Speech Acts Across Cultures

W
DE
G
Studies on Language Acquisition 11
Editor
Peter Jordens
Mouton de Gruyter
Berlin New York
Susan M. Gass and Joyce Neu
(Editors)
Speech Acts Across Cultures
Challenges to Communication
in a Second Language
Mouton de Gruyter
Berlin New York 1996
Mouton de Gruyter (formerly Mouton, The Hague)
is a Division of Walter de Gruyter & Co., Berlin.
The series Studies on Language Acquisition was formerly
published by Foris Publications, Holland.
@ Printed on acid-free paper which falls within the guidelines
of the ANSI to ensure permanence and durability.
The Library of Congress lists the hardcover edition as follows:
Library of Congress Cataloging-in-Publication Data
Speech acts across cultures ; challenges to communication in a
second language / Susan M. Gass and Joyce Neu (editors).
p. em. - (Studies on language acquisition; 11)
Includes bibliographical references and index.
ISBN 3-11-014082-9 (alk. paper) ISBN 978-3-11-019125-7
1. Second language acquisition. 2. Speech acts (Linguistics)
3. Intercultural communication. I. Gass, Susan M. II. Neu,
Joyce, 1950- III. Series.
PI18.2.S67 1995
303.48'2-dc20 95-40820
CIP
Copyright 1995 by Walter de Gruyter & Co., D-I0785 Berlin
All rights reserved, including those of translation into foreign languages. No part of this book
may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopy, recording or any information storage and retrieval system, without
permission in writing from the publisher. Printed in Germany.
Contents
Susan M. Gass
Introduction .
Part I
Methodological issues
Andrew Cohen
Investigating the production of speech act sets.
Noel Houck - Susan M. Gass
Non-native refusals: A methodological perspective.
1
21
45
Leslie M. Beebe - Martha Clark Cummings
Natural speech act data versus written questionnaire data:
How data collection method affects speech act performance . 65
Part II
Speech acts in a second language
Initiating and maintaining solidarity
Miriam Eistenstein Ebsworth - Jean "W: Bodman -
Mary ~ a r p e n t e r
Cross-cultural realization of greetings in American English. 89
Gayle L. Nelson - Waguida El Bakary - Mahmoud Al Batal
Egyptian and American compliments: Focus on second language
learners. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Michael L. Geis - Linda L. Harlow
Politeness strategies in French and English . . . . . . . . . . . . . 129
VI Contents
Naoko Maeshiba - Naoko Yoshinaga - Gabriele Kasper -
Steven Ross
Transfer and proficiency in interlanguage apologizing . . . . . . . 155
Face-threatening acts
Beth Murphy - Joyce Neu
My grade's too low: The speech act set of complaining. . . . . . . 191
Diana Boxer
Ethnographic interviewing as a research tool in speech act analysis:
The case of complaints". . . . . . . . . . . . . . . . . . . . . . . 217
Myra Goldschmidt
From the addressee's perspective: Imposition in favor-asking. . . . 241
Dale April Koike
Transfer of pragmatic competence and suggestions in Spanish
foreign language learning. . . . . . . . . . . . . . . . . . . . . . 257
Part III
Applications
Richard Schmidt - Akihiko Shimura - Zhigang Wang -
Hy-sook Jeong
Suggestions to buy: Television commercials from the U. S., Japan,
China and Korea. . . . . . . . . . . . . . . . . . . . . . . . . . 285
John L. Graham
Culture, negotiations and international cooperative ventures. . . . 317
Subject index .
Author index .
343
347
Introduction
Susan M. Gass
1. Introduction
This book investigates the notion speech act from a cross-cultural pers-
pective. That is, the starting point for this book is the assumption that
speech acts are realized from culture to culture in different ways and that
these differences may result in communication difficulties that range from
the humorous to the serious.
Early studies in speech acts stem from the field of philosophy (e. g.,
Austin 1962; Grice 1957, 1975; Habermas 1979, 1991 and Searle 1969,
1975, 1979, 1983, 1986, 1991) and have been extended and amplified on
by scholars from a number of different fields (e. g., linguistics - Sadock
1974; anthropology - Hymes 1974; Gumperz 1982; child language -
Ochs - Schiefflin 1979). What these studies have in common is the
assumption that fundamental to human communication is the notion of
a speech act, that is, the performance of a certain act through words (e. g.,
requesting something, refusing, thanking, greeting someone, compliment-
ing, complaining)l. Not only does the linguistic realization of the same
speech act differ, but the force of a speech act might differ. For example,
in some cultures to refuse an offer of something may necessitate much
"hedging" or "beating around the bush" before an actual refusal might
be made. In other cultures, a refusal may not necessitate as much mitiga-
tion. The result may, in some cases, be a misinterpretation of whether or
not an actual refusal has been made, but may also be a misunderstanding
of the intentionality of the refuser. In these latter instances, an individual
may be labelled as "rude", not because of the fact of refusal, but because
of the way the refusal was executed. Olshtain and Cohen cite the follow-
ing example of a misunderstanding due to the realization of the speech
act of apology:
One morning, Mrs. G, a native speaker of English now living in Israel, was
doing her daily shopping at the local supermarket. As she was pushing her
shopping cart she unintentially bumped into Mr. Y, a native Israeli. Her
2 Susan M. Gass
natural reaction was to say "I'm sorry" (in Hebrew). Mr. Y turned to her
and said, "Lady, you could at least apologize." On another occasion the
very same Mr. Y arrived late for a meeting conducted by Mr. W (a native
speaker of English) in English. As he walked into the room he said, "The
bus was late," and sat down. Mr. W, obviously annoyed, muttered to him-
self, "These Israelis, why don't they ever apologize!" (Olshtain - Cohen
1989:53)
In other instances, cross-cultural differences (and cross-gender
differences - cf., Tannen 1982, 1986, 1990) may reflect the degree of
indirectness (cf., Brown - Levinson 1978). For example, when some-
one says "I'm hungry", it often means something more than a mere
statement of fact. It can serve as a suggestion (let's go get something to
eat); or it can serve as an exultation (hurry up and finish so we can eat);
it can serve as a request for information (when will dinner be ready,
I'm hungry). While it may be the case that all languages/cultures have
the means to express a suggestion, an exultation or a request for
information, it is not necessarily the case that a statement of fact such
as "I'm hungry" will serve all of these functions. This book is dedicated
to the empirical study of a variety of speech acts in diverse cultural
settings and to the implications and applications of empirical speech act
data.
In this book we deal with three major areas of Speech Act research: 1)
Methodological Issues, 2) Speech Acts in an L2, and 3) Applications. In
the first section we deal with issues of methodology. As in any field and
clearly in all areas of second language research, issues of methodology are
central to an understanding of the phenomenon in question. A major
question is: to what extent can different methodologies contribute to
differential results? As Tarone - Gass - Cohen (1994: xiii) state: "The
validity of any discipline is predicated on the assumption that the
research methods used to gather data are sufficiently understood and
agreed upon." Speech act research is no exception, as all three chapters
in this section aptly illustrate. Human behavior and human interaction
are complex phenomena and are subject to many intervening variables.
Hence, any attempt to examine data and draw conclusions has to do so
fully aware of the multi-faceted nature of the data.
Introduction 3
2. Section One
Leading off in this section, Cohen considers both theoretical and applied
aspects of speech act research. He notes that a first step in speech act
research is a description of the sociocultural and sociolinguistic abilities
needed to produce a given speech act. A second step is the determination
of the research methodology. A third area that needs to be addressed
(in those instances when the data are from non-native speakers) is the
identification of interlanguage features in the data.
Most relevant to this section is his discussion of research methodology.
Many earlier discussions of various approaches to research methods have
focused on the advantages and/or disadvantages of one method over
another. Cohen argues against this approach and in favor of one that
combines different research methods. He argues that research methods
play different roles in the cycle of generating hypotheses, manipulating
variables, determining the range of speech acts and validation. For
example, ethnographic data are most important in generating initial
hypotheses; they are also useful when dealing with some speech acts,
particularly those that occur naturally in discourse. Ethnographic data
are less useful when investigating speech acts that do not occur frequently
and/or that are so sensitive to sociocultural constraints that the cons-
training variables could not be controlled. Similarly, role-plays, written
tests, verbal report data are all relevant, but all come with their own
baggage of advantages/disadvantages and appropriate and inappropriate
uses.
In addition to his discussion of theoretical issues surrounding the use
of various methods, Cohen considers more practical applications of
various methodologies drawn from his own data based on role plays and
follow-up interviews (using videos of the role play) of apologies,
complaints and requests. He makes the important point that particularly
when dealing with non-native speakers, as he was, it is crucial to separate
a learner's adeptness with the situation from his/her adeptness with the
language. For example, it may be the case in a role-play situation, that the
situation itself is foreign to the subject, making the linguistic production
more "unnatural" than it would be in a situation in which the learner felt
comfortable.
In his chapter, Cohen discusses some of the pros and cons of his
particular methodology. In addition, he focuses on a number of
aspects of the retrospective comments of his subjects. He points out that
through these comments we are able to gain additional information
4 Susan M. Gass
on what learners are doing when confronted with producing speech acts
in a second language. Among the issues discussed are the "din in the
head" phenomenon, self-debate, afterthoughts, formulaic speech,
omission, avoidance and simplification. It is through verbal self-report
data that we are able to learn about the options available to learners and
about the choices they make. However researchers select methodology,
Cohen reminds us that it is through triangulation that a more comprehen-
sive picture is able to emerge.
In Chapter Three, Houck and Gass respond to the fact that a signi-
ficant amount of research into speech acts performed by non-native
speakers uses discourse completion tests as a means of data elicitation.
They point out the well-known limitations of this methodology. Primary
among the limitations is the fact that the format used on discourse
completion tests constrains the type and amount of talk. A second
approach to the investigation of speech acts has been ethnographic in
nature. Within this framework, data are collected in naturally occurring
situations. This methodology, while alleviating certain problems that
have been apparent in discourse completion methodology, brings with it
other problems, namely the difficulty in controlling contextual variables
and the unpredictability of the occurrence of a particular speech act.
In their chapter they consider speech act research from a method-
ological and substantive perspective. In particularly, they focus on the
question of an adequate methodology for eliciting spoken speech acts and
provide detail on the ways in which research results may be dependent on
data collection procedures. The specific area of focus is refusals.
Refusals are a highly complex speech act primarily because they may
involve lengthy negotiations as well as face-saving maneuvers. Because
refusals normally function as second pair parts, they preclude extensive
planning on the part of the refuser. Following the work of Beebe -
Takahashi - Uliss-Weltz (1990), they investigate refusals to 1) invi-
tations, 2) suggestions, 3) offers and 4) requests. They depart from
previous work on speech acts in two important ways: 1) by using video-
taped data and 2) by basing their eliciting instrument on Scarcella's
conceptualization of socio-dramas (1978). Thus, the responses that are
given are not confined by either the printed page (e. g., the amount of
space provided on the page, the number of turns that the respondent is
expected to take) or by the closing response of the initiator of the inter-
action which, in many discourse completion tests, directs the refusal by
"sandwiching" it between a given opening remark and the subsequent
closing comment.
Introduction 5
The data-base consists of English responses by native speakers of
Japanese to 8 situations designed to evoke refusals. The subjects of the
study were given the contextual information surrounding each situation.
Following this introduction, each subject role-played the part with a
native speaker who had been instructed not to give up too easily in cases
in which the non-native speaker initially refused. What resulted were
often lengthy discussions in which each person negotiated his/her way
through to a final resolution.
The analysis of the data focuses on such aspects of the discourse as 1)
semantic tactics (sequencing and range), 2) turn length, 3) quantity and
quality of negotiations needed to effect the refusal or to abandon the
attempt to refuse, 4) amount of elaboration and repetition and 5) non-
verbal elements such as laughter and pausing. Their data reveal the
existence of a richer variety of semantic and pragmatic maneuvers than
has been documented in previous literature. Not only is there a difference
in maneuver types, but the methodology used allows for an analysis of
the discourse structure given the extensive negotiation which takes place.
Their chapter deals with substantive findings as well as the method-
ological implications of the differing results.
In a similar vein to that taken in the Houck and Gass article, Beebe
and Cummings question the use of more traditional speech act elicitation
(Discourse Completion Test) by comparing the results of speech act data
based on two different methodologies. They are quick to point out, as
have the authors of the other two chapters in this section, that each
methodology brings with it advantages and disadvantages in terms of the
actual collection of data and in the analysis of those data. For example,
naturalistic data or "notebook data" are valuable in that none of the
artificial constraints of data collection are present, but the data that
result are unconstrained in terms of the many variables that are known
to affect speech acts, such as status, socioeconomic background, age and
so forth.
In a particularly ingenious design, Beebe and Cummings set out to
directly compare comparable refusal data from Discourse Completion
Tests with those from naturally occurring recorded data. All subjects
were English as a Second Language teachers and were presented with the
same situation. The annual TESOL (Teachers of English to Speakers of
other Languages) conference was about to be held in New York City,
where all of the teachers worked. Because it was anticipated that the
conference turnout would be large, there was a need to get as many
volunteers as possible to help with local arrangements. Each participant
6 Susan M. Gass
was given either a written version of the request or was called on the
telephone.
The analysis of the data centers around word-counts as well as semantic
formulas. What they find is that the amount of talk is far greater in
the spoken refusals than in the written responses. Furthermore, in the
oral data, there is a greater display of elaboration on an excuse (a typical
part of a refusal). In face-to-face interactions, or, as in this case, telephone
interactions, it is sociolinguistically inappropriate to flatly refuse without
offering some sort of excuse. The written data do not display elabora-
tions in the same way as the oral data. Furthermore, the oral data allow
for learners to "negotiate" their way to the end of a refusal, rather than
"packing" their responses into the first turn after the request.
The results are related to Wolfson's "bulge theory". The responses
were less typical of strangers (even though in the case of the telephone
conversations, they were strangers) since both the requester and the re-
fuser shared a common profession and were members of the same pro-
fessional organization.
Interestingly, while the oral data reveal longer, more repetitive, and
more elaborated responses, the content of the semantic formulas used
was surprisingly similar (e. g., excuses, negative ability/willingness,
apologies). Thus, the value of Discourse Completion Tests in speech act
research can be validated. Nonetheless, Beebe and Cummings are quick
to point out that they are not a substitute for naturally occurring data.
3. Section Two
The first chapter in Section Two deals with greetings. Eisenstein
Ebsworth, Bodman and Carpenter point out the importance of greetings,
both in terms of the sociocultural significance as well as their timing in
most language classes. Despite their deceptive simplicity, they are
complex speech acts. Following the discussion in Cohen concerning the
need for multiple measures, Eisenstein Ebsworth, Bodman and Carpenter
collect data on greetings in more than one way. They begin through
observation of greetings by both native and non-native speakers of
English in naturally occurring situation, noting the kinds of greetings that
occurred in these situations. The observational data led to the creation of
a questionnaire to elicit data. As in the Beebe and Cummings chapter,
they find that while the data elicited from the questionnaire were more
Introduction 7
limited, they were similar in many respects to the naturally occurring
data.
The data base for their study is rich both quantitatively and
qualitatively. Their subject pool consisted of 50 native speakers of
American English and 100 non-native speakers of American English. This
latter group represented a wide range of native languages. All subjects
created dialogues for pre-specified greeting situations. The non-native
speakers created dialogues for the same situations in their native
languages. A second type of data came from (videotaped) role-plays of
the same situations. Finally, a subset of the subjects participated in open-
ended interviews following the role-plays.
In their chapter, Eisenstein Ebsworth, Bodman and Carpenter
challenge existing interpretations of greetings (in particular, the lack of
sincerity noted by Searle and others). They categorize and exemplify
greetings by native speakers into various types (greetings on the run,
speedy greetings, long greetings, intimate greetings, all-business greetings,
introductory greetings and re-greetings). Through a comparison of the
data gathered from non-native speakers in English with that of the same
speakers in their native language, many instances of native language
influences can be found. Not only can the native language influences be
noted from word-by-word translations of greetings (e.g., incorrect/
inappropriate use of titles, incorrect word choice, incorrect prosody), but
also from an understanding of the cultural norms or the context of the
greeting. Further, the authors discuss particular greeting types that are
problematic for learners and the resultant feelings and interpretations
that come from the different cultural backgrounds of speakers. The
authors also include in their discussion comments about pedagogical
issues and in particular make a plea for the inclusion of this complex
speech act in teaching materials.
Another speech act that indicates solidarity is that of compliments.
This is the focus of Chapter Six by Nelson, El Bakary and Al Batal.
Compliments vary considerably cross-culturally, not only in the words
chosen, but also in the context and frequency with which they are
offered. In fact, Nelson, El Bakary and Al Batal note that Egyptians are
often uncomfortable and embarrassed by the frequency with which
compliments are given in English. As an example, they note the im-
portance of the concept of the "evil eye" in Arabic, a concept that relates
to envy and potential harm coming to the individual whose person or
property is the object of the compliment. The study reported on in this
chapter considers American and Egyptian compliments focusing on the
8 Susan M. Gass
form of the compliment, the object of the compliment, the gender of the
compliment giver and the frequency with which compliments are given.
The data come from 20 American university students and 20 Egyptian
university students. Each was interviewed and asked to tell the most
recent compliment they had given, the most recent compliment that they
had received and the most recent compliment that they had observed.
The resulting analysis revealed a number of differences including
length of compliments (American compliments are considerably shorter)
and the use of comparatives (Egyptians use more similes and metaphors).
The syntactic patterns are limited in both American and Egyptian
compliments although the patterns are not identical. The analysis also
considers the different attributes that are complimented and the
relationship of the compliment giver in terms of gender to the attributes
they compliment. A discussion of the implications of this study for
classroom practices is included. Because the differences between the com-
pliments of the two cultures studied are slight and because the differences
may cause embarrassment and discomfort, it is crucial that the dif-
ferences be presented to learners.
Chapter Seven, the third chapter in the section on solidarity, by Geis
and Harlow, is concerned with the use of politeness strategies in a second
language. As in many other chapters in this volume, the authors point out
the importance of learning sociocultural aspects of language along with
linguistic ones. Their chapter investigates the pragmatic conditions (with
particular focus on politeness conditions) affecting how requests and
offers are communicated in French and English with a view to formulat-
ing these conditions in such a way as to allow them to be taught explicitly
to learners of French. Geis and Harlow's proposals are based on
experimental determination of how native English and French speakers
accomplish requests and offers and how these are done by learners of
French.
Drawing from previous work by Geis and his colleagues, the authors
note that information exchange in conversation occurs, not at the level of
literal meaning, but at the level of "gist" (which consists primarily of the
illocutionary force of an utterance). The form an utterance communicat-
ing a particular gist will take (i. e., syntax, morphology, etc.) is then deter-
mined by discourse context and by register, style, and politeness features.
The study consists of paired oral interactions - paired interactions
between native French speakers, paired interactions between native
English speakers, and paired interactions between English-speaking
learners of French. The 2 subjects in each experiment had to solve simple
Introduction 9
children's jigsaw puzzles for which each subject controlled pieces of
his/her partner's puzzle, necessitating verbal interaction to secure needed
puzzle parts. The authors were able to elicit natural language use in the
experiments, and argue that the insights gained from this experimental
context are generalizable to natural speech contexts.
The data indicate that native speakers of French and English tend to
frame requests somewhat differently, and that English-speaking learners
of French tend to fall somewhere in between, favoring pragmatic
strategies in their native language.
With a focus on politeness, Geis and Harlow present and defend a
theory of the politeness features for the English language and French
language cultures, show what the syntactic, morphological, and prosodic
consequences of these features are for the two languages and then discuss
how non-native speakers might be taught to communicate requests both
as and when French speakers do.
Yet another means of maintaining solidarity is through apologies. This
is the subject of Chapter Eight by Maeshiba, Yoshinaga, Kasper and
Ross. Non-native speakers have been noted to produce second pair parts
that lack coherence and/or cohesion with preceding first pair parts. While
lack of cohesion manifests itself in the choice of textually inappropriate
utterance structures, failure to establish coherence may affect the propos-
itional content, the illocutionary force, the politeness value of the
responding act, or a combination of the above (cf., Kasper 1984).
The realization patterns of the speech acts studied so far in inter-
language pragmatics, notably requests, apologies, complaints, refusals,
compliments, and expressions of gratitude, have been shown to depend
on such extralinguistic contextual factors as social distance and
dominance, and on factors pertaining to the act itself, for instance the
degree of imposition or offense involved in the act (e. g., Brown -
Levinson 1978/1987, Blum-Kulka - House - Kasper 1989). While all
of these factors will also determine the structure of responding acts, the
properties of the initiating act must be assumed to exert further
constraints on the function and form of the responding act.
Chapter Eight investigates responses to apologies, a speech act that has
been particularly well researched by scholars such as Cohen (Olshtain -
Cohen 1983), Olshtain (1983, 1989) and Holmes (1989). Specifically,
the way chosen apology patterns condition the functional and formal
properties of the responses is examined. The empirical material for this
chapter consists of responses to a Dialog Construction Questionnaire by
two groups of Japanese learners of English (intermediate and advanced)
10 Susan M. Gass
compared to responses by native speakers of English and Japanese. In
addition, information was gathered from the native speaker groups on
the likelihood of an apology being necessary and the likelihood of an
apology being accepted.
Their results suggest that as a function of proficiency, learners are less
likely to rely on their native speaker "guidelines" for transfer. However,
when faced with situations or circumstances with which they have little
experience, the advanced learners tend not to rely on their native
strategies, thereby supporting contentions by Kellerman (1979) regarding
language distance and the use of first language forms and/or functions.
The study of apologies, because they are a remedial verbal action to a
threatening (face or bodily) act serve as an appropriate bridge to the
following section which deals with face-threatening acts.
The first two chapters in this section by Murphy and Neu and by
Boxer deal with complaints. Murphy and Neu have a double purpose to
their study. The first, is to determine how native speakers of English and
Korean learners of English produce complaints and the second is to
understand how native speakers judge the speech act set of complaints.
Data were collected from native speakers of American English and
Korean learners of English through an oral Discourse Completion Task.
Subjects were asked to imagine themselves in a situation in which they
had to complain about a grade to a professor. First, the non-native
speakers of English were tape-recorded giving their response to the
situation. When these data were analyzed, Murphy and Neu discovered
that most of the Korean learners of English were producing criticisms
rather than complaints. To investigate the salience of these different types
or response, twenty-seven native speakers of American English were
asked to listen to a sample of the complaint and of the criticism speech
act produced by the non-native speakers and evaluate the content of the
response. The Americans judged the "criticizer" to be aggressive, dis-
respectful, and lacking credibility. The "complainer" was judged to be
respectful, credible, and not aggressive. This perception, that Korean
learners of English are placing the blame for a bad grade on the shoulders
of their professor, is certain to complicate the academic lives of Korean
non-native speakers of English.
The finding that non-native speakers of English may construct a
speech act so incorrectly that it becomes another speech act intirely may
help us gain a clearer handle on the distinctive features of a speech act.
Boxer, in her chapter, uses ethnographic interviews as a means of
corroborating data elicited through observation. She is concerned with
Introduction 11
indirect complaints as opposed to direct complaints. In her analysis,
Boxer argues that while complaints may be viewed as face-threatening
acts, indirect complaints may at times be a form of solidarity and may
involve rapport-building.
Boxer leads the reader through an open and frank discussion of ethno-
graphic interview techniques and provides specific suggestions as to how
interviews of this sort may result in a productive use of researchers' and
informants' times. Her informants provided remarkably similar percep-
tions of the differences between direct and indirect complaints. However,
gender differences did emerge when dealing with responses to indirect
complaints: men tend to offer advice, women tend to commiserate. In
addition, Boxer focuses on ethnic issues noting that within Jewish
culture, complaining appears to be widespread, particularly indirect com-
plaining.
As in other chapters in this volume, Boxer makes the important point
that accurate descriptions and functions of speech acts are crucial as
a basis for providing information (in the form of pedagogical interven-
tion) to language learners. Boxer aptly shows that complaints are not
necessarily what they seem to be on the surface. They serve an important
social function. It is therefore crucial that non-native speakers learn what
that social function is and how to interpret and respond to indirect com-
plaints.
The next chapter by Goldschmidt investigates the variables that deter-
mine how people ask favors of each other. Goldschmidt further attempts
to ascertain if the asking of favors in American English is a strategy-
dominated speech act that manifests the social structure of relationships.
In particular, she addresses the metalanguage used in favor-asking since
this speech act is often counter-intuitive, functioning either as a request
or as a directive. Asking a favor of someone is a potential imposition
depending crucially on the relationship of the asker and the asked.
Variables such as gender, interpersonal relationships, age, status and
degree of imposition are all important to an understanding of how people
respond to favor-asking.
The data for Goldschmidt's chapter come from a survey administered
to 200 people, varying in age, status (student vs. non-student) and gender.
The survey consisted of five favor situations. Subjects were asked to rate
each of these as to the degree of imposition involved. Her results
suggest that imposition is perceived to be great in instances when family
privacy is compromised and in situations in which a great deal of time
and/or effort may be involved. On the other hand, all participants
12 Susan M. Gass
(whether student or non-student, whether male or female and regardless
of age) similarly perceived the degree of imposition in the various situa-
tions.
A recurring theme in this volume has to do with the need to verify the
form, the function and the constraining variables of speech acts as a pre-
requisite for dealing with them in the classroom. It is not enough to
provide practice on "complaining" in a language classroom if we do not
have appropriate and accurate baseline data on which to base our
descriptions. Goldschmidt continues this line of argument by uncovering
the rules, patterns, and strategies of favor-asking and by arguing that this
is essential before we can teach non-native speakers to perform in the
target culture.
The final chapter in this section on face-threatening acts is by Koike
who investigates the speech act of suggesting by English speakers learning
Spanish. In particular, Koike questions 1) the extent to which suggestions
are understood as suggestions by learners at different levels of
proficiency, 2) the potential for misunderstanding and the resultant
possibility of negative reaction toward the suggester, 3) the types of
responses made to suggestions and 4) the degree of understanding of a
suggestion as a function of proficiency level.
The data for Koike's study come from responses by three groups of
learners of Spanish (native speakers of English) ranging from beginning
to advanced. Each subject was presented with a context and then
watched a videotape of a native speaker making a suggestion. The task
was to 1) respond to the suggestion as if the suggestion was being
addressed toward them, 2) identify the type of speech act and 3) evaluate
the speakers on a variety of personal characteristics. In general, Koike
found that speakers did rely on native language speech act patterns in
interpreting second language speech acts. For example, when the form of
an L2 speech act was similar to the form of the same speech act in the L1,
learners were more likely to understand the speech act. It was also noted
that misunderstandings frequently resulted in negative reactions.
Interestingly, negative elements in the form of the suggestion often yielded
a negative interpretation. Responses were given to suggestions by many
of the students even in those instances in which misinterpretation had
occurred. Even when confronted with a negative linguistic element or
with a misinterpretation, learners did not respond in a negative manner.
The fact that responses were given and the fact that those responses were
not negative leads to speculation that through responses and continued
negotiation learners will eventually work out the intended meaning. This
Introduction 13
study provides evidence, then, of the considerable use of the native
language in interpreting and responding to speech acts in a second
language. This chapter leads into the following section, the first chapter
of which also deals with suggestions, albeit suggestions of a different
sort.
4. Section Three
The last section of this volume deals with applications of speech acts. In
particular, the chapter by Schmidt, Shimura, Wang and Jeong illustrates
the way suggestions are made in different cultures in terms of TV
commercials. Graham similarly takes us into the business world by
considering international cooperative ventures.
Schmidt, Shimura, Wang and Jeong consider TV commercials within
the category of suggestions - that is, suggestions to viewers to buy a
particular product. Their rich data base comes from four countries:
United States, Japan, The People's Republic of China and South Korea,
all of which view the purpose of commercials in a slightly different
way.
They found that suggestions were more frequent in American televi-
sion commercials than in the other countries and that the preferred
linguistic mode of making a suggestion was the imperative. This was the
case more in American commercials than in the commercials of other
countries. Thus, American commercials tend to be more overtly
suggestive than those in the three Asian countries under investigation.
What is interesting is the difference among the three Asian countries.
Japanese and American advertising appear to be the most divergent with
Korean and Chinese commercials somewhere in the middle.
By considering the speech act of suggestion in commercial settings,
Schmidt, Shimura, Wang and Jeong propose that the language of adver-
tising is a result of a number of factors, among them are universal
pragmatic principles, cultural norms, market economy, and arbitrary
conventions established by the advertising industry. Other factors (such
as the use of comparative advertising) are determined by government
regulation. The cross-cultural study of commercials is thus a complex
endeavor. As Schmidt, Shimura, Wang and Jeong point out, it is perhaps
best to view the TV commercial not as containing the pure speech act of
a suggestion, but as being a hybrid of requests and suggestions. This
14 Susan M. Gass
chapter clearly shows that the production of speech acts in some
instances takes the perspective of the hearer into account to an even
greater extent than the perspective of the speaker.
Graham, in the final chapter, looks at cultural differences in business
relationships, considering in particular how cultural differences can cause
serious difficulties between or among participants. Graham's specific
focus is the investigation of differences in cultural styles of business
negotiations. His data base comes from videotaped simulated intra-
cultural negotiations involving business people from thirteen countries.
As in the study by Schmidt, Shimura, Wang and Jeong, Graham notes
differences in the negotiation behavior among the Asian countries in his
study. Differences can be found in the amount of use of "no", "you",
silent periods, interruptions and so forth.
As has been pointed out earlier in this chapter, baseline data are
important before designing and implementing pedagogical programs.
Similar comments are made by Graham although the context is different.
Before being able to design a program for training business people on the
differences in negotiation styles, one must first have accurate and detailed
descriptions of negotiations styles in the respective cultures.
5. Conclusion
Through data from a wide range of languages and through a wide range
of speech acts, this volume has set out to describe the ways in which
speech acts are similar and differ across languages and across cultures.
While this goal has been accomplished, there are a number of contexts in
which this work can be and needs to be extended. In particular, we hope
that the work presented herein will provide the impetus for pedagogical
materials and for training within different contexts.
Notes
1. In this introduction we do not deal with the issue of intentionality and
its relationship to the speech act itself (cf., Apel 1991, DeMulder 1993,
Habermas 1991, Leilich 1993 and Searle 1991). We refer the reader to the
works cited here for further elaboration on the theoretical underpinnings of
this issue.
Introduction 15
References
Apel, Karl-Otto
1991 "Is intentionality more basic than linguistic meaning?", in: Ernest
Lepore - Robert Van Gulick (eds.), 31-55.
Austin, John
1962 How to do things with words. Oxford, England: Calderon Press.
Bauman, Richard - Joel Sherzer (eds.)
1974 Explorations in the ethnography of speaking. New York: Cambridge
University Press.
Beebe, Leslie - Tomoko Takahashi - Robin Uliss-Weltz.
1990 "Pragmatic transfer in ESL refusals", in: Robin Scarcella - Elaine'
Andersen - Stephen Krashen (eds.), 55-73.
Blum-Kulka, Shoshana - Juliane House - Gabriele Kasper (eds.)
1989 Cross-cultural pragmatics: Requests and apologies. Norwood, NJ:
Ablex.
Brown, Penelope - Stephen Levinson
1978 "Universals of language usage: Politeness phenomena", in: Esther
Goody (ed.), 56-324.
Brown, Penelope - Stephen Levinson
1978/87 Politeness: Some universals in language usage. Cambridge:
Cambridge University Press.
Cole, Peter - Jerry Morgan (eds.)
1975 Syntax and Semantics 3: Speech Acts. New York: Academic Press.
Dechert, Hans - Manfred Raupach (eds.)
1989 Transfer in production. Norwood, NJ: Ablex.
De Mulder, Walter
1993 "Intentionality and meaning: A reaction to Leilich's 'intentionality,
speech acts and communicative action"', Pragmatics 3: 171-
180.
Gass, Susan - Larry Selinker (eds.)
1983 Language transfer in language learning. Rowley, MA: Newbury
House.
Goody, Esther (ed.)
1979 Questions and politeness. Cambridge: Cambridge University Press.
Grandy, Richard E. - Richard Warner
1986 Philosophical grounds of rationality. Oxford: Clarendon Press.
Grice, H. Paul
1957 "Meaning", Philosophical Review 66: 377-388.
Grice, H. Paul
1975 "Logic and conversation", in: Peter Cole and Jerry Morgan (eds.),
41-58.
Gumperz, John
1982 Discourse strategies. Cambridge: Cambridge University Press.
Habermas,Jurgen
1979 Communication and the evolution of society. Boston: Beacon
Press.
16 Susan M. Gass
lIabermas, Jurgen
1991 "Comments on John Searle: 'Meaning, communication, and repre-
sentation''', in: Ernest Lepore - Robert Van Gulick (eds.), 17-29.
lIolmes, Janet
1989 "Sex differences and apologies: One aspect of communicative
competence", Applied Linguistics 10: 194-213.
IIymes, Dell
1974 "Ways of speaking", in: Richard Bauman - Joel Sherzer (eds.),
433-451.
Kasper, Gabriele
1984 "Pragmatic comprehension in learner-native speaker discourse",
Language Learning 34: 1-20.
Leilich, Joachim
1993 "Intentionality, speech acts and communicative action: A defense of
J. lIabermas' & K. o. Apel's criticism of Searle", Pragmatics 3:
155-170.
Lepore, Ernest - Robert Van Gulick (eds.)
1991 John Searle and his Critics. Oxford: Basil Blackwell.
Gumperz, John
1982 Language and Social Identity. Cambridge: Cambridge University
Press.
Ochs, Elinor - Bambi Schiefflin (eds.)
1979 Developmental Pragmatics. New York: Academic Press.
Olshtain, Elite
1983 "Sociocultural competence and language transfer: The case of
apology", in: Susan Gass - Larry Selinker (eds.), 232-249.
Olshtain, Elite
1989 "Apologies across languages", in: Shoshana Blum-Kulka - Juliane
lIouse - Gabriele Kasper (eds.), 155-173.
Olshtain, Elite - Andrew Cohen
1983 "Apology: A speech act set", in: Nessa Wolfson - Elliot Judd (eds.),
18-35.
Olshtain, Elite - Andrew Cohen
1989 "Speech act behavior across languages", in: lIans Dechert -
Manfred Raupach (eds), 53-67.
Sadock, Jerrold
1974 Toward a linguistic theory of speech acts. New York: Academic
Press.
Scarcella, Robin
1978 "Socio-drama for social interaction", TESOL Quarterly 12: 41-46.
Searle, John
1969 Speech Acts. Cambridge: Cambridge University Press.
Searle, John
1975 "Indirect speech acts", in: Peter Cole - Jerry Morgan (eds.), 59-82.
Searle, John
1979 Expression and meaning: Studies in the theory of speech acts.
Cambridge: Cambridge University Press.
Introduction 17
Searle, John
1983 Intentionality: An essay in the philosophy of mind. Cambridge:
Cambridge University Press.
Searle, John
1986 "Meaning, communication and representation", In: Richard E.
Grandy - Richard Warner (eds.), 209-226.
Searle, John
1991 "Response: meaning, intentionality, and speech acts", In: Ernest
Lepore - Robert Van Gulick (eds.), 81-102.
Tannen, Deborah
1982 "Ethnic style in male-female conversation", in: John Gumperz (ed.),
217-231.
Tannen, Deborah
1986 That's not what I meant!: How conversational style makes or breaks
relationships. New York: Ballantine Books.
Tannen, Deborah
1990 You just don't understand. New York: Ballantine Books.
Tarone, Elaine - Susan Gass - Andrew Cohen (eds.)
1994 Research methodology in second language acquisition. Hillsdale, NJ:
Lawrence Erlbaum Associates.
Wolfson, Nessa - Elliot Judd (eds.)
1983 Sociolinguistics and language acquisition. Rowley, MA: Newbury.
Part I
Methodological Issues
Investigating the production of speech act sets 1
Andrew Cohen
This chapter will discuss both theoretical and applied issues regarding the
researching of speech acts, drawing in part from a recent research effort
to describe the processes involved in producing speech act utterances
(Cohen - Olshtain 1993). The chapter will end with some illustrative
findings from that study.
1. Theoretical issues
Speech acts have been investigated and described from a variety of
perspectives: Philosophical, social, linguistic and cultural. An effort has
been made to identify universal norms of speech behavior and to dis-
tinguish these from language-specific norms in order to better understand
and evaluate interlanguage behavior.
Given a speech act such as apologizing, requesting, complimenting, or
complaining, the first concern of the researcher is to arrive at the set of
potentially universal realization patterns, anyone of which would be
recognized as the speech act in question, when uttered in the appropriate
context. We have referred to this set of strategies as the speech act set of
the specific speech act (Olshtain - Cohen 1983). In order to arrive at the
speech act set, it is necessary to define the goals of the speech act in
question and to identify performative and semantic prerequisites for the
realization of these goals.
As an example, consider the speech act of requesting. It was necessary
to present a scale of impositives (i. e., degree of imposition; Olshtain -
Blum-Kulka 1984; Blum-Kulka 1989; Weizman 1989), moving from the
most direct to the most indirect request. For apologies, it was necessary
to separate the performative verbs from other semantic preconditions
that could result in acceptable apology realizations, such as an explana-
tion and justification for the offense, an offer of repair, and so forth.
22 Andrew Cohen
Hence, each speech act presents its unique set of preconditions and inter-
actional goals which have to be addressed in the realization patterns that
can act as the materialization of the speech act.
As of the 1980s it became clear that in order to adequately define and
describe such speech act sets, considerable empirical investigation both
within and across languages would be needed. Fortunately, over the last
decade there has been a wide range of empirical studies on speech act
behavior. One of the most comprehensive empirical studies of speech act
behavior, both for its breadth and depth, has been that of the Cross-
Cultural Speech Act Research Project (CCSARP) (Blum-Kulka - House
- Kasper 1989), which compared speech act behavior of native speakers
of a number of different languages with the behavior of learners of those
languages. The CCSARP project also produced useful instruments for
data collection and a coding scheme that has been widely replicated in
other speech act studies. Along with the empirical studies, several
excellent surveys of the research literature have appeared which help to
define and shape the field of investigation with respect to speech act
research (e.g., Wolfson 1989; Kasper - Dahl 1991).
In this first section, let us consider three theoretical areas of concern
with regard to speech act research. The first concerns the description of
the sociocultural and sociolinguistic abilities needed to perform a given
speech act. The second concerns the selection of research method for use
in gathering the speech act data. And in cases where the respondents are
non-natives, there is also a need to deal with the interlanguage features
present in the speech act data.
1.1. Sociocultural and sociolinguistic abilities
What has emerged both from the large-scale empirical studies and from
the comprehensive reviews of the literature is that successful planning
and production of speech act utterances depend on the sociocultural and
sociolinguistic abilities of the speaker. Speakers and hearers are successful
speech act users when they have control over the speech act sets for a
given speech act in the language in which they converse. Such control
calls for the ability to provide both socioculturally and sociolinguistically
appropriate behavior.
Sociocultural ability refers to the respondents' skill at selecting speech
act strategies which are appropriate given (1) the culture involved, (2) the
age and sex of the speakers, (3) their social class and occupations, and (4)
Investigating the production of speech act sets 23
their roles and status in the interaction. For example, in some cultures
(such as in the United States) it may be appropriate for speakers to use a
repair strategy by suggesting to the boss when to reschedule a meeting
that they had missed through their own negligence; however, in other
cultures (such as Israel), a repair strategy might be considered out of place
in that it would most likely be the boss who determines what happens
next. Thus, the sociocultural ability is what determines whether a speech
act set is used and which members of the set are selected for use.
Sociolinguistic ability refers to the respondents' skill at selecting
appropriate linguistic forms to express the particular strategy used to
realize the speech act (e.g., expression of regret in an apology, registra-
tion of a grievance in a complaint, specification of the objective of a
request, or the refusal of an invitation). Sociolinguistic ability is the
speakers' control over the actual language forms used to realize the
speech act (e. g., "sorry" vs. "excuse me", "really sorry" vs. "very
sorry"), as well as their control over register or formality of the utterance
from most intimate to most formal language. For example, when students
are asked to dinner by their professor and they cannot make it, the reply
"No way!" would be a phrase for use with an appropriate semantic for-
mula, namely, refusal. The problem is that sociolinguistically, this phrase
would constitute an inappropriate refusal, unless the students had an
especially close relationship with their professor and the utterance were
made in jest.
1.2. Research methodology relating to speech act description and
strategy selection
The complexity of speech act realization and of strategy selection requires
careful development of research methodology in this area. Rather than
choosing between ethnographic and elicited data methods, the combining
of different approaches to studying the same speech act may best enable
the researcher to reach useful and reliable descriptions of speech act
behavior. An ideal cycle of data collection could be perceived of as
following the different collection techniques presented in Olshtain and
Blum-Kulka (1985).
The researchers would start with the generation of initial hypotheses
based on ethnographic data collection of natural speech. Then they
would continue to simulate speech such as role-plays which can serve to
test the initial hypotheses. From there, they could go to a paper-and-
24 Andrew Cohen
pencil written completion test in order to focus on specific realizations
and manipulate the social and situational variables. If they are concerned
with the effect of the speech act on the listener, they might want to use
acceptability tests in order to validate the range of acceptability within a
speech community. Finally, it is advisable to validate findings by means of
further ethnographic data.
~ ethnographic
acceptability ~ -----.... role-play
written completion .-----------
Each of these data collection techniques has its own merits but it is the
use of more than one that provides us with important triangulation.
Ethnographic observation involves the collecting of naturally
occurring data. This method has proven effective in collecting data on
certain speech acts, such as compliments (see, for example, Wolfson
1989). Yet for other speech acts, such as apologizing, it may be extremely
time-consuming and not very productive. Aguilar Murillo, Aguilar, and
Meditz (1991), for example, found that even when they planted someone
crouched behind a door and videotaped the door being opened and the
person getting whacked by the door, the apology events were limited and
the data themselves not very useful.
When comparing native and non-native apologies, complaints or
other complex speech acts across a variety of situations, it would be
exceedingly time-consuming to gather natural data in all the desired
categories. It would also be virtually impossible to control all the
variables that role-play and written completion tasks can build into their
design - e. g., severity of the offense, familiarity/age/relative status/sex of
interlocutors, and so forth.
If a role-play situation is acted out theatrically (e. g., the respondent
bumps into another person), this would constitute genuine role-play.
Another, more popular format is that of a role-play interview in which the
respondents are requested to respond as they would in the given situation,
without acting it out. These two versions of role-play have been referred
to as semi-ethnographic in that they require the participants to take on
roles that are not always their own (Olshtain - Blum-Kulka 1985). If the
interlocutor involved is not aware that the event has been contrived for the
purposes of collecting data or is aware and agrees to cooperate (e.g., the
actual owner of a store hearing a complaint about merchandise from a
subject in a study), then the situation would be called real-play.
Investigating the production of speech act sets 25
Such role-play would most likely be audio- or video-taped. The taping
itself may introduce problems, depending on how intrusively it is done.
Even if the taping is relatively unobtrusive, it may still make some
respondents uncomfortable, at least for the first few minutes. Such taping
may even engender certain reactive effects; Stubbs (1983: 225) has
suggested that respondents might develop special verbal strategies for
dealing with tape-recorders.
The role-play may consist of a description of the situation, written in
the native or target language and/or read aloud, a prompt by the inter-
locutor (depending on the situation), and then the response. It can also be
specified that the interlocutor is to provide one or more rejoinders, to
turn the role-play into an interactive event.
There are two options for a written completion task. In both cases, a
situation is briefly described in writing, either in the target language or in
the native language. In the first type, that of open-ended elicitations, there
is a written prompt followed by a space for the respondent to provide a
written response. The second option is for what has been referred to as
the Discourse Completion Test (DCT) (Blum-Kulka 1982), whereby the
discourse is structured - part of it left open and part closed, providing
both for the speech act and a rejoinder. In fact, the rejoinder helps to cue
the respondent as to the appropriate nature of the speech act realiza-
tion - i. e., the level of formality, and a description of the roles and
relationships of the interlocutors.
The written approaches save enumerable time in data collection and
have been seen to provide reasonable projective measures of the same
kinds of data collected from oral role-play (Beebe - Cummings, this
volume; Cohen - Olshtain - Rosenstein 1986). In comparing talk over
the telephone to written questionnaire responses, Beebe and Cummings
(this volume) found there was four times as much spoken output as than
written. All the same, the results indicated that discourse completion tests
are an effective means of gathering a large amount of data quickly,
creating an initial classification of semantic formulas, and ascertaining
the structure of speech act(s) under consideration.
A problem with written responses is that certain kinds of information
are not collected this way, such as the prosodic and nonverbal features of
oral interaction. Furthermore, the respondent usually has more time to
respond when doing so in writing than when doing so orally. Also, the
very act of responding in writing as if speaking may inhibit the
respondent, producing a shorter response than would be the case in
speaking. One advantage of the discourse completion test over the open-
26 Andrew Cohen
ended format is that the former indicates the expected length of the
utterances while the latter does not.
Acceptability ratings as another means of testing for control of
speech act behavior involves the obtaining of respondents' judgments as
to how appropriate certain responses are for a given situation (Olshtain
- Blum-Kulka 1985). In this technique, a series of possible responses
are presented and the respondent has to select the most appropriate
of them for the given situation. Usually the responses are scaled on a
continuum according to some dimension. In the case of apology research,
the scale could go from the least intensified to the most intensified
apology.
In addition to the above-mentioned techniques which are useful for the
description of speech act behavior within a group, we can use research
techniques such as verbal report to give us insights regarding the choices
made by individuals in their speech behavior. By now it has been clearly
demonstrated that verbal report is not one measure, but rather encom-
passes a variety of measures, intended to provide mentalistic, data
regarding cognitive processing (Afflerbach - Johnston 1984; Olson -
Duffy - Mack 1984; Faerch - Kasper 1987). Such verbal reports in-
clude data that reflect self-report (learners' descriptions of what they do,
characterized by generalized statements about learning behavior), self-
observation (the inspection of specific, not generalized language behavior
introspectively or retrospectively), self-revelation (think-aloud, stream-of-
consciousness disclosure of thought processes while the information is
being attended to), or some combination of these (Cohen - Hosenfeld
1981; Cohen 1987).
Given the intrusive nature of verbal report techniques, it would be
unreasonable to ask speakers to provide such data while they are engaged
in the communicative act.
2
Yet once the interaction is over, subjects may
not be able to retrospect fully as to the strategy selection that they carried
out a few minutes prior to the intervention. For this reason, in the Cohen
and Olshtain study (1993), subjects were videotaped interacting in role-
play situations and then viewed the videotapes (one or more times) as a
means of jogging their memory as to their thought processes during the
interactions. Some illustrative findings from the study will be presented
later in the chapter.
Investigating the production of speech act sets 27
1.3. The study of speech act interlanguage
In second-language acquisition research, there is a concern for the way in
which learners learn and produce speech acts as part of the sociolinguistic
component of their communicative competence. It has been established in
previous studies that in speech act behavior, as in other language areas,
there is a discrepancy between a learner's receptive and productive
abilities. Thus, in a study done with immigrants in Israel, it was found
that while it might take as long as eight years to acquire native-like recep-
tion of speech acts, one may never truly acquire native-like production
(Olshtain - Blum-Kulka 1985).
When dealing with the production of speech acts the immediate
problem is the evaluation of interlanguage speech act behavior. The
questions that could be asked with regard to interlanguage features
include the following:
a. To what extent have learners acquired the sociocultural and socio-
linguistic abilities needed to realize the particular speech act?
b. To what extent is the learner's speech act behavior similar to or
different from a native speaker's behavior under the same circums-
tances?
c. What compensation strategies do learners use when their language is
inadequate?
d. What is the learners' selection route and decision making process with
respect to strategy preference, modification preference, content limit-
ation, and illocutionary intent?
While the first two questions relate to the evaluation of product, the last
two are concerned with process and require introspection and reflection.
There is a need to probe the actual decision-making and selection process
that learners at different levels of proficiency go through in order to
identify strategies that lead to the successful production of speech acts in
the target language.
2. Research design issues
Now that we have discussed some basic theoretical issues, let us look at
specific issues in research design. For the most part, the discussion will be
28 Andrew Cohen
based on issues that came up in the design and execution of the Cohen
and Olshtain (1993) study of speech act production.
The study sought to describe ways in which non-native speakers plan
and execute speech act utterances, and the relationship between choice
of processing strategies and successful execution of the utterance. The
subjects were fifteen advanced English foreign-language learners, all
Hebrew University undergraduates, eleven of whom were native speakers
of Hebrew and the remaining four advanced non-natives - native
speakers of French, Spanish, Portuguese, and Arabic.
The subjects were given six speech act situations (two apologies, two
complaints, and two requests) in which they were to role play along
with a native speaker. The interactions were videotaped, and after each
set of two situations of the same type, the tape was played back and the
respondents were asked both fixed and probing questions regarding the
factors contributing to the production of their response to that situation.
The subjects were interviewed in three sessions - after the apology, com-
plaint, and request situations respectively - instead of waiting until
after all six speech act situations, in order to obtain a more accurate
retrospective report of behavior. It was feared that the delaying of the
verbal report would reduce the reliability of the protocols, even using the
videotaped behavior as a memory aid.
2.1. The role-play interview as a research tool
The first issue to consider is the use of a semi-oral, role-play interview
(i. e., written situation and then role-play) as a simulation of actual
behavior. The question is whether such an elicitation technique is really
semi-ethnographic, as suggested in Olshtain and Blum-Kulka (1985).
What is the effect of having respondents take on a role they would not
assume in real life? In some instances in the Cohen and Olshtain study
(1993), respondents remarked that a given situation happened to them all
the time. In several cases, the respondents commented that they had
performed that speech act the previous day - e. g., requesting a neighbor
to turn down loud music late at night. In other cases, respondents made
it clear that it never happened to them.
In instances where the respondent had never had to react in such a
situation (e.g., apologizing for keeping a classmate's book two weeks
beyond the agreed date), it could be argued that the instrument forces
unnatural behavior and that if the respondent were not a good actor, the
Investigating the production of speech act sets 29
results might be problematic. The researcher's task would be to
distinguish respondents' language proficiency from their situational
adeptness. In the research study under discussion, the respondents were
not given the choice to opt out of the speech act. If they deflected the
stimulus, the interlocutor pursued the issue. This is not necessarily the
case in the real world, where a person may opt not to apologize, com-
plain, or request something (Bonikowska 1988).
Another effect of the situation might be the degree of planning it
activates in the speaker. In other words, the situation itself may have
properties that stimulate planning more than do other situations, regard-
less of the personal characteristics of the speaker. So, for example, if the
respondents feel that they are in the right, as in a complaint situation,
they may not plan as much as if they feel the need to, say, make amends,
as in an apology situation. This observation was made by several of the
respondents in the Cohen and Olshtain study.
Furthermore, an assumption was made in the Cohen and Olshtain
study that a sampling of three speech acts (apology, complaint, request)
in six situations could give a fair idea of how non-natives prepare and
execute utterances. It is possible that this was too small a sampling of
speech act production behavior through role-play. The study also
revealed that the speech act behavior was conditioned by the nature of
the situation. For example, a student's asking his/her teacher for a lift
home - where the inequality of status was found to play an important
role in the mind of the respondent - usually prompted a style shift, at
least after the interlocutor, playing the role of the teacher, replied,
"What?" in response to the student's initial request.
In addition, whereas an effort was made to select situations that were
cross-culturally appropriate - i. e., that had the same cultural weight in
different cultures (such as a neighbor playing loud music late at night), it
is possible that one or another of the situations could still have been
viewed by a respondent as not constituting an infraction. For example,
the situation of "being half an hour late to meet a friend to study for an
exam" may not be considered a serious offense in Latin America, and, in
fact, one of the respondents was originally from Argentina. In this case,
however, the respondent had lived most of her life in Israel.
Finally, the situations were written in the foreign language, English,
which thus provided clues for how to respond - for those respondents
who picked up on this. From time to time respondents lifted language
forms out of the text which described the situation - language forms
that were only partially or not at all in their productive knowledge. For
30 Andrew Cohen
example, in the "lift" situation, a respondent named Hava noted that she
lifted "my bus has just left" out of the text. Also, whereas she would
simply say "token," she requested a "phone token" in the "token"
situation because that was written in the text. Wassim also indicated
taking the expression "phone token" from the text. In that same situa-
tion, Yaakov said he had used the word "urgent" because the word
appeared in the description of the situation - that he would not have
used it otherwise. Likewise, Shlomit said she also used "urgent" because
"it was included in the situation." Finally, there was an instance of the
respondent's combining his own material with that contained in the text.
So, in the "lift" situation, Yaakov described how he arrived at asking
Debbie, "Can I come by your car?":
First I thought "with your car, with you" and that I would not mention the
car because I didn't know how to indicate hamixonit she/ax 'your car.' I
worried that she would think I wanted to go for a ride with her. "To get a
ride with you" would be an expression I wouldn't know how to use. "Can
I come" are words that I know how to use. After I heard Debbie read "by
car," I said "by your car."
The reverse was also true. There were numerous cases where respondents
did not make use of clues that were in the written descriptions of the
situation. For example, in the situation calling for a request from the
teacher for a "lift" home, there were respondents who disregarded this
clue and had difficulties finding a word in English for this request. The
Semi-Direct Oral Proficiency Interview (SOPI) (Stansfield - Kenyon -
Paiva - Doyle - DIsh - Cowles 1990), for example, gives the prompts
in English L1 rather than in the foreign language being assessed. This way
no clues are given concerning the response.
2.2. The role-play interview: Data collection issues
There are variables operating in the collection of role-play interview data
that can have considerable bearing on the reliability and validity of the
results. We will use the Cohen and Olshtain study (1993) to help illus-
trate these variables, the choices that were made, and the possible or
probable effects of these choices.
In administering the role-play interview in that study, it was decided
that the interlocutor would give the respondents an opportunity to read
the descriptions of two brief role-play situations at a time (two apologies,
two complaints, and two requests in all). Then she slowly read each situ-
Investigating the production of speech act sets 31
ation out loud to the respondents, giving them time to think of a
response, and then gave her opener and had the respondent role-play
with her. The interaction was videotaped and audiotaped as well.
The native English-speaking interlocutor determined whether the
interaction had reached its natural and logical end - usually after four
or five exchanges. Thus, in this case, flexible structuring of the role-play
was used. In a previous study that had been recorded only on audiotape,
the structure had been fixed - i. e., an opener was followed by a single
response from the respondent (Cohen - Olshtain 1981). It was felt that
this earlier approach put limitations on the depth and breadth of the data
available for analysis purposes. However, a problem that arose in the use
of flexibly structured role-play was that the interlocutor was not neces-
sarily consistent from one respondent to the next. During the pilot sessions
and in one or two instances at the beginning of the data collection, the
role-play interlocutor was perhaps too easy on the respondent. For
example, when the respondent was slow at making the request for a ride
explicit, the interlocutor offered, "Yeah, do you want a lift?" At other
times, the interlocutor was perhaps too tough on the respondents. They
would apologize, for instance, and she would not accept their apology.
Perhaps it could be argued that in one or two of these cases a native
speaker in a natural setting would accept the apology more readily.
The probing interviews conducted in this study were designed to
obtain retrospective self-observational data about the cognitive processes
involved in the production of speech act realizations. The interviewer's
probes were conducted in what was the native language for eleven of the
respondents, and a language of greater proficiency than English for the
other four respondents. Effort was made to have the respondents be
precise, and to give examples where possible. When the respondents were
not sure as to what they did and why, the interviewers played the relevant
portion of the videotaped session a second or even a third time. This
usually helped to jog the respondent's memory. In working with verbal
report, there is always the danger that if the interviewers suggest too
much, the respondents may fabricate inaccurate descriptions of what they
did to produce utterances. Another problem associated with the power of
suggestion in verbal report is that continued mention of a particular
behavior may do more than simply heighten awareness regarding it. Such
mention may actually cause the behavior to take place.
It is possible, for example, that when respondents were asked to
indicate the language in which they were thinking, it may have stimulated
them - especially the trilingual ones - to think in a language in which
32 Andrew Cohen
they had not been thinking. The question is the extent to which an effort
to heighten awareness about behavior that is taking place but is not
attended to, may inadvertently trigger that behavior. The Cohen and
Olshtain (1993) study seemed to reveal the three most common patterns
for language of thought in planning and responding to be "planning in
English and responding in English," "planning in Hebrew and translating
from Hebrew to English in the response," and "planning in Hebrew with
the response in English."
Furthermore, whereas the French, Portuguese, and Arabic speakers
reported that they tended to think in Hebrew rather than in their native
language, they indicated that they thought in their native language in one
or two situations: The French speaker for planning and producing his
request to his teacher for a lift home, the Portuguese speaker for planning
an apology after forgetting to return a book to a classmate and a com-
plaint after a peer's refusal to let her use her notes, and the Arabic speaker
for planning in the same "notes" situation. In the case of the Spanish
speaker whose English was weak, the patterns were reported to be most
complex, involving planning in Hebrew and Spanish simultaneously or in
a staggered fashion, and then translating from Spanish and/or Hebrew to
English. In reporting these cases of language choice, the researcher must
bear in mind the possible intrusive effects of the verbal report method
here.
There is always the danger that if interviewers make leading sug-
gestions in their efforts to elicit verbal report, the respondents may
fabricate inaccurate descriptions of what they did to produce utterances
(Cohen 1991). In addition, there is the possibility that the interviewer
might make false assumptions based on intuitions regarding the verbal
report and might put words in the respondent's mouth, as in the follow-
ing case from Cohen and Olshtain (1993): "I could see you were focusing
on grammar." In this instance, the informant indicated that he was not
doing so. On the plus side, verbal report interviews provide feedback
from respondents regarding aspects of their behavior that would other-
wise be left to the intuitions and speculations of the investigator.
Then there is still the issue as to the relationship between the reported
behavior and actual behavior. The use of immediate retrospection (imme-
diate playback of the tape after two situations) was intended to diminish
the likelihood of the retrospections being fabricated, but the possibility
still exists. No effort was made to investigate the relationship between the
respondents' report of planning and actual evidence of their planning
(e.g., pauses in delivery).
Investigating the production of speech act sets 33
In this study the respondent informants were not trained in giving
verbal report nor in the dynamics of speech act production. They were
thus naive informants - at least at the outset. It is likely that some, if not
many of them, became more aware of the phenomena being investigated
as they progressed from the first response session to the second and the
third. Hence, it is possible that our data were impaired by a lack of
training. Had we taken the measures of speech act production on which
we wanted verbal report and trained the respondents to pay particular
attention to them, perhaps the results would have been more informative.
Of course, there is then the risk that the training itself will implant certain
notions about "appropriate" behavior in the heads of the respondents
such that they no longer behave the way that they would have.
2.3. The use of multiple measures of speech act production
In the field of language assessment, there is a current emphasis on the
multi-method approach. The attitude is that anyone method would not
be assessing the entirety of the behavior in question. Would this also be
true with respect to determining speech act production behavior in oral
communication? The ethnographic approach would be difficult to
employ, unless respondents were somehow to record their production
strategies (e.g., in a journal) just after performing speech acts. Likewise,
a written discourse completion task would be at best a projective measure
of speaking.
3
It would be possible to gather acceptability data, both from
non-native peers and from natives. Such ratings could help to determine
the extent to which the speech act utterances themselves are appropriate
fOf the given situations.
3. Search, retrieval, and selection of language forms
Let us now take a look at some of the data produced by respondents
in the Cohen and Olshtain study, bearing in mind the methodological
problems raised regarding the collection of such data. We will consider
communication strategies and the concerns that one or more speakers
reported in searching for, retrieving, or selecting language forms to use in
their speech act utterances. These examples represent all the instances
34 Andrew Cohen
that were identified in the analysis of the verbal protocols for these fifteen
speakers. Eight of the categories reflect areas that have been much dis-
cussed in the communication strategy literature: din in the head, monitor,
use of formulaic speech, message omission or abandonment, lexical
avoidance or simplification, and approximation. The other four catego-
ries reflect insights gained from the use of verbal report protocols: Self-
debate, afterthoughts, partial delivery of a thought, and delivery of a
different thought.
3.1. Retrieval process - "din in the head"
Ricki noted after completing the first two situations that she had difficulty
in speaking English because of a long period of non-use: "When I start
speaking English after not speaking it for a long time, my vocabulary is
weak and it is hard to retrieve words from memory." Krashen (1985:
40-41) has called attention to the "din in the head" phenomenon where-
by the "din," or sense of having the language available for use, may take
anywhere from one to two hours of good input and may wear off after a
few days. In certain oral elicitation tasks, there may be a warm-up period,
but often this period is not long enough to activate the din in the head.
3.2. Self-debate before selection
In the "lift" situation, Hava debated between "to get a ride" and "to give
a lift," and finally asked whether she "could get a lift." Shalom debated
among "drive," "come," and "go," .and ended up with, "Can I come with
you?" Galit wanted to make a polite request and was uncertain as to
whether she could ask, "Do you have any room in the car?" As she put
it: "It has a lot of meanings and I wasn't sure that it was correct, so I
changed my tactic, and decided she would understand better if I said, 'I
want to drive with you.' I thought of 'lift,' but didn't know how to use
it in a sentence so I left it out." In the same situation, Lily debated
among three expressions, "in the same neighborhood/your same
neighborhood/in your neighborhood." She was translating from Spanish
and felt that the result was not good. Also with regard to the "lift" situ-
ation, Yaakov debated how to address Debbie - "Debbie," "Teacher,"
"Gveret ('lady')," or "Gveret Teacher." He decided to address her the
way he would in a high school class in Israel.
Investigating the production of speech act sets 35
3.3. Afterthoughts
In the "meeting" situation, Ricki used "very" as the intensifier in her
expression of apology, "very sorry," but reported thinking to herself
afterwards that she could have said "terribly sorry." She also used
"stopped" in that situation ("I'm very sorry, but I-I met some friends
and they stopped me and I couldn't go on ... ") and, as she put it, "I knew
it wasn't the correct word but I was already in the middle of things."
Sometimes the afterthoughts respondents have during a given speaking
task can, in fact, cause later communicative failure in that their minds still
engaged in some previous language form while they are being called upon
to perform a new task.
3.4. Awareness of using the monitor
Four of the respondents referred to their use or nonuse of monitoring.
With regard to the "meeting" situation, Lily commented, "I always think
about grammar and so my pace is so slow. I think about how to structure
the sentence correctly, verb tenses and other aspects. E. g., 'I haven't sleep
good' ~ 'I didn't sleep good.' I thought the first form wasn't correct." In
the "music" situation, Lily erroneously said, "you have listened to the
music very loud last night" and noted, "With this confusion, I wondered
whether to continue with the mistake or correct myself. I decided that it
was important to correct myself because if I am aware of an error and it
is possible to correct it, I want to do it." Ricki could also be viewed
as a consistent monitor user. With respect to the "music" situation, she
commented, "I am always thinking about grammar ... When I have
problems like 'not/don't,' I correct them. 'I was yesterday awake -' just
came out that way and I noted that it was not correct."
Hagar on the other hand would be viewed as an underuser of the
monitor. With regard to the same situation, she remarked, "I don't effort
at grammar. I am aware that it is bad. I focus on the idea, the message.
Grammar gets me stuck. I prefer not to know how grarpmatical I sound.
I depend on the listeners to see if they understand me, using facial expres-
sions and letting them complete my sentences for me." Wassim only
thought about grammar extensively in the "notes" situation in which it
was not spontaneous in that he was translating from Arabic. In the
"meeting" and the "book" situations, he reported: "When I first read the
situations, I thought that it would be good to think about my grammar,
36 Andrew Cohen
but I then forgot about it because it was more important for me that
Debbie understand me."
3.5. Use of formulaic speech
In the "lift" situation, Nogah used "I would love to-" in requesting a
ride, which sounded peculiar for the requesting party to use. Nogah
noted that she had heard this expression a lot and that is why it popped
up in her utterance. Although this was the only reported instance of an
unanalyzed phrase appearing in the respondent's data, it is likely that
such formulaic speech occurs with some regularity in the output of non-
natives (Ellis 1985).
3.6. Omission, avoidance, or simplification
There were also examples of respondents not saying what was intended
for lack of the appropriate forms or lack of certainty about them.
3.6.1. Omission
Two cases of omission of an utterance occurred in the data. In the
"meeting" situation, Lily thought of saying that she was late because of
a problem at home, but decided that it would be too difficult for her to
say it in English. Instead she chose to say that she usually comes late. She
also indicated that in general she chooses the easiest utterance - the one
for which she knows the verbs and the sentence structure, and can say it
directly "without having to express it in a round-about way." In the "lift"
situation, Shlomit debated whether she should address her teacher by
name, and then chose instead to say, "Excuse me, are you going home?"
because, as she put it, "it was a bit more formal - in general, when I
address a lecturer in Hebrew, I do it this way."
3.6.2. Abandoning a word or expression
Five instances of breakdown were identified in the data. In the "meeting"
situation, Galit said, "I really don't have any exc-" and stopped there. She
Investigating the production of speech act sets 37
said she got stuck because of the x. In the "book" situation, Shalom
asked, "Anything I can do to comp - something?" He said that he sort
of knew the word "compensate" receptively. In the "music" situation,
Hagar started the utterance, "Can't you just -" and stopped. She felt that
what she was starting to say was inappropriate and did not know how to
convey the correct message in English. In the same situation, Lily
produced, "I want you to - that -" and, in explanation, noted, "I
wanted to say that I didn't want that to happen again but stopped in the
middle because it was too complicated for me." In the "notes" situation,
Nogah wanted to indicate that she always gave her friend class notes if
she wanted them, but did not know how to say it: "I debated between
'often' and 'always' and I couldn't remember it, so I let it go." She simply
said, "When you need things I al - I give you" and made no further
attempt to supply the adverb.
3.6.3. Partial delivery of a thought
Two instances of partial delivery of an utterance were identified.
In the "notes" situation, Hagar was not sure whether she should just
continue requesting the notes or whether she should simply say
that she did not need any favors from her friend and thank her
anyway. She chose to be angry but commented that "anger doesn't
come out well in English." As she put it, "I started and got stuck
because of my English and so I chose a compromise." Her compromise
was to be sarcastic: "Well, you're very kind to me. I mean I gave
you in the past things and it's - uhm - alright, no thank you." In the
same situation, Nogah wanted to use strong language but did not know
how to say it in English in a way that would not sound too exaggerated,
so instead of saying the English equivalent of tov lada'at 'it's good to
know' or ani ezkor 'et ze 'I'll remember this,' she simply said, "I need
them too."
3.6.4. Delivery of a different thought
There were two examples found of a different thought being delivered.
4
In
the "meeting" situation, Hava wanted to indicate that the bus did not
come, but she reported that she did not find the words in English, so
instead she said, "I missed the bus." Galit, in looking for a reason that
38 Andrew Cohen
she needed a ride, said, "My bus is very late," which she saw right away
to be incorrect. As she explained it, "I meant that it wouldn't be leaving
until later in the evening, but grammatically the sentence was OK so I left
it. I let it go because it wasn't so bad - she would understand what I
meant."
3.6.5. Lexical avoidance or simplification
There was one identifiable instance of lexical avoidance and one of
simplification in the data. In the "music" situation, Shlomit wanted to
say that her neighbor's music was "too loud" but avoided the equivalent
English forms by saying, "Your music is - uhm - and I can't sleep with
your music." In the "notes" situation, Yaakov simplified his utterance in
order to execute it, "I really don't like - this." He explained as follows:
"I searched for something else like, "the way you act/your behavior," but
it didn't come to mind when I was answering. I used the easiest way out
at the moment."
3.6.6. Approximation
In five instances the word search ended in an approximation as the
speaker felt or knew the word was incorrect but could not come up with
an alternative. In the "book" situation, Jackie was looking for a word to
indicate repair but did not find it. He said, "I'm shocked, I'm sorry,"
but he was looking for lefatsot 'to compensate' and, in his words,
"had a blackout." Also in the "book" situation, Galit wanted to say the
English equivalent of xomer 'material,' and could not find a word
like "notebook," so she said "stuff": "I didn't find the - stuff." In
the "music" situation, she asked the neighbor to "reduce" the volume.
Her retrospective comment was as follows: "I had my doubts about
the word 'reduce'; it seemed like a literary word to me." When it was
noted that the interlocutor (Debbie) had in fact supplied the phrase
when she said, "I would have turned it down," Galit replied, "I was more
into my own words than into listening to Debbie's." In the same situ-
ation, Jackie wanted to ask that the neighbor "turn it down," and instead
he got stuck with "put it lower." Finally, in the "token" situation, Ricki
said she used "Listen -" as an opener "because I didn't have anything
else to use."
Investigating the production of speech act sets 39
4. Discussion and Conclusions
This chapter has covered both theoretical and applied issues with regard
to researching the production of speech act behavior. The chapter began
by calling attention to the sociocultural and sociolinguistic abilities neces-
sary for the production of speech acts. It was noted that the selection of
the appropriate speech act strategy is conditioned by a host of social,
cultural, situational, and personal factors. Then, the research cycle of
ethnography, role-play research, written completion tests, and accept-
ability checks was presented. It was indicated that each of these data
collection techniques has its own merits, but that it is the use of more
than one that provides us with important triangulation. It was suggested
that in addition to considering the above-mentioned techniques which are
useful for the description of speech act behavior within a group, the
researcher of speech act behavior also needs to better understand the
choices made by individuals and that here is where verbal reports can be
most valuable. Next, there was discussion of research design issues
relating to the role-play interview as a research tool and to specific data
collection issues. Then some findings generated by speech act production
research were presented. These related to the language of thought and to
the search, retrieval, and selection of language forms. Perhaps two of
these areas, namely, the debate before selection and afterthoughts,
warrant extra comment in that they especially provide us with a window
into the speech act production process.
The debate that goes on in the speaker's mind before selection, which
emerged from the verbal report data, seems to suggest that when faced
with role-play situations - and by extension, in the real world as well -
learners have their own individual spectrum of options from which to
choose. Some of these options relate to semantic elements, some to
grammatical features, and others to illocutionary intent. Among these
learner options, some would lead to appropriate responses while others
may lead to inappropriate ones. Learners make decisions based on those
options available to them at the moment, without knowing which may
lead to inappropriate results, while the native speaker makes choices
based only on acceptable realizations. It is only through verbal report
that we are able to tap these kinds of decision-making processes.
With regard to afterthoughts, it became apparent through the verbal
report protocols in the Cohen and Olshtain study that after completing a
speech act situation, learners continued to think about the degree to
which they were successful in their performance in that situation. Often
40 Andrew Cohen
these afterthoughts can lead to selfawareness and may affect future inter-
actions both for better and for worse. Furthermore, the verbal report
process itself can unintentionally trigger learners' awareness as to their
speech act performance. Whereas learners may be mistaken in how they
assess their speech act behavior, this extra awareness may, in fact, direct
them to proper use as well. Clearly more work will need to be done to
better understand the reactive effects of verbal report techniques in
speech production research.
5. Implications for the language learner and the language
classroom
At a time when teachers have been encouraged to give attention in their
instruction to communicative language which includes speech acts, there
is evidence that learners of a language may lack even partial mastery of
these speech acts and that this lack of mastery may cause breakdowns in
communication, much to the consternation of the speaker and hearer
(Wolfson 1989). The role of the researcher can be to determine the degree
of control that learners have over different speech acts through the
multiple measures suggested above - ethnography, role play, written
completion, and acceptability ratings. They can couple with this informa-
tion verbal report data which add insights regarding the cognitive
processes and conscious strategies used to interpret their role in an inter-
action and to produce appropriate speech act utterances.
Ideally, this information could then be used to prepare a course of
instruction that would teach to the gaps in language knowledge, and also
give tips as to strategies that may be useful for producing utterances. At
present there are only a few published studies dealing with the teaching
of speech act behavior, but the findings seem promising. For example, a
study of advanced English as a Foreign Language learners in Israel would
suggest that the fine points of speech act behavior such as (1) types of
intensification and downgrading, (2) subtle differences between speech
act strategy realizations, and (3) consideration of situational features, can
be taught in the foreign-language classroom (Olshtain - Cohen 1990;
1991).
Likewise, a study by Billmyer (1990) found that tutored English as a
Second Language learners produced a greater number of norm-appro-
priate compliments, produced spontaneous compliments (which the un-
Investigating the production of speech act sets 41
tutored group did not), used a more extensive repertoire of semantically
positive adjectives, and deflected many more compliments in their reply
than did untutored learners. Her conclusion was that formal instruction
concerning the social rules of language use given in the classroom can
assist learners in communicating more appropriately with natives outside
of the classroom.
Notes
1. Parts of this chapter are based on Cohen and Olshtain 1993.
2. On the other hand, verbal report data have been collected while informants
have been engaged in the other language skills - reading, writing, and
listening.
3. Robinson (1991) had twelve native Japanese speakers fill out a discourse
completion test of the ability to refuse requests or invitations, and had them
think aloud while they were doing so. She then played back their think aloud
data in an interview to get them to explain retrospectively their thoughts at the
time of completing the task. The verbal report was conducted in the target
language, English, and not in their native language. Although this study had
some innovative methods in it, such as having the respondents write down
their reactions to each situation at the time of the response, the language-
related data were somewhat limited. Respondents had more to say about
personality matters and about reactions to the given situations.
4. Note that this is more than simply omission because an alternate thought is
supplied.
References
Afflerbach, Peter - Peter Johnston
1984 "On the use of verbal reports In reading research", Journal of
Reading Behavior 16: 307-322.
Aguilar Murillo, Evelyn - Hellen Aguilar - Aimee Meditz
1991 "Teaching speech act behavior through video: Apologies", Athens,
OH: Linguistics Department, Ohio University. (Paper presented at
the Ohio TESOL Fall Conference, Ohio University, Athens, OH,
November 8-9, 1991.)
Billmyer, Kristine
1990 "'I really like your lifestyle', ESL learners learning how to compli-
ment", Penn Working Papers in Educational Linguistics 6.2: 31-48.
Blum-Kulka, Shoshana
1982 "Learning to say what you mean in a second language: A study
of the speech act performance of learners of Hebrew as a second
language", Applied Linguistics 3: 29-59.
42 Andrew Cohen
Blum-Kulka, Shoshana
1989 "Playing it safe: The role of conventionality in indirectness", in:
Shoshana Blum-Kulka - Juliana House - Gabriele Kasper (eds.),
37-70.
Blum-Kulka, Shoshana - Juliane House - Gabriele Kasper (eds.)
1989 Cross-cultural pragmatics: Requests and apologies. Norwood, NJ:
Ablex.
Bonikowska, Malgorzata P.
1988 "The choice of opting out", Applied Linguistics 9: 169-181.
Celce Murcia, Marianne (ed.)
1991 An introduction to teaching English as a second or foreign language.
2nd ed. Cambridge, MA: Newbury House/HarperCollins.
Cohen, Andrew D.
1987 "Using verbal reports in research on language learning", in: Claus
Faerch - Gabriele Kasper (eds.), 82-95.
Cohen, Andrew D.
1991 "Feedback on writing: The use of verbal report", Studies in Second
Language Acquisition 13: 133-159.
Cohen, Andrew D. - Carol Hosenfeld
1981 "Some uses of mentalistic data in second-language research",
Language Learning 31: 285-313.
Cohen, Andrew D. - Elite Olshtain
1981 "Developing a measure of sociocultural competence: The case of
apology", Language Learning 31: 113-134.
Cohen, Andrew D. - Elite Olshtain - David S. Rosenstein
1986 "Advanced EFL apologies: What remains to be learned?"
International Journal of the Sociology of Language 62: 51-74.
Cohen, Andrew D. - Elite Olshtain
1993 "The production of speech acts by EFL learners", TESOL Quarterly
27: 33-56.
Cohen, Andrew D. - Elite Olshtain
1994 "Researching the production of L2 speech acts''', in: Elaine E. Tarone
- Susan M. Gass - Andrew D. Cohen (eds.), 143-156.
Faerch, Claus - Gabriele Kasper
1987 "From product to process-introspective methods in second
language research", in: Claus Faerch - Gabriele Kasper (eds.),
5-23.
Faerch, Claus - Gabriele Kasper (eds.)
1987 Introspection in second language research. Clevedon, England:
Multilingual Matters.
Kasper, Gabriele (ed.)
1991 Pragmatics of Japanese as a native and target language. Technical
Report No.3. Honolulu, HI: Second Language Teaching and
Curriculum Center, University of Hawaii.
Kasper, Gabriele - Merete Dahl
1991 "Research methods in interlanguage pragmatics", Studies in Second
Language Acquisition 13: 215-247.
Investigating the production of speech act sets 43
Kieras, David E. - Marcel A. Just (eds.)
1984 New methods in reading comprehension research. Hillsdale, NJ:
Lawrence Erlbaum Associates.
Mac Mathuna, Liam - David Singleton (eds.)
1984 Languages across cultures. Dublin: Irish Association for Applied
Linguistics.
Olshtain, Elite - Andrew D. Cohen
1983 "Apology: A speech act set", in: Nessa Wolfson - Elliot Judd (eds.),
18-35.
Olshtain, Elite - Shoshana Blum-Kulka
1984 "Cross-linguistic speech act studies: Theoretical and empirical
issues", in: Liam Mac Mathuna - David Singleton (eds.), 235-248.
Olshtain, Elite - Shoshana Blum-Kulka
1985 "Crosscultural pragmatics and the testing of communicative
competence", Language Testing 2: 16-30.
Olshtain, Elite - Andrew D. Cohen
1990 "The learning of complex speech act behavior", TESL Canada
Journal 7: 45-65.
Olshtain, Elite - Andrew D. Cohen
1991 "Teaching speech act behavior to nonnative speakers", in: Marianne
Celce Murcia (ed.), 154-165.
Olson, Gary M. - Susan A. Duffy - Robert L. Mack
1984 "Thinking-out-Ioud as a method for studying real-time comprehen-
sion processes", in: David E. Kieras - Marcel A. Just (eds.),
253-286.
Robinson, Mary
1991 "Introspective methodology in interlanguage pragmatics research",
in: Gabriele Kasper (ed.), 29-84.
Stansfield, Charles - Dorry M. Kenyon - Ricardo Paiva - Fatima Doyle -
Ines DIsh - Maria A. Cowles
1990 "The development and validation of the Portuguese speaking test",
Hispania, 73: 641-651.
Stubbs, Michael
1983 Discourse analysis. Oxford: Basil Blackwell.
Tarone, Elaine E. - Susan M. Gass - Andrew D. Cohen (eds.)
1994 Research methodology in second-language acquisition. Hillsdale,
NJ: Lawrence Erlbaum Associates.
Weizman, Elda
1989 "Requestive hints", in: Shoshana Blum-Kulka - Juliane House -
Gabriele Kasper (eds.), 71-95.
Wolfson, Nessa
1989 Perspectives: Sociolinguistics and TESOL. Cambridge: Newbury
House.
Wolfson, Nessa - Elliot Judd (eds.)
1983 Sociolinguistics and language acquisition. Rowley, MA: Newbury
House.
Non-native refusals: A methodological
Noel Houck - Susan M. Gass
1. Introduction
Much research in recent years has been carried out in the area of inter-
language pragmatics. In this chapter we focus on one area of this research
domain, that of refusals, in the hopes of elucidating methodological
issues in non-native speech act research. A number of papers in this area
emphasize the point, with which we concur, that methodological issues
cannot be ignored, for it is not clear to what extent differences in
methodology yield differences in results (see, for example, Cohen this
volume; Cohen - Olshtain 1994).
Wolfson (1981) and Wolfson, Marmor and Jones (1989) have argued
that ethnographic data collection is the most reliable means of learning
about the social and linguistic constraints on a particular speech act. This
methodology allows for observation of naturally occurring speech events
with precise recording about the social setting, location, and the parti-
cipants, thereby providing information about the linguistic and social
constraints on the use of a given speech act (cf., Watson-Gegeo 1988 for
a discussion of methodological issues).1
However, as has been pointed out by a number of researchers (e.g.,
Rintell - Mitchell 1989; Kasper - Dahl 1991), there are limitations.
Not only can contextual variables not be controlled, but also the occur-
rence of a particular speech act cannot be predicted. If one is truly to
understand a given speech act, many occurrences are needed; this, of
course, is difficult when one must rely on instances when a particular
speech act is used by speakers who are unaware of being observed.
2
In the most detailed treatment to date on the issue of methodology in
second language speech act research, Kasper and Dahl (1991) review 39
studies of interlanguage pragmatics. They characterize the methods used
along two dimensions: 1) By the constraints each imposes on the data and
2) By the degree to which production or comprehension is studied. For
our purposes, we focus only on production data, although we bear in
46 Noel Houck - Susan M. Gass
mind Kasper's (1984) caution that many apparent production problems
are a result of a non-native speaker's inadequate comprehension of pre-
vious parts of the discourse.
Focussing on production data, Kasper and Dahl describe two major
data elicitation measures, discourse completion and role-play. Discourse
Completion Tests are written questionnaires consisting of a brief descrip-
tion of a situation followed by dialogue with a blank line where the sub-
ject is to put in what s/he believes to be an appropriate response. The
other major type of production data comes from role-play, both open and
closed. An example of a closed role-play comes from Rintell and Mitchell
(1989), in which subjects were given an oral version of the Discourse
Completion Test. In open role-plays, on the other hand, an entire dialo-
gue is observed and recorded.
1.1. Discourse completion tests
There are advantages and disadvantages to each of these types of data
collection. With regard to the Discourse Completion Test, perhaps the
most widely used methodology in interlanguage pragmatics, the advant-
ages are clear: Large amounts of data can be collected in a relatively short
amount of time. Furthermore, because of the consistency of the situation,
responses can be compared along a number of dimensions (e. g., age,
gender, ethnicity). On the other hand, there is the question of the extent
to which the data collected are a reflection of the sociolinguistic
constraints that operate on the speech act in question. This is similarly
argued by Wolfson, Marmor and Jones (1989), who point out that "short
decontextualized written segments" may not be comparable to what
takes place in actual interaction.
In fact, recent research on the comparability of Discourse Completion
Test data with data collected using other techniques has revealed some
important differences. Rose (1992) has shown that the frequency of
different types of response varies with the instrument. Rose compared
requests elicited by a Discourse Completion Test with responses to a
multiple choice questionnaire, both of which were administered in English
to native speakers of English and in Japanese to native speakers of
Japanese. He reports that while the most frequent response to all situations
on the Discourse Completion Test was conventionally indirect requests,
responses to the multiple choice questionnaire exhibited more contextual
variation, with respondents often choosing to opt out or to hint.
Non-native refusals 47
The richness of naturally occurring refusals cannot be adequately
captured with a formalized structure such as that represented by the
Discourse Completion Test. As we will show below, refusals are very
often filled with multi-turn responses involving negotiation, hedging and
even reversal. This has been further pointed out by Bardovi-Harlig and
Hartford (1990) and Hartford and Bardovi-Harlig (1992). Their work,
based on naturalistic data, focussed on rejections of advice in academic
advising sessions. In their data they found different strategies for refusals/
rejections than had been found in the Beebe, Takahashi, and Uliss-Weltz
(1990) study. For example, in Bardovi-Harlig and Hartford's data, the
subjects exhibited what they called verbal avoidance, evidenced in the
form of postponement ("Can I think about it?"), requests for repetition,
or requests for additional information. Hartford and Bardovi-Harlig
(1992) concluded that the Discourse Completion Test results in limited
data; specifically, their results show a more limited range of semantic for-
mulas, fewer status-preserving strategies, and none of the extended nego-
tiations which occurred in the natural data. Clearly, written responses,
especially those that are "sandwiched" between an opening statement
and a follow-up statement (as in Discourse Completion Tests), do not
allow a speaker to exhibit the strategies found in naturalistic data.
1.2. Role-play
Role-plays have the advantage of providing data in an oral mode rather
than a written mode (although below we will deal with differences in
these two channels). In a closed role-play (e.g., Walters 1980), subjects
are given a situation and are asked how they would respond. In Walters'
study children were asked to make a request to a puppet. These puppets
varied in age, sex, and race. However, as pointed out above, any type
of data that is "closed", in that it does not allow a free range of answers,
will suffer from the possibility of non-symmetry with naturally occurring
data.
Of the common data elicitation methods, open role-plays are the
closest to what we might expect to reflect naturally occurring speech
events. They have the advantage of allowing the researcher to set up
situations in which the occurrence of a particular speech act is likely in
circumstances in which the occurrence can be recorded and/or video-
taped, thus making possible the close analysis of long interaction
sequences of comparable data.
48 Noel Houck - Susan M. Gass
However, they are not problem free. They are cumbersome to
administer and time-consuming in both their administration and
analysis. Furthermore, role-plays are just that, role-plays, so again we
are left with the question of the degree to which they mirror the lingu-
istic behavior of individuals in the particular setting established by the
researcher.
With these caveats in mind, in this study we have adopted open role-
plays in order to study refusals. We have done this fully aware of the dis-
advantages, but also aware of the advantages particularly with regard to
refusals. As Edmondson (1981 cited in Beebe - Cummings 1985) points
out, "some speech acts are the result not of a single utterance but of a
negotiation, a cooperative achievement, or a conversational outcome
between two speakers." It is clear that to see this type of negotiation in
refusals, we must use an open format to elicit data.
2. Background
2.1. Methodological issues
We turn now to a discussion of the literature in which methodology
is the focus. In addition to Kasper and Dahl (1991), who present a survey
on methodological issues, there are two notable papers that present
results based on a comparison of methodologies. One is by Rintell
and Mitchell (1989) and the other by Beebe and Cummings (1985; this
volume).
Rintell and Mitchell used written and oral versions of a Discourse
Completion Test (eliciting apologies and requests) that were given to low
advanced learners of English and to native English speakers. They were
formulated as role-plays with a variety of social roles and situations
represented. Clearly some differences did exist between the two modal-
ities of elicitation. In particular, for the second language speakers, the oral
data were longer than the written data. This difference was not apparent
in the native speaker responses, leading the authors to conclude that it
was not so much the methodology that resulted in different responses,
but rather the way in which the two groups approached the tasks. In
general they found that the "language elicited ... is very similar whether
collected in written or oral form" (1989: 270). They argue that the
Non-native refusals 49
Discourse Completion Test is in actuality a role-play. That is, both the
written and spoken forms provide data that resemble spoken language
rather than written language.
Beebe and Cummings (1985; this volume) study refusals using two
types of data for their analysis: The Discourse Completion Test and tele-
phone requests. The data they collected were only from native speakers
of English. In both the written and the oral tasks, subjects were asked if
they would be willing to help with the local arrangements for the TESOL
convention in New York.
What they found was that in the oral data, there was more elaboration
of the refusal; in the written data, the layout on the page allowed for only
a minimum amount of data to be produced. Elaborations come as a result
of the "requester's" response. If the requester, upon hearing a refusal,
responds "all right, thank you" and then hangs up, there will be no
further need for elaboration. But if, on the other hand, there is silence or
some other attempt to keep the conversation going, the refuser will feel a
need to elaborate so as not to be offensive to the requester. Goffman
(1971) points out that the offending person (in this case the refuser) needs
to receive reassurance from the addressee that his/her offending remark is
not taken as a serious offense. Elaboration is what restores the offender
to his/her proper place in the eyes of the addressee. Beebe and Cummings
point out that the written test biases "the response toward less negotia-
tion, less hedging, less repetition, less elaboration, less variety and
ultimately less talk" (this volume: 71).
Beebe and Cummings analyzed the written and oral data in terms of
the types of responses given, finding that the written data reflect the
content of oral data (e.g., the use of "I'm sorry"; the frequency with
which excuses were offered; the frequency with which willingness or
ability was offered). Where the two modes differed was in what they call
the "psychosocial" domain. That is, when one refuses, one needs to take
a cue from the requester as to how offensive or how important the refusal
is. This will then dictate the degree to which further elaboration, hedging,
or apologizing is necessary.
2.2. Refusals
Refusals are a highly complex speech act primarily because they often
involve lengthy negotiations as well as face-saving maneuvers to accom-
modate the noncompliant nature of the speech act. Because refusals
50 Noel Houck - Susan M. Gass
normally function as second pair parts, they preclude extensive planning
on the part of the refuser.
2.3. Second language refusals
Two studies on second language refusals are relevant for our purposes.
The first is a study by Beebe, Takahashi and Uliss-Weltz (1990) in which
the major concern was the existence of pragmatic transfer. Four groups
of native speakers of Japanese and English (two native speaker controls
and two second language groups) filled out a Discourse Completion Test
involving twelve situations including refusals of requests, refusals of
invitations, refusals of suggestions and refusals of offers. Each situation
involved an initial segment of written speech followed by a blank and
then followed by a rejoinder that forced the subjects to write a refusal
in the preceding blank. In analyzing the results, the authors considered
the order of semantic formulas. Their preliminary conclusions suggest
evidence of pragmatic transfer although they are quick to caution us
regarding the limitations of the data elicitation methods used.
The second study we mention is that of Kinjo (1987), who examined
refusals to invitations and requests in English and Japanese. Data were
collected orally, with subjects responding to a taped invitation or request.
In her analysis, Kinjo considered the degree to which mitigators played a
role and the degree to which directness/indirectness reflected the
stereotypical notions one has of these two cultures. As with the Beebe
et al. study, Kinjo warns that the results that come about as a result of
this modified role-play method may not reflect naturally occurring
speech.
3. The study
In our attempt to investigate interlanguage refusals, we were primarily
concerned with the interaction involved in the refusal itself. Refusals are
played out events, rather than instances characterized by a brief exchange
or single utterance. That is, we begin with the notion that the modified
role-play, a typical means of gathering data, is insufficient to an under-
standing of the complete speech event of refusing.
Non-native refusals 51
3.1. Method
Following the work of Beebe et aI., we investigate refusals to four types
of situations: 1) Suggestions, 2) Offers, 3) Invitations and 4) Requests.
We depart from previous studies of refusals in two ways: 1) We use video-
taped data, and 2) We use full role-play situations rather than modified
role-plays, with the eliciting instrument based on Scarcella's (1978)
conceptualization of socio-drama. This allows participants to carry out
the refusal to its logical conclusion. Thus, the responses that are given are
not confined by either the printed page (e.g., the amount of space
provided on the page, the number of turns that the respondent is expected
to take) or by the closing response of the initiator of the interaction
which, in many Discourse Completion Tests, directs the refusal by "sand-
wiching" it between a given opening remark and the subsequent closing
comment.
Two situations requiring refusals were created for each of the four
refusal types so that in all eight situations existed (see Appendix). The
setting for each was the home of an American host family who asked the
guest to do something undesirable and quite unusual, such as 1) get a
strange haircut, 2) pierce their ears, 3) go skydiving, or 4) give a speech
at church.
3
3.2. Subjects
Our data-base consists of an interaction involving a native speaker of
English who was the person making the request, invitation, suggestion, or
offer and Japanese English as a second language (ESL) students at two
levels of proficiency. The subjects of the study were given the contextual
information surrounding each situation. Following this introduction,
each subject "role-played" the part with a native speaker who had been
instructed not to give up too easily in cases in which the non-native
speaker initially refused. We made certain that each subject understood
the situation before the session began. All sessions were videotaped and
a subset was transcribed. For each of the eight situations, data from four
non-native speaker-native speaker (two lower-level proficiency and two
higher-level proficiency) pairs were gathered. In analyzing the results for
this study, we focussed on a subset of the data collected.
4
In Houck and
Gass (in press) other data are analyzed from the perspective of non-verbal
communication.
52 Noel Houck - Susan M. Gass
4. Findings
Data collected using an open role-play differ from data collected
using a written or tape-recorded elicitation instrument in a number of
significant ways. The most obvious is that a real face-to-face encounter
results in a dynamic interaction. It is one thing to formulate a refusal
on paper; it is quite another to deliver that refusal to a person who
will respond to it. Not once in our data did the refusal interaction
terminate with the subject's initial response.
5
The role-plays resulted
in what were often lengthy interactions in which the participants
negotiated their way to a resolution. During this time, speakers hemmed
and hawed, cut each other off, requested clarification, self corrected,
modified and elaborated their positions, and generally became involved
in negotiating semantic, pragmatic, and social meaning.
6
Thus, our
role-play data differ from other data on refusals both quantitatively and
qualitatively.
4.1. Quantitative analysis
One quantitative consequence of using an open role-play is that the data
consist not of one response, but rather of a series of turns. To obtain a
quantitative measure of the data, we considered not only turn length, but
also the number of turns. Turn length often varied according to level of
English ability (Blum-Kulka - Olshtain 1986). In our data as well, less
proficient subjects had shorter turns. After eliminating back channels,
such as "mm" and "oh", and pause fillers, such as "uh", we found that
subjects with lower English proficiency averaged 3.5 words per turn.
Higher proficiency subjects were much more prolific, averaging 10.7
words per turn.
The total number of turns from the triggering speech act to the end of
the role-play varied from 7 to 18. On the average, subjects required 9.8
turns at talk to reach a resolution.7 As might be expected with real
negotiations, the outcomes differed considerably. Resolution was
achieved when the participants reached agreement, and: 1) The native
speaker accepted the non-native's refusal; 2) The native speaker and non-
native speaker reached a compromise; or 3) The non-native speaker
accepted the native speaker's offer, request, invitation, or suggestion, and
the role play ended with a few final comments or plans.
Non-native refusals 53
4.2. Qualitative analysis: Classifying the data
We also analyzed the refusal sequences, categorizing the responses made
by the non-native speakers. As a starting point, we applied a comprehens-
ive classification system of refusals, developed by Beebe, Takahashi
and Uliss-Weltz (1990) to each non-native speaker response. Several
categories in this system accounted for approximately 2/3 of the re-
sponses.
- Conventional nonperformative refusals (e. g., "I can't" "No")
- Statements of regret (e.g., "I'm sorry")
- Excuses/reasons/ explanations (e.g., "But 1 don't know yoU")8
- Proposals of alternatives (e.g., "Please wait in your car if you want to
meet him")
These responses also predominate in data from other research on
refusals, such as Kinjo (1987) and Bardovi-Harlig and Hartford (1990).
However, our data also contained non-native speaker responses that
did not correspond neatly with Beebe et al.'s classes. Three of these
were linguistic responses: Confirmations, requests for clarification/
information and agreements. An additional category we labelled non-
verbal responses.
4.2.1. Confirmations
Confirmations occurred frequently in the conversations of lower-
proficiency non-native speakers. When a non-native speaker began
groping for words or exhibiting signs of linguistic distress, the native
speaker often leapt in, checking assumptions and elaborating on minimal
utterances. The non-native speaker could then respond with a single
word, indicating that the native speaker was correct. The non-native
speaker was thus able to get away with a minimum of speech, as in (1)
where he is a guest at a weekend homestay. At breakfast, the native
speaker is inviting him to go skydiving with the family that day.
(1) Confirmation (skydiving)
1. NS: Do you like to skydive?
2. NNS: No
3. NS: No?
54 Noel Houck - Susan M. Gass
4. ~ NNS: ((nods)) Yes
5. NS: Why
6. NNS: Vh, I head(eh) headek headache
7. NS: Y, have ... headache?
8. ~ NNS: Headache
9. NS: Oh you have a headache oh no he has a headache
10. ~ NNS: Headache
In line 2 the non-native speaker indicates his lack of interest in skydiving,
and in line 4 he confirms it with a single word ("Yes") and no explan-
ation. It is the native speaker who requests an explanation (line 5) and,
having received one (line 6), repeats the non-native speaker's explanation
(lines 7 and 9), to which the non-native speaker offers little support
(a confirmatory repetition of his excuse in lines 8 and 10). While this
segment can be seen as an instance of negotiated meaning, it also
provides opportunities for the non-native speaker to elaborate on his
excuse or to add an apology. The non-native speaker seems to re-
cognize that a contribution is called for, but limits it to the single-word
repetitions in lines 8 and 10; the native speaker interprets the minimal
responses and adopts an appropriate attitude (line 9, "Oh no he has a
headache").
In this exchange, the native speaker and non-native speaker work out
the non-native speaker's excuse together, with the native speaker asking
questions and reacting to the information provided, while the non-native
speaker provides minimal answers and confirms the native speaker's
restatements.
4.2.2. Request for clarification
On the other hand, some non-native speakers formulated their own
requests for clarification, as in example (2).
(2) Request for information/clarification (skydiving)
1. ~ NNS: What means
2. NS: What is skydiving
In this example the non-native speaker has just been informed that she
will be going skydiving with the host family that day; she requests an
explanation of the term "skydiving" ("what means").
Non-native refusals 55
Thus, non-native speakers in these open role plays often spend some
of their time ostensibly in the negotiation of meaning - with confirma-
tions and requests for information, although it is likely that the clarifica-
tion requests were actually serving the function of verbal avoidance as
discussed by Bardovi-Harlig and Hartford (1990).
4.2.3. J\greennent
A third response type that differed from responses in most previous
studies of refusals is agreement. While Imai (1981), Rubin (1983) and
Beebe et al. mention general or unenthusiastic acceptance as types of re-
fusals, acceptances that evolve from initial refusals are not discussed.
In several of the role plays, when faced with a persistent native
speaker, the non-native speaker abandoned her attempt to refuse and
accepted. In example (3), the non-native speaker's hostess at a weekend
homestay has offered to give the non-native speaker a punk-style haircut
like her children's. This interaction occurs after the non-native speaker
has given two explicit refusals and a reason (only his barber, who is a hair
specialist, can cut his hair):
(3) Agreement (haircut)
1. NNS:
2. NS:
3. NNS:
4. NS:
5. NNS:
6. NS:
7. ~ N N S :
Ummmm ((laugh)) I like this barber
Uhhuh
Yeah
But but you like my children's haircuts, right?
Ummm
So I c'n I can cut your hair and you can feel
comfortable and cool?
Yeah please
In line 7, the non-native speaker abruptly changes his stance and agrees
to let the native speaker cut his hair. When asked afterwards if they
would really have agreed to having their hair cut, approximately half the
non-native speakers polled said that they would, because she was their
hostess.
56 Noel Houck - Susan M. Gass
4.2.4. Nonverbal responses
In addition to those responses in which the non-native speaker used, if
not propositions, at least lexical items from which a reasonable propos-
ition could be inferred, non-native speakers expressed their reactions and
responses through nonverbal signals. We contend that these nonverbal
signs often performed the same functions as turns with recoverable
propositions. In fact, they were often used to confirm a native speaker
statement or to request clarification or information.
Our transcripts contain a number of non-native speaker nonverbal
signals, such as the nod in (4) and raised eyebrows in (5), which clearly
carry intended communicative content. And, indeed, they can function by
themselves as a turn, performing an interactive function.
(4) Nonverbal: Confirmation (skydiving)
The non-native speaker has expressed fear of skydiving.
1. NS: ((to others )) She's afraid
2. NNS: ((nodding)) Mm
3. NS: But you're sure you don't want to go skydiving
4. ~ NNS: ((nods))
In line 1, the native speaker informs others that the non-native speaker is
afraid; and in line 2 the non-native speaker confirms this with an "Mm"
accompanied by nods. In line 3, the native speaker states an implication
of the non-native speaker's fear, i. e., that she does not wish to skydive, as
a request for confirmation. The non-native speaker confirms the native
speaker's understanding with a (nonverbal) nod (line 4). Thus, in line 4
the nod signals confirmation.
In example (5), as in example (4), the native speaker attempts a resta-
tement of the non-native speaker's previous statement, but in this case she
misinterprets the non-native speaker's meaning.
(5) Nonverbal: Request for information/clarification (speech at
church)
As the host family and the non-native speaker prepare to go
to church, the non-native speaker is informed that she has
been requested to give a talk about herself and her life at the
university.
Non-native refusals 57
1. NS:
2. NNS:
3. NS:
4. ~ N N S :
... they want you to give a speech to everybody in
the church. Is that OK?
Urn urn it's it's very short time for (me)
Oh OK you do not want to give a long speech
((raises eyebrows))
In this interaction, the non-native speaker apparently intended the
utterance in line 2 ("it's it's a very short time for (me)") to mean that she
would not have enough time to prepare a speech. However, the native
speaker understands her to be saying that she is willing to give a speech
if it is a short one (line 3). Thus, the native speaker's paraphrase does not
correspond to the non-native speaker's intentions.' The non-native
speaker's turn in line 4 is an opportunity to confirm or disconfirm the
native speaker's interpretation. Her raised eyebrows convey her un-
certainty and function as a question/request for further clarification.
Thus, in face-to-face interactions, non-native speakers may call on a
number of resources in negotiating refusals. And they may employ these
resources to convey different meanings, depending on the context.
5. Discussion and conclusion
To summarize, the use of open role plays illustrates that refusals often
require a number of turns to effect a response. The number of turns
required may reflect the natural need for conversationalists to interact to
solve a problem - e. g., through negotiation and elaboration of meaning.
The negotiation/elaboration may necessitate a greater number of turns
when a non-native speaker is involved than when only native speakers are
conversing. Or, it may indicate the persistence or stubbornness of the
individual native speaker interlocutor and the non-native speaker
respondent. A non-native speaker may also need to tryout more ploys to
resolve disharmony (see Bardovi-Harlig - Hartford 1990).
The use of open role plays has also shown that the performance of acts
such as refusals involves the use of resources not required or even appro-
priate in noninteractional role play. Thus, we identified three acts -
confirmation, request for clarification, and agreement - which have not
been included in most previous classification schemes.
These new classes of acts are special in that, unlike the acts convent-
ionally associated with refusals, they are characteristic of dynamic inter-
58 Noel Houck - Susan M. Gass
action in general. However, their occurrence in stressful negotiations is
especially appropriate and plays a crucial role in the non-native speaker's
negotiation of a response.
Additionally, these three classes are particularly effective because in
addition to their obvious speech act function, individual instances of the
acts can represent a discourse tactic or social maneuver designed to soften
the unpleasantness of a refusal. For instance, a refusal that develops into
a series of non-native speaker confirmations, as in example (1), may
allow the non-native speaker to build up solidarity with the native
speaker in a face-threatening situation.
It is to be further noted that a request for information or clarification
may function as an avoidance tactic. Bardovi-Harlig and Hartford (1990)
have pointed out that requests for information are employed by non-
native speakers as an indirect means of avoiding a refusal, in their case a
refusal of a suggestion. The interaction in example (2) took place after the
meaning of skydiving had been carefully explained and after a previous
role play on skydiving had been acted out in front of this particular non-
native speaker.
Interestingly, another subject, who had observed two skydiving role
plays also requested an explanation of the term "skydiving" during his
own skydiving role play, supporting the contention that these requests for
clarification may reflect the speaker's wish to avoid direct refusal rather
than a real need to negotiate meaning.
Finally, the fact that a change of heart took place and agreement
ultimately occurred in a number of cases represents the ultimate in refusal
alleviation. The fact that agreement occurred in refusal negotiations, all
of which began with a non-native speaker's clear disinclination to
comply, points to the interesting question of which contextual factors
facilitate agreement and which mitigate against it. Under certain condi-
tions, non-native speakers gave up refusing in favor of agreement; under
other conditions, non-native speakers were unmovable. For instance, in
those situations in which the offer or request was not dangerous or poten-
tially painful, or when subjects might be seen as disappointing their
hostess if they refused, they often eventually agreed. However, if a posi-
tive response meant that the subject or the host or host's family might be
put in a dangerous situation, the subject continued to refuse, no matter
how tricky a linguistic feat, and even at the risk of appearing ungracious
and impolite.
Thus, the inclusion of confirmations, requests for clarification and
agreements in a classification system of refusals is indicated as soon as we
Non-native refusals 59
consider a refusal not as a simple response to a static situation but as a
dynamic negotiated achievement.
A third practical methodological implication resulted from our use of
videotapes, which enabled us to capture the use of nonverbal resources to
negotiate meaning. Most importantly, under the appropriate circum-
stances a head movement or a raised eyebrow can clearly perform the
same function as a verbal "yes" or "oh?" (see Houck - Gass in press for
further discussion). And yet these gestures are available for incorporation
into analyses of speech act performance only when the researcher is
dealing with observed interactions. By including nonverbal signals as
intentional speech act moves, we recognize the wide range of resources
available for communicating a message.
Our data reveal the existence of a richer variety of meaningful
resources and maneuvers than has generally been documented in discus-
sions of non-native refusals. The negotiations we have described go far
beyond the notion of a simple response consisting of linguistically analyz-
able units. They involve art interaction not only between what the non-
native speaker wants to say and what her interlocutor wants her to say,
but how to say it - what grammar, gesture and discourse tactics to use to
carry out both her social obligations and her personal wishes effectively
in a particular situation.
Important work has been done collecting data on the selection and
realization of linguistic acts across cultures using written discourse
completion tests and closed role plays. The addition of interactional
data from open role plays can only enrich our understanding of speech
acts.
Appendix
Situations used
1. You are ready to leave the house to go to a party with the children of your host
family - Nathan, age 21, and Jennifer, age 23. They are telling you about
their friends and the things they usually do at parties. The more they talk, the
more you realize that everyone at the party will be using dangerous drugs.
Nathan picks up his car keys and starts for the door. [Invitation]
2. You are at your host family's home. Your host family, the Quentins, has gone
to a neighbor's house to discuss a business matter. They have left you at home
with specific instructions not to let anyone in the house, no matter what they
say. It could be dangerous. About 5 minutes after they leave, the doorbell
60 Noel Houck - Susan M. Gass
rings. It's a woman who says that she is Mr. Quentin's cousin from Detroit.
She is just passing through Lansing and says, "Can I come in and wait?"
[Request]
3. It is Saturday morning at your host family's home. At breakfast the family tells
you that they have made reservations at the airport for all of you to go sky-
diving this morning. The whole family - Mr. and Mrs. Cousins, Meg, and
Tim - are all getting ready to go. They ask you if you have ever gone skydiving
before. When you say no, they say, "Don't worry! It's easy!" [Invitation]
4. It is Sunday morning and you have agreed to attend church services with your
host family, the Jarvises. As you are getting ready to leave the house for
church, Mrs. Jarvis informs you that there are plans for you to give a short
speech about university life in Japan after the services. She says, "I hope you
won't mind." [Request]
5. It is 11:OOam Saturday morning at the home of your host family, the Larsons.
You arrived at the Larsons' home last night at about 8:00pm. You thought
that you would be having dinner with them, but they thought you had eaten,
so you had no dinner. This morning you had only a piece of toast and coffee.
You are now very hungry. Mrs. Larson walks into the room and tells you that
you will be going to an early barbecue for dinner. She suggests that because
you will be eating at about 5:00pm, you skip lunch today. But you are really
hungry. [Suggestion]
6. You are at the home of your host family, the Sumners. Both the children,
Charlie and Karen Sumner, have short, very ugly haircuts. At one point, they
ask you how you like their hair. You answer politely that it looks very cool and
comfortable. Mrs. Sumner announces proudly that she cuts their hair herself.
And because you like the style, she will be glad to cut your hair to look like
theirs. "Now where are my scissors ... ?", she asks. [Offer]
7. You are watching MTV with your host family on Saturday. You notice that
both men and women rock stars have at least 4 earrings in their ears. You
comment that this style is very interesting. Your host family's son Bob, age 22,
says, "Oh, I'm glad you like it. My girlfriend pierced my ears. Why don't you
get yours done, too? I'll call her right now, and she can be here in 20 minutes
to pierce your ears." Bob goes to the telephone to call. [Suggestion]
8. You are at your host family's home. Your host mother, Mrs. Boulware, is
admiring the expensive new pen that your family gave you before you left
Japan. Mrs. Boulware sets the pen down on a low table, and you and she go
into the backyard to look at her flowers. When you return to the room, the
Boulware's pet dog, Ruffy, is happily chewing on your pen. When Mrs.
Boulware gets the pen out of Ruffy's mouth, it is ruined. Mrs. Boulware says,
"Oh, I am so sorry. I'll buy you a new one." [Offer]
Notes
* We are grateful to Joyce Neu for helpful comments on an earlier version of this
chapter. We also thank the panelists and participants of the TESOL
Sociolinguistics Colloquium (1991), where these data were originally present-
ed, for their comments and suggestions.
Non-native refusals 61
1. An alternative to detailed ethnographic data is what Beebe (1993) refers to as
"notebook data". This data type consists of memorizing the core act (e. g.,
refusal) when it occurs, as well as any supporting moves that the researcher
can commit to memory; writing down immediately everything that she can
remember precisely, as well as any partially recalled speech or additional
information that may be relevant to a description of the interaction; and
marking notes ruthlessly to reflect which dialogue was recalled verbatim and
which was reconstructed. Although it is limited to capturing short inter-
actions, the method allows an alert observer to gather large amounts of data
on particuar types of acts in relatively short spans of time.
2. Bardovi-Harlig - Hartford (1992) analyze the advantages and limitations
of naturally occurring data collected in an institutional setting, in which
the interactants and situations are relatively invariant, and in which the
interactions are videotaped (and participants are aware of being observ-
ed).
3. One could argue that it is highly unlikely that a non-native speaker would
encounter situations such as the ones used in this study. However, that appears
not to be the case. The data for this study were collected immediately follow-
ing a home-stay weekend in which these students had visited an American
family. Some of these situations had quite coincidentally been encountered as
had even more bizarre ones, such as a suggestion to go to the morgue to see a
dead body. What is interesting and what will be discussed later in this chapter
is the extent to which the non-native speakers gave in to a request. This
appeared to be in large part determined by the extent to which the guest could
comply with the native speaker's request, offer, suggestion, or invitation with-
out putting herself at risk.
4. For the lower proficiency students we analysed the data from two of the sky-
diving situations, two of the speech at church situations and two of the hair-
cut situations. For the higher proficiency students, we analysed data from two
of the visiting cousin situations.
5. This may, of course, be due to the instructions given to them, but it also may
be a result of the methodology used, which did not allow for a comfortable
closure early on in the interaction.
6. The result is, of course, messy. In our data, the researcher determined the
nature of the initiating speech act, but had no real effect on the remaining
speech; in pure observational research, the researcher, of course, controls even
less (cf., Beebe 1993).
7. In addition to the native speaker's persistence, it might be expected that the
amount of negotiation reflected a japanese reluctance to refuse directly.
However, a look at the non-native speakers' use of the most direct linguistic
refusal "no" indicates that they were often willing to state refusals directly.
(In five of the eight role plays, the japanese subjects indicated refusal at
least once with "No.") This is corroborated by the data from Kinjo, who
found that her japanese subjects were more open and direct than her
American subjects.
The data from this study give a preliminary indication that japanese subjects
will give direct negative responses, at least in some situations. However, a
62 Noel Houck - Susan M. Gass
study of the factors responsible for determining the level of directness was not
within the scope of this project.
8. As Joyce Neu points out, there are other interpretations possible to a phrase,
such as "But I don't know you." For example, it may be a challenge to the
speaker's right to ask the hearer to do X. While this might be the case in inter-
actions between two native speakers, we are confident that in our data, given
the participants, given the intonation and given the body and facial expres-
sions, these are truly expressions of refusal.
References
Bardovi-Harlig, Kathleen. - Beverly S. Hartford
1990 "Learning to say 'no': Native and non-native rejections in English",
Paper presented at the conference on Pragmatics and Language
Learning. University of Illinois, April.
1992 "Natural conversations, institutional talk, and interlanguage prag-
matics", Paper presented at the Pacific Second Language Research
Forum. University of Sydney, July.
Beebe, Leslie
1993 "Rudeness: The neglected side of communicative competence?",
Paper presented at the Fourteenth Annual TESOL and Socio-
linguistics Colloquium. TESOL National Convention, Atlanta,
April.
Beebe, Leslie - Martha C. Cummings
1985 "Speech act performance: A function of the data collection proce-
dure", Paper presented at the Sixth Annual TESOL and Socio-
linguistics Colloquium. TESOL, New York, 1985.
Beebe, Leslie - Tomoko Takahashi - Robin Uliss-Weltz
1990 "Pragmatic transfer in ESL refusals", in: Robin Scarcella -Elaine
Andersen - Stephen Krashen (eds.), 55-73.
Blum-Kulka, Shoshana - ]uliane House - Gabriele Kasper (eds.)
1989 Cross-cultural pragmatics: Requests and apologies. Norwood, N]:
Ablex.
Blum-Kulka, Shoshana - Elite Olshtain
1986 "Too many words: Length of utterance and pragmatic failure",
Studies in Second Language Acquisition 8: 165-180.
Bouton, Lawrence - Yamuna Kachru (eds.)
1992 Pragmatics and language learning 3. Urbana-Champaign, IL:
University of Illinois.
Cohen, Andrew - Elite Olshtain
1994 "Researching the production of speech acts", in: Elaine Tarone
Susan Gass - Andrew Cohen (eds.), 143-156.
Edmondson, Willis
1981 Spoken discourse. London: Longman.
Goffman, Erving
1971 Relations in public: Microstudies of the public order. New York:
Basic Books.
Non-native refusals 63
Hartford, Beverly - Kathleen Bardovi-Harlig
1992 "Experimental and observational data in the study of interlanguage
pragmatics", in: Lawrence Bouton - Yamuna Kachru (eds.), 33-
52.
Houck, Noel - Susan Gass
in press "Nonverbal communication In non-native refusals", in: Adam
Jaworski (ed.).
Imai, Masaki
1981 Sixteen Ways to Avoid Saying No. Tokyo: Nihon Keizai Shinbun.
Jaworski, Adam (ed.)
in press Silence: Interdisciplinary perspectives.
Kasper, Gabriele
1984 "Pragmatic comprehension in learner-native speaker discourse",
Language Learning 34: 1-20.
Kasper, Gabriele - Merete Dahl
1991 "Research methods in interlanguage pragmatics", Studies in Second
Language Acquisition 13.2: 215-247.
Kinjo, Hiromi
1987 "Oral refusals of invitations and requests in English and Japanese",
Journal of Asian Culture 1: 83-106.
Rintell, Ellen M. - Candace J. Mitchell
1989 "Studying requests and apologies: An inquiry into method", in:
Shoshana Blum-Kulka - Juliane House - Gabriele Kasper (eds.),
248-272.
Rose, Kenneth
1992 "Speech acts and DCTs: How reliable is questionnaire data?" Paper
presented at TESOL, Vancouver, March.
Rubin, Joan
1983 "How to tell when someone is saying 'no' revisited", In: Nessa
Wolfson - Elliot Judd (eds.), 10-17.
Scarcella, Robin
1978 "Socio-drama for social interaction", TESOL Quarterly 12.1:
41-46.
Scarcella, Robin - Elaine Andersen - Stephen Krashen (eds.)
1990 Developing communicative competence in a second language. NY:
Newbury House.
Tarone, Elaine - Susan Gass - Andrew Cohen (eds.)
1994 Research methodology in second-language acquisition. Hillsdale,
NJ: Lawrence Erlbaum Associates.
Walters, Joel
1980 "Grammar, meaning and sociocultural appropriateness in second
language acquisition", Canadian Journal of Psychology 34:
337-345.
Watson-Gegeo, Karen Ann
1988 "Ethnography in ESL: Defining the essentials", TESOL Quarterly
22.4: 575-592.
64 Noel Houck - Susan M. Gass
Wolfson, Nessa
1981 "Compliments in cross-cultural perspective", TESOL Quarterly
15.2: 117-124.
Wolfson, Nessa - Elliot Judd (eds.)
1983 Sociolinguistics and language acquisition. Cambridge, MA:
Newbury House.
Wolfson, Nessa - Thomas Marmor - Steve Jones
1989 "Problems in the comparison of speech acts across cultures", in:
Shoshana Blum-Kulka - Juliane House - Gabriele Kasper (eds.),
174-196.
Natural speech act data versus written
questionnaire data: How data collection method
affects speech act performance~ ~
Leslie M. Beebe - Martha Clark Cummings
1. Introduction
In the early 1980's, when cross-cultural speech act research was
beginning to take hold (see Kasper 1992 for a complete review of this
literature), there was a debate raging about the preferred way to collect
data on speech acts. Manes and Wolfson (1980) claimed that the best
approach was to collect samples of spontaneous speech in natural settings
where none of the participants was aware of being observed or studied.
However, written role play questionnaires (called Discourse Completion
Tests) had been and continue to be used extensively to elicit speech act
data across different languages (e.g., Blum-Kulka 1982; Olshtain 1983;
Olshtain - Cohen 1983; Blum-Kulka - House - Kasper 1989; Beebe
- Takahashi - Uliss-Weltz 1990).
Since the present study was first presented in 1985, studies of cross-
cultural speech act realization, also known as interlanguage pragmatics
(Kasper - Dahl 1991), have still relied heavily on Discourse Completion
Tests to collect data. Kasper and Dahl rate discourse completion on the
lower end of production tasks used to collect such data, pointing out that
they are "a much used and much criticized elicitation format in cross-
cultural and IL [interlanguage] pragmatics" (1991: 221). Nevertheless,
they have been used exclusively to collect data in ten studies mentioned
in the Kasper and Dahl (1991) review, to wit: studies of requests (Blum-
Kulka 1982; Blum-Kulka - Olshtain 1986; House - Kasper 1987;
Faerch - Kasper 1989), complaints (Olshtain - Weinbach 1987),
refusals (Takahashi - Beebe 1987; Beebe - Takahashi - Uliss-Weltz
1990), corrections (Takahashi - Beebe 1993) and suggestions (Banerjee
- Carrell 1988).
As Kasper and Dahl (1991) point out, up to now, few attempts have
been made to compare data collection techniques. Rintell and Mitchell
(1989) compared data collected with Discourse Completion Tests and
closed role plays and found that they yielded very similar data. Kasper
66 Leslie M. Beebe - Martha Clark Cummings
and Dahl (1991) suggest that this is because neither data collection
procedure is interactive.
Bodman and Eisenstein (1988) compared data collected through
Discourse Completion Tests, open-ended role plays and field notes on
naturalistic data. The data differed in length and complexity, with
Discourse Completion Tests being shortest and least complex and
naturalistic data most complex.
Beebe and Takahashi (1989a; 1989b) used natural data as a supple-
ment to written discourse completion data in discussing performance of
face threatening acts between interlocutors of different status: disagree-
ment, chastisement, and giving embarrassing information, such as telling
others they have spinach in their teeth.
In an effort to establish the reliability of Discourse Completion Tests,
Rose (1992a) compared Discourse Completion Tests with multiple
choice questionnaire data and found significantly fewer hints on the
Discourse Completion Tests than on the Multiple Choice Questionnaires.
He did not compare questionnaire data and natural data. In a second
study, Rose (1992b) collected data using two types of Discourse
Completion Tests. "One form included hearer response, while the other
form did not. The two forms were identical in all other respects. The
results showed that although responses on the No Hearer Response
Discourse Completion Test tended to be slightly longer and use slightly
more supportive moves and downgraders, inclusion of hearer response
did not have a significant effect on requests elicited" (1992 b: 49).
Finally, Dahl (in progress) compared authentic discussions with open-
ended role plays and found, as reported in Kasper - Dahl (1991:
244):
The most important features that distinguished between authentic and
role play productions across discourse types were amount of talk and
directness in the performance of face-threatening acts. Amount of talk also
distinguished the two types of role plays from each other, with the inter-
active role plays producing less talk and the monologic role plays more talk
than their authentic counterparts.
As amount of talk typically distinguishes between different interlocutor
relationships (cf., Wolfson's [1989] bulge hypothesis), and directness inter-
acts with contextual factors in conveying politeness (see Kasper 1990, for
an overview), the discomforting conclusion suggested by Dahl's study is
that role plays are not representative of authentic interaction on these
measures.
However, Dahl emphasized that the way the role plays were elicited
implied a number of constraints that might have seriously reduced the
generalizability of her study. Moreover, she warned that the circumstances
Natural speech act data versus written questionnaire data 67
of the data collection might have introduced some extraneous factors that
could have impaired the validity of the role plays.
This brief review of the literature to date indicates that the debate
continues over the reliability and validity of Discourse Completion Test
data and that oral role plays, closed or open, do not solve all the
problems inherent in the collection of speech act data. We present our
study in support, with certain caveats, of the continuation of Discourse
Completion Test data collection.
We support the continuation of Discourse Completion Test data
collection, while in fact, we have reservations about all the methods that
have been used to collect data on speech act performance. As we see it,
each approach to data collection has strengths and weaknesses. Since
there have been concerns about the naturalness of discourse completion
data, we offer this chapter as a qualitative and quantitative analysis of the
differences between written Discourse Completion Test data and natural
spoken data from telephone conversations. In the end, we would like to
suggest that naturalness is only one of many criteria for good data and
that other approaches featuring natural data have drawbacks of other
kinds.
Beebe (1992) discusses the strengths and weaknesses of natural data
("ethnographic" and "notebook" data) in a paper on "questionable
questions" - expressions which are syntactically yes-no questions but
which function as criticisms, topic nominations, complaints, compliment
responses, suggestions, etc. Beebe argues that the weaknesses of written
questionnaire data have been widely discussed, but that less attention has
been paid to the problems that exist with "ethnographic" data.
Ethnographic data may be natural, and natural data may be good in that
they represent spontaneous natural speech as it really is. But ethno-
graphic data and notebook data are often unsystematic. The social
characteristics (e. g., age, socioeconomic status, ethnic group) of the in-
formants are frequently unreported and often unknown. There are vastly
different numbers of informants in each social category. The data are
unsystematically collected as well. Most, but not all, examples tend to
come from an undefined target population, and the sample population as
well, is often undefined.
The stated goal of ethnographic research on speech act performance
and social rules of speaking is to characterize the sociolinguistic norms of
a "speech community" (in the sense of Hymes 1972a; 1972b; e.g., in
Wolfson 1983; Daikuhara 1986). However, the family, colleagues,
friends, and acquaintances, not to mention the associated strangers,
68 Leslie M. Beebe - Martha Clark Cummings
around a researcher are not necessarily a "speech community." In a large
urban center, the population tends to be very mobile - geographically and
socially - and the circle of friends and colleagues of the researcher will
not necessarily share a speech variety. Furthermore, when the researcher's
graduate students participate in the data collection (e.g., Manes -
Wolfson 1981; Holmes 1988), the target and sample populations can
become even more problematic to define. It is circular to argue that a
group is a speech community because it shares a linguistic variety, and it
shares a linguistic variety because it is a speech community.
This concern is voiced in addition to the criticism that the so-called
"ethnographic" data, though natural, are not truly ethnographic and that
in the field of sociolinguistics, we have only begun to investigate the
social rules of speaking in their societal context. In our field, researchers
are currently studying spoken data and speculating about the ways they
reflect societal values, but we are not really investigating societal values
as anthropologists might.
To add to these difficulties, there are problems with tape-recording.
What can be taped with approval is a biased subset of the natural speech
that is spoken. Writing down data in a notebook solves these dilemmas
to some extent, but presents accuracy problems. Reconstructed dialog
and even memorized or immediately recalled data are more likely to be
accurately recorded in terms of pragmatic force than in terms of actual
linguistic structure. At best, with training and practice we can memorize
the core speech act and perhaps a few supporting expressions. The larger
linguistic context must be reconstructed approximately. (See Beebe 1994
for a complete discussion of this data collection method.)
2. Method
2.1. Data collection procedures
In the present study, our purpose was to compare data from written
Discourse Completion Tests and telephone conversation data tape-
recorded and transcribed with permission from the respondents. Once the
data were collected, we counted the number of words used by respon-
dents filling out two-turn Discourse Completion Tests and talking on the
telephone, then identified the semantic formulas they used according to
the categories developed in Beebe, Takahashi, and Uliss-Weltz (1990).
Natural speech act data versus written questionnaire data 69
2.2. Subjects
There were 22 subjects in the sample reported here. Eleven native
English-speaking teachers of English as a second language (ESL) in the
TESOL Program at Teachers College, Columbia University were
approached individually and asked to fill out a single written discourse
completion item. Eleven other native English-speaking teachers of English
as a second language, all members of New York State TESOL, were
called on the telephone and asked the same question. Some of these, it
turned out, were also students at Teachers College. All 22 subjects were
female and American.
2.3. The request
In both the Discourse Completion Test and the telephone call, the
requester asked the ESL teacher if she would be willing to help out at the
TESOL '85 convention in New York. On the Discourse Completion Test,
the hypothetical request read:
My name is Susan Miller. I'm calling on behalf of Jim Jenkins and the Local
Committee for TESOL '85 ... and we're really sort of desperate for volun-
teers to help out on-site at the convention here in New York. I was
wondering, if you haven't already volunteered, if you would like to now... ?
On the telephone, the requester said the same thing, using her own name
and the name of Jim Lydon, the Local Committee Chair. If the recipient
of the call volunteered (approximately twenty did), Cummings gave their
names and phone numbers to the TESOL '85 Local Committee. If the
recipient of the call refused, the requester informed her that she had tape-
recorded the call for TESOL study and a asked permission to use the
data. The requester continued to make telephone calls until 11 refusals
were collected. All eleven refusers gave their permission.
3. Results
3.1. Amount of talk
Table 1 indicates the differences in the amount of talk between the
written role plays and spoken responses. (See Appendix for classification
70 Leslie M. Beebe - Martha Clark Cummings
Table 1. Amount of talk in female native speakers' refusals to volunteer at
TESOL '85
Context of Request
Characteristics
of Refusals
Discourse Completion
Test (Written)
n= 11
Telephone Con-
versation (Oral)
n =11
Total Frequency
words 611 2719
sentences 60 229
semantic formulas 61 103
semantic repetitions 1 31
semantic elaborations 1 19
turns 20 85
Frequency in First Turn
words 497 386
sentences 39 26
semantic formulas 42 25
semantic repetitions 1 3
semantic elaborations 0 3
Average Frequency per Refusal
words 55.54 247.18
sentences 5.45 20.87
semantic formulas 5.54 9.36
semantic repetitions 0.09 2.81
semantic elaborations 0.09 1.72
turns 1.81 7.72
of refusals). On every measure the total amount of talk on the telephone
far exceeded the amount on the questionnaire. There were more than
four times as many words spoken, 3.8 times as many sentences spoken,
almost twice as many semantic formulas used, and more than four times
as many turns taken. A semantic formula, described by Fraser (1980) and
cited by Olshtain and Cohen (1983: 20) "consists of a word, phrase or
sentence which meets a particular semantic criterion or strategy,
and ... can be used to perform the act in question." Whereas repetitions
and elaborations were extremely infrequent in the written data (one
each), they were very common on the telephone (31 repetitions and
elaborations).
Natural speech act data versus written questionnaire data 71
One speaker, for instance, began by saying that the only reason she
couldn't help out was that she had made other plans. On her next turn
she said that the convention was scheduled at a bad time. Finally she indi-
cated that the organization might be at fault, scheduling conferences
when certain members were celebrating their religious holidays.
There were no such elaborations on excuses in the written data. On
the Discourse Completion Test, the layout of the written questionnaire
encouraged the respondents to imagine a conversation in which they
would have only two turns. Therefore, it was necessary to include every-
thing of importance they had to say on the imaginary "first turn."
Teachers filling out the questionnaire used more words, sentences, and
semantic formulas on turn one than did those responding over the
telephone. It could be argued that the lower overall amount of talk
is a function of the smaller number of turns and that if there were an
equal number of turns, there would be more comparable levels of talk.
On the other hand, it could be argued that the Discourse Completion
Test not only biases the respondent toward packing the whole refusal
into the first turn, but also that the written nature of the task, plus the
fact that it is test-like, and the fact that it is imagined, biases respondents
toward an answer that summarizes rather than elaborates and that
responds definitively rather than hedges and negotiates. _Thus, the
key formulas that are needed to fill the social requirements of the parti-
cipants are generally stated at once in the Discourse Completion Test
response.
There is evidence to suggest that the second analysis is the correct
one. Written role plays bias the response toward less negotiation,
less hedging, less repetition, less elaboration, less variety and ultimately
less talk. For one thing, the amount of talk (measured in number
of words, sentences, and formulas) sharply fell off in the second
turn of the Discourse Completion Test. There were no repetitions.
Only one elaboration occurred. And, out of the average of 55.5
words total per refusal, an average of 45.1 of them were used in the first
turn.
The evidence seems to point toward the testing instrument as a biasing
factor. Only one of the eleven Discourse Completion Test respondents
used all of the space provided on the questionnaire. This Teachers College
student might have been eliminated from the sample since she was a close
personal friend of the data collector and knew that the telephone study
was being conducted. She was kept in the study because she had not seen
or heard the data, nor was she acquainted with the classification of
72 Leslie M. Beebe - Martha Clark Cummings
semantic formulas. Interestingly, however, her response was both the only
questionnaire to use up all the space and the only one to use a joke for
avoidance. (The requester's second turn reads: "Oh, OK. I understand.
Well, thanks anyway. Maybe another time." To which the friend replied:
"Another time? Hmm. When is the next TESOL in New York? Will I
have reached retirement?") The friend's response was the only one to
request empathy from the hypothetical caller, and it was the only one to
repeat the requester's words as an avoidance strategy. Thus, it seems that
knowing the data collector and knowing the hypothetical item on the
questionnaire was in fact a real question being asked on the telephone by
her friend on behalf of the TESOL '85 Committee, the subject made a
more realistic response.
This leads us to discuss the Discourse Completion Test in terms of
what Wolfson (1981; 1985) called her "extremes follow similar pattern"
theory or her "bulge theory." Wolfson (1983: 125-126) (using the terms
of Brown and Gilman 1960) in her study of invitations, found that
"power, on the one hand, or inequality of status, favors direct invitations
and disfavors attempts at negotiation or expressions of good intent." She
had "no examples at all of ambiguous invitations given to a superior."
Her data showed that "solidarity which leads to reciprocity is, indeed, a
prerequisite to the initiation of invitation negotiations." She found that
another dimension - intimacy - was also important. For example, "in
cases where participants are intimates who share the same social status,
fear of rejection is minimized, and as a consequence, negotiation is often
unnecessary." Finally, it is with "nonintimates of approximately equal
social status" - exactly our situation on the telephone - that most
negotiation takes place. Wolfson (1983; 1985; 1988) claimed that in-
timates and strangers pattern similarly in all of her work on speech acts
between native speakers of English.
Our field notes on naturally occurring data support Wolfson's
hypothesis. Strangers are brief. If they want to say "no," they do so.
Intimates are also brief. It is friends and other acquaintances who
are most likely to get involved in long negotiations with multiple
repetitions, extensive elaborations, and a wide variety of semantic
formulas. In the telephone conversations reported here, although the
interlocutors were strangers, there was not one "no" refusal. Only one
occurred in the questionnaire data for the same request. We would
contend they were not strangers in the usual sense, because they shared
what Goffman (1967: 109) calls "equal and joint membership in a large
organization" .
Natural speech act data versus written questionnaire data 73
With regard to our data (see Table 1), we wish to make three claims:
1) The discourse Completion Test as a data collection method disfavors
the long negotiated sequences which occur in natural conversation.
2) Common membership in a known social network (such as New York
State TESOL) reduces social distance and lends the type of instability
to the relationship that acquaintances have. It thereby leads them in
natural conversation to negotiate in a long sequence of turns and to
talk the way they would in conversation with friends and other
acquaintances.
3) If subjects filling out a Discourse Completion Test substitute in their
imagination a known interlocutor for the stranger in the test situation
(as did the data collector's friend [see pages 71-72]), this will affect
the length, tone, and other features of the response.
Tables 2 and 3 lead us to further generalizations about the content and
tone of refusals in questionnaire versus telephone conversation data.
Analyzing the number of subjects who resorted to each formula (or
strategy), we see from Table 2 that the similarities are in many ways more
striking than the differences. No one used the performative "I refuse,"
and this, by the way, mirrors our natural participant observation data
where the performative verb, "refuse", is rare. In addition, there was only
one instance of a direct "no" in the entire corpus. The adjunct of positive
feeling, the expression of regret, the statement of negative ability or
willingness, and the excuse were the four formulas that both groups used
four or more times. A very brief response using all four of these formulas
would be, "I'd really love to help out [adjunct] but I'm sorry [regret] I
can't [negative ability] because my family and I are going upstate the
week of the convention [excuse]." These formulas seem to fulfill the
stereotypical American requirements for politeness and clarity in situa-
tions where specificity is needed and one wants to establish or maintain
some level of rapport. And, these four categories are the ones that are
relatively frequent (used by 1/3 or more of the subjects) for both
questionnaire and telephone conversations. Thus, the similarities between
natural spoken refusals and written questionnaire refusals are quite
strong - strong enough to suggest that Discourse Completion Tests are
a good way to discover what semantic formulas are frequently used (or
expected) in performance of a speech act.
Table 2 also leads us to see differences between questionnaire
and natural responses to a request, however. From the complete list of
semantic formulas, thirteen were never used by a single subject in writing;
74 Leslie M. Beebe - Martha Clark Cummings
Table 2. Number of female native speakers using semantic formulas in refusals to
volunteer at TESOL '85
Context of Request
Semantic Formulas and Discourse Completion Telephone Con-
Strategies for Refusal Test (Written) versations (Oral)
n = 11 n= 11
Direct
performative verb 0 0
no 1 0
negative ability/willingness 9 8
Indirect
regret 8 4
wish 0 2
excuse/reason 11 9
alternative
offer alternative 0 1
suggest alternative 1 2
condition for acceptance 2 3
promise 1 0
principle 0 2
philosophy 0 1
attempt to dissuade
guilt trip 0 1
criticism 0 3
request for empathy 1 4
avoidance
nonverbal avoidance 0 1
topic switch 0 0
joke for avoidance 1 0
repetition of request 1 0
postponement 1 2
hedging 0 7
Adjuncts to Refusals
positive feeling/opinion 6 5
empathy 0 2
gratitude 1 0
self-defense 1 1
Natural speech act data versus written questionnaire data 75
only eight were never used by a subject on the telephone. Thus telephone
responses were not only longer, more repetitive, and more elaborated,
they were more varied in terms of the number of different formulas and
strategies resorted to. Houck and Gass (this volume) similarly found that
oral role plays of refusals showed more complex responses than
Discourse Completion Test data.
Whereas Table 2 gives us the number of subjects who resorted to using
a formula, Table 3 indicates the number of times each formula was used by
the eleven subjects in both the written role play and the actual telephone
conversation settings. Again as in Table 2, the similarities are reassuring for
researchers who use Discourse Completion Tests. The frequency counts for
all formulas or strategies, with all the subcategories included, were very
similar. That is, in only 5 out of 27 formulas, strategies, or even sub-
categories was there a difference of three or more tokens.
This shows us that in many respects, written questionnaires accurately
reflect the content expressed in natural speech. Questionnaires yielded 17
excuses; telephone conversations contained 16. Questionnaires had 12
statements of negative ability/willingness; telephone conversations
contained 14. Questionnaires said "I'm sorry" 11 times; telephone
responses used it 9 times. The content was in many important ways very
similar.
We would like to argue that these findings legitimize the use of
Discourse Completion Test data for certain purposes in sociolinguistic
research. They indicate that native speakers of a language are in fact able
to write stereotypical responses that reflect the values of the native
culture. They write refusals which contain an almost formulaic core of
semantic content that meets the basic social requirements of politeness
and clarity. Tannen (1982: 9), in her comparison of recorded spontaneous
conversation and transcribed narratives, also found that spoken narrative
was more elaborated, giving more background information, and that
"the most striking difference is the increased integration or compactness
of the written text". Furthermore she found that the spoken versions
showed the speaker's attitude, not explicitly but through paralinguistic
cues and repetition, whereas the written texts tended to remain uneval-
uated and content-focused. This was also true in our data. Speakers tended
to repeat the same phrase four or five times, such as "That's the only
problem," or "I don't even know if I'll be here" whereas in writing, each
idea was stated only once.
We are not claiming, however, that Discourse Completion Test data are
in any way a substitute for data on natural speech. They are not the same.
76 Leslie M. Beebe - Martha Clark Cummings
Table 3. Total frequencya of semantic formulas used by female native speakers in
refusals to volunteer at TESOL '85
Context of Request
Semantic Formulas and Discourse Completion Telephone Con-
Strategies for Refusal Test (Written) versations (Oral)
n =11 n =11
Direct
performative verb 0 0
no 1 0
negative ability/willingness 12 14
Indirect
regret 11 9
wish 0 2
excuse/reason 17 16
alternative
offer alternative 0 1
suggest alternative 1 3
condition for acceptance 2 3
promise 1 0
principle 0 2
philosophy 0 1
attempt to dissuade
guilt trip 0 1
criticism 0 6
request for empathy 1 6
avoidance
nonverbal avoidance 0 1
topic switch 0 1
joke for avoidance 2 0
repetition of request 1 0
postponement 2 3
hedging 0 15
Adjuncts to Refusals
positive feeling/opinion 7 12
empathy 0 3
gratitude 1 0
self-defense 1 1
a Elaborations and repetitions excluded because there were none in the written
data.
Natural speech act data versus written questionnaire data 77
Nor are oral role plays or even spontaneous responses to a preplanned
question exactly the same.
We must consider the psychological domain in addition to the strictly
social or situational setting. It is particularly in the psychological domain
where the results of this study show differences between written
questionnaire and telephone data. Analysis of Table 3 indicated that
although most categories of refusal had very similar frequency counts,
there were five categories where the questionnaire group versus tele-
phone group displayed a difference in frequency of three to fifteen
tokens:
1) Avoidance by hedging (15 telephone/ 0 Discourse Completion Test)
Example: "I don't know what you mean by volunteering."
2) Request for empathy (6 telephone/ 1 Discourse Completion Test)
Example: "I'm very, very tired. I really, really am. I drink a gallon of
orange juice a day, I get so thirsty from saying, 'This is a book! This is
a book!'"
3) Expression of empathy (3 telephone/ 1 Discourse Completion Test)
Example: "This makes your job twice as hard."
4) Expression of positive feeling (12 telephone/ 7 Discourse Completion
Test)
Example: "I am a gung-ho proponent of ESL."
5) Criticism (6 telephone/ 0 Discourse Completion Test)
Example: "It seems to me we're dealing with so many different
cultures but we're really overlooking our own."
The differences are admittedly small (except in the case of hedging), but
the findings seem important for other reasons. They reflect the psycho-
logical (as opposed to the social or situational) domain. That is, they are
closely related to feelings. We would like to claim that the main reason
the spoken data are different from the Discourse Completion Test data is
that the Discourse Completion Test, a written hypothetical exercise, does
not bring out the "psycho-social" dynamics of an interaction between
members of a group.
The literature on apologies and remedial exchanges (e.g., Goffman
1971; Olshtain 1983; Owen 1983; Olshtain - Blum-Kulka 1985; Tros-
borg 1987; House 1988; Olshtain - Cohen 1989; Rintell - Mitchell
1989; Bergman - Kasper 1993) shows us that the refusals we collected
from the telephone were also examples of what Goffman (1971: 139)
calls "remedial work," "transforming what could be seen as offensive
into what can be seen as acceptable". As Goffman (1971: 119) puts it:
78 Leslie M. Beebe - Martha Clark Cummings
"When the individual provides an account or makes an apology, he
becomes needful of the addressee's providing a comment of some kind in
return; for only in this way can he be sure that his corrective message has
been received and that it has been deemed sufficient to re-establish him
as a proper person."
Let us look more closely at the categories that were used primarily in
the telephone conversations as opposed to the questionnaires. First of all,
hedging - a type of verbal avoidance - occurred 15 times on the
telephone, but never once on the questionnaire. In our data, hedging
appeared to be an avoidance of saying "I don't want to" or "I can't."
Seven teachers hedged, most claiming they didn't know where they
would be.
The expression of empathy occurred six times in the telephone data.
As one subject said, "Oh, Martha, I really appreciate your fix! (laughs) I
really do!" Then she proceeded to request empathy by saying she was
"over her head" in two similar situations. Later she expressed more
empathy, "I don't envy your task. I commiserate with you." Colorful
language, such as "I really appreciate your fix," is typical of real inter-
action. Only one person expressed empathy on the Discourse Completion
Test and that was the data collector's close friend. It wasn't seen as neces-
sary to express empathy toward a fictional character on paper. However,
the more formulaic expressions of positive opinion do occur, particularly
compliments about how wonderful other conventions were. These ex-
pressions are also more frequent with real interlocutors where the
psychological dynamics make the refuser want to re-establish rapport.
Although positive feelings seemed appropriate for both questionnaire
and real settings, criticism and "guilt tripping" (Example: "You caught
me at a bad time. It's Saturday night, you know. I'm trying to get ready
to go out.") occurred only in spontaneous telephone responses. Goffman
(1967) points out that when remedial work is attempted and no acknow-
ledgement seems to be forthcoming, the guilty party has no alternative
but to express indignation. One teacher, after insisting three times that
her only problem was that she had already made plans to go upstate,
decided to let out her real feelings. She argued for approximately six
minutes that TESOL had been insensitive in the scheduling of state and
national events on Jewish holidays. In one short excerpt from her
criticism, she said:
I think it's disgusting. I really do. I mean it's supposed to be this united
organization and it's turning ... Whatchamacallit ... It's alienating a whole
part of its constituency. I think that's sad. I really do. I, for one, am turned
Natural speech act data versus written questionnaire data 79
off by it myself and I'm not very religious, but I feel for my friends who are,
who are constantly being discriminated against. And I'll tell you I think I
would be more active myself if it wasn't for the way I'm feeling. I just have
this really indifferent feeling now and it's sad.
It is extremely unlikely that a hypothetical situation could evoke such
strong emotion as the actual scheduling of TESOL during Passover in
1985.
Finally, we feel that the telephone conversation data may have inadver-
tently been biased by us. At the outset of the study we decided to remain
neutral by interacting as little as possible with the telephone respondents,
giving only the minimal responses of "mm hm," "uh-huh," etc. However,
this was not the expected response in a remedial exchange. Owen (1983:
57) found that "if one speaker merely acknowledges that remedial work
has been performed, rather than accepting it, it ... has the effect of 'elicit-
ing' further remedial work," which he refers to as "elaborations" or
"recyclings. "
Schegloff (1982: 74), too, points out that if we make the sounds "uh
huh" and "mm hm," commonly referred to as "accompaniment signals"
or "backchannel actions" which he calls "continuers," the speaker
understands that we expect her to go on. Not only that, but "in passing
the opportunity to do a fuller turn," we are also "passing the opportunity
to do something in particular - the opportunity to do whatever might
have been relevantly done at that point" (1982: 87).
Tannen's (1982) investigations of different conversational styles
further support the notion that by saying "uh-huh" instead of some alter-
native, expected response, such as "Oh, that's OK," or "That's all right,"
or "It's not your fault," we may have inadvertently indicated that we
were not satisfied with the remedial work and wanted more - more
excuses, more hedging, more elaboration and justification. As Owen
(1983: 104) puts it, a pause immediately following an apology or excuse
leads the listener to feel "that an acknowledgement or acceptance is being
withheld, and the inference may be drawn that the remedial work offered
is being rejected." Our telephone respondents, then, may have spoken for
an unnaturally long time, or may have used a wider array of excuses than
would have been necessary had they received the expected response to
their remedial work.
80 Leslie M. Beebe - Martha Clark Cummings
4. Conclusions
Returning to our original research questions, we asked whether question-
naire data were an accurate reflection of spoken data or a useful research
method in other respects. In this chapter we argue that Discourse
Completion Tests are a highly effective research tool as a means of:
1) Gathering a large amount of data quickly;
2) Creating an initial classification of semantic formulas and strategies
.that will likely occur in natural speech;
3) Studying the stereotypical, perceived requirements for a socially
appropriate response;
4) Gaining insight into social and psychological factors that are likely to
affect speech and performance; and
5) Ascertaining the canonical shape of speech acts in the minds of speakers
of that language.
However, they are not intended to give us natural speech and they do not
"accurately reflect natural speech or even unselfconscious, elicited speech
with respect to:
1) Actual wording used in real interaction;
2) The range of formulas and strategies used (some, like avoidance, tend
to be left out);
3) The length of response or the number of turns it takes to fulfill the
function;
4) The depth of emotion that in turn qualitatively affects the tone,
content, and form of linguistic performance;
5) The number of repetitions and elaborations that occur;
6) The actual rate of occurrence of a speech act - e. g., whether or not
someone would refuse at all in a given situation.
Thus, we support the continued use of Discourse Completion Tests, while
acknowledging their many weaknesses. They do not give us natural
speech, nor do they claim to do so. To date, however, many studies of
natural speech have not given us scientifically collected speech samples
that represent the speech of any identifiable group of speakers. They do
not give us situational control, despite the fact that situation is known to
be one of the most influential variables in speech act performance.
Discourse Completion Test data do not have the repetitions, the number
of turns, the length of responses, the emotional depth, or other features
of natural speech, but they do seem to give us a good idea of the stereo-
Natural speech act data versus written questionnaire data 81
typical shape of the speech act - at least in this case of refusals. Since the
data from speech act studies are generally used by teachers and
researchers in TESOL and, more generally, cross-cultural communica-
tion, we believe that native speaker perceptions of what constitutes an
appropriate refusal, apology, or request is valuable information. We did
not discover a single semantic formula to amplify the classification of
semantic formulas as a result of collecting natural data. All the semantic
formulas had been found in earlier questionnaire data, though not all
were found in the questionnaire data for this study.
In the end, we advocate the comparison of data collected by different
data collection procedures, and we urge researchers of interlanguage and
native speaker pragmatics to gather data through multiple approaches
since each approach has its own strengths and weaknesses.
Appendix
Classification of refusals
I. Direct
A. Performative (e.g., "I refuse")
B. Non-performative statement
1. "No"
2. Negative willingness/ability ("I can't." "I won't." "I don't think so.")
II. Indirect
A. Statement of regret (e. g., "I'm sorry "; "I feel terrible ... ")
B. Wish (e.g., "I wish I could help you ")
C. Excuse, reason, explanation (e.g., "My children will be home that night."; "I
have a headache ... ")
D. Statement of alternative
1. I can do X instead of Y (e.g., "I'd rather ... " "I'd prefer ... ")
2. Why don't you do X instead of Y (e.g., "Why don't you ask someone else?")
E. Set condition for future or past acceptance (e. g., "If you had asked me earlier, I
would have ... ")
F. Promise of future acceptance (e.g., "I'll do it next time"; "I promise I'll ... " or
"Next time I'll ... " -using "will" of promise or "promise")
G. Statement of principle (e.g., "I never do business with friends.")
H. Statement of philosophy (e.g., "One can't be too careful.")
I. Attempt to dissuade interlocutor
1. threat or statement of negative consequences to the requester (e. g., "I won't
be any fun tonight" to refuse an invitation)
2. guilt trip (e.g., waitress to customers who want to sit a while -"I can't make
a living off people who just order coffee.")
3. criticize the request/requester, etc. (statement of negative feeling or opinion);
insult/attack (e.g., "Who do you think you are?"; "That's a terrible idea.")
82 Leslie M. Beebe - Martha Clark Cummings
4. request for help, empathy, and assistance by dropping or holding the request
5. let interlocutor off the hook (e. g., "Don't worry about it." "That's okay."
"You don't have to.")
6. self defense (e.g., "I'm trying my best." "I'm doing all 1 can do. " "1 no do
nutting wrong."
J. Acceptance which functions as a refusal
1. unspecific or indefinite reply
2. lack of enthusiasm
K. Avoidance
1. nonverbal
a. silence
b. hesitation
c. do nothing
d. physical departure
2. verbal
a. topic switch
b. joke
. c. repetition of part of request, etc. ("Monday?")
d. postponement (e.g., "I'll think about it.")
e. hedging (e.g., "Gee, 1 don't know." "I'm not sure.")
Adjuncts to refusals
1. Statement of positive opinion/feeling or agreement ("That's a good idea ... "; "I'd
love to ... ")
2. Statement of empathy (e. g., "I realize you are in a difficult situation.")
,3. Pause fillers (e.g., "uhh"; "well"; "oh"; "uhm")
Note: This appendix was originally Appendix C in Beebe, L. M., T. Takahashi and
R. Uliss-Weltz, 1990. "Pragmatic transfer in ESL refusals". In Scarcella, R. C.,
E. Andersen, and S. C. Krashen (eds.), Developing Communicative Competence in
a Second Language. Rowley, MA: Newbury House Publishers, Inc.
Notes
* The authors gratefully acknowledge the assistance of Frances Williams, a
student at Teachers College Columbia University, who helped us with the
tabulation of the data, and Ximena Waissbluth, a student at the Monterey
Institute of International Studies, and Naomi Fujita, a student at Teachers
Columbia University, who helped us with the preparation of the final draft of
the manuscript. We also thank Heinle and Heinle for their permission to use the
classification of semantic formulas first published as Appendix C in Beebe,
L.M., T. Takahashi, and R. Uliss-Weltz, 1990, "Pragmatic transfer in ESL refu-
sals", in Scarcella, R. C., E. Andersen and S. C. Krashen, (eds.), Developing
Communicative Competence in a Second Language. Finally, we thank Tomoko
Takahashi and Robin Uliss-Weltz who contributed to the development of a
system of classification of semantic formulas for refusals.
Natural speech act data versus written questionnaire data 83
References
Banerjee, Janet - Patricia L. Carrell
1988 "Tuck in your shirt, you squid: Suggestions In ESL", Language
Learning 38: 313-364.
Beebe, Leslie M.
1992 "Questionable questions", Paper presented at the TESOL
Convention, Vancouver, B. C.
Beebe, Leslie M.
1994 "Notebook data on power and the power of notebook data", Paper
presented at the TESOL Convention, Baltimore.
Beebe, Leslie M. - Martha Clark Cummings
1985 "Speech act performance: A function of the data collection
procedure?" Paper presented at the TESOL Convention, New York.
Beebe, Leslie M. - Tomoko Takahashi
1989a "Do you have a bag? Social status and patterned variation in second
language acquisition", in: Susan Gass - Carolyn Madden - Dennis
Preston - Larry Selinker (eds.), 103-125.
1989 b "Sociolinguistic variation in face-threatening speech acts", in:
Miriam Eisenstein (ed.), 199-218.
Beebe, Leslie M. - Tomoko Takahashi - Robin Uliss-Weltz
1990 "Pragmatic transfer in ESL refusals", in: Robin Scarcella - Elaine
Andersen - Stephen Krashen (eds.), 55-73.
Bergman, Marc - Gabriele Kasper
1993 "The interlanguage of apologizing: cross-cultural evidence", In:
Gabriele Kasper - Shoshana Blum-Kulka (eds.).
Blum-Kulka, Shoshana
1982 "Learning how to say what you mean in a second language: A study
of speech act performance of learners of Hebrew as a second
language", Applied Linguistics 3: 29-59.
Blum-Kulka, Shoshana - Juliane House - Gabriele Kasper (eds.)
1989 Cross-cultural pragmatics: Requests and apologies. Norwood, NJ:
Ablex.
Bodman, Jean - Miriam Eisenstein
1988 "May God increase your bounty: The expression of gratitude in
English by native and non-native speakers", Cross Currents 15:
1-21.
Brown, Roger - Albert Gilman
1960 "The pronouns of power and solidarity", In: Thomas A. Sebeok
(ed.),253-276.
Coulmas, Florian (ed.)
1981 Conversational routine: Explorations in standardized communi-
cation situations and prepatterned speech. Berlin: Mouton de
Gruyter.
Dahl, Merete
in progress Authentic and role play interaction: A study of method.
[Unpublished Ph.D. dissertation, University of Copenhagen.]
84 Leslie M. Beebe - Martha Clark Cummings
Daikuhara, Midori
1986 "A study of compliments from a cross-cultural perspective: Japanese
vs American English", The Penn Working Papers in Educational
Linguistics. 2: 103-134.
Dechert, Hans W. - Manfred Raupach (eds.)
1989 Transfer in language production. Norwood, NJ: Ablex.
Eisenstein, Miriam (ed.)
1989 The dynamic interlanguage. New York: Plenum Press.
Faerch, Claus - Gabriele Kasper
1989 "Internal and external modification in interlanguage request realiz-
ation", in: Shoshana Blum-Kulka - Juliane House - Gabriele
Kasper (eds.), 221-247.
Fine, Jonathan (ed.) .
1988 Second language discourse: A textbook of current research. Vol.
XXV. Norwood, NJ: Ablex.
Fraser, Bruce
1981 "On apologizing", in: Florian Coulmas (ed.), 259-271.
Gass, Susan - Carolyn Madden (eds.)
1985 Input in second language acquisition. Rowley, MA: Newbury House.
Gass, Susan - Carolyn Madden - Dennis Preston - Larry Selinker (eds.)
1989 Variation in second language acquisition: Discourse and pragmatics.
Clevedon: Multilingual Matters.
Gass, Susan - Larry Selinker (eds.)
1983 Language transfer in language learning. Rowley, MA: Newbury
House.
Goffman, Erving
1967 Interaction ritual: Essays on face to face behavior. Garden City:
Anchor Books.
Holmes, Janet
1988 "Paying compliments: A sex-preferential politeness strategy",
Journal of Pragmatics 12: 445-465.
House, Juliane
1988 "'Oh excuse me please ... ' apologizing in a foreign language", in:
Bernhard Kettemann - Peter Bierbaumer - Alwin Fill -
Annemarie Karpf (eds.), 303-327.
House, Juliane - Gabriele Kasper
1981 "Politeness markers in English and German", in: Florian Coulmas
(ed.),157-185.
1987 "Interlanguage pragmatics: Requesting in a foreign language", in:
Wolfgang Lorscher - Rainer Schulze (eds.), 1250-1288.
Hymes, Dell
1972a "Models of interaction of language and social setting", in: John
Macnamara (ed.), 8-28.
1972 b "On communicative competence", in: John B. Pride - John Holmes
(eds.),269-293.
Kettemann, Bernard - Peter Bierbaumer - Alwin Fill - Annemarie Karpf (eds.)
1988 Englisch als Zweitsprache. Tiibingen: Narr.
Natural speech act data versus written questionnaire data 85
Kasper, Gabriele
1990 "Linguistic politeness: Current research issues", Journal of
Pragmatics 14: 193-218.
1992 "Pragmatic transfer", Second Language Research 8: 203-231.
Kasper, Gabriele - Shoshana Blum-Kulka (eds.)
1993 Interlanguage pragmatics. New York: Oxford University Press.
Kasper, Gabriele - Merete Dahl
1991 "Research methods in interlanguage pragmatics", Studies in Second
Language Acquisition 13: 215-247.
Lorscher, Wolfgang - Rainer Schulze (eds.)
1987 Perspectives on language in performance: Festschrift for Werner
Hullen: Tiibingen: Narr.
Macnamara, John (ed.)
1972 "Problems of Bilingualism", Journal of Social Issues.
Manes, Joan - Nessa Wolfson
1981 "The compliment formula", in: Florian Coulmas (ed.), 115-132.
oIshtain, Elite
1983 "Sociocultural competence and language transfer: The case of
apology", in: Susan Gass - Larry Selinker (eds.), 232-249.
Olshtain, Elite - Shoshana Blum-Kulka
1985 "Degree of approximation: Non-native reactions to native speech act
behavior", in: Susan Gass - Carolyn Madden (eds.), 18-35.
Olshtain, Elite - Andrew Cohen
1989 "Speech act behavior across languages", in: Hans W. Dechert
Manfred Raupach (eds.), 53-68.
Olshtain, Elite - Liora Weinbach
1987 "Complaints: A study of speech act behavior among native and non-
native speakers of Hebrew", in: Jef Verschueren - Marcella
Bertuccelli-Papi (eds.), 195-208.
Owen, Marion
1983 Apologies and remedial interchanges. Berlin: Mouton Publishers.
Pride, John B. - Janet Holmes (eds.)
1972 Sociolinguistics. Harmondsworth: Penguin.
Richards, Jack - Richard Schmidt (eds.)
1983 Language and communication. London: Longman.
Rintell, Ellen - Candace J. Mitchell
1989 "Studying requests and apologies: An inquiry into method", in:
Shoshana Blum-Kulka - Juliane House - Gabriele Kasper (eds.),
248-272.
Rose, Kenneth
1992a "Speech acts and DCTs: How reliable is questionnaire data?" Paper
presented at the TESOL Convention, Vancouver, B. C.
1992 b "Speech acts and questionnaires: The effect of hearer response",
Journal of Pragmatics 17: 49-62.
Scarcella, Robin - Elaine Andersen - Stephen Krashen (eds.)
1990 Developing communicative competence in a second language. New
York: Newbury House.
86 Leslie M. Beebe - Martha Clark Cummings
Schegloff, Emanuel
1982 "Discourse as an interactional achievement: Some uses of 'uh huh'
and other things that come between sentences", in: Deborah Tannen
(ed.), 71-93.
Sebeok, Thomas (ed.)
1960 Style in language. Cambridge, MA: MIT Press.
Takahashi, Tomoko - Leslie M. Beebe
1987 "The development of pragmatic competence by Japanese learners of
English", JALT Journal 8: 131-155.
1993 "Cross-linguistic influence in the speech act of correction", In:
Gabriele Kasper - Shoshana Blum-Kulka (eds.), 138-157.
Tannen, Deborah
1982 "Oral and literate strategies in spoken and written narratives",
Language 58: 1-21.
1985 "Silence: Anything but", in: Deborah Tannen - Muriel Saville-
Troike (eds.), 93-111.
Tannen, Deborah (ed.)
1982 Analyzing discourse: Text and talk: Georgetown University round-
table on language and linguistics, 1981. Washington, D.C:
Georgetown University Press.
Tannen, Deborah - Muriel Saville-Troike (eds.)
1985 Perspectives on silence. Norwood, NJ: Ablex.
Trosberg, Anna
1987 "Apology Strategies in natives/non-natives", Journal of Pragmatics
11: 147-167.
Verschueren, Jef - Marcella Bertuccelli-Papi (eds.)
1987 The pragmatic perspective. Amsterdam: John Benjamins.
Wolfson, Nessa
1981 "Invitations, compliments and the competence of native speakers",
International Journal of Psycholinguistics 24. 7-22.
1983 "Rules of speaking", in: Jack Richards - Richard Schmidt (eds.),
61-87.
1985 Research methodology and the question of validity. Paper presented
at the TESOL Convention, New York.
1988 "The Bulge: A theory of speech behavior and social distance", in:
Jonathan Fine (ed.), 21-38.
1989 Perspectives: Sociolinguistics and TESOL. New York: Newbury
House.
Part II
Speech Acts in a Second Language
Initiating and maintaining solidarity
Cross-cultural realization of greetings in
American English
Miriam Eisenstein Ebsworth - Jean W. Bodman
Mary Carpenter
1. Introduction
Greetings are among the first speech acts that are learned by children in
their native languages. Dogancay (1990) identifies greetings among the
routines explicitly taught to children. Greetings commonly appear in the
speech of American English-speaking children between the ages of nine
months to eighteen months. Greeting is important in developing and
maintaining social bonds in all age groups. In fact, greeting rituals have
been found in nearly all cultures (Levinson 1983). In our study we invest-
igated the use of greetings by native and non-native English speakers in
an attempt to better understand their common function.
Simplified greetings are introduced early in most second language
courses and are often included in texts on cross-cultural communication
(Chan 1991; jupp - Hodlin 1983; Morgan 1990). Our research shows
that greetings are complex, involving a wide range of behaviors and a
sensitivity to many situational and psychosocial variables. Greetings in
American English are made up of a range of linguistic and non-verbal
choices which may include a simple wave or smile, a single utterance or
a lengthy speech act set which can involve complex rules
and take place over a series of conversational turns. This study reveals
that non-natives have significant difficulty in performing greetings in a
manner that is acceptable to native speakers of American English.
2. Background
With increased mobility of peoples throughout the world and the break-
down of small, egalitarian face-to-face societies (Gumperz 1982),
communicative conventions have become more important in establishing
understanding and acceptance. Greeting is one of the functions in
90 Miriam Eisenstein Ebsworth - Jean w: Bodman - Mary Carpenter
language that establishes a platform for acceptance creating a positive
social bond between interlocutors. When it is not performed well, it can
result in confusion, awkwardness and hostility.
Greetings have a high perceptual saliency because they often open
conversations. However, most speakers are consciously aware of only a
small number of high-frequency, ritualized semantic formulas that
contribute to greetings. Despite the fact that some descriptions of
greetings exist in the literature, there is a significant need for research on
how greetings are truly performed. Scarcella (1979) studied how Anglos
versus Hispanics get to know each other and Nine-Curt (1977) reported
on greeting behavior among Puerto Ricans. Scarcella found that topics
differed between the two groups she studied, and Nine-Curt found some
patterns that could cause potential misunderstanding. Searle (1969)
defines greeting as an illocutionary act, simpler than many other func-
tions because it lacks propositional content and thus has no requirement
for sincerity. Searle is not alone in viewing greeting as a simple utterance
expressing recognition of another person's presence, which, in English,
may also be followed by a phatic expression relating to health and well
being. In their 1985 book, TESOL Techniques and Procedures, Bowen,
Madsen and Hilferty claim that greetings are part of phatic communica-
tion. "Phatic expressions are mostly frozen sentence partials" (p.102).
"They communicate attitudes rather than just bare facts" (p. 102). Here,
as in other cases in the literature, greeting is discussed as a function
limited to the act of recognizing someone while making a comment which
is in fact devoid of content. Van Ek (1975) characterizes greeting as a
socializing function, but he limits his list to expressions such as "good
morning," "hello," "how are you?" Wilkins (1976) sees greeting as
straight-forward and highly routine. Bowen and his colleagues go so far
as to state that teaching greetings "involves little more than modeling and
practice" (p. 103).
3. Methodology
In the area of cross-cultural pragmatics, there exists a challenge for re-
searchers to capture the authenticity, creativity and richness of natural
speech while attempting to control the many variables inherent in
language use so that data from different individuals can be meaningfully
compared. Although we do not yet have an ideal solution to meet this
Cross-cultural realization of greetings in American English 91
challenge, we have come up with an approach that combines natural
observation and elicited data (Ebsworth 1992; Bodman - Eisenstein
1988).
We began by observing greetings among natives and non-natives as
they occurred in natural discourse. On the basis of these natural data, we
identified a great many situations in which different kinds of greetings
typically occurred.
We used this natural discourse as a resource for the creation of an
open-ended questionnaire in which we asked informants to construct
dialogues based on some of the situations we identified. Eisenstein and
Bodman (1986) and Bodman and Eisenstein (1988) demonstrated that
semantic formulas used in the realization of speech acts could be success-
fully elicited through carefully constructed role-plays and questionnaires.
While elicited data were more limited in demonstrating interactional
patterns, they did accurately represent some of the langu'age used in natu-
ral discourse.
Fifty native speakers of American English provided baseline data for
this study. Second language data came from two groups: One group was
composed of twenty bilingual graduate students, and the other was com-
posed of eighty adult, advanced level English as a Second Language (ESL)
students in a language program at the American Language Institute at
New York University. The first group was composed of individuals who
spoke English and one of the following languages: Spanish (two), French
(one), Hindi (one), Japanese (six), Mandarin (two), Taiwanese (three),
Greek (two), Hebrew (one), Malay (one), Arabic (one). Of the latter
group, there were twenty Japanese, thirteen Chinese, twelve Russian,
eight Spanish, seven Korean, five French/Creole, four Greek and eleven
speakers representing a variety of additional languages. The median age
of the informants (native and non-native) was twenty-three with a range
of nineteen to sixty-five years of age. The informants were predominantly
middle class.
Both the bilingual and the non-native English speaking groups were
asked to construct dialogues for seven typical greeting situations in
English (see Appendix A); then they were asked to construct dialogues for
the same situations in their native languages and to provide literal
translations into English. A total of 283 dialogues were collected. Finally,
we had native and non-native subjects role play the same situations on
videotape. A total of thirty sets of role-plays were transcribed: Ten dyads
consisted of native/native pairs; ten dyads were composed of native/non-
native pairs; and ten were non-native/non-native pairs.
92 Miriam Eisenstein Ebsworth - Jean mBodman - Mary Carpenter
We conducted twenty open-ended post-hoc interviews with partici-
pants from a subset of the populations described above to help provide
an informed interpretation of the data gathered.
4. Results
4.1. Natural data: Native English speakers
Our research leads us to question some of the assumptions in the
literature. While Searle states that there is no sincerity requirement for
greetings, we have found that greetings often convey feelings which are
reflected either in the words themselves or the tone of voice (e.g., "Oh,
it's nice to see you." or "Hi, how are you?" [warm tone]).
We found many examples in which greetings exhibited attitudes -
both sincere and insincere. We think the sincerity principle is operative
because the unmarked case presumes the speaker to be sincere. The
hearer commonly assumes that the speaker wants to be polite in
recognizing him or her. Greetings are often accompanied by a smile
and a warm and friendly tone. Yet, sometimes the feelings of the speakers
are not what they appear to be. Occasionally, because good manners in
public require politeness, cordiality will be feigned with a person who is
disliked.
Furthermore, our research does not uphold Searle and Bowen et al.'s
contention that greetings communicate "attitudes rather than just bare
facts" (p. 103). We have found that greetings contain both attitudes and
facts. As an example of their view, Bowen et al. point out that when
someone says "How are you?" an honest answer is not expected. But our
observations show that among friends and even acquaintances, an
honest answer is often given. When two professors encountered each
other in an elevator they greeted each other by Professor A saying,
"Hi, how's it going?" and Professor B responding, "Oh, this is my
worst day!" In a similar manner, when returning from a trip, a traveler
was greeted by a friend, saying, "Hi! How was your trip?" The
traveler responded, "Terrible." While the questions opening the
greetings in these exchanges mayor may not have been intended to be
phatic, the answers appear genuine since they convey both feeling and
content.
Cross-cultural realization of greetings in American English 93
Many of the writers we have quoted have suggested that greeting is
simple compared to other functions; yet, in looking at natural discourse,
we find that what people actually say is not always routine. While some
greetings that we observed were simple, short speech acts, many were
extensive and complex speech act sets. These observations were
analogous to those described for apologies by Cohen and Olshtain
(1981), and expressions of gratitude by Eisenstein and Bodman (1986)
and Bodman and Eisenstein (1988).
An analysis of our data has led us to identify a number of categories
of greetings. One style of common American greeting among intimates
and acquaintances is what we have called greetings on the run. Two
people see each other and exchange brief phatic statements or questions
which do not necessarily require responses. An example of this was
observed in the hallway of New York University between a male and a
female student who appeared to be in their late teens. This exchange
began as the speakers approached each other from different directions
and made eye contact. They exchanged smiles and began talking.
Throughout the entire exchange they continued moving past each other
until each turned and moved away. The female student began the
greeting by saying, "Hi, how ya doin'?" And the young student
responded, "Hi! Gotta run. I'm late for class." She smiled at him and said
cheerfully, "Okay!" As in the example above, greetings on the run
often contain an indirect excuse or apology for the shortness of the
communication. Greetings on the run may also contain a short utterance
that conveys reassurance that no slight is intended or that more
lengthy contact is hoped for in the future. Expressions like, "I'll call
you," "See you" or "Talk to you soon" are examples of this kind of
reassurance.
A second greeting that begins and ends abruptly, we have named the
speedy greeting. We differentiate it from greeting on the run because
information is exchanged. This dialogue was recorded between a female
and male who were colleagues. They were middle class professionals in
their 40's who had known each other for some years. The conversation
took place in an elevator on the way to their offices. The man greeted the
woman by saying,
He: Hi, how've you been?
She: Not bad. 'N you?
He: Oh, can't complain. Busy.
She: I know. Me, too.
94 Miriam Eisenstein Ebsworth - Jean ~ Bodman - Mary Carpenter
He: Oh well, gotta take off. See ya. *
She: Bye. Take care.
The third type of greeting that we identify is called the chat. This greeting
starts off like the speedy greeting, but includes a short discussion on a
topic or two before either leave-taking or the real purpose of the com-
munication is introduced. This conversation took place between two
female friends in their early 20's in an office lounge area.
Female 1: Hi!
Female 2: Hi.
Female 1: Howa ya doin'?
Female 2: All right - comfortable - pretty good. Oh! Got that letter, by
the way, that I said I was waiting for. I finally got it.
Female 1: Wow! That's great. That's pretty good.
Female 2: Look, I'll see you later.
Female 1: Okay. Bye.
The next kind of greeting that we observed we characterized as the long
greeting. One type of this greeting involves re-establishing bonds between
two people after a period of separation. This greeting is characterized by
a number of greetings separated by narrations of events that occurred
while the individuals were apart. In this example, two middle-aged,
female neighbors greet each other on the sidewalk between their houses
after about a month's separation.
Michelle:
Bea:
Michelle:
Bea:
Michelle:
Bea:
Michelle:
Bea:
Bea!
Michelle!
Where've you been? I haven't seen you around.
We were away. We just got back. What's new with you? What
have you been up to?
(Michelle reports on neighborhood news in detail.) We missed
you. How are you? It's so nice to see you. Where'd you go?
(Bea describes her vacation in detail.)
Well, I'm glad you're back. It's so nice to see you. I missed
talking to you.
Aw. Well, we're back! How have you been doing?
* In this and other examples the function of leave-taking is included to show the
brevity and context of the interaction.
Cross-cultural realization of greetings in American English 95
An interesting form of greeting occurs between people who know each
other well and speak frequently. We call this the intimate greeting. The
speakers may know each other so well that they may leave a great deal
implied or unsaid. Sometimes the greeting itself is omitted except for non-
verbal gestures. In one example a professor entered the office, walked
over to his secretary, leaned over and said to her, "Should I ask?" The
secretary shook her head "No." In another example, the husband walked
in, kissed his wife and said, "Well?" She responded, "Yes." He smiled and
said, "Great. What else did you do today?" She then began to narrate the
events of her day.
Another form of greeting that is sometimes characterized with a
very short preliminary greeting or, sometimes, no greeting at all, we
have labelled the all-business greeting. This typically occurs between
Americans who have a non-social relationship and, because they believe
that the other person's time is limited, show respect and consideration
by opening the conversation by immediately stating their business or
needs.
Client: Mr. Matone?
Joe Matone: Yes?
Client: I want to talk to you about Puerto Rico.
Joe Matone: ah? Come in. What about Puerto Rico?
The introductory greeting involves greeting between two people who are
meeting for the first time. It can be said that this greeting is one in which
the primary function of the interactions is to allow the parties to find a
connection (mutual friends, mutual experiences) or a topic of mutual
interest. The opening is not always a greeting or a formal introduction. It
can start with a comment.
A: Nice party.
B: Yes.
A: Who do you know here?
B: Bill. I work with him.
A: ah. Are you an accountant, too?
B: No, I'm in public relations.
A: ah. Well, I'm an old friend of Bill's.
There are many greetings that follow predictable patterns for special
occasions. These greetings may be highly ritualized in routine, frequently
96 Miriam Eisenstein Ebsworth - Jean ~ Bodman - Mary Carpenter
repeated contexts. In service encounters greetings may take the form of
"Yes? Can I help you?" (store); "Who's next?", or "Next?" (fast-food
restaurant); "Fill it up?" (gas station); "Good morning. AT!. [company
name] How may I help you?" (telephone receptionist).
A special category we call the re-greeting, involves acknowledging
someone you have greeted earlier or see repeatedly during the day. This
form of the re-greeting can involve a nonverbal gesture (a nod or a wave)
or a few quick words which refer to a shared topic. In this example, one
co-worker, having learned earlier in the day that her co-worker was
not feeling well, greets another by saying, "Mary? Feel better?" Mary
responds, "Yes. Thanks!"
4.2. Non-native results
We analyzed and reviewed written dialogues constructed by English
learners. In these we found that greeting poses many problems for non-
native speakers. A comparison of these data with translations of native
language dialogues of greetings showed evidence of transfer as well
as developmental problems and confusion. Role-plays were viewed and
transcribed. These data were consistent with the written data, but in-
cluded additional interactive and non-verbal aspects of the greeting
routines. This was also found in the study on expressions of gratitude
(Bodman - Eisenstein 1988) and in Beebe and Cummings' (1985) work
on refusals. Post hoc interviews with non-natives together with elicited
and observed data reflected many examples of cross-cultural dissonance
in the pragmatic systems of American English and other languages and
cultures.
The following are some general findings typical of the English
greetings of non-natives from a variety of cultures. Non-natives were able
to perform greetings in a manner that was often acceptable to native
listeners. However, sociopragmatic failure sometimes occurred. The fact
that we sometimes observed non-natives performing greetings in English
successfully may be due in part to the perceptual saliency of
greetings because they occur at the beginnings of interactions.
Slobin (1985) notes that the beginnings of utterances are easier to
remember; a parallel may exist between the beginnings of discourse.
Furthermore, greetings take place frequently, so there may be more
opportunities for learning how to perform them than other less
ubiquitous speech acts.
Cross-cultural realization of greetings in American English 97
Despite these successes there were many examples of greetings made
by non-natives that seemed a bit strange, unusual or "foreign". Also
some types of English greetings were not well understood by the non-
natives and, as a result, they did not know how to respond. This strange
or unusual "interlanguage-talk" was sometimes a result of the fact that
the other speaker said something that violated the conventions of the
non-native speaker's first language or surprised them in some manner. As
a result the non-native speaker chose to be creative, idiosyncratic or
unusual. In addition, native speakers of English exhibited a greater
variety in the types of greetings and creative language used in producing
the greeting than we might have expected from the literature on the
subject. This was particularly evident when native speakers engaged in
informal or intimate exchanges. Non-natives tended to follow rather
ritualized routines and remain formal; apparently, they lacked the reper-
toire for imitating informal repartee so common in native greetings.
Videotaped role-plays revealed that this verbal formality among non-
natives carried over into their non-verbal behavior. Non-natives took
fixed physical positions in role-plays. On the other hand, native speakers
varied their stance and gestures depending on the level of formality or
intimacy required by the setting and topic.
Sometimes the interactions of non-natives during role-plays were
judged by native speakers to be more like interrogations when comments
and expansions on the other speaker's utterances would have been more
appropriate. One non-native speaker tended to take the role of the
conversation leader; this speaker asked a question, the other speaker
answered briefly. After a short pause, while the first speaker thought up
another question, the interaction continued. Another non-native feature
of the role-plays was the presence of abrupt topic changes. The sum total
was that the interaction was awkward and unnatural. Wolfson (1989)
has commented on the same interactional problem among non-natives
who have difficulty responding to compliments; Rintell (1989) has made
a similar observation concerning the expression of emotion. The same
conversational feature can appear in the speech of native children. Ellis
(1984) noted that children exhibited avoidance behavior when they
found they were unable to keep a conversation going. They changed the
topic abruptly. This made their conversations sound like interrogations.
Two categories of greeting appear to present non-native speakers with
special problems. The non-natives found both the speedy greeting and
greeting on the run almost impossible to perform. They were unable to
make their greetings short enough. Even when they were given instruc-
98 Miriam Eisenstein Ebsworth - Jean w: Bodman - Mary Carpenter
tion to make their greetings shorter, they were unable to do so. Here is an
example from our role-play data between two non-native speakers:
A:
B:
A:
B:
A:
A and B:
I'll come in, President.
Yes, please.
By the way, what is you have something to discuss with me?
Before that, why don't you have a seat.
Oh. Thank you.
(perform a full greeting)
As receivers of these speedy greetings, they reported feeling that they were
being treated badly or rudely. Several informants pointed to these types
of greeting as evidence of Americans' lack of politeness, sensitivity and
interest in others. Although we reported above that there were fewer
instances than we expected in our data of the non-natives saying things
that were judged to be examples of sociopragmatic failure when they
expressed or responded to greetings in English, some types of English
greetings are received by non-native English speakers as a sociopragmatic
failure on the part of the native English speaker.
There were a number of instances of pragmalinguistic failure (Thomas
1983). In a few cases, students used the phrase "How do you do?" when
they meant to say "How are you?" An inappropriate use of titles was
also a problem. One example was a non-native speaker saying, "Hi,
President" to the head of a company; in another case, a non-native
speaker greeted a woman whose name was unknown by saying, "Hello,
Lady." Interactions in the workplace with persons in authority revealed
that native speakers chose language that was characterized by emotional
restraint. Occasionally, however, the non-natives who had not yet learned
the proper register used highly informal language. In several instances we
had reason to believe that the non-natives were trying to lighten a stress-
ful moment by joking, but did not have the sociolinguistic control to do
so successfully. Some subtle differences in the data showed approxima-
tion but not mastery of conventionalized language. An example of this
was, "Hi. I do not know you. My name is (name)." Our native infor-
mants felt that, "Hi, I don't think we've met. I'm (name)" would have
been more appropriate. In another example, a native speaker of Japanese
wrote the following dialogue:
A: Hello, my fellows. How's your work going?
B: Oh. President. So far, so good.
Cross-cultural realization of greetings in American English 99
Videotaped role-plays revealed that, in addition to the non-verbal
problems noted above, non-native speech often lacked appropriate
prosodic features. In particular, non-natives had difficulty approximating
the right tone and intensity. They often spoke in a monotone and lacked
the ability to project enthusiasm and warmth.
Our data, as interpreted with the help of native informants, indicate
some problems particular to certain linguistic and cultural groups in
learning to approximate American English greetings. In Puerto Rico,
greeting a friend or acquaintance is of such a high priority that on-going
conversations are often interrupted to greet passers-by. Nine-Curt (1977)
elaborates on this by noting that speakers are continually looking away
from each other to notice others in their vicinity. She refers to this as the
"rubber neck" syndrome of Puerto Ricans. When this behavior is trans-
ferred to American English, it can be distracting to the American or
perceived as impolite to the person with whom one is engaged in
conversation. In Puerto Rico, however, it would be socially inappropriate
if a person passing by was not acknowledged by his or her friend. Anglo-
Americans feel their first obligation is to the person who is speaking to
them and not in greeting the friend who is walking by; therefore, they
must maintain eye contact with the speaker in order to show interest and
be polite. Passers-by are expected to understand that if a friend is engaged
in a conversation, the friend and the conversation are not to be inter-
rupted. The friend, if he or she wishes to greet this person, will position
himself or herself just outside of listening distance and in the person's
line of sight and wait to be recognized when there is a pause in the
conversation.
As mentioned earlier, our data also demonstrate that the speedy
greeting as performed in American English presents difficulties for many
other cultures including Hispanics. When Americans must choose
between the competing obligations of greeting a friend or leaving the area
to do something else, they are able to abbreviate their greeting in a
manner that is acceptable to a person who is a member of the same
speech community. This short-cut is understood by both parties and a
slight is neither intended or, usually, felt. If the friend demands time to
talk or greet, the other interlocutor will usually quickly negotiate a time
to talk at greater length later that day or week. Hispanics appear to find
the speedy greeting next to impossible to perform. For them, friends in
one's presence take priority over other obligations, such as imminent
appointments (Ebsworth 1992). By extension a third party who is kept
waiting is expected to understand when the excuse is presented that the
100 Miriam Eisenstein Ebsworth - Jean ~ Bodman - Mary Carpenter
person was delayed b e c a ~ s e they had to say hello to a friend. American
English speakers are surprised that when they use the speedy greeting
with Hispanics, the Hispanics can experience it as disrespectful.
Just as length of greeting can vary cross-culturally, the choice of an
appropriate topic can be a source of difficulty. Certain topics that are
freely raised in American English greetings have different rules of use in
other cultures and languages. It is common to ask about the well-being of
the person being greeted as well as that person's family members in
English and many other languages. Among Arabs, Iranians and Afghans,
however, men may ask about the well-being of other male family
members, but are not as free to inquire about females as is commonly
done in the United States. Our Russian, Ukrainian and Georgian infor-
mants stated that greetings among co-workers and acquaintances did not
usually contain inquiries about well-being. When Americans greeted
them with expressions like, "Hi. How are you?" they sometimes found
the question so unexpected and startling that they responded with silence
and an embarrassed expression on their faces. They informed us that they
did not understand why the person wanted to know about their health
when they did not know this person well.
As with the Russians, Ukrainians and Georgians who are startled by
questions about well-being, speakers of American English sometimes find
themselves speechless when they are greeted in a number of countries in
Asia by two common greetings that translate, "Have you eaten?" and
"Where are you going?" One response we observed is that the native
speaker of English misinterpreted the greeting as an invitation. In the
other instance, American English speakers mistook the intent of the
greeting and, when spoken by a non-intimate, felt that the question was
inappropriate and a violation of their privacy.
Among some American subcultures, a frequent strategy for opening a
conversation with a stranger is not to greet them, but to state a poten-
tially-shared complaint. Indirect complaints have been identified by
Boxer (1993) as performing a bonding function for some Americans. This
strategy is shocking to members of other speech communities for whom
this alternative is not available.
The way that greetings are performed can vary from culture to culture.
In English, greeting usually involves serial turn-taking. In Afghanistan,
both parties often begin greeting each other simultaneously. Questions
about well-being are frequently not answered, but are overlapped by the
other speaker with a similar question about well-being. When a response
is given to "How are you?", it is often simply, "Thank you." English
Cross-cultural realization of greetings in American English 101
speakers find this confusing. For Afghans, these ritualized greetings are
obligatory each time two friends meet. While English speakers will per-
form the greeting on the run, Afghans almost always stop walking and
perform a full greeting each time they encounter each other during the
day. Americans typically wave, nod or say a word or two in subsequent
greetings unless they have a subject to discuss.
Complex rules of non-verbal behavior that accompany greeting such
as bowing, kissing, handshaking and touching vary from culture to
culture. Our Japanese informants confided that they were often highly
embarrassed when enthusiastically hugged or kissed by Americans. These
non-verbal behaviors often distinguish social classes and generational
distinctions within a particular culture.
A number of our non-native informants expressed anxiety about
greeting people in social settings. They often stated that they did not
know what to say. In examining the models that non-natives are typically
given in language classrooms, we found that the ritual of beginning with
a "hello", exchanging names and following with asking about well-being
was universally presented as a model of greeting. However, in our data,
much more variety occurs. For example, comments on the food, the
people, and/or the drinks often precede an exchange of names or personal
information. Little is available in textbook materials to show learners
how a topic of conversation is mutually developed or how native
speakers ease into formal introductions. There is not always a good fit
between American greeting rituals at parties and those common in other
cultures. For example, in Swedish, Bratt Paulston (1990) reports that a
guest is expected to go around and formally introduce him/herself to all
the other guests while in an American context introductions often
proceed under the guidance of the host or hostess or are an option, but
not a requirement, for the individual.
5. Conclusion
We have shown that greeting is a socially significant event in universal
terms, and that like other major speech acts its realization is language
specific. Greeting can consist of a single speech act or a speech act set.
Successful greetings may be simple or complex, phatic or meaningful,
formulaic or creative. Our data show that even relatively advanced non-
native English speakers experience difficulty with various aspects of
102 Miriam Eisenstein Ebsworth - Jean ~ Bodman - Mary Carpenter
American greetings on both productive and receptive levels. Challenges
for cross-cultural communication range from lexical choices to substan-
tial differences in cultural norms and values; thus, pragmalinguistic or
sociopragmatic failure may occur in cross-cultural greeting encounters.
A major implication for second language pedagogy is that models for
learning must be based on research into how greetings are actually per-
formed. Regrettably, few current texts for English as a second or foreign
language meet this criterion. Furthermore, the complexity and interactive
nature of greetings and how they are realized in different languages and
cultures must be considered.
In the area of research methodology, the mixed approach to data
gathering and analysis taken here is consistent with previous studies
indicating that the semantic formulas that appear in elicited written data
are parallel to those used in role-plays and found in natural observation.
The interactive nature of greetings and the combination of associated
verbal and nonverbal elements is most clearly illustrated in natural
language use but is effectively mirrored in role-plays. Written data are
helpful in determining the degree to which individuals are aware of
language and content appropriate for greeting others in particular socio-
linguistic contexts. The interpretation of data through interviews with
research participants representing the various cultural and linguistic
groups considered is extremely helpful in reaching an understanding of
potentially problematic group differences and in identifying areas of
shared values and perspectives. It is through such research that we will
continue to expand our awareness of language functions and improve our
understanding of cross-cultural communication.
Appendix A
Questionnaire
For each one of the situations below, write a short dialogue that represents typical
language that you and an English speaker would use if you found yourselves in
the situations described.
Situation #1: (Peers greeting each other) Two people who are friends are walking
toward each other. They are both in a hurry to keep appointments. They see each
other and say:
Situation #2: (Peers greeting each other) Two people who are friends see each
other. They are on their way to other places but are not in a hurry. They have a
minute to chat. They see each other and say:
Cross-cultural realization of greetings in American English 103
Situation #3: (Peers greeting each other) One friend has been invited to a dinner
party by another. One friend opens the door. They see each other and say:
Situation #4: (Higher and lower status) An employee has been asked to come to
the boss' office for an unknown reason. The employee knocks and enters the
room. They see each other and say:
Situation #5: (Higher and lower status) A boss happens to be passing by an
employee's work space. With no ulterior motive, the boss stops in. They see each
other and say:
Situation #6: (Stranger to stranger) Two students are sitting next to each other on
the first day of class. One turns to the other, and they say:
Situation #7: (Stranger to stranger) At a party, one person sees another person (of
the same sex) who looks friendly. They walk toward each other and say:
After you have completed these dialogues in English, please re-do the same dia-
logues in your native language. Then, if possible, provide a literal translation of
these dialogues in English.
Thank you.
Appendix B
Sample Written Responses
Non-native, Puerto Rican
A: Hi, I don't know you, but my name is Provy.
B: Are you enjoying yourself? Very well organized party, isn't it?
Native English speaker
A: Hi!
B: Hi, how you doin'?
A: Okay. You?
B: Okay. Catch you later.
Native English speaker
A: Ah, hello, (name of boss). You wanted to see me?
B: Yes. Come in. How are you?
A: Fine, thanks. How are you?
B: Fine. Have a seat.
Non-native, Israeli
A: Hi, Batya. How's school?
B: Baruch Hashem. How's your husband feeling?
104 Miriam Eisenstein Ebsworth - Jean w: Bodman - Mary Carpenter
A 42-year old native speaker of English:
A: Great party!
B: Yeah. The food is terrific.
A: The Smith's always go all out.
B: Oh? Is this an annual event?
A: Sort of. Are you new in the neighborhood?
B: Yeah. We just bought the house next door.
A: Well, welcome to the neighborhood. My name is Steve. I live at 677.
Sample natural data
Two students, native English speakers
A: How are you?
B: Fine.
A: Take care.
B: Okay.
Clerk, Customer, (strangers), native English speakers
A: Hi, how are you?
B: Good. It's nice to see the sunshine at last.
A: Right.
Sample transcripts of videotaped role-plays
Two Japanese women, college graduate students, non-native speakers
A: Hi.
B: Hi.
A: How're ya doing?
B: Thank you, I'm fine. How are you?
A: Yeah, I'm pretty good.
B: I haven't seen you around two years, right?
A: Is that so?
B: Yeah, that long. (they giggle) Where are you going?
A: I'm going to the Metropolitan Museum ...
Israeli, Jamaican women, non-native speakers
A: Hi. I never thought you have time for those romantic restaurants.
B: Aha - Oh, I love to eat. I love to go to restaurants. Fancy seeing you here.
A: I love eating, too. That's one of my main problems. You know that.
B: Yes, I do. Oh, this is my husband. Jack, this is (name).
Cross-cultural realization of greetings in American English 105
A: It's nice meeting you. That's my boyfriend.
B: Oh, nice meeting you.
A: That's my boss.
B: I see they're summoning us to the table.
A: Enjoy your meal.
B: Bye.
A: Bye.
Japanese, Israeli women, non-native speakers
A: Excuse me, ma'am, but do you have the time?
B: I see that you have the time.
A: (Looking at watch on her own wrist) Why, it's broken.
B: Oh ... It's broken. Well, it's seven past two.
A: Thank you very much. Do you like the picture, Picasso?
B: I like the picture, but it's not Picasso, I'm very sorry to tell you. It's Van Gogh.
A: Oh, I didn't know that.
B: There is a little difference between them.
A: I didn't recognize this picture. I like the picture of impressionist.
B: But this is not impressionist, I'm sorry to tell you. This is not, I'm sorry to tell
you. This is a classical picture from the nineteenth century.
A: Thank you very much for your kind lecture.
Native man and woman
A: Hi, Steve.
B: How are you doing, Adina?
A: Okay.
B: What's happening?
A: Vh ... vacation was really good.
B: Oh, yeah? Where'd you go?
A: Vh, skiing.
B: Lucky you!
A: Yeah, and you?
B: Just stayed in town, bored to tears.
A: That's too bad.
B: I know. I gotta get out of teaching. There's more money elsewhere. Oh, are you
taking a class this semester?
A: Oh, yeah, four ...
B: Four! Are you out of your mind?
A: Yep.
B: And working, too? I don't know ...
A: Yep.
B: It's crazy.
A: Yeah. Tell me about it.
106 Miriam Eisenstein Ebsworth - Jean w: Bodman - Mary Carpenter
References
Beebe, Leslie - Martha C. Cummings
1985 "Speech act performance: A function of the data collection
procedure?" Paper presented at the Sixth Annual Sociolinguistics
Colloquium, TESOL Convention, New York City.
Bodman, Jean - Miriam Eisenstein
1988 "May God increase your bounty: The expression of gratitude in
English by native and non-native speakers", Cross Currents 15:
1-21.
Bowen, J. Donald - Harold Madsen - Ann Hilferty
1985 TESOL Techniques and Procedures. Rowley, MA: Newbury House.
Boxer, Diana
1993 Complaining and commiserating: A speech act view of solidarity in
spoken American English. New York: Peter Lang Publishing.
Bratt Paulston, Christina
1990 "Linguistic and communicative competence", in: Robin Scarcella -
Elaine Andersen - Stephen Krashen (eds.), 287-302.
Chan, .Carole
1991 Latinos in the work force: Diversity and tradition. Los Angeles, CA:
Los Angeles County Commission on Human Relations.
Cohen, Andrew - Elite Olshtain
1981 "Developing a measure of sociocultural competence: The case of
apology", Language Learning 31.1: 113-34.
Dogancay, Seran
1990 "Your eye is sparkling: Formulaic expressions and routines in
Turkish", Pennsylvania Working Papers in Educational Linguistics
6.2: 49-64.
Ebsworth, Timothy J.
1992 Appropriateness of Puerto Rican/American cross-cultural
communication. [Unpublished Ph.D. dissertation, New York
University.]
Eisenstein, Miriam (ed.)
1989 The dynamic interlanguage. New York: Plenum Press.
Eisenstein, Miriam - Jean W. Bodman
1986 "I very appreciate: Expressions of gratitude by native and non-
native speakers of American English", Applied Linguistics 7.2:
167-85.
Ellis, Rod
1984 Understanding Second Language Acquisition. New York: Oxford
University Press.
Gumperz, John J.
1982 Discourse Strategies. Cambridge: Cambridge University Press.
JuPP, Thomas C. - Susan Hodlin
1983 Industrial English. London: Heinemann Educational Books Ltd.
Levinson, Stephen C.
1983 Pragmatics. New York: Cambridge University Press.
Cross-cultural realization of greetings in American English 107
Morgan, Graham
1990 "Anti-racism and language diversity: Raising metalinguistic aware-
ness to combat racism", Reading 24.3: 192-203.
Nine-Curt, Carmen J.
1977 Nonverbal communication. Cambridge, MA: National Assessment
and Dissemination Center for Bilingual Education, Lesley College.
Rintell, Ellen
1989 "The use of language to express emotion by second language
learners and native speakers", in: Miriam Eisenstein (ed.), 237-260.
Scarcella, Robin
1979 "On speaking politely in a second language", in: Carlos Yorio et al.
(eds.),275-287.
Scarcella, Robin - Elaine Andersen - Stephen Krashen (eds.)
1990 Developing communicative competence in a second language. New
York: Newbury House.
Searle, John R.
1969 Speech acts: An essay in the philosophy of language. Cambridge:
Cambridge University Press.
Slobin, Daniel
1985 The cross-linguistic study of language acquisition. Hillsdale, NJ:
Lawrence Erlbaum Associates.
Thomas, Jenny
1983 "Cross-cultural pragmatic failure", Applied Linguistics 4.2: 91-112.
van Ek, Jan A.
1975 The threshold level. London: Longman.
Wilkins, David A.
1976 Notional syllabuses. Oxford: Oxford University Press.
Wolfson, Nessa
1989 Perspectives: Sociolinguistics and TESOL. New York: Newbury
House.
Yorio, Carlos - Kyle Perkins - Jacquelyn Schachter
1979 On TESOL C79: The learner in focus. Washington, DC: TESOL.
Egyptian and Anterican contplintents: Focus on
second language learners)f-
Gayle L. Nelson - Waguida 1 Bakary -
Mahmoud Al Batal
1. Introduction
Not only did I need to know the right words. I needed to know the appro-
priate situations in which blessings were given and the appropriate
moments at which to give them (Spindel 1989: 215).
Carol Spindel lived in a small African village in the Ivory Coast and stu-
died the Dyula language in order to interact with those around her. The
view she presents, that language learners need to know more about the
target language than the phonological, morphological, and syntactic
rules, more than the "right words", is well known to applied linguists and
language teachers. In large part due to the work of Canale and
Swain (1980: 30-34), Hymes (1972: 269-288; 1974: 10-24), Wolfson
(1981: 117; 1983: 82-83; 1989: 32-53), and language teachers have
become aware of the notions of communicative competence, that
language must be not only linguistically accurate, but socially appro-
priate, and that sociolinguistic rules sometimes need to be explicitly
taught. This realization has led to the study of language in use, to the
study of the sociolinguistic rules as well as the linguistic rules of language.
One method of investigating the sociolinguistic rules of a given speech
community is to identify and study specific speech acts within that
community.
The term "speech act" has been defined as a minimal unit of discourse,
a basic unit of communication (Searle 1969: 16). Examples of speech acts
include giving compliments, making statements, asking questions,
apologizing, leave-taking, making introductions, making requests,
expressing gratitude, making refusals, and, as illustrated above, giving
blessings. This chapter presents the results of a cross-cultural study on the
speech act of complimenting, comparing various aspects of American
English and Egyptian Arabic compliments. With a focus on second
language learners, it extends an earlier discussion of these data (Nelson
- 1 Bakary - Al Batal 1993).
110 Gayle L. Nelson - Waguida El Bakary - Mahmoud Al Batal
2. Rationale
Within the communicative competence paradigm, a primary rationale
for studying speech acts is to obtain sociolinguistic knowledge of the rules
of the target language. According to this view, the study of speech
acts, such as compliments, is of interest to language teachers in order for
them to instruct students in the socially appropriate uses of compliments
in the target language. It may be difficult, however, for students to
achieve communicative competence in a second language due to the
transfer of sociolinguistic rules from their first language to their second
language.
The notion of transfer or interference, originally used to describe the
phenomena of phonological and syntactic transfer from a first language
to a second language, was adopted by sociolinguists as it became clearer
that "rules governing speech events may differ substantially from one
language group to another, thus leading to different rules and norms for
turn taking, amount of talking, speech act realizations, etc." (Schmidt -
Richards 1980: 146). Using the "rules governing speech events" from
one's first language speech community when interacting with members of
a second language speech community is referred to as pragmatic transfer.
Pragmatic transfer may lead to pragmatic failure, to a negative judgment
about a speaker such as his or her being impolite or uncooperative (Leech
1983: 281). It is hoped that by comparing speech acts across cultures,
miscommunication resulting from the pragmatic transfer of first language
rules to second language speech situations can be predicted and prevented
(Wolfson 1989: 140).
Speech act studies have been criticized as being ethnocentric in that
most have investigated variations of English (Blum-Kulka - House -
Kasper 1989: 10). The present study is valuable, in part, because it was
conducted in Arabic as well as English and the results contribute to an
understanding of the sociolinguistic rules of Arabic, an understudied area
of speech act research.
3. Compliments
The speech act of complimenting was selected for cross-cultural study for
two reasons. First, American compliments tend to be a "troublesome
aspect of English for learners from different cultural backgrounds"
Egyptian and American compliments 111
(Holmes - Brown 1987: 525). For example, non-native speakers of
English are often embarrassed by the frequency with which Americans
compliment (Holmes - Brown 1987: 525; Wolfson 1981: 123). A
second reason is that, although extensive research has been conducted on
American compliments, few cross-cultural studies have investigated
complimenting (exceptions include Barnlund - Araki 1985; Wolfson
1981). For the purpose of this study, a compliment is defined broadly as
an expression of praise or positive regard.
Wolfson and Manes (1980), using ethnographic methodology, have
collected over 1000 American compliments in a wide range of situations.
They (Wolfson 1981: 122; Wolfson 1983: 85; Wolfson - Manes 1980:
402-403) found that approximately 80 % of American compliments fall
into three syntactic patterns:
Syntactic Patterns
NP is/looks (intensifier) AD]
I like/love NP
PRO is AD] NP
Examples
Your shoes are great.
I love your perm.
These are great cookies.
Two of these three patterns depend on adjectives for their positive
semantic value and two thirds of the adjectival compliments use one of
five adjectives: nice, good, beautiful, pretty and great. With regard to
attributes praised, Americans most frequently compliment personal
appearance and ability.
Wolfson (1981: 120) has also noted cultural differences in
complimenting and observed that Iranian and Arabic speakers tend to use
proverbs and other precoded ritualized expressions when complimenting.
She gives the example of an Arabic speaker complimenting a friend's
child. The English equivalent is: "She [the child] is like the moon and she
has beautiful eyes".
Holmes and Brown (1987), also using ethnographic methodology,
collected 200 compliments in New Zealand. Their results were similar to
those of Wolfson and Manes. Almost 80 % of the compliments belonged
to one of the above three syntactic patterns; two thirds used one of five
adjectives: nice, good, beautiful, lovely and wonderful; and the most
frequently praised attributes were personal appearance and skill.
Barnlund and Araki (1985), using interviews and questionnaires,
investigated Japanese and American compliments. Interviewees were
asked to describe 1) the most recent compliment they had given and
received, 2) the relationship between the giver and recipient of the
112 Gayle L. Nelson - Waguida EI Bakary - Mahmoud Al Batal
compliment, 3) the attribute praised, 4) the exact words used in the
compliment, and 5) the day the compliment was given. Interview data
indicated that the japanese compliment much less frequently than
Americans. The compliments reported by the japanese occurred, on the
average, 13 days before the interview, whereas the compliments reported
by Americans occurred, on the average, 1.6 days earlier. Their findings
also indicate that japanese and Americans tend to compliment five
attributes: appearance, work and study, personal traits, skill and taste,
but with varying frequencies. japanese most frequently praise skill (31 %)
and work and study (19 0/0), whereas Americans most frequently praise
appearance (34 %) and personal traits (33 0/0).
Knapp, Hopper and Bell (1984) used a similar method and
asked subjects to describe recent compliments given and received and
also to provide information related to the compliment giver and
recipient. Data were analyzed for 1) attributes praised, 2) compliment
forms, and 3) relationships between giver and recipient of compliments.
Appearance, attire and performance were the most frequently praised
attributes. Compliment forms were analyzed according to syntactic
patterns and also along four dimensions: direct/indirect, specific/
general, comparison/no comparison and normal/amplified. Seventy-
five per cent of the compliments followed one of the three patterns
identified by Wolfson and Manes (1980: 402-403). Compliments
tended to be direct, general, non-comparative,
l
and normal (without
intensifier). In their analysis of the relationships between the givers
and recipients of compliments, Knapp et al. (1984: 26) found that
compliments are likely to be exchanged between individuals of the
same sex, and between individuals in close, rather than distant, relation-
ships.
Before describing the present study, it is important to discuss two
important facets of complimenting in Arabic. The first concerns the belief
in the evil eye. The evil eye refers to the "belief that someone can project
harm by looking at another's property or person" (Maloney 1976: v).
Frequently, the evil eye relates to "envy in the eye of the beholder" and is
most dangerous to pregnant women, children, and anyone who is
beautiful (Spooner 1976: 77). For example, if a person compliments a
mother on her child, the compliment, by causing the evil eye to notice the
child, may cause harm to visit the child. To counteract this effect, the
giver of the compliment invokes Allah to protect the child, saying, Allaah
yiHfazu ('May God protect him') or maa shaa'a Allaah! ('What God has
willed!'). In a study of pregnant women at the American University in
Egyptian and American compliments 113
Beirut hospital, Harfouche (1981: 87) found that 54.9 % believed in the
harmful effects of the evil eye.
2
Another important facet of Arabic complimenting is the practice of
offering the object of the compliment to the person who complimented.
For example, if Mohammed compliments Sami on his new cassette tape,
Sami might say basiita; itfaDDal! ('This is nothing; please take it!')
or m ~ a d d a m ('It is offered'). This practice of offering the object of the
compliment to the giver of the compliment seems, however, to be more
problematic for non-native Arabic speakers than native Arabic speakers.
Arabic speakers recognize this offering as a ritual and do not take it
literally whereas non-native Arabic speakers or English speakers fre-
quently accept the literal meaning and thus are either reluctant to
compliment, or embarrassed when the Arabic-speaker offers them the
object of the compliment.
4. The study
This study investigated Egyptian and American compliments to
determine similarities and differences in 1) compliment form, 2) attri-
butes praised, 3) gender of the compliment giver and recipient, and 4)
compliment frequency.
4.1. Method
An expanded version of Barnlund and Araki's (1985) procedure for
obtaining compliment data was used. Audiotaped interviews were
conducted with 20 Egyptian university students in Egypt and 20
American students in the United States. All students were between 18 to
25 years of age; half were males and half females. Interviewees were
asked to describe the most recent compliment they had given, received,
and observed, the relationship between the complimenter and the
recipient, the attribute praised, the exact words used in the compliment,
and the day the compliment was given. Interviews were conducted in the
native language of the interviewees. This procedure provided a corpus of
60 American and 60 Egyptian compliments.
This method of obtaining data has three advantages. First, the corpus
of compliments represents actual compliments given. Second, the
114 Gayle L. Nelson - Waguida EI Bakary - Mahmoud Al Batal
interviewee defines what constitutes a compliment. The interviewee may
be in a better position to define a compliment within a particular socio-
cultural context than researchers or coders. For example, one compliment
reported by an interviewee in this study, Show off macho, may have been
perceived as a criticism by researchers, but within the subculture of white
American male university students, this utterance was given as a comp-
liment. One male was complimenting another on his waterskiing.
The third advantage is that the interviewee, as the giver, receiver, and
observer of the compliment, decides when the compliment ends. This
ending point is not as apparent as one might expect. For example, an
American male reported the compliment in (1).
(1) You-'re really sweet and nice. I really appreciate your help.
(AM8)
The first part of this utterance seems an obvious compliment, but the
second part (I really appreciate your help) could be viewed as an expres-
sion of thanks. The interviewee, however, perceived the second segment
as part of the total compliment.
At the completion of the interviews, the audiotapes were transcribed;
the American compliments were transcribed in English and the Egyptian
in Arabic. The Arabic transcripts were also translated into English for the
analysis of attributes praised, gender of compliment giver and recipient,
and compliment frequency. The analysis of the form of the Arabic
compliments, however, was based on the Arabic transcripts, not the
English translations.
4.2. Results
4.2.1. Compliment form
Compliment form refers to the language used to express the compliment.
The compliments differed in their length, use of metaphor and compara-
tives, and to some extent, syntactic structure. They were also similar in
that both Egyptian and American compliments were primarily adjectival
in that an adjective was responsible for their positive meaning.
Egyptian and American compliments 115
The American compliments tended to be short, as in (2) through (4).
(2) You look great. (AF4)
(3) Your perm is nice. (AF7)
(4) Good job. (AM29)
In this particular corpus of American compliments, the average number
of words in the compliments reported by American females was 6.2 and
the average for American males was 4.3.
The Egyptian compliments were longer, as in (5).
(5) eeh sh-shiyaaka di! eeh l-fustaan da! bass, iHna ma-ntdarsh.
(EF21)
'What is all this chicness! What is this dress! Stop, we cannot
[take all that].'
The average compliment reported by Egyptian females contained 10.7
Arabic words and by Egyptian males 8.7 words.
The length of the Egyptian compliments appears to be related to two
features of Arabic discourse: 1) Repetition of almost the same idea with
a change in words and 2) The use of several adjectives in a series.
Examples of repeating a similar sentence are in (6) and (7).
(6) eeh l-Halaawa di! eeh sh-shiyaaka di! (EMI0)
'What is all this beauty! What is all this chicness!'
(7) waliid SaaHbi w-akhuuya. waliid ana caarfa a k h l a a ~ u . waliid
akhlaa'u kwayyisa w-mafiish aHsan minnu. shaab akhlaa'u
kwayyisa, sum'a kwayyisa, mafiish Hadd byitkallim calee
ghalaT. huwwa kamaan biy'aamil n-naasmu'amla SaHH.
mafihuush Haaga tit'aayib. l-'eeb l-waHiid illi fiih innu gaayiz
yi'mill-'amal wi ma y caddarsh eeh illi yiHSal. (EM17)
'Walid is my friend and [like] my brother. Walid, I know his
manners well. Walid has good manners and there is no one
better than him. He is a young man whose manners are good,
and no one says anything bad about him. He also treats people
well. He does not have any shortcomings. The only fault he has
is that he may do something and not consider the conse-
quences.'
116 Gayle L. Nelson - Waguida El Rakary - Mahmoud Al Ratal
An example of an Egyptian compliment that used a series of adjectives
is in (8).
(8) inti insaana 'add eeh kwayyisa wi zarifa wi wi Hakiima wi
razina wi zakiyya wi labiqa wi mish Haaga. (EF8)
'You are such a good and nice and sensible and wise and
serious and intelligent and diplomatic person and don't lack a
thing.'
In 15 out of 60 or 25 % of the Egyptian compliments, the speaker
repeated the same idea or used a series of adjectives when praising
another person. Only 5 % of American compliments used such patterns.
American and Egyptian compliments also differ in their use of
comparatives, particularly similes and metaphors; 11 % of the Egyptian
compliments used comparatives whereas comparatives were not used at
all in the American compliments.
Examples of Egyptian compliments using comparisons include (9)
through (11).
(9) huwwa zayy akhuuya. (EM17)
'He is like a brother to me.'
(10) A: fustaanik Hilwawi.
B: tasriHtik aHla. (EF30)
A: 'Your dress is very nice.'
B: 'Really, your hair style is nicer.'
(11) amiiSak Hilw zayy MiHammad (EF2)
'Your shirt is as pretty as [the singer] Mohammed Fouad's.'
Marriage is a common metaphor in Egyptian compliments as illustrated
by examples (12) through (14).
(12) shaklak 'ariis in-naharda. (EM8)
'You look like a bridegroom today.'
(13) eeh sh-shiyaaka di! iHna Hlawweena awi. fii miin fi-TTarit?
(EF4)
'What is all this chicness! We have become very pretty! Who is
on the way? [on the horizon]?'
Egyptian and American compliments 117
(14) mulfita li-n-naZar giddan! eeh sh-shiyaaka di! eeh l-fustaan da!
bass! iHna ma-ni'darsh (;ala kida. shaklik in-naharda zayy
l-'aruusa. (EF20)
'[You look] very attractive! What is all this chicness! What is
this dress?!
Stop! We cannot take this. You look like a bride today.'
Seven percent of the Egyptian compliments used a marriage metaphor.
No American compliments referred to marriage. An analysis of the
syntactic structure of the compliments indicated both similarities and
differences. Sixty-six per cent of the American compliments used one of
the three syntactic patterns identified by Wolfson and Manes in their
research on American compliments (NP is/looks AD]; I like/love NP;
PRO is AD] NP). Similarly, a limited number of syntactic patterns
account for the majority of the Egyptian compliments:
Syntactic Patterns
NP AD] (intensifier)
NP VP (intensifier) (AD])
eeh NP DEM. PRO.
Examples
inti shiik awi.
'You are very chic.'
(;ineeki Hlawwit awi.
'Your eyes have become very beautiful.'
eeh sh-shiyaaka di!
'What is all this chicness.'
Fifty percent of the Egyptian compliments used the syntactic pattern: NP
AD] (intensifier). This pattern is similar to the American pattern: NP
is/looks (intensifier) AD]. The Arabic version contains no verb "to be"
because in Arabic, the verb "to be" is not used in a nominal sentence;
thus, inti Hilwa means "you are beautiful". Fourteen percent of the
compliments used the second pattern: NP VP (intensifier) (AD]).
The third pattern (eeh NP Dem. Pro.), used in 14 % of the Egyptian
compliments, does not appear in the American data. It represents one
type of a precoded set of phrases that is used in particular situations that
demand complimenting; these phrases or formulas cannot be changed.
Three of the Egyptian compliments made reference to Allah (e.g., maa
shaa'a Allaah (;aleek; 'God's grace be upon you'), whereas no American
compliments referred to God.
The Egyptian and American compliments shared one major similarity:
they were primarily adjectival. Seventy per cent of the Egyptian compli-
ments and 73 % of the American compliments used adjectives. Four
118 Gayle L. Nelson - Waguida EI Bakary - Mahmoud Al Batal
Arabic adjectives, Hilw ('pretty'), Kwayyis ('good'), shiik ('chic'), and
Tayyib ('kind'), accounted for 66 % of the Arabic adjectives used. The
most frequent was Hilw which was used in 34 % of the adjectival com-
pliments. Three English adjectives, great, good, and nice were used in
74 % of the American adjectival compliments.
4.2.2. Attributes praised
In order to develop a means of classifying the compliments according to
the attributes praised, the researchers read the compliments and noted
possible classifications. Based on the researchers' notes, a classification
scheme with four categories (appearance, traits, skill, and work) was
piloted by two coders. The coders, however, were, at times, unable to
distinguish between the two categories "skill" and "work". Asa result of
the pilot test, the classification scheme was modified. The categories
"skill" and "work" were collapsed into one category, "skiIVwork". Next,
compliments were classified independently by two coders who were dif-
ferent from the first two coders. This second pair of coders classified the
compliments as belonging to one of the three categories: appearance,
traits, or skill/work.
The category, "appearance", referred to one's looks and included hair-
cuts, eyes, and clothing. "Skill/work" referred to the "quality of some-
thing produced through ... skill or effort: a well-done job, a skillfully
played game, a good meal" (Manes 1983: 101). "Traits" referred to
personality characteristics such as loyalty, kindness, maturity and intelli-
gence. An intercoder reliability of 97 % was determined by comparing the
classifications of the coders.
As shown in Table 1, personal appearance was frequently praised in
both cultures; it was praised in SO % of the Egyptian compliments and
43 % of the American. The attribute of physical appearance is, however,
not as straightforward as it first appears. Manes (1983: 99) found that
Americans tend to compliment "aspects of personal appearance which
are the result of deliberate effort, not simply natural attractiveness." Our
American data support Manes; only one compliment praised a natural
attribute. A male praised a female friend, saying, You have nice-looking
legs.
Egyptian and American compliments 119
Table 1. Attributes praised in Egyptian and American compliments
Egyptian
Appearance
(Natural)
30*
(15)
50%*
(250/0)
American
Appearance
(Natural)
26*
(1)
43%**
(1.6%)
Traits
20
33%
Traits
6
10%
SkillslWork
7
12%
SkillslWork
28
47%
Other
3
5%
Other
o
0%
Total
60
100%
Total
60
100%
*
number of occurrences out of 60 compliments
** percentage (based on 60 compliments)
However, half of our Egyptian compliments on appearance praised natu-
ral attributes. Some of the compliments are exemplified in (15) and (16).
(15) sha'rik gamiil. (EF25)
'Your hair is beautiful.'
(16) Hilwa awi. (EM12)
'Your eyes are very pretty.'
Consistent with the work of Manes (1983: 98), Americans in the present
study tended to compliment clothes and hair. Egyptians also compli-
mented clothes and hair, but in addition, they complimented eyes, skin,
and general attractiveness.
The second largest category of Egyptian compliments was "traits"
which accounted for 33 % of the Egyptian corpus. Examples are in (17)
through (19).
(17) shakhSiyyitik kwayyisa. (EMll)
'Your personality is good.'
(18) inti wi zakiyya wi diblumasiyya. (EF8)
'You are mature and intelligent and diplomatic.'
120 Gayle L. Nelson - Waguida EI Bakary - Mahmoud Al Batal
(19) inti Tayyiba wi shakhSiyyitik kwayyisa. (EF16)
'You are very kind and your personality is good.'
"Traits" accounted for only 10 % of American compliments. The largest
category of American compliments was "skills/work" which accounted
for 47% of the American corpus. Examples of American compliments on
"skill/work" include examples (20) through (23).
(20) You did a great job. (AF19)
(21) This paper is really good. (AM12)
(22) Nice catch. (AM21)
(23) You guys are doing a great job. (AM13)
Twelve percent of the Egyptian compliments praised skills or work.
4.2.3. Gender of compliment giver and recipient
As shown in Table 2, of the compliments reported, 61 % of those given
by Egyptian females and 62 % given by American females praised appear-
ance. Forty-three per cent of the compliments given by Egyptian males
praised appearance, whereas only 29 % of those given by American males
praised appearance. Both Egyptian and American males praised females
on appearance more frequently than they praised males on appearance.
Egyptian males praised personality traits in 41 % of the compliments and
American males praised skill or work in 590/0. American males were
twice as likely to praise males on "skill/work" as females.
4.2.4. Compliment frequency
The interview data indicate that the Americans complimented more
frequently than the Egyptians. Americans reported compliments that
were given, on the average, 1.6 days before the interview. Egyptians
reported compliments that were given, on the average, 8.6 days before
the interview.
Egyptian and American compliments 121
Table 2. Attributes praised according to nationality and gender of giver and
receiver of compliments
Egyptian
Females
Appearance Traits SkillslWork Other Total
Female to 8 (35%) 4 (180/0) 3 (13%) 1 (4%) 16 (690/0)
Female
Female to 6 (260/0) 1 (40/0) 0 (00/0) o(00/0) 7 (31%)
Male
Total 14 (61 %) 5 (22%) 3 (13%) 1 (4%) 23 (1000/0)
Males
Male to 12 (32 %) 8 (220/0) 0 (0%) 2 (5%) 22 (59%)
Female
Male to 4 (11 %) 7 (19%) 4 (11 %) 0(0%) 15 (41 %)
Male
Total 16 (430/0) 15 (41 0/0) 4 (11 0/0) 2 (50/0) 37 (100%)
American
Females
Appearance Traits SkillslWork Other Total
Female to 14 (54%) 0 (0%) 4 (15%) 0(0%) 18 (690/0)
Female
Female to 2 (80/0) 2 (8%) 4 (15%) o(00/0) 8 (31 %)
Male
Total 16 (620/0) 2 (8%) 8 (300/0) 0(0%) 26 (100%)
Males
Male to 6 (18%) 2 (6%) 6 (180/0) o(00/0) 14 (42%)
Female
Male to 4 (120/0) 2 (6%) 14 (41 0/0) o(00/0) 20 (59%)
Male
Total 10 (290/0) 4 (12%) 20 (590/0) o(00/0) 34 (100%)
122 Gayle L. Nelson - Waguida El Bakary - Mahmoud Al Batal
5. Discussion
In order for students to become communicatively competent in a second
language, they need to learn both the linguistic and sociolinguistic rules
of conversational discourse. However, achieving communicative comp-
etence may, at times, be complicated due to the transfer of rules from
students' first language to their second language. One of the goals of
cross-cultural studies such as this one is to predict the inappropriate
transfer of first language rules to second language situations and to
provide "a basis for determining which areas are most vulnerable to mis-
communication and should therefore be focused on" (Wolfson 1989:
140).
The results of this study provide linguistic and sociolinguistic inform-
ation about Arabic compliments that can be helpful to both instructors of
English as a second language who teach Arabic speakers, and to students
and teachers of Arabic as a second language. Linguistic and socio-
linguistic knowledge of Arabic compliments can help teachers of English
as a second language 1) predict areas where Arabic speakers may have
difficulty, 2) recognize when Arabic speakers are transferring native
language conventions to their use of English, and 3) understand the
reasons why Arabic speakers make certain linguistic and sociolinguistic
errors. Conversely, linguistic and sociolinguistic knowledge of English
compliments can help teachers of Arabic as a second language 1) predict
areas where English speakers may have difficulty, 2) recognize when
English speakers are transferring first language conventions to second
language situations, and 3) understand the reasons why English speakers
make certain linguistic errors or act inappropriately in some socio-
linguistic contexts. Such sociocultural knowledge may also result in less
pragmatic failure, fewer cross-cultural misunderstandings, and improved
intercultural communication.
It seems reasonable to assume that areas of similarity between
American and Arabic compliments will cause language learners less diffi-
culty than areas of difference. Similarities include the speech act itself;
complimenting seems common in both cultures. Both cultures also share
similarities in compliment form and attributes praised. Over 70 % of
both the Arabic and English compliments were adjectival; they depended
on an adjective for their positive semantic value, and both used a limited
number of adjectives. Consistent with the findings of other researchers
(Wolfson - Manes 1980: 400), this study suggests that Americans
frequently use the adjectives good, nice, and great. Common Arabic
Egyptian and American compliments 123
adjectives include Hilw, Kwayyis, shiik and Tayyib. Both Egyptian and
American compliments also tended to use a limited number of syntactic
patterns, and the most frequent pattern in both sets of data was similar:
NP is/look (intensifier) AD] for American compliments and NP AD]
(intensifier) for Egyptian compliments. Finally, in both cultures, males
and females tended to praise females on their personal appearance,
particularly on their clothes and hair. Students of English and Arabic can
use these similarities between Egyptian and American compliments to
their advantage by learning the limited number of adjectives used in com-
plimenting, the common syntactic patterns that are similar in both
languages, and the attributes that are praised in both cultures.
Although the two cultures appear to share some similarities in com-
pliment form and attributes praised, they also seem to differ in these two
areas. With regard to compliment form, the Egyptian compliments in this
study tended to be longer than the American compliments. This length is
related in part, to repetition, a feature of Arabic discourse (Suleiman
1973: 292). Arabic speakers use repetition to express their feelings; the
more something is valued, the more the repetition. With compliments, the
more the repetition, the better the compliment. Other facets of Arabic
contributing to compliment length are long arrays of adjectives (Shouby
1951: 291) and elaboration (Almaney - Alwan 1982: 83-84).
The findings of this study suggest that Egyptians also tend to use more
similes and metaphors than Americans. Again, similes and metaphors are
not particular to compliments; they are common in Arabic (Shouby 1951:
298-299). In this particular corpus of Egyptian compliments, marriage is
a frequent metaphor, perhaps suggesting not only the importance of
marriage in Egyptian society, but the centrality of the family.
The American compliments, on the other hand, are more likely to be
short and less often include metaphors and similes. For Arabic-speaking
students learning English, it may appear that American compliments
would be simple and easy to learn, but although the syntax seems simple,
such plain, unelaborate utterances may be difficult for an Arabic speaker
because they seem inadequate; they do not seem to express what he or she
wants to say. Compliments such as Nice shirt may be perceived by Arabic
speakers as flat and relatively meaningless. If Arabic speakers, in an
attempt to make compliments sound sincere to their own ears, use more
words than would a native English speaker, "pragmatic failure might
result from overindulgence in words," causing native speakers to sense a
lack of appropriateness (Blum-Kulka - Olshtain 1986: 175). English
speakers, on the other hand, may have difficulty with the repetition and
124 Gayle L. Nelson - Waguida EI Bakary - Mahmoud Al Batal
metaphor in Arabic compliments, feeling phony and insincere when
repeating the same compliment in several ways. In this case, pragmatic
failure may result not from too many words, but from too few.
Wolfson (1981: 19) notes that Arabs compliment in the form of
"proverbs and other precoded ritualized phrases." She uses a proverb as
an illustration:
Speaker 1: X is a nice girl and beautiful.
Speaker 2: Where is the soil compared with the star?
In line with Wolfson's findings, it was expected that the Egyptian data in
this study would contain a few proverbs. This was not the case. The data
did, however, illustrate the frequency of precoded set formulas in Arabic
compliments (e. g., eeh il-Halaawa di! eeh sh-shiyaaka di! 'What is all this
beauty! What is all this chicness.'). These qualitative findings are consis-
tent with the quantitative findings of Nelson, EI Bakary and Al Batal
(1993) in which the third most popular compliment form chosen by
Egyptians was "use a formulaic expression."
The results of this study also suggest sociolinguistic differences
between Egyptian and American compliments in factors such as attri-
butes praised, gender, and frequency of complimenting. Consistent with
other studies (Barnlund - Araki: 1985: 13-14; Knapp et al. 1984:
17-18; Manes 1983: 98-102), these findings suggest that Americans
tend to compliment skills and personal appearance, and are likely to com-
pliment someone of the same gender (Knapp et al. 1984: 26). This study
also suggests a relationship between the gender of the compliment giver
and recipient and the attribute being complimented. For example, the
American males tended to compliment other males on skills and work,
whereas American females tended to compliment other females on
appearance. Like Americans, Egyptians tended to compliment personal
appearance, but unlike Americans, Egyptians, particularly males, tended
to compliment males and females on personality traits. Neither American
males nor females were particularly likely to compliment individuals on
their personalities. Both second language Arabic and English teachers can
help non-native speakers become more communicatively competent by
assisting them in becoming more aware of appropriate attributes to
praise and the role of gender in complimenting.
A final difference between American and Egyptian compliments
appears to be the frequency with which individuals compliment. In this
Egyptian and American compliments 125
study, Americans seemed to compliment more frequently than Egyptians.
This tendency of Americans to compliment frequently is problematic to
many non-native speakers of English who are embarrassed by what they
perceive as excessive complimenting. If Egyptians compliment less often
than Americans, they may perceive American compliments to be
insincere. Such an interpretation may lead to a communication break-
down. Let's suppose, for example, that an American student or teacher,
in an attempt to be friendly and open a conversation, repeatedly com-
pliments an Egyptian in class. If the Egyptian student perceives the com-
pliment as insincere, he or she may feel uncomfortable and pull away
from the person who complimented. The result may be that, instead of
becoming friendly, the Egyptian becomes distrustful of the American.
Second language English teachers can help prevent such misunder-
standings by structuring classroom activities so that students learn that
Americans compliment frequently and that compliments serve functions
such as maintaining social harmony, opening conversations, and
greetings (Wolfson 1983: 87-90).
Ideally, speech acts should be studied in their natural context using
ethnomethodology (Wolfson 1983: 94). Ethnomethodology is difficult
for cross-cultural studies due to problems of comparability (Blum-Kulka
- House - Kasper 1989: 13) and a lack of ethnographers who belong
to speech communities other than English-speaking ones. Although the
compliments in this study were not observed under natural conditions,
our findings for native speakers of English are similar to compliments
collected by Wolfson and Manes in natural settings. This similarity
suggests the reliability of this study's methodology and data. However,
the method has limitations. Because these compliments were not collected
in a natural setting, it is difficult, if not impossible, to answer the
question: What functions do compliments serve in Arabic? Other ques-
tions also exist - How does power distance and status interact with
complimenting behavior? How does familiarity interact with compli-
menting? What are common responses to compliments?
In addition to the above questions, two other facets of Arabic com-
plimenting did not appear in the data. When an Arabic speaker
compliments an individual, particularly a pregnant woman or child, the
speaker often uses an invocation such as Allaah yiHfazu ('May God
protect him') to protect the individual from bad luck. For non-native
speakers of Arabic, the omission of such an invocation may produce
misunderstanding, and for non-native speakers of English, the transfer of
this rule to English may also result in misunderstandings. An ethno-
126 Gayle L. Nelson - Waguida EI Bakary - Mahmoud Al Batal
graphic study of compliments in various contexts would provide
information on this practice that would be helpful to both second
language Arabic and English students.
A second facet of Arabic complimenting that did not appear in the
data relates to the practice of offering the object of a compliment to the
person who complimented. Although it is assumed that this is a common
practice in Arabic-speaking countries (Almaney - Alwan 1982: 98-99),
little is known about the conditions surrounding its use. The ethno-
graphic study of Arabic compliments in context would contribute to our
understanding of these situations and would help students of Arabic learn
what these situations are, what to say, and how to respond.
These questions suggest fruitful areas of additional inquiry. This study
represents a first step in the study of complimenting in Arabic, but
additional steps are needed to paint a more complete picture of Arabic
complimenting, to understand the appropriate situations in which
compliments are given and the appropriate moments at which to give
them.
Notes
* This chapter is based on data from an ongoing research project comparing
Arabic and English speech acts such as compliments and compliment
responses. The authors would like to thank Joyce Neu and John Murphy for
their helpful comments on versions of this chapter.
1. Examples of comparative compliments are You play better than Jerry and
You ere prettier than ever.
2. We also have the evil eye phenomenon in the U. S. (and other Western cultures)
when we say "Knock on wood" to maintain good luck.
References
Almaney, A.J. - A.J. Alwan
1982 Communicating with Arabs. Prospect Heights, IL: Waveland Press,
Inc.
Barnlund, Dean C. - Shoko Araki
1985 "Intercultural encounters: The management of compliments by
Japanese and Americans". Journal of Cross-Cultural Psychology 16:
9-26.
Blum-Kulka, Shoshana - Juliane House - Gabriele Kasper
1989 "Investigating cross-cultural pragmatics: An introductory over-
view", in: Shoshana Blum-Kulka - Juliane House - Gabriele
Kasper (eds.), 1-34.
Egyptian and American compliments 127
"The influence of the Arabic language on the psychology of the
Arabs", Middle East Journal 5: 284-302.
1974
Blum-Kulka, Shoshana - Juliane House - Gabriele Kasper (eds.)
1989 Cross-cultural pragmatics: Requests and apologies. Norwood, NJ:
Ablex.
Blum-Kulka, Shoshana - Elite Olshtain
1986 "Too many words: Length of utterance and pragmatic failure",
Studies in Second Language Acquisition 8: 165-179.
Canale, Michael - Merrill Swain
1980 "Theoretical bases of communicative approaches to second language
teaching and testing", Applied Linguistics 1: 1-47.
Dundes, Alan (ed.)
1981 The evil eye: A folklore casebook. New York: Garland Publishing.
Harfouche, Jamal Karam
1981 "The evil eye and infant health in Lebanon", in: Alan Dundes (ed.),
86-106.
Holmes, Janet - Dorothy F. Brown
1987 "Teachers and students learning about compliments", TESOL
Quarterly 21: 523-546.
Hymes, Dell
1972 "On communicative competence", in John B. Pride - Janet Holmes
(eds.),269-293.
Foundations in sociolinguistics: An ethnographic approach.
Philadelphia: University of Pennsylvania Press.
Knapp, Mark L. - Robert Hopper - Robert A. Bell
1984 "Compliments: A descriptive taxonomy", Journal of Communica-
tion 34: 19- 31.
Leech, Geoffrey N.
1983 Principles of pragmatics. London and New York: Longman.
Manes, Joan
1983 "Compliments: A mirror of cultural values", in: Nessa Wolfson -
Elliot Judd (eds.), 96-102.
Maloney, Clarence (ed.)
1976 The evil eye. New York: Columbia University Press.
Nelson, Gayle L. - Waguida El Bakary - Mahmoud Al Batal
1993 "Egyptian and American compliments: A cross-cultural study",
International Journal of Intercultural Relations, 17: 293 -313.
Pride, John - Janet Holmes (eds.)
1972 Sociolinguistics. Harmondsworth, Middlesex: Penguin.
Prosser, Michael H. (ed.)
1973 Intercommunication among nations and people. New York: Harper
and Row.
Schmidt, Richard W. - Jack C. Richards
1980 "Speech acts and language learning", Applied Linguistics 1: 129-157.
Searle, John R.
1969 Speech acts. Cambridge: Cambridge University Press.
Shouby, E.
1951
128 Gayle L. Nelson - Waguida El Bakary - Mahmoud Al Batal
Spindel, Carol
1989 In the shadow of the sacred grove. New York: Vintage Books.
Spooner, Brian
1976 "The evil eye in the Middle East'" in: Clarence Maloney (ed.),
76-84.
Suleiman, Michael W.
1973 "The Arabs and the West: Communication gap", in: Michael H.
Prosser (ed.), 278-303.
Wolfson, Nessa
1981 "Compliments in cross-cultural perspective", TESOL Quarterly 15:
117-124.
1983 "An empirically based analysis of complimenting in American
English", in: Nessa Wolfson - Elliot Judd (eds.), 82-95.
1989 Perspectives: Sociolinguistics and TESOL. New York: Newbury
House/Harper & Row Publishers.
Wolfson, Nessa - Elliot Judd (eds.)
1983 Sociolinguistics and language acquisition. Rowley, MA: Newbury
House.
Wolfson, Nessa - Joan Manes
1980 "The compliment as social strategy", Papers in Linguistics:
International Journal of Human Communication 13: 391-410.
Politeness strategies in French and English
Michael L. Geis - Linda L. Harlow
1. Introduction
Anyone who wishes to learn to speak a second language faces two
problems. The first is to learn the language system. The second is to learn
to use the language. Since learning a language is a precondition on
learning to use it, much of the energy in second language education is
devoted to the former task. However, unless students are also taught how
to use the second language, we may expect that they will fall back on
language patterns acquired in the course of learning to use their first
language to guide how they use the second. The inevitable result is that
students will commonly employ forms that are contextually inappro-
priate in that they differ in style, politeness, and register from what native
speakers would employ.
The problem language learners face is that there is an extraordinary
amount of stylistic, politeness, and register variation in any language,
and, therefore, in any speech context, there will be a large number of
utterances available to him or her which communicate what he or she
wants to communicate, but which will differ in style, politeness, and
register, and therefore differ interactionally.
In this chapter, we report on a pilot study we have done concerning
how English, French, and English-speaking French learners do requesting
in an informal experimental conversational context (in which subjects
engage cooperatively in assembling jigsaw puzzles). The goal of the study
was to determine which English and French forms are politeness equiv-
alents and to what degree French learners employ forms used by French
speakers. Before discussing the study, and its implications for teaching
French to speakers of English, it will be necessary to present our theoret-
ical assumptions in some detail.
In order to understand the dimensions of the problem second
language learners face, it will be useful to consider a hypothetical
example drawn from English. Suppose John goes to a store to
130 Michael L. Geis - Linda L. Harlow
purchase "widgets," where he encounters Sue, who is a clerk, and the
conversation in (1) ensues:
(1) John: Ya got any widgets?
Sue: Yeah.
John: Why d o n ~ t you give me three?
Sue: Sure.
Now, at turn T1, John could have said any of the utterances of (2) by way
of determining whether or not Sue has widgets for sale.
(2) a. Have you got any widgets?
b. Ya got any widgets?
c. Got any widgets?
Or, he could have said any of the sentences of (3).
(3) a. Do you have any widgets?
b. Ya have any widgets?
c. Have any widgets?
The same is true of (4), and stylistic variants thereof.
(4) a. Do you carry widgets? (Ya carry widgets?, Carry widgets?)
b. Do you stock widgets? (Ya stock widgets?, Stock widgets?)
All of these sentences ask whether or not Sue has widgets available for
sale, and will be said therefore to be transactionally equivalent. However,
they differ significantly from each other in style and register, and some-
one learning English must learn when any particular form will and will
not be contextually felicitous.
The three utterances of (2) do not appear to differ at all in literal
meaning (L-Meaning), as is characteristic of stylistic variants. We might
represent the L-Meaning of these sentences informally as in (5), where
"Sp" stands for the speaker and "Ad" for the addressee.
(5) Sp inquires of Ad whether or not Ad possesses widgets.
The sentences of (3) also do not differ from each other in L-Meaning, and
arguably do not differ in L-Meaning from those of (2) either, which is to
Politeness strategies in French and English 131
say not only that the variation in (3) is stylistic, but that the choice
between using have and got is also stylistic.
Proposition (5), though it accounts for the literal meanings of (2) and
(3), does not adequately represent how they will be interpreted in con-
text, for John is not asking simply whether Sue possesses widgets but
whether she possesses them for sale. If one goes into a store that has light-
bulbs illuminating the store, but does not sell them, and asks whether
they have lightbulbs, the normal answer will be, "No," even though that
is technically false.
1
The reason is that clerks hear utterances like (2)
and (3) as having a meaning in context or S-Meaning (for contextual
significance), something like that represented in (6).
(6) Sp inquires of Ad whether or not Ad possesses widgets for sale.
In short, clerks hear utterances like (2) and (3) as having essentially the
same meaning in context as the utterances of (4), forms that are restricted
in use to the commercial register, despite the obvious differences in literal
meaning.
2
One of the reasons for positing the two levels of discourse meaning,
L-Meaning and S-Meaning, is that we cannot speak meaningfully of
linguistic variation unless we can say that we have something that varies
and something that stays the same. In the case of stylistic variation, what
stays the same is L-Meaning and what varies is linguistic form. However,
register variants clearly do not have the same L-Meaning (have and got
do not have the same L-Meaning as carry and stock) and so we must look
to a different level of meaning to express how register variants are the
same, namely S-Meaning. The same is true when we come to politeness
variation as we shall shortly show.
What was said of turn T1 goes equally for T3. Sp might have said any
of the sentences of (7) at T3 by way of signaling that it is three widgets
that Sp desires.
(7) a. I'll take three.
b. Give me three.
c. Why don't you give me three?
d. How about three?
e. Could I have three?
f. Could you give me three?
The utterances of (7) clearly differ from each other in L-Meaning.
However, they would do precisely the same work at turn T
3
, namely
132 Michael L. Geis - Linda L. Harlow
communicate that the number of widgets desired is three. As such, they
could be argued to have the same significance in context or S-Meaning
(i. e., are transactionally equivalent) were they to be uttered at turn T
3
.3
The forms in (7) are neither stylistic, nor register variants. Instead,
they differ in politeness. Utterances (7a) and (7e) differ from (7b)-(7d)
and (7f) in that the former are egocentric (the subject is I) and the latter
exocentric (the subject is you - an understood you in the case of (7b)).
All things being equal, exocentric forms are more polite than are ego-
centric forms. Utterances like (7c)-(7f) can be said to show deference to
the addressee in that they suggest that the addressee has an option as to
whether or not he or she will comply with the request, even though he or
she very well may not have such an option. These utterances are there-
fore more polite than utterances like (7a) and (7b) that do not show
deference. In English, utterances displaying deference are interrogative
(cf., Can I have three?), while those that do not display deference are
either imperative (Give me three) or declarative ( I ~ l l have three).
The politeness interrogatives of (7) are worthy of special note.
Questions can be used to request information, of course, but there are
very clear cases, such as these, in which they are being used actually to
provide information (namely, that it is three widgets that John wants),
not request it.
4
Suppose, as another example, that John goes into a frozen
yoghurt store and notes that the store offers large dishes of pineapple
yoghurt for sale at a price he can afford. At such a point, he could use
any of the sentence forms in (7) to frame his request (substituting "a large
pineapple" for "three," of course). In this case, it would be clear that
none of the interrogative utterances would be used to request information
- John has all the information he needs to place his order. Instead, he is
using these utterances to provide information - the information that he
wants a large pineapple yoghurt. It is when we find interrogatives being
used to convey information, rather than request it, that we have clear
cases of what we might call "politeness interrogatives".
As we have seen, someone learning English has to cope with a very
great deal of stylistic, register, and politeness variation. He or she must
learn what register, style, and politeness options (sociopragmatic options)
are available, how these options are realized morphosyntactically and
prosodically, and when any given option will be appropriate. The same
would be true, of course, in learning another language. To complicate
matters, different language cultures 5 may vary in what sociopragmatic
options are available, and even when the same options are available, they
may take different morphosyntactic and prosodic forms.
Politeness strategies in French and English 133
Suppose, for instance, that we have two students who are studying
together in informal circumstances and one wants to borrow a piece of
paper from the other. A French student might use either a negative or
positive declarative intonation question like those in (8) by way of
making a polite, deferential request, whereas an American student would
most likely use the interrogative form (9).6
(8) a. Tu n ~ a s pas de papier??
'You don't have any paper?'
b. Tu as du papier?
'You have some paper?'
(9) Do you have any paper?
This contrast between French and English is quite striking, for though the
French do have an inverted interrogative form like (9), namely (10), it is
very rarely used in informal contexts (Desirat - Harde 1988; Gadet
1989; Di Vito 1991, Valdman 1967), and the literal English equivalent to
(8a), namely (11), would never be used to make a polite request.
(10) As-tu du papier?
(11) You don't have any paper?
Instead it would be used to confirm a speaker's belief (presupposition)
that the addressee does not in fact have any paper. 8 Confusion may result
when an English-speaking learner of French is the recipient of a negative
declarative intonation question like (8 a). He or she may transfer native
language sociopragmatic knowledge and take it to be a confirmation of
the speaker's belief that he or she does not have the item in question
(which might be quite baffling) or, it may be interpreted as an abrupt,
somewhat impolite request (You d o n ~ t have any paper to give me? Well
you should!), rather than the polite request that it is in French.
Accordingly, if we are to teach students how to converse in a second
language we must teach them not only how to produce and understand
utterances in that language that are linguistically well-formed, but also
that are contextually felicitous. That is, students must learn to produce
utterances that are consistent with the discourse context and that employ
stylistic, register, and politeness features that are consistent with the
social context and the relative statuses and roles of participants. But
134 Michael L. Geis - Linda L. Harlow
before we do this, we must attempt to determine which forms in the first
and second languages are stylistic, politeness, and register equivalents.
Empirical studies produced by native speakers in context are necessary in
order to complement theoretical studies based on intuition (Blum-Kulka
- House - Kasper 1989: 3), for native speakers' intuitive understanding
of how speech acts function within their native languages have not always
been proven to reflect observed speech behaviors (Wolfson 1983: 83).
In our pilot study we were specifically interested in finding out which
syntactic patterns are employed in making requests by native French and
English speakers in an attempt to determine which features of context
(especially politeness features) seem most to govern utterance construc-
tion in the two languages in this context. We were then interested in
comparing utterances of native French speakers with those of English
speakers learning French to determine to what degree any differences
found could be attributed to sociolinguistic factors. Accordingly we were
less interested in what we might call "mechanical" errors of phonology,
morphology, and syntax than in differences of sociolinguistic interest,
especially cases of contextual infelicity.
In order to compare the productions of speakers of different lang-
uages it is necessary to have some means of determining when we have
cross-linguistic functional equivalents. In general, two utterances will be
what we might call "conversational equivalents" in two languages if they
have the same significance when used in the same context (S-Meaning).
We must therefore be clear on what it means for utterances to be
S-Meaning equivalents.
In many cases, especially when asking a straight-forward factual ques-
tion (What is the capital of Ohio?) or making a straight-forward as-
sertion (Columbus is the capital of Ohio), the significance or S-Meaning
of the utterance is essentially the same as its L-Meaning. Cases like these
will rarely present significant problems to language learners. The problem
arises in the case of utterances exhibiting significant stylistic, politeness,
or register influences. In what follows, we shall focus on one such class
of utterances, namely, utterances used in requesting that can be said to
have illocutionary significance. One way in which an utterance can have
illocutionary significance is by inquiring as to whether some condition on
a speech act is satisfied (Would you like a glass of water?) or asserting
that it is satisfied (I would like a glass of water). We shall refer to
illocutionary significance in this sense as "transactional illocutionary
significance". Another way in which an utterance can have illocutionary
significance is by doing"face work" (Goffman 1967), perhaps by redres-
Politeness strategies in French and English 135
sing a face-threat (Brown - Levinson 1978) associated with the action.
We shall refer to illocutionary significance in this sense as "interactional
significance". Let us briefly discuss these two notions.
As Searle (1969) observed, for a request to be felicitous, certain cond-
itions must be satisfied. His conditions on requesting (cf., (12)) will be
familiar to those who have read in the speech act literature.
(12) Searle's Conditions on Requests (H is the hearer and S is the
speaker)
Propositional Content Condition: Future act A of H.
Preparatory Conditions:
H is able to do A. S believes H is able to do A.9
It is not obvious to both Sand H that H will do A in the
normal course of events of his own accord.
Sincerity Condition: S wants H to do A.
Essential Condition: Counts as an attempt to get H to do A.
And, we would expect utterances used in requesting normally to address
these conditions in one way or another. Those that do can be said to have
transactional illocutionary significance.
Geis (1995) has argued that in order to apply speech act
theory to the analysis of how we do requesting in conversation,
including, in particular, how we do requesting in multi-turn inter-
actions like (1), it is necessary to revise Searle's speech act theory in
one quite critical way. Instead of saying what speech act a specific
utterance performs, a practice that sheds little light on multi-turn
interactions, we shall say what condition it addresses (if any). Geis also
argues that it is necessary to revise Searle's speech act structures in certain
ways.
Since multi-turn conversational sequences do not have a (single) pro-
positional content, we must abandon the propositional content condition
in favor of a statement of the domain of the request, which consists of a
specification of the properties of some requested thing or action. A
second major revision concerns Searle's sincerity condition. In some
cases, it identifies the psychological state from which the initiator of the
request launches an action, as is true of requests, but in other cases it does
not. Thus, the sincerity condition on asserting something is that the
speaker believe that what is being asserted is true, rather than that the
speaker desires to cause the addressee to believe that the assertion is true.
We shall adopt the view that speech acts 10 are subject to an initial state
136 Michael L. Geis - Linda L. Harlow
condition that always specifies the psychological state from which the
initiator initiates the action.
Searle's preparatory conditions are defective in a different way. Note
that Searle includes both a condition that the hearer be able to perform
the desired action and that the speaker believe the hearer to be able to
perform the action. The former condition is what we might call a "satis-
faction" condition, for it is concerned with speech act success; the latter
is a "felicity" condition in that it is concerned more with speech act
appropriateness. We shall adopt the view that speech act structures
should be restricted to satisfaction conditions, with the issue of felicity
being reserved to the issue of utterance planning (i. e., to the issue whether
or not the initiator of an act should initiate the act). Additionally, as is
common in speech act theoretic work, we shall add to the ability condi-
tion identified by Searle, a willingness condition stipulating that the
responder must be willing to perform an action before the request can be
successful.
One last revision is in order. Searle's essential condition identifies the
desired effect of a request - to cause the addressee to perform some
action. However, the actions we perform have other effects as well, in-
cluding, in particular, what we might call the interactional effects of the
acts. Brown and Levinson (1978) have argued that people have two face-
wants, a negative-face want not to have one's freedom of action impeded
and a positive-face want to be valued and to have what one values be
valued. They also argue that many types of speech acts are face threaten-
ing (FTA), and that when performing an FTA we have the option to
redress the face threat. Thus, requesting someone to do something
threatens their negative face. It also threatens the initiator's positive face
because of the possibility that the request be denied. We shall refer to the
face threats associated with an act as the interactional effects of the act.
We may now summarize our revisions of Searle's statement of the cond-
itions on requesting, where "Init" is the initiator 11 (i. e., the beneficiary)
of the request and "Resp" is the responder, and the action"A" involves
the responder's transferring some object to the initiator.
(13) Init Requests Resp to Do A
Initial State Condition: Init desires that Resp do A C 1
Illocutionary Effects:
Transactional Effect: To cause Resp to do A.
Interactional Effect: A threatens Resp's negative face and
Init's positive face
Politeness strategies in French and English 137
Satisfaction Condition:
Resp is able to do A C
2
Subcondition:
Res!) Possess X C
3
Resp is :J:illing to do A C
4
Domain (A):' Resp transfers X to Init Domain (X): ..... 12
The basic idea is thaf any utterance that occurs in a request sequence,
however short or long'it may be, will be expected to address one or more
elements of such a ~ t r u c t u r e . Utterances in different languages that
address the same elements of such a structure in the same way (including
interactionally) will be: said to be conversational equivalents.
Brown and Levinson (1978: 273) noted that theories of indirect speech
acts "appear to have lpissed the extremely systematic way in which the
rational strategies of face redress, like pessimism and hedging, are able to
predict the internal structure of polite indirect requests." Brown and
Levinson do not provJlde a formal account of how what we might call
"politeness features" determine the internal structures of utterances.
What we shall argue is that the internal structures of utterances involved
in requesting, like those in (7), are a compositional function of what
the speaker wishes to communicate - which elements of speech act
structures the utterance addresses - and politeness features. We shall
restrict our attention here to the morphological, syntactic, and lexical
properties of utterances.
What politeness features operate in English and French is a largely
open question. We shall assume the system in (14).
(14) A. Orientation Features: determine the subject of the utterance.
1) Init-oriented. For cases of utterances instantiating the initial-
state condition, of which there are two subtypes in English.
a) [nit-specific. Utterances that are desire-specific will be
those like [ want a corner piece, [ would like a corner
piece, and [ need a corner piece, that fairly literally
instantiate the initial-state condition. They are egocentric
if the initiator is speaking (and are therefore not very
polite) and are exocentric if the responder is speaking
(Do you want a corner piece?)
b) Object-specific. Utterances (which in English employ
either have or take) which communicate that the speaker
is selecting something from available alternatives. On this
138 Michael L. Geis - Linda L. Harlow
view, ['11 have a hot chocolate. S-Means "I hereby select a
hot chocolate." These forms are also egocentric if uttered
by the initiator and are not considered especially polite.
c) Object-only. This feature is used for utterances with no
subject (cf., How about a corner piece?). We do not have
a good understanding of this class of cases. We treat these
forms as init-oriented because they are restricted in use to
identifying what the initiator desires.
2) Resp-oriented. These are utterances that instantiate either
the willingness condition or the ability condition. Resp-
willingness forms are more polite than resp-ability forms, all
other things being equal, but are a problem for the initiator
in that a resp-willingness rejection (I won't do that) more
greatly threatens the initiator's positive face than does a
resp-ability rejection ([ can't do that).
a) Resp-ability. These are utterances that address the ability
satisfaction condition and are relatively polite if uttered
by the initiator - cf., Can you give me that corner piece?
b) Resp-willingness. These are utterances that address the
willingness satisfaction condition and are relatively polite
if uttered by the initiator. Imperatives like Give me that
corner piece will be treated as nondeferential resp-
willingness utterances (as is consistent with the traditional
view that they have understood you subjects and are
volitional - cf., Give me that corner piece, won't you?)
3) Domain-oriented. This is for utterances that address condi-
tion C
3
such as Do you have a corner piece?
B. Power and Deference Features:
1) Power features. These forms are involved in lexical selection
for modal verbs and main verbs.
a) [nit-up. This is for utterance forms that suggest that the
initiator enjoys greater social power over the responder.
It is for contrasts between cases like ['11 take a corner
piece (init-up) versus ['11 have a corner piece (no-init-up)
or [ want a corner piece (init-up) versus [ need a corner
piece (no-init-up).13
b) No-init-up. This feature pairs with"init-up" to define the
contrasts just discussed.
2) Deference Proper. These features determine sentence type in
concert with other features.
Politeness strategies in French and English 139
a) Deference. (Can I have a corner piece?:J Would you give me
a corner piece?)
b) No-Deference. (I:Jll have a corner piece:J Give me a corner
piece)
These features do not exhaust the set of politeness features, nor are we
fully confident of their being correct.
Now, if we wanted to teach someone to use English felicitously, we
could try to pair up sentences with contexts with instructions like "Say
such-and-such type of sentence if you are in a context having property C
and have social relationship of type R with your addressee by way of
making a request for something of type X." However, given the rather
large array of utterance types available,14 this would be a daunting task.
We believe, instead, that it would make more sense to teach students
what sociopragmatic distinctions are made in a language culture and how
to map sociopragmatic features into morphosyntactic and prosodic
features. One might tell students of English that in such-and-such
context, one needs to show deference to an addressee and one should
therefore use an interrogative sentence. Or that in another context, one
must show deference and have an ability-condition-specific, addressee-
oriented utterance, the result being an interrogative sentence with you as
the subject and can or could as the modal (Can you give me a corner
piece). Obviously, if we are to provide instruction like this, we must find
out what sociopragmatic features are operative in the native language
and in the target language and what their morphosyntactic and prosodic
implications are.
2. Pilot study
In this chapter, we report on a pilot study we have done on how
native French- and English-speaking subjects do requesting in a specific
experimental context and how English-speaking learners of French do
requesting in French in the same context. Our objective was to determine
what politeness features seem to be at work in the two language cultures
and how the various politeness features found are realized morpho-
syntactically and prosodically in the two languages.
Three groups of subjects participated in the experiment: 26 pairs of
French native speakers from the University of Nantes, who were spending
140 Michael L. Geis - Linda L. Harlow
a quarter at The Ohio State University (OSU) to complete a certificate in
the Business program, 23 pairs of English native speakers drawn largely
from an introductory linguistics course at OSU, and 27 pairs of English-
speaking learners of French enrolled in intermediate conversation courses
at OSU.15
For each run of the experiment, two subjects within the same language
group were seated at opposite ends of a small table. Each subject was
asked to put together a children's jigsaw puzzle. Each was given a set of
pieces to his or her puzzle (at the subject's right hand side) and a set of
pieces to his or her partner's puzzle (at the subject's left hand side).
A "starter piece" was set before each subject. Two pieces to each puzzle
were hidden in a box to be retrieved when it became clear that neither
subject held the piece. In one experimental context, a barrier was placed
between the subjects to make it impossible for subjects to see the pieces
controlled by the other subject. This was done to encourage multi-turn
sequences (as was our hiding two pieces of each puzzle from view).
Subjects were instructed to assist each other by handing over needed
pieces, subject to the constraint that the exchange had to result from
some sort of verbal interaction. (We were careful not to use the words
request or offer or other language (e.g., ask your partner for a piece) in
our instructions that might bias the specific language forms subjects
used.) In general, subjects seemed to find the task to be fun, and relatively
colloquial language resulted. All experimental runs were recorded on
videotape and audiotape and were transcribed, including relevant non-
verbal behavior.
3. Results and discussion
After completion of the experiment, ten experimental runs from each
group of subjects were randomly selected for analysis, with the only
stipulation being that within each language group, there would be ap-
proximately ten females and ten males, divided equally between the
barrier and non-barrier conditions. Demographic data on the 60 paired
subjects (see Table 1
16
) revealed that the groups were evenly matched for
sex. Although the age of the subjects ranged from 18-69, each group was
fairly even in age, considering that the median age for each group was 21
and that 80 % of the English group, 90 % of the French learners group,
and 100 % of the French natives were in the age range of 18-29.
Politeness strategies in French and English 141
Table 1.
Demographic data English French French
Learners
Sex 2 FF 2FF 3FF
2MM 2MM 2MM
6FM 6FM 5FM
Age Range 18-69 20-22 18-45
Mean Age 28 21 23
Median Age 21 21 21
F = Female; M = Male
It was a given of this experiment that subjects were to assist each
other. Put in satisfaction condition terms, condition C
4
on requesting
was understood to be satisfied or true at the outset. One of the
predictions we made, therefore, was that there would be relatively few
interrogative utterances that instantiate this condition (i. e., whose
S-Meaning can be expressed in terms of this condition). (If a condition
is understood to be satisfied, there is no need to address it further.)
In fact, of the 245 English request initiations (105 in the barrier-down
(NB) condition and 139 in the barrier-up (B) condition), we find only
those in (15).
(15) No Barrier
Fl: Will you hand me that piece that's closest to you?
F2: ((Points))
Fl: Right there. Yeah
F2: ((Laughs and passes piece)) Let's see here.
Fl: And that corner piece. ((Points))
F2: ((Picks up piece))
Fl: Umhum
We counted this as two instances of a willingness-condition specific
utterance, but it should be clear that these did not represent genuinely
independent instantiations of this condition. Among the French request
initiations, only 7 out of the 243 spoke to the willingness of the addressee
142 Michael L. Geis - Linda L. Harlow
to hand over pieces. These were all made by the same speaker within the
same conversation, and were of the variety found in (16).
(16) Tu veux me passer celui ou il yale plus de beige, enfin, d'ocre?
Celui-la. Voila, exactement ~ a .
'Do you want to pass me the one where there is the most beige,
well, dark yellow? That one. That's it, that's it exactly.'
On the other hand, imperative utterances, which, as we noted, we are
treating as no-deference, resp-willingness forms, can occur when the
willingness of the addressee to perform the desired action can be
assumed. Indeed, this is a precondition on their use.
We were interested in evoking two kinds of utterances - utterances
employed to identify whether or not some desired piece was held by the
other party and utterances employed to indicate a desire for some piece.
What we found was that there was variation in the types of utterances
used in both circumstances. Thus, in cases in which the existence of a
desired piece was unknown (sometimes because the initiator was unclear
how to describe it), we would get domain-oriented forms like Do you
have a corner piece? (English) or Tu n'as pas de coin? (French), and
variants thereof (utterances addressing condition C
3
of the ability cond-
ition C
2
) but we also got utterances addressing other conditions a.s well,
such as, init-oriented forms like I need a corner piece (English) or ]'ai
besoin d'un coin (French). Thus, even in this case subjects made poli-
teness choices in how they framed these request initiations.
In circumstances in which the initiator could easily identify a desired
piece (as when the barrier was down), it would be apparent that the
responder had the piece (C
3
was true) and therefore the responder was
able to provide the piece (C
2
was true). Since C
4
was given as true, all the
responder-specific satisfaction conditions would be true. At such a point,
the only thing that is not known is what specific piece the initiator wants.
It is at that point that virtually all request forms (all of the forms of (7),
and more) become available to the speaker and speakers are forced to
make politeness choices. This corresponds to a real world situation such
as we find in a frozen yoghurt store where the sizes and flavors offered
and the prices charged are posted in easy view. In such a circumstance the
only thing that is not known is what the customer wants and, therefore,
any of the forms of (7), and more, become available.
We organized the data obtained according to the orientation and
deference choices the speaker made, with the orientation features
Politeness strategies in French and English 143
representing the highest level of organization (since they equate to the
satisfaction condition the utterance addresses). In what follows, we
provide a summary of our findings.
17
In all cases, we give forms uttered
by initiators, for we were interested in requesting, not offering.
I. Init-oriented forms (Utterances addressing the initial-state condition
C
1
) - See Table 2.
18
A. Init-specific forms:
19
I need a corner ai besoin coin 20
B. Object-specific: Forms (whose S-Meanings are "I hereby select X")
are restricted in use to cases in which the initiator can see that the res-
ponder has the desired piece.
1. No-init-up forms:
a. No-deference: I'll have that corner piece/(No French Equivalent)
b. Deference: Can I have a corner piece?/Est-ce que je peux avoir un coin?
2. Init-up forms:
a. No-deference: I'll take a corner piece/Ie prendrai un coin
b. Deference: Can I take a corner piece?/Est-ce que je peux prendre un
coin?
Table 2.
I. Init-Oriented forms
A 41 39 80 (32.7) 40 28 68 (28)
B.l.a 0 0 o (0) 0 0 o (0)
B.l.b 1 20 21 (8.6) 0 1 1 (0.4)
B.2.a 0 2 2 (0.8) 1 0 1 (0.4)
B.2.b 0 0 o (0) 0 0 o (0)
Totals 42 61 103 (42.1) 41 29 70 (28.8)
English
B NB T(%)
French
B NB T(%)
French Learners
B NB T(%)
57 11 68 (28)
0 0 o (0)
0 9 9 (3.7)
0 0 o (0)
0 0 o (0)
57 20 77 (31.7)
B = Barrier; NB = No Barrier; T = Total
A = init-specific forms
B.l.a = Object-specific, no init-up, no deference
B.l.b = Object-specific, no init-up, deference
B.2.a = Object-specific, init-up, no deference
B.2.b = Object-specific, init-up, deference
144 Michael L. Geis - Linda L. Harlow
The primary init-oriented forms are the LA forms (cf., Table 2), for they
employ some verb that means "desire" or "need," which is cognate with
the predicate of the initial-state condition. We find init-specific forms in
both the "barrier up" (B) and "no-barrier" (NB) conditions. One major
difference between English and French is that what we are calling object-
specific, init-oriented forms (cf., I. B) were virtually never used by French
natives in our experiment.
II. Resp-oriented forms (Utterances addressing the willingness and
ability conditions.)
A. Resp-willingness.
1. No-deference: Give/Hand me a corner piece/Donne/Passe-moi un coin
2. Deference: Will you give me a corner piece?/Tu me passeras un coin?
B. Resp-ability.
1. No-deference: You could give me a corner piece (now)/Tu peux/
pourrais me donner un coin
2. Deference: Could you give me a corner piece?/Est-ce que tu
peux/pourrais me donner un coin?
There was a striking contrast between the English and French speakers
with respect to the use of imperative utterances (see Table 3). Over 100/0
Table 3
II. Resp-oriented forms
English French French Learners
B NB T(%) B NB T(%) B NB T(%)
A.l 0 1 1 (0.4) 2 24 26 (10.7) 0 6 6 (2.5)
A.2 0 2 2 (0.8) 0 7 7 (2.9) 0 1 1 (0.4)
B.l 0 0 o(0) 0 0 0(0) 0 0 o(0)
B.2 0 5 5 (2) 3 9 12 (4.9) 0 7 7 (2.9)
Totals 0 8 8 (3.2) 5 40 45 (28.5) 0 14 14 (5.8)
B =Barrier; NB =No Barrier; T =Total
A.l = Resp-willingness, no deference
A.2 =Resp-willingness, deference
B.l =Resp-ability, no deference
B.2 =Resp-ability, deference
Politeness strategies in French and English 145
of the French natives' requests were in the imperative form, compared
with just 0.4 % (1 request) of the English speakers. It seems that in an
informal situation such as the one in the current study (helping someone
perform some task in an informal atmosphere), using the direct form of
the imperative is not considered impolite in French. The addition of
lexical softeners such as s:lil te plait ('please'), tiens ('well'), and done
('so') in more than one third of the imperative requests also served to
reduce the abruptness of the imperative form. The fact that the French
learners patterned themselves more after the American English speakers
probably reflects the English taboo against the use of the imperative. This
supports research reported by Kellerman (1979; 1983) which indicates
that error due to interference from the native language is more likely the
less the perceived distance between the target language and the native
language. It is difficult to extrapolate the use of the imperative to a large
number of other French contexts without further study, but certainly
teachers could encourage students to use it in similar informal contexts,
particularly in conjunction with lexical softeners.
There were no particularly striking differences among the language
groups in regard to resp-ability utterances, utterances addressing condi-
tion C
2
They were relatively rare, we think, largely because utterances
addressing its precondition, C
3
, were so common, as we shall shortly see.
A positive response to C
3
utterances implies a positive response to C
2
,
given the conditions of the experiment, and so such utterances will
diminish in frequency as C
3
utterances increase (and conversely, of
course).
The last class of utterances we shall consider are domain-oriented
utterances, so-called because they inquire as to whether some desired
thing is available (in requests for things) or whether someone is available
to do something (in requests for actions). In English, the principle
domain-oriented utterances are Yes-No-Questions like those in (2)
employing have in the case of requests for things and questions like You
doing anything? or Whateha doing?, and the like, for requests for actions.
As we noticed earlier, French domain-oriented utterances commonly
consist of positive and negative declarative intonation questions, as well
as inverted questions like those used in English. Negative declarative
intonation questions seem to be more polite than the various types of
semantically positive French question forms. The explanation for this
seems to be that they are being used to confirm a speaker presupposition
that the addressee does not have the desired thing and therefore provide
the addressee with a "built-in" excuse for rejecting the request. In Table
146 Michael L. Geis - Linda L. Harlow
4, we combine French Est-ce que questions, declarative intonation
questions, and interrogative (inverted) sentences in row A.1 for they
either do not presuppose the existence of the desired thing or express a
positive presupposition. They are contrasted, then, with the negative
declarative intonation questions (A.2), which convey a negative pre-
supposition. We then go on to show the distribution of the various
question forms in Table 5.
III. Domain-oriented forms. (Utterances addressing Subcondition C
3
of
the ability condition.) - See Table 4.
A. Deference:
1) No-Presupposition/Pos-Presupposition: Do you have a corner
piece?/Tu as un coin?, Est-ce que tu as un coin?, As-tu un coin?
2) Neg-Presupposition: You wouldn't have a corner piece?/Tu n'as pas
de coin? 21
B. No-Deference: You (must) have a corner piece/Tu dais avoir un coin.
As Table 4 indicates, the contrasts between English and French in regard
to C
3
utterances were quite striking. As noted in Table 5 and in connec-
tion with examples (8) and (11), French speakers rarely use the inverted
interrogative form in making information questions, preferring, instead,
to use Est-ce que questions or either positive (Tu as ... ?) or negative (Tu
n'as pas ... ?) declarative intonation questions. Interestingly, the French
Table 4.
III. Domain-Oriented Forms
English French French Learners
B NB T(%) B NB T(%) B NB T(%)
A.1 51 17 68 (27.8) 41 12 53 (21.8) 86 46 132 (54.3)
A.2 0 8 8 (3.3) 32 24 56 (23) 1 1 2 (0.8)
B 0 0 0 (0) 0 9 9 (3.7) 0 1 1 (0.4)
Totals 51 25 76 (31.1) 73 45 118 (48.5) 87 48 135 (55.5)
B =Barrier; NB =No Barrier; T =Total
A.l = Deference, no-presupposition/possible presupposition
A.2 = Deference, negative-presupposition
B =No deference
Politeness strategies in French and English 147
Table 5.
French question forms
French Natives French Learners
B NB T (%) B NB T (%)
Est-ce que tu as ... 43 20 63 (47.4) 50 42 92 (59.0)
Tu as ... 33 36 69 (51.9) 3 2 5 (3.2)
As-tu ... 1 0 1 (0.8) 39 20 59 (37.8)
Totals 77 56 133 (100) 92 64 156 (100)
B =Barrier; NB =No Barrier; T =Total
were just as likely to inquire about the nonexistence of a puzzle piece than
the affirmative existence of it. We speculate that in using the negative
form, French speakers were providing a built-in excuse for the addressee
to not provide the piece, that is to say, a face-saving device for both
speakers even though the speaker clearly expected to receive the piece.
Quite strikingly, the French learners virtually exclusively employed
Est-ce que questions and inverted interrogative forms, almost totally
ignoring positive declarative intonation forms and both negative Est-ce
que and negative declarative intonation forms. We offer the following
explanation for the learners' behavior. Regarding the excessive use of
inversion for question format, we feel that this may be due to a special
emphasis on it in the language classroom. Since intonation per se is so
easy to use, teachers tend to concentrate their teaching efforts on Est-ce
que questions. They also focus on the inverted forms for they are
grammatically more difficult. In addition, studies have found that French
textbooks frequently do not clarify sufficiently differences in style and
frequency of use regarding Est-ce que, intonation, and inversion (Di Vito,
1991; 1992; Walz 1986). Even if inversion is described as useful in
primarily written contexts, most textbooks have exercises that require
students to practice using inversion for forming questions in spoken con-
texts as well. To make matters worse, inversion is often used in textbooks
to provide directions for oral and written exercises and activities.
Although the use is correct since the textbook represents written language
and students do need to become familiar with the form, it may nonethe-
less give students the impression that inversion is used more frequently
than it really is.
148 Michael L. Geis - Linda L. Harlow
Moreover, English speakers usually do not use declarative intonation
questions as "ordinary" questions. They tend to be relatively context-
bound (e.g., used when the speaker believes that the propositional
content expressed by the question may be false or is surprising - You
saw John? How odd. I thought he was in Chile!) Thus, it seems that
learners were transferring their native language speech act strategies in
this case and thereby avoided the commonly used intonation pattern
preferred by the French.
With regard to the lack of negative request forms, English speakers do
not employ negative declarative intonation questions simply to request
information (or, as in this experiment, to request things). Instead they are
employed to confirm a speaker belief that the propositional content
expressed by the proposition is true. (You didn't see John? He was here!
I saw him several times!) Moreover, as Fraser and Nolen (1981: 103)
point out, negative interrogatives are perceived as less deferential by
Americans than positive interrogatives, a view that we can expect would
be carried over to a second language.
4. Conclusions / Implications
Our study of American English, French, and French learner requesting
behavior is unusual, we believe, in that it was based on how each
language group did requesting in precisely the same experimental
conversational context. This method gives us a reliable measure, we
believe, of what counts as French and English functional equivalents for
doing requesting in relatively informal conversational contexts, and a
means of assessing possible sociopragmatic interference in language
learning by our French learners. Further research is needed to confirm or
revise our findings in this pilot study. Larger sample sizes, different con-
texts, and eventually comparisons between English and French and other
languages will shed further light on the nature of pragmatic language
interference.
There is an important theoretical conclusion to be drawn from this
study, and this is that different language cultures can vary along two
dimensions sociopragmatically. First, different language cultures may
differ in the sociopragmatic distinctions they make. Thus, whereas in
English, colloquial domain-oriented utterances tend to be restricted to
positive inverted forms like Do you have a corner piece? and stylistic
Politeness strategies in French and English 149
variations thereof (cf., our discussion of example (1)), the French have
three forms: inverted questions, positive and negative declarative intona-
tion questions, and Est-ce que questions. This strongly suggests that the
French language culture makes more sociopragmatic distinctions than
does the American English language culture, with stylistic and politeness
considerations each playing a role. Second, even when two language
cultures make the same politeness distinctions, the forms they use in a
particular context may be quite different syntactically and semantically.
We are inclined to think, for instance, that the French declarative intona-
tion question, Tu as un coin?, is the closest politeness equivalent to the
English form, Do you have a corner piece?, though our study is mute on
the point.
The results of this study suggest several implications for the teaching
of French. First, although questions formed by inversion need to be
taught in the classroom along with Est-ce que and intonation, the data in
this study suggest that the importance of the use of simple intonation
questions in conversation is not at all clear to the learner and needs to be
made so. Language textbooks need to reduce the amount of inverted
forms used, clarify the limited use of inversion, and refrain from obliging
students to use inversion in oral activities.
Second, in addition to engaging in a certain amount of sociopragmatic
teaching that stresses that the declarative intonation question is one of the
most commonly used French devices for requesting in informal contexts,
the data suggest that teachers stress that the negative intonation form is
especially polite because it allows the addressee a face-saving basis for
rejection of the request in that it provides a built-in excuse because of
the negative presupposition it communicates (You don't have a corner
piece -, and so, therefore you can't pass it over.) This is as true, of
course, of negative Est-ce que questions as of negative declarative
intonation questions. In teaching the negative declarative intonation,
teachers might note that they seem to parallel in use the English
form You wouldn't have a green piece?, which, as line (A.2) of
Table 4 shows, does occur in English in roughly the same context as the
French negative declarative.
Third, the imperative form, in conjunction with the use of lexical
softeners, was preferred in over 10 % of the French natives' requests and
yet the French learners eschewed the imperative form, perhaps because
there is a taboo against using it in English in most circumstances. This
suggests that learners should be taught that the imperative form is not as
impolite in French as it is in English, at least in some informal contexts.
150 Michael L. Geis - Linda L. Harlow
Further studies will indicate whether the imperative form can be applied
appropriately in polite requests in other contexts.
Finally, this study, has attempted to add to the growing body of
research into sociopragmatic choices available to speakers of languages. It
is becoming more and more clear that teachers need to focus their efforts
actively not only on teaching contrasting linguistic structures in English and
the second language, but also on how to make the correct sociopragmatic
choices in conversation (the correct style, politeness, and register choices).
In so doing, language teachers will greatly facilitate learners' successful
interactions with native speakers of the second language.
Notes
1. We have actually carried out this experiment informally in a few stores to see
what would happen. The answers, so far, have all been negative, except for
one case in which the clerk said, Do you mean "ever?"
2. Note, for instance, that while any of the utterance-types in (2) and (3) might
be used between friends in requesting a beer (Do you have any beer?), the
utterances of (4) could not (Do you carry beer?).
3. We would expect that these utterances could have different significances or
S-Meanings were they to be uttered in different contexts.
4. We shall take the position here that there are just two basic things that
utterances do: provide information and request information. Such actions as
requesting, suggesting, inviting are treated as being epiphenomenal in
character - as the product of implicature (Grice 1975), if one likes.
5. The notion "language culture" is defined as a pattern of social expectations
associated with a particular, homogeneous subgroup of speakers of a
language. Such a notion is required if we are to understand how different
subgroups of those speaking a single language differ in how they use
language.
6. Since "intonation questions" of the sort that we find in French are
declarative rather than interrogative in form, in that they do not exhibit
inversion, we shall refer to them as "declarative intonation questions."
7. The"?" at the end of this and other declarative (in form) intonation questions
in French and English is intended to represent "question intonation."
8. In French one can also use the negative interrogative form in a presupposi-
tion-checking manner. The difference in the languages is that French, unlike
English, makes this form available as a polite form.
9. Searle's providing alternate characterizations of this condition - one in
- "objective" terms and one in terms of speaker beliefs - will not do. Geis
(1995) argues that felicity conditions should be stated objectively, with the
issue of speaker beliefs arising only in utterance planning.
10. Geis (1995) demonstrates that so-called speech acts are social, as opposed to
linguistic, actions in that they can commonly be performed nonverbally and
Politeness strategies in French and English 151
the factors that distinguish them are invariably social or psychological, ra-
ther than linguistic in nature. With the understanding that speech acts are
actually communicative social actions, not linguistic actions, there is no
harm in continuing to use the older term.
11. The initiator of an act is the one whose desires or needs are the cause of the
initiation of the act and is not to be equated with the person who speaks first.
We would say that the cases of a customer who goes to a store to purchase
something (an attempt to buy) and of a door-to-door salesperson who tries
to sell something to a customer (an attempt to sell) are speech acts theoret-
ically equivalent in that in both cases the buying-selling actions will be
successful just in case the initiator (buyer) comes to desire the thing being
offered for sale and the responder is able and willing to sell it, that is, has the
thing and finds the price the initiator is willing to pay acceptable.
12. Here would be spelled out the properties of the thing desired. If the initiator
wishes to purchase one pint jar of blackberry jam, the Domain would be as
in (1).
(1) Domain(x): blackberry-jam(x) & size(x, pint) and number(x, one)
13. Note that people who are socially inferior to others (children versus parents
or employees versus employers) tend to use the less "pushy" forms I need a
new bike/I need a raise over the more pushy forms I want a new bike/I want
a raise, whereas superiors have no difficulty using the pushy forms I want
you to clean up your room/I want you to finish the report before you go
home.
14. The examples of (7) barely touch the surface of the problem.
15. A minimum of 5 quarters of French study were a prerequisite for entering the
course. The average student had taken the equivalent of 8 or 9 quarter
courses prior to this one.
16. Only 5 FM pairs in the learner group participated in the entire experiment.
Thus, an additional pair of FF had to be used.
17. The percentages given in the tables below are based on 245 English, 243
French, and 245 French learner request initiations.
18. In the tables that follow, we shall use the roman numerals employed in this
classification to identify forms.
19. The deference/no-deference contrast does not apply to init-specific
utterances, for we never make requests saying, Do I want a hot chocolate?
20. In what follows, we collapse all instances of an instantiation of a particular
condition into a single type. So, in the case of (LA), we collapse I
need/want/would like P together and, looking ahead to (LB.1.b), collapse
such variants as Could/Can/May I have P into a single type of utterance.
Obviously, there are politeness differences among these variants.
21. We place the English conditional negative declarative intonation question
You wouldn't have a corner piece? alongside the French simple negative
declarative form. It must be noted that these are not literal equivalents.
Instead, they are functional (politeness) equivalents.
152 Michael L. Geis - Linda L. Harlow
References
Blum-Kulka, Shoshana - Juliane House - Gabriele Kasper
1989 "Investigating cross-cultural pragmatics: An introductory over-
view", in: Shoshana Blum-Kulka - Juliane House - Gabriele
Kasper (eds.), 1-34.
Blum-Kulka, Shoshana - Juliane House - Gabriele Kasper (eds.)
1989 Cross-Cultural pragmatics: Requests and apologies. Norwood, NJ:
Ablex.
Brown, Penelope - Stephen C. Levinson
1978 "Universals in language usage: Politeness phenomena", in: Esther
Goody, (ed.), 56-311.
Desirat, Claude - Tristan Harde
1988 La langue franfaise au 20e siecle. Paris: Bordas.
Cole, Peter - Jerry L. Morgan (eds.)
1975 Syntax and semantics III: Speech acts. New York: Academic Press.
Di Vito, Nadine O.
1991 "Incorporating native-speaker norms in second language materials",
Applied Linguistics 12: 383-396.
1992 "Sensitizing teaching assistants to native-speaker norms in the
communicative classroom", in: Joel C. Walz (ed.), 171-189.
Fraser, Bruce - William Nolen
1981 "The association of deference with linguistic form", International
Journal of the Sociology of Language 17: 93-109.
Gadet, Fr a n ~ o i s e
1989 Le franfais ordinaire. Paris: Armand Colin.
Gass, Susan - Larry Selinker (eds.)
1983 Language transfer in language learning. Rowley, MA: Newbury
House.
Geis, Michael L.
1995 Speech acts and social actions: Toward a theory of conversational
competence. Cambridge: Cambridge University Press.
Goffman, Ervin
1967 Interaction ritual: Essays on face-to-face behavior. Garden City, NY:
Doubleday
Goody, Esther (ed.)
1978 Questions and politeness: Strategies in social interaction.
Cambridge: Cambridge University Press.
Grice, H. Paul
1975 "Logic and conversation", in: Peter Cole - Jerry L. Morgan (eds.),
41-58.
Kellerman, Eric
1979 "Transfer and non-transfer: Where we are now", Studies in Second
Language Acquisition 2: 37-57.
1983 "Now you see it, now you don't", in: Susan M. Gass - Larry
Selinker (eds.), 112-134.
Politeness strategies in French and English 153
Searle, John R.
1969 Speech acts: An essay in the philosophy of language. Cambridge:
Cambridge University Press.
Valdman, Albert
1967 "Norme pdagogique: Les structures interrogatives du
International Review of Applied Linguistics 5: 3-10.
Walz, Joel C.
1986 "Is oral proficiency possible with today's French textbooks?",
Modern Language Journal 70: 13-20.
Walz, Joel C. (ed.)
1992 Development and supervision of teaching assistants in foreign
languages. Boston: Heinle & Heinle.
Wolfson, Nessa
1983 "An empirically based analysis of complimenting in American
English", in: Nessa Wolfson - Elliot Judd (eds.), 82.-95.
Wolfson, Nessa - Elliot Judd (eds.)
1983 Sociolinguistics and language acquisition. Rowley, MA: Newbury
House.
Transfer and proficiency in interlanguage
apologizing
Naoko Maeshiba - Naoko Yoshinaga - Gabriele Kasper
- Steven Ross
1. Pragmatic transfer
The influence of non-native language users' linguistic and cultural back-
ground on their performance of linguistic action in a second language has
been a focal concern in interlanguage pragmatics (Kasper 1992, for
review). Transfer effects have been noted at the sociopragmatic and
pragmalinguistic level (cf., Leech 1983; Thomas 1983, for the distinction
between sociopragmatics and pragmalinguistics). Sociopragmatic transfer
has been found to operate in learners' perceptions of contextual factors,
such as interlocutors' relative social status (e.g., Beebe - Takahashi -
Uliss-Weltz 1990; Takahashi - Beebe 1993); assessment whether
carrying out a particular linguistic action is socially appropriate (e. g.,
Robinson 1992), and the overall politeness style adopted in an encounter
(e.g., Blum-Kulka 1982; Garcia 1989; Olshtain - Cohen 1989).
Pragmalinguistic transfer has been noted in learners' use of conventions
of means and form, affecting the illocutionary force and politeness value
of interlanguage utterances (e.g., House - Kasper 1987; Bodman -
Eisenstein 1988; House 1988; Beebe - Takahashi - Uliss-Weltz 1990).
Because of its potential for miscommunication, focus has been given to
negative transfer, the projection of first language-based sociopragmatic
and pragmalinguistic knowledge onto second language contexts where
such projections result in perceptions and behaviors different from those
of second language users. Yet positive transfer, the projection of first
language-based sociopragmatic and pragmalinguistic knowledge where
such projections result in perceptions and behaviors consistent with those
of second language users, have also been attested. For instance,
conventionally indirect forms for requesting (e. g., Blum-Kulka 1982;
House - Kasper 1987; Frerch - Kasper 1989; Takahashi - DuFon 1989)
were successfully transferred to English from Hebrew, German, Danish,
and Japanese. Meaning conventions were transferred in the apology
strategies used in German-English (House 1988) and Thai-English
156 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
(Bergman - Kasper 1993) interlanguage. In all of the above cases, we are
justified to assume that positive transfer from first language to second
language pragmatic knowledge was operative because the matching
patterns apply to specific pairs of native and target languages and
cultures and not to others. In most instances, however, it is difficult to
disentangle positive transfer from learners having recourse to universal
pragmatic knowledge and inferencing strategies (Blum-Kulka 1991).
In addition to ascertaining where pragmatic transfer occurs and
whether it leads to perceptions and behaviors divergent from or
consistent with second language users', interlanguage pragmaticists have
attempted to identify the conditions for transfer to occur, and the factors
which mediate its operation. Transferability constraints posited in the
literature include learners' psychotypology in the sense of Jordens (1977)
and Kellerman (1977), and their perceptions of sociopragmatic and
pragmalinguistic knowledge as specific for a given culture and language
or as culturally and linguistically "neutral". An example of differential
psychotypologies is seen in the request patterns of Danish learners, who
made freer use of their first language when requesting in German than in
English (House - Kasper 1987). Universal versus culture-specific percep-
tions distinguished the need to apologize as expressed by Russian and
English-speaking learners of Hebrew, the Russians perceiving the need to
apologize as determined by the nature of the committed act, whereas the
English-speaking learners made apologizing contingent on cultural con-
text (Olshtain 1983). Japanese female informants stated that refusing
offers, requests, or invitations was much more acceptable in American
than in Japanese society; hence they felt that transfer from their Japanese
norms of interaction would not be successful in an American context
(Robinson 1992). At the pragmalinguistic level, German learners of
English avoided the use of the mitigator 'I mean' because they considered
the German equivalent ich meine as language-specific (Kasper 1982). The
transferability of conventionally indirect request strategies from Japanese
to English was shown to be highly context-dependent, and varied with
learner factors such as proficiency and familiarity with the situation
(Takahashi 1992).
Non-structural factors interacting with pragmatic transfer include
learner-external factors such as learning context and length of residence
in the target community, and learner-internal factors such as attitude
towards the native and target community, and second language
proficiency. Takahashi and Beebe (1987) found that transfer of Japanese
refusal strategies, while occurring in the refusal patterns of English as a
Transfer and proficiency in interlanguage apologizing 157
second language and English as a foreign language learners, was more
prevalent in the English as a foreign language than in the English as a
second language learners' production. According to informants' self-
reports, their culture- and language-specific perceptions of refusal
strategies and pragmalinguistic function, noted by Robinson (1992) and
Kasper (1982) (see above), could partly be attributed to explicit teaching.
We are not aware of any studies which specifically examine the effect of
length of residence on pragmatic transfer. However, since some studies
suggest that length of stay influences second language pragmatic behavior
in a non-linear fashion (Olshtain - Blum-Kulka 1985; Blum-Kulka -
Olshtain 1986), it can be assumed that this factor has an impact on
pragmatic transfer also. What exactly this influence is remains to be
studied. A quantitative measure like amount of exposure alone or, for
that matter, proficiency cannot account for the fact that highly proficient
long-term residents often preserve some of their first language communi-
cative style, and even pass it on to the next generations of immigrants
(e.g., Clyne 1979; Blum-Kulka - Sheffer 1993). Such an "intercultural
style", seen, for instance, in American immigrants to Israel, hardly
reflects "deficient" second language communicative competence (Blum-
Kulka 1991). Rather, it appears to express language users' need for
disidentification, or maintaining their cultural identity as separate from
the community at large. Pragmatic divergence of this kind can best be
accounted for in an accommodation-theoretical framework (e.g., Giles -
Johnson 1987). Of the learner-internal factors, then, social-psychological
orientation is a potential determinant of pragmatic transfer. The other,
"cognitive" factor, second language proficiency, has been found to
constrain pragmatic transfer in requesting (Blum-Kulka 1982) and
apologizing (Olshtain - Cohen 1989). In both studies, it was found that
learners' limited second language knowledge prevented them from
transferring complex conventions of means and form from their first
language. These findings are thus consistent with Takahashi and Beebe's
(1987) hypothesis that second language proficiency is positively correlat-
ed with pragmatic transfer. While their own study on refusals performed
by Japanese learners of English at two different proficiency levels did not
demonstrate the predicted proficiency effect, Blum-Kulka's (1991) and
Olshtain and Cohen's (1989) studies support Takahashi and Beebe's view.
However, since these studies do not look at the performance of learners
at different proficiency levels, they do not provide conclusive evidence for
or against the effect of proficiency on transfer. In this chapter, therefore,
we shall put Takahashi and Beebe's transfer hypothesis to another test.
158 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
2. Native and non-native apology
Second only to requests, apologies are the next-best studied speech act in
descriptive, cross-cultural, and interlanguage pragmatics. This is so for
good reasons. In any speech community, participants need to be able to
engage in remedial verbal action upon committing an offense, that is, to
apologize. While the speech act of apologizing can thus be regarded as a
pragmatic universal, the conditions which call for apology are clearly not.
Speech communities differ in what counts as an offense, the severity of
the same offensive event, and appropriate compensation. These percep-
tions will in turn be mediated by social factors such as the interlocutors'
relative status and familiarity. Non-native speakers have to learn what the
specific conditions for apology are in the target community, what the
strategies and linguistic means are by which apology can be implemented,
and how to make contextually appropriate choices from the apology
speech act set.
The supposition of an apology speech act set is supported by a large
body of studies examining native and non-native speakers' apologizing
patterns. This notion, first proposed by Olshtain and Cohen (1983) and
empirically sustained in a series of studies by these authors (Olshtain
1983; 1989; Olshtain - Cohen 1989) entails that apologies can be
carried out by a finite set of "conventions of means", or strategies, all of
which are related to the offensive act and serve as the speaker's attempt
to "make it go away": either by conveying regret and proposing remedy,
or by diminishing the offense or the speaker's responsibility for it. Two
strategies, offering an explicit apology and assuming responsibility for
the offense, were found to be used in remedy of most offenses, irrespec-
tive of the specific contextual circumstances. The remaining strategies,
upgrading apologetic force, downgrading the severity of the offense or
speaker's responsibility, offering repair, and placating the offended party
by different kinds of verbal redress are clearly cross-culturally available,
yet their use is highly sensitive to contextual conditions, and subject to
cross-cultural variation (Olshtain 1989; Bergman - Kasper 1993).
Selections from the apology speech act set are determined by a variety
of context-internal and context-external factors. One of the context-
internal factors is the nature of the offense. Borkin and Reinhart (1978)
suggested that "excuse me" is used to remedy "a breach of etiquette or
other light infraction of a social rule", whereas "I'm sorry" is used as an
expression of dismay or regret about "a violation of another person's
right or damage to another person's feelings" (1978: 61). Their observa-
Transfer and proficiency in interlanguage apologizing 159
tion compares well to Goffman's (1971) distinction between ritual and
substantive apology.
Within the category of substantive apology, the offender's obligation
to apologize affects the choice of apologetic formula (House 1988) and
intensification of apologetic force (Cohen - Olshtain 1981; House 1988;
Vollmer - Olshtain 1989; Bergman - Kasper 1993). The factor that has
been shown to have the strongest effect on apology realization is the
severity of the infraction. Comparison of apology in Hebrew, Australian
English, Canadian French, and German with assessments of contextual
factors in different offense contexts suggests that "severity of offense is
the representative contextual factor in the socio-pragmatic set of the
apology" (Olshtain 1989: 160). In the case of significant injury or
inconvenience, Fraser (1980) observed a shift from the pattern apology +
account to apology + offer of compensation. Tanaka (1991) noted that
both native speakers of Japanese and of Australian English increased
apology intensification and formality of apologetic formulae with higher
offense severity. Holmes (1989) noted that severity of offense has a
differential impact on female and male offenders' use of apology
strategies. Female New Zealand offenders apologized most to light
offenses whereas men apologized most to infractions of medium severity.
Whereas native speakers of English intensified apologetic force dependent
on severity of offense, non-native speakers were found to take less
account of severity in their choice of apology intensification (Bergman -
Kasper 1993).
According to most studies, apology performance is affected by the
context-external factors social power and social distance. The lower the
offender's status vis-a-vis the offended person, the more the perpetrator
is prone to apologize by means of an explicit apologetic formula (Vollmer
- Olshtain 1989), intensify apologetic force (Fraser 1980; Olshtain
1989; Vollmer - Olshtain 1989), and choose a more formal apology
strategy (Cohen - Olshtain 1981; Olshtain - Cohen 1983). However,
Holmes (1989) found a non-linear relationship between social power and
apology in her New Zealand data: most apology was offered in equal
status relationships, lower status offenders apologizing second most, and
higher status offenders least frequently. In American and Thai apolo-
gizing, social power did not influence offender's selection of apology
strategy (Bergman - Kasper 1993). Barnlund and Yoshioka (1990)
found that Japanese offenders varied forms of apologizing more
according to participants' status than American perpetrators did in
comparable contexts.
160 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
The impact of social distance on apology behavior varies across
studies. Except for a limited tendency towards a negative correlation
between social distance and use of an explicit apology formula, Olshtain
(1989) did not establish any relationship between social distance and use
of apology strategy. Bergman and Kasper (1993) found that the closer the
interlocutors, the more likely the offender was to expressly assume
responsibility for the offensive act. This finding is contrary to that of
Wolfson, Marmor, and Jones (1989), in whose study most responsibility
was expressed between acquaintances and equally little at the two oppo-
site ends of the social distance continuum. This finding was interpreted as
further evidence in support of Wolfson's bulge hypothesis (1989).
However, neither Wolfson et al. (1989) nor Bergman and Kasper's (1993)
investigation include intimate interlocutor relationships. Hence neither of
the two studies has demonstrated evidence for or against the bulge
hypothesis.
Studies of interlanguage .apologies include the language pairs Hebrew
first language-English second language (Cohen - Olshtain 1981),
English and Russian first language-Hebrew second language (Olshtain
1983), Danish first language-English second language (Trosborg 1987),
German first language-English second language (House 1988), Spanish
first language-English second language (Garcia 1989), and Thai first
language-English second language (Bergman - Kasper 1993). Of these,
only Trosborg's (1987) study examined proficiency effects on learners'
performance of apology, and therefore had potential implications for a
theory of pragmatic development in adult second language learners.
However, the only developmental effect she found was that the use of
modality markers increased with higher proficiency. It is difficult to say
whether this pattern truly reflects a development of pragmalinguistic
competence or merely an extension of the learners' lexical repertoire. The
learners' strategy use differed in some respects from that of the native
speakers of Danish (first language) and British English (second language),
but there were no proficiency effects on the learners' use of apology
strategy.
Previous research has offered descriptive accounts of transfer and
proficiency in interlanguage users' speech act performance. Considering
the effects of contextual factors on strategy selection as reported in the
literature, it seems plausible to assume a relationship between contextual
factors and transfer of apology strategies. This study will therefore
examine whether pragmatic transfer is constrained by contextual factors,
and whether it is affected by learners' proficiency level.
Transfer and proficiency in interlanguage apologizing 161
3. Method
3.1. Subjects
Four groups of subjects participated in this study:
1) 30 Japanese learners of English (Intermediate) (JEI)
2) 30 Japanese learners of English (Advanced) (JEA)
3) 30 Native speakers of English (E)
4) ,30 Native speakers of Japanese (J)
JEI were students enrolled in the English Foundation Program at the
Hawai'i Pacific University (HPU) at the time of the study. Their average
age was 22.8. Their TOEFL scores ranged between approximately 400
and 500. JEA were undergraduate or graduate students enrolled at the
University of Hawai i at Manoa (UHM) at the time of the study. Their
average age was 27.5. Their TOEFL scores ranged from 510 to 627 (aver-
age 579.2). E and J were undergraduate and graduate students at UHM
(average age NSE: 30.4, NSJ: 25.9). J participated as non-native speakers
in JEI or JEA.
3.2. Materials
A 20-item Dialog Construction Questionnaire was prepared in English
and Japanese. The items represented different social domains and inter-
locutor role relationships in terms of gender, social distance and relative
social status, and differing degrees of severity of the committed offense.
The content of the items is listed in (5) through (24).
(5) A and B are friends. A damaged B's car while backing up.
(Damaged Car)
(6) A and B are friends. A borrowed a magazine from B and spilled
coffee over it. (Ruined Magazine)
(7) At a staff meeting, teacher A contradicted teacher B.
(Contradiction)
162 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
(8) At a staff meeting, teacher A accused teacher B of being a poor
teacher. (Poor Teacher)
(9) At an office, a junior colleague forgot to pass on a personal
message to a senior colleague. (Personal Message Low-High)
(10) At an office, a senior colleague forgot to pass on a personal
message to a junior colleague. (Personal Message High-Low)
(11) At an office, a junior colleague forgot to pass on an important
business message to a senior colleague. (Business Message
Low-High)
(12) At an office, a senior colleague forgot to pass on an important
business message to a junior colleague. (Business Message
High-Low)
(13) At a restaurant, a customer changed her mind after the order
had already been taken. (Order Change)
(14) At a restaurant, a waiter spilled food on a customer's clothes.
(Food on Customer)
(15) At a restaurant, a waiter brought the wrong order. (Wrong
Order)
(16) At a restaurant, a customer spilled food on a waiter. (Food on
Waiter)
(17) At the airport, a customs official has messed up a traveller's
suitcase. (Messed-up Bag)
(18) At the airport, a traveller has been caught trying to smuggle a
Bonsai tree into Japan. (Bonsai Tree)
(19) At the airport, a customs official has broken a legally purchased
statue when searching a traveller's suitcase. (Broken Statue)
(20) At the airport, a traveller is unable to produce a customs form.
(Customs Form)
Transfer and proficiency in interlanguage apologizing 163
(21) A professor has not yet graded a term paper which a student
was supposed to pick up. (Ungraded Paper)
(22) A student forgot a book she was supposed to return to her
professor. (Borrowed Book)
(23) A professor misplaced a student's term paper and failed the
student. (Failed Student)
(24) A student plagiarized from a published book and is found out
by the professor. (Cheating Student)
Subjects were asked to supply both the offender's and the offended party's
turn (see Appendix 1 for a sample item). For this study, only the first pair
parts were analyzed.
In order to examine the relationship between contextual factors and
strategy use, an Assessment Questionnaire was prepared, including the
same offense contexts as the Dialog Construction Questionnaire. Each
context was rated on a five-point scale for five context-internal factors
(severity of offense, offender's obligation to apologize, likelihood for the
apology to be accepted, offender's face loss, offended party's face loss)
and two context-external factors (social distance and dominance) (see
Appendix 2 for sample item). Since both questionnaires were adapted
from a previous study (Bergman & Kasper 1993), the japanese version of
the questionnaires was prepared by first translating the English
questionnaires into japanese and then back into English. Adjustments to
the japanese version were made based on a comparison of the original
and translated English versions. The translations were provided by a
graduate student who is a native speaker of Japanese, bilingual in
japanese and English, and not an author of this chapter.
3.3. Procedure
Items in all questionnaires were randomized. For the Dialog Construction
Questionnaire, subjects were instructed to fill in what they would say in
each of the twenty contexts. The intermediate and advanced learners (jEI
and JEA) filled in the English and japanese version of the Dialog
Construction Questionnaire in counterbalanced order. At least one week
elapsed before the second Dialog Construction Questionnaire was
164 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
administered. The Japanese and English version of the Assessment
Questionnaire was completed by the native speakers of Japanese (J) and
English (E), respectively. No time limits were imposed on completing the
Dialog Construction and Assessment questionnaires.
3.4. Analysis
The Dialog Construction data were coded into the major categories (from
Bergman - Kasper 1993) listed in (25) through (29). IFID: Illocutionary
Force Indicating Device, specifying the force of apology ("I'm sorry",
"I'm afraid").
(25) Upgrader: Element increasing apologetic force ("I'm terribly
sorry", "I really didn't mean to hurt you").
(26) Taking on responsibility: Speaker admitting the offense, including
self-blame ("How stupid of me"), lack of intent ("I didn't mean to
do this"), and admission of fact ("I haven't graded it yet").
(27) Downgrading responsibility or severity of offense: (a) utterance
reducing speaker's accountability for the offense, including excuse
("My watch had stopped"), justification ("I was suddenly called to
a meeting"), claiming ignorance ("I didn't know you were expect-
ing me"), problematizing a precondition ("we weren't supposed to
meet before 12"), or denial ("I didn't do it"); (b) utterance reduc-
ing severity of offense ("I'm only ten minutes late").
(28) Offer of repair: Speaker offering to remedy damage inflicted on
offended party by specific compensation for the offense ("I'll pay
for the damage", "I'll have it marked tomorrow").
(29) Verbal redress: Speaker showing concern for offended party ("I
hope you weren't offended"), efforts to appease ("Let me buy you
a drink") or promise of forbearance ("It won't happen again").
Interrater-reliability was established through consensus coding by three
raters (E, JEI, JEA data) and two raters (J data). For the analyses in
Section 4, differences between the four groups are reported in percentages
of positive and negative transfer in the JEI and JEA groups. Since this
phase of the study was essentially exploratory, no inferential statistics
were applied. Statistical tests of the transfer analyses in Section 5 were
based on total strategy frequencies tallied for the intermediate and
Transfer and proficiency in interlanguage apologizing 165
advanced ESL speakers, and the American baseline groups. All Chi-
square tests were calculated with a correction for continuity.
4. Results
4.1. Contextual effects on pragmatic transfer
Comparison of the contextual assessments provided by the native
speakers of English and japanese showed that there was strong agree-
ment in the two groups' perception of status, obligation to apologize, and
likelihood of apology acceptance. On each of these factors, only two con-
texts received different assessments. The assessment of likelihood of
apology acceptance parallels the one found by Bergman and Kasper
(1993). In their study, ratings of the same contexts obtained from native
speakers of American English and Thai revealed likelihood of apology
acceptance to be the factor on which both groups agreed most, only three
contexts receiving diverging ratings. In contrast, the findings for status
and obligation to apologize in the present study deviate considerably
from previous research. Beebe and Takahashi's studies of face-threatening
acts in japanese and English (e.g., Beebe - Takahashi 1989; Takaha-
shi - Beebe 1993) consistently demonstrated a much stronger different-
iation of status-relationships in japanese than in American speakers'
performance of such acts. We had therefore suspected that a similar
difference would show up in japanese and American raters' perceptions
of status in offense contexts. Obligation to apologize was the factor on
which Thais and Americans differed most - eleven out of the twenty
offense contexts received different scores from these groups (Bergman -
Kasper 1993). In the present study, the most different ratings were given
on offenders' face loss (7 contexts), offended party's face loss (9 con-
texts), and social distance (11 contexts).
Because previous research had demonstrated that context assessment
affects the selection of apology strategies, we reasoned that pragmatic
transfer can preliminarily be predicted from similarities and differences of
native speakers' contextual assessments. Thus, it was assumed that
similar native speaker ratings predict positive transfer of apology
patterns, whilst different ratings predict negative transfer. A context was
Table 1. Transfer predictions based on contextual assessment
~
0\
0\
Context Social Social Severity Obligation Likelihood Off. H'S Transfer
Distance Power Face-Loss Face-Loss Prediction
~
~
Damaged Car + - + + - -
fZJ
~
+ ~
~
Ruined Mag. - + - + + + - fZJ
;::s--
~
Contradiction + + + + + - + positive ~
Poor Teacher + + - + + + + positive I
Pers.Mess.L-H - + + + + + - positive ~
Pers.Mess.H-L + + + + + + + positive
0<
Bus.Mess L-H positive
~
+ + + + + - -
;::s--
Bus.Mess H-L - + + + + - -
fZJ
~ .
~
Order Change - - + + - + +
fZJ ~
Food on Cust. - + + + + + + positive I
Wrong Order - + + + - + + positive
0
Food on Wait. + + + + + + + positive
~
Messed Up Bag + + + - + + - fZJ ~
~
Bonsai Tree - + + + + + + positive
~
""'t
Broken Statue + + + + + - + positive I
Customs Form - + + + + + + positive ~
Ungraded Pap. - + + + + + + positive
~
0
Borrowed Book + + - - + - - negative
~
~
Failed Student - - + + + + -
fZJ
Cheating Student - + - + + + + positive
+ =same rating by J and E
- =different rating by J and E
off. face-loss =offender's face-loss
Transfer and proficiency in interlanguage apologizing 167
categorized as "similar" when ratings did not differ on five or more
factors, and as "different" when more than four factors were rated
differently. The categorizations were based on separate multivariate
analyses of variance for each of the twenty contexts. The dependent
variables in these analyses were the seven contextual factors, and the
independent variable was membership in the native Japanese or American
groups. Table 1 shows for each context whether it was rated the same or
different on each of the seven factors, and the transfer predictions based
on these ratings.
Since the American and Japanese raters in this study generally showed
more agreement than disagreement in their context assessment, predic-
tions of positive transfer by far outnumbered those of negative transfer.
Only Borrowed Book received different ratings on four factors, and was
therefore expected to elicit negative transfer. For six contexts, same and
different ratings counterbalanced each other; hence no transfer predic-
tions could be made. Thirteen contexts were rated similar and thus
predicted to elicit positive transfer of apology strategies.
4.2. Transfer of apology strategies
In order to determine whether pragmatic transfer was operative, a
modified version of Selinker's (1969) operational definition of language
transfer was adopted from Kasper (1992). According to this definition,
lack of statistically significant differences in the frequencies of a prag-
matic feature in the first language, second language, and interlanguage
can be operationally defined as positive transfer. Statistically significant
differences in the frequencies of a pragmatic feature between inter-
language-second language and first language-second language and lack of
statistically significant difference between interlanguage and first
language can be operationally defined as negative transfer. Applied
to the identification of transfer of apology strategies, positive transfer
obtains when there is no statistically significant difference in the use
of an apology strategy between E and J, E and JEIIJEA, and J and
JEIIJEA. Negative transfer requires statistically significant differences
in strategy use between E-J and E-JEIIJEA and no statistically significant
differences between J-JEIIJEA. Because in this study, J was a subset of
JEI and JEA (i.e., the same subjects served as learners and first language
native speakers), comparisons between J and JEIIJEA were not carried
out.
168 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
Table 2. Predicted positive transfer of apology strategies
Context JEI% JEA%
Contradiction 67 67
Poor Teacher 83 67
Personal Message L-H 83 83
Personal Message H-L 50 67
Business Message L-H 83 100
Food on Customer 50 67
Wrong Order 83 83
Food on Waiter 17 33
Bonsai Tree 83 100
Broken Statue 100 83
Customs Form 100 100
Ungraded Paper 17 67
Cheating Student 83 83
The transfer predictions established through comparison of contextual
factors in the previous section were matched against the actual occurrence
of transfer on each apology strategy in each offense context. There was
high correspondence between the predictions of positive transfer and its
occurrence. Table 2 displays the percentage of strategies which were trans-
ferred positively according to the context-based transfer predictions.
In the majority of contexts where positive transfer had been predicted,
the converging social perceptions of Japanese and American subjects was
reflected in the same use of apology strategies by both learner groups.
This match was even greater in the case of the advanced learners, who
outperformed the intermediate learners in six contexts. The intermediate
learners, however, did better than the advanced learners in Poor Teacher
and Broken Statue. There were only two contexts where the prediction of
positive transfer was not borne out. In Food on Waiter, both learner
groups apologized differently from the American native speakers on all
but one measure (JEI: Taking on Responsibility) and two measures,
respectively (JEA: IFID and Taking on Responsibility). In Ungraded
Paper, the only strategy which the intermediate learners used in the same
way as the Americans was upgrading apologetic force. Table 3
summarizes how the prediction of positive transfer was matched by the
actual use of apology strategies.
Again, the advanced learners' performance compared better to the
transfer predictions than the intermediate learners' in their choice of four
Transfer and proficiency in interlanguage apologizing 169
Table 3. Positively transferred apology strategies (all contexts)
IFID 0/0 UG% TR% DG% Rep 0/0 VR 0/0
JEI
77 77 85 54 62 62
JEA
77 85 92 54 77 85
IFID = Illocutionary Force Indicating Device
UG = Upgrading
TR = Taking on Responsibility
DG = Downgrading
Rep = Offer of Repair
VR = Verbal Redress
apology strategies. The strategy where both learner groups displayed least
positive transfer was Downgrading apologetic force.
Overall, native speakers' social perceptions proved to be an excellent
predictor of positive pragmatic transfer: Where Japanese and Americans'
contextual assessments converged, Japanese learners of English were
prone to use the same strategies in their interlanguage apologizing as both
native speaker groups. This was even more true of the advanced learners
than of the intermediate learners, which makes sense in light of the
assumption that advanced learners are likely to be more acculturated
than intermediate learners, and have the linguistic facility to transfer
pragmatic strategies from their native language where they perceive this
as consistent with target use.
A requirement for positive transfer is obviously that the native speaker
groups display the same kind of behavior. In the instances of positive
transfer noted above, the similarity of native speakers' strategy use could
be seen as the behavioral correlate to their converging social perceptions
of contextual factors. Interestingly, native speakers also preferred the
same apology strategies where they did not agree in their context assess-
ment, and in these contexts, the learners displayed the same strategy
choices as the native speakers did. In other words, positive pragmatic
transfer of apology strategies occurred even in contexts where this was
not predicted by the contextual assessment. Table 4 summarizes the con-
texts in which apology strategies were positively transferred contrary to
the transfer prediction.
Curiously, Borrowed Book, the only context for which negative
transfer had been predicted, achieved the highest scores on actual positive
170 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
Table 4. Positive transfer contrary to prediction
Context JEI% JEA%
Damaged Car 83 67
Ruined Magazine 67 33
Business Message H-L 83 83
Order Change 67 83
Messed-up Bag 67 50
Borrowed Book 83 100
Failed Student 83 83
transfer. In Ruined Magazine and Messed-Up Bag, the advanced learners'
strategy choices differed most from the native speakers' apology patterns
and were thus more in accordance with the prediction of zero transfer.
The advanced learners followed the prediction of zero transfer in
their selection of IFIDs, the intermediate learners on Repair. On all other
strategies, the learners converged in their strategy selection with the
native speakers and did not differ in terms of proficiency. Since only the
native speakers' sociopragmatic perceptions were elicited, we have no
way of knowing whether the learners viewed the seven contexts similarly
to the American native speakers and selected their apology strategies in
accordance with their contextual assessment, or whether they assessed
these contexts differently from the American judgements but nonetheless
followed the same pattern of strategy selection.
The same mismatch between transfer predictions and transfer
occurrence was observable in the few instances of negative transfer. Just
Table 5. Unpredicted positive transfer of apology strategies (all contexts)
JEI
JEA
IFID%
86
29
UG%
71
86
TR%
86
86
DG%
100
86
Rep 0/0
29
71
VR 0/0
71
71
IFID =Illocutionary Force Indicating Device
UG =Upgrading
TR =Taking on Responsibility
DG =Downgrading
Rep =Offer of Repair
VR =Verbal Redress
Transfer and proficiency in interlanguage apologizing 171
as the only prediction of negative transfer (in the Borrowed Book context,
see above) contrasted with actual positive transfer of strategy choice,
negative transfer occurred in contexts were positive transfer had been
predicted. Table 6 lists the strategies which were transferred negatively
from Japanese, and the frequencies by which these strategies were used
by the learners and native speakers.
Each strategy was thus transferred negatively at least once. In the
Ungraded Paper context, a professor promised to have a student's paper
graded but has not done it yet when the student comes to pick it up. In
response to the student's request "Can I have my paper back now?", the
intermediate learners apologized less by means of an explicit apology
(IFID) than the American native speakers and advanced learners. Typical
responses given by the four language groups are listed in (31) through (37).
Table 6. Negative transfer of strategy choice
IFID in Ungraded Paper
J%
53
JEI%
57
A%
83 JEI < A
Upgrading in Food on Customer
J% JEl%
50 48
Upgrading in Food on Waiter
J% JEI%
20 23
J % JEA 0/0
20 43
Taking on Responsibility in Contradiction
J% JEI%
63 66
A 0/0
77
A%
73
A%
73
A%
37
JEI < A
JEI < A
JEA<A
JEI > A
JEI < A
A%
83
Downgrading Responsibility or Severity in Cheating Student
J % JEI % A0/0
83 71 48 JEI > A
J% JEA% A%
83 79 48 JEA > A
Offer of Repair in Food on Customer
J% JEI%
3.3 55
Verbal Redress in Contradiction
J% JEI%
30 31
A%
70 JEI < A
172 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
(31) J: Chotto matte, sugu tsukeru kara
"Wait just a minute. I will mark your paper right away"
JEI: Not yet. Have a seat and wait a minute.
A & JEA: Sorry, I haven't quite finished it. Could you come
back tomorrow?
In Food on Customer, a waiter spills food on a customer's dress. While
the Americans and advanced learners intensified the waiter's apology to
the customer, the intermediate learners followed the Japanese native
speakers by upgrading apologetic force less in this context, as is
illustrated in (32). Mooshiwakearimasen here might be taken as an up-
graded apology since its literal meaning is 'inexcusable'. However,
mooshiwakearimasen is the appropriate form for a waiter at an expen-
sive restaurant to apologize to the customer. In upgraded responses,
intensifiers such as taihen 'very/ awfully' or hontooni 'truly/really' are
explicitly added to mooshiwakearimasen. Therefore, the expression
mooshiwakearimasen without any intensifiers was counted as non-up-
graded form.
(32) J: Mooshiwakearimasen, okyakusama. "It IS inexcusable,
sir/ma'am" .
JEI: Oh, I'm sorry.
A & JEA: Oh no! I'm so incredibly sorry!
The same offense in reverse role relationships is represented in Food on
Waiter, where a guest in a restaurant knocks off a waiter's tray when
getting up and the food spills all over the waiter. While the American
subjects had the guest apologize to the waiter just as profusely as the
waiter to the guest in the previous situation, most respondents in both
learner groups and the Japanese native speakers apologized less strongly
by not intensifying their expression of regret.
(33) J: Gomenasai. "I'm sorry."
JEI & JEA: I'm sorry. Are you okay?
A: Oh, my God! I'm terribly sorry. I'm such a klutz.
However, almost twice as many of the advanced learners as the interme-
diate learners upgraded apologetic force, suggesting that they are on their
way to abandoning the native Japanese pattern of differentiating apology
intensification according to interlocutor status in favor of the more
egalitarian target usage.
Transfer and proficiency in interlanguage apologizing 173 '
In Contradiction, a teacher contradicts something that another teacher
said at a staff meeting, and hurt his colleague's feelings. Most of the
intermediate learners and Japanese native speakers redressed this offense
by explicitly assuming responsibility for it. The American respondents
preferred to offer sympathy in this situation, as will be seen below. The
contrasting patterns are illustrated by the responses in (34).
(34) J: Konoaida wa gomen. Warugi wa nakattan da.
"I am sorry for the other day. I didn't mean it".
JEI: I'm sorry. I hurt your feelings, I think.
A & JEA: I hope you didn't take what I said personally.
In the role of a student who plagiarized for a term paper and is found out
by his professor (Cheating Student), the learners and Japanese native
speakers downplayed the offense by finding excuses or claiming
ignorance. The Americans used this strategy less, probably on the
assumption that attempts to downplay the offense would make their case
worse rather than better, cf., (35).
(35) J: Sumimasen. Tesuto toka iroiro to isogashikute, shimekiri ni
maniawazu sono tame ni shikata ga nakatan desu. Hi wo
aratamete teishutsu shitemo ii desu ka.
"I sorry. I was busy preparing tests, so in order to prepare them
in time it was the only thing I could do. May I submit it later?"
JEI: I'm sorry for copying, but I was busy.
JEA: I'm sorry. I tried to write the essay myself, but the state-
ment in the text was so nicely written, and I didn't know
you would consider if we copy from the text.
A: Well, actually, I did get some of my ideas from a book.
In the Food on Customer context, the American and advanced
respondents had the waiter offer repair to the customer, such as
promising to have her dress cleaned. Half of the intermediate learners did
not offer repair, and thus assumed a medium position between American
and Japanese usage. Only a single Japanese respondent offered repair to
this offense. Typical responses are given in (36).
(36) J: Taihen mooshiwake gozaimasen. "I am very sorry"
JEI: Oh, I'm sorry. Are you all right?
A & JEA: Oh, I'm terribly sorry. We'll have the suit cleaned for
you.
174 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
Finally, the intermediate learners followed the Japanese pattern of not
expressing much concern for the insulted teacher in Contradiction. Most
of the American respondents offered tokens of concern for the offended
colleague's feelings, cf., (37). Half of the advanced learner group
expressed concern; the advanced learners thus placed themselves between
the Japanese/intermediate and American respondents in this context.
(37) J: Warukatta keredo are wa boku no sochokku na iken de atte,
kimi wo kizutsukeru tsumori wa nakatta.
"Sorry, but that was my honest opinion and I didn't mean to
hurt you."
JEI: I know that I hurt you, but that which I said was my
opinion. I think I was right.
A: I'm sorry, Jennifer. Did my words upset you?
Most of the negative transfer occurred in contexts with a high power
differential between the interlocutors, regardless whether the offender
was the higher status participant (Food on Waiter, Ungraded Paper) or in
the lower status position (Food on Customer, Cheating Student). In Food
on Customer and in both of the student-professor contexts, social
distance was perceived differently by Japanese and American raters. It is
possible, therefore, that despite the overall agreement in context percep-
tion, the diverging assessment of social distance is contributive to the
differences in strategy use. This interpretation is consistent with previous
studies, which demonstrated a complex interaction of contextual factors
and choice of apology strategies. As noted above, each context factor has
a different weight as trigger of strategy choice, and the same factor affects
the selection of some strategies but not others. In previous studies, social
distance was shown to affect offender's assumption of responsibility for
the offense (Bergman - Kasper 1993) and, more tentatively, their choice
of IFID (Olshtain 1989). This study suggests that diverging perceptions
of social distance can account for different choices of IFID, Upgrading of
apologetic force, and offer of Repair. It does not explain, however, why
different assessments of social distance affect some but not all contexts.
A closer look at two pairs of contexts which differ only in the inter-
locutors' power relationship indicates that the direction of the status dif-
ferential is the crucial factor. Japanese and American raters did not differ
in their assessment of social power in Food on Customer / Food on
Waiter and Borrowed Book / Ungraded Paper. However, in both of these
paired contexts, Japanese and Americans differed in their actual use of
Transfer and proficiency in interlanguage apologizing 175
the IFIDs in the student-professor situations and of Upgrading in the
waiter-customer contexts. Americans used IFIDs equally often in the
student-professor situation no matter whether the offender was the
professor or the stuent (83 0/0), and they upgraded their apologies in the
customer-waiter contexts to the same extent regardless of offender's
status (75 %). The japanese respondents, by contrast, made their selec-
tion of both strategies contingent on the direction of the status relation-
ship. In the high to low contexts, they apologized less by means of an
explicit formula (professor ~ student (53 0/0)) and intensified apologetic
force less (customer ~ waiter (20 %)) than in the corresponding low to
high contexts (student ~ professor: 87 % IFID; waiter ~ customer: 50 %
upgrading). This contrast in power differentiation is consistent with
Barnlund and Yoshioka's (1990) observation that japanese offenders are
more status-sensitive in their choice of apology strategy than Americans.
The learners who followed the native model thus transferred the status-
differential apology pattern from japanese to English in these contexts.
As an instance of sociopragmatic transfer, this finding fits in well with
Takahashi and Beebe's (e.g., 1993) work on face-threatening acts in
japanese-English interlanguage. Their studies showed that compared to
Americans, native speakers of japanese employ a more distinctly status-
differentiating approach to corrections and refusals. Mitigators such as
softeners and expressions of regret were used more frequently by the
status lower to the status higher interlocutor than vice versa by both
Americans and japanese; however, the japanese respondents accentuated
the status difference more than the Americans did. just as the japanese
learners of English transferred the status-differential patterns of mitiga-
tion to their performance of face-threatening acts in English, the inter-
mediate learners in this study aggravated apologetic force according to
the status-differential first language model.
5. Extensions to recent studies of Japanese-American
apology
The foregoing discussion of transferred apology strategies was based on
Dialog Construction codings devised by Bergman and Kasper (1993). In
order to examine the generality of pragmatic transfer, however, it is
important to compare the results of the present study with other cross-
176 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
cultural studies of apology in japanese and American contexts. Barnlund
and Yoshioka (1990) found, for example, that japanese were more likely
to offer several types of apology for a transgression than Americans
rating the same situation. They were also more likely to offer significantly
more repair by suggesting some form of compensation for their trans-
gressions than the Americans. In contrast, Americans were found to be
more likely to provide a rationale - more excuses, justifications and
downgrading of the severity of the transgression. japanese appear to be
equally direct as Americans in apologizing for perceived wrongdoings. As
Barnlund and Yoshioka used different units of measurement (ratings) and
a different classificatory system, we will equate the Dialog Construction
Questionnaire results with the most obvious of the Barnlund and
Yoshioka findings in order to establish comparability between the cross-
cultural contrasts identified by Barnlund and Yoshioka and instances of
negative transfer in our study.
In the Barnlund and Yoshioka study, the variance in preferred
strategies is taken to represent major pragmatic differences between the
two cultural norms. Such differences can potentially result in negative
pragmatic transfer, as defined above.
Four apology strategies involving pragmatic contrasts in the Barnlund
and Yoshioka research were matched with the most comparable five
strategy types in the present study. These were "explaining the situ-
ation"/downgrading, "saying directly 'I am very sorry'''/IFID and up-
grading, "offering to do something for the other person"/ repair, "apol-
ogizing directly; several ways several times"/ (multiple apologies for a
single transgression).
Instances of apology strategies for the same three groups JEI, JEA, and
A were reanalyzed for the twenty independent Dialog Construction
Questionnaire contexts. As the preceding section of this study has
suggested, there should be a greater likelihood that the lEIs would use
japanese strategies than the JEAs. The results of the comparisons are
presented and discussed individually below.
5.1. "Explaining the situation"/ downgrading
In Barnlund and Yoshioka, "explaining the situation" is used less
frequently by japanese than by American offenders. Contrary to
Barnlund and Yoshioka's findings, the present study revealed that in only
one context out of 20 is there a significant difference between Americans
Transfer and proficiency in interlanguage apologizing 177
and Japanese in the downgrading/rationalizing category. Specifically, the
JEIs used the preferred American strategy of rationalizing the transgres-
sion more than the American group. There was no such difference
between the JEAs and the native speakers of English. The sole difference
was for Poor Teacher.
In this situation teacher A accuses teacher B of being a poor teacher
at a staff meeting (Chi-square = 6.38; p = .0118). Such a difference
could indicate a transitional period during which the Japanese learners
over-accommodate toward what they perceive the American norm
to be, and in a sense 'out-American' the Americans. In no context do
the advanced learners differ from the Americans in terms of the
frequency of downgrading/rationalization, a finding not in agreement
with the pragmatic contrasts established by Barnlund and Yoshioka.
This discrepancy may stem from the fact that all of the Japanese
respondents were residents of Honolulu, and had perhaps had
sufficient exposure to the Hawaiian-American norms of downgrading
and reference to circumstantial causes of the transgression instead of
taking responsibility, whether such taking of responsibility is warranted
or not.
5.2. "Offering to do something for the other person"/repair
According to Barnlund and Yoshioka, Japanese offenders are more prone
to offer compensation for an infraction than Americans. We therefore
predicted that the learners would offer more repair than the American
native speakers. However, with one exception, the learners differed from
Americans in their repair offers by providing less rather than more repair.
The intermediate learners were much more prone to undersupply repair
than the advanced learners were. The contexts in which the JEIs differed
from the Americans in their offers of repair are listed in (38) through
(44).
(38) Damaged Car: A has had an accident with a car borrowed
from B.
Chi-square = 4.31 p = 0.037
(39) Ruined Magazine: A borrowed a magazine from B and spilled
coffee over it.
Chi-square =4.31 p =0.037
178 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
(40) Business Message Low to High: At an office, a junior colleague
forgot to pass on an important business message to a senior
colleague.
Chi-square = 7.18 p = 0.007
(41) Business Message High to Low: At an office, a senior colleague
forgot to pass on an important business message to a junior
colleague.
Chi-square = 7.06 p = 0.007
(42) Food on Customer: At a restaurant, a waiter spills food on a
customer's clothes.
Chi-square =3.88 p =0.048
(43) Food on Waiter: At a restaurant, a customer spills food on a
waiter.
Chi-square = 7.12 p = 0.007
(44) Failed Student: A professor misplaced a student's term paper
and failed the student.
Chi-square = 5.42 P = 0.019
Since the contexts in which the intermediate learners offer less repair
involve different status relationships and degrees of social distance, their
divergent apology behavior cannot be explained in terms of context
external factors. However, with the exception of Ruined Magazine, the
offenses are all high severity infractions. For the American subjects,
repair offers are appropriate ways of redress to these kinds of trans-
gressions. The learners' failure to offer compensation on a regular basis
in these contexts suggests that they underdifferentiate their selection of
repair offer according to high versus low severity offenses.
In contrast to the intermediate learners, the frequency of repair offers
by the advanced ESL speakers is very similar to the native speakers of
English. Only in Damaged Car (Chi-square = 4.13; p = 0.037), Food on
Waiter (Chi-square =7.12; p =0.007), and Ungraded Paper (Chi-square =
4.31; p = 0.037) do the advanced learners differ from native speakers.
In Damaged Car and Food on Waiter, like their intermediate
proficiency counterparts, the advanced learners provide significantly less
apology than the native speakers for the transgressions. This may indicate
that while advanced learners are in general familiar with American
Transfer and proficiency in interlanguage apologizing 179
apology strategies, in highly marked and unfamiliar contexts such as
Damaged Car, they know they cannot revert to japanese strategies but do
not have the experiential basis to extrapolate from their repertoire of
second language pragmatic strategies. Familiarity with social contexts has
been shown to influence interlanguage pragmatic performance generally
(Eisenstein - Bodman 1986), and pragmatic transfer specifically
(Takahashi 1992).
Food on Waiter represents the influence of role differentiation aware-
ness between Americans and japanese. An infraction such as the one in
Food on Waiter does not warrant an offer of compensation from the
japanese because of the role/status differential implicit in customer/
service employee relations. Ungraded Paper is the only instance of
negative transfer (Chi-square 4.31; p = 0.037) in the lEAs' repair offers.
We note that the lEIs did not differ from the native speakers of English
in this context (a professor failing to grade a student paper on time).
Here, the advanced japanese are more inclined to see repair offer from
the professor as appropriate, whereas their low-proficiency counterparts
do not. What may appear to be negative transfer by the advanced
japanese in this context may actually be indicative of more subtle and
complex pragmatic influences. They may, for instance, realize that the
second language status differential does not require high to low repair.
The advanced japanese would therefore be less inclined to consider
no repair as appropriate. They might not however realize that
the American norm is not built on a single egalitarian principle,
and that it might be subject to real world constraints such as the fact
that professors are notoriously tardy and expect that students understand
this.
5.3. "Direct apology"/IFID and upgrading
This category of apology strategy involves the speaker providing a clear
and direct apology for an infraction. The directness is most overtly
expressed as some variant of the IFID "I am sorry", and is here
considered distinct from an indirect form of apology such as "it is a
shame it had to turn out that way". Barnlund and Yoshioka find that for
both Americans and japanese, direct apology is the most highly preferred
form of redress. japanese are even more inclined to employ direct
apology. Learners can therefore be expected to use equal or surpass
Americans in their use of direct apology.
180 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
This expectation was confirmed. There was only one context, Food on
Waiter, in which the JEIs offered direct apology less often than the
American native speakers (Chi-square =7.32; p =0.006). The status
differential between customer and waiter does not seem to require direct
apology in the view of the intermediate learners.
5.4. Multiple apologies
The Barnlund and Yoshioka study found that the Japanese are more in-
clined to provide multiple apologies or apologetic paraphrases than
Americans. The function of the multiple apology in the Japanese milieu is
to demarcate the speaker's responsibility for the transgression, and to
provide signals of sincerity for the apology. Multiple apology should
therefore be a prime candidate for negative transfer in the dialog con-
struction tasks used in this study. Indeed, the JEI group should transfer
multiple apologies more than the advanced learners. However, the results
suggest that differences in the frequency of multiple apology is not as
common as expected. For the JEIs, multiple apology was significantly dif-
ferent from the native English speakers in only two contexts. In Messed
Up Bag, the JEIs provided more multiple apologies than the Americans
(Chi-square = 4.27; p = 0.038), but in Ungraded Paper, they chose this
strategy significantly less than the American NS (Chi-square = 4.56;
p = 0.017).
For the advanced speakers, the likelihood of negative transfer of the
multiple apology strategy can be considered less than that for the inter-
mediate learners. This fact is borne out in the present study. Only in one
context did the JEAs differ from the native speaker of English. Here, the
Japanese provided significantly more multiple apologies (Chi-square =
5.82; p =0.015). The transgression in Ruined Magazine is one that can
be considered avoidable. Presumably, since the perpetrator was remiss
in not preventing the mishap, the severity of the transgression becomes
more acute in the minds of the advanced English as a second language
speakers.
6. Discussion
Compared to the pervasive effect of positive transfer, negative transfer
was infrequently at work in the learners' apology performance. Yet, two
Transfer and proficiency in interlanguage apologizing 181
important findings emerge from the analysis of negative transfer. For one
thing, in only two instances did the advanced learners transfer their
apology behavior from japanese to English when japanese and American
apology patterns differed, whereas the intermediate group did the same
in six instances. From this, it follows that the advanced learners have a
better ability to emulate American apology behavior than the inter-
mediate learners. Secondly, except for taking on responsibility in
Contradiction, the intermediate learners differed from the Americans in
that they provided fewer instances of the apology strategy in question.
Their negative transfer of apology strategies thus consisted in adopting a
less elaborated, first language-based approach to redress offenses than the
American native speakers and the advanced learners. This study, then,
does not lend support to Takahashi and Beebe's (1987; 1993) contention
that advanced learners display more negative pragmatic transfer because
'they have the rope to hang themselves'. Rather, when advanced japanese
learners provide responses to exceptional situations for which they have
little experience to rely on, they are inclined not to transfer first language
strategies which they suspect to be insufficient for the context. The
advanced learners may still not have developed the pragmatic where-
withal to provide the same responses to the subtleties of such situations
as do the native speakers of American English.
There are noteworthy similarities and differences in the transfer
behavior of the intermediate learners in this study and the intermediate
Thai learners of English in Bergman and Kasper (1993). The Thai
learners' performance suggested negative transfer of Upgrading and
Repair in only one context and of IFID and Taking on Responsibility in
two contexts. Their patterns of negative pragmatic transfer on these
strategies was thus quite similar to that of the japanese intermediate
learners. However, the Thai learners transferred negatively on their use of
Downgrading in three contexts where the japanese intermediate learners
did not transfer negatively at all and on Verbal Redress in as many as six
contexts. Furthermore, these negative transfers were the result of over-
supplying the strategy in question, rather than under-using it, which is
what the japanese learners were inclined to do.
Previous research has demonstrated that negative pragmatic transfer is
more prevalent in foreign language contexts than in second language
contexts (Takahashi - Beebe 1987). This difference in learning contexts
can partly account for the variance in negative transfer between the Thai
and japanese learners: the Thai learners were in an English as a foreign
language context, whereas the japanese learners were in an English as a
182 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
second language environment. This generalization is also borne out in the
comparison of the Barnlund and Yoshioka predictions with the second
language acquisition patterns observed here. The richer opportunities for
input and productive use of English in the English as a second language
context surely put the japanese learners at an advantage. Furthermore,
the variety of English which served as target norm in both the Thai and
japanese studies was Hawai'i Standard English, which was consistent
with the input variety received by the japanese in Honolulu, but not by
the Thais in Bangkok. Another reason for the differences between the
Barnlund and Yoshioka conclusions and those of the present study relates
to differences in the sociolinguistic norms of the three speech communi-
ties. Since the Americans in the Barnlund and Yoshioka study were
presumably from two distinct homogeneous populations in japan
and on the U. S. mainland, whereas the Americans and japanese in the
present study were from a single heterogeneous speech community in
Honolulu, we might surmise that there was a greater likelihood for
exposure, accommodation and convergence in the Hawaiian milieu.
Given that the focus of this study is apology, it is perhaps most appro-
priate to apologize for the obvious limitations of the study itself. With the
act of apology as the center of much cross-cultural miscommunication, and
its status in perceptions of duty, responsibility and liability in American and
japanese societies, it is of particular importance to continue the investiga-
tion of apology across a wide variety of communicative domains.
Appendix 1
A sample item from the Dialog Construction Questionnaire (Ruined Magazine)
At a friend's home
Ann and Bill are both 35 years old and are good friends. Ann borrowed a
computer magazine from Bill. Unfortunately, Ann spilled coffee on the magazine
and damaged it. She is now returning it to Bill.
Bill: What happened to my magazine?
Ann:
Bill:
Appendix 2
A sample item from the Assessment Questionnaire (Ruined Magazine)
At a friend's home
Transfer and proficiency in interlanguage apologizing 183
Ann and Bill are both 35 years old and are good friends. Ann borrowed a
computer magazine from Bill. Unfortunately, Ann spilled coffee on the magazine
and damaged it. She is now returning it to Bill.
1. How CLOSE are Ann and Bill in this situation?
1 2 3 4
very close
5
very distant
2. What is the STATUS RELATIONSHIP between Ann (A) and Bill (B)?
1 2 3 4 5
A>B A=B A<B
3. How SERIOUS is Ann's OFFENSE?
1 2 3 4 5
very serious not serious
4. Does Ann have the OBLIGATON to apologize?
1 2 3 4 5
strong obligation no obligation
5. Is Bill LIKELY to ACCEPT Ann's apology?
1 2 3 4 5
very likely unlikely
6. Is this situation EMBARRASSING to Ann?
1 2 3 4 5
not embarrassing very embarrassing
7. Is this situation EMBARRASSING to Bill?
1 2 3 4 5
not embarrassing very embarrassing
References
Barnlund, Dean C. - Miho Yoshioka
1990 "Apologies: Japanese and American styles", International Journal of
Intercultural Relations 14: 193-206.
Beebe, Leslie M. - Tomoko Takahashi
1989 "Sociolinguistic variation in face-threatening speech acts", in:
Miriam Eisenstein (ed.), 199-218.
Beebe, Leslie M. - Tomoko Takahashi - Robin Uliss-Weltz
1990 "Pragmatic transfer in ESL refusals", In: Stephen D. Krashen -
Robin Scarcella - Elaine Andersen (eds.), 55-73.
Bergman, Marc L. - Gabriele Kasper
1993 "Perception and performance in native and nonnative apology", in:
Gabriele Kasper - Shoshana Blum-Kulka (eds.), 82-107.
184 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
Blum-Kulka, Shoshana
1982 "Learning how to say what you mean in a second language: A study
of speech act performance of learners of Hebrew as a second
language", Applied Linguistics 3: 29-59.
1991 "Interlanguage pragmatics: The case of requests", in: Robert
Phillipson - Eric Kellerman - Larry Selinker - Michael Sharwood
Smith - Merrill Swain (eds.), 255-272.
Blum-Kulka, Shoshana - Elite Olshtain
1986 "Too many words: Length of utterance and pragmatic failure",
Studies in Second Language Acquisition 8: 165-180.
Blum-Kulka, Shoshana - Hadass Sheffer
1993 "The metapragmatic discourse of American-Israeli families at
dinner", in: Gabriele Kasper - Shoshana Blum-Kulka (eds.),
196-223.
Blum-Kulka, Shoshana - Juliane House - Gabriele Kasper (eds.)
1989 Cross-cultural pragmatics. Norwood, NJ: Ablex.
Bodman, Jean - Miriam Eisenstein
1988 "May God increase your bounty: The expression of gratitude in
English by native and nonnative speakers", Cross Currents 151:
1-21.
Borkin, Ann - Susan M. Reinhart
1978 '''Excuse me' and 'I'm sorry"', TESOL Quarterly 12: 57-70.
Clyne, Michael
1979 "Communicative competence in contact", ITL 43: 17-37.
Cohen, Andrew D. - Elite Olshtain
1981 "Developing a measure of sociocultural competence: The case of
apology"; Language Learning 31: 113-134.
Dechert, Hans W. - Manfred Raupach (eds.)
1989 Transfer in language production. Norwood, NJ: Ablex.
Eisenstein, Miriam (ed.)
1989 The dynamic interlanguage. New York: Plenum.
Eisenstein, Miriam - Jean W. Bodman
1986 '''I very appreciate': Expressions of gratitude by native and non-
native speakers of American English", Applied Linguistics 7:
167-185.
Frerch, Claus - Gabriele Kasper
1989 "Internal and external modification in interlanguage request realiza-
tion", in: Shoshana Blum-Kulka - Juliane House - Gabriele
Kasper (eds.), 221-247.
Fraser, Bruce
1980 "Conversational mitigation", Journal of Pragmatics 4: 341-350.
Garcia, Carmen
1989 "Apologizing in English: Politeness strategies used by native and
non-native speakers", Multilingua 8: 3-20.
Gass, Susan - Carolyn Madden (eds.)
1985 Input in second language acquisition. Rowley, MA: Newbury
House.
Transfer and proficiency in interlanguage apologizing 185
Gass, Susan - Larry Selinker (eds.)
1983 Language transfer in language learning. Rowley, MA: Newbury
House.
Giles, Howard - Patricia Johnson
1987 "Ethnolinguistic identity theory: A social psychological approach to
language maintenance", International Journal of the Sociology of
Language 68: 69-99.
Goffman, Erving
1971 Relations in public. Harmondsworth: Penguin.
Holmes, Janet
1989 "Sex differences and apologies: One aspect of communicative
competence", Applied Linguistics 10: 194-213.
House, Juliane
1988 "'Oh excuse me please ... ': Apologizing in a foreign language", in:
Bernhard Kettemann - Peter Bierbaumer - Alwin Fill -
Annemarie Karpf (eds.), 303-327.
House, Juliane - Gabriele Kasper
1987 "Interlanguage pragmatics: Requesting in a foreign language", in:
Wolfgang Lorscher - Rainer Schulze (eds.), Perspectives on
language in performance 1250-1288.
Jordens, Peter
1977 "Rules, grammatical intuitions, and strategies in foreign language
learning", Interlanguage Studies Bulletin 2: 5-76.
Kasper, Gabriele
1982 "Teaching-induced aspects of interlanguage discourse", Studies in
Second Language Acquisition 4: 99-113.
1992 "Pragmatic transfer", Second Language Research 8: 203-231.
Kasper, Gabriele (ed.)
1992 Pragmatics of Japanese as native and target language. (Second
Language Teaching & Curriculum Center Technical Report #3).
Honolulu, HI: University of Hawai'i Press.
Kasper, Gabriele - Shoshana Blum-Kulka (eds.)
1993 Interlanguage pragmatics. New York: Oxford University Press.
Kellerman, Eric
1977 "Toward a characterization of the strategy of transfer in second
language learning", Interlanguage Studies Bulletin 2: 58-145.
Kettemann, Bernhard - Peter Bierbaumer - Alwin Fill - Annemarie Karpf
(eds.).
1988 Englisch als Zweitsprache [English as a second language}.
Tiibingen: Narr.
Krashen, Stephen - Robin Scarcella - Elaine Andersen (eds.)
1990 On the development of communicative competence in a second
language. Cambridge, MA: Newbury House.
Leech, Geoffrey N.
1983 Principles of pragmatics. London: Longman.
Lorscher, Wolfgang - Rainer Schulze (eds.)
1987 Perspectives on language in performance. Tiibingen: Narr.
186 N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross
Olshtain, Elite
1983 "Sociocultural competence and language transfer: The case of
apology", in: Susan Gass - Larry Selinker (eds.), Language transfer
in language l e a r n i n g ~ 232-249.
1989 "Apologies across languages", in: Shoshana Blum-Kulka - Juliane
House - Gabriele Kasper (eds.), 155-173.
Olshtain, Elite - Shoshana Blum-Kulka
1985 "Degree of approximation: Nonnative reactions to native speech
act behavior", in: Susan Gass - Carolyn Madden (eds.), 303-
325.
Olshtain, Elite - Andrew Cohen
1983 "Apology: A speech act set", in: Nessa Wolfson - Elliot Judd (eds.),
18-35.
1989 "Speech act behavior across languages", in: Hans W. Dechert -
Manfred Raupach (eds.), 53-67.
Phillipson, Robert - Eric Kellerman - Larry Selinker - Michael Sharwood
Smith (eds.)
1991 Foreign/second language pedagogy research. Clevedon: Multilingual
Matters.
Robinson, Mary Ann
1992 "Introspective methodology in interlanguage pragmatics research",
in: Gabriele Kasper (ed.), 27-82.
Selinker, Larry
1969 "Language transfer", General Linguistics 9: 67-92.
Takahashi, Satomi - Margaret A. DuFon
1989 Cross-linguistic influence in indirectness: The case of English direc-
tive performed by native Japanese speakers. [Unpublished
manuscript, University of Hawaii at Manoa.]
Takahashi, Satomi
1992 "Transferability of indirect request strategies", University of Hawaii
Working Papers in ESL 11.1 : 69-124.
Takahashi, Tomoko - Leslie M. Beebe
1987 "The development of pragmatic competence by Japanese learners of
English", JALT Journal 8: 131-155.
Takahashi, Tomoko - Leslie M. Beebe
1993 "Cross-linguistic influence in the speech act of correction", in:
Gabriele Kasper - Shoshana Blum-Kulka (eds.), 138-157.
Tanaka, Noriko
1991 "An investigation of apology: Japanese in comparison with
Australian", Meikai Daigaku Gaikokugo Gakubu Ronshu 4:
35-53.
Thomas, Jenny
1983 "Cross-cultural pragmatic failure", Applied Linguistics 4: 91-112.
Trosborg, Anna
1987 "Apology strategies in natives/non-natives", Journal of Pragmatics
11: 147-167.
Transfer and proficiency in interlanguage apologizing 187
Vollmer, Helmut J. - Elite Olshtain
1989 "The language of apologies in German", in: Shoshana Blum-Kulka
- Juliane House - Gabriele Kasper (eds.), 197-218.
Wolfson, Nessa
1989 Perspectives. Cambrigde, MA: Newbury House.
Wolfson, Nessa - Elliot Judd (eds.)
1983 Sociolinguistics and language acquisition. Rowley, MA: Newbury
House.
Wolfson, Nessa - Thomas Marmor - Steve Jones
1989 "Problems in the comparison of speech acts across cultures", in:
Shoshana Blum-Kulka - Juliane House - Gabriele Kasper (eds.),
174-196.
Face-threatening acts
My grade's too low:
The speech act set of complaining
Beth Murphy - Joyce Neu
1. Background
Since Hymes' (1972) introduction of the concept of communicative
competence, there has been a heightened awareness that one of the most
important goals in acquiring a second language is learning the appro-
priate use of linguistic forms. That is, the acquisition of communicative
competence involves knowledge of the form as well as the appropriate
context in which to say it. Yet although many second language re-
searchers recognize the importance of communicative competence in
second language acquisition, research on the specific components of com-
municative competence remains scarce. In fact, Hymes (1972) called for
a broader framework for linguists to examine the social aspects of
language as well as the structural. In his definition of communicative
competence, Hymes suggested that we look not only at grammaticality,
but also at the feasibility and the appropriateness of utterances. Through
his investigation of speech acts, Searle (1971; 1975) also demonstrated
the importance of context in examining the relationship between
language and action. Speech acts are the acts we perform when we speak:
e. g., congratulating, thanking, requesting. These acts, in order to be per-
formed satisfactorily, must take place in a specified context of situation.
Thus, research on speech acts is crucial in that it can provide us with the
social context and the appropriate sociocultural rules surrounding native
speaker utterances.
This chapter is concerned with one aspect of communicative
competence - the performance of the speech act set
l
of complaint.
Performing this speech act set, which usually involves a face-threatening
act, appears to be challenging even for native speakers of English who
often pre-plan how they will go about making a complaint. Given the
delicate nature of this speech act set, non-native speakers of English may
unintentionally express inappropriate complaints. These complaints, in
turn, may not only sound non-native, but may serve to alienate the inter-
192 Beth Murphy - Joyce Neu
locutor. The objectives of the study reported in this chapter are to isolate
the components of the complaint speech act set as performed by
American native speakers of English and by Korean non-native speakers
of English and to analyze how American native speakers perceive Korean
non-native speakers' performance of complaint speech act sets.
2. Speech act production
The foundation for the study of speech acts was laid by Austin (1962)
and Searle (1971; 1975). A departure from Chomskyan linguistics, their
work situated language within a social context, providing us with a
greater awareness of the importance of sociolinguistic knowledge in the
production of speech. By focusing on speech acts rather than on isolated
sentences, Austin found that a class of verbs, called performative verbs,
function as the accomplishment of an action by their being spoken. That
is, by uttering "I apologize," the act of apologizing is performed.
Contributing to the development of speech act theory, Searle (1971)
defined speech acts as the smallest units of rule-governed meaningful
communication.
Researchers such as Manes (1983) and Wolfson (1983) have drawn
from Austin and Searle's development of speech act theory and applied
it to the analysis of a specific speech act - the compliment. The studies
by Manes and Wolfson reveal that American English speakers com-
pliment on appearance, new acquisitions, and effort. These serve the
functions of maintaining solidarity and reinforcing social values.
Wolfson's 1983 study further demonstrates that the status relationship
between the participants plays an important role in the topic of the
offered compliment.
Wolfson, D'Amico-Reisner, and Huber (1983) investigated the socio-
cultural rules of invitations in middle-class white American society. Their
study demonstrates that interpreting this speech act may be troublesome
for non-native speakers since invitations are ambiguous much of the time.
Unambiguous invitations are produced when the speaker refers to a time
and/or place/activity as well as requests a response from the addressee.
The invitation, "Do you want to go to the movies tomorrow night?"
contains all three components that make up an unambiguous invitation.
However, Wolfson, D'Amico-Reisner, and Huber found that these un-
ambiguous invitations occurred in only one-third of the data. Ambiguous
The speech act set of complaining 193
invitations, on the other hand, which provide for negotiation between
interlocutors, were found to be more representative of how native
speakers of English arrange for social commitments. These invitations
contain a "lead" that is a question or comment that opens up the
possibility for an unambiguous invitation to follow. For example, "Are
you busy tomorrow night?" is a lead that serves to establish the avail-
ability of the person before the issuance of an unambiguous invitation.
Thus, although leads often precede invitations, they do not in themselves
constitute an invitation. Consequently, the distinction between leads and
full invitations may result in misinterpretations between native speakers
and non-native speakers.
Cohen and Olshtain (1981) expanded the concept of the speech act in
their analysis of apologies. They found that semantic formulas, whether
in combination or alone, can be used to perform an act of apology. For
example, a speaker may express an apology, "I'm sorry"; acknowledge
responsibility for a perceived wrong, "It's my fault"; offer a repair for the
wrong, "I'll pay for it"; promise forbearance, "It won't happen again";
or explain the situation, "There was a traffic jam." Because each of these
formulas is in itself a speech act, they make up the speech act set of
apology. In another study, Olshtain (1983) used this apology speech act
set as a framework for her intercultural research. She discovered that
some cultures preferred one or another formula, or combination or
formulas, to express an apology. American English speakers, for example,
tend to express an apology and follow it with an explanation of the situa-
tion whereas Hebre'Y speakers tend to give an explanation only.
Furthermore, Olshtain noted that these language-specific preferences may
cause a second language learner to sound inappropriate in the target
language. By providing just an explanation and no apology, Hebrew
speakers who transfer this formula will undoubtedly sound rude in
English.
While Cohen and Olshtain referred to the speech act set, Ferrara
(1985), drawing on van Dijk (1977) explained the need to talk about
macro speech acts. Although these studies concentrated on a single
speech act, Ferrara (1985) has argued that speech act theory must be
extended to capture' the core action of discourse. He claims that there is
a distinction between understanding the text semantically (what the talk
means) and understanding the text pragmatically (what the talk does).
According to Ferrara, capturing "what the talk does" involves identify-
ing the set of macro speech acts that "underlies the entire text and insures
its pragmatic coherence" (1985: 149). Although the macro speech act is
194 Beth Murphy - Joyce Neu
composed of myriad single speech acts, it can only be determined by
reference to the dominant speech acts in the text. Ferrara (1985) thus
argues for a broader unit of analysis, the macro speech act, as a way of
more effectively investigating the relationship between language and
action. Second language investigators have more commonly referred to
the "speech act set," a term that appears synonymous with Ferrara's
macro speech act and van Dijk's 1977 use of "macrostructure."
3. Speech act acceptability judgments
To ascertain whether the language-specific preferences noted by Olshtain
result in socially inappropriate utterances, native speaker judgments are
needed. To concentrate only on the productive aspect leaves the re-
searcher with a partial picture of the consequences of speech act per-
formance. In fact, Olshtain and Cohen (1983) highlight the importance
of sociolinguistic acceptability judgments in their discussion of method-
ological issues concerning the study of speech acts. They argue that
native speakers' judgments of non-native speakers' performance are
needed to determine whether or not communication has been successful.
In a later article, Cohen and Olshtain (1985) again focus on the use of
acceptability judgments as one way of capturing and examining speech
act behavior more effectively. The study reported in this chapter has used
acceptability judgments as a critical component necessary to expand our
understanding of how and when non-native speakers fail to communicate
effectively. Cohen and Olshtain discuss ways in which non-native
speakers' performances of the apology speech act set may be deviant.
They say that this deviance may be due to "a lack of compatibility
between [the] speaker's intent and [the] hearer's standards of accept-
ability" (1985: 178). The conclusion is that we must investigate perfor-
mance both from the speaker's perspective and from the listener's as
well.
Although few in number, some studies have concentrated on how non-
native speakers' production is perceived by native speakers. Olshtain and
Blum-Kulka (1985), for example, conducted a study in which 172 native
speakers of English, 160 native speakers of Hebrew, and 124 non-native
speakers of Hebrew judged the appropriateness of request and apology
strategies in Hebrew. They found that as the length of stay in the target
speech community increased, non-native speakers' acceptability judg-
The speech act set of complaining 195
ments became increasingly similar to native speakers' judgments. Carrell
and Konneker (1980) compared politeness judgments of American
English native speakers and non-native speakers of English from different
language backgrounds. The subjects judged and ranked eight different
request strategies in English in three specially contextualized situations in
terms of levels of politeness. Their study revealed that although there was
a high correlation in their politeness judgments of native speakers and
non-native speakers, the non-native speakers of English perceived a
greater number of levels of politeness than did the native speakers. The
researchers concluded that the English as a second language learners'
greater number of politeness distinctions may be a result of their over-
sensitivity to syntactic-semantic features.
Few studies appear to have examined speech acts both from the
speakers' and the listeners' perspectives. The present study examines the
productive aspect of the complaint speech act set as performed by
American native speakers and Korean non-native speakers of English.
The listener perspective was investigated through the use of American
native speaker acceptability judgments of the Koreans' performances.
4. Rationale and research questions
The present study has two objectives: First, to compare the components
of the speech act set produced by American native speakers and Korean
non-native speakers of English; and second, to ascertain how these speech
act sets were judged by native speakers on a number of factors (listed
below). The questions are as listed in (1 a)-(8).
Production:
(1 a) Given the context of expressing disapproval to a professor (i. e.,
someone of higher status), will A m e r i ~ a n native speakers of
English produce a complaint speech act set?
(1 b) Give the same context, will Korean non-native speakers of
English produce a complaint speech act set?
(2a) If American native speakers produce a complaint speech act set,
what are the components of this set?
(2b) If Korean non-native speakers of English produce a complaint
speech act set, what are the components of this set?
196 Beth Murphy - Joyce Neu
(3) Do the speech act sets of complaint produced by American
native speakers of English and Korean non-native speakers of
English differ? If so, how?
Judgments:
(4) Will American native speakers of English judge the Korean non-
native speakers' productions as aggressive?
(5) Will American native speakers of English judge the Korean non-
native speakers' productions as respectful?
(6) Will American native speakers of English judge Korean non-
native speakers' productions as credible?
(7) Will American native speakers of English judge the Korean non-
native speakers' productions as appropriate?
(8) Will American native speakers of English say that, given the
same situation, they would not produce the same kind of
complaint as the Korean non-native speakers of English?
5. Methodology
To answer the questions (1a) through (8), a two-part study was under-
taken to analyze both the production of the complaint speech act set by
native speakers and by non-native speakers and to examine native
speakers' judgments of the non-native speakers' productions.
5.1. Subjects
For the productive part of the study, 14 male American graduate students
and 14 male Korean graduate students participated (hence the use of the
pronoun he/his throughout). All were graduate students in various
departments at Penn State University. Twenty-seven American students
recruited from an advanced undergraduate course in American-English
phonetics participated in the receptive part of the study-the acceptabili-
ty judgments. Twenty three of these subjects were undergraduates; four
were graduate students. All of these subjects received extra credit in their
course for their participation in the study and subjects in both parts of
the study signed informed consent forms agreeing to participate in the
study.
The speech act set of complaining 197
5.2. Instruments
The speech act data were collected via an oral discourse completion task
that consisted of a hypothetical situation typed on a sheet of paper. The
situation placed the subject in the position of a student whose paper had
been unfairly marked by a professor. The subject was directed to "talk"
to the professor (see Appendix A for the situation).
The instrument in the acceptability judgments part of the study
consisted of a questionnaire composed of ten "yes-no" questions and one
open-ended question. Of the ten "yes-no" questions, five were distractor
items. The five "yes-no" questions that were analyzed measured the
native speakers' perceptions of the Korean non-native speakers' responses
in the context of a student expressing disapproval about an undeserved
grade (see Appendix B for the questionnaire and Appendix C for the
written transcripts of the complainer and the criticizer). The open-ended
question asked the subjects to explain their answers to one question: "If
you were the student in this situation, would your approach be different
from the student you've just heard? Please explain your answer for both
speaker-student 1 and speaker-student 2."
5.3. Procedures
For the first component of the study, the subjects were instructed to read
the instrument carefully and to voice their reaction into a tape recorder.
They were told to give as complete a response as if they were actually
conversing with the professor. All of the subjects indicated that they had
either experienced the situation or could easily imagine it. Each subject
was alone during the discourse completion task. The speech data were
then transcribed into written form.
In part two of the study, the questionnaire was distributed to the 27
American students. The subjects were told they were going to listen to
two non-native speakers of English role-playing as students who are
expressing their disapproval to a professor about an undeserved grade.
Because it was found that most Korean non-native speakers produced a
criticism rather than a complaint speech act set, to determine the salience
of this difference for native speakers of English, one complaint and one
criticism were selected for native speaker judgment. For ease of reference
in this chapter, the Korean non-native speakers who complained and who
criticized are referred to as "the complainer" and "the criticizer,"
198 Beth Murphy - Joyce Neu
respectively. To validate the researchers' assumption that the com-
plaint and the criticism were indeed distinct speech act sets, the 27
American students were asked to evaluate the Korean non-
native speakers' responses. Seventy-four percent of the Americans
judged the complainer's response to be a complaint and the criticizer's
response to be a criticism. The American subjects were then instructed
to answer the questionnaire by judging only the content of the response
in terms 'of the situation, not the grammar or tone of voice. They
first listened to and judged the complainer's response. They then heard
and judged the criticizer's performance. The American native speakers
were given a written transcript of each speaker to distract them
from focusing on the Korean non-native speakers' grammar or pro-
nunciation problems. Both of the non-native speakers' responses were
randomly selected from the speech data base obtained in the first part of
the study.
5.4. Methods for data analysis
The speech data elicited for the first part of the study were examined
using Cohen and Olshtain's (1981) definition of speech act set. While
Cohen and Olshtain investigated the semantic formulas of the speech act
set of apology, this study attempted to identify the semantic formulas of
the speech act set of complaint.
For the second part of the study, a .05 alpha level of significance was
set for a Chi-square analysis of the American students' responses to the
five "yes-no" questions. For all differences between responses analyzed,
the Yate's Correction for Continuity was used.
The free responses to the open-ended question were compiled to
examine the reasons why the subjects' approaches would or would not
differ from those taken by the two Korean non-native speakers.
6. Results
The first research question asked if American native speakers, given the
situation presented in this study of a student expressing disapproval
about a grade received on a paper to a professor, would produce a com-
The speech act set of complaining 199
plaint speech act set. The answer appears to be yes. A description of this
speech act set is in section 6.1.
The second research question asked if Korean non-native speakers,
given the same context, would also produce a complaint speech act set.
Only three of the 14 Korean non-native speakers-produced speech act
sets contained a component that could be termed a complaint. Therefore,
most Korean non-native speakers do not produce a complaint speech act
set in this situation.
6.1. Semantic components of the speech act set of complaint/criticism
The speech act sets produced by the American native speakers and the
Korean non-native speakers contained certain core components, with the
notable difference between American native speakers and Korean non-
native speakers occurring in the complaint element of the speech act set.
An explanation of purpose was provided by each of the 14 American
male subjects before actually stating the complaint. That is, each
American subject explained the purpose of his presence to the inter-
locutor (the professor). Give the context of venturing into the office of his
professor, the student must explain why he is there. Thus, a student may
say, "Excuse me, I just dropped by to talk about my grade." Through the
explanation of purpose, the student sets the stage and gives cause for his
presence. Common examples from the Americans' responses include
samples (a) and (b).
(a) Subject A5
2
: Hello, Professor Filano. Vh, I got my paper back here
and after looking through it ...
(b) Subject A7: Vh, Dr. Smith, I just came by to see if I could talk about
my paper.
An explanation of purpose was produced by 13 out of the 14 Korean
subjects. Their responses paralleled those of the American subjects in that
the focus is also on the grade/paper as the cause of their presence. Thus,
the Korean non-native speakers appear to express the explanation of
purpose appropriately and in much the same manner as the American
native speakers, as shown in samples (c) and (d).
(c) Subject K1: Good afternoon Professor. Vh, I have something to talk
to you about my paper ...
200 Beth Murphy - Joyce Neu
(d) Subject K4: Hello, Dr. Brown. I came by to ask if some question
about the papers I handed-I hand in to you.
A complaint was produced by each of the American subjects only after
they explained their purpose. A complaint is initiated when the speaker
perceives he has been treated unfairly by the instructor. Moreover, in this
context of a student-professor relationship, the Americans are keenly
aware of the status difference and are thus constrained from producing a
criticism. Thus, to express disapproval, it appears that the most socially
appropriate option available to the Americans is to complain. The topic
of all of the Americans' complaints concerns the grade of the paper/test,
as shown in samples (e) and (f).
(e) Subject A1: I think, uh, it's my opinion maybe the grade was a little
low.
(f) Subject AS: I was kind of upset with my grade. Uh, I know that
a lot of problems are mine but there are certain areas that I uh
wasn't totally in agreement with what you said were my problems
here.
A complaint seems to occur in only three of the 14 Korean subjects'
responses. These three complaints are similar to the Americans' in that
they reflect concern and discontent with the grade, as shown in samples
(g) and (h).
(g) Subject KS: Professor, I little bit disappointed uh in my grade ...
(h) Subject K9: I was very upset at uh seeing my grade on my research
paper.
However, the other 11 Korean non-native speakers appear to have
produced not complaints, but criticisms. This appears to be a serious
deviation from the speech data obtained from the American native
speakers. Thus, the criticisms may indicate inappropriate behavior on the
part of the Korean non-native speakers. Examples of criticisms produced
by the 11 Korean subjects are shown in samples (i) and (j).
(i) Subject K6: But you just only look at your point of view and uh you
just didn't recognize my point.
(j) Subject K8: ... but I think my point of view is more relevant to the
subject and that has to be emphasized ...
The speech act set of complaining 201
A justification occurs in each of the American native speakers responses.
The justification demonstrates that the complainer can support his
claims, that he can explain the reasons why he is complaining. In this
sense, the justification lends credence to the complaint. It appears that if
a student complains to a professor, he must justify it by providing
well-developed reasons as support. In this situation, the topic of 13
out of 14 of the Americans' justifications focused on the time, work,
or effort that went into producing the paper, as shown in samples (k)
and (1).
(k) Subject A2: I put a lot of time and effort in this ...
(1) Subject A3: Uh, but a great deal of work went into that ...
A justification was similarly given by all of the Korean non-native
speakers and like the American native speakers, most of the Korean non-
native speakers' justifications referred to the amount of time, effort, or
work put into the paper. Examples of the Korean non-native speakers'
responses include samples (m) and (n).
(m) Subject K5: ... and urn I read my articles and readings and also I uh
put down all your lectures and so I tried to answer to my best ...
(n) Subject K6: ... I-I spent a lot of times to do research on this
paper ...
A candidate solution: request appears to be the last speech act produced
by all of the American native speakers. A candidate solution: request is a
proposed course of action offered by the complainer as a way to resolve
the problem. From the data, it is also evident that the American native
speakers propose the candidate solution in the form of a request rather
than as a demand. Thus, an asymmetrical status relationship between the
complainer and the recipient appears to constrain how American native
speakers can offer a solution to a professor. In this situation, all of the
Americans' candidate solutions involved requesting the professor to
reconsider an/or discuss the grade/paper. Representative examples are
shown in samples (0) and (p).
(0) Subject A9: I would appreciate it if you would reconsider my grade.
(p) Subject A12: ... so, I'd like to perhaps set up a time when we can get
together and discuss in detail what you felt- what was missing or
what was lacking.
202 Beth Murphy - Joyce Neu
A candidate-solution: request was produced by 12 of the Korean non-
native speakers. These requests are similar to those of the American
native speakers in that the focus is on asking the professor to reconsider
the grade/paper, as shown in samples (q) and (r).
(q) Subject K9: And I hope you can reconsider my grade and urn I hope
you can reevaluate my research paper.
(r) Subject K10: So, I honestly ask you to reconsider my paper ...
Two of the Korean non-native speakers, however, demanded a candidate
solution instead of requesting one, as shown in samples (s) and (t).
(s) Subject K7: Your grading is not fair and uh so it must be changed.
(t) Subject K8: I hope you give me some plausible explanation where
defects or handicaps that the points that I missed that should be
covered.
Tables 1 and 2 represent a summary of the complaint and criticism speech
act sets produced by American native speakers of English and Korean
non-native speakers of English, respectively.
Thus, the first research question is answered affirmatively: American
male native speakers of English produce a complaint speech act set in
response to the context of expressing disapproval to a professor. All 14
Table 1. Complaint speech act set produced by American NS of English
Components
Explanation of
purpose (EOP)
Complaint
(CaMP)
Criticism (CRIT)
Justification
(JUST)
Candidate
solution: request
Candidate
solution: demand
Frequency Examples
of Use
100 % I just came by to see if I could talk about my
paper.
100 % I think, uh, it's in my opinion maybe the grade
was a little low.
0%
100 % I put a lot of time and effort into it.
100 % I would appreciate it if you would reconsider
my grade.
0%
The speech act set of complaining 203
Table 2. Criticism speech act set produced by Korean NNS of English
Components Frequency Examples
of Use
EOP 93% I have something to talk to you about my
paper.
CaMP 21 0/0 I little bit disappointed uh in my grade.
CRIT 79% But you just only look at your point of view
and uh you didn't recognize the my point.
JUST 1000/0 I spent a lot of times to do research on this
paper.
CS: request 79% I honestly ask you to reconsider my paper.
CS: demand 21 0/0 Your grading is not fair and uh so it must be
changed.
American native speakers produced a complaint speech act set. The data
indicate that the complaint speech act set consists of four basic compo-
nents: Explanation of purpose, complaint, justification, and candidate
solution: request.
The second research question that asks whether Korean non-native
speakers of English will produce a complaint speech act set given the
same context is not answered in the affirmative. Rather, the findings of
this study indicate that 11 out of 14 Korean non-native speakers (79 %)
produce a criticism in the slot where the American native speakers
produce a complaint. This deviation is a critical one in that the complaint
speech act is a fundamental element of the entire speech act set. Excluding
the complaint component and substituting a criticism component may
change the illocutionary force of the set into a criticism. In the following
section, we will describe the linguistic formulas that differentiate the
complaint from the criticism component.
6.2. Linguistic features of the speech act set of complaint/criticism
When faced with expressing disapproval to a professor about an un-
deserved grade, each of the 14 American native speakers produced a
204 Beth Murphy - Joyce Neu
complaint. A complaint serves the function of assuming some of the
responsibility of the perceived wrongdoing and may contain any of the
characteristics listed in I-V.
I. Use of pronoun "we" in two primary ways: 1) to indicate that both
parties share the blame: e. g., "I know we have a different point of
view on this subject" (Al), "I know through class discussions we've
disagreed with each other ... " (A2); and 2) as a way of negotiating
the problem (the grade): e.g., "I hope that we could sit down and
discuss the paper ... " (A12), "I'd like to perhaps set up a time when
we can get together and discuss in detail what you felt was
missing ... " (A12).
II. Use of questioning to ask for advice, for permission to explain one-
self, or to get the listener to reconsider or discuss the problem.
Thirteen of the 14 American native speakers used the modals
"could" or "would" in their questions, thereby indicating politeness
or hesitation, for example, "I'd like any advice you could give me of
the problems I have ... " (AS), "Do you have a minute so that we
could go over the paper together?" (A14).
III. Depersonalization of the problem, transferring blame from the
professor to the problem. That is, the focus centers on "the paper,"
"it," "the grade" rather than on a person. Examples include "I feel
this grade may reflect a difference of opinion" (A9), " ... but it seems
like the grade and the effort kind of don't mix" (A8).
IV. Use of mitigators to soften the complaint. All of the American native
speakers used one or more of the following mitigators: "maybe,"
"just," "a little," "kind of," "perhaps," "really," "you know,"
"I don't know." In the complaint "I think uh it's just in my
opinion maybe the grade was a little low," virtually every word is a
mitigator.
V. Acceptance of partial responsibility for the problem: " ... and uh
perhaps it wasn't quite as polished as both of us would have liked,
but the content is there ... " (A3), "Uh, I know that a lot of the
problems are mine ... " (AS), "Uh, I'm wondering whether it was just
lack of explanation on my part or if you had overlooked a few things
which I had presented" (A8).
As stated earlier, most of the Korean non-native speakers did not produce
a complaint but rather a criticism. A criticism abdicates responsibility for
the problem and places the blame on another party, in this situation, the
The speech act set of complaining 205
professor. The data reveal the formulas I - III used in the Korean non-
native speakers' criticisms.
I. Use of second person + modal "should" that indicates that the
speaker is in a position to dictate the behavior of the listener (the
professor). Examples include" uh you should take account uh into
uh I'm a foreigner ... " (K7), " you should be equal to everybody"
(K11).
II. Personalization of the problem, placing the blame on the other, e. g.,
"But you urn decreased my grade ... " (K4), " ... but you just only
look at your point of view and you just didn't recognize the my uh
point" (K6).
III. Refusal to accept responsibility for the problem. For example, "My
point of view is more relevant to the subject and that has to be
emphasized than yours" (K8), "I don't think that uh in that in my
paper there is any fault or any errors" (K11), "I think my thought
my idea is totally correct" (K13).
From the data, then, complaint and criticism appear to be two distinct
speech act sets that contain different components. The speech act set of
complaint, as produced by American native speakers, allows the speaker
to assume partial responsibility for the problem. In fact, given the
asymmetrical status and power relationships in the situation analyzed in
this study, the American native speaker appears to be constrained to
accept responsibility as a matter of politeness and concern for the
ongoing relationship. By assuming partial responsibility, the complaint
facilitates a negotiation between the student and the professor. The
complaint speech act set, given an asymmetrical situation, exhibits
features such as 1) acceptance of partial responsibility, 2) depersonaliza-
tion of the problem, 3) use of questions, 4) use of mitigators, and 5) use
of the pronoun "we."
On the other hand, the speech act set of criticism functions to deny the
responsibility of the speaker for the problem. Consequently, the criticisms
given by 11 of the Korean non-native speakers place the blame for the
poor grade on the professor. The speech act set of criticism is character-
ized in this instance by 1) abdication of responsibility, 2) personalization
of the problem, and 3) use of second person + the modal "should." While
sharing many components, these two speech act sets contain such
different features that they appear to be two distinct speech act sets (see
Table 3).
206 Beth Murphy - Joyce Neu
Table 3. Complaint and criticism speech act sets: A comparis<;>n
Complaint
1. acceptance of partial responsibility
2. depersonalization of the problem
3. questioning techniques that use
modals "would" and/or "could"
4. use of mitigators
5. use of pronoun "V\Te"
Criticism
1. denial of responsibility
2. personalization of the problem
3. use of the modal "should"
7. Sociolinguistic acceptability judgments
This section presents the results of the American English native speakers'
acceptability judgments of a complaint produced by a Korean non-native
speaker (referred to as "the complainer") and a criticism produced by a
Korean non-native speaker (referred to as "the criticizer").
Analysis of the acceptability judgments data revealed significant
differences in the Americans' perceptions of the complainer and the
criticizer in terms of 1) the speakers' aggressiveness (Research Question
3); 2) the speakers' respectfulness (Research Question 4); the speakers'
credibility (Research Question 5); and the speakers' appropriateness
(Research Question 6) (see Table 4).
7.1. Aggressiveness
Significantly more Americans perceived the crItIcIzer to be aggressive
(X
2
=12.96, df =1, P < .01). Over 65 % of the Americans judged the
speaker who criticized the professor to be aggressive, whereas only 150/0
judged the speaker who complained to be aggressive. The answer to
research question 3 is yes: Korean non-native speakers who produce
a criticism in lieu of a complaint in this setting are perceived as
more aggressive than the Korean non-native speaker who produced a
criticism.
7.2. Respectfulness
A significant difference was found in native speaker perceptions between
the complainer and the criticizer (X2 = 14.84, df = 1, p < .01). While over
The speech act set of complaining 207
Table 4. American NS acceptability judgments of the complainer and the criticizer
QUESTION Com- Criti- X
2
(N=27) plainer Clzer
YES NO YES NO
1. Is the student aggressive? 4 23 18 9 12.9
150/0 850/0 670/0 330/0 6**
2. Is the student respectful? 23 4 8 19 14.8
85
0
/0 15% 30% 700/0 4**
3. Does the student present a credible 19 8 8 19 7.41
case for obtaining his goal? 70% 30% 30% 70% **
4. Is the student's response appropriate 20 7 11 16 4.85
for the situation? 74% 26% 41% 59% *
5. If you were the student in this 20 7 20 7 0.00
situation, would your approach be 74% 26% 74% 26% ns
different from the student you have
just heard?
*
p<0.05
**
p <0.01
85 % of the American native speakers judged the complaint as respectful,
about 30 % rated the criticism as respectful. Over 70 % judged the
criticizer as disrespectful, thus answering the fifth research question in the
negative: American native speakers do not perceive Korean non-native
speakers' criticisms as respectful in this setting.
7.3. Credibility
When asked, "Does the student present a credible case for obtaining his
goal?," the American native speakers' perceptions towards the com-
plainer and the criticizer were again significantly different (X
2
= 7.41,
df = 1, P < .01). Over 70 % of the native speakers judged the complaint as
credible and the criticism as not credible. The answer to the sixth research
question is negative: American native speakers do not perceive most of
Korean non-native speakers' productions as credible.
208 Beth Murphy - Joyce Neu
7.4. Appropriateness
A significant difference was also found in the native speakers' judgments
towards the complainer and the criticizer (X
2
= 4.85, df = 1, p < .05). A
considerably greater number of American native speakers rated the com-
plaint as more appropriate (74.1 %) than the criticism (40.70/0). Thus,
research question 6 is negative:. Most Korean non-native speakers'
responses to the situation were not perceived as appropriate by American
native speakers.
7.5. Differences in approach to the situation
The last research question regards whether the American native speaker
judges would take a different approach to the situation than had the
Korean non-native speakers they had just heard. No significant difference
was found in the American native speakers' perceptions of the two non-
native speakers, but they did respond that their approach would be
different: 74 % of the American native speakers said their approach
would be different; 26 % responded that their approach would not be
different.
7.6. Results of the open-ended question
The questionnaire asked the American native speaker judges to explain
their answers for one of the yes-no questions, the only one, in fact, that
turned out not to be significant in the difference between the complainer
and the criticizer.
For the complainer, numerous answers were given. No general theme
appeared in the responses. However, for the criticizer, 21 of the 27
American judges responded that if they were the student in that setting,
they would not have been so aggressive and critical of the professor. Nine
of the American native speakers selected one sentence produced by the
criticizer as highly inappropriate: " ... my point of view is more relevant
to the subject and that has to be emphasized than yours in this area- in
this specific area ... " One American wrote, "I would be much more
careful not to attack the prof or accuse him/her of grading (sic). It's
dangerous to say, 'my point of view is more relevant!' All these things
evoke a reaction that is negative, if it wasn't there before!" Another judge
The speech act set of complaining 209
wrote, "No prof would consider regrading his paper - he's too rude."
Ten of the responses indicated that the criticism given by the non-native
speakers not only challenged the professor, but would put him/her on the
defensive. One subject wrote, "His (the non-native speaker) approach
will certainly result with the professor responding in a defensive way; this
is going to hurt him in obtaining his goal." Another native speaker judge
stated, "He straight out challenges the prof which is not the right thing
to do in order to get a better grade."
8. Discussion
The findings of this study are that when placed in the position of expres-
sing disapproval to a professor about a grade, American native speakers
produce a complaint. Korean non-native speakers in this same situation
produce a criticism, not a complaint. Although the components of ex-
planation of purpose, justification, and candidate solution: request are
parallel in the American native speakers' and the Korean non-native
speakers' performances, the component of criticism instead of complaint
is produced in the Korean non-native speakers' utterances. An earlier
study of American native speakers (Murphy 1989) revealed that an asym-
metrical status relationship seems to prevent American students from
criticizing a professor face-to-face. In the context of a student com-
plaining to a professor about a poor grade, the criticisms produced by the
Korean non-native speakers demonstrate inappropriate sociolinguistic
behavior.
This study's analysis of the characteristics of the complaint and
criticism speech act sets provides some insights into why the Korean non-
native speakers' productions are a serious deviation from appropriate
behavior. The American native speakers, by presenting a complaint, not
only accept partial responsibility for the grade, but also facilitate a
potential negotiation between themselves and the professor. To achieve
their goal of getting the professor to reconsider the grade, the American
native speakers depersonalized the problem, incorporated politeness and
hesitation markers such as the use of modals and mitigators, and used the
inclusive pronoun "we." Using these t:lative speaker strategies as a basis
for comparison, the source of inappropriateness of the Korean non-native
speakers' productions becomes more apparent. By personalizing the
problem, by not accepting any part of the responsibility for the problem
210 Beth Murphy - Joyce Neu
and telling the professor what action he/she "should" take, the Korean
non-native speakers place the blame on the professor. Criticizing a
superior, when a complaint is called for, was perceived as aggressive,
challenging, lacking credibility, and inappropriate. It is unlikely that the
Korean non-native speaker who produces a criticism in this context will
achieve his/her goals. Given a setting in which a student disagrees with a
professor's evaluation, American native speakers judged the complaint
speech act to be acceptable; the criticism was perceived not only as un-
acceptable, but as a dangerous deviation from the accepted norm.
9. Limitations
The study focused on the tape-recorded production of a response to one
written situation by a small number (28) of male Korean non-native
speakers. While we believe that this open-ended type of oral discourse
completion task allowed the subjects to respond as they wished, there
was no interaction between speakers, as would happen in a normal
conversation. Ideally, data would be collected in settings in which com-
plaints would be naturally elicited. As others have noted, the difficulty
lies in how much data one would have to collect to assure the production
of an adequate number of the speech act under analysis. Another concern
is that we focused on only one situation - that of a student complaining
to a professor. This situation places the subject in an asymmetrical status
relationship where the subject is of lower status than the person she is
asked to complain to. Further research should investigate if there are
differences in complaints lodged against peers and persons of higher or
lower status. We would assume there would be. More data, using
different status situations, males and females, and non-native speakers
from different language backgrounds, would help to clarify the findings
of this study.
For the second part of the study that focused on American native
speaker subject judgments of the Korean non-native speakers' produc-
tions, all of the American subjects were from the same class at
Pennsylvania State University (Speech Communication 410 - a course in
Introductory Phonetics). This course is a requirement for all speech
communication majors at Pennsylvania State University and students in
the course are sensitized to non-native discourse in a way that most
Americans would not be. This may have led the American subjects to
The speech act set of complaining 211
distinguish between the non-native speakers' complaints and criticisms in
a way that others might not have. Another factor in the American native
speakers' judgments is that although the American native speakers were
told to judge only the content of the Korean non-native speakers'
responses, and not the grammatical accuracy or suprasegmentals such as
intonation, these may have influenced the American native speakers'
attitudes about the appropriateness of the non-native speakers' responses.
One alternative would be to give the American native speaker judges
transcriptions of the non-native speakers' responses rather than have
them listen to the taped responses.
10. Implications and conclusion
This study has implications for both the methodology of second language
research and English as a second/foreign language instruction. First, this
study demonstrates the importance of getting native speaker acceptability
judgments of non-native speaker speech data. Native speaker perceptions
about non-native speakers' productions provide support for researchers'
intuitions about appropriateness and acceptability.
The study also provides insight into what American students consider
appropriate behavior in an academic setting. Specifically, the speech data
produced by the Americans as well as the American native speakers'
judgments about non-native speakers' productions indicate that to com-
plain to a professor is acceptable and appropriate behavior; however, to
criticize a professor to her face is considered aggressive and disrespectful.
The distinction between a complaint and a criticism is clearly a salient
distinction for American native speakers although clearly a difficult one
for Korean non-native speakers to make. These "distinctive features"
make up speech acts and for speakers of different languages, different
features are salient. Where there is a disjuncture between what is salient
in one language and what is not in another, the non-native speaker either
"hears" a difference that doesn't exist in the target language or doesn't
"hear" a difference that native speakers do hear.
For teachers of English as a second/foreign language, these distinctive
features need to be taught. In the case of our study, the findings indicate
that international students (Koreans only?) need to learn how to assert
their rights in an academic setting without being perceived as threatening
or disrespectful. To assist international students in achieving this goal,
212 Beth Murphy - Joyce Neu
and by doing so, to avoid potential misunderstanding and even censure,
this study has isolated several appropriate native speaker strategies avail-
able to English as a second/foreign language learners who find themselves
needing to complain to a professor.
Speech act research has shown us how difficult it is for a non-native
speaker to acquire communicative competence in a second language. Out
of necessity do international students try to learn how to be perceived as
appropriate by their American peers and professors. In our "global
village," perhaps it is past time to also argue for the instruction of
American native speakers in the understanding of, and tolerance for, non-
native speaker productions.
Appendix A: Oral discourse completion task
Instrument A: Low status situation
You are handed back a paper by your professor. However, you are startled by
your grade and feel that you have been marked down for disagreeing with the
professor's point of view rather than on any flaws in your content and analysis.
You are particularly upset since you have spent weeks researching this paper and
feel the professor has ignored your effort through simple bias. You decide you
must speak to him/her about this. So, after class, you go to the professor during
office hours and say:
The speech act set of complaining 213
Appendix B: American NS Acceptability Judgment Questionnaire
Note: Please judge only the content of the Student Student
response, not the grammar or tone. 1 2
Remember to answer in the context of the
role situation (see instructions).
YES NO YES NO
1. Is the student aggressive (pushy)?
2. Is the student friendly?
3. Is the student respectful?
4. Is the student intelligent?
5. Is the student informal?
6. Is the student to the point?
7. Is the student confident?
8. Does the student present a credible
case for obtaining his goal?
9. Is the student's response appropriate
for the situation?
10. If YOU were the student in this
situation, would your approach be different
from the student you have just heard?
Please explain your answer to question 10, for both student 1 and student 2.
Continue on the reverse side if extra room is needed.
Student 1:
Student 2:
#1 #2
11. In an overall assessment of the speaker's CRIT COM CRIT COM
response, is the student criticizing or
complaining to the professor?
214 Beth Murphy - Joyce Neu
Appendix C: The Complainer and the Criticizer
The Complainer:
Hello, Dr. My name is ad uh I was very upset at uh seeing your my
grade on my research paper. I read that the I didn't fully follow uh your directions
during class but you know this is graduate level class and I think the crea-
creativity is as important as the other things like I um I gotta follow uh your
research directions.
The Criticizer:
Professor, can I ask you there ask you why I got this grade for this paper? I spent
several weeks at least several weeks to think about and planning and researching
for this paper and I devoted myself wholly to this project and I don't-don't under-
stand why you gave me the lower grade than I expected. I hope you give me some
plausible explanation or some defects or handicaps that the points I missed that
should be covered. (pause) I know you taughted the class a little bit different
point of views than uh but I think the my point of view is more relevant to the
subject and that has to be emphasized than yours in this area this specific area.
So, I took this um point and developed it to the fullest length and elaborated on
it. (pause) Thanks, I understand. But I don't think my work because a little bit
different from yours my point and my work has to graded lower than I expected
only because of-of the fact that uh my point of views is different from yours. I
think this a little bit unfair.
Notes
1. A speech act set is a combination of speech acts that, taken together, make up
a complete speech act. That is, it is frequently the case that one utterance alone
does not perform a speech act. Some examples are apologies and invitations
where several utterances are necessay for the intended illocutionary act to be
accomplished.
2. The coding system used was A1-14 for the 14 American subjects and K1-14
for the 14 Korean subjects.
References
Austin, John L.
1962 How to do things with words. Oxford: Oxford University Press.
Black, Max (ed.)
1965 Philosophy in America. London: Allen & Unwin Ltd.
Carrell, Patricia - Beverly Konneker
1981 "Politeness: Comparing native and nonnative judgments", Language
Learning 31.1: 17-30.
The speech act set of complaining 215
Cohen, Andrew - Elite Olshtain
1981 "Developing a measure of sociocultural competence: The case of
apology", Language Learning 31.1: 113-134.
1985 "Comparing apologies across languages," in: Kurt Jankowsky (ed.),
175-184.
Cole, Peter - Jerry Morgan (eds.)
1975 Syntax and semantics. New York: Academic Press.
Ferrara, Alessandro.
1985 "Pragmatics", in: Teun van Dijk (ed.), 137-157.
Gass, Susan - Carolyn Madden (eds.)
1985 Input in second language acquisition. Rowley, MA: Newbury
House.
Gass, Susan - Larry Selinker (eds.)
1983 Language transfer in language learning. Rowley, MA: Newbury
House.
Hymes, Dell
1972 "On communicative competence," in: John B. Pride - Janet Holmes
(eds.),269-293.
Jankowsky, Kurt (ed.)
1985 Scientific and humanistic dimensions of language. Amsterdam: John
Benjamins.
Manes, Joan
1983 "Compliments: A mirror of cultural values", in: Nessa Wolfson -
Elliot Judd (eds.), 96-102.
Murphy, Beth
1989 Influence of status on the elicitation of two speech act sets:
Complaint and criticism. [Unpublished seminar paper, Pennsylvania
State University.]
Ochs, Elinor
1979 "Transcription as theory", in: Elinor Ochs - Bambi Schieffelin
(eds.),43-72.
Ochs, Elinor - Bambi Schieffelin (eds.)
1979 Developmental pragmatics. New York: Academic Press.
Olshtain, Elite
1983 "Sociocultural competence and language transfer: The case of
apology", in: Susan Gass - Larry Selinker (eds.), 232-249.
Olshtain, Elite - Shoshana Blum-Kulka
1985 "Degree of approximation: Nonnative reactions to native speech act
behavior", in: Susan Gass - Carolyn Madden (eds.), 303-323.
Olshtain, Elite - Andrew Cohen
1983 "Apology: A speech act set", in: Nessa Wolfson - Elliot Judd (eds.),
18-35.
Pride, John B. - Janet Holmes (eds.)
1972 Sociolinguistics. Harmondsworth, England: Penguin Books.
Searle, John
1965 "What is a speech act?", in: Max Black (ed.), 221-239.
1975 "Indirect speech acts", in: Peter Cole - Jerry Morgan (eds.), 59-82.
216 Beth Murphy - Joyce Neu
van Dijk, Teun (ed.)
1977 Text and context: Explorations in the semantics and pragmatics of
discourse. London: Longman.
Wolfson, Nessa
1983 "An empirically based analysis of complimenting in American
English", in: Nessa Wolfson - Elliot Judd (eds.), 82-95.
Wolfson, Nessa - Elliot Judd (eds.)
1983 Sociolinguistics and language acquisition. Rowley, MA: Newbury
House.
Wolfson, Nessa - Lynne D'Amico-Reisner - Lisa Huber
1983 "How to arrange for social commitments in American English: The
invitation", in: Nessa Wolfson - Elliot Judd (eds.), 116-128.
Ethnographic interviewing as aresearch
tool in speech act analysis: The case of complaints
Diana Boxer
The ethnographic interview as a research tool has been virtually ignored
as a means of tapping the norms of communities both in research on
speech act usage among native speakers of particular languages and
research on non-native speaker pragmatic transfer. With the exception of
very few such studies (e. g., Katriel 1985), scholars have either concen-
trated on 1) analyzing spontaneous speech without the corroboration of
insight from members of the speech community; 2) analyzing speech
behavior based on only native speaker impressions derived from dis-
course completion questionnaires (Discourse Completion Tests) without
the corroboration of spontaneous speech; or 3) analyzing speech be-
havior through some combination of 1) and 2). However, Discourse
Completion Tests or more traditional survey techniques uncover only
intuitions that native speakers have about how they should speak. They
do not allow the researcher to discover the actual speech patterns of
native speakers. By combining the use of Discourse Completion Test
questionnaires with the analysis of spontaneous speech, scholars can
compare the use of a specific speech act with the canonical shape of the
patterning of that speech act in the minds of community members (Beebe
1985; Beebe - Cummings this volume). But while such a combination
allows for a system of checks on whether people indeed speak as they
think they ought to, it falls short of uncovering the tacit knowledge that
we all have about how and why we speak as we do.
Research into the speech behavior of native speakers of English is
important not only for establishing descriptions of how we perform
verbally in our day-to-day interactions with other native speakers, but
also for the purpose of making use of this baseline information in second
language learning. From the viewpoint of ethnolinguistic study, it is
important to discover what we as native speakers are doing when we use
certain speech acts in order to learn about our values and social system.
From the viewpoint of second language acquisition, such knowledge is a
necessary underpinning for the successful teaching of sociolinguistic
218 Diana Boxer
rules. It is now true that there are approximately as many non-native
speakers of English in the world as native speakers. We must have
effective communication among all English speakers if we are to avoid
unwanted misunderstandings. Thus the need for sociolinguistic descrip-
tion and the application of sociolinguistic findings to TESOL is more
critical than ever. As TESOL becomes a burgeoning endeavor inter-
nationally, both in English as a second language settings and English as a
foreign language settings, the sociolinguistic component of language
learning must be attended to. It is now imperative that descriptive
analyses of the sociolinguistic rules of English-speaking speech communi-
ties be carried out and widely replicated for increased generalizability and
applicability.
While it is true that the debate continues as to whether rules of
speaking can or should be taught in English as a second language/English
as a foreign language (Kachru 1988; Widdowson 1988), it is clear that
where the English as a second language is not a non-native institution-
alized variety, the ability to communicate with native speakers according
to native norms is important. Native speakers often forgive phonolog-
ical, syntactic and lexical errors as clear signs that a speaker does not
have native control of a language. Native speakers, however, typically
interpret sociolinguistic errors as rudeness rather than as the transfer of
different sociolinguistic rules (Ervin-Tripp 1972; Thomas 1983; Wolfson
1981; 1983; 1989).
There is a rapidly increasing population of adult learners studying
English as a second language in the United States. For these learners the
acquisition of sociolinguistic competence is at least as important as
linguistic competence for successful language learning. These adult
learners are finding themselves ever more involved in communication
with native speakers in settings including the academic, business,
diplomatic, scientific and technical spheres. Since the potential for mis-
communication is very great, it is of extreme timeliness that descriptive
analyses of native speaker sociolinguistic behavior be undertaken and
applied.
This chapter reports on baseline findings on rules for the realization
and underlying social strategies of a specific speech act sequence.
Corroborating evidence in the form of ethnographic interviews is dis-
cussed in depth. Described here are the results of two sets of interviews,
one structured and one open-ended, that were conducted with the aim of
tapping the knowledge of native informants about a speech event termed
"troubles telling" (Jefferson - Lee 1981; Jefferson 1984), "troubles
Ethnographic Interviewing 219
talk" (Tannen 1990) or "troubles-sharing" (Hatch 1992). More specifi-
cally, it delves into what is frequently the initiating move of the speech
event, namely the speech act of "griping" or what is termed here "indirect
complaining." The term "indirect complaint" is taken from the work of
D'Amico-Reisner (1985) on disapproval exchanges. Indirect complaints
are juxtaposed by D'Amico-Reisner with instances of direct complaint or
disapproval. Indirect complaints differ from instances of direct complaint
in that the addressee is not held responsible for a perceived offense.
Indirect complaint will be defined here as the expression of dissatisfaction
to an interlocutor about oneself or someone/something that is not
present. The exchange in (1) will illustrate the nature of the
griping/troubles-telling exchange:
(1) Two female graduate students in a course they both dislike.
a: I sat through yesterday's class with total non-comprehen-
sion!
b: Oh, yesterday was the worst!
In the exchange in (1) the speaker signals to the addressee her feelings by
using an indirect complaint about a class they are both taking. By agree-
ing, the addressee demonstrates to the speaker a mutual sentiment. On
this basis alone, an opening for further conversation and relationship-
building is provided in which they might go on to discover precisely their
common areas of interests, purposes or sympathies.
The spontaneous speech data of this study consists of 533 troubles-
telling exchanges that were tape-recorded or recorded in the form of field
notes. The methodological basis of data analysis was that of the ethno-
graphy of speaking (Hymes 1962). Categories emerged from the analysis.
Six categories of complaint responses emerged as major categories of
ways in which people in this community respond to indirect complaints:
1) Response or topic switch; 2) Questions; 3) Contradiction; 4) Joke/
teasing; 5) Advice/lecture; and 6) Agreement/commiseration.
Almost half of the responses to the initiating complaint moves fell into
one of six possible types of responses, that which was termed "commisera-
tion." The apparent preponderance of responses of this one type was
cause to speculate that the underlying social strategy of much complaining
behavior in the community is not negative but positive in nature. In order
to check this suspicion it seemed necessary to conduct an informal inter-
view with members of the speech community in order to tap their native
speaker tacit knowledge on the functions of this speech act/event.
220 Diana Boxer
Table 1. Indirect complaint responses
Category % of corpus Comments
1. 0 or topic switch 10.190/0 rhetorical, deliberate topic
switch, repeated
backchanneling
2. Question 11.700/0 request for elaboration;
challenge question
3. Contradiction 14.720/0 intimates, status unequals,
addressee wants distance
4. Joke/teasing 6.23% making light of situation
5. Advice/lecture 13.58 % platitudes, specific advice,
moralizing
6. Commiseration 43.58% agreement, reassurance,
exclamations, finishing
speaker's sentence
The interviews that are reported on here do indeed illustrate that com-
plaint sequences can often work toward establishing solidarity when they
are part of the troubles-talk event. Of course, this fact is easily apparent
for other more positively evaluated speech acts (e.g., compliments, offers,
invitations). However, the speech act with the semantic label "complaint"
is not typically thought of as rapport-inspiring speech behavior. Griping
sequences are examined here precisely because, while it is not intuitively
obvious, they do often fit into the category of speech behaviors that can
lead to the establishment of relationships between interlocutors.
2. Ethnographic interviewing as corroborating data for
speech act/event analysis
Since the ethnographic interview is a method of getting people to talk
about what they know - of discovering what human behaviors mean to
the individuals participating in those behaviors - it differs greatly from
the traditional interview or questionnaire in that it seeks to uncover not
only knowledge that is explicit but also knowledge that is tacit. The tacit
knowledge that informants have about behavior is brought out only after
a rapport has been established between the researcher and the informant.
Ethnographic Interviewing 221
Because ideal informants in studies of speech acts/events are socio-
linguistically naive, it is often possible to bring their tacit knowledge to a
state of explicitness through gentle questioning by the researcher within
the setting/context in which the speech behavior typically occurs. The
reverse is also true. Researchers themselves who are native speakers of the
community they are studying also possess knowledge that is tacit, and by
interviewing other native speakers, their own tacit knowledge can be
made explicit. Hence, by combining the researcher's own analysis of
spontaneous speech with information gleaned from native informants
through an ethnographic interview, a more complete analysis of the
specific speech behavior can be made than that which results from a
reliance on more traditional interviews or questionnaires.
The knowledge of how to go about conducting an ethnographic
interview cannot be gained simply by reading about how these are best
carried out. Although there are numerous sources available that deal
directly with ethnographic interviewing (e.g., Spradley 1979; Briggs
1986), the best way to learn how to conduct such an interview is by
doing one. I include here a description of a first attempt at such an
interview and how it went awry. Although as researchers we rarely hear
about what goes wrong in the process of collecting data, I believe it is
fruitful for methodological papers to detail such trials and errors. The
benefit accrues to those who are interested in employing new tools such
as this type of interview, but who do not have firsthand experience with
their use.
1
The first attempt at an ethnographic interview in this study was
initiated by careful thought about who the best informants would be,
what questions would be most fruitful, and how long each interview
should last. First, given that the best informants are those with whom the
ethnographer has built a rapport, ten people who were acquaintances of
the researcher but who knew little about the research under investigation
were chosen as informants. After locating ten informants who met these
criteria (five males and five females), interview times were scheduled with
the plan of asking eight structured questions of each informant. This was
followed by having the informants listen to and comment on a reading of
fourteen different sequences taken from the data. The intention was to
find out what the informants thought about various aspects of troubles-
telling as well as to get them to discuss what they saw as the social func-
tions within the actual data sequences.
Asking the very same questions of each informant turned out to be a
mistake. Although there seemed in theory to be nothing wrong with
222 Diana Boxer
conducting the interview in a structured manner, adhering to a fixed
agenda of questions led to an inflexibility that defeated the purpose of the
interview: To uncover tacit as well as explicit knowledge about troubles
talk. The fixed agenda did not allow the informants to lead the researcher
to unplanned questions based on their answers. A second mistake was in
trying to achieve too much within a relatively brief time allocation: Forty-
five minutes to one hour. Attempting to have each informant listen to
fourteen sequences of data as well as answer eight structured questions
imposed a heavy burden on the informants. One probable reason for the
feeling of imposition might have been precisely the structured nature of
the interviews. Apparently, the fact that they were more like interviews
and less like conversations caused the informants to want to "get it over
with." Many of the informants, knowing that they would also have to
give their impressions on spontaneous speech data, seemed to want to
quickly finish with the structured questions. Indeed, several rushed
through their responses, giving them less thought than they might have
had the demands of the interview been different. Another drawback of
this first attempt at interviewing was the desire to get brief, concise
answers to the structured questions so that they could be easily tabulated.
Because of this, the interviewees were less likely to elaborate. As a result,
their responses offered very little insight on the various issues brought
up.
Because there was no follow-up on issues brought up, the informants'
short replies were accepted as satisfactory. This resulted in an inability to
infer much about what they really knew. While their explicit knowledge
was tapped, a point was never reached in being able to ascertain what
they tacitly knew about indirect complaining. Spradley, in discussing
ethnography states: " ... a large part of any culture consists of tacit know-
ledge. We all know things that we cannot talk about or express in direct
ways. The ethnographer must then make inferences about what people
know by listening carefully to what they say, by observing their behavior,
and by studying artifacts and their use" (Spradley 1979: 9).
Given that the results of this interview were less than satisfactory, it
was apparent that a second interview would have to be conducted in a
more informal, open-ended manner. Of course, this would mean finding
ten new informants, as those who participated in the first interview were
no longer naive. Before re-doing the interview some very good advice
came forth from an ethnographer who had just been through a similar
process of trial and error. Micheau (personal communication) had the
following guidelines to offer: 1) Do as little talking as possible; 2) Pick
Ethnographic Interviewing 223
three or four areas to cover; introduce the issues and let the informants
talk; 3) Get the informants to do narratives and then chain their
narratives together.
Micheau's advice was followed in planning for the second interview.
The ten new informants were offered a short introduction on the purpose
of the interview. The introduction was an invitation to the informants to
talk about an instance/instances in which they were either the speaker or
addressee in a troubles telling exchange. The informants' tales were inter-
spersed with the researcher's questions on some of the issues under
investigation in this study. Thus the informants talked while their cues
were taken to ask gently probing questions at appropriate junctures in
their narratives. In such a manner the researcher is able to touch upon the
various issues of concern. One of the drawbacks in such an approach is
the inability to ask the very same questions of each informant. In this
case, not all ten informants addressed each issue. One of the advantages,
on the other hand, is that each informant is able to become introspective
in response to the questions asked of them - questions that emerge
directly from their narratives. Moreover, informants were more engaged
in their talk and hence more willing to carryon with their narratives.
Six key issues on which it seemed important to discover community
norms had emerged from the analysis of spontaneous speech data: 1) The
first issue dealt with perceptions of the delineation between direct and
indirect complaining. While a direct complaint is a face-threatening
activity (Brown - Levinson 1978), an indirect complaint does not hold
the addressee responsible for a perceived offense. Why is it, then, that
these two speech behaviors share the same semantic label? It would seem
that their underlying social functions differ greatly. 2) Given that there
appear to be various options in responding to the indirect complaint
(e.g., contradiction, topic switch, advice, commiseration), it seemed
important to gather informants' perceptions on how these differing
responses function to bring about differing outcomes in the troubles-
telling speech event; 3) The analysis of the spontaneous speech data indi-
cated vast differences in the way intimates respond to indirect complaints
as opposed to non-intimates. Therefore, it seemed important to gather
perceptions on how the social distance variable affects responses within
the troubles-telling exchange; 4) Gender emerged as a very strong
indicator of the propensity of addressees to respond with a reply that in
some way enhanced solidarity between the interlocutors: women were
twice as likely to commiserate as men; men were twice as likely to give
advice (this finding corroborates those of Tannen 1990). This difference
224 Diana Boxer
indicated a call to tap informants' perceptions on the gender variable in
troubles talk; 5) Related to the gender difference is perception on how
indirect complaints are used to open and support interactions; 6) A sixth
issue that emerged had to do with ethnicity and the propensity to use
indirect complaints as an initiating move in troubles talk. This sixth issue
was not previously dealt with in the research, but which appeared, from
an initial analysis of the speech data, to require attention. It was not
immediately apparent in the analysis of the data for this study that the
ethnic makeup of the 295 interlocutors was heavily Jewish. This ethnicity
factor gradually emerged throughout the data analysis and became a
clearer variable only after the quantitative analysis of the study was com-
pleted. One of the disadvantages of conducting interviews after under-
taking a quantitative analysis is precisely the possibility of missing an
important variable when looking at social distribution of any speech act.
Thus while ethnicity was not taken as an independent variable in the
quantitative analysis of the larger study (Boxer 1991; 1993a), it appeared
to have important implications in complaining behavior. These implica-
tions were discussed in some detail by the informants in the course of our
conversations, serving to shed light on the ethnicity issue from the native
informant's viewpoint.
Each of these issues was discussed with most of the informants during
our open-ended ethnographic interview. Two additional issues emerged
from the informants' narratives, and these were added to the above six.
They were: 7) Troubles talk in academia; and 8) The community's image
of troubles talk behavior. Table 2 indicates the various issues discussed,
the number of informants who commented on them, and a summary of
their comments. It also demonstrates how the results of such qualitative
information can be quantified.
The ten informants were either students, staff or faculty members at a
large university in the northeastern U. S. or alumni of this university
working outside of the campus setting. These were individuals with
whom the researcher had some kind of acquaintance or friendship. The
group consisted of: 1) Two graduate students, one male and one
female; 2) Three faculty members, all male; 3) Three staff members,
all female; and 4) Two alumni, one male and one female. Because of the
preponderance of faculty, staff and alumni, the average age of the
informants is estimated to be late thirties to early forties. Informants were
interviewed either in their offices on campus or were invited to the
researcher's home. A detailed examination of the results of the interviews
follows.
Ethnographic Interviewing 225
Table 2. Informants' views on troubles-telling issues
Issues # of informants Number in Comments
questioned agreement
1. D/IC distinction 9 9 All 9 said ICs are
more supportive
2 out of 9 said ICs
are less constructive
in problem solving
2. Responses
Agreeing 10 10 ICs seek agreement
o response 6 5 5 of 6 claim to ignore
chronic complainers
Disagreeing 6 5 Less likely with
strangers, friends, and
acquaintances
Advice 6 5 From men and among
intimates
3. Social Distance 7 7 All said they are more
a g ~ e e a b l e w/friends
and strangers than
intimates
4. Gender 8 8 Women complain
more and both sexes
use ICs more to other
women. Women are
more supportive
5. Ethnicity 10 9 9 strongly agreed that
Jews complain more
6. ICs as openers 8 8 ICs used as openers
w/strangers
7. ICs in academia 4 4 Academics like to talk
and ICs playa part
8. Image of complaint 8 8 All agree on negative
image, but often
positive reality
(IC: Indirect complaint; D: Direct complaint)
226 Diana Boxer
3. Perceptions of direct and indirect complaining
The informants were in general agreement that people within the com-
munity hesitate to participate in confrontational activities such as direct
complaining. Although most thought immediately of direct complaining
when hearing the term 'complaint', the majority later indicated that gri-
ping, grumbling, or indirect complaining is a more commonly occurring
activity than direct complaining within the community. However, only
two of the nine informants who addressed this issue even thought of in-
direct complaints upon being presented with the term "complaint." One
female informant, a departmental administrative assistant stated: "I
thought of direct complaints only. But I guess they're much more rare.
And in fact most people won't confront directly. I'm timid myself." A
male graduate student stated a similar idea: "One tends to think of the
complaint department' or the complaint box.'"
In sum, nine informants addressed the issue of the distinction between
direct and indirect complaints. Among these nine all agreed that indirect
complaining is less confrontational as a verbal activity. The informants
indicated that indirect complaints offer the complainer a way to let off
steam, a means of leaking the complaint to the responsible party, and
the possibility of establishing a common bond with the recipient of the
indirect complaint.
4. Perception of indirect complaint responses
4.1. Agreeing/commiserating as an indirect complaint response
The possibility of establishing a commonality through troubles-telling
exchanges is realized only if agreement is in some way obtained in the
response to the initial indirect complaint. Indeed the seeking of agreement
appears to be a widespread goal of indirect complaints. All ten infor-
mants mentioned either agreement as a goal of indirect complaint
utterances or the fact that indirect complaint/ agreement exchanges typ-
ically serve to forge a common bond between speakers and addressees.
Some of their thoughts on the subject are illustrated in examples (1)
through (3).
Ethnographic Interviewing 227
(1) I find a lot of people complain to me about other people
because they know I agree with them.
(2) I guess what causes the camaraderie among people IS the
empathy in agreeing.
(3) There's a common bond. That we're both subjected to the same
obstacles and by dealing with it we share something.
Community members perceived the expression of shared sentiments to be
one of the important elements of indirect complaint exchanges. Shared
feelings forge a common bond between interlocutors that may have the
potential for fostering feelings of closeness - feelings that may eventually
lead to the establishment or the deepening of a friendship between the
speaker and addressee.
Immediately striking is the question of why it is that we employ
negative evaluations to seek agreement or forge a common bond. The
notion that complaints are used as strategically positive speech acts is
counterintuitive. Several informants touched on this issue. Example (4) is
from one of them" a woman.
(4) If you walk over to someone and say "Gee, I'm so lucky" or
"What a wonderful day", they'd think you're weird.
4.2. Responding to chronic complainers
One of the several possible types of responses to indirect complaints is
offering no response or changing the subject. Slightly more than 10 % of
the responses in the corpus fell into this category. One situation in which
this response type frequently occurred was in reply to chronic complainers.
Six of the informants brought up the subject of chronic complaining,
and in so doing were able to shed some light on how they respond to
chronic complainers, offering insight into the response type that is here
referred to as 0 response/topic switch. Of these six informants, five
agreed that they tend to ignore or shun complaints from those they know
to overuse such speech behavior. Examples are given in (5) and (6).
(5) My mother was a constant complainer. I'd ignore it. My dad
tunes it out because all she does is complain, but nobody listens
to her.
228 Diana Boxer
(6) You let it roll off your back - ignore it. Which can lead to a
situation of "crying wolf."
Although this question was not asked of all ten informants, five out of the
six who did speak about this subject agreed they would typically respond
to a person whom they knew to be a chronic complainer with either no
response or an attempt to change the subject. It may indeed be the case
that there is a critical level of complaints that is surpassed by those
perceived as chronic complainers.
4.3. Advice as an indirect complaint response
The data from face-to face interactions in troubles talk indicated that
men gave advice almost three times as often as did women; women com-
miserated approximately twice as much as did men. There was general
agreement among the informants that advice is largely a male response to
indirect complaints. Whereas women are more likely to commiserate,
men are more likely to give advice, especially in response to indirect
complaints by female speakers. Of the six informants who specifically
addressed this issue, five thought that they tend to give advice as a com-
plaint response to intimates, particularly males to female intimates. Both
male and female informants indicated that men give advice because they
have been conditioned to think "more logically" and tend to want to
"solve problems." While women want to provide emotional support,
men want to get to the root of the problem and make it better.
5. The social distance variable
5.1. Disagreeing!contradicting as an indirect complaint response
The findings from the analysis of the spontaneous speech data indicated
that contradiction is a rare response to a troubles talk opener among
people who are not well acquainted. While approximately 15 % of the
responses overall fell into this category, among strangers the incidence of
contradiction responses was less than 10 % as compared with over 30 0/0
among intimates. Six of the informants discussed this response type, and
Ethnographic Interviewing 229
of these six, five agreed that they would be most likely to contradict
indirect complaints by intimates. Examples are given in (7) through (9).
(7) My husband and I disagree about everything. His typical
response is, "Oh, quit worrying.'" It's so unsupportive. It ticks
me off.
(8) With your spouse or kids. I might contradict my wife on some-
thing like the salad dressing in a restaurant, but I'd probably
agree with the guy at the next table.
(9) With intimates you're more likely to disagree, with friends you
want to be liked. With intimates your relationship is more
established, you can vent more honestly.
To summarize the informants' thoughts on disagreeing/contradicting,
there was general consensus that in this community the tendency is for
disagreement to occur as responses to indirect complaints from intimates.
The reason for such a disparity seems evident, and has been put forth in
Wolfson's Bulge theory (1988). It has to do with the relative certainty of
our relationships with our intimates as compared with the uncertainty of
our relationships with more distant friends and acquaintances. With this
latter group we are more conscious of being inoffensive (see Boxer 1993 b
for further discussion of how the "Bulge" is skewed).
The social distance variable was discussed with seven of the ten
informants. All seven agreed that they tend to behave differently
with strangers than with close friends or intimates. An example is given
in (10).
(10) Some of this stuff goes on in transient conversation, in a store,
with a salesperson, with a parking attendant. Particularly when
the weather is bad. It's just a way of communicating and you
might start with something like an indirect complaint. I think
people just want to be nice to each other - just want to talk to
each other.
Community members intuited that they behave differently with intimates
than they do with friends, acquaintances and strangers. They claimed to
be more agreeable with the latter than with those people closest to them.
One informant's statement summarizes: "On almost every level you'd
speak differently with people you know very well."
230 Diana Boxer
6. The gender variable
The eight community members questioned for this portion of the study
indicated that men and women behave very differently with respect to
both complaining and responding to complaints. There was total agree-
ment among the eight informants that women commiserate much more
than men and thus tend to be more supportive of speakers' complaints.
Examples are given in (11) through (13).
(11) I would tend to be less tolerant of a male that complained as
much as my wife does. I'm much more likely to give advice than
to commiserate. (male)
(12) Women listen better in general and have more concern, that
kind of thing. I would say men are more stubborn and "know
it all" (not me, of course) being dominant, you know. I think
women are more sensitive in general and therefore not as
concerned with being in control, like cut-throat. (male)
(13) Among women there are certain things we have in common,
like raising children, which is not always joyous. They change
your life and it's not all pleasant. (female)
Informants intuited that women participate more in indirect complaining
than men and individuals of both sexes complain more to other women
than men. This is due to the increased likelihood of obtaining a satis-
factory response, that is, one of agreement or commiseration. Women not
only hesitate to complain to men, but find more common ground with
other women about which to complain. One reason for this is that
women are the principal caregivers of children, a common cause of com-
plaint. This is true even for women of diverse ethnic and racial back-
grounds. Women always have the children or husband issue in common,
and these provide fertile subjects for indirect complaints. Whereas men
tend to want to tell the complainer how to go about solving the problem,
women just try to "be there" for the person complaining, providing the
feeling that they understand because they have had similar experiences.
Thus, for women, indirect complaining provides one important way for
female bonding to take place (Boxer 1993c).
Ethnographic Interviewing 231
7. Ethnicity
Recall that during the analysis of data for this project it became evident
that a large majority of the interlocutors recorded during the data collec-
tion were Jewish. As such, ethnicity gradually emerged as a factor
affecting the way indirect complaints are used in the community. Thus it
seemed clear that the ethnicity variable could not be ignored. Given this
ethnic slant to the data, the question that arises is the following: Is
indirect complaining/troubles-telling a result of the ethnicity of the
segment of the community studied? It seemed that the best way to
ascertain the effect of Jewish ethnicity on complaining behavior would be
to tap into the perceptions of the members of the community who served
as informants for this study.
Informants of mixed ethnic/religious backgrounds were chosen to get
to the heart of the interaction of Jewish ethnicity and complaining, just
as a mixed group of male and female informants were selected in order
to more clearly understand how gender and indirect complaints interact.
Some of their comments are in (14) through (16).
Jewish informants:
(14) I think complaining is part of the Jewish culture, but it's
partially true also that Jews are generally taught to, if they find
something wrong, to express that. Certain other ethnic groups
are more restrained.
(15) Maybe it has something to do with not getting a kinehura [a
non-literal translation of this Yiddish term means 'let the evil
eye not be cast']. I remember my mother and father would never
say "things are wonderful." It's been passed on in our culture.
My father in business would never tell anyone how well things
were going. It's definitely a Jewish thing, this complaining.
Particularly in the generation before. How much has been
passed on, I don't know.
Non-Jewish informant:
(16) My girlfriend [who is Jewish] complains a lot ... There's the
religious perspective too. There's sort of a saintly attitude like
you don't complain. I know tons of people like that in the
Christian religion, particularly women who take all kinds of
232 Diana Boxer
abuse and turn the other cheek. They have very difficult lives,
some of these women I teach, but they never complain.
There was 100 % agreement among the informants that Jews participate
heavily in complaining in general, and indirect complaining/troubles-
telling in particular. Some insights were given as to why this may be true.
It was stated by more than one informant that Jews are more vocal than
some other groups (e.g., White Anglo-Saxon Protestants) and tend to
express their feelings and emotions more openly. One informant, when
speaking about her friend of Italian origin, expressed the perception that
other ethnic groups may also be equally prone to complaining. Because
of the relative paucity of data from that particular ethnic group in the
corpus, no tentative statements can be made. Perhaps a replication of this
study in an ethnic Italian neighborhood would bear out the validity of
this intuition.
It may well be the case that indirect complaint/troubles telling
behavior is not merely an ethnic/religious issue but a regional one as well.
As the data for the present study is limited in regional scope, replication
of the study might be fruitful in different regions of the U. S. as well as
other ethnic communities.
8. Indirect complaints as conversational openers
The subject of how indirect complaints are used to open the troubles-
telling event and thus to open and support conversations and interactions
came up either directly or indirectly with eight out of the ten informants
in response to several different questions and in various parts of our
conversations. All eight informants said that they often use indirect com-
plaints to open conversations with strangers or little-acquainted addres-
sees. Some of their comments, in (17) and (18) illustrate.
(17) You see those innocuous ones [indirect complaints] in the
elevator all the time. It's a tremendous entree for more
conversation which IS a very safe entree to really express
feelings somehow.
(18) If you're waiting for service, for example. Someone behind me
complained and I sort of agreed, I was waiting too. She was
upset enough to say something to me and I agreed. It does
happen that way ... It's sort of an opener.
Ethnographic Interviewing 233
Informants repeatedly brought up the issue of creating a common bond
through the use of indirect complaints, and they noted that with strangers
particularly indirect complaints are frequently used to open and sustain
conversations. With interlocutors of extreme social distance (i. e.,
strangers, casual bystanders) it is not the serious complaints that are used
as openers but the small innocuous ones. These encompass such com-
plaints as those about the weather, the bus or train being late, or waiting
in lines. With little-acquainted interlocutors these small negative evalua-
tions serve to find some point of common interest. The informants indi-
cated that negative evaluations appear to be more commonly used as
openers than positive evaluations. It may be the case that positive
evaluations, such as compliments, threaten the negative face (Brown -
Levinson 1978) of addressees of extreme social distance more than
negative evaluations.
9. Troubles-telling and the academic community
There is indication that talk about troubles, or complaining, is a
prevalent verbal activity within the specific speech community studied
here. As stated earlier, indirect complaints were found to be ubiquitous
in spontaneous social conversations among students, faculty and staff
on campus. The ethnic and regional factors were discussed by some
of the informants in the preceding sections as possible contributing
factors. In addition to these factors, however, several of the informants
indicated that academia is fertile ground for indirect complaints due
to its inherent intellectual atmosphere. Examples are given in (19)
through (21).
(19) People at [this University] are great talkers and it goes along
with the intellectual atmosphere, leading to talking more than
action. There's an atmosphere of that.
(20) Students always complain about professors. Students bond by
giving each other the feeling that we're all in this together ... It's
cooler and more fashionable to be cynical. It's more hip.
(21) I have to watch myself, I think academics do because we tend
to be argumentative and at dinner parties where other people
aren't that way you can get into arguments. Other people aren't
as fond of that.
234 Diana Boxer
Indirect complaining may in fact be a commonly occurring speech
behavior in many communities; however, in the particular community
that is the focus of study here several factors combine to increase their
incidence of occurrence. The ethnic variable has been discussed as one
important contributing factor. The regional factor, that the University is
a large, urban one in the northeastern part of the country, may in fact be
a second contributing factor. Last, but perhaps not least in importance,
the simple fact that it is an academic community may be of significance
in the ubiquity of indirect complaints.
10. The negative image of complaining
The information gleaned from the ten informants offered strong indica-
tion that they perceive indirect complaints as a normal part of everyday
communication. These community members indicated that indirect com-
plaints serve to open and support conversations, that indirect complaints
are often used with the purpose of seeking agreement, and that in
obtaining agreeable responses indirect complaints aid in forging a
common bond between speakers and addressees. These perceptions,
however, run counter to the image community members have of "com-
plaining." In order to grapple with this contradiction, eight of the infor-
mants were asked to give their views on their image of complaining.
Examples are given in (22) through (25).
(22) Everybody complains. I can't imagine not being able to do that,
how you feel inside. But when you say the word 'complain' it
seems so negative.
(23) Oh [it's] negative, very negative. But what we're talking about
isn't so negative. You're bonded with the other person who
agrees with you.
(24) ... It's not a good thing to do. But people do it all the time.
Especially when you look at the small gripes... Most
complaints serve a different purpose from what you tradition-
ally think of.
Ethnographic Interviewing 235
Can you think of a better term for it?
(25) "Complain" may be a little too strong for it. Maybe just like
"venting. "
Of the eight informants directly questioned about the image that com-
plaint conjures, all were in agreement that it definitely has a negative
image. Notwithstanding this image, each had something to say on how
what we had been discussing did not seem so negative. Part of the
problem seems to lie in the fact that the semantic label "complain" covers
a broad spectrum of negative evaluation. The term "complaint" or
"indirect complaint" may be somewhat misleading. Troubles-telling,
troubles-talk, or troubles-sharing all refer to the larger speech event for
which indirect complaints is typically the initiating speech act. These
terms seem to be more adequately descriptive and less connotationally
negative than "indirect complaint."
What is so ubiquitous in the ordinary social conversation within the
community are not those serious complaints about things that adversely
affects people's lives but the commonplace small gripes about things not
working right, the weather not being fair, or the professor not teaching
well. It is these innocuous, everyday indirect complaints that are so
frequently used to make small talk, that aid to establish common ground,
and that may eventually evolve into getting to know another person
better.
11. Conclusions
The preceding discussion has had a two-fold focus. First, a comparison
has been drawn between two interviews undertaken for this study, one
structured and one open-ended in nature. The results of these have been
contrasted in order to illuminate the benefits of the ethnographic inter-
view in uncovering both tacit and explicit knowledge on community
norms regarding speech behavior. Second, the results of the ethnographic
interview were discussed in detail. The information gleaned from
informally interviewing ten members of the community yielded data that
served to elucidate several issues emerging from the analysis of spontan-
eous indirect complaint data.
The issue of how to go about uncovering the norms of a speech
community is exceedingly important in studies of the analysis of speech
236 Diana Boxer
in face-to-face interaction. Interviewing native speakers can yield data
that either corroborates or disconfirms the researcher's own analysis of
spontaneous speech data. However, the manner in which informants are
interviewed bears significantly on the information they offer to the re-
search. Informants should be relatively naive about the nature of the
study so that they can offer a fresh perspective on the issues in question.
Moreover, informants should be individuals with whom the researcher
has developed a rapport that enables them to speak openly and at length
on the subject. Individuals who are comfortable doing much of the
talking make the best informants. Specific questions should emanate
naturally from the narratives of the informants rather than be precon-
ceived by the researcher. Clearly, the researcher will have to decide on the
issues of importance during his/her discussions with informants; however,
cues about other issues typically come from what the informants have to
say. In such a way and only in such a way can tacit knowledge be un-
covered.
Ethnographic interviewing techniques can sometimes be learned only
by trial and error. The trials and errors reported in the above discussion
enabled a comparison to be made between the results of two interviews
that were very different in nature. By letting informants speak freely in
the second, informal interview and allowing the questions to emerge from
the informants' ideas, information was gleaned that added greatly to the
analysis of troubles talk and its functions in the community.
Whereas most members of the community immediately perceived
"complaint" as direct complaint, they indicated that indirect complaining
is a more prevalent and positive activity. The informants were able to
expound upon their perceptions about indirect complaints by talking
about how indirect complaints were employed in their own conversa-
tions. They indicated that indirect complaints are typically used as an
important part of "small talk" and with the underlying strategy of
obtaining agreement. With strangers and non-intimates, agreement or
commiseration is the preferred response, with the end of establishing
some kind of commonality, albeit brief, that makes encounters more
pleasant.
Women participate more in troubles-talk than men and are the
recipients of more indirect complaints, since they are perceived to be
more supportive in general than men. Men tend to give advice as an
indirect complaint response to both male and female speakers.
Informants generally agreed that it is a male tendency to want to take
control of such situations and solve the problem. Religion/ethnicity is
Ethnographic Interviewing 237
perceived to playa role in indirect complaint behavior, with some religi-
ous/ethnic groups participating in indirect complaint exchanges more
than others (e.g., Jews, Italians). Academia is also perceived to be fertile
ground for indirect complaint exchanges, as people involved in philoso-
phical issues of all kinds are more likely to take part in the sort of debate
that is sometimes characteristic of indirect complaining.
The community norm is that complaining is a negative speech
behavior. Notwithstanding this explicit perception, the tacit knowledge
of the informants is that indirect complaints more often than not have a
positive function in the everyday social conversation of members of the
community. Thus given certain requirements that are met by setting and
interlocutor characteristics, what is explicitly thought of as negative is
tacitly known to be a positive speech behavior.
These baseline findings on the underlying social strategies of troubles-
telling among native speakers of u.s. English have important implications
for applied linguists. Before we can go about teaching rules of speaking
to language learners, we need to have information on what native
speakers do with regard to specific speech acts and events. Research in
applied linguistics over the past twenty years has given us much infor-
mation on speech act realization and pragmatic transfer in rules of
speaking. What is suggested here is to go a step beyond the gathering of
data through questionnaires and/or recorded spontaneous speech. It is
possible to add a deeper dimension to our insight about speech behavior
through corroborating evidence gleaned from ethnographic interviews of
members of the target language speech community.
If indirect complaints function to open and sustain troubles-telling
sequences, and if, as indicated here, such speech behavior has a positive
underlying social strategy, this information can be extremely important
for language learners whose rules for the realization of the speech
act sequence differ from our own. As we have seen, the ethnographic
interview has an important role to play in the analysis of speech
behavior. Before we can apply findings on speech act/speech event
patterning, we need to tap into the tacit knowledge of native speakers
of the language. Only in this way can we hope to uncover the true
functions of language forms as they are used among members of a speech
community.
238 Diana Boxer
Notes
1. Details of the initial trial at ethnographic interviews for this project were
presented at the 1994 Conference on Pragmatics and Language Learning at
the University of Illinois (parasession on research methodology). The audience
welcomed remarks on what went wrong, indicating that such a description
could be very helpful to researchers who plan to use this method in their future
work.
References
Atkinson, Jean - John Heritage (eds.)
1984 Structures of social action: Studies in conversation analysis.
Cambridge: Cambridge University Press.
Beebe, Leslie - Martha Clark Cummings
1985 Speech act performance: A function of the data collection procedure?
Paper presented at the 18th Annual TESOL Convention, New York,
April.
Boxer, Diana
1991 A descriptive analysis of indirect complaint sequences among
speakers of American English. [Unpublished Ph. D. dissertation.,
University of Pennsylvania].
1993 a Complaining and commiserating: A speech act view of solidarity in
spoken American English. New York: Peter Lang Publishing.
1993 b "Social distance and speech behavior: The case of indirect com-
plaints", Journal of Pragmatics 19: 103-125.
1993c "Complaining and commiserating: Exploring gender issues", Text
13.3: 371-395
Briggs, Charles
1986 Learning how to ask: A sociolinguistic appraisal of the role of the
interview in social science research. Cambridge: Cambridge
University Press.
Brown, Penelope - Stephen Levinson
1978 "Universals in language usage: Politeness phenomena", in: Esther
Goody (ed.), 56-289.
D'Amico-Reisner, Lynne
1985 An ethnolinguistic study of disapproval exchanges. [Unpublished
Ph. D. dissertation, University of Pennsylvania].
Dascal Marcelo (ed.)
1985 Dialogue in interdisciplinary approaches. Amsterdam: John Benjamins.
Ervin-Tripp, Susan
1972 "On sociolinguistic rules: Alternation and co-occurrence", in: John
Gumperz - Dell Hymes (eds.), 213-250.
Fine, Jonathan (ed.)
1988 Second language discourse: A textbook of current research.
Norwood, NJ: Ablex.
Ethnographic 239
Gladwin, Thomas - William C. Sturtevant (eds.)
1962 Anthropology and human behavior. Washington, DC:
Anthropological Society of Washington.
Goody, Esther (ed.)
1978 Questions and politeness. Cambridge: Cambridge University Press.
Gumperz, John - Dell Hymes (eds.)
1972 Directions in sociolinguistics. New York: Holt, Reinhart and
Winston.
Hatch, Evelyn
1992 Discourse analysis and language education. Cambridge: Cambridge
University Press.
Hymes, Dell
1962 The Ethnography of Speaking, in: Thomas Gladwin and William C.
Sturtevant (eds.), 13-53.
Jefferson, Gail
1984 "On stepwise transition from talk about a trouble to inappropritely
next-positioned matters", in: Jean Atkinson and John Heritage
(eds.), 191-222.
Jefferson, Gail - John R. E. Lee
1981 "The rejection of advice: Managing the problematic convergence of
a "troubles-telling" and a "service encounter", Journal of
Pragmatics 5.5: 399-423.
Kachru, Braj
1988 "Teaching world Englishes", ERIC/CLL Bulletin 12.1: 1-4.
Katriel, Tamar
1985 "Griping as a verbal ritual in some Israeli discourse", in: Marcelo
Dascal (ed.), 367-381.
Spradley, John
1979 The ethnographic interview. New York: Holt, Reinhart and
Winston.
Tannen, Deborah
1990 You just understand: Women and men in conversation. New
York: William Morrow and Company.
Thomas, Jenny
1983 "Cross-cultural pragmatic failure", Applied Linguistics 4.2: 91-109.
Widdowson, Henry
1988 "Language, context and culture in the classroom", ERIC/CLL
News Bulletin 12.1: 6-7.
Wolfson, Nessa
1981 "Invitations, compliments and the competence of the native
speaker", International Journal of Psycholinguistics 24.4: 7-22.
1988 "The Bulge: A theory of speech behavior and social distance", in:
Jonathan Fine (ed.), 21-38.
1989 Perspectives: Sociolinguistics and TESOL. Rowley, MA: Newbury
House.
Frolll the addressee's perspective: Illlposition in
favor-asking
Myra Goldschmidt
1. Introduction
It has been well documented by now that conversation is more than static
discourse. Indeed, since Hymes published The Ethnography of Speaking
(1962) there has been a burgeoning interest in the way verbal discourse
is organized in particular speech communities (Gumperz 1971: 219-231;
Hymes 1974: 47; Labov 1972: 120-121). More and more, empirical
research is uncovering and making available patterns of speech behavior
within target language communities. However, immersed within verbal
repertoire is not only the speech behavior of a speaker, but also the
impact this speech behavior has on an addressee in terms of action and
reaction. In speech act research particularly, researchers are now looking
at the ensuing talk after the head act and finding that both the speaker's
perspective and the hearer's perspective (Blum-Kulka - Levenston 1987:
156-158) are important with regard to gaining insight into the rules of
speaking (Hymes 1972: 35-71).
It is the intent of this chapter to focus on a particular speech act, that
of favor-asking, and the impact this act has on an addressee in terms of a
specific reaction - imposition. This will be done by looking at the results
of a survey given to two hundred speakers of American English between
the ages of 17 and 69, in order to determine, first of all, how imposed
upon people feel when certain types of favors are asked, and second, what
the pedagogical implications for non-native learners inherent in such an
analysis might be. The major reason for devising a survey on imposition
was to address the "person spoken to" who is as important as the speaker
in this dynamic speech act. Very little research has been done from this
perspective, yet it is one that cannot be overlooked. As in the case of other
speech acts, favor-asking involves negotiated segments of conversation
which go beyond the philosophic tradition put forth by Austin (1962:
150-163) and Searle (1969: 23-24) where talk was dealt with as a static
product, rather than a dynamic process.
242 Myra Goldschmidt
With regard to the objectives being sought in doing this survey, they
were threefold. The first objective was to establish an underlying value
system in terms of what causes people to feel "put upon" in a particular
target community. The second objective was to see if the social variables
of role (in terms of student vs. non-student), gender, and age condition
imposition. These three variables were chosen based on the investigator's
native speaker intuition that any differences in the findings may be
evidenced by the role, age, and gender of the subjects. Specifically, it was
thought that men and women would perceive imposition differently
because they often perceive favors, themselves, differently (Goldschmidt
1993: 140). In addition, the age of the individuals, as well as whether or
not they were students, was intuited to condition perceived imposition.
Finally, it was hoped that the results from the first two objectives of the
determination of an underlying value system and the influence of social
variables would help to clarify any implications for TESOL regarding
how imposed upon people may feel when asked certain types of favors.
Favor-asking denotes a speech act in which the motive or purpose
behind the act itself is to get an addressee to do a specific task for the
speaker. In doing so, it encompasses different social role relations be-
tween individuals as well as varying intensities of volition. Specifically,
favor-asking is characterized by four defining features:
1. Favor-asking is a speech act which involves asking for something
"outside" of the addressee's daily routine.
2. Favor-asking entails doing activities that require some time and/or
effort on the part of the addressee or involves a good belonging to the
addressee.
3. Favor-asking entails no role-related obligation on the part of the
addressee to fulfill the task.
4. Favor-asking implies the notion of reciprocity in terms of a return
favor
(Goldschmidt 1993: 157-169).
Favor-asking also imposes (to some extent)1 on an interlocutor to ful-
fill a future act which is potentially costly to the addressee, thus is face
threatening (Brown and Levinson 1987: 13-15) and calls for redressive
action. This redressive action is often in the form of mitigation in order
to compensate for any imposition. Interestingly, Ervin-Tripp and Gordon
(1986) show that sensitivity to degree of imposition or intrusiveness in
requesting is often learned by school age, to the extent that children
become aware of what might be disruptive or difficult from the hearer's
Imposition in favor asking 243
point of view and vary their language accordingly. Furthermore, Brown
and Levinson (1987: 74- 84) state that imposition is ranked in propor-
tion to the expenditure of service (in the case of favor-asking), including
the provision of time. However, they caution that the ranking is deter-
mined by a whole set of variables which can alter the degree of impos-
ition, including whether actors have specific rights or obligations to
perform the act, or whether they have specific reasons for not performing
it, or whether the actors are known to enjoy being imposed upon in some
way.
Thus, it can be said that favor-asking, like most speech acts, is
comprised of participants engaged in a particular speech situation as a
means to an end. It is also one in which Hymes' (1972: 35- 71)
mnemonic SPEAKING (S =Situation; P =Participants; E =Ends; A =Act
sequence; K =Keys; I =Instrumentalities; N =Norms; and G =Genres)
would condition every favor-asking situation. In other words, one's mood
and physical circumstances at the time of asking could heavily impact on
how an addressee might react to the favor (Ervin-Tripp 1976: 50-63).
The data reported on in this chapter consider the extent to which
imposition is perceived depending on the gender, age and student/non-
student status of individuals.
2. Method
A survey was given to two hundred subjects from the Philadelphia area
(one hundred students at Villanova University between the ages of 17 and
21 and one hundred "non-students" between the ages of 20 and 69), who
rated the amount of imposition in five data-generated situations. (See the
"Results" section for an account of the five situations).2
The selection of the participants in the survey questionnaire was
according to the criteria of age (adult speakers) and native language, and
was accomplished in the following ways. First, for the student partici-
pants, colleagues of the investigator in the English department at
Villanova University were asked if they would administer the surveys to
their English classes during the 1991 spring and fall semesters. The
professors agreed to do so and obtained permission from the students to
administer the surveys. The completed surveys were collected and
returned to the investigator.
244 Myra Goldschmidt
Selection of the non-student participants was carried out over a period
of eight months. The investigator distributed the surveys to people
(friends, acquaintances and strangers) at various places in and around the
Philadelphia area. After a brief explanation about the survey, permission
was secured and the questionnaires were completed (out of the presence
of the investigator), collected and analyzed.
The favor situations used in the survey were chosen from a corpus of
200 favors. In deciding which tokens would be used, the investigator
picked those examples from the data which seemed to represent favors
typically asked and which seemed to depict a range in terms of degree of
imposition. She then showed the data to another sociolinguist who also
chose those tokens which seemed within the norm of favors she has asked
or has heard asked. The questions which were decided upon were those
which overlapped between the two linguists.
Each of the five favor situations required that the respondents rate
each favor according to imposition in terms of great imposition (GI),
some imposition (51), little imposition (LI), and no imposition (NI). The
term imposition, itself, was not defined but left to the discretion of each
informant as to what it meant. Imposition literally means "the act of
putting a burden on"; however, the definition varies from person to
person in terms of what it constitutes. In fact, some folk definitions
collected by the investigator included, "When you ask someone to do
something they're not ready to do. It can require time, m<:>ney, and/or
effort - though effort is open to interpretation"; "When I feel I'm being
taken advantage of, or when it's over and above a 'normal' favor in
'normal' circumstances, excluding an emergency, or if it were never
reciprocated"; and finally, "When I feel obligated to do something that I
don't really wish to do." In view of these differences of opinion, it was
decided not to impose a definition, but to leave it open.
The questionnaires were analyzed according to the student and non-
student populations, age and gender since it was thought that these
variables could be determinants with regard to the responses. A statistical
analysis for chi square and probability was undertaken for each of the
five situations (the alpha level was set at 0.05). After analyzing the results
which emerged from the data, reply profiles (or the proportion of replies
that fell into the four given categories) were established in terms of
imposition - what types of favors seem to incite the perceived feeling of
great imposition, some imposition, little imposition and no imposition?
From these profiles, tables were established to determine if there were any
recurring patterns with regard to imposition in favor-asking. Following
Imposition in favor asking 245
are the results of the survey which provide insight from the perspective of
the addressee in order to understand more fully the process of favor-
asking with a focus on imposition.
3. Results
The respondents to the questionnaire were divided into two groups, non-
students (N =100) and students (N =100). The age distribution of these
two groups is given in Table 1. For analytical purposes the non-student
group was further divided into three groups: < 32 years (N =26), 32-45
years (N =52), and> 45 years (N =22). Smaller ranges would have given
smaller numbers per group, making statistical evaluation difficult. The
student population was not further subdivided because of the very small
age range of 17- 21 years.
The gender of the two groups was also determined and is given in
Table 2. When comparing the two groups there were significantly more
(p > 0.001) female non-students compared to male non-students
(X
2
=19.3; df =1).
A statistical analysis of the five situations for chi square and
probability (p) was then undertaken. The following are the results:
Table 1. Age distribution of survey respondents
Mean Age
Std. Deviation
Median Age
Range
Non-Student (N = 100)
39.4 years
10.5 years
39.0 years
20-69 years
Student (N = 100)
18.4 years
0.9 years
18.0 years
17-21 years
Table 2. Gender distribution of survey respondents
Non-student Student Total
Female 78% (N=78) 48 % (N = 48) 63 % (N = 126)
Male 22% (N=22) 52% (N=52) 370/0 (N = 74)
Total N= 100 N= 100 N= 100
246 Myra Goldschmidt
1. Favor situation: Husband to wife (who was home from work because
of a snow storm) on his way to work: "If you have some time today, do
me a favor and shovel the driveway" (90 feet long). (Table 3.1)
Table 3.1. Ql by student
NI LI SI GI Total
Student 10(10%)
Non-student 17 (17%)
23 (23 %) 32 (32 %) 35 (35 %)
17(17%) 38(38%) 28(28%)
100(50%)
100(50%)
Total 27(13.5%) 40(20%) 70(35%) 63(31.5%) 200(100%)
NI = No imposition, LI = Little imposition, SI = Some imposition, GI = Great
imposition. Xl= 4.0; df =3; p =0.26
No significant differences were found between the responses of the
students and non-students in this situation. (Table 3.1)
Table 3.2. Ql by gender
NI LI SI GI Total
Female 20 (15.9%) 24 (19.1 %) 46 (36.5 %) 36 (28.5%) 126 (63 %)
Male 7 (9.5%) 16 (21.6%) 24 (32.4%) 24 (36.5 %) 74 (37%)
Total 27 (13.5%) 40 (20%) 70 (35%) 63 (31.5%) 200 (100%)
Xl= 2.72; df = 3; p = 0.43
No significant differences were found between the responses of females
and males in this situation. (Table 3.2)
Table 3.3. Ql by age
NI LI SI GI Total
Students 10 (10%) 23 (23 %) 32 (32 %) 35 (35%) 100(50%)
< 32 yrs 7 (26.9%) 4 (15.4%) 7 (26.9%) 8 (30.8 %) 26 (13%)
32-45 yrs 6 (11.5 %) 12 (23.1 %) 20 (38.5%) 14 (26.9%) 52 (26%)
> 45 yrs 4 (18.2 %) 1 (4.6%) 11 (50%) 6 (27.3 %) 22 (11 %)
Total 27 (13.5%) 40 (20%) 70 (35%) 63 (31.5 %) 200 (100%)
Xl= 11.68; df =9; p =0.23
Imposition in favor asking 247
No significant differences were found between the students and the three
non-student age groups in this situation (Table 3.3).
Both the non-student and student populations considered this situa-
tion to be imposing [SI + GI > NI + LI]. Similarly, females, males, and all
age groups considered this to be an imposing situation.
2. Favor situation: One person to a good friend who is not working out-
side the home: "Could you please watch my children (3) tomorrow for a
few hours? I have to go to work, and they'd rather be with your kids than
a babysitter."
Table 4.1. Q2 by student
N1 L1 SI G1 Total
Student 13 (13 %) 37 (370/0) 43 (43 %) 7(7%) 100(50%)
Non-Student 14 (140/0) 18 (18 %) 42 (42%) 26 (260/0) 100(50%)
Total 27 (13.5%) 55 (27.5%) 85 (42.50/0) 33 (16.5%) 200 (100%)
X
2
= 17.55; df = 3; p = 0.001
There is a considerable statistically significant difference between the
two groups. The non-students found this to be a more imposing situation
than did the students (GI = 26 % vs. 70/0). (Table 4.1)
Table 4.2. Q2 by gender
N1 L1 SI G1 Total
Female 21 (16.7%) 33 (26.2%) 47 (37.3%) 25 (19.8%) 126 (630/0)
Male 6 (8.1%) 22 (29.7%) 38 (51.40/0) 8 (10.8%) 74 (370/0)
Total 27 (13.5%) 55 (27.50/0) 85 (42.5%) 33 (16.5%) 200 (100%)
X2= 7.21; df = 3; p = 0.065
No significant differences were found between the females and males
in this situation. Thus, gender was not a significant variable in this situa-
tion. (Table 4.2)
248 Myra Goldschmidt
Table 4.3. Q2 by age
NI LI SI GI Total
Students 13 (13 %) 37 (370/0) 43 (43 %) 7(7%) 100(50%)
< 32 yrs 4 (15.30/0) 6 (23.1 0/0) 10 (38.5 0/0) 6 (23.1 0/0) 26 (130/0)
32-45 yrs 7 (13.5%) 7 (13.5%) 25 (28 %) 13 (25%) 52 (26%)
> 45 yrs 3 (13.7%) 5 (22.7%) 7 (31.8 %) 7 (31.8%) 22 (11 %)
Total 27 (13.5%) 55 (27.5%) 85 (42.5%) 33 (16.5 %) 200 (100%)
X
2
= 20.09; df = 9; p = 0.017
There is a statistically significant difference among the various age
groups (Table 4.3). The non-students considered this to be a more impo-
sing situation compared to the students. In order to determine if anyone
of the non-student age groups considered this a more imposing situation,
these three groups [< 32 years, 32- 45 years, and> 45 years] were analyz-
ed separately for chi square and p values (Xl= 2.76; df =6; p =0.84).
After further analysis, the age of the non-student respondent is not of sig-
nificance in this situation.
Both the student and non-student groups considered this an imposing
situation (SI + GI >NI + LI). Similarly, both females and males and all the
age groups considered this an imposing situation.
3. Favor situation: One neighbor asks another neighbor who is a good
friend: "If I can't find anyone else, do you think I can count on you to
take care of our animals while we are away? It would just be Saturday
afternoon and evening, and Sunday morning for you to feed them and let
them out for a few minutes."
Table 5.1. Q3 by student
NI LI SI GI Total
Student 21 (21 %) 43 (43 %) 25 (25 %) 11 (11 %) 100(50%)
Non-Student 33 (33 %) 26 (26%) 28 (28 %) 13 (13 %) 100(50%)
Total 54 (27%) 69 (34.5%) 53 (26.5%) 24 (12 %) 200 (100%)
X2= 7.19; df = 3; p = 0.07
Imposition in favor asking 249
No statistically significant differences were found between the
responses of the students and non-students in this situation. (Table 5.1)
Table 5.2. Q3 by gender
NI LI SI GI Total
Female 45 (35.7 0/0) 38 (30.2 0/0) 31 (24.6%) 12 (9.5 0/0) 126 (63%)
Male 9 (12.2 %) 31 (41.9%) 22 (29.70/0) 12 (16.2 %) 74 (370/0)
Total 54 (27%) 69 (34.5 0/0) 53 (26.5 0/0) 24 (120/0) 200 (100%)
X
2
= 13.64; dt = 3; p = 0.003
Males found this situation to incur somewhat more imposition than
did the females. This might suggest that males are asked to do this type
of task less often than are females, thus find it to be more imposing. Also,
males may perceive this type of task to require an inordinate amount of
time and effort, which again, may contribute to their belief that this
particular task constitutes more imposition. (Table 5.2)
Table 5.3. Q3 by age
NI LI SI GI Total
Students 21 (10.5 0/0) 43 (21.5 %) 25 (12.5 %) 11 (6.50/0) 100 (50%)
< 32 yrs 13(50%) 6 (23.1 0/0) 6 (23.1 %) 1 (3.8%) 26 (13 0/0)
32-45 yrs 14 (26.9 0/0) 15 (28.9%) 14 (26.9 0/0) 9 (17.30/0) 52 (26%)
>45 yrs 6 (27.3%) 5 (22.70/0) 8 (36.4%) 3 (13.6%) 22 (11 0/0)
Total 54 (27%) 69 (34.5 0/0) 53 (26.5%) 24 (120/0) 200 (100%)
X2= 9.75; dt = 9; p = 0.37
No significant differences were found between the students and the
three non-student age groups in this situation (Table 5.3). Overall, the
student and non-student groups considered this an unimposing situation
(NI+LI>SI+GI).
Question 3 also may be considered to be a "fuzzy" situation in that
the data showed no strong reaction in anyone of the four types of
250 Myra Goldschmidt
imposition, though the trend is certainly towards little and no imposition
(61.5 %) as opposed to some and great imposition (38.5 %).
What makes question 3 so interesting, however, is not necessarily the
numbers, but rather the situation itself (caring for animals) compared
with question 2 (caring for children). Twice as many people thought that
looking after animals incurred no imposition (27 %) when compared
with looking after children (13.5 %). This is particularly noticeable in the
non-student group, 33 % of whom said looking after animals incurred no
imposition while only 14 % said looking after children incurred no
imposition. Similarly, 21 % of the student group believed that looking
after animals incurred no imposition, but only 13 % said that looking
after children incurred no imposition. These percentages suggest
that people appear to be more willing to look after animals than
children. Perhaps most people have had personal experience with pets
rather than children to account for these findings. Or perhaps, more
people believe that there is less time and effort involved with looking after
pets.
Furthermore, three times as many females than males found this situa-
tion to incur no imposition. Again, perhaps this is a situation that more
women do for others with regard to favor-asking. In terms of age, the
primary group to consider this situation as causing no imposition was the
non-student group less than 32 years of age, especially when compared
with their looking after children (50 % and 15 % respectively).
4. Favor situation: An acquaintance asks the following of someone she
hasn't seen for several months, and who just got out of the hospital
because of leg surgery. It is 3pm on Saturday, Christmas Eve: "I called to
ask you a favor. We're being kicked out of our house tonight because
David (the eldest of 3 sons) wants to cook dinner for his girlfriend and
Table 6.1. Q4 by student
NI LI 51 GI Total
Student 1 (1 %) 5 (5%) 14 (14%) 80(80%) 100(50%)
Non-Student 1 (1 0/0) 2 (20/0) 7 (70/0) 90 (900/0) 100 (500/0)
Total 2 (1 %) 7 (3.5%) 21 (10.5 %) 170 (85%) 200 (100%)
Xl= 4.21; df =3; p =0.24
Imposition in favor asking 251
we need some place to go. Can we (4 people) come to your house for the
evening and for dinner? We can pick up something like a pizza."
There are several groups in this analysis [Student/No imposition and
Non-student/No imposition/Little imposition] that have small numbers
5). This may invalidate the chi square test. The two groups NI and LI
for both students and non-students were combined to increase the group
size, and the chi square and p values were recalculated (X
2
= 3.92; df = 2;
p = 0.41).
No significant differences were found between the responses of the
students and non-students in this situation. (Table 6.1)
Table 6.2. Q4 by gender
NI LI 51 GI Total
Female 0(0%) 4 (3.2%) 9 (7.1%) 113 (89.7%) 126 (63%)
Male 2 (2.7%) 3 (4.1%) 12 (16.2%) 57 (77%) 74 (37%)
Total 2 (1%) 7 (3.5%) 21 (10.50/0) 170(85%) 200 (100%)
X
2
= 8.04; df = 3; p = 0.05
As stated above, the chi square and p values were recalculated because
of small numbers (X
2
= 5.86; df = 2; p = 0.05).
Gender was found to be a statistically significant factor in the way the
respondents viewed this situation. Females found this situation to incur
great imposition more than did the males. This might suggest that
females, especially, consider "private family time" to be extremely
important and something that is not to be imposed upon. (Table 6.2)
Table 6.3. Q4 by age
NI LI SI GI Total
Students 1 (0.5%) 5 (2.5%) 14 (7%) 80 (40%) 100 (500/0)
< 32 yrs 1 (3.9 0/0) 0 3 (11.5 %) 22 (84.6%) 26 (13%)
32-45 yrs 0 2 (3.9%) 4 (7.7%) 46 (88.5%) 52 (260/0)
>45 yrs 0 0 0 22 (100%) 22 (11 %)
Total 2 (1%) 7 (3.5%) 21 (10.50/0) 170 (85%) 200 (100%)
X2= 10.03; df = 9; P = 0.35
252 Myra Goldschmidt
As stated above the chi square and p values were recalculated because
of small numbers (X
2
=8.41; df =6; p =0.38).
Although there is no statistically significant difference between the
students and the three non-student age groups, there is a trend evident
(Table 6.3). With increasing age, the degree of imposition increases; all
respondents over 45 years of age considered this situation to incur great
imposition. All groups considered this a very imposing situation with
more than 85 % of the respondents deeming this a great imposition.
5. Favor situation: One graduate student asks another graduate student
known only by sight: "I was wondering. Well, 1have these questionnaires
for people doing a study of students here, and 1 was wondering if you'd
mind filling one out?"
Table 7.1. Q5 by student
NI LI SI GI Total
Student 41 (41 0/0) 41 (41 0/0) 12 (12 %) 6 (60/0) 100 (500/0)
Non-Student 48 (48 %) 32 (32 %) 16(16%) 4 (4%) 100(50%)
Total 89 (44.5%) 73 (36.5%) 28 (14%) 10 (5%) 200 (100%)
X
2
= 2.63; df = 3; p = 0.45
No significant differences were found between the responses of the
students and the non-students in this situation. (Table 7.1)
Table 7.2. Q5 by gender
NI LI SI GI Total
Female 64 (50.8 %) 38 (30.2 %) 18 (14.2 %) 6 (4.8%) 126 (63%)
Male 25 (33.8 %) 35 (47.3 %) 10 (13.5 %) 4 (5.4%) 74 (37%)
Total 89 (44.5%) 73 (36.5%) 28 (14%) 10 (5%) 200 (100%)
X2= 6.84; df = 3; p = 0.08
Imposition in favor asking 253
No significant differences were found between the responses of
females and males in this situation. (Table 7.2)
Table 7.3. Q5 by age
NI LI 51 GI Total
Students 41 (41 %) 41 (41 %) 12 (12 %) 6 (60/0) 100(50%)
< 32 yrs 13 (500/0) 5 (19.2 %) 5 (19.20/0) 3 (11.6%) 26 (13%)
32-45 yrs 24 (46.2 %) 18 (34.6 0/0) 9(17.3%) 1 (1.9%) 52 (26%)
>45 yrs 11 (50%) 9 (40.9%) 2 (9.1 %) 0 22 (11 %)
Total 89 (44.5%) 73 (36.5 0/0) 28 (14%) 10 (5%) 200 (100%)
Xl= 9.57; df = 9; p = 0.39
No significant differences were found between the students and the
three non-student age groups in this situation (Table 7.3).
Both the student and non-student groups considered this an un-
imposing situation (NI + LI > SI + GI). Similarly, both females and males
considered this an unimposing situation.
Question 5 showed a strong trend towards 'little' and 'no' imposition
(81 0/0). This is a situation which appears to entail little time and/or effort.
It is also one that students do for other students.
4. Discussion and implications of the results
From the results of the survey, it can be said that the three objectives
sought were, in fact, achieved. First of all, with regard to an underlying
value system in the targeted speech community, people, in general, feel
imposed upon in the following favor-asking situations:
1) In situations where family privacy (or special times relating to the
family) is intruded upon.
2) In situations involving a great deal of time and/or effort.
In the first type of situation, most of the subjects vehemently felt impo-
sed upon and objected to the situation concerning Christmas Eve since
this is a special "family" time for most people. Not only were the majo-
254 Myra Goldschmidt
rity of great impositions given for this situation, but also a large number
of comments were elicited as well. These ranged from, "This is outra-
geous!" to "Who needs friends like this?" Clearly, the most indignation
and annoyance surfaced in this situation.
In the second type of situation, people generally felt imposed upon
whenever they were asked to do something that required a great deal of
time and/or effort. Though favors, in general, require varying degrees of
time and/or effort, it seems that the greater the amount of perceived time
or effort involved in a favor, the greater is the amount of perceived
imposition. Certainly, time is one of the most precious commodities
people have today, and they don't like to spend too much of it doing
favors for others. Likewise, any situation which infringes on a person's
time, also entails some effort on that person's part. Therefore, the
questions about shoveling a 90 foot driveway and taking care of children
garnered a number of great and some imposition responses.
With regard to the results of the survey for the second objective, it
became very clear that the investigator's preconceived hypotheses
concerning the social variables of age and gender and role as conditioning
factors in imposition were not fully substantiated. Not only were the
overall results between the males and females similar with respect to
imposition intensity, but also the results between the students and the
non-students, as well as the age of the individuals proved to be, for the
most part, comparable. What this points to is that there exist unwritten,
but understood "parameters" within speech communities concerning
what is and what is not thought of as imposing. These parameters seem
to transcend all other variables and are important to learn as a new
member of a particular speech community.
Finally, the results confirm implications for TESOL. Clearly, speakers
must be cognizant of the rights and obligations they have towards others
which include the notion of imposition. In other words, learners should
recognize that since favor-asking requires some time and/or effort on the
part of an addressee, the speaker, by asking the favor, is impinging on the
rights of the addressee. Also, learners of a language need to know what
constitutes imposition, how people in a particular culture define the level
or degree of imposition, and when and how it is acceptable to impose
upon someone. These are very important to communicate to learners
because if favor-asking is defined as a speech act where people are not
obligated by their role or status to do the act, then degree of imposition
or what can be asked or expected of an addressee is critical for learners
to know about.
Imposition in favor asking 255
This type of information is important for learners to be truly
competent members of the target language community. Interestingly, non-
native learners often know parts of social routines, but fail in the overall
delivery of them which may result in an undesirable impact on an addres-
see. Example (1) illustrates an utterance by a non-native learner to a
teacher he had had three years earlier.
(1) NNS
NS
==> NNS
I have a favor to ask you.
Sure, what can I do for you?
You need to write a recommendation for me.
This student's asking of the actual favor was inappropriate and could
have caused a breakdown in communication with a teacher (or other
addressee) who was not aware of the student's lack of understanding of
this type of speech behavior. A typical reaction and/or response might
have been, "I don't need to do anything for you," which would have
caused ill-feelings on the part of both the student and the teacher.
By focusing on both the speaker and the addressee in certain convers-
ational situations, teachers do a great service to new learners of a
language. Everyone needs help at times, and the best way of ensuring that
this help is attended to is by asking favors in a way which presents the
speaker in a good light. Since deviation from social norms can be inter-
preted as impertinent, sarcastic, or rude, learners need to be taught the
social norms for a particular culture in order to comprehend both how to
ask a favor on the one hand, and what the listener expects to hear, on the
other. In other words, learners need to be sensitive, not only to the
speaker's situation, but to the addressee's situation as well.
Notes
1. Perceived imposition varies from person to person. Whereas one person may
perceive a favor as causing great imposition, another person may perceive the
same favor as causing little or no imposition. Each favor-asking situation is
conditioned by several variables which could affect the overall amount of
perceived imposition.
2. The original survey had twelve situations; however, for the purpose of this
chapter, a representative sampling was used.
256 Myra Goldschmidt
References
Austin, John L.
1962 How to do things with words. Oxford: Clarendon Press.
Blum-Kulka, Shoshana - Edward Levenston
1987 "Lexical-grammatical pragmatic indicators", Studies in Second
Language Acquisition 9: 155-170.
Brown, Penelope - Stephen Levinson
1978 Politeness: Some universals in language usage. Cambridge:
Cambridge University.
Ervin-Tripp, Susan
1976 "Is Sybil there? The structure of American English directives",
Language in Society 5.1: 25-66.
Ervin-Tripp, Susan - David Gordon
1986 "The development of requests", in: R.L. Schiefelbusch (ed.), 61-95.
Fishman, Joshua A. (ed.)
1968 Readings in the sociology of language. The Hague: Mouton.
Gladwin, Thomas - William C. Sturtevant (eds.)
1962 Anthropology and human behavior. Washington, DC:
Anthropological Society of Washington.
Goldschmidt, Myra
1993 For the favor of asking: A sociolinguistic analysis. [Unpublished
Ph. D. dissertation, University of Pennsylvania.]
Gumperz, John
1971 Language in social groups. Palo Alto, CA: Stanford University Press.
Gumperz, John - Dell Hymes (eds.)
1972 Directions in sociolinguistics. New York: Holt, Rinehart and
Winston.
Hymes, Dell
1962 "The ethnography of speaking", in: Thomas Gladwin and William
C. Sturtevant (eds.), 15-53.
[1968] [Reprinted in: Joshua A. Fishman (ed.), Readings in the sociology of
language. The Hague: Mouton, 1968, 99-138.]
1972 "Models of the interactions of language and social life", in: John
Gumperz - Dell Hymes (eds.), 35-71.
Labov, William
1972 Sociolinguistic patterns. Philadelphia: University of Pennsylvania
Press.
Schiefelbusch, Richard L. (ed.)
1986 Language competence: Assessment and intervention. San Diego, CA:
College-Hill.
Searle, John R.
1969 Speech acts. Cambridge: Cambridge University Press.
Transfer of pragmatic competence and suggestions
in Spanish foreign language learning
Dale April Koike
1. Introduction
The transfer of knowledge of a first language in the learning of a foreign
language is a common process. This knowledge ~ a n reflect aspects of any
component of language, including syntax, phonology, morphology,
semantics, and pragmatics. The basic question that is addressed in this
study is the following: Do foreign language learners use strategies of
transfer to help them comprehend information? The context of learning
examined here is that of the aural comprehension of the speech act of sug-
gestions.
In a previous study, I discuss the notion of a "pragmatic competence"
in a learner's interlanguage, involving knowledge and use of rules of
appropriateness and politeness that dictate the way the learner under-
stands and formulates utterances, such as requests (see Koike
1989: 279-281). Data obtained from several experiments suggest that
first language pragmatic knowledge transferred to the foreign language
speech act situation causes learners to attempt to produce utterances that
they believe would be pragmatically appropriate to the context. When
faced with producing a speech act that is more difficult than students feel
competent in formulating, however, many choose to employ a less appro-
priate but syntactically simpler form, showing a change of strategy in
actual production.
This pragmatic competence is part of interlanguage, a system that
represents dynamic stages in the learning process and that are subject to
continual change and modification (see Selinker 1972: 229-230; Corder
1975: 410-411). One may ask whether changes in the way knowledge
from the first language is transferred to the foreign language can be
observed at different stages of learning. The present study examines this
question through data from native English speakers who are learners of
Spanish to see (1) whether adult foreign language learners of various
levels of proficiency can recognize the speech act of suggestions in the
258 Dale April Koike
foreign language; and (2) to what extent first language pragmatic
information is transferred in the comprehension of a foreign language
speech act, and what variables can affect this transfer.
The data used in this investigation come from responses to a question-
naire by adult U. S. native speakers of English who are Spanish learners
at varying levels of study at the University of Texas at Austin. On the
basis of the results of this questionnaire, I suggest that students' levels of
proficiency affect the way in which they transfer first language speech act
knowledge to understand a complex foreign language speech act. When
learners are advanced enough to begin to analyze the components of
complex speech act utterances, they can sometimes misunderstand the
speaker's intent, especially if the speech act utterance in question has
different connotations in the two languages. In other words, learners first
need to be proficient enough to understand the overall pragmatic intent
of complex utterances. When such is the case and they can also pay
attention to individual elements in the utterance, they may transfer
knowledge from the first language concerning these elements at that
point. In the following sections, I will discuss transfer in foreign
language learning, the speech act of suggestions in both Spanish and
English, the questions to be addressed, the experiment, and the results
and conclusions.
2. Transfer in foreign language learning
Many studies have addressed the use of transfer as a strategy in foreign
language learning, examining evidence of the phenomenon largely in oral
production (see especially Gass - Selinker 1983; 1992). All agree that
transfer from the first language does occur in foreign language learning,
under various constraints. Corder (1983: 95 and 1992: 29) states that the
first language acts as a tool in the discovery of the formal properties of
the foreign language, facilitating especially in the learning of those
features that resemble those of the first language. Corder, qS well as
Kellerman (1977; 1983), stress that the greater the similarity perceived by
the learner' between the first language and the foreign language, the
greater the likelihood that transfer will occur in foreign language ac-
quisition. Kellerman (1983: 117) claims a first language structure will be
treated as language-specific and not transferable to a foreign language, or
language-neutral and thus transferable. Blum-Kulka, House, and Kasper
Transfer of pragmatic competence 259
(1989: 26-27) found that learners avoid transfer of language-specific
structures, revealing an awareness of transferability of pragmatic con-
straints.
Schachter (1983: 104; 1992:38-39) outlines the possible outcomes of
learner hypotheses about the target language based on first language
information. One possibility she describes is that the learner may choose
the correct domain or abstract category within the target language (my
example, reflexive pronouns) but the wrong hypothesis about the input,
either because of a somewhat mistaken analysis of the input or because
the learner correctly equates the domains of the two languages but
assumes a hypothesis that is appropriate for the first language and not for
the foreign language. In this case, a transfer error is produced. The other
possibilities are, of course, that the transfer will cause no problem, or will
be understood in spite of the application of the hypothesis to the wrong
domain.
Since they were looking for evidence of transfer, all of the studies cited
above were focused on foreign language or second language production,
and not on a receptive skill such as listening comprehension. One cannot
doubt that transfer, a cognitive process, also occurs in listening com-
prehension and reading activities. A study by Gass (1989: 196) examined
the extent of transfer in the reading interpretation of simple sentences.
One of her conclusions was that second language learners seek an initial
hypothesis regarding the second language on the basis of their first
language. When there is an incongruity between the two languages, how-
ever, learners may resort to fundamental universal properties of language,
such as that of canonical word order. The question arises whether these
claims and findings about foreign language transfer also apply ~ n the
same way to the listening comprehension skill. If they do, then one must
ask what the consequences are for the foreign language learner who, as
Schachter describes, makes a wrong hypothesis about the input in a
correct domain in attempting to understand the foreign language.
I am assuming, as does Gass (1989: 183), that syntax, semantics, and
pragmatics simultaneously interact in the way a learner interprets input,
along with other features of the grammar. As Gass points out, there are
many linguistic elements available to cue a learner to interpret input, and
because of factors such as frequency, ease of interpretation, and informa-
tion value, not all of the elements are equally used during on-line proces-
sing. She adds that if there are cross-linguistic differences in cue usage,
then a learner must know not only the appropriate cues of the target
language (e. g., word order) but also the strengths of those cues. In
260 Dale April Koike
learning a foreign language, a learner may learn the new cues or rely on
knowledge of the first language.
3. Spanish and English suggestions
Suggestions are speech acts that are made presumably in the best interest
of the listener, usually to help the listener toward some goal that the latter
desires or is assumed to desire. Suggestions are sometimes made in the
best interest of the speaker as well. As a type of directive, or a way to get
someone to do something (Searle 1979: 13), suggestions require a future
effort by the listener (Blum-Kulka - House - Kasper 1989: 12), and call
for strategies to mitigate their force. They represent a complex speech act
whose intent is sometimes misunderstood even by native speakers.
Many forms of suggestions in Spanish and English are alike and easily
transfer between the two languages, such as in (1) through (3).
(1) Sugiero que leas este libro.
'I suggest that you read this book'
(2) 2No puedes leer este?
'Can't you read this one?'
(3) 2Por que no lees este?
'Why don't you read this one?'
There are differences, however, in the formulation of some suggestions in
Spanish as opposed to English, particularly in the use of negatives in
interrogative suggestions, for example as in (4) through (9).
(4) Have you thought about reading this book?
(5) 2No has pensado en leer este libro?
'Haven't you thought about reading this book?'l
(6) #2Has pensado en leer este libro?
'Have you thought about reading this book?'
(7) Should you read this book?
Transfer of pragmatic competence 261
(8) 2No deberias/tienes que leer este libro?
'Shouldn't you/don't you have to read this book?'2
(9) #2 Deberias/tienes que leer este libro?
'Should you/do you have to read this book?'
To convey the illocutionary force of a suggestion in Spanish, the Spanish
equivalents of (4) and (7) must be expressed negatively, as in (5) and (8).
The utterances in (6) and (9) convey purely yes-no confirmation
questions. To an English speaker, however, the English translations of (5)
and (8) convey a much stronger force than their Spanish counterparts; the
force can come across as almost a reproach to the listener.
Aside from the question of transfer of speech act knowledge, there are
differences in the grammatical formulation of suggestions in English as
opposed to Spanish that can confound the comprehension process. Such
is the use of the negative. I base my ideas on data drawn from two studies
of speech acts in Spanish, specifically, of directives (Hobbs 1990; Koike
1994). In the two studies, a total of 83 native Spanish speakers from
Mexico were asked what they would say in a potentially face-threatening
situation in which they had to ask a child, a peer, or a non-intimate,
Declarative
Deberias 'You should/ought to'
Puedes/Podrias 'You can/could'
No estaria mal si 'It wouldn't hurt if you'
Mejor/Seria mejor (mas rapido, buena idea, etc.) si 'It would be better
(faster, a good idea, etc.) if'
Sugiero/creo (que) 'I suggest/think (that)'
Si fuera Ud. 'If I were you' (I would)
Vamos a 'Let's'
Interrogative
i,Por que no 'Why don't youlWhy not'
i,No deberias 'Shouldn't/Should you'
i, Que tal/Que te parece si 'How about/What do you think (if)'
i,No has pensado en 'Ever think about/Have you thought about/
considered'
i,No puedes 'Can/Can't you'
Imperative
Trata de 'Try to'
Figure 1. Commonly used formulaic expressions in Spanish suggestions
262 Dale April Koike
higher-ranking stranger to move from a chair reserved for someone else.
Figure 1 displays the suggestion forms used by the informants in the two
studies, as divided into three types according to syntactic and lexical
considerations. The types are declarative, interrogative, and imperative.
Of the declarative type of suggestions, seven forms were used, three of
which include a si- 'if' clause, conveying an irrealis condition. Five inter-
rogative forms were employed, four of which are negated, and one in-
cludes a si- clause.
One must remember that a suggestion is a kind of directive in which
the speaker normally wants to minimize to the greatest extent the possi-
bility that the listener will be offended at the suggestion. The data show
that Spanish usually requires the negative when formulating interrogative
suggestions. English, on the other hand, allows both negated and non-
negated forms in conveying a suggestion intent, as in examples (10) and
(11) spoken in response to someone who said she was running out of time
to turn in an article.
(10) Have you thought about asking for an extension?
(11) Haven:lt you thought about asking for an extension?
Example (10) is ambiguous in that it can have three implications: (a)
The speaker is asking a simple information question to find out
if the listener has thought about X; (b) The speaker is asking to find
out the listener's reaction to X; or (c) As a suggestion, the speaker is
letting the listener know s/he (the speaker) believes X should be done
by asking about it. Example (11), however, implies that the speaker
believes the listener should think about X, assumes that the listener
has not thought about it, and questions that assumption. Thus, the
force of the utterance for English speakers is rather strong, since the
speaker is not allowing the listener "freedom of action unhindered," as
Brown and Levinson (1987: 129-130) describe negative politeness
(for further discussion, see Koike 1992: 21-31).3 Use of the negative
in English suggestions is not common, especially in interrogative
suggestions. The negative sentence can sound like an insult or a reproach,
as if the listener overlooked the option mentioned by the speaker. In
sum, the differences between Spanish and English interrogative
suggestions lead to the possibility of problems of miscommunication
and misunderstanding between the English and Spanish languages and
cultures.
Transfer of pragmatic competence 263
4. The experiment
4.1. Questions and hypotheses
The purpose of this study is to examine the comprehension and reactions
of English-speaking students of Spanish at different levels of foreign
language proficiency upon hearing these negated suggestions. The speci-
fic questions to be addressed are:
(a) Do English-speaking students of different levels of Spanish
language proficiency understand the intent of the suggestion when the
Spanish form is similar but different from the English one, and expresses
a different intent than that of the English form?
The question implies that learners may transfer their speech act know-
ledge from the first language to the foreign language. In the case of
negated interrogative suggestions, one may ask whether learners perceive
the difference in the forms regarding the negative element. If they do
understand the suggestion intent and do not notice the negative element,
they will attach the suggestion illocutionary force to the utterance, rely-
ing on the context and their knowledge of suggestion forms from the first
language. On the other hand, if they do perceive the difference in form,
it is hypothesized that learners may simply understand the utterance as a
suggestion formulated according to Spanish constraints (as opposed to
English ones), or misunderstand the intent.
(b) Do perceptions of the differences in Spanish and English forms
cause misunderstanding and negative reactions to the speaker?
If learners do perceive the differences in form, they may misunderstand
the intent and react negatively to the speech act, based on first language
speech act knowledge. This interpretation could account for some mis-
communication and negative attitudes toward native Spanish speakers.
(c) How do the learners respond verbally to the speech act?
If the students misunderstand the intent and react negatively, they may
respond verbally in a negative manner, thereby causing further mis-
communication.
(d) Is there a marked difference in understanding of the speech act
according to level of foreign language proficiency?
It is hypothesized that there is an increase in ability to identify the
speech act and to reproduce it correlated with higher levels of Spanish
foreign language proficiency. It is predicted that lower-level students will
understand the global intent alone, if they understand at all, and will not
264 Dale April Koike
hear the negation. It is also predicted that higher-level students will
understand the intent and that some will hear the negation and react in a
negative manner to the utterance.
In order to seek answers to the research questions posed above, an
experiment was carried out with English-speaking or bilingual students at
different levels of Spanish language study during a summer session in a
university program.
4.2. Subjects, instrumentation, and procedure
The 114 students who participated in this study were of three different
levels of language study, as follows: 46 first-year students in intensive
Spanish classes at the beginning of the equivalent of the second semester;
34 second-year students, at the end of their third semester of Spanish; and
34 advanced students, most in their third or fourth year of university
Spanish courses. The advanced students included both non-native (68 0/0)
and bilingual Spanish-English Chicano students (32 %). The number of
bilingual speakers in the other groups was negligible. Two classes taught
by different non-native instructors for each level were involved. The
informants were given a context for each situation, and were then asked
to watch a very short portion of a videotaped speech act by a native
speaker. There were seven speech acts in total (see Appendix One). The
seven native speakers were given a script from which they could vary
except for the "key" sentence that explicitly expressed the speech act in
question.
4
They were asked to speak directly to the camera, and try to act
as naturally as possible. A videotape was used instead of an audio-
cassette tape because it was believed that the communication of speech
acts is much more effective with "holistic" information, including not
only the actual utterance but also body movements and facial expres-
sions. Included in the tape were one rebuke, one request, one information
question, and four suggestions. One of the suggestions contained an
irrealis clause in declarative form (e. g., "If I were you, I would ... "), and
two others represented negated interrogative suggestions. Another was a
non-negated interrogative suggestion, which does not convey a sugges-
tion intent.
Students watched each of the seven situations and, immediately
following each one, answered three basic questions on the questionnaire
shown in Appendix Two. First, they were asked to respond, in English or
Spanish, to the speaker they heard and saw, as if they were answering
Transfer of pragmatic competence 265
back to the speaker. Second, informants were to identify the type of
speech act expressed (e. g., suggestion, apology) and, if possible, to re-
produce how it was expressed. Third, students were asked to evaluate the
speakers, using a Likert-scale, in terms of degrees of aggressive/passive,
rude/polite, non-communicative/communicative, strong/weak, and un-
friendly/friendly.
4.3. Results
4.3.1. Listening comprehension and responses
The results from the experiment were tabulated and run through the
Statistical Analysis System (SAS) for analysis. Table 1 reveals the results
in percentages of frequency of occurrence for all groups in all situations
regarding the targeted identification of the speech act.
5
In general, only about one third of the first-year and one fourth of the
second-year students understood the intent of the speech acts (except
Situation Five), revealing that it is a skill that relatively few students at
these levels can perform. On the average, a little over half of the advanced
students understood the intent. The results were also run through Tukey's
Studentized Range (HSD) Test to compare the performances of the three
groups. The comparisons reveal that the first- and second-year students
were not significantly different but the advanced group was significantly
Table 1. % of correct answers for each group by situation
Group
First year
Second year
Advanced
Average score
of all groups
Situation
1 2 3 4 5 6 7
33 28 37 22 17 43 26
47 21 29 21 6 32 29
56 53 50 62 3 59 68
44 33 39 33 10 44 40
(df =2; Sum of Squares =56.37; Mean Square =28.18; F =14.65; P <0.001)
n sizes:
First year = 46
Second year =34
Advanced =34
266 Dale April Koike
Table 2. % of answers by each group for each option in identification of speech
act (correct answers underlined)
Option Situation
1 2 3 4 5 6 7
First year classes (n =46)
Invitation 2 0 0 17 13 20 0
Apology 2 7 0 7 0 0 7
Request 2 4 9 15 0 43 4
Suggestion 33 35 37 22 22 0 26
Order 0 7 2 2 0 2 0
Information Question 41 2 20 11 17 22 9
Mild rebuke 9 28 17 2 4 4 17
Other 2 2 7 2 30 4 20
NR 9 15 9 22 14 5 17
Second year classes (n =34)
Invitation 9 0 0 9 6 29 0
Apology 0 0 3 3 3 0 3
Request 0 9 0 9 12 32 9
Suggestion 47 32 29 21 12 6 29
Order 0 6 0 0 0 0 0
Information Question 32 12 6 15
2
24 6
Mild rebuke 6 21 18 0 6 3 15
Other 0 3 24 21 15 3 18
NR 6 18 21 22 41 3 21
Advanced classes (n =34)
Invitation 9 0 0 9 15 0 0
Apology 0 0 3 6 0 3 0
Request 0 6 0 9 0 59 0
Suggestion 56 15 50 62 53 0 68
Order 0 0 0 0 3 0 9
Information Question 21 15 12 9 3 32 3
Mild rebuke 9 53 6 3
.3.
0 12
Other 0 3 6 3 15 6 6
NR 6 9 23 0 9 0 2
Transfer of pragmatic competence 267
different from the other two (mean scores first year = 2.07; second
year =1.85; advanced =3.50).
The results, however, should be examined with consideration of
factors in the situations themselves. Table 2 shows the answers given in
each situation by each of the three groups of students. I will focus on the
results of Situations One, Three, Four, and Five, which involve negated
and non-negated interrogative suggestions. In Situation One, many
students thought that a simple information question was being asked in
2No has pensado en estudiar con tus coLegas de La cLase ahora? 'Have(n't)
you thought about studying with your colleagues in the class now?' While
the intent of the speaker is clearly that of a suggestion, learners sometimes
misunderstand speech acts, especially when they are not prefaced by sug-
gestion formulaic expressions, such as "Why don't you" and "How
about." Only a few students in each of the three groups thought the
utterance was a mild rebuke. Thus, the data indicate that most of the
students did not hear the negative element in this utterance, or if they did,
they did not associate it with the English interpretation.
The frequencies of appropriate responses to the speech act itself, dis-
played in Table 3, show that several first- and second-year students and
about half of the advanced group produced a logical and appropriate
answer to the utterance, in spite of the fact that relatively few could
reproduce the utterance accurately, as shown in Table 4.
This pattern, in which many informants could respond to the speech
act appropriately but could reproduce the utterance only on occasion,
held in all the situations. In general, this indicates that even though many
students understood only partially or not at all, they could many times
say something that would encourage further interaction and more input
from the speaker, which presumably would lead them to understand the
intent.
Table 3. % of logical and appropriate responses to the speech act by all three
groups in all situations
Group Situation
1 2 3 4 5 6 7
First year 13 11 22 33 46 63 22
Second year 21 18 18 12 3 32 18
Advanced 56 76 32 71 65 79 82
268 Dale April Koike
Table 4. % of students who reproduced the targeted utterance verbatim
Group Situation
1 2 3 4 5 6 7
First year 2 4 0 4 2 0 9
Second year 0 3 0 3 0 0 0
Advanced 29 65 9 41 21 29 29
The data from Situation Three are interesting for their variety (see
Column 3, Table 2). Some students of the first year (20 %) and advanced
(12 %) levels thought that the speaker was asking a simple information
question with 2No deberias hablar con la profesora y pedir que te de
credito extra por ese trabajo? 'Shouldn't you talk with the teacher and ask
that she give you extra credit for that work?' Some first-year (17 %) and
second-year (18 %) students thought they heard a mild rebuke, which was
the predicted outcome from a transfer of the English interpretation of the
target utterance. However, 24 % of the second-year students believed they
heard another kind of speech act, usually labelled "Complaint" or
"Statement of a Problem." Twenty-one percent of the advanced students
could not identify the speech act at all. The responses to the speaker and
the reproductions of what the speaker said in Tables 3 and 4, respectively,
show that most students of all levels could not reproduce the utterance
accurately, and only about 20 % of the first- and second-year group and
32 % of the advanced students responded appropriately.
Situation Four, with the utterance 2Por que no copias las hojas de
Raquel? 'Why don't you copy Rachel's pages?' also revealed some mis-
understanding, especially by the first- and second-year students. Several
first-year students thought they heard an invitation to have coffee (17 0/0),
while others believed they heard a request to borrow some notes (150/0)
(see Table 3). Some second-year informants thought the speaker asked an
information question (15 0/0) or thought they heard another speech act,
most commonly labelled "Giving Information" (21 %). These misinter-
pretations indicate that the students probably inferred the speech act
from the context when they were unsure of what they had heard. Hearing
someone talk about coffee and the class homework, and having twenty
minutes before class could lead some to think they were invited to have a
cup, or asked to lend their notes. Almost half the advanced students
(41 %) were able to reproduce the utterance exactly as in the situation,
Transfer of pragmatic competence 269
which was prefaced by the 2Por que no? expression, as seen in Table 4.
This time, however, the expression did lead more of them who relied on
it to interpret the speech act correctly (62 %). So, as in Situation Two, it
appears that many students rely on the formulaic expression to help them
in speech act comprehension.
In Situation Five, the speaker asked 2Deberias conocer a su mejor
amigo, Francisco? 'Should you meet his best friend, Francisco?' which
represents a yes-no question and not a suggestion in Spanish. The speaker
made an effort to use question intonation in expressing this question, but
the rise in pitch seems to have been too subtle for all levels. Moreover, the
situation lends itself to a suggestion, since the speaker is saying that she
knows the listener is shy and wants to talk to Eduardo. Suddenly, she asks
if the listener should perhaps get to know his best friend, Francisco, who
is very nice. Without the question intonation, this utterance expresses a
suggestion, and the speaker did not make the interrogative intonation
clear enough, as verified by some native speakers of Spanish who also
saw the videotape. Some students in all three groups heard the utterance
as a suggestion (first-year: 22 0/0; second-year: 12 %; advanced: 53 % -
see Table 3). Thirty percent of the first-year students heard another
act, usually labelled "description/information" or "opinion," which in-
dicates that they comprehended the statements about the friend more
than the suggestion to meet him. Almost half of the second-year in-
formants did not understand at all. Some of the advanced students (15 0/0)
also heard the same speech act as some of the first-year group, labelled
"description" or "opinion." In Column 5 of Table 4, the 21 % exact
reproduction of the utterance is misleading, since most of these in-
formants did not use any punctuation at all in writing their replies.
Therefore, they were probably reporting a suggestion rather than a
question.
In general, the data suggest that the more advanced students, who are
more proficient in listening comprehension, are better at understanding
the global intent of the speech act, as expected. As far as noticing the
negative element in some of the interrogative suggestions, it appears that
only some of the advanced students notice this element, or at least only
some can notice it and retain it long enough in their memory to write it
down. These students were most competent in reproducing the mild
rebuke of Situation Two, which was also the shortest speech act, followed
by the suggestion of Situation Four. Both of these utterances were
preceded by the 2Por que no? expression. This indicates that students rely
heavily on these expressions to help them understand utterances. It is
270 Dale April Koike
interesting that in Situation Two, many students reproduced the utter-
ance exactly (
CC
2
Por que no viniste a verme and four of those
students still identified it as a suggestion, showing that they ignored the
verb tense cue.
4.3.2. Judgments of the speaker
Table 5. Mean reaction by group in each situation
Situation 1 aggr rude .,.com
strong -friend
First year 2.3 2.1 2.5 3.1 2.3
Second year 2.6 2.2 2.5 3.5 2.3
Advanced 2.2 1.6 1.9 3.3 1.7
Situation 2
First year 3.2 3.0 2.8 3.4 3.2
Second year 3.2 3.0 2.7 3.6 3.1
Advanced 3.7 3.5 2.8 3.9 3.8
Situation 3
First year 3.1 2.8 2.9 3.7 3.0
Second year 3.5 2.9 3.1 3.5 3.0
Advanced 2.7 2.0 2.2 3.2 2.1
Situation 4
First year 2.2 1.9 2.2 3.1 2.0
Second year 2.4 1.6 2.3 3.1 2.0
Advanced 1.9 1.6 1.8 2.8 1.6
Situation 5
First year 2.2 1.8 2.3 2.9 1.7
Second year 2.4 2.0 2.9 2.6 1.9
Advanced 2.2 1.6 1.7 3.5 1.5
Situation 6
First year 3.2 2.6 2.3 3.5 2.3
Second year 3.3 2.5 2.1 3.6 2.0
Advanced 3.5 2.3 2.0 3.6 1.9
(n =85; First yr. =37; Second yr. =25; Adv. =23)
Key:
aggr =aggressive
rude =rude
-com =non-communicative
strong =strong (personality)
-friend =unfriendly
Transfer of pragmatic competence 271
Table 5 shows the mean score given to each speaker by the three groups
of students based on the five characteristics of aggressiveness, rudeness,
communicativity, strength, and friendliness. The top score was 5,
denoting the negative characteristic, and the bottom score was 1,
denoting the most positive. Only those students who rated all six
speakers (Situation Seven was not rated) were included in the tabulations
of results (n =85). In a general overview of the results, it appears that the
first- and second-year students had a similar perception of the speakers,
and the advanced students differed from them, at times, significantly. If
this is true, then it indicates that people's opinions of a speaker change
when they understand the linguistic intent of the speaker's message.
When they do not understand or misunderstand the message, they are
probably reacting to what they thought they heard or are basing their
judgments on other features of the utterance, such as intonation, or the
speaker's physical appearance or gestures.
Returning to the question of how students reacted to the Spanish
negative interrogative suggestions when they heard them, the answers of
the advanced students were examined in detail since these students were
more likely to have understood the intended speech act. Of the ten
advanced students who correctly reproduced the negative interrogative
suggestion in Situation One, three were non-native Spanish learners and
six were bilingual Chicano Spanish speakers. Two of the non-native
students rated the speaker toward the positive end of the scale with low
scores, except for the "strong" category. These two students gave the
speaker a 4, tending toward "strong." The mean score by the three non-
natives for this item was 3.7. Three of the six Chicano students also rated
the speaker with a 4 in the same item. Two students also rated the speaker
with 4's in the non-communicative category, and one found the speaker
very unfriendly. The mean score for these Chicano students in the
"strong" item was 3.0. The mean for all advanced students was 3.3. The
results may reflect some variance in interpretation of the "strong/weak"
characteristic. That is, some students may have believed that it is more
positive to have a "strong" as opposed to a "weak" image, and therefore
rated the speaker as strong when they wanted to convey a positive rating.
If this were the case, then the scores for this category should reflect a
score on one end of the scale when most of the other categories reflected
scores on the opposing end. While the scores for the "strong" item are
sometimes higher than all. the others, they are not consistently so, i. e.,
when students seemed to be reacting negatively towards a speaker, the
"strong" scores would tend toward the higher end of the scale. Another
272 Dale April Koike
interpretation of the results is that the students may have reacted to the
negative interrogative suggestion, and found the speaker to be strong in
the sense of "strong-willed" or domineering. This interpretation seems to
be more viable for these informants.
None of the ten advanced students who understood the utterance in
Situation One and also noticed the negative element responded in a
negative way, and in fact, they generally agreed with the speaker. Four
first-year students thought they heard a rebuke, and two of them
responded with answers that can be described as defensive ("The last test
was not that easy"; and "I attend class and study very hard.") but not
negative. The two second-year and two advanced students who thought
the speaker was expressing a rebuke did not react negatively. Of course,
it is not expected that students would respond to a teacher's suggestion
or rebuke in an openly negative way, given the deference factor and
power differential. In general, then, the responses were not negative, even
when the students believed they were being scolded.
Those questionnaires that indicated that the speech act in Situation
Three (2No deberias hablar con la profesora ... ?) was considered a mild
rebuke were examined. Of the eight first-year students who heard a mild
rebuke, two wrote no response. The others did not write negative
responses (e.g., "I have a class"; "I will work harder."). Six second-year
students also interpreted the utterance as a mild rebuke, and two had no
response. The other answers were not negative in nature, ranging from a
request to repeat the utterance to statements regarding the teacher and
the way she helps the students. The two advanced students, however, did
react somewhat negatively, responding with "I don't feel sorry for you for
getting in trouble with the teacher" (a response that is inappropriate for
the utterance given) and "Hey, I wanted sympathy, not a reprimand."
Thus, the more proficient, advanced students seem more likely to respond
verbally to a peer in a negative manner if they misunderstand the
utterance to convey a negative meaning, probably because they are more
capable of responding. Of those three students who did understand the
utterance correctly and also noticed the negative element in the sugges-
tion, none gave a negative response.
A Repeated Measures Analysis of Variance was performed, examining
factors of the situation and the characteristic in question, and the
performance of the three groups. The results, seen in Table 6, show that
there is a significant difference in how the informants of the three groups
rated the situations and characteristics, but the difference is correlated to
the particular situation and the characteristic in question.
Transfer of pragmatic competence 273
Table 6. Source of variance (Repeated Measures Analysis of Variance - Tests of
hypotheses)
df 5S MS F
P
Within groups
Situation 5 370.30 74.06 54.21 0.0001
Character 4 362.71 90.68 113.86 0.0001
Between groups
Group 2 22.22 11.11 3.82 0.03
Error 82 238.74 2.91
5. Conclusions
Regarding the research questions posed at the beginning of this study, the
data show the following:
(a) Do English-speaking students of different levels of Spanish
language proficiency understand the intent of the suggestion when the
Spanish form is similar but different from the English one:} and expresses
a different intent than that of the English form?
The first- and second-year students had difficulty in compre-
hending the global intent of the speech acts in these situations while
more of the advanced students were more successful. The advanced
students performed better in those situations in which the head
act was expressed in a form using the same formulaic expressions
as those in English, namely "Why don't you" and "If I were you, I
(would)." They were slightly less successful in Situations One and
Three in which the forms were similar but different from English in
that they were negated in Spanish. All groups performed poorly in
comprehending the intent of Situation Five, in which the head act was
uttered with interrogative instead of declarative intonation, and repre-
sented a speech act other than one anticipated from the context. The
fault, however, probably lies in the fact that the speaker did not make the
intonational difference clear enough. All the results point to the fact that
students of all three groups transferred their first language speech act
knowledge in differing degrees in understanding the foreign language
speech act.
(b) Do differences in form cause misunderstanding and negative reac-
tions to the speaker?
274 Dale April Koike
The data showed that there was misunderstanding of the intent of the
speech act by about half of the advanced students and about 60 to 75 %
of the first- and second-year students, who performed similarly.
In Situation Three, the first- and second-year students tended to judge
the speaker as more aggressive, non-communicative, and strong. It should
be remembered that some of the informants in these two groups also
heard the speech act (i. e., 2No deberias hablarle a la profesora ... ?) as a
mild rebuke (18 % - see Table 2) or a complaint (24 %). In this situation,
some correlation could be drawn between the misinterpretation of the
speech act and the tendency to react negatively to the speaker for some
characteristics.
As for negative reactions to speakers due to the negative element in
interrogative suggestions, the data are too few to draw any definitive con-
clusions. The preliminary finding in this study, however, is that advanced
non-native speakers who do hear the negative element do seem to react
more negatively than the bilingual Chicano students who also hear the
negation. If this finding is valid, then there is a possible source of negative
reaction to Spanish speakers based on this lexical difference in inter-
rogative suggestions.
(c) How do the students respond to the speech act?
In general, many students of all levels were able to say something in
response to all the speech acts that would lead to further interaction and
input from the speaker, which would probably then lead to an under-
standing of the speaker's true intent through negotiation for meaning.
Even when learners noticed the negative element in the negative inter-
rogative suggestions or misinterpreted the speech act according to an
English interpretation, they were not likely to respond verbally in a
negative manner and thereby cause further miscommunication and
misunderstanding.
(d) Is there a marked difference in understanding of the speech act
according to the level of foreign language proficiency?
The data indicate that the advanced students were much more
competent in understanding the true intent of the speech acts than the
first- and second-year students.
In conclusion, the results of this limited experiment indicate, on the
positive side, that learners usually try to keep the interaction going with
the speaker through their responses to speech acts even when they only
partially understand or really don't understand at all. In all probability,
this continued interaction would eventually lead to comprehension of the
speaker's intent. On the negative side, the results suggest that speech act
Transfer of pragmatic competence 275
comprehension is difficult for first- and second-year language students,
even when the utterances are formulated in the same way in both
languages.
6
Comprehension is more likely to occur at the higher levels of
language study when students are able to understand larger units of dis-
course in a context. A few of the responses to the negated interrogative
suggestions, however, suggest that some non-native advanced students
misunderstand the utterance due to an analysis of the components of an
utterance and a transfer of first language speech act knowledge and the
lexical and syntactic elements associated with particular speech acts in
the first language.
Regarding transfer in the listening comprehension skill, these data
reveal that the transfer strategy is applied by learners at different levels of
proficiency, leading to some correct and some incorrect hypotheses about
the input. The students seem to transfer pragmatic knowledge in
matching what they can understand of the utterance to the context and
other cues such as intonation. The more advanced learners, who can
begin to analyze the input more closely (i. e., pay attention to details of
lexicon and structure) can sometimes make incorrect hypotheses about
those details they find if the structural/lexical items match those of the
first language and if they transfer their first language knowledge of those
items to the foreign language situation. Thus, a secondary type of transfer
is demonstrated.
Finally, on a pedagogical note, foreign language educators should be
aware that possibilities for misunderstanding and miscommunication
exist for students, even more so in the advanced levels of language study
when students are more likely to analyze the components of utterances
they hear. Many times it is assumed that if the message looks and sounds
the same grammatically and lexically in both languages, then students
should be able to understand through transfer and analogy. The results in
this study indicate that the comprehension process is much more
complex, especially in the case of speech acts, since they are so context-
dependent for meaning. That is, learners must not only learn to decode
at the utterance level, but also learn to read the context and its relation-
ship to the utterance. They must also learn differences between the first
and target language formulations of the speech acts, such as the use of the
negative in interrogative suggestions. This process implies a knowledge of
the target language speech acts at both the grammatical/lexical level as
well as the pragmatic level of use. This in turn implies an ability to make
pragmatic adjustments in situations of language contact (Blum-Kulka
1989: 65; Weizman - Blum-Kulka 1987: 65), whether they be in the
276 Dale April Koike
classroom or in a "real life" language interaction. Students of Spanish
need to be exposed not only to the language itself, but to a contextual-
ized, interactive language, such as through videotapes. Only through
exposure to contextualized language at all levels will students become
truly proficient in language use, and understand the target language ways
of speaking.
7
Appendix one
Note: All scripts below represent what the native speakers actually said in the
videotapes.
Situacion 1: (suggestion)
Professor sitting in office. (female - Colombia)
Professor: jHola! Pasa. Si, tengo tiempo para hablarte ahora. preocupado
con la nota que sacaste en el examen parcial? Pero no fue un examen muy dificil.
has pensado en estudiar junto con tus colegas de la clase ahora? 'Hi!
Come in. Yes, 1 have time to talk to you now. Are you worried about the grade
that you got on the midterm exam? But it wasn't a very hard exam. Urn - have
you thought about studying together with your fellow students in the class now?'
Situacion 2: (rebuke)
Professor in office. (male - Cuba)
Professor: No se si tu problema tiene que ver con ellibro de texto, que probable-
mente te cuesta mucho trabajo entender. Bueno, que no viniste a hablarme
antes? 'I don't know if your problem has to do with the textbook, which probably
is very difficult for you to understand. Well, why didn't you come to speak to me
earlier?'
Situacion 3: (suggestion)
Young man at table. (male - Spain)
Chico: Si, yo tambien creo que es injusto que la profesora nos de tanto trabajo.
Pero es mas injusto que ella te de un trabajo extra. No puedo creerlo. 2No
deberias ir a hablarle y pedir que te de credito extra por ese trabajo? Asi tal vez
suba tu nota en la clase. 'Yes, I, too, believe that it's unfair for the teacher to give
us so much work. But it's even more unfair for her to give you extra work. 1can't
believe it. Shouldn't you go and talk to her and ask her to give you extra credit
for that work? That way maybe your class grade will go up.'
Situacion 4: (suggestion)
Young man in room drinking coffee. (male - Puerto Rico)
Chico: jHola! Estoy aqui tomando un cafecito antes de ir a la clase. Me dices que
necesitas las hojas de tarea que la profesora nos dio la semana pasada, pero no
tengo las mias. Por que no copias las hojas de Raquel? Ella esta por aqu t cerca.
(Looks at watch). Tenemos unos veinte minutos. 'Hi! I'm here drinking some
Transfer of pragmatic competence 277
coffee before going to class. So you say that you need the worksheets that the
teacher gave us last week, but I don't have mine. Why don't you copy Raquel's?
She's around here. We have about twenty minutes.'
Situacion 5: (question)
Girl in room. (female - Guatemala)
Chica: no te pongas triste. Se que tienes muchas ganas de conocer a
pero que eres demasiado timida para hablarle primero. tratar
de conocer a su mejor Francisco? Es muy simpatico. 'Look, don't be sad.
I know you really want to meet Eduardo, but you're too shy to say something to
him first. Should you try to meet his best friend, Francisco? He's really nice.'
Situacion 6: (request)
Girl coming into room. (female - Peru)
Chica: lMaria! lHola! te vengo a devolver ellapicero que dejastes en mi casa
anoche. iQue bonita tu blusa! (No podrias prestarmela para esta noche? 'Maria!
Hi! Hey, I've come to return your pen, which you left at my house last night.
What a pretty blouse! Could you lend it to me for tonight?'
Situacion 7: (suggestion)
Young woman at table. (female - Panama)
Chica: S yo se que es muy pero tienes que hacer algo. Marta no te puede
tratar asi. Si yo fuera Ie diria que tienes los mismos derechos que los
mismos derechos que cualquier ser humano. 'Yes, I know it's hard, but you have
to do something. Marta can't treat you like this. If I were you, I would tell her
that you have the same rights as she does, the same rights as any other human
being.'
Appendix two
Cuestionario
Directions: You will see a series of videotaped interactions, in which a speaker is
supposedly talking to you. After each interaction, the tape will be stopped and
you will have a few minutes to answer the questions below. The first question
asks you to write the first thing you would say to the speaker on the tape in
response to what he or she says. The next question asks you to identify the main
gist of what the speaker said, and if possible, reproduce it, and the last question
asks you to react to the speaker.
Situation 1: You go to see your Spanish instructor in her office because you are
having some trouble.
1. Please write down a response in English or Spanish to what the speaker has
just said to you.
2. What was the main gist of what the speaker said? (Circle one)
A. Invitation
B. Apology'
278 Dale April Koike
C. Request
D. Suggestion
E. Order
F. Information question
G. Mild rebuke
H.Other--------------------------
If you can remember how the speaker said the last sentence, please write it here
in English or Spanish.
3. Please rate the speaker in terms of each characteristic, which you can base on
your reaction to the speaker. Circle one number for each item.
a. aggressive passive/gentle
5 4 3 2 1
b. rude polite
54321
c. non-communicative communicative
5 4 3 2 1
d. strong weak
5 4 3 2 1
e. unfriendly friendly
5 4 3 2 1
Situation 2: You go to see your other Spanish instructor, because you are also
having problems in that class. (Same Questions 1-3)
Situation 3: You are in the library and run into one of your classmates and com-
plain to him. (Same Questions 1-3)
Situation 4: You go to the Union and see one of your classmates and ask him a
question about the work.
Situation 5: You (a young woman) sit down to chat with one of your friends
about another person.
Situation 6: You (a young woman) are in your room when one of your friends
drops in.
Situation 7: You sit down to chat with one of your friends about another person.
Notes
1. Investigations by Schumann (1979) and Zobl (1980a; 1980b) found the use
of preverbal negation by Spanish-speaking learners of English (e. g., "I no
want that.") as evidence for an order of acquisition of morphemes in the
second language or foreign language, based on transfer from the first language.
I mention these studies here because they are relevant to the topic of the
acquisition of negation. The negation in the sentences examined in this study,
Transfer of pragmatic competence 279
however, does not change the truth value of the speech act utterances. This
difference could be crucial in the way learners deal with negation in acquisi-
tion.
2. Two forms in Spanish are offered here, corresponding to what appears to be
dialectal variation. I thank Ninfa Burgos-Kohler for pointing this out to me.
The symbol # is used to denote a sentence that, pragmatically speaking, is not
appropriate for the context.
3. The difference discussed here is really a difference between polite and deferent
interaction, discussed in Koike (1992: 30-34). Briefly, in deferent interaction,
social distance to the point of little or no direct interaction is emphasized. In
polite interaction, the element of respect while maintaining direct interaction
is present.
4. Although they were allowed this freedom in expressing themselves, the
speakers actually strayed very little from the script given to them. Their speech
acts were checked for extra softeners of illocutionary force, but none were
found. Two elements that could not be controlled, however, were intonation
and facial expression. All were asked, however, to be serious and not to smile.
S. Since the purpose of this study is to examine the suggestions in the data as a
whole, the tables do not show categories of the data according to register.
6. These findings contrast to those of a previous study (Koike 1989: 282-283)
in which students showed that almost all students of first, second, and
advanced levels demonstrated an understanding of speech acts. It should be
noted, however, that (a) those speech acts were simpler in language; (b) they
were read to the students by their own teachers, instead of by a series of un-
familiar native speakers; and (c) the rate of speaking was probably slower than
in normal speech.
7. I thank the graduate assistant instructors and undergraduate students of the
Spanish classes at the University of Texas at Austin who graciously agreed to
participate in this experiment.
References
Blum-Kulka, Shoshana
1989 "Playing it safe: The role of conventionality in indirectness", in:
Shoshana Blum-Kulka - Juliane House - Gabriele Kasper (eds.),
37-70.
Blum-Kulka, Shoshana - Juliane House - Gabriele Kasper
1989 "Investigating cross-cultural pragmatics: An introductory over-
view", in: Shoshana Blum-Kulka-Juliane House-Gabriele Kasper
(eds.), 1-34.
Blum-Kulka, Shoshana - Juliane House - Gabriele Kasper (eds.)
1989 Cross-cultural pragmatics: Requests and apologies. Norwood, NJ:
Ablex.
Brown, Penelope - Stephen Levinson
1987 Politeness: Some universals In language usage. Cambridge:
Cambridge University Press.
280 Dale April Koike
Larry
Larry Selinker
Larry Selinker
1983
1992
1992
1983
1992
1994
1992
Corder, S. Pit
1975 "The language of second-language learners: The broader issues",
Modern Language Journal 59: 409-413.
"A role for the mother tongue", in: Susan Gass
{eds.),85-97.
"A role for the mother tongue", in: Susan Gass
(eds.), 18-31.
Gass, Susan
1989 "How do learners resolve linguistic conflicts?" in: Susan Gass -
Jacquelyn Schachter (eds.), 183-199.
Gass, Susan - Jacquelyn Schachter (eds.)
1989 Linguistic perspectives on second language acquisition. Cambridge:
Cambridge University Press.
Gass, Susan - Larry Selinker (eds.)
1983 Language transfer in language learning. Rowley, MA: Newbury
House.
Language transfer in language learning. Amsterdam: John Ben-
jamins.
Hatch, Evelyn (ed.)
1979 Second language acquisition: A book of readings. Rowley, MA:
Newbury House.
Hobbs, Dianne
1990 Gender-based strategies in issuing directives in Mexican Spanish.
[Unpublished Ph. D. dissertation, The University of Texas at Austin.]
Kellerman, Eric
1977 "Towards a characterization of the strategy of transfer in second
language learning", Interlanguage Studies Bulletin 2.1: 58-145.
"Now you see it, now you don't", in: Susan Gass - Larry Selinker
(eds.), 112-134.
Koike, Dale
1989 "Pragmatic competence and adult L2 acquIsItIon: Speech acts in
interlanguage", Modern Language Journal 73.3: 279-289.
Language and social relationship in Brazilian Portuguese: The prag-
matics of politeness. Austin, TX: University of Texas Press.
"Negation in Spanish and English suggestions and requests:
Mitigating effects?" Journal of Pragmatics 21: 513-526.
Schachter, Jacquelyn
1983 "A new account of language transfer", In: Susan Gass
Selinker (eds.), 98-111.
"A new account of language transfer", In: Susan Gass - Larry
Selinker (eds.), 32-46.
Schumann, John
1979 "Second language acquisition: The pidginization hypothesis", in:
Evelyn Hatch (ed.), 256-271.
Searle, John
1979 Expression and meaning: Studies in the theory of speech acts.
Cambridge: Cambridge University Press.
Transfer of pragmatic competence 281
Selinker, Larry
1972 "Interlanguage", IRAL 10: 209-231.
Weizman, Elda, - Shoshana Blum-Kulka
1987 "Identifying and interpreting translated texts: On the role of prag-
matic adjustment", Indian Journal of Applied Linguistics, 13.2:
61-73.
Zobl, Helmut
1980a "The formal and developmental selectivity of L1 influence on L2
acquisition", Language Learning 30.1: 43-57.
1980b "Developmental and transfer errors: Their common bases and
(possibly) differential effects on subsequent learning", TESOL
Quarterly 14.4: 469-479.
Part III
Applications
Suggestions to buy: Television commercials from
the U. S., Japan, China, and Korea
Richard Schmidt - Akihiko Shimura - Zhigang Wang -
Hy-sook Jeong
1. Introduction
This chapter deals with television commercials as suggestions to viewers
to buy consumer products, comparing television commercials from the
United States, Japan, the People's Republic of China, and South Korea
from this perspective.
1
It is intended to complement and augment other
studies of differences in the realization of speech acts across cultures, but
is different from most other studies in several respects.
1) It is often assumed that the most problematic issue both in speech
act theory and in cross-cultural communication is the determination of
what is meant by what is said. It is commonly pointed out that our goal
in conversation is to convey our intentions in socially appropriate ways
(Aston 1993; Grice 1975; Searle 1969; 1976) and that failure to convey
or interpret intentions may be the most important source of cross-cultural
communication breakdown (Gumperz, 1982).
However, the goal of a television commercial is obvious; it is to
persuade consumers to buy a specific product. Successful communication
is less a matter of getting television viewers to recognize this illocutionary
force (indeed, advertisers may attempt to mask this goal) than a function
of the persuasive impact of the commercial, including the linguistic and
nonlinguistic strategies used to persuade.
2
2) Speech act analyses have been based on several kinds of data,
including native speaker intuitions, interviews, naturally occurring
utterances in face-to-face interaction, role play, spoken or written lan-
guage elicited through discourse completion tests, and various perception
tests such as card sorts, paired comparisons and rating scales (see Kasper
- Dahl 1991 for review). However, we are unaware of any other study
of speech acts in different cultures that is based on data from television
commercials. Our data are not just unique but also represent artful rather
than naturally occurring discourse, carefully scripted by professional
writers, although the television commercial is naturally occurring lan-
286 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
guage in another sense, not experimentally elicited for the purpose of
linguistic analysis.
3) Very useful work has been done in the cross-cultural comparison of
speech act behavior by focusing on patterned variation in speech act
realizations. The distribution of linguistic strategies for performing
particular speech acts has been matched with speaker and hearer vari-
ables such as age, sex, social distance, and relative power, together with
the degree of imposition of the speech act involved, factors which are,
according to the theory of politeness (Brown - Levinson 1987), the
primary determinants of linguistic choices in speech act realization. In the
case of the television commercial, we might expect similar variation in
speech act realization when characters on screen talk to one another or
when particular segments of the viewing audience are targeted (e.g.,
children, adolescent males, etc.), but the central notions of speaker and
hearer are problematic when applied to the language of advertising.
Goffman (1981) has criticized the commonsense notion of speaker,
pointing out that a speaker may be the one who speaks the message, the
one who has encoded it, or the one who is committed to the beliefs ex-
pressed. In ordinary conversation, these three typically coincide; in role
play, there may be no committed speaker; and in commercials, these roles
are distributed among actors, copywriters, producers and directors, and
the product manufacturer. As for the hearer, Lakoff suggests that in
advertising language and in persuasive language in general, there is no
addressee, but only an audience (Lakoff 1982:31). In addition, while
commercials may use a particular kind of language in order to influence
specific target audiences and may exploit such roles as celebrity/fan, these
devices are often used strategically, creating situations and relationships
rather than being sensitive to them.
4) Many studies of cross-cultural variation in speech act realization
have direct relevance for second language learners whose goal is to inter-
act effectively with native speakers of the target language under different
social constraints. We do not claim such implications for our work, but
we do see this study as basic research for the teaching of language in
business contexts, a field for which there is great demand and little
empirical research (Johns 1986).
Television commercials from the u. and Korea 287
2. Television commercials as suggestions
Television commercials provide easily obtainable data that are relevant
for a number of sociolinguistic concerns. Commercials have been
analyzed as expressions of cultural codes and mythologies (Barthes 1972;
Hall - Saracino - Resh 1979; Mueller 1987; Sherry - Camargo 1987);
as a medium-specific example of the register of advertising language with
a focus on novel uses of language (Agoston - von Raffler-Engel 1979;
Bhatia 1987; Leech 1966; Masavisut - Sukwiwat - Wongmontha
1987; Moeran 1985; O'Barr 1979); as n:anipulative or deceptive lan-
guage (Bolinger 1973; 1980; Coleman 1990; Harris 1983; Vestergaard -
Schreder 1985); and as an exemplar of the broader category of persuasive
discourse (Lakoff 1982; Schmidt - Kess 1985; 1987). Geis (1982) has
perceptively analyzed a number of the pragmatic aspects of American
television commercials, including the ways in which product claims are
interpreted through reference to conversational maxims, but no study to
date has dealt with television commercials from a speech act perspective.
As a speech act, the television commercial is clearly some sort of direc-
tive (Searle 1976) or impositive (Leech 1983). The essential point of a
television commercial, the reason an advertiser purchases time, is that it
is an attempt to get some hearer or audience, viewers in their role as
consumers, to perform some future action, that is, to buy a product. We
propose that television commercials are best viewed as suggestions to
buy, however, rather than some other species of directive, such as
requests, orders, or hints. Commercials do not seem to be requests,
because they rarely attempt to engage the hearer's compliance on the
ground that the speaker wants or needs the act to be done. They are
not orders, because advertisers cannot expect consumers to buy a product
as a consequence of the advertiser's or manufacturer's authority. They are
not hints, because their illocutionary force is transparent (Weizman
1993). A commercial can only suggest or recommend, persuading the
viewer "to consider the merits of taking the action in virtue of the
speaker's belief that there is sufficient reason to act" (Fraser 1983: 40).
Geis has argued that syntactic imperatives, observed to be common in
advertising (Leech 1966), are to be viewed as suggestions rather than
orders (Geis 1982:19). However, in analyzing commercials as
suggestions, we are somewhat hampered by the lack of detailed studies of
this speech act, particularly from a cross-cultural perspective. Requests
have been investigated extensively (see Blum-Kulka - House - Kasper
1989 for summary), but the speech act of suggestion, a cousin of the
288 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
request, has been much less studied. We have located only two data-based
cross-cultural studies, Rintell's (1979) brief comparison of suggestions in
Spanish and English and Banerjee and Carrell's (1988) comparison of sug-
gestions by native and non-native speakers of English.
Although our discussion of television commercials so far has assumed
that a commercial as a whole is a speech act with a unifying illocutionary
point, it may be preferable to view the commercial as a whole as a speech
event, the internal structure of which consist of a sequence of utterances
that may differentially support such functions as suggesting, informing,
entertaining, and the like. Our analysis will be based on a distinction
between head acts within the discourse of television commercials, those
utterances or parts of utterances that directly realize the act of suggesting,
and various supporting moves that provide grounds or reasons for some-
thing to be done or remove objections to the proposed action (Blum-
Kulka - House - Kasper 1989; Blum-Kulka - Olshtain 1984;
Edmondson - House 1981). We view the underlying discourse structure
of a commercial as in example (1).
(1) Head act
Consumer should buy,
use (etc.) the product
+
because
Supporting moves
Product is effective, stylish
(etc.) Product will
make consumer
happy, healthy, young
(etc.)
Commenting on the application of the distinction between head act and
supporting moves to data derived from discourse completion tests, Blum-
Kulka and Olshtain (1984) and Edmondson and House (1981) have ob-
served that distinguishing between the two is difficult and subjective,
because what may be a supporting move in one case may function as the
head act elsewhere, for example when a hint does not preface a request but
by itself conveys the force of requesting. We have attempted to minimize
this problem by providing as strict a separation as possible between head
acts and supporting moves or reasons. Operationally, we have defined the
head act of suggestion in television commercials as any utterance or part of
an utterance that linguistically refers to the viewer or some other consumer
buying an advertised product or interacting with the product in some other
way, such as using it or enjoying the benefits of owning it.
We therefore allow the possibility of commercials that have more than
one utterance classified as a head act, as well as commercials that have no
Television commercials from the U.S., Japan, China, and Korea 289
head act as we have defined it, while admitting that much of the
remaining language in a commercial may be suggestive in a broader
sense. Even this relatively strict definition, however, does not result in the
exclusive assignment of each turn or utterance to only one category,
either head act of suggestion or supporting move. Consider, for example,
E37 from our sample of English commercials:
E37 !'m gonna take what doctors would take, wouldn't you?
In this example, the only linguistic material that directly refers to a
desired action to be taken by the viewer is the question tag wouldn't you?
However, by our definition, there are two other utterance parts we
identify as additional head acts. The actor uttering the lines of the
commercial, acting as a surrogate consumer, states that he is going to take
the product (!'m gonna take) and will presumably have to buy it in order
to do so. He also asserts that other consumers would do the same (what
doctors would take). These different strategies for suggesting are
common in our data, and we therefore code for three instances of the
head act in this case, although the claim that doctors would use the
product is also clearly presented as a reason (supporting move) for the
viewer to act.
3. The study
In this chapter, we look at American television commercials and compare
them with commercials from three Asian countries. Japan, the People's
Republic of China, and South Korea. There are reasons to think that a
comparison of American and Asian commercials may uncover some
interesting differences. A great deal of advertising research supports the
claim that American advertising (in all media) is primarily persuasive
rather than informative (Dowling 1980; Hong - Muderrisoglu -
Zinhan 1987; Hunt 1976; Kaynak - Mitchel 1981; Laczniak 1979;
Madden - Caballero - Matsukubo 1986; Resnik - Stern 1977;
Stern - Krugman - Resnik 1981). But the view from Asia is rather
different.
In China, advertising was banned during the Cultural Revolution
(1966-1976), and a modern advertising infrastructure has been
developed only since 1978, as part of a rapid shift from Marxist socialism
290 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
to market socialism (Rice - Lu 1988; Tse - Belk - Zhou 1989). The
official view is that the functions of advertising are to promote produc-
tion, invigorate the economy, increase consumer convenience and guide
consumption, develop international economic activities, and promote
socialist moral standards (Central Administration for Industry and
Commerce, cited in Rice - Lu 1988). Ho and Sin (1986) found
that Chinese managers hold that the main purpose of advertising is to
inform.
In both japan and Korea, Miracle (1987) claims that advertisers rely
on a feel-do-learn strategy, in which the primary goal of advertisers is to
entertain and establish feelings and moods that are transferred to the
product, as opposed to the predominant Western learn-do-feel sequence,
in which advertising presents reasons for buying, with positive feelings
the result rather than the precursor of purchase. japanese advertising
practices have been quite thoroughly studied, and from a number of
impressionistic studies (Fields 1983; Miracle 1985; 1987; Yamaki n. d.) a
consistent picture has emerged. japanese television commercials are said
to be evocatively filmed, but extremely indirect in approach. In a study of
print advertising, Sherry and Camargo report that most japanese ads
"neither preach, promise nor praise; some don't even portray product
attributes" (Sherry - Camargo 1987: 181).
Based on the literature on advertising in the U. S., China, japan and
Korea, our initial hypothesis was the following:
H1 Suggestions will be more frequent in commercials from the u.S.
than in those from any of the three Asian countries.
We are equally interested in knowing what linguistic strategies are used
to convey suggestions in commercials. As noted above, previous studies
have commented on the high frequency of imperatives in English adver-
tising, linking the syntactic imperative to the function of suggesting, but
why this should be so is not entirely clear. Both Rintell (1979) and
Banerjee and Carrell (1988) observe that suggestions (for the benefit of
the hearer)are more likely to be expressed directly than are requests (for
the benefit of the speaker). If one salient feature of suggestions is that the
maker of a suggestion assumes no special authority over the addressee
(Green 1975), then perhaps there is little potential loss of face involved
in the use of the imperative, which is normally viewed as not allowing
options However, Banerjee and Carrell found that imperatives were not
common among suggestions elicited by discourse completion question-
Television commercials from the U. S., Japan, China, and Korea 291
naires, accounting for fewer than 10 % of the responses offered by both
native and non-native speakers (Banerjee - Carrell 1988: 331) and were
used only in situations requiring immediate attention. Leech (1966) sug-
gested that the high frequency of imperatives in British advertising might
be a function of several factors: the fact that commercials are necessarily
obvious in declaring their intention, plus the need for attention value,
memorability, and selling power.
Whatever principle may explain the occurrence of imperatives in
American and British advertising best, we do not expect imperatives to be
found as often in Asian television commercials. Even though language
may function pragmatically in rather different ways in persuasive
discourse than in ordinary conversation (Lakoff 1982), there is every
reason to expect cross-cultural differences in persuasive language, with
considerable influence form the norms of face-to-face interaction. For
japanese, it has been said that the bald, unadorned imperative is hardly
ever used, and that "the form is considered even too forceful for recipes
and other regular instruction" (Matsumoto 1988: 420). Takahashi (1987)
developed a taxonomy. for comparing levels of directness in japanese and
English directives (using role plays to elicit data), finding that japanese
employed a more indirect approach in general and were especially likely
to employ hinting strategies. While there is clearly a danger in exaggerat-
ing the stereotype of direct American usage as opposed to indirect
japanese speech act performance (Beebe - Takahashi 1989), observations
that japanese advertising is more evocative than persuasive also suggest
that imperatives will be infrequent in japanese television commercials and
that less direct strategies for making suggestions will be used.
For Korean and Chinese commercials, we have less information upon
which to base our hypothesis, since there have been no comparative
studies of suggestions or other directives in these languages that we are
aware of. Both Koreans and Chinese are often observed to be more
forthright and direct than the japanese, but Koreans and Chinese, like
japanese, are said to emphasize harmonious social relationships and to
disprefer any appearance of presumptuousness or over-eagerness,
preferences sometimes linked to discourse strategies such as a steady
buildup of information before making a request or arriving at the impor-
tant message (Kaplan 1966). We therefore hypothesize the following:
H2 Imperatives will be the most frequent form used to realize
suggestions in American television commercials, but not in any
of the three Asian languages.
292 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
3.1. Method
3.1.1. Data
The data for this study consist of one full evening of television broad-
casting in Honolulu (NBC), Tokyo (Fuji), Seoul (Korean Broadcasting),
and Beijing (CCTV) during the first week of March, 1988. Commercials
were recorded from the most watched station in each location at peak
audience times.
These video types yielded slightly more than one hour of commercials
in each language, from which a sample of 50 commercials was drawn for
each country. Our original intent was to balance the samples for products
advertised, but this proved impossible, because the most commonly
advertised products are not the same in each country. The Chinese data
contained numerous advertisements for washing machines and television
sets, not found in any of the other language databases. The English and
japanese tapes both contained a high proportion of automobile ads
(March is the traditional season for car sales in both countries), while the
Korean database contained fewer commercials for consumer durables of
all types than the other three language samples (Keown - Schmidt -
jacobs - Ghymn 1992). Because we were interested only in consumer
product advertising, the sample analyzed in this chapter consist of the
first 50 commercials from each database, after elimination of repeats,
public service announcements, commercials clearly of local rather than
national origin, film trailers, and promotional spots advertising other
television programming. We also eliminated from the Chinese sample a
number of commercials showing heavy industrial equipment, on the
grounds that few viewers could be considered potential consumers of
such products.
The samples also vary in length. Commercials in the U. S. vary between
10 and 30 seconds (occasionally longer), in five second increments.
japanese and Korean commercials tend to be shorter than those from the
U.S., while Chinese commercials are longer than those from the U.S. and
more than twice as long as those from japan. For the samples analyzed
here, the mean length of commercials for each language was as follows:
japanese, 13 sec.; Korean, 18 sec.; English, 22 Sec.; Chinese, 28 sec.
We believe that our samples are reasonably representative of prime-
time consumer advertising on major channels in the four countries, but
do not claim that our sample is representative of all television advertising
in any of them. For example, we would expect some important
Television commercials from the u. and Korea 293
differences in the language of American television commercials broadcast
at different times of the day, such as Saturday mornings or late at night.
We would also expect differences in commercials that are not nationally
distributed (e.g., local used car commercials) and for products sold
through mail-order and advertised primarily on cable stations.
3.1.2. Analysis
All commercials were transcribed, including spoken, printed and sung
language, and three translations were produced: a morpheme-by-
morpheme translation, a literal translation into English, and an idiomatic
or free translation. Except in cases where linguistic form is at issue,
examples are presented only in romanized transcription and free trans-
lation.
Each of the four authors of this chapter was responsible for the
analysis of commercials in his or her native language and the initial
assignment of utterances to categories, after which the examples were
discussed by all four researchers until consensus was reached on each
categorization.
3
Chi-square analysis was used to test hypotheses, and the alpha level
for significance was set at .05. Reported frequencies represent the number
of instances of a category in the whole set of 50 commercials for each
language (which may include more than one instance from a single
commercial), not the number or percentage of commercials in which
utterances of a particular type occurred. In addition to results addressing
our specific hypotheses, we will also present a number of post-hoc and
qualitative analyses, together with examples illustrating the tone of
commercials from each country.
3.2. Results and discussion
As indicated in Table 1, the hypothesis that suggestions would be more
common in American commercials than in those from any of the three
Asian countries was supported. Considering all types of suggestions
coded from our data - including suggestions made to either viewers or
on screen characters, testimonials and reported behavior referring to
other consumers - the set of English commercials contained a signif-
icantly higher number of suggestions than the Japanese, Chinese or
294 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
Table 1. Frequency of suggestions in the four language samples
Eng ]pns Chns Kor
Total number of suggestions: 78 35 35 53
All: X
2
= 24.73 df = 3 p<0.05
Ex] X2= 15.86* df = 1 p<0.05
ExC X2 = 15.86* df = 1 p<0.05
ExK X2= 4.27* df = 1 p<0.05
] xC
X2=
0.00 df = 1 n.s.
]xK X2= 4.18* df = 1 p<0.05
CxK X2=
4.18 df = 1 p<0.05
Suggestions to viewer only: 54 27 26 30
All: X2 = 15.44 df = 3 p <0.05
Ex]
X2= 8.50* df = 1 p<0.05
ExC X2 = 9.30* df = 1 p<0.05
ExK X2 = 6.36* df = 1 p<0.05
] xC X2 = 0.52* df = 1 n.s.
] xK X2 = 0.66* df = 1 n.s.
CxK X2 = 0.78* df = 1 n.s.
* Corrected value, because df =1 (Hatch - Lazaraton 1991: 405-406)
Korean samples. Korean commercials contained more suggestions than
either Chinese or Japanese commercials.
The distribution of linguistic forms in suggestions is shown in Table 2,
comparing imperatives (including those with please or a tag question
appended) to all other forms, but only including those suggestions
addressed directly to the viewer. We expect that suggestions addressed by
on screen characters to one another might exhibit some interesting
differences, possibly varying as a function of speaker and hearer sex, age,
status and role, but suggestions to on screen characters occurred too
infrequently in our data to permit meaningful analysis. Some types of
suggestions (reports of what other consumers have done) would not
permit the imperative at all.
As shown in Table 2, the hypothesis that imperatives would be the
preferred form for suggestions in American television commercials and
would be used less frequently in commercials from the three Asian
commercials was also supported. Japanese commercials also contained
significantly fewer imperatives than either Chinese or Korean, which
were not significantly different from each other.
Television commercials from the U.S., Japan, China, and Korea 295
Table 2. Suggestions to the viewer, Imperatives vs. other forms
Eng Jpns Chns Kor
Imperatives
Other forms
Total
Imperatives, All:
Imperatives, E x J
Imperatives, E x C
Imperatives, E x K
Imperatives, J x C
Imperatives, J x K
Imperatives, C x K
* Corrected values
Xl =19.25
Xl = 24.50*
Xl = 10.50*
Xl =10.80*
Xl = 4.07*
Xl = 4.27*
Xl = 0.55*
df =3
df =1
df = 1
df =1
df =1
df = 1
df = 1
p<0.05
p<0.05
p<0.05
p<0.05
p<0.05
p<0.05
n.s.
33
21
54
3
24
27
11
15
26
10
20
30
These two measures indicate that American television commercials
are more overtly suggestive than commercials from japan, Korea, and
China. One measure, the overall frequency of suggestions of
all types (Table 1), indicates that Korean ads are somewhat more
suggestive than commercials from japan or China, whereas the
frequency of imperatives (Table 2) indicates that japanese commercials
are the least suggestive in that respect. One possible interpretation
of these findings is that the persuasive function of television is
emphasized in American ads, whereas other functions are emphasized
in Asian commercials, such as simply providing information to
consumers. Another possible interpretation is that the function of
suggestion is accomplished just as effectively though less directly in Asian
commercials.
A caveat must be raised regarding the identification of syntactic
imperatives as a basic measure of directness in suggestions. Although
the syntactic imperative is the most direct possible way to make sugges-
tions in each of the four languages, we cannot assume functional
equivalence across languages; in fact, we know that the japanese imper-
ative is virtually a tabu form. Moreover, indirection in suggestions is not
one-dimensional. While attempts have been made to rank directive
utterances along a single scale of directness or politeness (Takahashi
1987), we find that in television commercials there are at least four ways
in which suggestions can be conveyed indirectly. We have already
recognized that what we consider supporting moves (reasons to buy, such
296 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
as product attributes) can function indirectly as suggestions in a broad
sense, just as hints function as indirect requests in face-to-face inter-
action. In addition, we find examples in our data of what we will call
indirection by participant shift, indirection by action shift, and linguistic
indirection.
3.2.1. Indirection by participant shift
Suggestions in commercials that are addressed directly to the viewer as a
potential consumer and that refer to the desirability of the viewer buying
the product or doing something with it (direct) can be distinguished
from suggestions that are directed at some other addressee or
simply report what other consumers have done (indirect). As already
indicated (see Table 1), suggestions to the viewer were the most common
type in our samples. In each language, we find suggestions addressed
to the viewer such as in sample E49, E36, J40, J8, C23, C31, K16,
and K42
E49 Consider the Acura Legend Coupe. (male announcer)
E36 So get confused. Shop where you but start at your
Buick dealer. (on screen actor, female)
J40 Pipp Erekiban 0 kau 18-kin to daia de dekita sutekina
puchi pendanto ga chuusen de atarimasu.
'If you buy Pipp's Erekiban now, you may win a petite pendant
made of 18K gold and diamonds.' (male announcer)
J8 Gankai no shiji ni shitagai tadashiku goshiyoo kudasai.
'Please follow your eye doctor's directions and use them
properly.' (printed message)
C23 Xinqiu nin de lixiang.
'Xinqiu stereo system, your ideal choice.' (male announcer)
C31 Qing nin fuyong tongrentang shengchan de kanglaoyannianwan.
'Please use the Kanglao Yannian produced by Tongrentang
Medicine Factory.' (female announcer)
Television commercials from the U. S., Japan, China, and Korea 297
K16 Hyokwalul senthaykhaseyyo.
'Choose the effect.' (on screen actor, female)
K42 Philyohal ttayman cokumssik ccaseyyo.
'When needed, press a little.' (on screen actor, male)
It is interesting to note who makes the suggestion to the viewer. In
English, such suggestions may be made by an on screen character,
through printed messages, or through song lyrics, but the most frequently
used strategy by far is the device of using the voice of an off screen an-
nouncer to make the suggestion (34 of 54 suggestions to viewers).
Overwhelmingly (in 33 of 34 cases), the off screen announcer is an adult
male. Although the observed frequencies are too small to permit tests of
statistical significance, interesting patterns can be seen in each of the
other languages. Only the commercials from China are like the American
ones in strongly preferring the off screen announcer as the primary source
of suggestions (18 of 26 suggestions), and in the Chinese commercials
two thirds of those suggestions were made by female announcers. In the
japanese commercials, the most common way to present suggestions was
through printed rather than spoken messages (13 of 27 suggestions to the
viewer). On screen actors made suggestions to the viewer as often as did
an off screen announcer (7 cases each). However, when an off screen
announcer makes the suggestion, japanese commercials are just as
gender-biased as American ones. All off screen voices except two were
male in our japanese sample. One was a child's voice, and one
commercial used a female off screen voice to represent the thoughts of a
character. The Korean commercials did not use printed messages for any
suggestions, used an off screen announcer for just over half of all
suggestions (16 of 30), and - like the Chinese commercials - did not
show gender bias; 9 off screen announcers were female and 7 were male.
The remaining suggestions to the viewer in the Korean commercials were
either spoken by on screen actors or presented in song lyrics.
A less direct way to ~ a k e a suggestion is to make it to an on screen
character, an actor appearing as a surrogate consumer. Such suggestions
might be made by an off screen voice, but we found no examples of this
strategy in any of our language samples. In each of the language samples,
there are examples of actors making suggestions to other actors, such as
samples E42, j2, C33 and Kll.
E42 Here, try this one. (male actor to female actor)
298 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
J2 Ojii-chan, Tansu ni gon katte kite kudasai na?
'Grandpa, please go and buy Tansu ni Gon, won't you?'
(daughter-in-law to father-in-Iaw
4
)
C33 Ni chi dian jianpixiaoshi wan jiu hao le, haoma?
'You will feel better after taking Jianpi Xiaoshi Wan, OK?'
(mother to son)
Kll Enni twuthongyakul tusil ilici.
'Sister, why don't you take a headache medicine?' (female actor
to female)
It would be interesting to compare commercials from these different
countries to see which role relationships are used most often in such cases
and who gives advice to whom, but it would take a larger sample than
ours to identify patterns.
Alternatively, still in their role as surrogate consumers, on screen
actors can indirectly convey a suggestion to the viewer or another on
screen actor by reporting that they have used a product and benefited
from it. We found examples of such testimonials from each country, such
as in E36, J37, Cl, and K32.
E36 Thank goodness for Pine Sol. (female actor)
J37 Boku wa itsumo hakuchuu doodoo suwan nan desu yo.
'I always drink Swan boldly in broad daylight.' (male celebrity)
Cl Wo young guo, xiaoguo hai bu chuone.
'I've tried it and found its effects not bad.' (female actor)
K32 Yocum daewoo patko salayo.
'We live now receiving good treatment' (male actor)
Finally, instead of having an on screen character report his or her
purchase or satisfaction with the product, advertisers may suggest in-
directly by reporting what other consumers have done. Again, we find
examples in each language sample as in E17, J41, C7, and K23.
E17 Since the ]ohnsons got their Mitsubishi Mirage, t h e y ~ v e been
driving a lot more. (male announcer)
Television commercials from the U.S., Japan, China, and Korea 299
J41 Katakori no tonari no Gen-san Pipp katte, nonde, kiita.
'Gen-san, our next door neighbor with the stiff shoulder,
bought Pipp and took it.' (on screen male actor to female)
C7 Luotuo jin wanjia, wan ja huanle duo.
'Camel comes to thousands of families and brings them more
happiness.' (female announcer)
K23 I taykeyto ce taykeyto soykoki Masna.
'At this home and that home too, Soykoki Masna.' (song lyrics)
Between-language differences in the distribution of suggestions to the
viewer, suggestions made to on screen characters, testimonials, and reports
were not significant. However, there is one additional strategy for convey-
ing suggestions indirectly that we have found only in the U. S. commercials,
a variant of the testimonial. This is modeled behavior, in which the on
screen "consumer" does not report past actions but is shown purchasing
the product or indicates that he or she is going to buy or use the product.
Five U.S. commercials used this strategy, as in samples E6 and E25.
E6 Oh! I'll take it home now. (female actor)
E25 Wear-dated, please. (female actor)
3.2.2. Indirection by action shift
Any suggestion involves some threat to an addressee's face, because
people do not, in general, want to be told what to do. The suggestion to
buy involves a literal cost as well. In advertising, various strategies are
used to present messages that emphasize the benefits and minimize the
costs to the buyer. Price mayor may not be mentioned. If it is, the price
may be presented as a reason for buying, either because the price is lower
than that of the competition or because it will be higher after a limited
offer expires. Other strategies involve the formulation of price ($ 39.95
instead of $ 40.00) or stating a base price in large print while mentioning
restrictions and exclusions in small print. (Both of these strategies were
found only in the U. S. commercials).
The suggestion that the consumer is to buy something, i. e., part with
money, can also be conveyed less directly by referring to consumer
300 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
actions other than buying itself. This is done in commercials from all
four countries. We identified the following categories of suggested
action:
Suggestion to buy: An utterance that lexically refers to buying, selling,
or ordering the product, or a related financial transaction, as in samples
25, J5, C38, and K20.
25 If y o u ~ r e one of the thousands of people who asked for a free
sample of Wear-Dated carpet with Stainblocker and tried it:l
y o u ~ d probably like to know what steps to take to buy it. (male
announcer)
J5 Nyuu konseputo gokoinyuu no kata ni Goto Kumiko orijinaru
terehon kaado purezento chuu.
'Giving away original Kumiko Goto telephone cards to those
who purchase New Concept.' (printed message)
C38 Dianhua dinghu:l shonghuo shangmen.
'Order by phone, delivered to your home.' (printed message)
K20 Ne fin Ramyon hana te saollay?
'Would you go buy one more Jin Ramyon?' (on screen actor,
father to daughter)
Suggestion to get: An utterance that lexically refers to some action
which implies the possibility or likelihood of purchase, for example,
going to a store or calling for information, as in samples 25, C9,
and K44.
25 Call for the dealer near you. (male announcer)
C9 Quing jizhu wo de shangbiao:l weili pai xiyiji.
'Please remember my brand, Weili washing machine.' (cartoon
character)
K44 Niksaykpyengul chacuseyyo.
'Look for the green bottle.' (female announcer)
Suggestion to use: An utterance that refers to a consumer using the pro-
duct in some way, as in samples 25, J39, Ci, and K12.
Television commercials from the U. S., Japan, China, and Korea 301
25 Throw in your dirtiest clothes, then toss in a Fab i-Shot pack.
(male announcer)
J39 Poora no Dei-ando-dei, Massaaji Kuriimu wa mizu de araina-
gaseru kara kantan ni tsukaemasu.
'As Pola's Day And Day massage cream can be washed off with
water, you can use it easily.' (off screen voice of female
character)
C1 Ni key fangxin de yong.
'You can use it without worry.' (on screen female actor, to female)
K12 Wuli emanun Heinz.
'Our mother uses Heinz.' (song lyrics)
Suggestion to enjoy benefits: An utterance that refers to the consumer
experiencing the benefits of owning or using the product, as in samples
43, J13, C35, and K24.
43 Instead of spending your evening creating a classic, you can
spend it enjoying one. (male announcer)
J13 Saa, hajimemasen ka, ii iki no shuukan?
'Let's start the habit of new breath, shall we?' (on screen male
celebrity)
C35 Ji jiang jianmei you xiang koufu.
'You can keep fit and enjoy gourmet's luck at the same time.'
(male announcer)
K24 Santtushan masulo kiekhaseyyo.
'Remember as a fresh taste.' (female announcer)
Suggestion with unspecified action: An utterance that implies that a
viewer will interact with the product in some way, but leaves the desired
action unspecified or makes a metaphorical suggestion, as in samples
41, J17, and K27.
41 Listen to the heartbeat of America. (song lyric)
302 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
J17 Kotoshi wa fain na nama biiru.
'For this year, fine draft beer.' (on screen male celebrity and
printed message)
K27 Memohaseyyo.
'Do the Memo.' (female announcer). [Note: the product name
is "Memobis"]
Table 3 shows the distribution of suggestions to buy versus all others.
Explicit references to buying (ordering, purchasing, etc.) or to selling (the
reciprocal of buying) are not favored in any of the four languages invest-
igated. In each sample, reference to other actions such as getting, using, or
benefiting from owning the product are emphasized. Korean television
commercials appear to represent the extreme in avoiding mention of
buying and selling, while Japanese commercials are apparently the least
reluctant to mention financial transactions. However, between-language
differences in the frequencies of suggestions to buy are not statistically
significant. It is also worth noting that all but one of the Chinese ads in
this category simply referred to where products are sold, which might
be taken as a simple statement of availability rather than a suggestion
to buy.
Table 3. Content of suggestions
Eng Jpns Chns Kor
Suggestions to buy 7 10 7 2
All other actions 71 25 28 51
Total 78 35 35 53
Category x Language, All: X
2
= 14.18 df = 3 p<0.05
Category x Language, English X2 = 52.05* df = 1 p<0.05
Category x Language, Japanese
X2= 5.92* cif = 1 p<0.05
Category x Language, Chinese X2 = 12.10* df = 1 p<O.OS
Category x Language, Korean X2 = 44.80* cif = 1 p<0.05
* Corrected values
Television commercials from the U. S., Japan, China, and Korea 303
3.2.3. Linguistic indirection and politeness phenomena
As was indicated in Table 2, syntactic imperatives were the most common
form of suggestions in the American television commercials but were
uncommon in Japanese, Chinese, and Korean commercials. Table 4
presents an expanded tabulation of linguistic forms found in our data,
again limited to suggestions directed at the viewer. The observed
frequencies are too small to permit statistical analysis and are reported
here simply to illustrate the range of forms found. We were also interested
in seeing whether television commercials contain other suggestion forms
such as those identified for face-to-face communication in English by
Edmondson and House (1981), as in example (2).
(2) Why not ... ?
I suggest that you ...
You should/ought to/must ...
Maybe you could .
The thing to do is .
No examples of expressions such as you should, you ought to or I suggest
that were found in any of the four language samples, though in American
commercials a number of somewhat similar forms were found:
Table 4. Linguistic form of suggestions to viewer
Eng ]pns Chns Kor
Bare imperative 33 0 7 9
Imperative + please or tag 0 3 4 1
Elliptical imperative (no verb) 4 14 0 11
Negative question 1 2 0 0
Nominalization 4 4 3 0
Embedded suggestion 4 1 0 0
Conditional 2 1 1 0
Consumer as object 3 0 6 0
Ability/possibility statements 1 2 1 0
Passive 0 0 4 0
Propositives (e. g., let's go) 0 0 0 4
Other forms 2 0 0 5
Total 54 27 26 30
304 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
E21 Why not the best?
E16 Why cook in oil, margarine, or butter?
E43 ... you can spend it enjoying one.
E25 You-'d probably like to know...
E13 Isn-'t it nice to know...
E9 It-'s a good time for the great taste at MacDonald-'s.
E5 It-'s gotta be a Dodge.
Other forms found in our data that have not been previously identified in
the literature as forms for suggestions included E17 and E30.
E17 Suddenly, the obvious choice. (nominalization)
E30 This one-'s gonna turn your head around. (consumer as object).
We suspect that a larger sample of commercials would yield more
suggestion forms. American commercials occasionally contain need-
statements, both those referring to hearer-need (If you-'re a frequent flyer,
maybe you need a new credit card [hypothetical example]) and speaker-
need (At Friendly Auto Sales, we have to get rid of 200 cars and trucks
this weekend [hypothetical example]), neither of which were found in our
sample. As indicated in Table 4, a range of suggestion forms was also
found in each of the other languages investigated, and presumably a
larger sample would also increase these inventories.
There are some major problems in attempting to compare the incidence
of such forms across languages. We think it unlikely that an etic grid
could be devised that would include all forms and provide a universally
valid ranking of forms by level of directness and politeness. However,
three minimal assumptions seem reasonable: (1) imperatives are more
direct and less polite than all other forms; (2) imperatives with tag
questions or overt politeness markers such as please or polite address
forms are more polite (though no less explicit) than bare imperatives; and
(3) suggestions in which the addressee is the linguistic subject and in
which the verb represents the action recommended are more direct than
those in which the recommended action is omitted or nominalized or sug-
Television commercials from the U. S., Japan, China, and Korea 305
gestions cast as passives. These assumptions are insufficient to rank com-
mercials from the four languages in terms of directness or politeness in a
completely precise way, but some interesting patterns do emerge for each
language.
American television commercials are, as we have already noted, most
direct according to principle (1) and are also (as seen in Table 4) the least
polite by principle (2). English advertising is alone in its extreme
preference for bare imperatives, with no mitigating devices. However, our
English commercials also contain less explicit forms such as those
mentioned in principle (3) and (as discussed previously) often make
suggestions less explicit through participant shift and action shift, so it is
not the case that the U. s. commercials are unremittingly aggressive.
There is also an apparent interaction in the u.s. commercials between
these different ways of being indirect. Although English prefers the bare
imperative for suggestions in commercials and although nearly 10 % of
all suggestions are suggestions to buy, we do not have a single case in
which either the viewer or an on screen character is bluntly told to buy
the product. Explicit mention of the exchange of money appears to
require either participant shift (actors as surrogate consumers may say
that they bought a product or some other consumer did) or linguistic
indirection (viewers may be told that a product is a better buy or told
what steps to take to buy it.
5
This tradeoff between the different dimensions of indirectness in the
U. S. commercials was not found in our samples from Asia. The Chinese
and Korean commercials in our sample contained few suggestions to buy
(regardless of the type of linguistic encoding), and the japanese sample
contained very few imperatives of any sort. However, it is interesting to
note that one of the two examples of imperatives in the japanese sample
was an imperative to buy:
J9 Minna katte ne?
'Everybody buy, won't you.' (child's voice)
However, this utterance occurred in a commercial directed at children
and a child's voice was used to make the utterance, conveying an
intimate, cute tone in japanese.
In the Chinese television commercials, the most striking aspect of the
form of suggestions is the use of please and polite pronouns. In Chinese,
the tone of an imperative may be softened by adding particles ba, Ie, or
la at the end of a sentence (none of which are found in our data), using
verbs with less directness of action (as discussed under indirection by
306 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
action shift), or by adding please to the imperative. Our Chinese
sample contains four instances of imperative plus please, of which C33 is
typical.
C33 Nin yao xiangyao nin de xiao baobao huobokeai ma? Na qing
nin gei tamen fuyong tongrentangzhiyaochang shenchan de
jianpixiaoshiwan.
'You want your children to be as healthy and lovely [as these]?
Please give them jianpi Xiaoshi Wan produced by Tongrentang
Medicine Factory.' (female announcer)
Pronouns of address also indicate politeness in Chinese. For the addres-
see, there is a choice between ni (regular, informal) and nin (polite,
honorific). In our data, nin is always used when the suggestion is
addressed to the viewer (as in C33 above, for example), although the
sample contains several commercials in which on screen actors use the ni
form to each other. The use of polite pronouns gives suggestions in
Chinese commercials a polite and formal tone.
Korean commercials do not favor imperatives with please, but
japanese commercials do. Of the three imperatives in our japanese
sample, one uses a question tag, and the other two are printed messages
with kudasai 'please', again sounding more like requests (by English
norms) than suggestions. Suggestions in japanese commercials also con-
tain negative questions (two to the viewer and two to on screen actors),
which have been identified as preferred japanese request forms
(Takahashi 1987).
For both japanese and Korean, the most interesting finding with
respect to linguistic form is the very high incidence of elliptical imper-
atives, related to the strategy of indirection through action shift. Several
examples of elliptical imperatives also occur in English, as in E16 and
E24.
E16 Pam cooking spray, because how you cook is as important as
what you cook. (male announcer)
E24 Pine S o l ~ because you care about clean. (male announcer)
Elliptical imperatives have been noted for English requests, e. g., salt,
meaning 'pass the salt' when said at a dinner table, and are generally
considered to be at the explicit, direct end of the continuum of directive
Television commercials from the U. and Korea 307
types (Ervin-Tripp 1976). Elliptical imperatives such as those in E16 and
E24 are not direct in this context, however, but are ambiguous. The
utterance Pine Sol, because you care about clean could be paraphrased as
'We created Pine soil because you care about clean' or as 'Buy Pine Sol,
because you care about clean'.
What is a relatively minor suggestion strategy in English commercials
turns out to be a major strategy in Japanese and Korean. Examples from
the Japanese sample include J14, J16, and J43.
J14 Kondo no do nichi wa ochikaku no matsuda e.
'Next Saturday and Sunday, to Mazda (stores) near (you).'
(male announcer)
J16 Odekake mae ni ichi kapuseru.
'Before going out, one capsule.' (printed message)
J43 Migaki-arai wa Kaneyon.
'For polishing, Kaneyon.' (female actor and printed message)
Korean examples include K9, K25, and K40.
K9 Ttenaki cen nal pamey Kwimitheytlul.
'At night before leaving, Kwimithey.' (male announcer)
K25 Iceypwuthen, Shiny Fresh Brown.
'From now on, Shiny Fresh Brown.' (song lyrics)
K40 Ismomi nappulttayn, Insatol.
'When gums are bad, Insatol.' (male announcer)
Native speakers of Japanese and Korean judge most of the examples of
elliptical imperatives in our data to be typical of advertising language, not
language that would be used in face-to-face interaction. What is
interesting about this is that there is a connection between this conven-
tionalized use of language in advertising and grammatical and pragmatic
principles that function more generally in these languages. Both Korean
and Japanese are discourse sensitive languages that permit the deletion of
any constituent if it is recoverable from context, although verbs are less
likely than nouns to be deleted. Takahashi (1987) has reported that
Japanese subjects produced directives in discourse completion tasks that
308 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
did not refer explicitly to the action to be taken, making the recipient of
the directive responsible for guessing what was wanted.
4.0. Conclusions
Previous studies of cross-cultural advertising practices have reported that
American advertising is essentially persuasive in nature, while Asian
advertising emphasizes other functions, informativeness in the case of
Chinese advertising and entertainment value and the establishment of
positive feelings in the case of Japanese and Korean advertising.
Advertising research of this type is typically based on the intuitive reac-
tions of native speaker judges, and results are reported without reference
to the language used in commercials or other forms of advertising. This
study has shown that prime-time consumer product advertising on televi-
sion in the U. S. is more overtly persuasive than similar advertising in three
Asian countries in terms of the frequency of suggestions and the frequency
with which imperatives are used to make such suggestions. Indeed, as
Leech has pointed out, statements such as "Asian advertising is less per-
suasive than American advertising" only make sense if they can be relati-
vized in terms of the pragmalinguistic strategies used in different
communities and situations (Leech 1983: 231). We have provided some
pragmalinguistic evidence that Japanese advertising practices are at the
opposite pole from those in the u.s. with respect to the speech act of
suggesting, while Korean and Chinese commercials are nearer the middle
of a continuum.
This study has uncovered a number of other intriguing facts about the
ways suggestions are made in television commercials in these four
countries. Perhaps the most interesting general question that can be asked
is whether the language of advertising is a function of universal pragmatic
principles, a reflection of cultural norms, the result of the requirements of
selling in a market economy (Hall - Saracino-Resh 1979), or just
a reflection of arbitrary conventions established by the advertising
industry.
While our evidence is fragmentary, we have reason to think that each
of these forces plays a role and that no one of them is entirely responsible.
Our hypotheses were based on the assumption that general cultural
norms regarding directness in language would be reflected in advertising
language. At the same time, these cultural differences are manifested
against a background of universal principles. The suggestion to buy
Television commercials from the Uo So;, Japan;, China;, and Korea 309
implies a cost to the viewer, and we find that Leech's maxim of tact, to
minimize cost and minimize benefits to a hearer (Leech 1983: 132) is
reflected in commercials from all four countries. Suggestions referring to
purchase are less common than those that stress benefits to the consumer
and avoid mention of the actual exchange of money for goods. In the u. S.
data, we also found an apparent trade-off between the use of the
imperative, the most direct linguistic form, and the choice of the action
recommended, a nice illustration of one of the corollaries of Leech's tact
maxim, that the more transparent the cost to the hearer in terms of
prepositional content the greater will be the need for optionality and
indirectness in expression of the impositive (Leech 1983: 126).
There are a number of ways in which the commercials in our samples
reflect the economies in which they are embedded, most obviously in the
distribution of products that are advertised. Other differences among the
commercials from the four countries that we have not mentioned so far
in this chapter reflect governmental regulation. These differences are less
noticeable with respect to the head act of suggestion than to the types of
supporting moves used, the reasons presented to the viewer to buy a
product. One example is the use of comparative advertising, which is
proscribed (more by culture and tradition than by law) in both Japan and
Korea, but not in China. Only the u.S. sample contained comparative
statements in which competing brands were mentioned by name (in 6 of
50 commercials). However, a wide variety of other comparative devices
was found, including strategies that can be paraphrased as better than
others (unspecified), the nothing better, and winner. Only
the Japanese commercials are non-comparative in this wider sense, while
Chinese commercials are quite comparative, frequently mentioning the
number of prizes a product has won in government sponsored quality
competitions (14 out of 50 commercials). Another aspect of commercials
subject to governmental supervision is the type of claims allowed. In the
U. S., FTC regulations prohibit the making of statements that cannot be
substantiated (Geis 1982). This may account for the fact that U. S.
commercials make very few explicit claims and are full of puffery, state-
ments that sound important but that actually make few if any claims.
Chinese commercials, by contrast, make very strong claims that would
not be allowed under U. S. advertising regulations (e. g., promising that a
particular medicinal product will cure a long list of diseases, restore
youth, and bring the user success and prosperity). By this measure, Chinese
advertising is much more direct than advertising in the U. S., Japan, or
Korea. However, the language of advertising cannot be completely a
310 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
function of the requirements of selling in a particular economy under a
particular set of government regulations. The u.s. and Japanese com-
mercials were most different with respect to the aspects of language
examined in this chapter, although the economies in which they are
situated are the most similar in many respects.
There are several ways in which the realization of suggestions to buy
in commercials reflects the potential of the medium of television. The
clearest example of this is the use of indirection by participant shift,
making suggestions directly to a viewer through an off screen voice, print-
ed message or song lyric, having on screen actors make suggestions to
each other, and so forth. The technique of having on screen characters
directly model the action of purchase, found only in the U. S.
commercials, may be simply an innovation in the use of the medium
which may spread across national and linguistic borders in time. In
commercials from each country, we also find some apparent examples of
register-specific conventionalized language. For U. S. advertising, the high
frequency of imperatives may be partly viewed as a convention of
advertising, since it cannot be attributed to universal requirements of
selling (imperatives are infrequent in commercials from the other
countries) or the function of suggesting versus requesting (imperatives are
not the most common form for suggestions in face-to-face interaction).
The finding that the formal pronoun nin is used to address the viewer in
Chinese commercials conflicts with the claim of Fang and Heng (1983)
that nin has been replaced almost completely by ni since the Cultural
Revolution, so this may be a Chinese advertising convention.
(Alternatively, the claim by Fang and Heng, for which no empirical
sources were cited, may be incorrect). Probably the best example of
conventionalized language in advertising we have found in our data is
the use of elliptical imperatives in Korean and Japanese commercials,
though as we have noted, this draws upon both the linguistic resources of
those languages (topic-comment structure, with optional deletion of
constituents) as well as the pragmatic preference in commercials for
avoiding reference to the desired action.
One of the most interesting issues for continued cross-cultural study of
the pragmatics of advertising and the language of business in general is
the need to further elucidate relationships among the language used,
universal pragmatic principles, cultural norms, and the more strictly
conventionalized aspects of advertising register. Another line of fruitful
inquiry raised by the data examined here concerns the nature of speech
acts such as suggestions and requests. We do not have a fully satisfactory
Television commercials from the U. S., ]apan,China, and Korea 311
explanation for the occurrence of request-like forms in commercials from
China, Japan, and (to a lesser extent) Korea, especially those with please
and formal politeness markers that are not found in U. S. commercials at
all. One possible line of explanation could be that requests and sugges-
tions may be less differentiated speech act categories in Japanese and
Korean than in English. An analysis along these lines is suggested by
Coulmas' (1981) observation that apologies and expressions of gratitude
are less differentiated as speech act categories in Japanese than in many
other languages. Banerjee and Carrell (1988) reported that their non-
native speakers of English (of Chinese and Malay language backgrounds)
sometimes used requesting strategies inappropriately for suggestions in
English. However, even in English there is great overlap among the ling-
uistic forms used for requesting and suggesting, and these categories of
speech acts are not completely distinct at a conceptual level. There are
clear cases of requests directed at actions desired by a speaker with no
benefits for a hearer, as well as suggestions in cases where only the
hearer's benefit is at issue and the speaker making the suggestion has no
interest at all in whether or not the hearer carries out the action. However,
the business concept of a free transaction between a willing seller and a
willing buyer implies a balance between the costs and benefits accruing to
both parties. Perhaps our initial analysis, that commercials should be
understood as suggestions, was incorrect. If commercials are more
properly analyzed as hybrids containing elements of both request and
suggestion, then the U. S. preference for the linguistic forms typical of
suggestions and avoidance of those typical of requests might be seen not
as a reflection of the fact that commercials are suggestions in essence but
as another manifestation of the maxim of tact, a manipulative strategy
designed to minimize apparent costs by using linguistic forms that are
appropriate when recommending actions that are clearly to a hearer's
benefit.
Notes
1. The videotapes of Chinese, Korean and Japanese commercials used in this
research were provided by Charles Keown and Lawrence Jacobs. Initial trans-
cription of the English commercials was done by Johanna Guth. An informa-
tion content analysis of a larger sample of commercials drawn from the same
database appears in Keown, Jacobs, Schmidt, and Gymn (1992). Useful
comments on an earlier draft of this paper were provided by Carl James,
Gabriele Kasper, Rajendra Singh, Peter Schmidt, and Keiichi Morita, none of
whom are responsible for any weaknesses.
312 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
2. We do not discuss visual strategies in this chapter. While television is often
considered to be primarily a visual medium in which language plays only a
secondary role, Geis (1982) has provided ample evidence that television is no
less an auditory medium than is radio.
3. An alternative to using consensus among the authors for coding decisions
would have been to have more than one rater categorize examples from each
language and compute inter-rater reliability coefficients. While this would
enhance confidence in the analysis of each individual language sample, this
method would not guarantee comparability of coding across languages, which
we believe was achieved better using discussion and consensus.
4. It is common for a Japanese women to address her father-in-law as ojiichan
'grandpa', especially in front of her children. He is not her grandfather, but he
is grandfather to her children and grandfather is seen as his primary role with
respect to the family unit.
5. An exception to the generalization that imperatives are not used in U. S.
advertising with explicit reference to the exchange of money occurs in
advertisements for mail-order products (none in the sample analyzed in this
chapter), at the end of which a viewer might be told: Call 1-800-xxx-xxxx,
Have your credit card ready, Or send $ 9.95 plus $ 3 shipping and handling
to... [hypothetical example]. A possible explanation for these forms
may be that unless the viewer takes note of the address or phone number
when the commercial is being broadcast (or, as a minimum, is primed to do this
when next hearing the same commercial), the commercial will not fulfill its
goaL
References
Agoston, Thomas - Walburga von Raffler-Engel
1979 "A linguistic analysis of some commercial television advertisements",
in: Robert St. Clair (ed.), 224-242.
Alatis, James E. - G. Richard Tucker
1979 Georgetown University round table on languages and linguistics
1979. Washington, DC: Georgetown University Press.
Aston, Guy
1993 "Notes on the interlanguage of comity", in: Gabriele Kasper -
Shoshana Blum-Kulka (eds.), 224-250.
Banerjee, Janet - Patricia L. Carrell
1988 "Tuck in your shirt, you squid: Suggestions In ESL", Language
Learning 38: 313-364.
Beebe, Leslie M. - Tomoko Takahashi
1989 Do you have a bag?: Social status and patterned variation in second
language acquisition. In: Susan Gass - Carolyn G. Madden -
Dennis R. Preston - Larry Selinker (eds.), 103-125.
Barthes, Roland
1972 Mythologies. New York: Hill and Wang.
Television commercials from the U. S., Japan, China, and Korea 313
Bhatia, Tej.K.
1987 "English in advertising: Multiple mixing and media", World Englishes
6: 33-48.
Blum-Kulka, Shoshana - Juliane House - Gabriele Kasper (eds.)
1989 Cross-cultural pragmatics: Requests and apologies. Norwood, NJ:
ABLEX.
Blum-Kulka, Shoshana - Elite Olshtain
1984 "Requests and apologies: A cross-cultural study of speech act
realization patterns (CCSARP)", Applied Linguistics 5: 196-213.
Bolinger, Dwight
1973 "Truth is a linguistic function", Language 49: 539-550.
1980 Language: The loaded weapon. London: Longman.
Brown, Penelope - Stephen C. Levinson
1987 Politeness. Cambridge: Cambridge University Press.
Cole, Peter - Jerry L. Morgan (eds.)
1977 Speech acts (Syntax and semantics 3). New York: Academic Press.
Coleman, Linda
1990 "The language of advertising" (review article), Journal of Pragmatics
14: 137-145
Coulmas, Florian
1981 "Poison to your soul. Thanks and apologies contrastively viewed",
in: Florian Coulmas (ed.), 69-91.
Coulmas, Florian (ed)
1981 Conversational routine. The Hague: Mouton.
Dowling, Graham R.
1980 "Information content in U. S. and Australian television advertising",
Journal of Marketing 44 (Fall): 34-37.
Edmondson, Willis - Juliane House
1981 Lefs talk and talk about it. Munich: Urban & Schwarzenberg.
Fang, Hanquan - J.H. Heng
1983 "Social changes and changing address norms in China", Language in
Society 12: 495-507.
Feasely, F. (ed.)
1987 Proceedings of the 1987 conference of the American Academy of
Advertising.
Fields, George
1983 From bonsai to Levi:Js. New York: New Yotk Library.
Fraser, Bruce
1983 "The domain of pragmatics", in: Jack C. Richards - Richard W.
Schmidt (eds.), 29-59.
Gass, Susan M. - Carolyn G. Madden - Dennis R. Preston - Larry Selinker (eds.)
1989 Variation in second language acquisition: Discourse and pragmatics.
Clevedon: Multilingual Matters.
Geis, Michael
1982 The language of television advertising. New York: Academic Press.
Goffmann, Erving
1981 Forms of talk. Philadelphia: University of Pennsylvania Press.
314 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
Green, Georgia M.
1975 "How to get people to do things with words: The whimperative
question", in: Peter Cole - Jerry L. Morgan (eds.), 107-141.
Grice, H. Paul
1975 "Logic and conversation", in: Peter Cole - Jerry L. Morgan (eds.)
41-58.
Gumperz, John J.
1982 Discourse strategies. Cambridge: Cambridge University Press.
Hall, Dennis R. - Linda Saracino-Resh
1979 "Advertising as cultural language",in: Robert St. Clair (ed.) 191-211.
Harris, Richard J. (ed.)
1983 Information processing research. Hillsdale, NJ: Erlbaum.
Hatch, Evelyn - Anne Lazaraton
1991 The research manual: Design and statistics for applied linguistics.
New York: Newbury House.
Ho, Suk-ching - Yat-ming Sin
1986 "Advertising in China: Looking back and looking forward",
International Journal of Advertising 5: 307-316.
Hong, Jae W. - Aydin Muderrisoglu - George M. Zinhan
1987 "Cultural differences and advertising expression: A comparative
content analysis of Japanese and u.S. magazine advertising", Journal
of Advertising 16: 55-68.
Hunt, Shelby D.
1976 "Informational vs. persuasive advertising: An appraisal", Journal of
Advertising 5 (Summer): 5-8.
Johns, Ann M.
1986 "The language of business", Annual Review of Applied Linguistics 7:
3-17.
Kaplan, Robert B.
1966 "Cultural thought patterns in inter-cultural education", Language
Learning 16: 1-20.
Kasper, Gabriele - Shoshana Blum-Kulka (eds.)
1993 Interlanguage pragmatics. New York: Oxford University Press.
Kasper, Gabriele - Merete Dahl
1991 "Research methods in interlanguage pragmatics", Studies in Second
Language Acquisition 13: 215-247.
Kaynak, Erdener - Lionel A. Mitchel
1981 "Analysis of marketing strategies in diverse cultures", Journal of
Advertising Research, 213: 25-32.
Keown, Charles F. - Lawrence W. Jacobs - Richard W. Schmidt -
Kyung-Il Gymn
1992 "Information content of advertising in the United States, Japan,
South Korea, and the People's Republic of China", International
Journal of Advertising 11: 257-267.
Laczniak, Gene R.
1979 "Information content in print advertising", Journalism Quarterly 56:
324-327, 345.
Television commercials from the U.S., Japan, China, and Korea 315
Lakoff, Robin Tolmach
1982 "Persuasive discourse and ordinary conversation, with examples
from advertising", in: Deborah Tannen (ed.) 25-42.
Leech, Geoffrey
1966 English in advertising. London: Longman.
1983 Principles of pragmatics. London: Longman.
Madden, Charles S. - Marjorie Caballero - Sinya Matsukubo
1986 "Analysis of information content in U. S. and Japanese magazine
advertising", Journal of Advertising 15.3: 38-45.
Masavisut, Nitaya - Mayuri Sukwiwat - Seri Wongmontha
1987 "The power of the English language in Thai media", World Englishes
5: 197-207.
Matsumoto, Yoshiko
1988 "Reexamination of the universality of face: Politeness phenomena in
Japanese", Journal of Pragmatics 12: 403-426.
Miracle, Gordon E.
1985 "Advertising regulations in Japan and the USA: An introductory
comparison", Waseda Business and Economic Studies 21: 335-69.
1987 "Feel-do-Iearn: An alternative sequence underlying Japanese
commercials", in: F. Feasely (ed.) R73-R78.
Moeran, Brian
1985 "When the poetics of advertising become the advertising of poetics.
Syntactical and semantic parallelism in English and Japanese adver-
tising", Language and Communication 5: 29-44.
Mueller, Barbara
1987 "Reflections of culture: An analysis of Japanese and American
advertising appeals", Journal of Advertising Research (June/July):
51-59.
O'Barr, William M.
1979 "Language and advertising", in: James E. Alatis - G. Richard
Tucker (eds.) 272-286.
Resnik, Alan - Bruce L. Stern
1977 "An analysis of information content in television advertising",
Journal of Marketing 41: 50-53.
Rice, Marshall D. - Zaiming Lu
1988 "A content analysis of Chinese magazine advertisements", Journal of
Advertising 17.4: 43-48.
Richards, Jack C. - Richard W. Schmidt (eds)
1983 Language and communication. London: Longman.
Rintell, Ellen
1979 "Getting your speech act together: The pragmatic ability of second
language learners", Working Papers on Bilingualism 17: 97-106.
Schmidt, Rosemarie - Joseph F. Kess
1985 "Persuasive language in the television medium: Contrasting adver-
tising and televangelism", Journal of Pragmatics 9: 287-308.
1987 Television advertising and televangelism: Discourse analysis of per-
suasive language. Amsterdam: Benjamins.
316 R. Schmidt - A. Shimura - Z. Wang - H. Jeong
Searle, John R.
1969 Speech acts. Cambridge: Cambridge University Press.
1976 "A classification of illocutionary acts", Language in Society 5: 1-23.
Sherry, John F. - Eduardo G. Camargo
1987 "May your life be marvelous: English language labeling and the
semiotics of Japanese promotion", Journal of Consumer Research
14: 174-188.
St. Clair, Robert (ed)
1979 Perspectives on applied sociolinguistics. Lawrence, Kansas:
Coronado Press.
Stern, Bruce L. - Dean M. Krugman - Alan Resnik
1981 "Magazine advertising: Analysis of its information content", Journal
of Advertising Research 21.2: 39-44.
Takahashi, Satomi
1987 A contrastive study of indirectness exemplified in Ll directive speech
acts performed by American and Japanese. [Unpublished master's
thesis, The University of Illinois at Urbana-Champaign, Urbana IL]
Tannen, Deborah (ed)
1982 Analyzing discourse: Text and t a l k ~ Georgetown University round
table on languages and linguistics. Washington DC: Georgetown
University Press.
Tse, David K. - Russel W. Belk - Nan Zhou
1989 "Becoming a consumer society: A longitudinal and cross-cultural
content analysis of print ads from Hong Kong, the People's Republic
of China, and Taiwan", Journal of Consumer Research 15: 457-472.
Vestergaard, Torben - Kim Schreder
1985 The language of advertising. Oxford: Basil Blackwell.
Weizman, Elda
1993 "Interlanguage requestive hints", in: Gabriele Kasper - Shoshana
Blum-Kulka (eds.) 123-137.
Yamaki, Toshio
n. d. International comparison of television commercials [Unpublished
M.S.]
Culture, negotiations and international cooperative
ventures
John L. Graham
1. Introduction
"You've all heard the story about the invention of copper wire - two
Dutchmen got a hold of a penny." This anecdote was served up during a
dinner speech by the American president of a joint venture owned by an
American multinational company and a comparable Dutch firm. At one
level the story is a friendly gibe, although the professor from a Dutch
university sitting nearby did not appreciate the American's remarks in
general or the ethnic joke in particular. Indeed, at another level the story
is stereotyping of the worst sort.
However, at an even deeper level there is an important lesson here for
all managers and students of joint ventures and international cooperative
arrangements in general. Culture can get in the way. The American
president was in his "humorous" way attributing part of the friction
between him and his Dutch associates to differences in cultural values. He
might have blamed personality differences or clashing "corporate"
cultures, but instead he identified national cultural barriers to be a major
difficulty in joint venture management. And although I (also) do not
appreciate his humor, I certainly agree that cultural differences between
joint venture partners and managers can cause divisive, even decisive
problems.
Harrigan (1987) suggests that a crucial aspect of joint ventures is the
negotiation of the original agreement. The seeds of success or failure are
often sown at the negotiation table where not only financial and legal
details are agreed to, but perhaps more importantly, the ambiance of
cooperation is established. Indeed, as Harrigan indicates, the legal details
and the structure of joint ventures are almost always modified over time,
and usually through negotiations. But the atmosphere of cooperation
established at the negotiation table persists or the venture fails.
The purpose of this chapter is to present selected results from a
program of research investigating differences in cultural styles of business
318 John L. Graham
negotiations. The study has involved more than 1000 business people in
seventeen countries and cultures. The analyses reported below comprise
some of the most interesting findings of the project. Other results are
reported in a series of complementary articles (cf., Graham 1980; 1983;
1985 a; 1985 b; 1992). Here six business people from each of fourteen
countries were videotaped during simulated intracultural negotiations.
The content of their negotiation strategies and linguistic structural
aspects of their conversations were analyzed. Our findings suggest that
substantial differences exist in negotiation styles across the thirteen
countries. Further, it is our supposition that such differences can cause
friction, suspicion, and even failure in otherwise mutually beneficial inter-
national joint ventures.
The remainder of this chapter is organized into three sections. First,
the theoretical perspective is briefly discussed. Next, methods of data
collection are described. Finally, the results are summarized, conclusions
are drawn, and hypotheses for future testing are suggested.
2. Theoretical perspective
Despite the increasing importance of cross-national commercial rela-
tionships, business negotiations in different countries have received little
attention. During the 1970s, a few articles appeared in business journals
(for example, Jastram 1974; Kapoor 1974; Van Zandt 1970; Wells
1972), but they were primarily descriptive and often anecdotal. Recently,
more systematic studies of negotiations in foreign countries have been
undertaken. Tung (1982) considered business negotiations between
American and Chinese executives. Harnett and Cummings (1980) com-
pared bargainers' characteristics and behaviors across several cultural
groups. Graham, Mintu and Rodgers (1994), investigated the deter-
minants of business negotiation outcomes in the United States, and ten
foreign countries. Weiss has provided in-depth reports of case studies of
major international business negotiations (1987; 1990). Francis (1991)
has considered the importance of adaptation in international business
negotiations. These studies have proven valuable, but most are limited in
their use of questionnaire items as measures of negotiation processes.
Most recently, the outcomes of business negotiations have been
hypothesized to be the result of several factors that can be classified into
three categories or kinds of theoretical constructs - individual
Culture, and international cooperative ventures 319
characteristics, situational constraints, and process measures (see Rubin
Brown 1975; Sawyer - Guetzkow 1965). Many empirical measures
of both individual characteristics and situational constraints have been
tested in previous research. Examples of such individual characteristics
might be intelligence, self-esteem, credibility, attractiveness, and cultural
background. Examples of situational constraints might include company
goals, time limitations, or unequal power relations.
2.1. Process measures
A few studies have fo'cused on the process of business negotiations (for
example, Dwyer - Walker 1981; Lewis - Fry 1977; Pennington 1968;
Pruitt - Lewis 1975). Graham (19.83: 82) has defined process measures
as "qualitative and quantitative descriptions of the activities involved in
a business negotiation for example, bargaining strategies." Based on an
extensive review of the negotiation literature, Rubin and Brown (1975)
conclude that the behaviors of bargainers during the negotiation process
affect negotiation outcomes. The kinds of behaviors they list are opening
moves, countermoves, types of appeals, demands, and the like. But little
work has been done to investigate relationships among process measures
and negotiation outcomes, individual characteristics and situational con-
straints.
Moreover, because researchers have only recently turned their atten-
tion to process measures, concepts and operational definitions remain
vague and 'relations not adequately specified. The present study focuses
on the development of operational definitions of process measures using
observational methods. Further, special attention is given to the influence
of national culture on these process measures.
2.2. differences in business negotiation processes
Culture has been a difficult concept to deal with in any consistent way.
Anthropologists and sociologists have been arguing over definitions for
years. Culture has appeared in the marketing literature primarily as a
determinant of consumer behavior (for example, Engel - Blackwell
but operational definitions seem to have varied from study to
study. Perhaps the most widely accepted definition is that professed by
Linton: "A culture is a configuration of learned behaviors and results of
320 John L. Graham
behavior whose component parts are shared and transmitted by
members of a particular society" (Linton 1945: 5). The important part
of the definition for the present research is the idea that behaviors are
shared by members of a particular culture. Or as Spiro put it, "members
of a given society behave in uniform and predictable ways" (Spiro 1950:
20). A central goal of the study is to discover what shared behaviors
manifest themselves during business negotiations in the thirteen
countries.
In addition to the bargaining behaviors being consistent within
cultures, several authors have suggested that negotiation processes differ
across cultures (for example, Condon 1974; Frake 1972; Kay 1970;
Sawyer - Guetzkow 1965; Van Zandt 1970). Therefore, a second
purpose of this work is to identify how bargaining processes in several
countries might differ from one another.
2.3. Content versus context
Aside from the cultural differences in negotiation behaviors discovered,
perhaps the most important implication of the study regards the "content
versus context" issue. Social psychologists have focused on the verbal
content of negotiation in their research. Alternatively, linguistic theory
holds that consideration of only verbal content yields inadequate under-
standings of interpersonal interactions. Sociolinguists emphasize the
importance of the context of communication of nonverbal and structural
aspects of language. Our results suggest that the linguists are correct.
Simply stated, the content of conversation is what is said, while the
context is how it is said. The distinction is both theoretically and
practically a "fuzzy" one. Several researchers have developed schemes
for categorizing the what aspects of negotiations (e.g., Angelmar - Stern
1978; Bales 1950; Bonoma - Felder 1977; Pennington 1968; Pruitt
Lewis 1975; Walton - McKersie 1965), and used these schemes to
analyze the verbal content of bargaining interactions. More recently, the
how of meaning has also been considered. Take, for example, the inter-
action described and interpreted in The Wall Street Journal:
The japanese executive sucks in air through his teeth and exclaims, "Sa!
That will be very difficult!" What he really means is just plain "no." But
the japanese consider an absolute "no" to be offensive and usually seek a
euphemistic term. That's why in japan, the "difficult" really may be impos-
sible. The American on the other side of the negotiation table knows none
Culture, negotiations and international cooperative ventures 321
of this and presses ahead to resolve the "difficulty." The Japanese finds this
inexplicable persistence to be abnormally pushy. The atmosphere deterio-
rates, and sure enough, the big deal falls through (Ricklefs 1978: 4).
How do we understand the meaning of the word "difficult?" Does it
mean "no" or "maybe?" Ethnomethodologists emphasize the importance
of context as well as content for establishing a shared meaning of
communication. The idea is that communication must be considered as
an integrated whole, content and context; and context has often been
"taken for granted" in previous negotiation studies.
Certainly the reality of any particular situation provides much of the
context for making decisions about meaning. And so does all previous
communication between actors. Gumperz (1979) has posited that
humans, in the course of interaction, also indicate context for interpret-
ation of verbal communications through the use of contextualization
cues. He explains:
Our hypothesis is that conversational inference, i. e., the process by which
speakers interpret what is intended by a conversational contribution, is in
part determined by a system of conventional discourse-level verbal and
non-verbal signals. These signals, termed "contextualization cues," serve to
signal the way in which any conversational contribution is to be under-
stood, in light of the participants' expectations and the situation at hand
(Gumperz 1979: 2).
An example of a contextualization cue might be a rise in tone of voice to
indicate or underline an important point. Gumperz and his associates
have also found that contextualization cues vary across cultures. They are
behaviors learned in the course of the individuals' socialization. Further,
he suggests that these differences are often the cause of misunder-
standings that can have serious consequences (e.g., failed negotiations) in
cross-cultural interactions.
Authors in other fields also emphasize the importance of context. For
example, Bonoma and Felder (1977) and Soldow and Thomas (1984)
offer alternative definitions of context, non-verbal behaviors and rela-
tional communication, respectively. Hall (1976), Cateora (1983), and
Graham (1987) argue that the influence of context varies across cultures.
Cateora states, "communication in a high-context culture depends
heavily on the context or non-verbal aspects of communication, whereas
the low-context culture depends more on explicit, verbally expressed
communication" (1983: 133). In the present study both content and
context are considered.
322 John L. Graham
3. Methods
3.1. Sample
The participants in the experiment are six business people from each of
the fourteen cultures (japanese, Korean, Taiwanese, Chinese, Russian,
German, English, French, Spanish, Brazilian, Mexican, French Canadian,
English Canadian, and American (D. S. )). All were participating in
executive or Master's of Business Administration (MBA) programs and
all volunteered for this study. All received college educations and all have
had at least two years' business experience in their respective countries.
The sample was limited to experienced business people because Fouraker
and Siegel (1963) reported differences in the bargaining behavior of
students and businesspeople.
3.2. Laboratory setting
The negotiation simulation, developed by Kelley (1966) and used by
Pruitt and Lewis (1975) and Lewis and Fry (1977) involves bargaining
over three related issues. Differing amounts and types of background
information can be included with the basic pay-off matrices, depending
on the focus of the research. The simulation is simple enough to be
learned quickly, but complex enough to provide usually one half hour of
face-to-face interaction. Forty-two negotiations were conducted - three
for each group..
3.3. Data collection
The forty-two interactions were videotaped using a wide-angle perspec-
tive to capture postures, body movements, and interpersonal distances.
Participants were asked to evaluate the obtrusiveness .of the setting on
questionnaires followirig the .negotiation game and they reported a mini-
mum of discomfort.
3.4. Verbal behaviors
As mentioned, a primary purpose in this exploratory work is the identif-
ication and clarification of process measures. Consequently, the discussion
Culture, negotiations and international cooperative ventures 323
in the sections to follow are organized as a "list" of process measures.
Associated with each item on the "list" are operational definitions, a brief
account of the method of measurement, and mention of apparent
differences among the thirteen cultural groups.
The first step in the measurement and analysis of verbal behaviors
during the business negotiations is the transcription of the audio portion
of the videotapes. This is a potential source of error in measurement. A
complete check of the transcript revealed some minor mistakes, and these
were corrected.
The second step in the measurement and analysis process consisted of
translation of the non-English interactions. With one exception native
speakers of the foreign languages were instructed to "make the transla-
tions as literal as possible while still communicating the intended
meaning." The quality of the English grammar, etc. was not the primary
consideration. The exception to the native speaker rule regarded the
Russian translation, wherein a 20 year American resident of Moscow did
the work.
3.5. Content analysis
Angelmar and Stern (1978) have described a content analysis scheme
developed specifically for the analysis of bargaining communications in
business settings. Utterances by participants are classified into twelve
categories. The categories and definitions are listed in Table 1. Angelmar
and Stern report positive results from a reliability and validity assessment
of the system applied to written communications. The present study is
one of the few to apply the scheme to transcripts of conversations.
Coding transcribed conversations is a more difficult undertaking; spoken
words are the only channel of communication. Transcripts do not include
. information communicated through other channels such a proxemics,
prosody, kinetics, or facial expression. Theory indicates that these
channels also may be important for accurate interpretation and measure-
ment of conversational contributions.
Two coders were employed in classifying segments of the conversation
into twelve bargaining categories. The author coded all forty-two inter-
actions and research assistants (ignorant of the theory and hypotheses
involved in the study) coded three interactions to provide a reliability
check. The author is cognizant of the possible biases involved in using
coders informed about the theory applied in the research. However,
324 John L. Graham
Table 1. Verbal negotiation tactics (The "what" of communications)
Bargaining Behaviors and Definitions Cultures
(Anglemar and Stern, 1978) (in each group, n = 6)
JPN KOR TWN
Promise. A statement in which the source indicated his 7* 4 9
intention to provide the target with a reinforcing
consequence which source anticipates target will
evaluate as pleasant, positive, or rewarding
Threat. Same as promise, except that the reinforcing 4 2 2
consequences are thought to be noxious, unpleasant,
or punishing.
Recommendation. A statement in which the source 7 1 5
predicts that a pleasant environmental consequence
will occur to the target. Its occurrence is not under
source's control
Warning. Same as recommendation, except that the 2 0 3
consequences are thought to be unpleasant.
Reward. A statement by the source that is thought to 1 3 2
create pleasant consequences for the target.
Punishment. Same as reward, except that the 1 5 1
consequences are thought to be unpleasant.
Positive normative appeal. A statement in which the 1 1 0
source indicates that the target's past, present, or future
behavior was or will be in conformity with social norms.
Negative normative appeal. Same as positive normative 3 2 1
appeal except that the target's behavior is in violation of
social norms.
Commitment. A statement by the source to the effect 15 13 9
that its future bids will not go below or above a
certain level.
Self-disclosure. A statement in which the source reveals 34 36 42
information about itself.
Question. A statement in which the source asks the
target to reveal information about itself. 20 21 14
Command. A statement in which the source suggests that
the target perform a certain behavior. 8 13 11
)!- Read "7% of the statements made by Japanese negotiators were promises."
a northern China (Tianjin and environs)
C u l t u r e ~ negotiations and international cooperative ventures 325
Table 1 (contd.)
CHNa RUSS GRM UK FRN SPN BRZ MEX FCAN ECAN USA
6
1
2
5
3
4
7
3
5
11 5
3 5
6 3
11
2
4
3
2
5
7
1
8
8
3
5
6
o
4
8
4
4
1 0 1 1 3 1 1 2 5 0 1
1 3 4 5 3 3 2 1 1 3 2
0 1 2 0 3 2 3 0 2 1 3
1 0 0 0 0 0 0 0 1 0 1
0 0 1 1 0 1 1 1 2 1 1
10 11 9 13 10 9 8 9 8 14 13
36 40 47 39 42 34 39 38 42 34 36
34 27 11 15 18 17 22 27 19 26 20
7 7 12 9 9 17 14 7 5 10 6
326 John L. Graham
resource constraints necessitate this less-than-ideal state. Significantly,
analysis of discrepancies in coding between the two coders revealed this
source of bias to be minimal. Intercoder reliability was 63 %, comparable
to Angelmar and Stern (1978) wherein they report 66 % agreement for
coding written negotiations.
The data presented in Table 1 represent the percentage of each
category of behavior used averaged across the six negotiators in
each country. For example, the number of promises used by a single
American negotiator was divided by the total behavior coded for
that negotiator, then the average percentage of promises across
the six American negotiators was calculated and reported in the upper
left corner of Table 1. Such a procedure allows for comparisons
across the cultural groups controlling for differences in time spent
negotiating.
3.6. Structural Aspects ("no" and "you")
Graham (1985b) suggests that the simple counting of these two words
may shed light on subtle differences in cultural styles of persuasion. He
found substantial differences between the frequency of the use of the
word "no" by Brazilian bargainers as opposed to American and
Japanese. Several authors (e.g., Nakane 1970; Veda 1974; Van Zandt
1970) indicate that Japanese negotiators seldom use the word "no"
during negotiations. Graham (1985 b) also notes a Brazilian propensity to
speak more frequently in the second person using the pronoun "you."
The number of times each word was used was tallied for each negotiator
then multiplied by the time of the negotiation in minutes and then divided
by thirty minutes to provide a frequency measure which might be com-
pared across the various groups. Intercoder reliability (calculated
using Guetzkow's 1960 formula for marginal reliability, the difference
in the number of units between coders as a percentage of the sum
of the units) was calculated for three of the interactions and found
to be 1 %.'
3.7. Nonverbal behaviors
In this section of explorations into the process of buyer-seller negotia-
tions' nonverbal aspects of the videotaped interactions were considered.
Culture, negotiations and international cooperative ventures 327
First the rhythm of the conversations will be discussed, specifically
examining silent periods and conversational overlaps. Next, gaze direc-
tion of the participants will be considered. Lastly, findings related to
touching during negotiations are presented. All measurements in this
section have been derived irrespective of the verbal content of the inter-
actions to avoid potential bias, that is, the tapes could be coded without
knowledge of the language spoken. Reliability of the coding was
calculated for all the nonverbal behaviors by having a second
assistant code three interactions using a marginal reliability approach
(difference in the tallies of the two coders divided by the sum of the
two coders). These numbers are reported at the end of each section to
follow.
3.8. Conversational coordination
Communication theory suggests that when two people are effectively
sharing ideas, their communication behaviors - both verbal and non-
verbal-will be rhythmically coordinated (Condon 1968; Erickson 1976;
Gumperz 1979). Here two measures of conversational coordination,
"silent periods" and "conversational overlaps," are operationally
defined, and findings are reported below.
(1) Silent Periods. Silent periods are defined as gaps in conversations
ten seconds or more in duration. The time period of ten seconds
was selected somewhat arbitrarily, but it is a long enough period of
silence to appear unnatural to most American observers. The tapes
were searched for gaps in conversations of 10 seconds or more, and
these gaps were noted on the transcripts and tallied (see Table 2).
Once again the frequency of occurrence was calculated by multiplying
the number of silent periods by the duration at each negotiation divided
by thirty minutes (marginal reliability < 1 0/0).
(2) Conversational Overlaps. The concept of "interactional sychrony,"
the unconscious coordination of verbal and nonverbal behaviors of two
or more participants in a conversation, is discussed at length by Graham
(1980). One possible measure of this construct is the number of
conversational overlaps or interruptions during a conversation.
Conversational overlaps are defined here as periods when both speakers
are talking simultaneously, or when the conversational contribution of
one speaker overlaps that of the other speaker. Identification of such
overlaps is independent of the verbal content of the interactions. In the
328 John L. Graham
Table 2. Linguistic aspects of language and nonverbal behaviors ("How" things
are said)
Bargaining Behaviors (per 30 minutes) Cultures
(in each group, n =6)
JPN KOR TWN
Structural Aspects
"No's." The number of times the word "no" was
used by each negotiator.
"You's." The number of times the word "you"
was used by each negotiator.
Nonverbal Behaviors
Silent Periods. The number of conversational gaps
of 10 seconds or longer.
Conversational Overlaps. Number of interruptions.
Facial Gazing. Number of minutes negotiators
spent looking at opponent's face.
Touching. Incidents of bargainers touching one another
(not including handshaking).
a northern China (Tianjin and environs)
1.9
31.5
2.5
6.2
3.9
o
7.4
34.2
o
22.0
9.9
o
5.9
36.6
o
12.3
19.7
o
present work, the videotapes were searched for overlaps, and such inter-
ruptions in the flow of conversation were noted on the transcripts.
Frequencies were calculated as above and reported in Table 2 (marginal
reliability =10%).
3.9. Facial gazing
The third nonverbal variable to be considered is facial gazing. Other
researchers have found significant relationships between facial gazing and
outcomes of negotiations (Lewis - Fry 1977). Moreover, several authors
have suggested differences in facial gazing behavior across cultures
(Argyle - Cook 1976).
In this study, facial gazing is defined as the percentage of time
a bargainer gazes at the face of his opponent. Ten-minute video-
tape excerpts of each of the forty-two interactions served as data
Culture, negotiations and international cooperative ventures 329
Table 2 (contd.)
CHNa RUSS GRM UK FRN SPN BRZ MEX FCAN ECAN USA
1.5 2.3 6.7 5.4 11.3 23.2 41.9 4.5 7.0 10.1 4.5
26.8 23.6 39.7 54.8 70.2 73.3 90.4 56.3 72.4 64.4 54.1
2.3
17.1
3.7 0
13.3 20.8
2.5 1.0 0 0 1.1
5.3 20.7 28.0 14.3 10.6
0.2
24.0
2.9
17.0
1.7
5.1
11.1 8.7 10.2 9.0 16.0 13.7 15.6 14.7 18.8 10.4 10.0
o o o o 0.1 0 4.7 o o o o
here. Using a stopwatch, two observers recorded the time each
participant spent gazing at his opponent's face. The method used
was very similar to that reported by Lewis and Fry (1977), except that
here videotapes were reviewed rather than real-time interactions. Using
videotapes is a more reliable technique, allowing reviews and reliability
checks.
3.10. Touching
Finally, the number of times a negotiator touched a partner (excluding
beginning and ending handshaking) was recorded for each interaction
(marginal reliability = 0 %).
330 John L. Graham
4. Discussion
4.1. Results
The results from the analyses are presented in Tables 1 and 2. The unex-
pected similarities among the fourteen groups are perhaps more striking
than the differences. Particularly with regard to Anglemar and Stern's
(1978) content analysis scheme, negotiation styles appear to be surpris-
ingly consistent across the fourteen cultural groups. Negotiations in
all cultures studied are composed primarily of information exchange
tactics - questions and self-disclosures.
4.2. Nonverbal behaviors
Reported in Table 2 are the analyses of some linguistic aspects and non-
verbal behaviors for the fourteen videotaped groups, as in Graham
(1985 b). While our efforts here merely scratch the surface of these kinds
of behavioral analyses, they still provide indications of substantial
cultural differences. Note that the japanese are at, or next to, the end of
almost every dimension of the behaviors listed in Table 2. Their facial
gazing and touching are the least among the fourteen groups. Only the
northern Chinese used the words "no" less frequently and only the
Russians used more silent periods than did the japanese.
A broader examination of the data in Tables 1 and 2 reveals a more
meaningful conclusion. That is, the variation across cultures is greater
when comparing linguistic aspects of language and nonverbal behaviors
than when the verbal content of negotiations is considered. For example,
notice the great differences between japanese and Brazilians in Table 1
vis-a-vis Table 2.
Following are further descriptions of the distinctive aspects of each of
the fourteen cultural groups we have videotaped. Certainly, we cannot
draw conclusions about the individual cultures from an analysis of only
six business people in each, but the suggested cultural differences are
worthwhile to consider briefly.
Japan. Consistent with most descriptions of japanese negotiation
behavior in the literature, the results of this analysis suggest their style of
interaction to be the least aggressive (or most polite). Threats, com-
mands, and warnings appear to be deemphasized in favor of the more
negotiations and international cooperative ventures 331
posItIve promises, recommendations, and commitments. Particularly
indicative of their polite conversational style was their infrequent use of
"no" and "you" and facial gazing, as well as more frequent silent
periods.
Korea. Perhaps one of the more interesting aspects of this study is the
contrast of the Asian styles of negotiations. Non-Asians often generalize
about the Orient. Our findings qemonstrate that this is a mistake. Korean
negotiators used considerably more punishments and commands than did
the japanese. Koreans used the word "no" and interrupted more than
three times as frequently as the japanese. Moreover, no silent periods
occurred between Korean negotiators.
China (northern). The behaviors of the negotiators from northern
China (i. e., in and around Tianjin) are most remarkable in the emphasis
on asking questions (34 0/0). Indeed, 70 % of the statements made by the
Chinese negotiators were classified as information exchange tactics.
Other aspects of their behavior were quite similar to the japanese - the
use of "no" and "you" and silent periods.
Taiwan. The behavior of the business people in Taiwan was quite
different from that in China and Japan but similar to that in Korea. The
Chinese on Taiwan were exceptional in the time of facial gazing, on the
average almost 20 'out of 30 minutes. They asked fewer questions and
provided more information (self-disclosures) than did any of the other
Asian groups.
Russia. The Russians' style was quite different from that of any other
European group, and, indeed, was quite similar in many respects to the
style of the Japanese. They used "no" and "you" infrequently and used
the most silent periods of any group. Only the japanese did less facial
gazing, and only the Chinese asked a greater percentage of questions.
Germany. The behaviors of the western Germans are difficult to
characterize because they fell toward the center of almost all of the
categories. However, the Germans were exceptional in the high percent-
age of self-disclosures at 47% and the low percentage of questions at
11 0/0.
United Kingdom. The behaviors of the British negotiators are remark-
ably similar to those of the Americans in all respects.
I Spain. Diga is perhaps a good metaphor for the Spanish approach to
negotiations evinced in the data. When you make a phone call in Madrid,
the usual greeting on the other end is not hola ('hello') but is, instead,
diga ('speak'). The Spaniards likewise used the percentage of
commands (17 %) of any of the groups and gave comparatively little
332 John L. Graham
information (self-disclosures, 34 0/0). Moreover, they interrupted one
another more frequently than any other group, and they used the term
"no" and "you" very frequently.
France. The style of the French negotiators is perhaps the most aggres-
sive of all the groups. In particular, they used the highest percentage of
threats and warnings (together, 80/0). They also used interruptions, facial
gazing and "no" and "you" very frequently compared to the other
groups, and one of the French negotiators touched his partner on the arm
during the simulation.
Brazil. The Brazilian businesspeople, like the French and Spanish,
were quite aggressive. They used the highest percentage of commands of
all the groups. On average, the Brazilians said the word "no" 42 times,
"you" 90 times, and touched one another on the arm about 5 times
during 30 minutes of negotiation. Facial gazing was also high.
Mexico. The patterns of Mexican behavior are good reminders of the
dangers of regional or language-group generalizations. Both verbal and
nonverbal behaviors are quite different than those of their Latin
American (Brazilian) or continental (Spanish) cousins. Indeed, Mexicans
answer the telephone with the much less demanding bueno. In many
respects, the Mexican behavior is very similar to that of the negotiators
from the United States.
Francophone Canada. The French-speaking Canadians behave
quite similarly to their continental cousins. Like the negotiators from
France, they, too, used high percentages of threats and warnings, and
even more interruptions and eye contact. Such an aggressive interaction
style would not mix well with some of the more low-key styles of some
of the Asian groups or with English speakers, including Anglophone
Canadians.
Anglophone Canada. The Canadians who speak English as their first
language used the lowest percentage of aggressive persuasive tactics (that
is, threats, warnings and punishments totaled only 1 %) of all fourteen
groups. Perhaps, as communications researchers suggest, such stylistic
differences are the seeds of interethnic discord as witnessed in Canada
over the years. With respect to international negotiations, the
Anglophone Canadians used noticeably more interruptions and "no's"
than negotiators from either of Canada's major trading partners, the
United States and Japan.
United States. Like the Germans and the British, the Americans fell in
the middle of most categories. They did interrupt one another less fre-
quently than all the others, but that was their sole distinction.
Culture, negotiations and international cooperative ventures 333
These differences across cultures are quite complex. The key here is to
be aware of these kinds of differences so one doesn't misinterpret the
Japanese silence, the Brazilian "no, no, no ... ", or the French threat.
4.3. Content versus context
The findings of this study clearly suggest that our understanding of
negotiation processes is incomplete and perhaps inadequate if we rely
solely on analyses of verbal content. As Gumperz (1979) and others
suggest, the context of communication is crucial. Indeed, at the level of
Angelmar and Stern's (1978) content analysis, it is most difficult to
distinguish between cultural styles of bargaining. However, consideration
of both the structural aspects of language and nonverbal behaviors yields
substantial differences among the groups. That is, cultural background of
the negotiators affects the "contextual" more than the content-related
aspects of the negotiation process. This is consistent with the findings
of Neu and Graham (in press) who report that context variables have
stronger influences on negotiation outcomes than do content
variables.
5. Conclusions
These results, of course, are not definitive. The small sample sizes do not
allow for tests of statistical significance. How representative the partici-
pants are is problematic. They are all experienced business people and
citizens and permanent residents in each of their countries, which is an
improvement over most other business negotiation research where
students are used as surrogates for bargainers. But how well six business
people represent a "cultural style" cannot be determined. Finally,
external validity of the experimental setting is questionable. However,
this work represents an improvement over most other business negotia-
tion research by the use of face-to-face communication instead of written
or electronic means. The value and strength of this study are the
observational methods used to measure the negotiation process.
Videotaping allows for multiple observers and multiple observations
concurrent with the bargaining process. Thus, the reliability and validity
of the process measures developed does not depend on a priori
334 John L. Graham
experimental manipulations or post hoc partIcIpant self-reports. The
methods developed in this study are time consuming and expensive but,
as these findings suggest, potentially fruitful.
Cross-cultlJ.ral interactions were not explicitly considered here.
Graham (1980; 1985b) and Adler and Graham (1989) report findings
from such studies. Generally, the findings indicate that such cultural
differences in bargaining process as described above are potential sources
for friction alfd misunderstandings between bargainers that often result
in increased transaction costs in international commercial relationships.
For example, frequent interruptions of American negotiators by Brazilian
counterparts can lead to irritation and to inaccurate attributions of
rudeness when Brazilian executives are just conforming to Brazilian
norms for interactions. Likewise, lack of eye contact from Japanese
partners during negotiations' may lead to Americans' suspicions and
attributions of Japanese secrecy or even dishonesty. And such problems
can destroy cooperative relationships and preclude otherwise mutually
beneficial commercial agreements.
Perhaps the most worrisome of our findings is that greater differences
were related to how things were said than to what was said. That is, the
negotiation styles of the fourteen cultural groups are surprisingly similar
at the level of content: The clearest contrasts ~ e t w e e n groups were found
in structural and nonverbal aspects of conversational styles. Such
differences are generally not consciously perceived by negotiators. These
"hidden" problems lead not only to ethnic jokes, but worse yet to cross-
cultural disharmony, prejudices, and perceptions and feelings of ill will.
Thus, that necessary condition for joint venture success- an ambiance
of cooperation - can be lost for no apparent reason other than cultural
misunderstanding.
The findings of this study suggest that substantial differences in bar-
gaining styles exist across cultures. Reliable and valid measures of
negotiation processes have been developed. This exploratory work de-
serves follow-up research with larger sample sizes. Increased statistical
power associated with larger samples would allow for investigations of
not only the culture ---7 process relationship but also more complex rela-
tions such as culture ---7 process ---7 negotiation outcomes. The findings of
such studies will hold important implications for training business
executives and students to manage more efficiently the international
relationships of the future, increasingly taking place in a more global
marketplace.
Culture, negotiations and international cooperative ventures 335
References
Adler, Nancy J. - John L. Graham
1989 "Cross-cultural interaction: The international comparison fallacy",
Journal of International Business Studies Fall: 515-537.
Angelmar, Reinhard - Louis W. Stern
1978 "Development of a content analytic system for analysis of bar-
gaining communication in marketing", Journal of Marketing
Research 15: 93-102.
Archibald, Kathleen (ed.)
1966 Strategic interaction and conflict. Berkeley Institute of International
Studies: University of California Press.
Argyle, Michael - Mark Cook
1976 Gaze and mutual gaze. Cambridge: Cambridge University Press.
Bales, Robert F.
1950 Interaction process analysis: A method for the study of small groups.
Cambridge, MA: Addison-Wesley Press.
Bonoma, Thomas V. - Leonard C. Felder
1977 "Nonverbal communication in marketing: Toward a communica-
tional analysis", Journal of Marketing Research 14: 169-180.
Campbell, Nigel - Alain Jolibert - Hans-Gunther Meissner - John L. Graham
1988 "A comparison of marketing negotiations in France, Germany, the
United Kingdom, and the United States", Journal of Marketing 52:
49-62.
Cateora, Philip R.
1983 International marketing. (5th edition.) Homewood, IL: Richard D.
Irwin.
Condon, John C.
1968 "Linguistic-kinesic research and dance therapy", A.D.1:A. conven-
tion proceedings.
1974 "Introduction: A perspective for the conference", in: John C.
Condon - M. Saito (eds.), 3-14.
Condon, John C. - Mitsuko Saito (eds.)
1974 Intercultural encounters in Japan. Tokyo: Simul Press.
Dwyer, Robert F. - Orville C. Walker
1981 "Bargaining in an asymmetrical power structure", Journal of
Marketing 45: 104-115.
Ehlich, Konrad - Johannes Wagner
in press The discourse of international negotiations. Berlin: Mouton de
Gruyter.
Engel, James F. - Roger D. Blackwell
1982 Consumer behavior. Chicago: The Dryden Press.
Erickson, Fred
1976 "Talking down and giving reasons: Hyper-explanation and listening
behavior in interracial interviews", paper delivered at the Inter-
national Conference on non-verbal behavior, Ontario Institute for
Studies in Education, Toronto, Canada, May 11, 1976.
336 John L. Graham
Fouraker, Lawrence E. - Sidney Siegel
1963 Bargaining behavior. New York: McGraw-Hill.
Frake, Charles o.
1972 "Struck by speech: The Yakan concept of litigation", in: John J.
Gumperz - Dell Hymes (eds.), 106-129.
Francis, June N.P.
1991 "When in Rome? The effects of cultural adaptation on intercultural
business negotiations", Journal of International Business Studies
22.3: 403-428.
Graham, John L.
1980 Cross-cultural sales negotIatIons: A multilevel analysis. [Un-
published Ph. D. dissertation, School of Business, University of
California, Berkeley.]
1983 "Brazilian, Japanese, and American business negotiations",
Journal of International Business Studies Spring/Summer:
81-96.
1985 a "The influence of culture on the process of business negotiations: An
exploratory study", Journal of International Business Studies Spring:
81-94.
1985 b "Cross-cultural sale negotiations: A laboratory experiment",
Marketing Science 4: 130-146.
1987 "Deference given the buyer: Variations across twelve cultures", in
Peter Lorange - Farok Contractor (eds.), 473-486.
Guetzkow, Harold
1960 "Unitizing and categorizing problem in coding qualitative data",
Journal of Clinical Psychology 6: 45-53.
Gumperz, John J.
1979 "Sociocultural knowledge in conversational inference", 28th Annual
Roundtable Monograph Series on Languages and Linguistics,
Georgetown University.
Gumperz, John J. - Dell Hymes (eds.)
1972 Directions in sociolinguistics. New York: Holt, Rinehart and
Winston.
Hall, Edward T.
1976 Beyond culture. Garden City, NY: Anchor Press/Doubleday.
Harrigan, Katherine R.
1987 "Why joint ventures fail", Euro-Asia Business Review 6.3:
20-26.
Harnett, Donald L. - Larry L. Cummings
1980 Bargaining behavior: An international study. Houston, TX: Dane
Publications, Inc.
Jastram, Rothdow W.
1974 "The Nakado negotiators", California Management Review 17.2:
86-90.
Kapoor, Ashok
1974 "MNC negotIatIons: Characteristics and planning implications",
Columbia Journal of World Business 9: 121-130.
C u l t u r e ~ negotiations and international cooperative ventures 337
Kay, Paul
1970 "Some theoretical implications of ethnographic semantics", Current
Directions in Anthropology Bulletins of the American Anthro-
pological Association 3:2.
Kelley, Harold H.
1966 "A classroom study of the dilemma in interpersonal negotiations",
in: Kathleen Archibald (ed.), 49-73.
Kelman, Herbert C. (ed.)
International behavior: A social-psychological analysis. New York:
Holt, Rinehart and Winston.
Lewis, Steven A. - William R. Fry
1977 "Effects of visual access and orientation on the discovery of inte-
grative bargaining alternatives", Organizational Behavior and
Human Performance 20: 75-92.
Linton, Ralph
1945 The cultural background of personality. New York: Appleton-
Century-Crofts.
Lorange, Peter - Farok Contractor (eds.)
1987 Cooperative strategies in international business. Lexington, MA:
Lexington Books.
Nakane, Chie
1970 Japanese society. Berkeley, CA: University of California Press.
Neu, Joyce - John L. Graham
in press "An analysis of language use in negotiations: The role of context and
content", in: Konrad Ehlich - Johannes Wagner (eds.).
Pennington, Allan L.
1968 "Customer-salesman bargaining behavior in retail transactions",
Journal of Marketing Research 5: 255-262.
Pruitt, Dean G. - Steven A. Lewis
1975 "Development of integrative solutions in bilateral negotia-
tion", Journal of Personality and Social Psychology 31.4:
621-633.
Ricklefs, Richard
1978 "For a businessman headed abroad, some basic lessons", The Wall
Street Journal. January 16: 4.
Rubin, Jeffrey Z. - Bert R. Brown
1975 The social psychology of bargaining and negotiation. New York:
Academic Press.
Sawyer, Jack - Harold Guetzkow
1965 "Bargaining and negotiation in international relations", in: Herbert
C. Kelman, (ed.), 464-520.
Soldow, Gary F. - Gloria Penn Thomas
1984 "Relational communication: Form versus content in the sales inter-
action", Journal of Marketing. 48: 84-93.
Spiro, Melford E.
1950 "Culture and personality: The natural history of a false dichotomy",
Psychiatry 14.1: 19-46.
338 John L. Graham
Tung, Rosalie L.
1982 "D. S.-China trade negotiations", Journal of International Business
Studies Fall: 25-38.
Deda, Keiko
1974 "Sixteen ways to avoid saying 'no' in japan", in: john C. Condon -
Mitsuko Saito (eds.), 185-192.
Van Zandt, Howard F.
1970 "How to negotiate in japan", Harvard Business Review November-
December: 45-56.
Walton, Richard E. - Robert B. McKersie
1965 A behavioral theory of labor negotiations. New York: McGraw-Hill.
Weiss, Stephen E.
1987 "Creating the GM Toyota joint venture: A case in complex negotia-
tion", Columbia Journal of World Business 22: 23-37.
1990 "The long path to the IBM Mexico agreement An analysis of the
microcomputer investment negotiations, 1983-86", Journal of
International Business Studies 21.4: 565-96.
Wells, Louis T.
1972 "Negotiating with third world governments", Harvard Business
Review january-February: 72-80.
List of Contributors
Mahmoud Al Batal
Department of Near Eastern Studies
Emory University
Atlanta, GA 30322
Leslie M. Beebe
Program in Applied Linguistics
Teachers College
525 West 120the Street
Columbia University
New York, NY 10027
Jean "U7. Bodman
American Training Institute and School of Education
Department of Teaching and Learning
239 Green Street
Room 635 East Building
Washington Square
New York University
New York, NY 10003-6674
Diana Boxer
Program in Linguistics
University of Florida
112 Anderson Hall
PO Box 115454
Gainesville, FL 32611
Mary Carpenter
School of Education
Department of Teaching and Learning
239 Green Street
Room 635 East Building
Washington Square
New York University
New York, NY 10003-6674
340 List of Contributors
Andrew Cohen
Institute of Linguistics & Asian & Slavic Languages & Literatures
196 Klaeber Court
University of Minnesota
320 16th Avenue SE
Miineapolis, MN 55455
Martha Clark Cummings
Monterey Institute of International Studies
Department of TESOL
425 Van Buren Street
Monterey, CA 93940
Miriam Eisenstein Ebsworth
School of Education
Department of Teaching and Learning
239 Green Street
Room 635 East Building
Washington Square
New York University
New York, NY 10003-6674
Waguida El Bakary
American University in Cairo
PO Box 2511
Cairo Egypt
Susan M. Gass
English Language Center
Michigan State University
East Lansing, MI48824-1035
Michael L. Geis
Department of Linguistics
222 Oxley Hall
1712 Neil Avenue
Ohio State University
Columbus, OH 43210
Myra Goldschmidt
English Department
St. Augustine Center
Villanova University
Villanove, PA 19085
John L. Graham
Graduate School of Management
University of California at Irvine
Irvine, CA 92717
Linda L. Harlow
Department of French and Italian
248 Cunz Hall
1841 Millikin Road
Ohio State University
Columbus, OH 43210
Noel Houck
Temple University of Japan
Nagaoka Building #1
4-40-12 Takadanobaba
Shinjuku-ku
Tokyo 169
Japan
Hy-sook Jeong
61 Tong Eui Dong
Jong Ro Gu
Seoul, Korea
Gabriele Kasper
Department of English as a Second Language
1890-East-West Road
University of Hawai'i at Manoa
Honolulu, HI 96822
Dale April Koike
Spanish and Portuguese Department
Batts Hall 110
University of Texas at Austin
Austin, TX 78712
Naoko Maeshiba
Department of Theatre and Dance
University of Hawai'i at Manoa
1170 East-West Road
Honolulu, HI 96822
Beth Murphy
1305 Redbud Hill Apartments
Bloomington, IN 47406
List of Contributors 341
342 List of Contributors
Gayle Nelson
Department of Applied Linguistics and English as a Second Language
Georgia State University
University Plaza
Atlanta, GA 30303
Joyce Neu
Conflict Resolution Program
The Carter Center
One Copenhill
Atlanta, GA 30307
Steven Ross
School of Policy Studies
2-1 Gakuen
Sanda, Hyogo 669-13
Japan
Richard Schmidt
Department of English as a Second Language
1890-East-West Road
University of Hawai'i at Manoa
Honolulu, HI 96822
Akihiko Shimura
3-26-23-208 Minowa
Kohoku, Yokohama 233
Japan
Zhigang Wang
3123 Herbert St. #A3
Honolulu, HI 96815
Naoko Yoshinaga
Department of Linguistics
1890 East-West Road
University of Hawai'i at Manoa
Honolulu, HI 96822
Subject index
addiction of responsibility 205
addressee 241-242, 245, 254, 255
see also speech act, reception
advice/lecture 219
34-35,40
age 242,244,248,252,254
agreement/commiseration 219, 226,
230
Americans 95-101,109,111-112,
120,124-125,165,167,171,
172-177,178-181,192-193,
195-196, 198-211
Anglos 90
apology
form of 164
see also speech act, apology
approximation 34, 38
Arabs 100,111, 124
Australian English 159
avoidance 74, 76-77, 97
verbal 47, 55, 58, 74, 78
blessings
see speech act, blesings
British English 160
bulge theory 6, 72, 160, 229
Canadian French 159
communication 102, 285
communicative competence 109-110,
122, 191
communicative conventions 89
complaint
see speech act, complaint
compliment form 114
Arabic 112-113, 124
Egyptian 114-120, 123
confirmation
see speech act confirmation
context 321
continuers 79
contradictions 219, 228-229
criticism
see speech act, criticism
Cross-Cultural Speech Act Project 22
cross-cultural 1-2, 65, 90, 100,
109-110,122,158,175-176,308,
310,334
differences 2, 13, 286, 291, 317,
320
dissonance 96
cultural 21
values 317
culture 1,12,22-23,75,101-102,
285,317,319
Danish 160
data collection 4,65-81,90-91,
102,161,197,210,221,292
declarative intonation 148-149
negative 133, 146, 148
positive 133, 146-147
delivery of a different thought 34, 37
"din in the head" 4, 34
Discourse Completion Test 4-5,
46-51,59,65-69,71-78,80,
217
see also role play
Egyptian Arabic 109,113,124
elaboration 49, 52-54, 57, 70-72,
79-80
empathy 78
English 134,137,139,142-149,
161,163-165,181-182,
194-195,260-264,267,273
as a foreign language 40, 51, 91,
157
ethnicity 224, 231
ethnographic data collection 45
344 Subject index
ethnographic interview 217,
220-221,224,235-236
ethnographic observation 24
"evil eye" 7, 112-113
explanation
see speech act, explanation
explanation of purpose 195, 203, 209
face-saving 49, 134
face-threatening 12, 58, 65, 136, 165,
175,191,223,242,299
facial gazing 328, 331-332
favor-asking
see speech act, favor-asking
formulaic speech 4
use of 34, 36
French 133-134,137,139,
141-149
gender 223-224, 230, 242,
244-245, 251
Georgians 100
German 156, 159
head act 288-289
hearer 286
see also speech act, reception
Hebrew 159-160, 193-194
hedging 1, 77-79
Hispanics 90,99-100
I. F. I. D. (Illocutionary Force Indicating
Device) 164,170-171,174-176,
179, 181
illocutionary 134-135, 155, 203,
261,263,285,288
immediate retrospection 32
imperatives 290-291, 294-295,
303-308
imposition 241, 244, 250, 252-254
interlanguage 257
interlanguage pragmatics 45-46,
65
intimacy (of participants) 72
invitations
see speech act, invitation
Iranians 100, 111
Japanese 111-112, 124-125, 156,
159,161,163-165,167, 169,
172-177,179,180-181
Jewish 224, 231-232
see also ethnicity
joke/teasing 219
justification 203, 209, 210
Korean 192,195-211
language habits 129
language of thought 32
language system 129
lexical avoidance/simplification 34,
38
linguistic form 131, 191, 203
variation 131
mesage omission/abandonment 34,
36
mitigators 50, 175
monitor 34-35
multiple choice questionnaire 46
native speaker 7, 10, 46, 52, 56-58,
69,72,81,89,91,97-99,101,
125,158-159,165,169-170,
178-179,192-196,199,206,208,
211,217-218,242,291,323
natural data 5,66-67,91-92,
102
negative 262, 267
reaction 274
use of 260-261
negotiation 14,47-49,52,57-59,
71-72,193,274,319-320,322,
327,330-332,334
New Zealand 111, 159
non-native speaker 3, 5, 7, 46,
52-59,89,91,96-99,101,111,
125,155,158-159,170,191-196,
199-200,208,211-212,
217-218,241,255,271,275,
291
non-verbal 5,51,53-56,59,95-96,
99,101,140,320-321,326-327,
330,332-334
open-ended elicitations 25
overlaps 387
partial delivery of a thought 34, 37
personalization of the problem 205
phatic expression 90, 92-93, 101
politeness 8, 75, 137, 139, 142,
149-150,155,195,257,303-304,
306,311
pragmalinguistic 102, 155-157, 160,
308
failure 98
pragmatic competence 257
Puerto Ricans 90, 97
qeustions 219
real play 24
response/topic switch 219, 227
requests
see speech act, requests
role 242, 298
role play 3, 5, 24-25, 29, 31, 46-49,
51-52,65-67,69,91,96,98-99,
102
closed 46-47, 59
interview 28, 30
open 46-48,52,55,59,66
oral 75, 77
written 71, 75
Russians 100
"satisfaction" conduction 136, 142
second language
acquisition 191
learners 40, 109, 129, 149-150,
157, 160
pedagogy 102
problems of 129
157, 160
self-control 52
self-debate 34
Semi-Direct Oral Proficiency Interview
30
semi-ethnographic 24, 28
silent periods 327, 330-331
sincerity principle 92, 135
Subject index 345
social differentiation 174
social distance 159-160, 174, 178,
223,228-229
social power 159, 175, 272
sociocultural
ability 3, 22, 27
knowledge 122, 124, 192
sociolinguistic
ability 23, 27, 98, 110, 155
acceptability 194, 206
behavior 209
errors 218
knowledge 122, 124, 192
rules 109-110, 122, 218
sociopragmatic 96, 98, 102, 132,
139,148-150,155,156,170
solution
candidate request 201-203, 209
Spanish 160, 260'-264, 273, 276
speaker 286
see also speech act, production
speech act 1-5,7,13,21-22,24,29,
33,40,45,52,58-59,67,72,80,
91, 101, 125, 134-137, 191-194,
217,220,242,257-258,260,
263-265,267,273-275,285,
287-288
agreements 52, 55, 57-58
alternatives 53
apologies/apologizing 9-10, 21,
28-29,48-49,54,77-79,93,
156,158-160,165,167-172,
174-176,178-182,192-194
direct 179
ritual 159
substantive 159
blessings 109
complaints 1, 28-29, 100, 191,
196-211,223,226,235-236
indirect 219, 223-224,
226-227,230,232-236
compliments 1,7-8,21,109-120,
122-126,192,220
confirmations 53-58
criticism 197-198, 200,
202-211
explanations 53-54
346 Subject index
favor-asking 11, 242-244,
253-254
see also speech act, requests
goals 21-22
greetings 6-7, 89-92, 94-102
all-business 95
chat 94
intimate 95
introductory 95
long 94
on the run 101
re-greeting 96
speedy 93,97-100
griping 220, 226
invitations 50-52, 72, 100, 192,
220
ambiguous 192
unambiguous 192-193
offers 51-52, 220
production 27-29,33,45,102,
109-110,194
provide information 132
reasons 53
reception 27, 102
refusals 4, 45, 47, 49-53, 57-58,
65,70,73-75,77,157
repetitions 70, 72, 80
restatement 56
requests 1, 48, 50-52, 55, 58,
69-70, 77, 129, 132-133,
135-137, 140, 142, 145, 150,
155,158,201,310-311
suggestions 2, 12-13, 50-52, 58,
260-264,267,271-272,285,
287-290,293-299,303,305,
310
to buy 300, 302
to enjoy 300
to get 300, 302
to use 300
unspecified 300
thanking 1
speech act behavior 22, 27, 39-40,
194,217,241,286
speech act event 288
speech act set 10-11, 21-22, 93,
158,193,195,198-199,205
speech act strategies 22-23, 39,
74-76,80,158,163,175
speech community 67-68,217,219,
236-237,241,253-254
spontaneous speech 221, 235-237
status
non-student 242, 244-245, 248,
254
student 242, 245, 248, 254
stylistic 130-132, 134
see also linguistic variations
supporting 288-289
Swedish 101
TESOL 49, 69, 72-73, 78-79, 81,

Thai 160, 165, 181
touching 329, 332
transfur 10, 110, 156, 160, 168,
257-259,261,268,275
error 259
negative 110, 165, 167, 169-171,
174,176,179-181
positive 167-169,180
pragmatic 110,155-157,165,
167,175,217,237,257,259
strategies 257-258
see also speech act, strategies
zero 170
triangulation 4, 39
"troubles telling" (aka talk or sharing)
218-219,222-224,226,228,232,
235,237
turns 52, 56-57, 2\89
length 5, 52
number of 52, 70
Ukrainians 100
use of second person \ 205
verbal report 26, 28, f2-33, 39-40
written discourse completion 33
Author index
Adler, Nancy J. 334
Afflerbach, Peter 26
Agoston, Thomas 287
Aguilar Murillo, Evelyn 24
Al Batal, Mahmoud 109, 124
Almancy, A.J. 123, 126
Alwan, A.J. 123, 126
Angelman, Reinhard 320, 323, 326,
330,333
Araki, Shoko 111, 124
Argyle, Michael 328
Aston, Guy 285
Austin,John 1,192,241
Bales, Robert F. 320
Banerjee, Janet 65, 288, 290-291,
311
Bardovi-Harlig, Kathleen 47, 53, 55,
57-58
Barnlund, Dean C. 111, 124, 159,
175-177,179-180,182
Barthes, Roland 287
Beebe, Leslie 4, 47-51, 53, 55,
65-68,96,155-157,165,175,
181,217,291
Belk, Russel W. 290
Bell, Robert A. 112
Bergman, Marc 77, 155, 158-160,
163-165,174-175,181
Bhatia, T. K. 287
Billmyer, Kristine 40
Blackwell, Roger D. 319
Blum-Kulka, Shoshana 9,21-28,
52,65,77,110,123,125,134,
155-157,194,241,258,260,275,
287-288
Bodman, Jean 66, 91, 93, 96, 155,
179
Bolinger, Dwight 287
Bonoma, Thomas V. 320-321
Borkin, Ann 158
Bowen, J. Donald 90, 92
Boxer, Diana 100, 224,
229-230
Bratt Paulson, Christian 101
Brown, Bert R. 319
Brown, Dorothy F. 111
Brown, Penelope 2, 9, 135-137,
223,233,242-243,262,
286
Brown, Roger 72
Caballero, Marjorie 289
Camargo, Eduardo G. 287, 290
Canale, Michael 109
Carrell, Patricia 65, 195, 288,
290-291,311
Cateora, Philip R. 321
Chan, Carole 89
Clyne, Michael 157
Cohen, Andrew 1-2, 9, 21,
25-26,28-32,40,45,65,70,
77, 93, 155, 157-160,
193-194,198
Coleman, Linda 287
Condon,JohnC.320,327
Cook, Mark 328
Corder, S. Pit 257, 258
Coulmas, Florian 311
Cowles, Maria A. 30
Cummings, Larry L. 318
Cummings, Martha C. 48-49, 96,
217
D'Amico-Reisner, Lynne 192, 219
Dahl, Merete 22, 45 -46, 48, 65-66,
285
Daikuhara, Midori 67
Di Vito, Nadine O. 133, 147
Dogancay, Seran 89
348 Author index
Dowling, Graham R. 289
Doyle, Fatima 30
Du Fon, Margaret A. 155
Duffy, Susan A. 26
Dsirat, Claude 137
Dwyer, Robert F. 319
Ebsworth, Timothy J. 91, 99
Edmondson, Willis 48, 288, 303
Eisenstein, Miriam 66, 91, 93, 96,
155,179
El Bakary, Waguida 109, 124
Ellis, Rod 97
Engel, James F. 319
Erickson, Fred 327
Ervin-Tripp, Susan 218, 242-243,
307
Faerch, Claus 26, 155
Fang, Hanguan 310
Felder, Leonard C. 320-321
Ferrara, Alessandro 193-194
Fields, George 290
Fouraker, Lawrence E. 322
Frake, Charles o. 320
Francis, June N.P. 318
Fraser, Bruce 70, 148, 159, 287
Fry, William R. 319,322,328-329
Gadet, 133
Garcia, Carmen 155, 160
Gass, Susan 2, 51, 59, 258-259
Geis, Michael L. 135, 287, 309
Giles, Howard 157
Gilman, Albert 72
Goffman, Erving 72, 77-78, 134,
159,286
Goldschmidt, Myra 242
Gordon, David 242
Graham, John L. 318-319, 321,
326-327,330,333-334
Green, Georgia M. 290
Grice, H. Paull, 285
Guetzkow, Harold 319-320, 326
Gumperz, John 1, 87, 241, 285, 321,
327,333
Gymn, Kyng-Il 292
Habermas, Jiirgen 1
Hall, Dennis R. 287, 308, 321
Hard, Tristan 133
Harnett, Donald L. 318
Harrigan, Katherine R. 317
Harris, Richard J. 287
Hartford, Beverly 47, 53, 55,
57-58
Hatch, Evelyn 219
Heng, J.H. 310
Hilferty, Ann 90
Ho, Suk-ching 290
Hobbs, Dianne 261
Hodlin, Susan 89
Holmes, Janet 9, 68, 111, 159
Hong, Jae W. 289
Hopper, Robert 112
Houck, Noel 51, 59
House, Julianne 9,22,65,77,110,
125,134,155-156,159-160,258,
260,287-288,303
Huber, Lisa 192
Hunt, Shelby D. 289
Hymes, Dell 1, 67, 109, 191, 219,
241,243
Jacobs, Lawrence W. 292
Jastram, Rothdow W. 318
Jefferson, Gail 218
Johns, Ann M. 286
Johnson, Patricia 157
Johnston, Peter 26
Jones, Steve 45-46,160
Jordens, Peter 156
JuPP, Thomas C. 89
Kachru, Yamuna 218
Kaplan, Robert B. 291
Kapoor, Ashok 318
Kasper, Gabriele 9, 22, 26, 45 -46,
48,65-66,77,110,125,134,
155-160,163-165,167,
174-175,181,258,260,285,
287-288
Katriel, Tama 217
Kay, Paul 320
Kaynak,Erdenor 289
Kellerman, Eric 145, 156, 258
Kelley, Harold H. 322
Kenyon, Dorry M. 30
Kess, Joseph F. 287
Keown, Charles F. 292
Kinjo, Hiromi 50, 53
Knapp, Mark L. 112, 124
Koike, Dale 257, 261-262
Konneker, Beverly 195
Krugman, Dean M. 289
Labov, William 241
Laczniak, Gene R. 289
Lakoff, Rubin Tolmach 286-287,
291
Lee,JohnR.E.218
Leech, Geoffrey N. 110, 155, 287,
291,308-309
Levenston, Edward 241
Levinson, Stephen 2,9,89,135-137,
223,233,242-243,262,286
Lewis, Steven A. 319-320, 322,
328-329
Linton, Ralph 319-320
Lu, Zaiming 290
Mack, Robert L. 26
Madden, Carolyn 289
Maloney, Clarence 112
Manes, Joan 65,68,111-112,
117-119, 122, 124, 192
Marmoor, Thomas 45-46, 160
Masavisut, Nitaya 287
Matsukubo, Sinya 289
Matsumoto, Yoshiko 291
McKersie, Robert B. 320
Miracle, Gordon E. 290
Mitchel, Lionel A. 289
Mitchell, Candace J. 45-46, 48, 65,
77
Moeran, Brian 287
Morgan, Graham 89
Muderrisoglu, Aydin 289
Murphy, Beth 209
Nakane, Chie 326
Nelson, Gayle L. 109, 124
Author index 349
Neu, Joyce 333
Nine-Curt, Carmen J. 90, 99
Nolen, William 148
O'Barr, William M. 287
Ochs, Elinor 1
Olshtain, Elite 1-2, 9, 21, 23-32,
40,45,52,65,70,77,93,123,
155-160,174,193-194,198,288
Olson, Gary M. 26
Owen, Marion 77,79
Paiva, Riccardo 30
Pennington, Allan L. 319, 326
Pruitt, Dean G. 319-320, 322
Reinhart, Susan M. 158
Resnik, Alan 289
Rice, Marshall D. 290
Richards, Jack 110
Ricklefs, Richard 321
Rintell, Ellen M. 45-46, 48, 65, 77,
97,288,290
Robinson, Mary 155-157
Rose, Kenneth 46, 66
Rubin, Jeffrey Z. 319
Rubin, Joan 55
Sadock, Jerrold 1
Saracino-Resh, Linda 287, 308
Sawyer, Jack 319-320
Scarcella, Robin 4, 51, 90
Schachter, Jacquelyn 259
Schegloff, Emanuel 79
Schiefflin, Bambi 1
Schmidt, Richard W. 110
Schmidt, Rosemarie 287, 292
Schroeder, Kim 287
Searle, John 1, 90, 92, 109, 135-136,
191-192,241,260,285,287
Selinker, Larry 167, 257-258
Sheffer, Hadass 157
Sherry, John F. 287, 290
Shouby, E. 123
Siegel, Sidney 322
Sin, Yat-ming 290
Slobin, Daniel 92
350 Author index
Soldow, Gary F. 321
Spindel, Carol 109
Spiro, Melford E. 320
Spradley, John 221, 222
Stansfield, Charles 30
Stern, Bruce L. 289, 320, 323, 326
Stern, Louis W. 330, 333
Stubbs, Michael 25
Suleiman, Michael W. 123
Sukwiwat, Mayor 287
Swain, Merrill 109
Takahashi, Satomi 155-156, 179,
291,295,306-307
Takahashi, Tomoko 4, 47, 50, 53,
65-66,68,155-157,165,175,
181,291
Tanaka, Noriko 159
Tannen,Deborah 2,75,79,219,223
Tarone, Elaine 2
Thomas, Gloria Penn 321
Thomas, Jenny 98, 155, 218
Trosberg, Anna 77, 160
Tse, David K. 290
Tung, Rosalie L. 318
Ueda, Keiko 326
Uliss-Weitz, Robin 4, 47, 50, 53, 65,
68, 155
Ulsh, Ines 3
Valdman, Albert 133
Van Dijk, Teun 193-194
Van Zandt, Howard F. 318, 320,
326
Vestergaard, Torben 287
Vollmer, Helmut J. 159
Von Raffler, Engel 287
Walker, Orville C. 319
Walters, Joel 47
Walton, Richard E. 320
Walz, Joel C. 147
Watson Gegeo, Karen Ann 45
Wells, Louis T. 318
Weiss, Stephen E. 318
Weizman, Elda 21, 275, 287
Widdowson, Henry 218
Wolfson, Nessa 22, 24, 40, 45-46,
65 -68, 72, 97, 109-112, 117,
122, 124-125, 134, 160, 192,
218,229
Wongmontha, Seri 287
Yamaki, Toshio 290
Yoshioka, Miho 159,175-177,
179-180,182
Zhou, Nan 290
Zinhan, George M. 289