Speech Acts Across Cultures: Challenges to Communication in a Second Language 9783110219289, 9783110140828, 9783110191257

Citation preview

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, 1950III. 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 .

1

Part I Methodological issues

Andrew Cohen Investigating the production of speech act sets.

21

Noel Houck - Susan M. Gass Non-native refusals: A methodological perspective.

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 ~arpenter 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 .

343

Author index .

347

Introduction Susan M. Gass

1. Introduction This book investigates the notion speech act from a cross-cultural perspective. 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, complimenting, 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 mitigation. 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 following 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 himself, "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 someone 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 constraining 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 comprehensive picture is able to emerge. In Chapter Three, Houck and Gass respond to the fact that a significant 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 methodological 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) invitations, 2) suggestions, 3) offers and 4) requests. They depart from previous work on speech acts in two important ways: 1) by using videotaped 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 interaction 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) nonverbal 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 methodological 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 elaborations 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 refuser shared a common profession and were members of the same professional 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 openended 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 importance 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 compliments of the two cultures studied are slight and because the differences may cause embarrassment and discomfort, it is crucial that the differences 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 formulating 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 communicating a particular gist will take (i. e., syntax, morphology, etc.) is then determined 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 propositional 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 interlanguage 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, disrespectful, 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 ethnographic 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 perceptions 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 complaining. 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 intervention) 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 complaints. The next chapter by Goldschmidt investigates the variables that determine how people ask favors of each other. Goldschmidt further attempts to ascertain if the asking of favors in American English is a strategydominated 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 situations. 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 prerequisite 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 television 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 advertising 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 intracultural 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: 171180. 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 representation''', 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 distinguish 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 explanation 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 interactional 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 CrossCultural 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, registration 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 formula, 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. ~

acceptability

~

ethnographic

- - - - - . . . . 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 interlocutor (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 realization - 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 encompasses 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 include data that reflect self-report (learners' descriptions of what they do, characterized by generalized statements about learning behavior), selfobservation (the inspection of specific, not generalized language behavior introspectively or retrospectively), self-revelation (think-aloud, stream-ofconsciousness 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 roleplay 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 reception 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 sociolinguistic 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 circumstances? 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 limitation, 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, complaint, 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, complain, 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, regardless 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 situation, 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 illustrate 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 necessarily 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 complaint 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 suggestions 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 following 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 otherwise 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 (immediate 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 discussed 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 categories reflect insights gained from the use of verbal report protocols: Selfdebate, 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 whereby 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" situation, 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 expressions 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 nonnatives (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 situation, 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 necessary 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 acceptability 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 interactions 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 information verbal report data which add insights regarding the cognitive processes and conscious strategies used to interpret their role in an interaction 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-appropriate 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 languagerelated 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 compliment", 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 comprehension 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

perspective~1-

Noel Houck - Susan M. Gass

1. Introduction Much research in recent years has been carried out in the area of interlanguage 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 participants, 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 occurrence 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 previous 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 description of a situation followed by dialogue with a blank line where the subject 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 dialogue 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 advantages 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 formulas, fewer status-preserving strategies, and none of the extended negotiations 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 videotaped, 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 linguistic behavior of individuals in the particular setting established by the researcher. With these caveats in mind, in this study we have adopted open roleplays in order to study refusals. We have done this fully aware of the disadvantages, 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 modalities 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 telephone 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 negotiation, 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 accommodate 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 understanding 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 videotaped 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 "sandwiching" 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 nonnative 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 comprehensive 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 responses. -

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 nonverbal responses.

4.2.1. Confirmations Confirmations occurred frequently in the conversations of lowerproficiency 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. 5. 6. 7. 8. 9. 10.

~

NNS: NS: NNS: NS: ~ NNS: NS: ~ NNS:

((nods)) Yes Why Vh, I head(eh) headek headache Y, have ... headache? Headache Oh you have a headache oh no he has a headache 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 explanation. 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 recognize 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. 2.

~

NNS: What means 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 confirmations and requests for information, although it is likely that the clarification 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 refusals, 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. 2. 3. 4. 5. 6.

NNS: NS: NNS: NS: NNS: NS:

7.

~NNS:

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 proposition 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. 2. 3. 4.

~

NS: NNS: NS: NNS:

((to others )) She's afraid ((nodding)) Mm But you're sure you don't want to go skydiving ((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 restatement 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: ... they want you to give a speech to everybody in the church. Is that OK? 2. NNS: Urn urn it's it's very short time for (me) NS: Oh OK you do not want to give a long speech 3. 4. ~NNS: ((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 uncertainty 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 appropriate 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 conventionally 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 nonnative 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 nonnative 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 conditions, 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 potentially painful, or when subjects might be seen as disappointing their hostess if they refused, they often eventually agreed. However, if a positive 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 circumstances 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 discussions of non-native refusals. The negotiations we have described go far beyond the notion of a simple response consisting of linguistically analyzable units. They involve art interaction not only between what the nonnative 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

3.

4.

5.

6.

7.

8.

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] 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 skydiving 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] 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] 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] 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] 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] 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 presented, 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 interactions, 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 observed). 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 following 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 without putting herself at risk. 4. For the lower proficiency students we analysed the data from two of the skydiving situations, two of the speech at church situations and two of the haircut 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 interactions between two native speakers, we are confident that in our data, given the participants, given the intonation and given the body and facial expressions, 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 pragmatics", 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 Sociolinguistics Colloquium. TESOL National Convention, Atlanta, April. Beebe, Leslie - Martha C. Cummings 1985 "Speech act performance: A function of the data collection procedure", Paper presented at the Sixth Annual TESOL and Sociolinguistics 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.), 3352. 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 crosscultural 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 crosscultural 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 (BlumKulka 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 supplement to written discourse completion data in discussing performance of face threatening acts between interlocutors of different status: disagreement, 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 openended 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 interactive 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 interacts 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 ethnographic data and notebook data are often unsystematic. The social characteristics (e. g., age, socioeconomic status, ethnic group) of the informants 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 taperecorded and transcribed with permission from the respondents. Once the data were collected, we counted the number of words used by respondents 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 volunteers 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 taperecorded 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 Conversation (Oral) n =11

Total Frequency words sentences semantic formulas semantic repetitions semantic elaborations turns

611 60 61 1 1 20

2719 229 103 31 19 85

Frequency in First Turn words sentences semantic formulas semantic repetitions semantic elaborations

497 39 42 1 0

386 26 25 3 3

Average Frequency per Refusal words sentences semantic formulas semantic repetitions semantic elaborations turns

55.54 5.45 5.54 0.09 0.09 1.81

247.18 20.87 9.36 2.81 1.72 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 indicated 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 everything 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 participants 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 intimates 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 situations 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 Strategies for Refusal

Direct performative verb no negative ability/willingness

Discourse Completion Test (Written) n = 11

Telephone Conversations (Oral) n= 11

0 1 9

0 0 8

8 0 11

4

alternative offer alternative suggest alternative condition for acceptance promise principle philosophy

0 1 2 1 0 0

1 2 3 0 2 1

attempt to dissuade guilt trip criticism request for empathy

0 0 1

1 3 4

avoidance nonverbal avoidance topic switch joke for avoidance repetition of request postponement hedging

0 0 1 1 1 0

1 0 0 0 2 7

Adjuncts to Refusals positive feeling/opinion empathy gratitude self-defense

6 0 1 1

5 2 0 1

Indirect regret wish excuse/reason

2 9

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 subcategories 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 unevaluated 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 Discourse Completion Test (Written) n = 11

Telephone Conversations (Oral) n = 11

Direct performative verb no negative ability/willingness

0 1 12

0 0 14

Indirect regret wish excuse/reason

11 0 17

2 16

alternative offer alternative suggest alternative condition for acceptance promise principle philosophy

0 1 2 1 0 0

1 3 3 0 2 1

attempt to dissuade guilt trip criticism request for empathy

0 0 1

1 6 6

avoidance nonverbal avoidance topic switch joke for avoidance repetition of request postponement hedging

0 0 2 1 2 0

1 1 0 0 3 15

Adjuncts to Refusals positive feeling/opinion empathy gratitude self-defense

7 0 1 1

12 3 0 1

Semantic Formulas and Strategies for Refusal

9

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 telephone 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 psychological (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; Trosborg 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 interaction. Only one person expressed empathy on the Discourse Completion Test and that was the data collector's close friend. It wasn't seen as necessary 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 expressions 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 acknowledgement 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 inadvertently 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 'eliciting' 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 alternative, 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 questionnaire 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 communication, 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 refusals", 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 1989 a "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 communication 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 realization", 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. o Ishtain, 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 nonnative 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 SavilleTroike (eds.), 93-111. Tannen, Deborah (ed.) 1982 Analyzing discourse: Text and talk: Georgetown University roundtable 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 investigated 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 int~ractional 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 breakdown 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 functions 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 communication. "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 researchers 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 successfully 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 natural 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 composed 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/nonnative pairs; and ten were non-native/non-native pairs.

92

Miriam Eisenstein Ebsworth - Jean

m Bodman -

Mary Carpenter

We conducted twenty open-ended post-hoc interviews with participants 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: She: He: She:

Hi, how've you been? Not bad. 'N you? Oh, can't complain. Busy. I know. Me, too.

94

Miriam Eisenstein Ebsworth - Jean

He: She:

Oh well, gotta take off. See ya. * Bye. Take care.

~

Bodman - Mary Carpenter

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 communication is introduced. This conversation took place between two female friends in their early 20's in an office lounge area. Female Female Female Female

1: 2: 1: 2:

Hi! Hi. Howa ya doin'? 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:

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: (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.) Bea: Michelle: Well, I'm glad you're back. It's so nice to see you. I missed talking to you. Bea: 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 nonverbal 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: Joe Matone: Client: Joe Matone:

Mr. Matone? Yes? I want to talk to you about Puerto Rico. 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: B: A: B: A: B: A:

Nice party. Yes. Who do you know here? Bill. I work with him. ah. Are you an accountant, too? No, I'm in public relations. 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 nonnative 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 included 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 nonnatives 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 repertoire for imitating informal repartee so common in native greetings. Videotaped role-plays revealed that this verbal formality among nonnatives 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:

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. A and B: (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 stressful moment by joking, but did not have the sociolinguistic control to do so successfully. Some subtle differences in the data showed approximation but not mastery of conventionalized language. An example of this was, "Hi. I do not know you. My name is (name)." Our native informants 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 transferred 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. AngloAmericans 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 interrupted. 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 beca~se 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 informants 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 potentially-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 perform 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 nonnative 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 substantial 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 performed. 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 sociolinguistic 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 dialogues 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: B: A: B:

Hi! Hi, how you doin'? Okay. You? Okay. Catch you later.

Native English speaker A: B: A: B:

Ah, hello, (name of boss). You wanted to see me? Yes. Come in. How are you? Fine, thanks. How are you? 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: B: A: B: A:

Great party! Yeah. The food is terrific. The Smith's always go all out. Oh? Is this an annual event? 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: B: A: B:

Hi. I never thought you have time for those romantic restaurants. Aha - Oh, I love to eat. I love to go to restaurants. Fancy seeing you here. I love eating, too. That's one of my main problems. You know that. Yes, I do. Oh, this is my husband. Jack, this is (name).

Cross-cultural realization of greetings in American English A: B: A: B: A: B: A:

105

It's nice meeting you. That's my boyfriend. Oh, nice meeting you. That's my boss. I see they're summoning us to the table. Enjoy your meal. Bye. Bye.

Japanese, Israeli women, non-native speakers A: B: A: B: A: B: A: B: A: B:

Excuse me, ma'am, but do you have the time? I see that you have the time. (Looking at watch on her own wrist) Why, it's broken. Oh ... It's broken. Well, it's seven past two. Thank you very much. Do you like the picture, Picasso? I like the picture, but it's not Picasso, I'm very sorry to tell you. It's Van Gogh. Oh, I didn't know that. There is a little difference between them. I didn't recognize this picture. I like the picture of impressionist. 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: B: A: B: A: B: A: B: A: B: A: B:

Hi, Steve. How are you doing, Adina? Okay. What's happening? Vh ... vacation was really good. Oh, yeah? Where'd you go? Vh, skiing. Lucky you! Yeah, and you? Just stayed in town, bored to tears. That's too bad. 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 nonnative 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 awareness 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)fGayle 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 appropriate 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 studied 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 appropriate, 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. Seventyfive 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, relationships. 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~addam ('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 frequently 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) attributes 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 sociocultural 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 compliment. 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 expression 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 comparatives, 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 akhlaa~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 consequences.'

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 'a~la wi Hakiima wi razina wi zakiyya wi labiqa wi mish na~Sik 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: el-Ha~ii~a tasriHtik aHla. (EF30) A: 'Your dress is very nice.' B: 'Really, your hair style is nicer.'

(11)

amiiSak Hilw zayy bitaa~ MiHammad Fu~aad. (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 (14)

117

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 compliments 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 compliments. 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 different 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 haircuts, eyes, and clothing. "Skill/work" referred to the "quality of something 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 intelligence. 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*

Traits

SkillslWork

Other

Total

20

7

3

60

5%

100%

(15)

50%*

33%

12%

(250/0)

American Appearance (Natural) 26*

Traits

SkillslWork

Other

Total

6

28

o

60

10%

47%

0%

100%

(1) 43%** (1.6%)

* number of occurrences out of 60 compliments ** percentage (based on 60 compliments)

However, half of our Egyptian compliments on appearance praised natural attributes. Some of the compliments are exemplified in (15) and (16). (15)

sha'rik gamiil. (EF25) 'Your hair is beautiful.'

(16)

~ineeki 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 complimented 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 ~a'la, wi zakiyya wi diblumasiyya. (EF8) 'You are mature and intelligent and diplomatic.'

120

(19)

Gayle L. Nelson - Waguida EI Bakary - Mahmoud Al Batal

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 appearance. 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.

121

Egyptian and American compliments

Table 2. Attributes praised according to nationality and gender of giver and receiver of compliments Egyptian Females Appearance

Traits

SkillslWork

Other

Total 16 (690/0)

Female to Female

8 (35%)

4 (180/0)

3 (13%)

1 (4%)

Female to Male

6 (260/0)

1 (40/0)

0

(00/0)

o (00/0)

14 (61 %)

5 (22%)

3 (13%)

1 (4%)

23 (1000/0)

Male to Female

12 (32 %)

8 (220/0)

0

(0%)

2 (5%)

22

(59%)

Male to Male

4 (11 %)

7 (19%)

4 (11 %)

0(0%)

15

(41 %)

16 (430/0)

15 (41 0/0)

4 (11 0/0)

2 (50/0)

37 (100%)

Appearance

Traits

SkillslWork

Other

Total

14 (54%)

0

(0%)

4 (15%)

0(0%)

18

(80/0)

2

(8%)

4 (15%)

o (00/0)

8 (31 %)

16 (620/0)

2

(8%)

8 (300/0)

0(0%)

26 (100%)

Male to Female

6 (18%)

2

(6%)

6 (180/0)

o (00/0)

14 (42%)

Male to Male

4 (120/0)

2

(6%)

14 (41 0/0)

o (00/0)

20

4 (12%)

20 (590/0)

o (00/0)

34 (100%)

Total

7

(31%)

Males

Total American Females

Female to Female Female to Male Total

2

(690/0)

Males

Total

10 (290/0)

(59%)

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 competence 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 miscommunication and should therefore be focused on" (Wolfson 1989: 140). The results of this study provide linguistic and sociolinguistic information 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 sociolinguistic 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 sociolinguistic 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 difficulty 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 complimenting, 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 compliment 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 consistent 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 attributes 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 compliment 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 breakdown. Let's suppose, for example, that an American student or teacher, in an attempt to be friendly and open a conversation, repeatedly compliments an Egyptian in class. If the Egyptian student perceives the compliment 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 misunderstandings 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 questions also exist - How does power distance and status interact with complimenting behavior? How does familiarity interact with complimenting? What are common responses to compliments? In addition to the above questions, two other facets of Arabic complimenting 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 ethnographic 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 overview", in: Shoshana Blum-Kulka - Juliane House - Gabriele Kasper (eds.), 1-34.

Egyptian and American compliments

127

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. 1974 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 Communication 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 "The influence of the Arabic language on the psychology of the Arabs", Middle East Journal 5: 284-302.

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 inappropriate 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 equivalents 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 theoretical 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: Sue: John: Sue:

Ya got any widgets? Yeah. Why don~t you give me three? 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 someone 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 context, 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 lightbulbs 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. b. c. d. e. f.

I'll take three. Give me three. Why don't you give me three? How about three? Could I have three? 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 egocentric 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 therefore 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~ll 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~as 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 don~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 construction 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 languages 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 question (What is the capital of Ohio?) or making a straight-forward assertion (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 conditions 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 interactions 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) propositional 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 "satisfaction" 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 condition 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, including, in particular, what we might call the interactional effects of the acts. Brown and Levinson (1978) have argued that people have two facewants, 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 threatening (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 conditions 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 C1 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

Satisfaction Condition: Resp is able to do A Subcondition: Res!) Possess X Resp is :J:illing to do A Domain (A):' Resp transfers X to Init Domain (X): ..... 12

137

C2 C3 C4

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 ~tructure. 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 initialstate 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

2)

3) B. 1)

2)

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. Resp-oriented. These are utterances that instantiate either the willingness condition or the ability condition. Respwillingness 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 respwillingness 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?) Domain-oriented. This is for utterances that address condition C 3 such as Do you have a corner piece? Power and Deference Features: 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. 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, addresseeoriented 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 morphosyntactically 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 Englishspeaking 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 nonverbal 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 approximately 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

Age Range

2 FF 2MM 6FM 18-69

2FF 2MM 6FM 20-22

3FF 2MM 5FM 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) Fl: F2: Fl: F2: Fl: F2: Fl:

No Barrier Will you hand me that piece that's closest to you? ((Points)) Right there. Yeah ((Laughs and passes piece)) Let's see here. And that corner piece. ((Points)) ((Picks up piece)) 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 condition 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 politeness 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 (C2 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 piece/J~ai besoin d~un 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 responder 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 English B

NB

A B.l.a B.l.b B.2.a B.2.b

41 0 1 0 0

39 0 20 2 0

Totals

42

61

French

French Learners

B

NB

T(%)

B

NB

T(%)

80 (32.7) o (0) 21 (8.6) 2 (0.8) o (0)

40 0 0 1 0

28 0 1 0 0

68 (28) o (0) 1 (0.4) 1 (0.4) o (0)

57 0 0 0 0

11 0 9 0 0

68 (28) o (0) 9 (3.7) o (0) o (0)

103 (42.1)

41

29

70 (28.8)

57

20

77 (31.7)

T(%)

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 objectspecific, 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

A.l A.2 B.l B.2

0 0 0 0

1 2 0 5

1 (0.4) 2 (0.8) o (0) 5 (2)

2 0 0 3

24 7 0 9

26 (10.7) 7 (2.9) 0(0) 12 (4.9)

0 0 0 0

6 1 0 7

6 (2.5) 1 (0.4) o (0) 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

T(%)

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 condition 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 presupposition. 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 connection 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(%)

9

53 (21.8) 56 (23) 9 (3.7)

86 1 0

46 1 1

132 (54.3) 2 (0.8) 1 (0.4)

45

118 (48.5)

87

48

135 (55.5)

B

NB

T(%)

B

NB

A.1 A.2 B

51 0 0

17 8 0

68 (27.8) 8 (3.3) 0 (0)

41 32 0

12 24

Totals

51

25

76 (31.1)

73

T(%)

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 Learners

French Natives B

NB

Est-ce que tu as ... Tu as ... As-tu ...

43 33 1

20 36 0

Totals

77

56

B = Barrier; NB

T (%)

T (%)

B

NB

63 (47.4) 69 (51.9) 1 (0.8)

50 3 39

42 2 20

92 (59.0) 5 (3.2) 59 (37.8)

133 (100)

92

64

156 (100)

= 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 contexts 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 nonetheless 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 contextbound (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 contexts, 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 intonation 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 intonation 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 presupposition-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

11.

12.

13.

14. 15.

16. 17. 18. 19. 20.

21.

151

the factors that distinguish them are invariably social or psychological, rather 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. 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 theoretically 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. 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) 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. The examples of (7) barely touch the surface of the problem. 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. Only 5 FM pairs in the learner group participated in the entire experiment. Thus, an additional pair of FF had to be used. The percentages given in the tables below are based on 245 English, 243 French, and 245 French learner request initiations. In the tables that follow, we shall use the roman numerals employed in this classification to identify forms. The deference/no-deference contrast does not apply to init-specific utterances, for we never make requests saying, Do I want a hot chocolate? 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. 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 overview", 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, Fran~oise 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 fran~ais", 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 - Steven Ross

Naoko Yoshinaga -

Gabriele Kasper

1. Pragmatic transfer The influence of non-native language users' linguistic and cultural background 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 perceptions 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 context (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' selfreports, 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 communicative 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 (BlumKulka 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 correlated 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 perceptions 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, irrespective 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 contextinternal 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 apologizing, 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 opposite 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) 2) 3) 4)

30 Japanese learners of English (Intermediate) (JEI) 30 Japanese learners of English (Advanced) (JEA) 30 Native speakers of English (E) ,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 (average 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 interlocutor 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 expecting me"), problematizing a precondition ("we weren't supposed to meet before 12"), or denial ("I didn't do it"); (b) utterance reducing 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 Chisquare 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 agreement in the two groups' perception of status, obligation to apologize, and likelihood of apology acceptance. On each of these factors, only two contexts 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; Takahashi - Beebe 1993) consistently demonstrated a much stronger differentiation 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 contexts), 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 Context

Damaged Car Ruined Mag. Contradiction Poor Teacher Pers.Mess.L-H Pers.Mess.H-L Bus.Mess L-H Bus.Mess H-L Order Change Food on Cust. Wrong Order Food on Wait. Messed Up Bag Bonsai Tree Broken Statue Customs Form Ungraded Pap. Borrowed Book Failed Student Cheating Student

Social Distance

Social Power

Severity

+

+ + + + + + + +

-

+ +

+ +

-

+ + +

-

+ + + + + + + +

+

+ + + + + + + + + +

+ + + + + + + + + + + +

-

+ +

+ + + +

-

+ +

+

+

+

-

-

+

-

+

-

+ = same rating by J and E - = different rating by J and E off. face-loss = offender's face-loss

Obligation

0\ 0\

Off. Face-Loss

H'S Face-Loss

Transfer Prediction

+ + + + + + + +

-

-

fZJ

+ -

-

fZJ

+ +

positive positive positive positive positive

-

+ + + + + +

Likelihood

+

+ + + + + + + + +

+ + +

-

+ -

-

~ ~

~ ~

fZJ

+ + + +

positive positive positive

-

fZJ

+ +

+ + + +

-

-

positive positive positive positive negative

+ +

-

fZJ

+

positive

-

~

fZJ

;::s-~ ~

I ~

0
A JEA > A

JEI < A

JEI < A

172 (31)

N. Maeshiba -N. Yoshinaga - G. Kasper - S. Ross

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 upgraded apology since its literal meaning is 'inexcusable'. However, mooshiwakearimasen is the appropriate form for a waiter at an expensive 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-upgraded form. (32)

J: Mooshiwakearimasen, okyakusama. "It sir/ma'am" . JEI: Oh, I'm sorry. A & JEA: Oh no! I'm so incredibly sorry!

IS

inexcusable,

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 intermediate 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 statement 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 perception, 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 interlocutors' power relationship indicates that the direction of the status differential 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 selection of both strategies contingent on the direction of the status relationship. 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 statusdifferential 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 statusdifferentiating 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 mitigation to their performance of face-threatening acts in English, the intermediate 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 transgressions 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 crosscultural 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 situation"/downgrading, "saying directly 'I am very sorry'''/IFID and upgrading, "offering to do something for the other person"/ repair, "apologizing 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 transgression 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 transgressions. 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 awareness 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 inclined 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 construction 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 different 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 intermediate 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 intermediate 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 wherewithal 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 oversupplying 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 communities. 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 appropriate 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 investigation 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 A45 yrs

1 (0.5%) 1 (3.9 0/0) 0 0

5 (2.5%) 0 2 (3.9%) 0

14 (7%) 3 (11.5 %) 4 (7.7%) 0

80 22 46 22

Total

2 (1%)

7 (3.5%)

21 (10.50/0)

170 (85%)

X2= 10.03; df = 9; P = 0.35

Total (40%) (84.6%) (88.5%) (100%)

100 26 52 22

(500/0) (13%) (260/0) (11 %)

200 (100%)

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, 1 have 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 Non-Student

41 (41 0/0) 48 (48 %)

41 (41 0/0) 32 (32 %)

12 (12 %) 16(16%)

6 (60/0) 4 (4%)

100 (500/0) 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

Female Male

64 (50.8 %) 25 (33.8 %)

38 (30.2 %) 35 (47.3 %)

18 (14.2 %) 10 (13.5 %)

Total

89 (44.5%)

73 (36.5%)

28 (14%)

X2= 6.84; df = 3; p = 0.08

GI

Total

6 (4.8%) 126 (63%) 4 (5.4%) 74 (37%) 10 (5%)

200 (100%)

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

51

LI

NI Students < 32 yrs 32-45 yrs >45 yrs

41 13 24 11

(41 %) (500/0) (46.2 %) (50%)

Total

89 (44.5%)

41 5 18 9

(41 %) (19.2 %) (34.6 0/0) (40.9%)

73 (36.5 0/0)

12 (12 %) 5 (19.20/0) 9(17.3%) 2 (9.1 %) 28 (14%)

GI

Total

100(50%) 6 (60/0) 3 (11.6%) 26 (13%) 52 (26%) 1 (1.9%) 22 (11 %) 0 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 unimposing 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 imposed 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 outrageous!" 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, nonnative learners often know parts of social routines, but fail in the overall delivery of them which may result in an undesirable impact on an addressee. 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 conversational 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 interpreted 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 ~an 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 suggestions. 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 understands 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 appropriate 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 questionnaire 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 acquisition. 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 constraints. 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 comprehension 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, however, 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 information value, not all of the elements are equally used during on-line processing. 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)

# 2 Has pensado en leer este libro?

'Have you thought about reading this book?' (7)

Should you read this book?

Transfer of pragmatic competence

(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?

261

'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 interrogative forms were employed, four of which are negated, and one includes 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 possibility 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 nonnegated 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 specific 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 knowledge 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, relying 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 miscommunication 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 miscommunication. (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 audiocassette 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 expressions. 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 suggestion 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

265

Transfer of pragmatic competence

back to the speaker. Second, informants were to identify the type of speech act expressed (e. g., suggestion, apology) and, if possible, to reproduce 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 unfriendly/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 47 56

28 21 53

37 29 50

22 21 62

17 6 3

43 32 59

26 29 68

44

33

39

33

10

44

40

(df = 2; Sum of Squares n sizes: First year = 46 Second year = 34 Advanced = 34

= 56.37; Mean Square = 28.18; F = 14.65; P