English: Research and Teaching, New York: Modern
Language Association of American, pp. 85—95.

Garrett, P., Coupland N. and Williams A. (in press) "City harsh" and "The
Welsh version of RP": Some ways in which teachers view dialects of
Welsh English. Language Awareness.

Grootaers, W. (1959) Origin and nature of the subjective boundaries of
dialects. Orbis, 8, 355-84.

Inoue, F. (1986) Sociolinguistic aspects of new dialect forms--language change
in progress in Tokyo. International [ournal of the Sociology of
language, 58, 73-90.

Kuiper, L. (1996) The parisian view of regional French. Paper presented at
AILA (11th World Congress of Applied Linguistics). vaéiskyla,
Finland, August 4-9.

Kramer, C. (1977) Perceptions of Female and Male Speech. language and
Speech, 20 (2), 151-61.

Kurath, H. (1949, reprint 1966) A word geography of the Eastern United States.
Ann Arbor: University of Michigan.

Labov, W. (1966) The Social Stratification of English in New York City,
Washington, DC: Center for Applied Linguistics.

Labov, W. (1991) The three dialects of English. In P. Eckert (ed.), New Ways of
Analyzing Sound Change, New York: Academic Press, pp. 1-44.

Long, D. (1996) Perceptions of regional variation in Japan. Paper presented at
AILA (11th World Congress of Applied Linguistics). Jyvaskyla,
Finland, August 4-9.

1 l6
Moreno, I. (1996) Linguistic attitudes of the city of Madrid. Paper presented at
AILA (11th World Congress of Applied Linguistics). vaaskyla,
Finland, August 4-9.

Newcombe, N. and Arnkoff, DB. (1979) Effects of speech style and sex of
speaker on person perception. Iournal of Personality and Social
Psychology, 37, 1293-1303.

Paltridge, I. and Giles, H. (1984) Attitudes towards speakers of regional accents

of French: Effects of regionality, age and sex of listeners. Linguistische
Berichte, 90, 70-85.

Pederson, L.A., S.L. McDaniel, G. Bailey, and M. Bassett (1986) Linguistic Atlas
of the Gulf States, Athens, CA: University of Georgia.

Preston, D. (1982) Perceptual dialectology: Mental maps of United States

dialects from a Hawaiian perspective. Working Papers in Linguistics, 14
(2), 5-49.

Preston, D. (1985) Southern Indiana perceptions of "Correct" and "Pleasant"
speech. In H. Warkentyne (ed.), Methods V/ Me’thodes V (Papers from
the Prfl'h International Conference on Methods in Dialectology), British
Columbia: University of Victoria, pp. 387-411.

Preston, D. (1986) Five visions of America. language and Society, 15 (2), 221-
40.

Preston, D. (1993a) Folk Dialectology. In Preston (ed.), American Dialect
Research, Amsterdam: Benjarnins, pp. 333-77.

Preston, D. (1993b) Two heartland perceptions of language variety. In T.
Frazer, (ed.), "Heartland" English, Tuscaloosa: University of Alabama,
pp. 23-47.

Preston, D. (1996) Personal communication.

Preston, D. and Howe, G. (1987) Computerized studeis of mental dialect maps.
In K. M. Denning, S. Inkelas, F. C. McNair-Knox & LR. Rickford, (eds.),
Variation in language: NWAV-XV at Stanford. Stanford: Department
of Linguistics, pp. 361-78.

Reed, C. (1956) Washington words. PADS, 25, 3-11.

Reed, C. (1957) Word geography of the Pacific Northwest. Orbis, 6, 86-93.

117

Reed, C. (1961) The pronunciation of English in the Pacific Northwest.
language, 37, 559-64.

Reed, C. (1977) Dialects of American English, Cambridge: University of
Massachusetts.

Rensink, W. (1955) Dialectindeling naar opgaven van medewerkers.
Amsterdam Dialectbureau Bulletin, 7, 20-3.

Ryan, E.B., H. Giles, and RJ. Sebastian (1982) An integrative perspective for
the study of language attitudes toward language variation. In E.B. Ryan

8: H. Giles (eds.), Attitudes towards language variation, London:
Arnold, pp. 1-19.

Shuy, R.W. (1983) On discovering language attitudes. In E. Dudley & P. Heller
(eds.), American Attitudes Toward Foreign languages and Foreign
Cultures, Bonn: Grundmann, pp. 71-93.

SYSTAT: Statistics, Version 5.2 Edition (1992) Evanston, IL: SYST AT, Inc.,
724 pp.

Trudgill, P. (1974) Sociolinguistcs: An Introduction to language and Society,
London: Penguin.

Trudgill, P. (1992) Dialect contact, dialectology and sociolinguistics. In K.
Bolton and H. Kwok (eds.), Sociolinguistics Today: International
Perspectives, London: Rutludge, pp. 71-9.

Williams, F. (1972). Relations between language attitudes and teacher
expectancy. American Educational Research Iournal, 9 (2), 263-77.

Williams, F. (1974) The identification of linguistic attitudes. International
Iournal of the Sociology of language, 3, 21-32.

Williams, A., P. Garrett, and N. Coupland (in press) Perceptual dialectology,
folklinguistics and regional stereotypes: Teachers' perceptions of
variation in Welsh English. Multilingua.

Wolfram, W. (1991) Dialects and American English, Englewood Cliffs, NJ:
Prentice Hall Regents.

 

"trill"lint