:- m "-3. ,-., m if ‘\

- 9-... _z_..", 'Is-‘;",‘ I
51"-? ’._.._.._ tt'fsau'z.~
h-—-L ——l~~# FE} “I? .. Q {1‘

B- O 9!

 

 

 

This is to certify that the

thesis entitled

INFLUENCES ON THE USE OF
FOREIGNER TALK AND FOREIGNER REGISTER
BY NATIVE ENGLISH SPEAKERS

presented by

Susan Kathleen Kitao

1

has been accepted towards fulfillment
of the requirements for

. Communication
degree 1n

9

 

 

WL‘M-

Major professor

 

 

Date June 27, 1986

 

07639 MS U is an Affirmative Action/Equal Opportunity Institution

 

MSU

 

 

 

RETURNING MATERIALS:
Place in booE drop to
remove this checkout from

 

 

UBRARms
.l..c.... your record. FINES will
be charged if book is
returned after the date
stamped below.
47— 2. 32.. J, ‘
.W « r . V /’/ xvi“)
W' ewé‘wébg
558 K220 “Q
I
fEB-GE‘BBQS ,.

 

 

INFLUENCES ON THE USE OF
FOREIGNER TALK AND FOREIGNER REGISTER
BY NATIVE ENGLISH SPEAKERS

BY

Susan Kathleen Kitao

A THESIS

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

”ASTER OF ARTS

Department of Communication

1986

ABSTRACT
INFLUENCES ON THE USE OF
FOREIGNER TALK AND FOREIGNER REGISTER
BY NATIVE ENGLISH SPEAKERS
By

Susan Kathleen Kitao

Foreigner t§L_ (FT) and foreigner registgg (FR) are
simplified language used to address foreigners perceived to
be deficient in the target language. FT is often ungram-
matical by the standards of the target language. while FR,
though simplified, follows the grammatical rules of the
target language.

I studied the effects of the grammatical proficiency
and fluency of a Japanese non-native English speaker on
native English speakers use of FR and FT. 1 also looked at
how prejudice against Japanese, ethnocentrism, experience
with non-native speakers. and experience using a foreign
language affected the choice to use FT or FR.

The results of this study indicate that differences in
grammatical proficiency and fluency did not affect use of FR
and FT. However, prejudice and experience with NNSs are
negatively correlated with the use of FR and FT; ethno-

centrism and experience with foreign language are positively

correlated with use of FR and FT.

For my husband, Kenji Kitao.
and
my parents. Menno and Ruth Ediger,

with deepest appreciation

iii

ACKNOULEDGHENTS

I would like to thank my advisor. Dr. Jack Bain. and
the members of my committee. Dr. Nary Bresnahan. Dr. Paul
Munsell and Dr. James Stiff for their assistance, advice.
and support. I would also like to express my appreciation
to Dave Collins, Kazuko Okamoto and Laura Santos for making
the video tape used in my study and Nancy Lesperance, Janice
Otto. Melissa Stewart, and Lisa Orechkin for their help in
coding. Last. I would like to Kenji Kitao for his support

and encouragement.

iv

TABLE OF CONTENTS

LiSt Of Tables. 0 o o o 0

Chapter I--Review of Literature

Introduction . . . .
Foreigner Talk . . .
Foreigner Register .

Strategies in FT and FR.

0

Triggers of FT and FR.

Types of Studies . .
Hypotheses . . . . .

Chapter II--Methods .
Overview . . . .
Participants . .
DeSign o o o o 0
Procedures . . . .

s

O 0 O O

Stimulus Material
Manipulations. .

Chapter III--Results. . .
Manipulation Checks.
ReliabilitieSo o o o
Scatterplots . . . .
Analysis of Variance
Correlations . . . .

Chapter IV--Discussion. .
Manipulation Checks.
Reliabilities. . . .
Major Hypotheses . .

O O O O O O

O
O
0
O

Correlational Analyses

Conclusions. . . . .

0

Suggestions for Future

0
O
O

O O 0

O O 0 O O
O O O O O
O O O 0 0
O O O O

O O O O O
O O 0 O O
O O O O O
0 O O O O
O O O O 0
O O O O O
O O O O O
0 O O O O
O O O O O
0 O O O O
O O 0 O O
O O O O O
0 O 0 O O
O 0 O O O
0 O O O O
O 0 0 0 O
O O O O O
O O 0 O O
O 0 O O O
0 O O O O
O O O O O
O O O 0 0
Research.

Appendix A--Scenario Questionnaire. .

Appendix B--Basic Dialogue.

0 O O O O O O O O O O O O O O O O O O O O O

O O O C O O O

O O O O O O O O O O O O O O O O O O 0 O O O

O O O O O O O

O

O

O O O O O O O O O O O O O O O O O O O O O O

O O O O O O O

O O O O O O O O O O O O O O O O O O O O O O

O O O O O O 0

Appendix C--Dialogue with Interference Errors

List of References. . . .

0 O O O O

O O O O O O O O O O O O O O O

O O O O O O O O O O O O O O O O O O O O O O O O O O O O O

O O O O O O O O O O O O O O O O O O O O O O

O O O O O O O

O

O O O O O O O O 0 O O O O O O

O O 0 O O O

O O O O O O O

O O O O O O O O 0 O O O O O O

O O O O O O

O O O O O O O

O O O O O O O O O O O O O O O O O O O O O

LIST OF TABLES

Table 1
Cell means with ratings of grammatical proficiency as
D.V.

Table 2
Cell means with ratings of fluency as D.V.

Table 3
Cell means with FT/FR as D.V.

Table 4
Source table for ANOVA with manipulation as I.V. and
FT/FR scale as D.V.

Table 5
Cell means with FT/FR. Item 6 as D.V.

Table 6
Source table for ANDVA with percieved proficiency as
I.V. and FT/FR. Item 6 as D.V.

Table 7
Pearson correlations
Table 8
Fourth-order partial correlations. from Pearson

correlations corrected for measurement error

vi

INFLUENCES ON THE USE OF
FOREIGNER TALK AND FOREIGNER REGISTER
BY NATIVE ENGLISH SPEAKERS
Chapter I
Review of Literature
tro t on

Ferguson (1971) has suggested that all speech communi-
ties have varieties of simplified speech regarded as appro-
priate for use when hearers do not have a full command of
the language. These include baby talk. ways of talking to
people who are hard of hearing. and ways of speaking to non-
native speakers of the language judged deficient in language
proficiency. These codes are conventionalized and are
learned by the same process of cultural transmission as
other language behaviors (Corder, 1975).

The purpose of this study is to look at some variables
that have been posited to trigger the simplified language
that is directed at non-native speakers (NNSs) of that
language.

Egcgigner Talk

The term foreigner talk (FT) was first used by Ferguson
(1971) to describe speech used with foreigners perceived to
be deficient in proficiency in the target language.
Ferguson speculated that FT was a representation of the way
native speakers (NSs) thought non-native speakers spoke.
since FT closely resembles the language elicited when NSs

are asked to imitate foreigners (Ferguson, 1975). In fact,

2
in English, secondary uses of FT include reporting the
speech of foreigners (Ferguson. 1981). Also, in German and
in Dutch, correlations have been found between the errors
made in NNSs’ speech and the errors used in the FT of their
NS partners (Meisel. 1977 and Snow, van Eeden and Muysken,
1981).

Some characteristics of FT in English that Ferguson
(1975) described included absence of the copula (be verb):
absence of certain morphological markers (e.g., possessives.
past tense. and 3rd person -s): and presence of words that
do not usually appear in English spoken to other NSs (e.g.,

avv as a verb and hggg [meaning much or veryJ). Other
characteristics include use of full forms instead of
contractions; short sentences: repetition of words;
selection of one all purpose form (e.g., mg used for 1, mg,
mlflg. and my in English or infinitive for all non-past verb
forms in Italian); and feedback devices. such as tag
questions (Ferguson and DeBose. 1977). These linguistic
features generally include the most stereotypical features
of the language of NNSs known to the community of NSs
(Valdman. 1981). FT often differs from what is acceptable
in standard language. as the characteristics listed by
Ferguson and DeBose would indicate.

Some non-linguistic features of FT have also been
described. They include slowing down and enunciating words

carefully; increased volume, especially for important

3
content words; and use of exaggerated gestures (Ramamurti,
1980). As Hammerly (1982) pointed out, these features have
the effect of distorting speech. It is also difficult for
NNSs who have come to depend on slowed. carefully enunciated
speech to become accustomed to language spoken at natural
speed.

Ferguson (1971) posited that FT contributed to the
development of pidgins. since it prevented non-native speak-
ers from hearing well-formed, standard language. Comprehen-
sible language is necessary for acquiring a target language
(Krashen and Seliger, 1975). If the comprehensible language
that a NNS hears consists of non-standard language, it is
non-standard language that is repeatedly reinforced.

Ferguson believed that FT was intended to derogate the
person it was addressed to. Valdman (1981) agreed and
suggested that NSs also use FT to maintain social distance
from NNSs considered socially inferior or subservient. The
NS recognizes his/her native language as an emblem of mem-
bership in the society and symbolically protects it from the
outsider. In this way, the NS exerts control by presenting
an impoverished variety of the target language. This denial
of the opportunity to assimilate has the effect of reducing
the foreign learner’s motivation and fossilizing the

foreigner’s language at a highly deviant stage.

Foreigner Register

In contrast to FT, there is a second variety of lan-
guage for speaking to NNSs. This variety is called
fgreigngr register (FR), a term coined by Arthur, Ueiner,
Culver. Lee. and Thomas (1979). FR was defined as a sim-
plified register which follows the rules of the standard
language. used by NSs in speaking to NNSs. The main dif-
ference between FT and FR is that FR follows the rules of
the standard language while FT does not.

Characteristics of FR that have been identified include
use of high frequency vocabulary, simplified grammatical
structure. and avoidance of idioms (Gass and Varonis. 1985).
FR also has a higher percentage of yes/no questions (e.g.,
Are you going to class?) and or/choice questions (e.g., Are
you going to class or to the cafeteria?) than wh- questions
(e.g., Uhere are you going?) (Long, 1981). Native speaker
register (NSR). used by NSs to speak with other NSs. has
more wh- questions than yes/no questions. Uhile FT can
generally be distinguished from FR or NSR by its pidgin-like
characteristics. the differences between FR and NSR are. in
many cases. quantitative rather than qualitative. There-
fore, this distinctions between FR and NSR are not always
clear and precise. There is extensive overlap between syn-
tactic forms that are used in FR and those used in NSR. The
same structure may occur in both registers. For example.

wh- questions and yes/no questions appear in both FR and

5
NSR; the difference is in the percentage of each (Long,
1981). This fact has sometimes made it difficult to
distinguish FR from NSR.

Not all researchers make the distinction between FR and
FT. Some (Henzl, 1979; Tarone. 1980) consider both to be
part of the same phenomenon. However. the distinction ap-
pears to be useful and will be employed in this paper.
Also. these terms are not universally used for the distinc-
tion. For example. Long (1981) uses the term gcammatigal
{urgigngc talk for FR.

It has been hypothesized that an NNS’s linguistic in-
sufficiency triggers FT or FR (Freed, 1981). or more ac-
curately. the NS’s perception of the NNS’s linguistic insuf-
ficiency (Gass and Varonis. 1985; Schinke-Llano. 1983).
Gass and Varonis speculated that in listening to and inter-
acting with the NNS, the NS forms an impression about
whether the NNS can understand NSR. If the NS forms the
impression that the NNS will not understand NSR. he/she may
use FT or FR to make messages easier for the NNS to under—
stand (Ferguson. 1971). i.e., to lighten the NNS’s interac-
tional burden.

Sinatggigs in FT and FR

This effort to lighten the interactional burden can
take several forms (Long, 1981). The most obvious is that
of simplifying grammatical structures. This notion of

simplifying can be nebulous, since it is not always easy to

6

determine whether one grammatical structure is "simpler"
than another (Ferguson and DeBose, 1977). Strategies for
simplification include omitting material (e.g., copulas and
articles): reducing irregularity; and using short, simple
sentences rather than longer ones with relative or sub-
ordinate clauses (e.g., "His aunt came to visit him. She
lives in California," rather than. “His aunt, who lives in
California. came to visit him.").

NSs may also simplify the lexicon and use high fre-
quency words words likely to be within the vocabulary range
of the NNS. Yes/no questions and or/choice questions are
more frequent in FT and FR. because it is considered easier
for the NNS to respond with a "yes” or "no" or to repeat one
of the choices given in the question rather than to respond
with completely new information (Long and Sato. 1983).

Other strategies involve not so much simplifying as
clarifying the message (Ferguson and DeBose, 1977). This
may involve adding redundancy to the message. supplying
material that is normally omitted (e.g., "You close the
door." rather than "Close the door.") and using full forms
instead of contractions or reductions.

NSs may also use different strategies for certain
functions when speaking to a NNS. For example, questions
have a higher frequency as topic-initiating moves in FT or
FR than in NSR. Long (1981) found that 96% of topic-

initiating moves were in the form of questions in NS-NNS

7
interaction but only 62% in NS-NS interaction. Long specu-
lated on several reasons for this. Questions lead naturally
to answers. while statements do not. Questions give the NNS
more encouragement to participate in the conversation. Lin-
guistic markers associated with questions (e.g., subject-
auxiliary inversion [changing “He can go." to ”Can he 90?")
and rising intonation) can signal the NNS that a speaking
turn is approaching. Questions encode at least part of the
expected response and therefore make responding easier
(e.g., it is easier to respond to "Uhere do you come from?"
than "I can see that you’re not an American."). It also
seems easier for the NNS to understand what response is
expected to a question than to a statement.

W of FT and FR
NNS traits that have been posited to trigger FR or FT

 

include pronunciation. grammatical proficiency, vocabulary,
fluency (ability to speak smoothly, without unnecessary
pauses). demonstration of lack of comprehension. and an
interaction among these (Gass and Varonis, 1985). NNSs
obviously vary widely in their ability to pronounce English.
and pronunciation may be taken as an indication of overall
proficiency. If the NNS’s production is not fluent. i.e.,
if the rate of speaking is slow or if the speech is
punctuated by long or inappropriate pauses. this is inter-
preted as indicating time required to process the grammati-

cal structures or to search for a lexical item (Butterworth,

8
1975). The NNS’s grammatical and lexical proficiency in
speaking may be taken as an indication of the level of
proficiency at which he/she can comprehend the language. If
the NNS demonstrates a lack of comprehension, the NS may
respond by switching from NSR to FR or FT, recognizing an
initial overestimation of the NNS’s proficiency.

NS traits may also affect the use of FR or FT. How-
ever, while numerous studies look at the effects that NNS
traits have on use of FT or FR. I could not find any that
systematically looked at the traits of the NS. One possible
factor is familiarity with non-native speech (Gass and
Varonis. 1985). Uhile Gass and Varonis do not try to ex-
plain this relationship. one possible explanation is that
familiarity with non-native speech may make the NS more
familiar with possible simplification strategies. which
he/she may put to use in FR or FT. Another explanation is
that the NS assumes that the NNS can comprehend what he/she
can produce. and bases FR or FT on utterances that NNSs
Produce.

Another possible variable is experience dealing with
NNSs (Hatch, Shapira, and Gough, 1978). Presumably. a per-
son who has had experience interacting with NNSs will be
more sensitive to their areas of difficulty and have well-
developed simplification and clarification strategies for
dealing with those difficulties. It was my experience in

conversing in Japanese that people who knew me seemed more

9
likely to use FR than strangers did. Strangers seemed more
likely to use NSR.

Another influence on the choice to use FT or FR may be
experience speaking a foreign language with native speakers
of that language. A person who has had the experience of
being the object of FR or FT in another language may be
aware of simplification or clarification strategies and
their usefulness in facilitating communication. I could not
find any studies that mentioned this relationship. However,
it was my experience that I became more conscious of the FR
strategies I used in English as a result of the FR that
native Japanese speakers used in speaking Japanese to me.

It is not clear what triggers FT rather than FR. Among
the traits that might affect the choice to use of FT or FR
are ethnocentricity (belief in the superiority of one’s own
ethnic group) and prejudice (an adverse opinion about a
certain group. formed without direct knowledge) against
foreigners or against persons of a particular nationality
(Valdman. 1981). As mentioned above, Ferguson (1971) and
Valdman (1981) have suggested FT indicates a desire to
maintain distance from foreigners. to indicate to them that
their assimilation in the target culture is not desirable.
If this is true, then ethnocentric N85 and N85 who are
prejudiced against foreigners would be more inclined to use

FT.

10
Types of Studies

Studies of FT and FR have fallen into three categories:
interactional. ideological, and fictional (Valdman, 1981).
Interactional studies involve observing NSs speaking with
NNSs in natural settings, recording the conversation. and
analyzing the results. Ideological studies are studies of
what NSs report that they think they would say in interac-
tions with NNSs. Fictional studies are studies of represen-
tations of FT or FR in literature. Each type of study has
have both advantages and disadvantages.

Interactional studies are obviously ideal. because they
provide examples of N85 actually using FR and FT. However,
they generally involve narrow linguistic descriptions of
language and lack a sociological perspective (Valdman.
1981). According to Valdman. most interactional research
involves studying language use in isolation and does not
take into account non-linguistic characteristics of the
participants. such as prejudice or ethnocentrism.

Another problem is that, while the general character-
istics of the stimulus utterances can be described, (e.g.,
heavily accented correct language or heavily accented broken
language). it is impossible to guarantee that all partici-
pants get exactly the same stimulus. even if the same NNS is
used. Because the participants get different stimuli. it is
difficult to isolate the characteristics of the language or

of the NNS that they are responding to. If more than one

11
NNS provides the stimulus utterances, this problem is
exacerbated. because there are even more differences among
the stimuli.

In a field study (as interactional studies typically
seem to be [Gass and Varonis. 1985: Long, 1983; Long and
Sato, 1983: Meisel. 1971: Hatch, Shapira. and Gough, 1978:
Henzl, 1981; and Ramamurti, 1980]), it is difficult or
impossible to measure NS attribute variables. Uhile the
language produced can be analyzed, the attributes of the NS
who produced it are unknown and may have effects on the
production of FR or FT that researchers do not recognize.

Ideological studies solve some of these problems, even
as they create others. In an ideological study, it is
possible to control the input from the NNS with a video or
audio tape and to measure the attributes of the partici-
pants. However. the study is dependent on the accuracy of
the participants’ reports of what they would say, rather
than on samples of what they actually do say. Valdman
(1981) argued that, because of the unconscious nature of the
use of FT and FR. these self-reports do have value, though
they might not show the entire picture. In spite of this,
he suggested that ideological studies and fictional studies
(which share many of the advantages and disadvantages of
ideological studies) were useful, at minimum. for iden-
tifying possible relationships that could be more fully

researched in interactional studies. As Ferguson (1971)

11
NNS provides the stimulus utterances, this problem is
exacerbated. because there are even more differences among
the stimuli.

In a field study (as interactional studies typically
seem to be [Gass and Varonis. 1985: Long. 1983; Long and
Sato. 1983; Meisel, 1971; Hatch, Shapira, and Gough, 1978;
Henzl. 1981: and Ramamurti, 1980]), it is difficult or
impossible to measure NS attribute variables. Uhile the
language produced can be analyzed, the attributes of the NS
who produced it are unknown and may have effects on the
production of FR or FT that researchers do not recognize.

Ideological studies solve some of these problems. even
as they create others. In an ideological study. it is
possible to control the input from the NNS with a video or
audio tape and to measure the attributes of the partici-
pants. However. the study is dependent on the accuracy of
the participants’ reports of what they would say, rather
than on samples of what they actually do say. Valdman
(1981) argued that. because of the unconscious nature of the
use of FT and FR. these self-reports do have value, though
they might not show the entire picture. In spite of this,
he suggested that ideological studies and fictional studies
(which share many of the advantages and disadvantages of
ideological studies) were useful. at minimum, for iden-
tifying possible relationships that could be more fully

researched in interactional studies. As Ferguson (1971)

12

pointed out, FR and FT are conventionalized in each language

and

are learned by speakers in the same way that the native

language is learned.

Hypgtheses

U

0\

The null hypotheses considered in this study were:

FR or FT will be used least with the NNS who has both
interference errors and exaggerated pauses.

FT or FR will be used less with the NNS with exag-
gerated pauses but no interference errors, than the NNS
who has interference errors but no exaggerated pauses.
FR or FT will be used least with the NNS who is rated
low in both grammatical proficiency and fluency.

FT or FR will be used less with a NNS rated low in
fluency than one rated low in grammatical proficiency.
Ethnocentric participants will adjust their language
less than participants who are not ethnocentric.
Participants prejudiced against Japanese will adjust
their language less than participants who are not prej-
udiced against Japanese.

Participants with experience with NNSs will adjust
their language more than participants with out ex-
Perience.

Participants with experience using a foreign language
will adjust their language less than participants with-

out experience.

Chapter II
Methods

mm

This experiment focused on the effect of two variables,
fluency and grammatical proficiency, on use of FT and FR.
Participants were shown video tapes of a NNS exhibiting
different levels of grammatical proficiency and fluency in
conversing with a NS in an effort to find out whether one of
these. or a combination. triggers FT or FR. In addition.
four other variables, experience with NNSs. experience with
foreign language, prejudice against Japanese. and ethnocen-
trism. were measured and their correlations to use of FT and
FR calculated.
Pa ' ants

Participants were 201 students from four Communication
100 classes at Michigan State University. The study was
done on a day set aside for research. Participants were
mostly freshmen and sophomores of various majors. There
were approximately fifty students in each class. Each class
was shown a different video tape.
Design

This study utilized a control group posttest design.
The independent variables (grammatical proficiency and
fluency) were presented to participants in four conditions:
(1) high fluency and high grammatical proficiency, (2) high

fluency and low grammatical proficiency, (3) low fluency and

13

14

high grammatical proficiency, and (4) low fluency and low
grammatical proficiency. Participants were administered a
questionnaire to determine their responses to FT, FR and NSR
in each condition. In addition, the questionnaire included
scales to measure the participants’ experience with NNSs.
experience with foreign language, prejudice against Japanese
people. and ethnocentrism.
Pr ures

The study took place in the participants’ classrooms
during the regular class time. Participants were shown a
video tape approximately one minute in length of a native
Japanese speaker and a native English speaker conversing.
After listening to the tape, participants filled out a
questionnaire that described different scenarios and asked
them to choose among utterances written according to NSR,
FT. and FR style. The questionnaire also asked them to rate
the native Japanese speaker’s pronunciation. fluency, gram-
matical proficiency. and overall English proficiency. Be-
cause of the possibility of testing effect. half of the
participants were given the scenarios first and the other
half were given the proficiency scales first. Finally.
participants answered questions about their experiences with
NNSs and foreign languages and questions intended to find
out whether they were ethnocentric and whether they were
prejudiced against Japanese people. (See the questionnaire

in Appendix A.)

 

15
m l ri ls

Part I of the questionnaire. the FT/FR scale, included
ten scenarios (see Appendix A. questions 1-10). Partici-
pants were asked what they would say in those situations.
They were asked to chose from among examples of FT. FR. and
NSR the response closest to what they would say in that
situation. They were, in addition. given an opportunity to
specify another response and asked to write out exactly what
they would say in that situation. if none of the alterna-
tives were close. If they chose to do this. the specified
response was classified as FT. FR or NSR by coders blind to
the group to which the participant belonged. Two coders
looked at the specified responses and classified it with the
FT. FR or NSR response that it most closely resembled. when
the coders agreed on the classification, the specified
response was changed to FT. FR. or NSR. If the coders could
not agree on the classification. the participant was
eliminated from consideration. Of 243 responses. the coders
were able to agree on classifications for 88% of them.

In order to get examples of NSR for these situations. I
gave a questionnaire to forty business communication stu-
dents. mostly juniors and seniors. asking them what they
would say to an American during an initial interaction in
each of the situations. Responses were chosen that seemed
representative of the responses given to that question and

that had characteristics not likely to occur in FR or FT.

16
These utterances were changed to FT using the characteris-
tics of FT described by Ferguson (1971) and Ferguson and
DeBose (1977). The utterances were changed to FR using the
characteristics described by Gass and Varonis (1985) and
Long (1981).

Participants were asked to rate the NNS on overall
English proficiency. pronunciation. fluency. and grammatical
proficiency (see Appendix A. questions 16-21).

This study considered four NS variables: experience
with NNSs. including experience with Japanese people: expe-
rience using a foreign language: ethnocentricity: and prej-
udice against Japanese. The first two were measured by
asking respondents about their background and experiences
(see Appendix A. questions 16-24).

Ethnocentricity and prejudice against Japanese people
were measured using two scales. The scale measuring prej-
udice is derived from one developed by Greenberg and Rosen-
field (1979) (see Appendix A. questions 32-41). Since I was
unable to find measures related to prejudice against Asians
in general or Japanese people in particular. I modified a
scale originally used to measure prejudice against Blacks.

The original scale consisted of bipolar adjectives. It
was rewritten as a Likert scale to make it easier for parti-
cipants to respond to. The scale consists of five general
statements about Japanese and the same five statements about

Americans. Since three of the dimensions of the original

17

scale (industriousness. politeness. and intelligence) seemed
more applicable to prejudice against Blacks in the United
States than against Japanese people. I substituted
creative/imitative. open/underhanded. and fair/unfair dimen-
sions. These seem more likely to be dimensions of prejudice
against Japanese people than the original dimensions. The
other two dimensions (reliability and honesty) were
retained. The scale is applied by subtracting the score for
Americans from the score for Japanese.

The scale of ethnocentricity was derived from one
developed by Bredemeier. Bernstein. and Oxman (1982). Of
the twenty-two questions included in the original scale.
eight were chosen that seemed to reflect ethnocentricity in
relation to people from other cultures in particular (see
Appendix A. questions 25-32).

In addition to the above scales. two questions intended
to measure sympathy for a NNS speaking English were included
(see Appendix A. questions 33-34). The intention of these
questions was to see whether sympathy for the NNS influenced
ratings on proficiency. i.e., whether participants who were
sympathetic toward the NNS’s efforts to speak English were
more lenient in rating proficiency.

[152.122.131.925

Of the possible NNS variables that Varonis and Gass

mentioned (pronunciation. grammar. vocabulary. fluency.

demonstration of lack of comprehension). I looked at two:

18
fluency and grammar. and at the interaction between the two.
Specifically. I looked at the effects that the Japanese
NNS’s interference errors (errors caused by direct transla-
tion from the NNS’s native language) and exaggerated pauses
have on NS’s use of FT or FR.

Exaggerated pauses were operationalized as long pauses
at sentence junctures and between phrases. According to
Nakasone (1979). these are the major differences between
pauses in the speech of native English speakers and native
Japanese speakers speaking English. In addition. long
switching pauses (pauses between the utterance of the cur-
rent speaker and that of the non-current speaker) were
inserted between the NS’s question and the NNS’s answer.
Switching pauses are a function of encoding difficulty or
the extent to which a response is readily available.
Switching pauses of more than three seconds in a dyadic
encounter. the focus of which is conversation. are con-
sidered "awkward" in the sense that it reflects on the
competency of the communicator (McLaughlin. 1984). If the
information contained in the response is readily available.
presumably a long switching pause is caused by encoding
difficulty. or. in the case of a NNS. possibly a combination
of difficulty with decoding the question and difficulty with
encoding the answer.

Interference errors were chosen because they are typi-

cal of the grammatical errors that NSs of Japanese make when

19

speaking English (see Appendix C for text of dialogue with
interference errors). Most articles (e. e3, and Lee) and
plural morphemes (-s) were deleted from Kazuko’s utterances.
In addition. two mistakes related to modifiers of countable
nouns (mush bus and much hendicre£_) and two mistakes
related to prepositions (fleer from Osake and graduation of
ESQ) were inserted. All of these are typical of
interference errors that Japanese people make in speaking
English (personal communication, M. Komiya. December 23.
1985: Bryant. 1984).

I made four video tapes of one-minute conversations
between a Japanese and a native English speaking American.
The content of the conversations was the same. The conver-
sations were an initial interaction between two female
classmates arriving early for class. They introduce them-
selves. and Laura. the American. asks Kazuko. the Japanese.
questions about where she is from, whether she lives off
campus. and what her major is. Kazuko answers the questions
and gives additional details (see Appendix B for text of
dialogue). These topics were chosen for the conversation.
because they are common in initial interactions (Kellermann,
Broetzmann. Lim. and Kitao. 1986). Laura’s utterances were
kept to a minimum so that the amount of FT. FR or NSR would
not prejudice the participants’ responses. Kazuko spoke
more than Laura. to give the participants a larger sample of

her language. and to allow her an opportunity to make a

20
variety of mistakes.

Both the video tapes and the questionnaire were pre-
tested with Communication 100 students. The questionnaire
was revised slightly in response to their comments.

The part of Laura was played by a 23-Year-old American
undergraduate. The part of Kazuko was played by a 40-year-
old Japanese graduate student. This age difference did not
seem to be a problem. since the students who viewed the tape
during the pretest estimated Laura’s age to be between
twenty and twenty-two and Kazuko’s age to be approximately
one year older.

In four conversations. the conditions were I. no inter-
ference errors and no exaggerated pauses. II. interference
errors and no exaggerated pauses. 111. no interference er-
rors and exaggerated pauses. and IV. interference errors and

exaggerated pauses.

Chapter III

Results
Danieuletign Qhecke
Pilgt Study. In order to identify research partici-

pants’ perceptions of the grammatical proficiency and fluen-
cy of the actor in each stimulus tape manipulation. checks
were conducted. To check the grammatical proficiency
manipulation. the grammatical proficiency ratings from the
pilot study were subjected to an analysis of variance
(ANOVA) with manipulation levels of fluency and grammatical
proficiency as the independent variables. Ratings of gram-
matical proficiency and fluency were used as dependent
variables.

Participants exposed to the stimulus with interference
errors rated the source’s grammatical proficiency sig-
nificantly lower (F = 13.11. p < .05. df = 1.23. h 2 = .35)
than participants exposed to the stimulus with no inter-
ference errors. Participants exposed to the stimulus with
exaggerated pauses also rated the source’s grammatical pro-
ficiency lower (F = 5.09, p < .05. df = 1,23. ll2 = .14)
than participants exposed to the stimulus with natural
pauses. Although the fluency manipulation seems to have had
some effect. these data clearly suggest that the grammatical
proficiency manipulation was effective. The cell means for
the grammatical proficiency ratings are found in Table 1.

Higher means indicate perceptions of higher proficiency.

21

22

Grammatical Proficiency

 

 

Interference No Interference
Errors Errors

F

l Natural

u Pauses 3.56 2.00

e

n

c Exag. 4.17 3.10

y Pauses

 

 

 

 

Table 1--Cell means with ratings of grammatical
proficiency as D.V.

when ratings of fluency were used as the dependent
variable. participants exposed to the stimulus with natural
pauses did not give the source significantly different
fluency ratings (F = .44. p > .05. df = 1.23. N2 = .03)
than participants exposed to the stimulus with exaggerated
pauses. Also. participants exposed to the stimulus with
interference errors did not give significantly different
fluency ratings (F = .09, p > .05, df = 1.23, [72 = .00)
from participants exposed to the stimulus with no inter-
ference errors. This indicates that the fluency manipula-
tion was not effective. as measured by the fluency ratings
of the participants. This problem will be discussed in
greater detail in Chapter IV. Cell means for fluency rat-
ings are found in Table 2. Higher means indicate higher

ratings of fluency.

23

Grammatical Proficiency

 

 

Interference No Interference
Errors Errors
F
l Natural l
u Pauses 2.89 4.00
e
n
c Exag. 3.17 2.50
y Pauses

 

 

 

 

Table 2--Cell means with ratings of fluency as D.V.

Because participants in the first pretest mentioned
that it was sometimes difficult to remember the NNS’s pro-
ficiency after filling out the FT/FR scale. a second pretest
was done. In this second pretest. participants only rated
the grammatical proficiency. fluency. pronunciation ability.
and overall proficiency of the Japanese speaker. Using
ANOVA to analyze the results. no significant main effects or
interaction effects were found. This indicates that par-
ticipants in the second pretest did not perceive any dif-
ferences among the proficiency levels of the NNS. and sug-
gests that the manipulation was not effective.

Cgfieee. In order to check the actual grammatical pro-
ficiency and fluency produced by the manipulation. as op-
posed to the perceived proficiency. two coders blind to the
purpose of the study evaluated the tapes. They counted the
number of grammatical errors and the number of pauses and
timed the length of the pauses. The reliability of the two
coders was 1.00 for grammatical errors. .99 for number of

pauses. and .94 for length of pauses.

24

Comparison of the mean frequencies of these coded be-
haviors served as another form of manipulation check. These
comparisons using one-tailed t-tests at the .05 level sug-
gest that the actor manipulated the desired behavior ef-
fectively. The number of pauses in the natural pause condi-
tion was less than the number in the exaggerated pause
condition (t = 2.64. p < .05). Also. the total length of
pauses in the natural pause condition was less than the
total length of pauses in the exaggerated pause condition
(t = 21.23. p ( .05). Finally. the number of interference
errors in the no errors condition was less than the number
of errors in the interference errors condition (t = 27.08.
p < .05).

The combined results of these manipulation checks pro-
vide evidence that the actor was able to manipulate the
intended behavior. The coders who observed these stimuli
perceived the differences in behavior produced by the
manipulations. However. these differences were not neces-
sarily perceived by participants rating the proficiency of
the NNS.

R 1 es

Reliabilities were calculated for all scales. For the
scale using FT. FR. and NSR (FT/FR scale), alpha was equal
to .48. For the proficiency scale. alpha was equal to .75.
Alpha for experience with NNSs equal to was .30. when Item

3 (See Appendix A. Item 18) was deleted. alpha increased

25

to .43. Because Item 3 had a low correlation with other
items. it was deleted for the purposes of these analyses.
Alpha for the experience with a foreign language scale
was .68. For the prejudice scale alpha was found to be
equal to .61. For ethnocentricity. alpha was found to be
equal to .40. Since Item 5 (See Appendix A. Item 29) was
not well correlated with the other items. it was deleted for
the purposes of analysis. This increased alpha to .45.

None of these reliabilities are more than moderately
high. Reliabilities for experience with NNSs. the FT/FR
scale. and ethnocentricity were particularly low. This
indicates that not all of the items of each scale were
measuring the same construct. or there were not enough
items.

mm

A scatterplot was made with the FT/FR scale as the
dependent variable and proficiency as the independent
variable. In addition. scatterplots were made with the
FT/FR scale as the dependent variable and individual ratings
for grammatical proficiency. fluency. and overall proficien-
cy as the independent variables. Scatterplots showed random
relationships between the pairs of variables. The correla-
tion between the proficiency scale and FT/FR was .04.
between fluency and FT/FR was .01. grammatical proficiency
and FT/FR .01. and between FT/FR and overall proficiency was

-.02. The relationships do not appear to be curvilinear.

26

In addition. there was no evidence of other forms of non-
linearity in these relationships.
agelzsis 9f Variance

An independent groups ANOVA was calculated with two
levels of the independent variables grammatical proficiency
(without interference errors. with interference errors) and
fluency (with natural pauses. with exaggerated pauses). The
dependent variable was the scores on the FT/FR scale. The
cell means are presented in Table 3. Higher scores indicate
greater adjustment of language.

Grammatical Proficiency

 

 

No Errors Errors
F
1 Natural
u Pauses 17.05 17.02
e
n
c Exag. 16.78 16.75
y Pauses

 

 

 

 

Table 3--Cell means with FT/FR as D.V.

The results of the analysis (Table 4) indicate that
manipulation levels of fluency and grammatical proficiency
had no effect on participants use of FT/FR. The effects for
fluency (F = .45. df = 1.172. p > .05.‘i\2 = .01). grammati-
cal proficiency (F = .01. df = 1.172, p > .05. 3? 2 = .00)
and the fluency by grammatical proficiency interaction
(F = .00, if = 1.172. p > £5,302 = .00) were all non-

significant and trivial in size. These results indicate

27
that the levels of grammatical proficiency and fluency did

not influence the participants’ use for FT, FR or NSR.

 

ss df ns F p eta2
Main Effects
Fluency 3.20 1 3.20 .45 >.05 .005
Gram. Prof. .04 1 .04 .01 >.05 .00
2-Uay Inter.
Fluency x Gram.
PPO‘FO 000 1 000 000 >005 000

Uithin cell
variance 616.95 172 3.59
Total 620.16 175 3.54

 

Table 4--Source table for ANOVA with manipulation as
I.V. and FT/FR scale as D.V.

 

 

 

Because it is possible that perception of proficiency
mediates between the NNS’s proficiency and the use of FT or
FR (Gass and Varonis. 1985). another ANOVA was computed with
participants’ perceptions of grammatical proficiency and
fluency as the independent variable and scores on the FT/FR
scale as the dependent variable. Participants were assigned
to high/low levels of grammatical proficiency and fluency
using median splits of individual ratings for fluency and
grammatical proficiency. ANOVAs were also computed using
the entire FT/FR scale as a dependent variable and each of
the individual items of the scale as a dependent variable.
No significant effects were found for the analysis involving
the FT/FR scale as the dependent variable. In addition.

when analyzed separately as a dependent variable. only one

28

of the ten items from the FT/FR scale produced significant
effects. Item 6 of the FT/FR scale (see Appendix A. Item 6)
was the only item that with a significant F-ratio. For Item
6. there was a significant F-ratio for the interaction
between fluency and grammatical proficiency. The cell means
and source table are presented in Tables 5 and 6. Higher

means indicate greater adjustment of language.

Grammatical Proficiency

 

 

Interference No Interference
Errors Errors
[7:
l Natural l
u Pauses 1.88 1.83
e
n
c Exag. 1.74 1.97
y Pauses

 

 

 

 

Table 5--Cell means with FT/FR. Item 6 as D.V.

 

 

ss df ns F p eta2
Main Effects
Fluency .08 1 .08 .61 >.05 .00
Gram. Prof. .21 1 .21 1.59 >.05 .01
l2-Uay Inter.
Fluen. x Gram.
i PPOTO 067 1 067 5012 (005 003
within cell
variance 22.54 172 .13
Total 23.45 175 .13
Table 6--Source table for ANOVA with perceived
proficiency as I.V. and FT/FR. item 6 as D.V.

 

 

 

29

This interaction does not appear to be a ”magic cell"
interaction. Special effect coding for a magic cell is not
significant. Examination of the cell means indicates that
the interaction is a disordinal one. There are no dif-
ferences in adjustment of language for the level of gram-
matical proficiency (with interference errors. without
interference errors) when pauses are natural. However, when
pauses are exaggerated. there was more adjustment of lan-
guage with interference errors than without interference
errors. The least adjustment was found in the exaggerated
pauses. no interference errors cell: the greatest adjustment
was found in the exaggerated pauses, no interference errors
cell.

rr l ' ns

Correlations were calculated among scales for FT/FR.
proficiency. experience with non-native speakers. experience
speaking a foreign language. prejudice and ethnocentrism.
Pearson correlations and their probabilities are reported in

Table 7.

29

This interaction does not appear to be a ”magic cell”
interaction. Special effect coding for a magic cell is not
significant. Examination of the cell means indicates that
the interaction is a disordinal one. There are no dif-
ferences in adjustment of language for the level of gram-
matical proficiency (with interference errors. without
interference errors) when pauses are natural. However. when
pauses are exaggerated. there was more adjustment of lan-
guage with interference errors than without interference
errors. The least adjustment was found in the exaggerated
pauses. no interference errors cell: the greatest adjustment
was found in the exaggerated pauses, no interference errors
cell.

rr l ns

Correlations were calculated among scales for FT/FR.
proficiency. experience with non-native speakers. experience
speaking a foreign language. prejudice and ethnocentrism.
Pearson correlations and their probabilities are reported in

Table 7.

30

 

FT/FR ethno- profi- exper. exper. prej-
centrism ciency with with udicel
FT/FR 1.00 NNSs foreign
language
ethnocen- .01 1.00
trism (.02)
PrOfi" 002 ‘00? 1000
Ciency (003) (-012)
exper. -.03 -.15 t .04 1.00
with (-007) (’034)* (007)
NNSs
exper.
with .09 -.28 * -.O2 .20 t 1.00
foreign (.16)* (-.50)* (-.03) (.37)*
language
Prerd- -008 020 * 004 -014 * -010 1000

ice (-.15)* (.38)* (.12) (-.27)* (-.16)*

 

Table 7--Pearson correlations
( ) = correlations corrected for attenua-
tion due to measurement error
* = p < .05

 

 

 

The correlations corrected for attenuation due to
measurement error include several interesting correlations
that are significantly different from zero. Ethnocentrism
has strong negative correlations with both experience with
NNSs and experience with foreign language. Prejudice has a
somewhat weaker. but still significant. negative correla-
tions with these variables. These correlations indicate
that participants who had experience with NNSs and with
foreign language tended not to be prejudiced or ethnocen—
tric. The correlation between ethnocentrism and prejudice

was also significantly different from zero.

31

Partial correlations between FT/FR and the other five
variables were calculated using the Pearson correlations
corrected for attenuation. In each of the five cases. the
other four variables were held constant. The results are

presented in Table 8.

 

ethno- profi- exper. exper. prej-
centrism ciency with with udice
NNSs foreign
language
FT/FR .16* .02 -.16* .24: -.20*

 

Table 8--Fourth-order partial correlations. from
Pearson correlations corrected for measure—
ment error
* = p < .05

 

 

 

Correlations between FT/FR and ethnocentrism. ex-
perience with NNSs. experience with a foreign language. and
prejudice. with four other variables held constant. were
significantly different from zero. Uhile these correlations
are not extremely strong, they do indicate that the NS
variables have an influence on the extent to which a speaker
uses FT or FR. This will be discussed further in Chapter
IV.

Chapter IV
Discussion
Man l on Checks

Peeteets. The results of the manipulation check showed
that the differences in pauses and interference errors among
the tapes were not reflected in the ratings of proficiency
made by the participants. In the first pretest. partici-
pants’ ratings of the NNS’s grammatical proficiency were
higher for the conditions where there were no errors than
where there were interference errors. However. these re-
sults were not replicated in the second pretest. where no
significant main or interaction effects were found.

It is difficult to draw definite conclusions from these
data. since contradictory results were found. Participants
of the first pretest seem to have been more aware of gram-
matical proficiency than fluency. Grammatical errors may
stand out more than exaggerated pauses. since participants
are likely to be more accustomed to thinking of ”good" and
“bad“ English in terms of grammatical correctness rather
than fluency. Exaggerated pauses may be attended to only
when the pauses are so long as to interfere with
comprehension.

These results indicate that a NNS who speaks with
exaggerated pauses is rated as being more grammatically
proficient than one who speaks with natural pauses. One

reason for this might be that. because pauses are taken to

32

33
indicate cognition (McLaughlin. 1984). a NNS who appears to
be pausing and thinking carefully about his/her words is
assumed to be producing better grammatical structures than
one who does not.

There are several possible explanations for the lack of
correspondence between the manipulation of the levels of
grammatical proficiency and fluency and the ratings of pro-
ficiency. One is that participants tend to be sympathetic
toward a person experiencing difficulty in communicating in
a foreign language. However. this explanation can be ruled
out by the small correlation between the sympathy scale and
the ratings of proficiency were low (r = .04, p > .05).

A more likely explanation is that it is difficult for
untrained participants to reliably rate the language pro-
ficiency of a NNS. Organizations that test the language
uproficiency of NNSs require evaluators to take hours of
training. (For example. the raters who will evaluate the
speaking proficiency of foreign teaching assistants at
Michigan State University next Fall must undergo ten hours
of training. much of which is spent listening to tapes of
NNSs and discussing differences among different levels of
proficiency [personal communication. Barrett, R.P., May.
1986]). This indicates that there may be problems in using
paper and pencil evaluations to reliably evaluate the lan-
guage proficiency of NNSs. so evaluations of untrained par-

ticipants may not reflect the ”real" proficiency of the NNS.

34

An additional complicating factor is that it may have
been difficult for participants to keep the proficiency of
the NNS in mind while filling out the scale. since they did
not fill out the proficiency scales until the tape had
finished. Since the participants are not likely to have had
experience in filling out such evaluations. the time lapse
may have decreased the validity and reliability of the
ratings.

Another explanation is that NSs may. in listening to a
NNS. be concentrating more on communication than on
proficiency. If there is no interference with communica-
tion. the NS may not pay much attention to the grammatical
proficiency or fluency of the NNS. Therefore. participants
may not distinguish between the grammatical proficiency and
fluency of NNSs of different levels as long as they can
understand the NNS.

These results have implications for future research on
triggers of FT or FR. Proficiency ratings of untrained
observers may not reflect the actual proficiency of the NNS.
Since training participants would be a lengthy process.
researchers may have to depend on the results of a
standardized test of speaking proficiency to measure to
relative proficiency of confederates. The training of these
raters would include differentiating among different aspects
of language proficiency. Therefore. their ratings of gram-

matical proficiency. for example. would be less likely to be

35

influenced by the confederate’s fluency. If a study in-
volves participants interacting with a NNS confederate,
confederates of different levels of proficiency could be
chosen according to the results of a standardized test.

gggeeg. In spite of the problems in the ratings of
proficiency by participants. significant differences were
found by the coders. The coders counted the number of
grammatical errors. the number of pauses. and the total
length of the pauses. There was very high reliability
between the two coders. They found significant differences
between the number of grammatical errors in the cells with
and without interference errors and significant differences
in the number and total length of pauses in in the cells
with natural pauses and exaggerated pauses. This indicates
that there were indeed differences among the four cells.
even if these differences apparently did not strongly in-
fluence the participants’ ratings of the NNS’s proficiency.
3.1mm

One problem with this study was low reliabilities among
the items of the scales. Alpha for the FT/FR scale was .48.
In writing the FT/FR scale. I attempted to make use of
various characteristics of FT and FR. including avoidance of
slang and idioms. predominance of or/choice and yes/no ques-
tions rather than wh- questions. and use of coordination in
preference to subordination. Possibly the use of so many

different simplification and clarification strategies

36
reduced the reliability of the scale. Also. it is difficult
to look at FR and NSR in particular outside of the context
of a conversation. since in many cases. it is the percentage
of certain structures rather than the incidence of those
structures that differentiates FR from NSR.

More reliable scales need to be used in further re-
search. If a pencil and paper measure is used in other
studies. it may yield more reliable results if only one
aspect of FT or FR. for example. simplification of grammati-
cal structures. is tapped.

As mentioned above. reliability of the proficiency
scale could be improved by using trained raters rather than
participants to rate the proficiency of the NNS.

Some of the other scales were unreliable as well.
Experience with NNSs was a particularly unreliable scale.
with alpha of .43 with the least reliable item deleted.
Ethnocentrism was also an unreliable scale. with alpha of
only .45. even with the most unreliable item deleted.

For the ethnocentrism scale. as mentioned in Chapter
11. items that seemed related to ethnocentrism in relation
to people from other countries were chosen from a larger
scale. Apparently. these items are not as closely related
as they appeared to be. By replacing two or three more of
the items that are not highly correlated with the others.
the reliability of the scale would be increased. Item 2

(See Appendix A. Item 25). for example. does not have high

37
correlations with other items and could be deleted.

Similarly. the reliability of the prejudice scale ap-
pears that it would be increased if the first two items of
the American scale and the first two of the Japanese scale
(See Appendix A. Items 35. 36. 40 and 41) were replaced.
since they are less well correlated than the others.

Mejec flypgtheses

The menigulated levele of fluency and grammatical pro-
ficiency had no significant effects on use of FT and FR.
These findings fail to support any of the major hypotheses
of the study. There are several possible explanations for
this.

The manipulation involved four different levels of
proficiency on the part of the NNS--no interference errors.
natural pauses; interference errors. natural pauses: no
interference errors. exaggerated pauses: and interference
errors. exaggerated pauses. However. in all cases. the NNS
was able to understand the questions asked by the NS and was
able to respond appropriately to them. One reason for the
lack of differentiation in the responses to the different
conditions may be due to the fact that comprehension did not
vary much from condition to condition. Speakers may adjust
their language when they realize that they are speaking to a
NNS in comparison to speaking to 3 NS. However. they may
not differentiate among levels of proficiency in NNSs unless

there are problems in communication. i.e., unless the NNS

38
does not understand what the NS says and/or has difficulty
responding to it appropriately. In fact. other research had
used misunderstanding on the part of the NNS to elicit FT or
FR (Gass and Varonis. 1985).

Another possible explanation is related to the salient
characteristics on which NS base their judgment about the
NNSs proficiency. It is possible that NSs do differentiate
among levels of proficiency of even NNSs who do not show
evidence of lack of comprehension. However. perhaps gram-
matical proficiency and fluency are not the characteristics
of the NNS’s speech that the NS uses to estimate the NNS’s
proficiency. Pronunciation is another variable that has
been proposed to trigger FT/FR. and it may be pronunciation
that NSs use to differentiate among NNSs who don’t demon-
strate difficulty with comprehension.

Another explanation has to do with the methodology of
this study. In Valdman’s (1981) terms. this was an ideolo-
gical study. one where participants report what they would
say in a given situation. Since participants were not
actually interacting with the NNS. they were not constantly
reminded of the NNS’s proficiency. In an interactional
study. participants have constant reminders of the pro-
ficiency of the NNS.

Uhen retigge for fluency and grammatical proficiency
were used as independent variables. no significant effects

were found using the FT/FR scale as the dependent variable.

39

Uhen individual items were used as dependent variables.
there was one significant interaction effect. This in-
dicates a linear interaction between ratings of grammatical
proficiency and of fluency. Since so many different
analyses were done. it would be expected that at least one
would be significant. and it is therefore difficult to draw
any definite conclusions.

In future research. lack of comprehension on the part
of the NNS should be used as a variable. This would help
determine whether the above explanation--NSs adjust their
language when NNSs demonstrate lack of comprehension. but
they do not differentiate among levels of proficiency on the
part of the NNS--is accurate. Also. other NNS variables.
such as pronunciation. should be included to see whether
they have more salience to NSs in their adjustment of the
language they used in speaking to NNSs.

Also. controlled interactional studies should be done
to further investigate the findings of this study. Uhile
ideological studies are useful. they do involve what
participants report that they would say rather than what
they actually say. By definition. an interactional study
would have greater external validity.
ggeeeletiggel Anelysee

Several variables were related to the use of FR and FT.
Uhile these relationships are not strong. the partial cor-

relations between the FT/FR scale and ethnocentrism.

39

Uhen individual items were used as dependent variables.
there was one significant interaction effect. This in-
dicates a linear interaction between ratings of grammatical
proficiency and of fluency. Since so many different
analyses were done. it would be expected that at least one
would be significant. and it is therefore difficult to draw
any definite conclusions.

In future research. lack of comprehension on the part
of the NNS should be used as a variable. This would help
determine whether the above explanation--NSs adjust their
language when NNSs demonstrate lack of comprehension. but
they do not differentiate among levels of proficiency on the
part of the NNS--is accurate. Also. other NNS variables.
such as pronunciation. should be included to see whether
they have more salience to NSs in their adjustment of the
language they used in speaking to NNSs.

Also. controlled interactional studies should be done
to further investigate the findings of this study. Uhile
ideological studies are useful. they do involve what
participants report that they would say rather than what
they actually say. By definition. an interactional study
would have greater external validity.

Qgereletignel Anelyeee

Several variables were related to the use of FR and FT.

Uhile these relationships are not strong. the partial cor-

relations between the FT/FR scale and ethnocentrism.

40

experience speaking a foreign language. experience with
NNSs. and prejudice against Japanese people. are all sig-
nificant. As mentioned in Chapter I. most literature on FT
and FR emphasizes the characteristics of FT and FR and the
roles that the proficiency of the NNS and other NNS charac-
teristics play in triggering FT and FR. while some authors
discussed NS characteristics such as familiarity with non—
native speech. prejudice against NNSs. and ethnocentrism
(Valdman. 1981: Ferguson. 1971: Gass and Varonis. 1985:
Hatch. Shapira and Gough. 1978). 1 could not find any re-
searchers who had studied these relationships in any
systematic way and very few who discussed possible explana-
tions for the relationships. Therefore. these results must
be treated as preliminary and the suggested explanations as
speculative.

Pr e ainst a anese. The partial correlation
between the FT/FR scales and the prejudice scale was -.20
(p < .05). while experience with NNSs. experience speaking a
foreign language. ethnocentrism and ratings of proficiency
were held constant. This indicates that the more prejudiced
a person is against Japanese people. as measured by the
prejudice scale. the less likely they are to adjust the
language that they use when they speak to NNSs.

This contradicts the speculations of Valdman (1981) and
Ferguson (1971) about the relationship between prejudice and

FT. Valdman and Ferguson would suggest that participants

41
prejudiced against Japanese people would be more likely to
use FT. However. this study indicates that people who are
prejudiced against Japanese people report that they are less
likely to use FT and. in fact. are less likely to report
that they adjust their speech when speaking to a Japanese
person. Apparently. NSs who are prejudiced against NNSs do
not. as Valdman speculated. attempt to prevent NNSs from
hearing well-formed language. Instead. it appears that NSs
who are prejudiced against Japanese people are less likely
to adjust their language at all.

Since the purpose of adjustment for NNSs is to improve
communication. it could be that people who are prejudiced
against Japanese people are less concerned about the commu-
nication and whether the communication is successful than
people who are not prejudiced. Therefore. people who are
prejudiced against Japanese may be less likely to adjust
their language.

Ethngcentriem. The partial correlation between the
FT/FR scale and ethnocentrism was .16 (p < .05). with
ratings of proficiency. experience with NNSs. experience
speaking a foreign language. and prejudice against Japanese
people held constant. This indicates that participants who
score high on the ethnocentrism scale are more likely to
adjust their language for a NNS.

Participants who consider their own culture to be bet-

ter than any other appear to be somewhat more inclined to

UE

Pi

42
use FT. or at least less likely to use NSR. than partici-
pants that are more open to other cultures. as measured by
the ethnocentrism scale.

Uhile Valdman (1981) and Ferguson (1971) did not
directly mention ethnocentrism in their discussions of the
sociological aspects of FT. this is the relationship that
they probably would have predicted. People who think that
their culture is better than any other seem to be more
inclined to use FT. thus reducing the amount of well-formed
language. the badge of group membership. that the NNS is
able to hear. and making it more difficult for the NNS to
acquire the language as NSs speak it.

Eggeeience with NNSs. The partial correlation between
experience with NNSs and the FT/FR scale was -.16 (p < .05).
holding ratings of proficiency. ethnocentrism. experience
speaking a foreign language. and prejudice against Japanese
people. constant. Participants who had experience with NNSs
were somewhat less likely to adjust their language.

A possible explanation for that is that N85 with ex-
perience with NNSs may recognize that the NNS’s production
may not reflect receptive ability. i.e., NNSs can generally
understand more complex grammatical structures than they can
produce.

a f re n lan u . The partial
correlation between experience speaking a foreign language

and the FT/FR scale was .24 (p < .05). holding prejudice.

43
ethnocentrism. experience with NNSs. and ratings of pro-
ficiency. constant. Participants who had experience speak-
ing a foreign language were more likely to adjust their
language for NNSs.

One possible explanation for this is that people with
experience speaking a foreign language have themselves been
the object of FT or FR and may be more aware of the benefits
to successful communication derived from using these. Hav-
ing had FT or FR directed at them. they may also be more
aware of the strategies that NSs of the foreign language
use. and may therefore apply those strategies when they
themselves speak to a NNS of their native language. Sub-
jects who have had experience may have more awareness of
different registers. since they may hear both language that
is modified--the language that is directed toward them-~and
language that is not modified--the language that NSs of that
language use among themselves. the language that they hear
in the mass media. etc.

on l 'ons

Hzegtheses related to NNS variable . The first two

null hypotheses were accepted. No main effects or interac-

 

tion effects were found when the manipulations of grammati-
cal proficiency and fluency were used as the independent
variables. The third and fourth null hypotheses would also
have to be accepted. Only one interaction was found between

ratings of grammatical proficiency and fluency. Considering

44
the number of analyses performed. this cannot be given much
weight. The analysis did not indicate that there were main
effects. and only one interaction effect was found for one
item of the FT/FR scale.

Hypotheses releted to NS cheracteristics. The fifth
null hypothesis was rejected. A positive correlation was
found between the ethnocentrism of a participant and the
amount of adjustment he/she made for a NNS. This indicates
that ethnocentric NSs are more likely to adjust their lan-
guage when speaking to NNSs than NSs who are not
ethnocentric.

The sixth null hypothesis was accepted. In fact. a
difference in the opposite direction of what was predicted
was found: participants who were prejudiced against Japanese
people reported that they were less likely to make adjust-
ments of their language in speaking to a Japanese than
participants that were not prejudiced. This contradicts
speculation by Valdman (1981) and Ferguson (1971) that NSs
who are prejudiced use FT in order to exclude them from the
community of N85.

The seventh null hypothesis was accepted. Again. the
study found an effect that was the opposite of the one
predicted. Experience with NNSs and use of FT/FR were
negatively correlated. Participants with experience with
NNSs were less likely to adjust their language for NNSs.

The eighth null hypothesis was rejected. The study

44
the number of analyses performed. this cannot be given much
weight. The analysis did not indicate that there were main
effects. and only one interaction effect was found for one
item of the FT/FR scale.

Hyeotheses releted to NS characteristics. The fifth
null hypothesis was rejected. A positive correlation was
found between the ethnocentrism of a participant and the
amount of adjustment he/she made for a NNS. This indicates
that ethnocentric NSs are more likely to adjust their lan-
guage when speaking to NNSs than NSs who are not
ethnocentric.

The sixth null hypothesis was accepted. In fact. a
difference in the opposite direction of what was predicted
was found: participants who were prejudiced against Japanese
people reported that they were less likely to make adjust-
ments of their language in speaking to a Japanese than
participants that were not prejudiced. This contradicts
speculation by Valdman (1981) and Ferguson (1971) that NSs
who are prejudiced use FT in order to exclude them from the
community of NSs.

The seventh null hypothesis was accepted. Again. the
study found an effect that was the opposite of the one
predicted. Experience with NNSs and use of FT/FR were
negatively correlated. Participants with experience with
NNSs were less likely to adjust their language for NNSs.

The eighth null hypothesis was rejected. The study

45

found that participants with experience using a foreign
language were more likely to adjust their language.

.fley findinge; The most important and interesting find-
ings of this study are the importance of the roles that NS
variables play in influencing the use of FT/FR. while these
variables have been. to a great extent. neglected in the
study of FT and FR. they appear to have an influence. per-
haps a stronger influence than the proficiency of the NNS.
Experience with NNSs and prejudice have negative correla~
tions with the use of FT/FR. while ethnocentrism have posi-
tive correlations with use of FT/FR.
Suggestions for Future Research

Generally. studies of FT and FR have been directed at
describing their characteristics and at finding out what NNS
characteristics trigger them. The study failed to reject
the null hypotheses of this study related to the relation-
ships between interference errors. pause behavior and the
choice to use FT. FR. or NSR. However. connections between
use of FT/FR and four NS variables (ethnocentrism. prejudice
against Japanese. experience with NNSs. and experience with
a foreign language). while not extremely strong. were
significant.

In the future. more research on influences on FT and FR
that involves interactional and ideological methods is
needed. This study. done using an ideological method.

should be replicated using an interactional method. Also.

46
future research on FR and FT should include measurement and
analysis of NS characteristics such as prejudice. ethnocen-
trism. experience with NNSs. and experience with a foreign

language.

APPENDICES

APPENDIX A

1.

APPENDIX A

Scenario Questionnaire
Imagine that you were the American talking to Kazuko in
the video that you saw. Remember that this is an infor-
mal conversation with a student that you just met. Uhat
would you say to Kazuko in each of the following situa-
tions? Choose the response that is closest to what you
would say. Choose your answer according to the form of
the response (how it is said) rather than the content of
the response (what is said). If none of the responses
are close to what you would say. mark g and specify
exactly what you would say in the conversation. in every-
day language. All are legitimate responses and might be
used in conversation. There is no "right" or "wrong"

answer 0

1. Kazuko takes out a pack of cigarettes. lights one. and

sets the pack down on a desk. Assuming that you smoke
and that you would like to borrow a cigarette from her.
how would you request a cigarette?

a. Could I bum a cigarette?

b. Could I borrow a cigarette?

c. Could I borrow cigarette?

d. Other (specify): 1 would say: "

 

47

48
2. You took a course last term that Kazuko is planning on
taking next term. How do you tell her that it is a very
easy course?
a. That’s an easy class.
b. That’s a blow off class.
c. That is easy class.

d. Other (specify): I would say: "

 

3. Due to various problems in your life. you have been
experiencing a lot of tension this term. How do you
describe the tension you have been feeling?

a. I really tense these days.
b. I’m really stressed out these days.
c. I’m really tense these days.

d. Other (specify): I would say: "

 

4. You thought that the midterm in the class you and Kazuko
are taking was very difficult. How do you say that?
a. Uasn’t that test difficult?
b. Uasn’t test difficult?
c. Uasn’t that test a killer?

d. Other (specify): I would say: "

 

49
5. You just realized that you forgot to bring a pen to
class. How do you ask Kazuko if you can borrow a pen?
a. Do you have an extra pen?
b. Borrow a pen?
c. Could I borrow a pen?

d' Other (specify): 1 would say: ”

 

6. You have heard that Professor Smith. one of the profes-
sors that teaches a class Kazuko is taking next term. is
tough. How do you tell her to avoid that section?

a. Don’t take Professor Smith’s class.
b. You don’t take Professor Smith’s class.
c. Stay away from Professor Smith.

d. Other (specify): I would say: "

 

7. Kazuko gets up to put out her cigarette in an ashtray in
the hall. You want to ask her to close the door. what
do you say?

3. Please close door on way back.
b. Uould you mind closing the door on your way back?
c. Please close the door on your way back.

d. Other (specify): I would say: "

 

5.

49
You just realized that you forgot to bring a pen to
class. How do you ask Kazuko if you can borrow a pen?
a. Do you have an extra pen?
b. Borrow a pen?
c. Could I borrow a pen?

d’ Other (5P8C1ty): I would say: ”

 

You have heard that Professor Smith. one of the profes-
sors that teaches a class Kazuko is taking next term. is
tough. How do you tell her to avoid that section?

a. Don’t take Professor Smith’s class.

b. You don’t take Professor Smith’s class.

c. Stay away from Professor Smith.

d. Other (specify): I would say: ”

 

Kazuko gets up to put out her cigarette in an ashtray in
the hall. You want to ask her to close the door. Uhat
do you say?

a. Please close door on way back.

b. Uould you mind closing the door on your way back?

c. Please close the door on your way back.

d‘ Other (specify): I would say: "

 

8.

9.

10.

50
The class that you and Kazuko are taking is one mostly
taken by seniors. How do you find out what year Kazuko
is?
a. You senior?
b. Uhat year are you?
c. Are you a senior?

d. Other (specify): I would say: "

 

How do you find out how long Kazuko has been in
Michigan?

a. Have you been in Michigan long?

b. How long have you been in Michigan?

c. You in Michigan long?

d. Other (specify): 1 would say: "

 

How would you ask Kazuko whether she has a job?
a. Are you working or just taking classes?

b. You work or just take class?

c. Are you working while taking classes?

d’ Other (specify): I would say: "

 

II.

11.

12.

13.

14.

III.

15.

51
Rate Kazuko’s language proficiency in the following
areas. For example. if you think that her proficiency
is the same as a native speaker’s. mark “1". If you
think she has no proficiency in that area. mark "7". If
you think that her proficiency is between these two
extremes. mark "2". ”3”. ”4”. "5" or ”6".
fluency (ability to speak smoothly. without unnecessary
Pauses)
native speaker 1 2 3 4 5 6 7 no proficiency
grammatical proficiency
native speaker I 2 3 4 5 6 7 no proficiency
pronunciation
native speaker 1 2 3 4 5 6 7 no proficiency
overall English proficiency

native speaker 1 2 3 4 5 6 7 no proficiency

Mark the answers to the following questions.

Is English your native language?

a. yes

b. no

c. I am equally proficient in English and another lan-

guage/other languages.

52
16. How often do you talk with people who are not native
speakers of English?
a. daily
b. two or more times a week
c. two or more times a month
d. ten or more times a year
e. five or fewer times a year
f. one or fewer times a year
17. How many friends have you had who were not native speak-
ers of English?
a. none
b. one
c. two
d. three
e. four or more
18. How often do you talk with Japanese people (born in
Japan. not in the United States)?
a. daily
b. two or more times a week
c. two or more times a month
d. ten or more times a year
e. five or fewer times a year

f. one or fewer times a year

 

 

 

 

53

19. How many Japanese friends do you have?

20.

21.

22.

a. none

b. one

c. two

d. three

e. four or more

How many foreign languages have you studied. both in
high school and in college?

a. none

b. one

c. two

d. three or more

In total. how long have you studied foreign languages?
a. never

b. less than one academic year

c. one to two academic years

d. two to three academic years

e. three or more academic years

How many times have you spoken to native speakers of the
language(s) that you studied? (If you have never
studied a foreign language. mark "never".)

a. never

b. one to five times

c. six to ten times

d. ten to fifteen times

e. more than fifteen times

 

23.

24.

IV.

SA A
1 2

54
Have you ever visited another country. other than an
English-speaking country?
a. yes
b. no
How much did you use a foreign language while you were
there?
a. 752 of the time or more
b. 50-75% of the time
c. 25-502 of the time
d. 252 of the time or less

e. not at all

If you strongly agree (SA) with the following state-
ments. mark "1“. If you agree (A) with it. mark "2”.
If you are neutral. (N) mark ”3”. If you disagree (0).

mark "4”. If you strongly disagree (SD). mark "5".

N D 80

3 4 5 25. The greater variety of different points
of view a person can sympathize with. the
more weak-willed that person probably is.

N 0 SD

3 4 5 26. without the friendly cooperation of many
other nations. the United States would

not survive for very long.

3 4 5 27. No matter how different another group’s
culture is. it can probably be justified

as readily as ours.

28.

29.

30.

31.

32.

33.

34.

55

America may not be perfect. but the
American Day has brought us about as
close as human beings can get to a per-

fect society.

No reasonable person can avoid feeling
personally enriched when they are with
people who think and act differently from

them.

It is tempting to stick with one’s kind
for the comfort and assurance it brings.

but the cost is self-narrowing.

It is tempting to try to understand dif-
ferent points of view for the interest it

brings. but the cost is self-confusion.

One should always be on guard against the
temptation to think that tolerance of

different views is always a good thing.

Uhen speaking to a foreigner who doesn’t
speak English well. I tend to get im-

patient with him/her.

If a foreigner is trying hard to speak
English. I feel sympathetic toward
him/her.

V.

56
If you strongly agree (SA) with the following statements.
mark "1". If you agree (A) with it. mark “2”. If you
are neutral. (N) mark "3”. If you disagree (D). mark
"4". If you strongly disagree (SD). mark “5”. The
statements refer to Japanese people born in Japan. not

Americans of Japanese descent.

A N 0 SD

2 3 4 5 35. The average Japanese is honest.

A N 0 SO

2 3 4 5 36. Japanese people are. in general. fair in
their dealings with others.

A N O SO

2 3 4 5 37. The Japanese are imitative rather than
creative pe0ple.

A N 0 SD

2 3 4 5 38. Japanese people are unreliable.

A N 0 SO

2 3 4 5 39. In general. the Japanese are underhanded.

A N 0 SD

2 3 4 5 40. The average American is honest.

A N 0 SO

2 3 4 5 41. Americans are. in general. fair in their
dealings with others.

A 0 SD

2 4 5 42. Americans are imitative rather than crea-
tive people.

A N D SD

2 3 4 5 43. Americans are unreliable.

A N 0 SD

2 3 4 5 44. In general. Americans are underhanded.

56
If you strongly agree (SA) with the following statements.
mark "1". If you agree (A) with it. mark “2”. If you
are neutral. (N) mark "3". If you disagree (D). mark
”4”. If you strongly disagree (SD). mark “5”. The
statements refer to Japanese people born in Japan. not

Americans of Japanese descent.

A N 0 SD

2 3 4 5 35. The average Japanese is honest.

A N D SD

2 3 4 5 36. Japanese people are. in general. fair in
their dealings with others.

A N 0 SD

2 3 4 5 37. The Japanese are imitative rather than
creative people.

A N 0 SD

2 3 4 5 38. Japanese people are unreliable.

A N 0 SO

2 3 4 5 39. In general. the Japanese are underhanded.

A N 0 SD

2 3 4 5 40. The average American is honest.

A N O SO

2 3 4 5 41. Americans are. in general. fair in their
dealings with others.

A 0 SD

2 3 4 5 42. Americans are imitative rather than crea-
tive people.

A N D SD

2 3 4 5 43. Americans are unreliable.

A N 0 SO

2 3 4 5 44. In general. Americans are underhanded.

APPENDIX B

Laura:
Kazuko:
Laura:
Kazuko:
Laura:
Kazuko:
Laura:

Kazuko:

Laura:

Kazuko:

Laura:

Kazuko:

APPENDIX 8

Basic Dialogue
Hi. how are you?
Fine. how are you?

Fine. My name is Laura.

My name is Kazuko.
Uhere are you from?
I come from Kyoto. Japan.

I don’t know Kyoto. Is that near Tokyo?

No. it is near Osaka. It is a very old city. It

is about one thousand two hundred years old. and

it was the capital before Tokyo. It is famous for

temples and for handicrafts like weaving and

pottery.
Do you live on campus?

No. I live off campus. I live in an apartment.

It is about two miles from campus. I take the bus

to school every day. but there are not many bus-

ses. It is not very convenient.
what is your major?
My major is business. I want to go home and work

for a bank or a large company after graduation

from MSU.

57

APPENDIX C

 

 

 

 

APPENDIX C

Laura:
Kazuko:
Laura:
Kazuko:
Laura:
Kazuko:
Laura:

Kazuko:

Laura:

Kazuko:

Laura:

Kazuko:

APPENDIX C

Dialogue with Interference Errors
Hi. how are you?
Fine. how are you?
Fine. My name is Laura.
My name is Kazuko.
where are you from?
I come from Kyoto. Japan.
I don’t know Kyoto. Is that near Tokyo?
No. it is near from Osaka. It is very old city.

It is about one thousand two hundred years old.

and it was capital before Tokyo. It is famous for

temple and for much handicraft like weaving and
pottery.
Do you live on campus?

No. I live off campus. I live in apartment. It

is about two miles from campus. I take bus to

school every day. but there are not much bus. It
is not very convenient.

what is your major?

My major is business. I want to go home and work

for bank or large company after graduation of MSU.

58

Laura:
Kazuko:
Laura:
Kazuko:
Laura:
Kazuko:
Laura:

Kazuko:

Laura:

Kazuko:

Laura:

Kazuko:

APPENDIX C

Dialogue with Interference Errors
Hi. how are you?
Fine. how are you?
Fine. My name is Laura.
My name is Kazuko.
where are you from?
I come from Kyoto. Japan.
I don’t know Kyoto. Is that near Tokyo?
No. it is near from Osaka. It is very old city.

It is about one thousand two hundred years old.

and it was capital before Tokyo. It is famous for

temple and for much handicraft like weaving and
pottery.

Do you live on campus?

No. I live off campus. I live in apartment. It

is about two miles from campus. I take bus to

school every day. but there are not much bus. It
is not very convenient.

Uhat is your major?

My major is business. I want to go home and work

for bank or large company after graduation of MSU.

58

LIST OF REFERENCES

LIST OF REFERENCES

Arthur. 8.. Ueiner. R., Culver. M., Lee. Y.J.. and Thomas.
D. (1980). The register of impersonal discourse to
foreigners: Verbal adjustments to a foreign accent. In
D. Larsen-Freeman. (Ed.). Discourse Anelysis in Second
Language Reseerch. (pp. 111-124). Rawley. MA: Newbury
House.

Bredemeier. M.E.. Bernstein. G., and Oxman. U. (1982). Be
Fa Ba Fa and dogmatism/ethnocentrism: A study of atti-
tude change through simulation-gaming. Simulation egg

Bryant. U.H. (1984). Typical errors in English made by
Japanese ESL students. JALT Journel. Q. 1. 1-18.

Butterworth. B. (1975). Hesitation and semantic planning
in speech. Jouggel of Psycholinguistic Research. 5(1).
75-87 0

Corder. S.P. (1975). "Simple codes" and the source of
second language learners initial heuristic hypothesis.

Studies in Second Language Acquisition. ;(1). 1-10.

Ferguson. C.A. (1971). Absence of copula and the notion of
simplicity: A study of normal speech. baby talk.
foreigner talk. and pidgins. In D. Hymes. (Ed.).
Pidginization end Creolizetion of Languages. (pp. 141-
150). Cambridge: Cambridge University Press.

 

Ferguson. C.A. (1975). Toward a characterization of English
foreigner talk. Anthroeological Linguistics. l1. 1-14.

Ferguson. C.A. (1981). "Foreigner talk" as the name of a
simplified register. International Journel of the

Sociolggy of Language. 2Q. 9-18.

Ferguson. C.A. and DeBose. C.E. (1977) Simplified
Registers. Broken Language. and Pidginization. In A.
Valdman. (Ed.). Pidgin end Cgeole Linguistics. (pp. 99-
125). Bloomington: Indiana University Press.

Freed. B.F. (1981). Foreigner talk. baby talk. native
talk. n ernation l Journal f 80 l and an ua e.
2&9 19‘390

Gass. S.M.. and Varonis. E.M. (1985). Variation in native
speaker modification to non-native speakers. Studies in

Secgnd Lenggage Acguisition. 2. 37-58.

59

60

Greenberg. J. and Rosenfield. O. (1979). Uhites’ ethnocen-
trism and their attributions for the behavior of blacks:
A motivational bias. Journal of Personality. ‘fiZ(4).
643-657.

Hammerly. H. (1982). Synthesis in second language
teachingei An introduction to languistic . Blaine. UA:
Second Language Publications.

Hatch. E., Shapira. R. and Gough. J. (1978). Foreigner
talk discourse. ITL Revieegof Applied Linguistics.
£10.! 39-60 0

Henzl. V.M. (1979). Foreign talk in the classroom.
International Reviegfof Applied Lin uistics. 12(2). 159-
167.

Kellermann. K.. Broetzmann. S.. Lim T.S.. and Kitao. K.
(November. 1986). Paper presented at the conference of
Speech Communicaiton Association. Chicago. IL. Under
review at Discourse Processes.

Krashen. 8.0. and Seliger. H. The essential characteristics
of formal instruction in adult second language learning.

TESOL Quarterly. 2. 173-183.

Long. M.H. (1981). Questions in foreigner talk discourse.

Langgage Learning. 31(1). 135-157.
Long. M.H. (1983). Linguistic and conversational adjust-
ments to non-native speakers. Studies in Second

 

Language Acggisition. 5(2). 177-193.

Long. M.H. and Sato. C.J. (1983). Classroom foreigner talk
discourse: Forms and functions of teachers’ questions.
In Seliger. H.U. and Long. M.H., (Eds.). Classrgom
Oriented Research in Second Language Acquisition.
Rawley. MA: Newbury House. pp. 268-285.

McLaughlin. M.L. (1984). Conversation: How talk is
arganizeg. Beverly Hills: Sage.

Meisel. J.M. (1977). Linguistic simplification: A study
of immigrant workers’ speech and foreigner talk.
ll ue e lin uisti ue a li uee e Neuchatel. 1975.

 

 

88-113.
Nakasone. H. (1979). Crags-egltural study on pause
eharaeteriatics in on-ggigg eeeech by USA students and

 

 

 

Japanese studente. Unpublished master’s thesis.
Michigan State University. East Lansing. MI.

61

Ramamurti. Rajam. (1980). Strategies involved in talking
to a foreigner. Penn Revieg of Linguistics. 5. 84-92.

Schinke-Llano. L.A. (1983). Foreigner talk in content
classrooms. In Seliger. H.U. and Long. M.H.. (Eds.).
Classroom. Oriented Research in Second Language

 

A s 'on. Rawley. MA: Newbury House. pp. 146-165.
Schumann. J.H. (1976). Second language acquisition: The
pidginization hypothesis. yorkpapers in Teaehing

 

n lish a Second an ua e. t9. 127-140.

Snow. C.E.. van Eeden. R., and Muysken. P. (1981). The
interactional origins of foreigner talk: Municipal
employees and foreign workers. Integgational Journal of

Sociology and Language. 28. 81-91.

Tarone. E. (1980). Communication strategies. foreigner
talk. and repair in interlanguage. Language Learnieg.
39(2). 417-431.

Valdman. A. (1981). Sociolinguistic aspects of foreigner

talk. International Journal of Sociology and Language.
2&9 41-520

Laura:
Kazuko:
Laura:
Kazuko:
Laura:
Kazuko:
Laura:

Kazuko:

Laura:

Kazuko:

Laura:

Kazuko:

APPENDIX 8

Basic Dialogue
Hi. how are you?
Fine. how are you?
Fine. My name is Laura.
My name is Kazuko.
Uhere are you from?
I come from Kyoto. Japan.

I don’t know Kyoto. Is that near Tokyo?

No. it is near Osaka. It is a very old city. It

is about one thousand two hundred years old, and

it was the capital before Tokyo. It is famous for

temples and for handicrafts like weaving and

pottery.
Do you live on campus?

No. I live off campus. I live in an apartment.

It is about two miles from campus. I take the bus

to school every day. but there are not many bus-

ses. It is not very convenient.
Uhat is your major?
My major is business. I want to go home and work

for a bank or a large company after graduation

from MSU.

57

APPENDIX C

Laura:
Kazuko:
Laura:
Kazuko:
Laura:
Kazuko:
Laura:

Kazuko:

Laura:

Kazuko:

Laura:

Kazuko:

APPENDIX C

Dialogue with Interference Errors
Hi. how are you?
Fine. how are you?
Fine. My name is Laura.
My name is Kazuko.
where are you from?
I come from Kyoto. Japan.
I don’t know Kyoto. Is that near Tokyo?
No. it is near from Osaka. It is very old city.
It is about one thousand two hundred years old.
and it was capital before Tokyo. It is famous for
temple and for much handicraft like weaving and
pottery.
Do you live on campus?
No. I live off campus. I live in apartment. It
is about two miles from campus. I take bus to
school every day. but there are not much bus. It
is not very convenient.
Uhat is your major?
My major is business. I want to go home and work

for bank or large company after graduation of MSU.

58

L I ST OF REFE RENCES

LIST OF REFERENCES

Arthur. B.. Ueiner. R., Culver. M.. Lee. Y.J.. and Thomas.
D. (1980). The register of impersonal discourse to
foreigners: Verbal adjustments to a foreign accent. In
D. Larsen-Freeman. (Ed.). Discourse Analysis in Second
Langnage Research. (pp. 111-124). Rawley. MA: Newbury
House.

Bredemeier. M.E.. Bernstein. G., and Oxman. U. (1982). Ba
Ea Ba Fa and dogmatism/ethnocentrism: A study of atti-
tude change through simulation-gaming. Simulation and

fiamee. i§(4). 413-436.

Bryant. U.H. (1984). Typical errors in English made by
Japanese ESL students. JALT Journal. Q. 1. 1-18.

Butterworth. B. (1975). Hesitation and semantic planning
in speech. Joucnal of Psycholinguistic Research. 3(1).
75-87 0

Corder. S.P. (1975). "Simple codes" and the source of
second language learners initial heuristic hypothesis.

Studies in Second Language Acguisition. i(1). 1-10.

Ferguson. C.A. (1971). Absence of copula and the notion of
simplicity: A study of normal speech. baby talk.
foreigner talk. and pidgins. In D. Hymes. (Ed.).
Eidginization and Creolization of Languages. (pp. 141-
150). Cambridge: Cambridge University Press.

 

Ferguson. C.A. (1975). Toward a characterization of English
foreigner talk. Anthropolo 1 al Lin uistics. 11. 1-14.

Ferguson. C.A. (1981). ”Foreigner talk” as the name of a
simplified register. international Journal of the

Sncinlngy nf Langeage. 28. 9-18.

Ferguson. C.A. and DeBose. C.E. (1977) Simplified
Registers. Broken Language. and Pidginization. In A.

Valdman. (Ed.). Pidgin and Creole Linguistics. (pp. 99-

125). Bloomington: Indiana University Press.

Freed. B.F. (1981). Foreigner talk. baby talk. native
talk. n rna i nal o rn l f So iolo an Lan ua e.
2&9 19-390

Gass. S.M.. and Varonis. E.M. (1985). Variation in native
speaker modification to non-native speakers. Studies in

Seennd Language ncgnisitinn. 2. 37-58.

59

60

Greenberg. J. and Rosenfield. D. (1979). Uhites’ ethnocen-
trism and their attributions for the behavior of blacks:
A motivational bias. Journal of Personality. §Z(4).
643-657.

Hammerly. H. (1982). Synthesis in second language
teaching: An introduction to languistic . Blaine. UA:
Second Language Publications.

Hatch. E.. Shapira. R. and Gough. J. (1978). Foreigner
talk discourse. ITL Revienfiof Applied Linguistics.
32:59. 39-60.

Henzl. V.M. (1979). Foreign talk in the classroom.
International Revien of Applied Linguistics. 12(2). 159-
167.

Kellermann. K.. Broetzmann. S.. Lim T.S.. and Kitao. K.
(November. 1986). Paper presented at the conference of
Speech Communicaiton Association. Chicago. IL. Under
review at Discourse Processes.

Krashen. 8.0. and Seliger. H. The essential characteristics
of formal instruction in adult second language learning.

TESOL QuartQPIZg 2. 173—1830

Long. M.H. (1981). Questions in foreigner talk discourse.

Lflflsgflse ngrnins’ 31(1), 135—1570

Long. M.H. (1983). Linguistic and conversational adjust-
ments to non-native speakers. Studies in Secnnd

Lansgasg QESQISitiQng :(2)g 177-1930

Long. M.H. and Sato. C.J. (1983). Classroom foreigner talk
discourse: Forms and functions of teachers’ questions.
In Seliger. H.U. and Long. M.H.. (Eds.). Claesrnnm

Oriented Research in Second Language Acquieitin .
Rawley. MA: Newbury House. pp. 268-285.

 

McLaughlin. M.L. (1984). Conversation: How talk is
nrganized. Beverly Hills: Sage.

Meisel. J.M. (1977). Linguistic simplification: A study
of immigrant workers’ speech and foreigner talk.

Cnllegne de linguistique appliq ee de Ne ch tel. 1975.
88-113.

 

Nakasone. H. (1979). Crnss-enitural study on pause
eharaeterietiefi Lin on-going seeech by USA students and

Jananeee students. Unpublished master’s thesis.
Michigan State University. East Lansing. MI.

 

61

Ramamurti. Rajam. (1980). Strategies involved in talking
to a foreigner. Benn Reyien of Linguistics. 5» 84-92.

Schinke-Llano. L.A. (1983). Foreigner talk in content
classrooms. In Seliger. H.U. and Long. M.H., (Eds.).
Classroom Oriented Reeearch in Second Language
Aegnisition. Rawley. MA: Newbury House. pp. 146-165.

Schumann. J.H. (1976). Second language acquisition: The
pidginization hypothesis. Uorkpaners in Teaching
n l h a Second an ua e. i9. 127-140.

Snow. C.E.. van Eeden. R., and Muysken. P. (1981). The
interactional origins of foreigner talk: Municipal
employees and foreign workers. international Journal of

Sociology and Language. 28. 81-91.

Tarone. E. (1980). Communication strategies. foreigner
talk. and repair in interlanguage. Language Learning.
39(2). 417-431.

Valdman. A. (1981). Sociolinguistic aspects of foreigner

talk. Intennatinnal Journal of Sociology and Language.
_2_§g 41-520

MICHIGAN STATE UNIV. LIBRARIES
1NWWW111111“lllllllllllllllllllllllllllllllllllll
31293106953148