Style Shifting in Japanese (Pragmatics and Beyond New Series)

Style Shifting in Japanese

Pragmatics & Beyond New Series (P&BNS) Pragmatics & Beyond New Series is a continuation of Pragmatics & Beyond and its Companion Series. The New Series offers a selection of high quality work covering the full richness of Pragmatics as an interdisciplinary field, within language sciences.

Editor Andreas H. Jucker

University of Zurich, English Department Plattenstrasse 47, CH-8032 Zurich, Switzerland e-mail: [email protected]

Associate Editors Jacob L. Mey

University of Southern Denmark

Herman Parret

Jef Verschueren

Susan C. Herring

Emanuel A. Schegloff

Belgian National Science Foundation, Universities of Louvain and Antwerp

Belgian National Science Foundation, University of Antwerp

Editorial Board Shoshana Blum-Kulka Hebrew University of Jerusalem

Jean Caron

Université de Poitiers

Indiana University

Masako K. Hiraga

St.Paul’s (Rikkyo) University

University of California at Los Angeles

Deborah Schiffrin

David Holdcroft

Georgetown University

Sachiko Ide

Kobe City University of Foreign Studies

Sandra A. Thompson

Thorstein Fretheim

Catherine KerbratOrecchioni

John C. Heritage

Claudia de Lemos

Teun A. van Dijk

Marina Sbisà

Richard J. Watts

Robyn Carston

University College London

Bruce Fraser

Boston University University of Trondheim University of California at Los Angeles

University of Leeds Japan Women’s University

University of Lyon 2 University of Campinas, Brazil University of Trieste

Volume 180 Style Shifting in Japanese Edited by Kimberly Jones and Tsuyoshi Ono

Paul Osamu Takahara

University of California at Santa Barbara Pompeu Fabra, Barcelona University of Berne

Style Shifting in Japanese Edited by

Kimberly Jones University of Arizona

Tsuyoshi Ono University of Alberta

John Benjamins Publishing Company Amsterdam / Philadelphia

8

TM

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences – Permanence of Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data Style shifting in Japanese / edited by Kimberly Jones, Tsuyoshi Ono. p. cm. (Pragmatics & Beyond New Series, issn 0922-842X ; v. 180) Includes bibliographical references and index. 1.  Japanese language--Style.  I. Jones, Kimberly. II. Ono, Tsuyoshi. PL635.S79

2008

495.6--dc22 isbn 978 90 272 5425 2 (Hb; alk. paper)

2008031616

© 2008 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa

Table of contents Acknowledgements

vii

The messy reality of style shifting Kimberly Jones and Tsuyoshi Ono

1

Style shifts in Japanese academic consultations Haruko Minegishi Cook

9

Interpersonal functions of style shift: The use of plain and masu forms in faculty 39 meetings Naomi Geyer Speech style shift as an interactional discourse strategy: The use and non-use of desu/-masu in Japanese conversational interviews Shoko Ikuta

71

Playing with multiple voices: Emotivity and creativity in Japanese style mixture Senko K. Maynard

91

Riyuu ‘reason’ for nai desu and other semi-polite forms Mutsuko Endo Hudson

131

Masen or nai desu – That is the question: A case study into Japanese conversa161 tional discourse Satoshi Uehara and Etsuko Fukushima The power of femininity: Can Japanese gender variation signify contradictory social meanings? Yuka Matsugu Tuning speech style and persona Yoshiko Matsumoto

185

213



Style Shifting in Japanese

Speech style and the use of regional (Yamaguchi) and Standard Japanese in conversations Shigeko Okamoto

229

“Involved” speech style and deictic management of spatio-temporal and textual 251 reference: A case of ko/so-deictics in Japanese Kuniyoshi Kataoka Variation in prosodic focus of the Japanese negative nai: Issues of language specificity, interactive style, and social situations Shoji Takano

285

Name index

329

Subject index

333

Acknowledgements A collaborative book such as this leaves its editors indebted to many people. We are particularly grateful to the several organizations that funded the Japanese Speech Style Shift Symposium, thus enabling us to bring together a number of pioneering researchers in this area: Japan Foundation, National Science Foundation, and the Office of the Vice-President for Research and the College of Humanities at the University of Arizona. Many people assisted us, first as we organized and implemented the symposium, then later as we prepared this volume. With apologies to anyone we have overlooked, we would like to thank a few of the most significant supporters by name. Brian McKnight was the head of the East Asian Studies department when we first began planning the symposium, and we are grateful for his understanding of the potential significance of this event and for his support as we searched for funding. When Tim Vance later became department head, his support was equally invaluable as we organized and hosted the symposium. Kuniyoshi Kataoka and Junko Mori worked with us as co-organizers, and Yuka Matsugu and Takako Ogawa served as research assistants for the symposium. We would also like to thank all of the participants in the original symposium, both those who contributed to this volume and those who did not; the students who took the style shift seminar we taught that semester, and who by their thoughtful comments and questions helped us prepare intellectually; and the East Asian Studies staff members who helped with a myriad of organizational details both major and minor, Brenda Fraker and Janet Kania. We are especially grateful to an anonymous reviewer for this series who gave us the benefit of an extremely thorough and balanced reading of the manuscript we submitted, offering many valuable suggestions that greatly improved the final product. Naoko Witzel’s editorial assistance was also invaluable; she read the entire book not once, but at two different points during the editorial process, and helped us weed out many errors and inconsistencies. Andreas Jucker was encouraging and supportive as we prepared the manuscript. Of course, we take responsibility for any remaining shortcomings. Finally, we thank all of the contributors to this volume for their thought-provoking chapters and for their patience (and occasional gentle nudges) as we struggled to keep the book from becoming sidetracked by our various other scholarly and administrative responsibilities. It is thanks to them most of all that our dream of a book that would bring together multiple perspectives on the phenomenon of style shift in Japanese has now become a reality.

The messy reality of style shifting Kimberly Jones and Tsuyoshi Ono

University of Arizona, University of Alberta

The Japanese language is renowned for having a variety of styles of speech available to speakers. Among the many examples of different styles in Japanese are gendered forms of language, regional dialects, and the use or non-use of forms traditionally treated as “polite” forms, such as various types of honorifics. Traditionally, speakers’ and writers’ use of different styles has been explained, both by linguists and by teachers of Japanese, as a way of expressing various social distinctions, such as the relative rank of two or more interlocutors, the degree of social distance between them, and “formal” versus “informal” contexts. These explanations of various styles, which generally stem from speakers’ intuitions about how they use the language, are unlikely to be contradicted by approaches to linguistics or language pedagogy that rely on constructed data and intuition. However, recent decades have seen more and more linguists examining naturally occurring speech and written texts in order to further our understanding of the structure and the use of language. Approaches such as conversation analysis, discourse analysis, linguistic anthropology, functional linguistics, and sociolinguistics have provided us with discoveries which could not have come from working solely with constructed data. Once researchers began examining examples of actual language use, it became obvious that style shifting was not at all uncommon. By style shifting, we mean the use of two or more styles, even ostensibly mutually exclusive styles, within a single speech event or written text. The factors traditionally considered to result in the use of a particular style, such as the formality of the occasion or text, the relative rank of the interlocutors (or authors and readers), and the overall social distance between them are unlikely to change significantly over the course of a single interaction or within a single written text. Thus, researchers have had to look for other explanations of why different styles are employed at different points. In addition to an ever-increasing interest in studying language as it is actually used, recent years have also seen a growing recognition of the mutually constitutive nature of language and context, particularly in the case of spoken language (e.g., Bachnik and Quinn 1994, Duranti and Goodwin 1992). Blanket assertions that a given context makes a given style appropriate are no longer adequate. Instead, we now realize



Kimberly Jones and Tsuyoshi Ono

that speakers are actively shaping context by their use of particular styles of language at particular points in their conversations. Proposed explanations of why speakers might shift between “polite” and “plain” styles include the idea that style shifting may be a way to convey how closely connected different parts of a conversation are to each other and/or to negotiate affect or psychological distance at a given moment (Cook 1998, Ikuta 1983, Makino 2000, Maynard 1991, Sukle 1994, Yamazaki 2000). Wetzel (1988) has looked at how speakers shift between different verbs of giving and receiving in order to signal a shift of perspective, identifying more closely first with one, then with another, group of participants in an interaction. Other areas that have been explored include shifting between a regional dialect and standard Japanese (Okamoto 1998) and gender- and age-related social dialects (Inoue 1998, Matsumoto 1999, 2001; Okamoto 1996; Okamoto and Smith 2004). Although not a “proceedings” volume, this book has its roots in the Japanese Speech Style Shift Symposium held at the University of Arizona in 2002. That symposium, made possible by funding from Japan Foundation, National Science Foundation, the University of Arizona Office of the Vice-President for Research, and the University of Arizona College of Humanities, brought together many of the pioneering researchers who have looked at the phenomenon of style shifting in spoken and written Japanese. Some of the papers in this book were originally presented at the symposium and have since been rewritten for publication. Others are new papers by symposium participants, and still others were contributed by scholars in the field who did not attend the symposium. The authors represent a diverse range of approaches to the study of language, including conversation analysis, discourse analysis, cognitive linguistics, functional linguistics, linguistic anthropology, and sociolinguistics. Though diverse in approach, all share the belief that it is valuable to study language in the contexts in which it is used. For spoken language, this necessitates analyzing recordings of naturally occurring conversation or other types of talk. Another common thread running through these chapters is the understanding that language use is inextricably linked to both context and language structure, in mutually constitutive relationships. Given the culturally rooted nature of style shifting, it is critical to examine this phenomenon within a single culture. And because of the acknowledged variety of different styles of speech that are available to speakers of Japanese, the Japanese language is a particularly fruitful one within which to examine style shifting. Given the newness of this area of research, there is still much work to be done. Nevertheless, Japanese is almost certainly the best-studied language as far as style shifting is concerned. (Examples of similar research for other languages can be seen in Iwasaki and Horie 2000, Kulick 1992, and McElhinny 1992, among others.) The first several chapters of this book all consider shifting between so-called “polite” predicates ending in desu or -masu, and “plain” (non-desu/masu) predicates. Haruko Minegishi Cook examines interactions between college students and their professors and considers how the participants negotiate their social identities, collaboratively constructing multiple social personae through their strategic use of these



The messy reality of style shifting

forms. Professors alternate between using desu/-masu predicates, thus constructing a professional and egalitarian social relationship, and using plain predicates, thus creating a more casual mood and/or presenting the professor in the role of a friendly and supportive “coach.” Students, on the other hand, tend to speak to their professors using desu/-masu forms, but Cook also describes contexts in which they do not do so. Students contribute to a hierarchical relationship if they respond to their professors’ plain-style utterances by themselves using desu/-masu. However, at times they avoid the appearance of hierarchy by using sentence fragments or by co-constructing sentences, embedding their professors’ plain-style utterances in a sentence with a desu/masu predicate. Naomi Geyer first discusses certain “core properties” of desu/-masu forms and plain forms. She argues that the core meanings of desu/-masu forms are deference, formality, and the presentation of a public self, while plain forms are used to show the lack of such features. Geyer then looks at the use of these two forms in faculty meetings, including cases where participants shift between the two. When used in specific contexts, the basic meanings of the two forms can be used to convey a variety of different pragmatic effects, some of which might seem to be rather at odds with their core meanings. Thus, desu/-masu forms can be used not only to display deference or an impersonal and official frame, but also, in certain contexts, to enhance solidarity between the participants. On the other hand, the use of plain forms can enhance solidarity by showing that the speaker does not feel it necessary to express deference or formality, but those same plain forms can also be used to accomplish various types of mitigation, such as mitigating disagreement or hedging. Geyer shows that a detailed analysis of how the forms are used along with other factors in specific contexts is necessary for understanding the diversity of their potential functions. Similarly, Shoko Ikuta also emphasizes the multi-functionality of desu/-masu forms and plain forms. In her analysis of talk during interviews, she focuses on how speakers shift between these forms for discourse organization purposes. Shifting to the plain form can mark a more embedded section, or a “subspace,” of the discourse. In doing this discourse organization work, participants shape and reinforce different personae for themselves, making it clear that the interviewer is the emcee and prompter, and the interviewee is the main speaker and is responsible for providing information. By shifting to plain forms at certain points in the discourse, interviewers show that their contributions are to be considered subsidiary and supportive utterances which complement, but do not usurp, the interviewees’ turns at talk. Ikuta’s chapter is similar to Cook’s in her emphasis on how speakers use styles – and especially, style shifting – to portray particular social personae. By showing that shifting to the plain form can, seemingly paradoxically, be used as a politeness strategy, Ikuta also mirrors Geyer in showing the limitations of taking an overly simplistic view of the meaning of a particular style and failing to consider how it is used in particular contexts. In the final chapter dealing with shifting between -desu/masu and plain forms, Senko K. Maynard extends the consideration of this type of style shifting to written





Kimberly Jones and Tsuyoshi Ono

discourse. Maynard analyzes a variety of contemporary written genres: essays, fiction, nonfiction, and internet bulletin board entries. She looks at authors’ incorporation of features characteristic of spoken language, focusing particularly on the use of the [DQ + to yuu + N] structure, in which a noun is modified by a clause comprised of a direct quotation followed by the quotative marker to yuu. Maynard shows that, much as speakers use style shifting to manipulate different social personae, authors use shifting to manipulate different voices. A given text typically has what Maynard terms a “base-line style,” and when the author shifts from that style, the reader seeks a motivation for the shift. Style shifting in these written texts is shown to be a creative and poetic strategy, one which authors use to express and communicate a variety of emotional attitudes. The authors of the next two chapters consider the difference between different types of desu/-masu style forms – what Mutsuko Endo Hudson terms the “polite” and the “semi-polite” styles. Using a recent novel as data, and focusing especially on polite and semi-polite negative forms (-masen and -nai desu, respectively), Hudson argues that the use of the semi-polite style has two functions – to signal a certain psychological distance between speaker and hearer and to imply that the speaker is explaining or assessing a situation. Satoshi Uehara and Etsuko Fukushima take up the same issue and, again focusing on shifting between -masen and -nai desu, examine the alternation between these forms in spoken data from first-encounter conversations. They find that -masen forms dominate at the beginning of the conversations and after breaks in the conversations, or at other times when the speakers are foregrounding “politeness.” On the other hand, -nai desu forms are used at other points in the conversation, as a way of expressing a growing familiarity between the speakers. Gendered styles are the topic of the next two chapters. Yuka Matsugu challenges the idea that “women’s language” necessarily and only indexes softness, powerlessness, and politeness, in contrast to “men’s language,” which is said to index qualities such as assertiveness, power, and directness. These binary and stereotypical distinctions fail to adequately capture the totality of what is going on when women shift styles and use forms that are considered to be extremely feminine. Analyzing situations in which women use such forms to convey displeasure or frustration, Matsugu suggests that the speakers are invoking their status as fully adult women to convey a powerful and authoritative social stance. Based on this, she argues that particular forms can convey multiple social meanings, some overt and some covert. Yoshiko Matsumoto also looks at female speakers’ use of gendered forms, examining the speech of middle-aged, middle-class stay-at-home mothers, who might be thought to be the quintessential users of so-called “Japanese women’s language.” Nevertheless, these speakers too shift between styles, using neutral and even more masculine styles at times. Matsumoto suggests that style can be employed to index a social identity such as gender, a speaker’s stance and attitude at a particular time, or a combination of the two. That is, speakers may choose expressions (and listeners may interpret the expressions used) as demonstrating an affiliation with a particular gender, as



The messy reality of style shifting

expressing oneself in a more forceful or delicate way, or as simultaneously achieving both types of goals at once. As Matsumoto puts it, “it is such layers of interpretations available in expressions that provide speakers with resources for creating their own personae through shifts in styles in conversation.” Shigeko Okamoto also wrestles with the issue of how particular forms may index a multiplicity of meanings, some directly and some indirectly. Analyzing shifting between Standard Japanese and Yamaguchi dialect, she finds that the two dialects are not used as separate systems, but are mixed in varying proportions in different speech situations. The use of different forms may index either affiliation with a particular dialect or how formal a tone the speaker wants to take, or perhaps both. Different ratios of mixing allow for “fine-tuning” of the effect that a speaker’s utterances create. The final two chapters of the book push research on Japanese style shifting in new directions. Kuniyoshi Kataoka looks at the role that shifting plays in organizing the discourse and creating a sense of involvement in monologic narratives by rock climbers describing climbing routes. The specific forms he analyses are ko-series, or proximal, deictic expressions and so-series, or distal, deictic expressions. Kataoka finds that so-series deictics are the preferred choice for most of the narrators in his study, while ko-series deictics are preferred when speakers are referring to the peak of involvement. This often comes when they describe the most strenuous or technical parts of the climb or particularly notable holds and moves on the route. For two of the participants in his study, this pattern was reversed, with -ko-series deictics being the commonly used forms, and -so-series deictics used at more salient or highly involved sections of the discourse. Though the former pattern is more common, and may be more natural, Kataoka argues that it is the fact that a shift occurs that is most relevant in the discourse, rather than the direction of that shift. Finally, Shoji Takano calls for research in style shifting to move beyond a consideration of shifting between various linguistic forms to a consideration of paralinguistic features such as prosody. In his look at how negation is prosodically realized in various Japanese contexts – casual conversation, televised debates, and the reading of television news – Takano captures the complex interplay of various linguistic and social factors and their influence on the expression of negation. He convincingly argues for the importance of looking at language-specific factors, and also of looking at additional genres beyond the casual conversation on which so much work is based. Takano suggests that Japanese style shifting is multi-layered, and that in addition to the morphosyntactic variation that has primarily been the focus of scholars’ attention so far, prosodic variation is another important resource that speakers use to express their thoughts and feelings and portray a variety of stances. It has been a pleasure to be able to learn from the insights in such a diverse group of papers as we worked to make them available in one volume. We hope that bringing together the variety of perspectives and methodologies represented here and taking a detailed look at the phenomenon of style shifting in one specific language will advance our understanding of the complex phenomenon of style shifting not only in Japanese,





Kimberly Jones and Tsuyoshi Ono

but also more generally, contributing to our growing understanding of various types of speech style shifting, some of the factors that influence shifting, and theoretical and methodological approaches to understanding this phenomenon.

References Bachnik, J. M. and Quinn, Jr., C. J. (eds) 1994. Situated Meaning: Inside and Outside in Japanese Self, Society, and Language. New Jersey: Princeton University Press. Cook, H. M. 1998. “Situational meanings of Japanese social deixis: The mixed use of the masu and plain forms.” Journal of Linguistic Anthropology 8 (1):87–110. Duranti, A. and Goodwin, C. (eds) 1992. Rethinking Context: Language as an Interactive Phenomenon. New York: Cambridge University Press. Ikuta, S. 1983. “Speech level shift and conversational strategy in Japanese discourse.” Language Sciences 5:37–53. Inoue, F. 1998. Nihongo Wotchingu (Japanese-Watching). Tokyo: Iwanami. Iwasaki, S. and Horie, P.I. 2000. “Creating speech register in Thai conversation.” Language in Society 29 (4): 519–554. Kulick, D. 1992. “Anger, gender, language shift and the politics of revelation in a Papua New Guinean village.” Pragmatics 2: 281–296. Makino, S. 2002. “When does communication turn mentally inward? – A case study of Japanese formal-to-informal switching.” In Japanese/Korean Linguistics 10, N. M. Akatsuka and S. Strauss (eds), 121–131. Stanford: CSLI Publications. Matsumoto, Y. 1999. “Sedai to kotoba no sentaku (Stylistic choices in Japanese across generations).” In Jitsuyooteki Gengoriron no Koochiku o Mezashite (Linguistics and Japanese Language Education), Y. Alam (ed.), 33–48. Tokyo: Kurosio. Matsumoto, Y. 2001. “Gender identity and the presentation of self in Japanese.” In Gendered Practices in Language, S. Benor, M. Rose, D. Sharma, J. Sweetland and Q. Zhang (eds), 339–354. Palo Alto, CA: Center for the Study of Language and Information Publications. Maynard, S. 1991. “Pragmatics of discourse modality: A case of da and desu/masu forms in Japanese.” Journal of Pragmatics 15: 551–582. McElhinny, B. 1992. “‘I don’t smile much anymore’: Affect, gender, and the discourse of Pittsburgh police officers.” In Locating Power: Proceedings of the Second Berkeley Women and Language Conference, K. Hall, M. Bucholtz, and B. Moonwomon (eds), 386–403. Berkeley: Berkeley Women and Language Group. Okamoto, S. 1996. “Indexical meaning, linguistic ideology, and Japanese women’s speech.” Proceedings of the 1996 Annual Meeting of the Berkeley Linguistics Society: 290–301. Okamoto, S. 1998. “The use and non-use of honorifics in sales talk in Kyoto and Osaka: Are they rude or friendly?” Japanese/Korean Linguistics 7: 141–157. Stanford: Center for the Study of Language and Information. Okamoto, S. and Smith, J. S. S. (eds) 2004. Japanese Language, Gender, and Ideology: Cultural Models and Real People. New York: Oxford University Press. Sukle, R. 1994. “Uchi/soto: Choice in directive speech acts in Japanese.” In Situated Meaning: Inside and Outside in Japanese Self, Society, and Language, J. M. Bachnik and C. J. Quinn Jr. (eds), 113–142. NJ: Princeton University Press.



The messy reality of style shifting Wetzel, P. 1988. “Japanese social deixis and discourse phenomena.” Journal of the Association of Teachers of Japanese 22 (1):7–27. Yamazaki, M. 2000. “Carving a ‘playful space’ out of ‘honorifics’: A case study of the Japanese desu/masu predicate-final form in a ‘Driver-Navigator’ discourse.” Unpublished manuscript.



Style shifts in Japanese academic consultations Haruko Minegishi Cook

University of Hawaii at Manoa

1. Introduction: style shifts The question of why speakers shift styles has been one of the central issues in sociolinguistics (e.g., Eckert and Rickford 2001). It is universal that language has stylistic variations (e.g., regional dialects, social dialects, registers, and polite or rude ways of speaking), and that no speaker of a language speaks only in a single speech style. In other words, language has many ways of saying “the same thing.” For example, a pair of utterances said in the standard language and a regional dialect containing the same propositional content do not mean the same thing in terms of social meaning. Sociolinguists have studied when and why speakers shift speech styles and have provided cognitive and social explanations. There are three major theoretical positions with regard to style shifts: the Labovian cognitive view, the view that attributes shifts to external social factors, and social constructivism. Labov’s cognitive view considers “demographic” aspects of speakers’ identities (e.g., sex, age, and socio-economic class) to be correlated with speech variants in a fairly static way. Labov (1966) claims that since linguistic variables are correlated to social identities, speakers can consciously manage to control their social identities if they are aware of how they speak. Thus, style shifts are triggered by speakers’ relative attention to speech. In order to elicit different speech styles from his informants during his sociolinguistic interviews, Labov used a range of contexts from word lists to a near-death story. The underlying assumption is that more attention to speech produces careful speech whereas less attention produces casual speech. In contrast to the Labovian cognitive view, some sociolinguistic studies take a position that considers style shift as a social phenomenon. The underlying assumption shared among these studies is that speakers change their speech style in reaction to a change in some external social factors such as the social situation, the topic of talk, and/or the addressee/audience. For example, Blom and Gumperz (1972) found that between standard and rural varieties of Norwegian, code switching occurred both when the social situation changed (situational code-switching) and when topics changed (metaphorical code-switching). They reported that formal school lectures were given in the standard variety, but when the follow-up discussion took place,



Haruko Minegishi Cook

speakers shifted to the local variety. This is a case of situational code-switching, which is governed more or less by social norms. In metaphorical code-switching, the switch occurs when the topic of conversation shifts (Gumperz 1982). Even when there is no change in the social situation or topic of talk, speakers may shift styles in order to accommodate the addressee and/or audience. The notions of situational and metaphorical code-switches do not account for these cases. Studying the language of New Zealand’s radio news, Bell (1984) noted that the same newsreader used a non-standard variety on a local community radio station and the standard variety on the prestigious national radio station. Drawing on the speech accommodation theory in the field of social psychology (Giles and Hewstone 1975), Bell proposed that speakers shift styles so that the speech is more acceptable to the audience (i.e., “audience design”). Bell (1997: 243) states, “Speakers design their style primarily for and in response to their audience.” These studies attribute style shifts to the shifts of external social factors, whether it be the social situation, topic, or addressee/audience. In this view, external social factors influence language use. In contrast, social constructivist theory proposes that social categories such as social identities are not given a priori in the social structure but are created and negotiated in social interaction (Bucholtz 1999; He 1995; Ochs 1993; Sacks 1992; Schegloff 1991). Language is a tool available to the members of society for constructing their social identities. An individual has multi-faceted identities and the relevant identity is negotiated in interaction (Sacks 1972; Schegloff 1991). In this view, language use defines the social identity of the speaker, and speakers shift styles in order to index a shift in their social identities. Thus, through moment-by-moment interaction, speakers may shift from one social persona to another by shifting speech styles. Coupland (2001) refers to such shifts as “persona management.” Social identities are fluid, and speakers may display various social identities while interacting with the same addressee(s) in a given social situation. Taking the point of view of social constructivism, this paper discusses shifts between the Japanese so-called ‘addressee honorific’, masu form and its non-honorific counterpart, the plain form, in university academic consultation sessions. In most previous studies, shifts between these two linguistic forms have been directly linked to shifts in social situations and addressees. In particular, the shift to the masu form has been associated with a shift to a formal social situation and/or to a higher-status addressee. A conversation between a professor and a student has been depicted as a prototypical example of a status-different interaction, in which the student must give the masu form to the professor and receives the plain form (see Ide 1982). The present paper attempts to show that shifts between the two forms are not necessarily motivated by a change in the social situation or addressee, but by the participants’ displays of different social identities. In this paper, I adopt Ochs’ definition of social identity (1993: 288), which covers “a range of social personae, including social statuses, roles, positions, relationships, and institutional identities one may attempt to claim or assign in the course of social life…” An academic consultation session between a professor and



Style shifts in academic consultations

a student is ideal for proving this point because the social situations and addressees are held constant in this social event. The paper demonstrates that the masu and plain forms are linguistic resources available to the participants to construct their multiple social personae, or in Bakhtin’s (1981) term, “voices” in an academic discourse.

2. Shifts between the Japanese masu and plain forms 2.1

The masu and plain forms

Speakers of Japanese must choose between the honorific masu form and the non-honorific plain form at the sentence-final position. The term masu form includes the present (-masu) and past (-mashita) tense forms, the gerund form (-mashite), and the copular present (desu) and past tense (deshita) forms.1 The plain form includes the present (-u or -ru) and past (-ta) tense forms, the gerund (-te), and the copular present (da) and past tense (datta) forms, as well as bare nominals, which are considered to be derived by copula deletion (hon da/desu ‘is a book’ → hon Ø ‘is a book’). Table 1 summarizes the masu and plain forms. Table 1.  Three clause types and gerund in the masu and plain forms Clause type

Masu form

Plain form

Verbal

Verb stem -mas-u (present) Verb stem -mas-en (negative) Verb stem -mashi-ta (past) Verb stem -mashi-te (gerund) Adj + copula des-u (present) deshi-ta (past) deshi-te (gerund) Nom + copula des-u (present) deshi-ta (past) deshi-te (gerund)

Verb stem -(r)u (present) Verb stem -nai (negative) Verb stem -ta (past) Verb stem -(t)te (gerund)

Adjectival

Nominal

Adj stem -i (present) Adj stem -kat-ta (past) Adj stem -ku-te (gerund) Nom + copula da (present) dat-ta (past) de (gerund)

1. The tentative form of desu (deshoo) is not included in the analysis. Since the plain counterpart daroo sounds rough even in informal conversation, it is typically used more by men than by women. Therefore, although deshoo is the masu form, it is normally used even in informal conversation by both women and men. Since a contrast between deshoo (the masu form) and daroo (the plain form) is not salient in most talk, I did not include deshoo as a masu form in the present analysis.





Haruko Minegishi Cook

Examples (1a) and (1b) are identical in referential meaning but differ in social meaning because the main predicate ends with the masu form in (1a) and with the plain form in (1b). Typically, the masu form is considered as a polite form, and the plain form as a non-polite form. (1) a. Taro ga Hanako to dekake-masu. (masu form) Taro SUB Hanako with go out ‘Taro goes out with Hanako.’ b. Taro ga Hanako to dekake-ru. (plain form) Taro SUB Hanako with go out ‘Taro goes out with Hanako.’ In (1a) the verb ‘to go out’ is in the masu form (dekakemasu) while in (1b) it is in the plain form (dekakeru). When a sentence is a complex sentence, the verbal of the main clause is responsible for the contrast in social meaning. Whether the verbal of the main clause is in the masu or plain form, the verbal of the subordinate clause remains in the plain form unless the sentence is in “hyper-polite” style (Jorden and Noda 1997; Makino 2002).2 (2) a.

[[Taro ga Hanako to dekake-ru toki] itsumo ame ga Taro SUB Hanako with go out when always rain SUB furi-masu] (masu form) fall ‘When Taro goes out with Hanako, it always rains.’

b.

[[Taro ga Hanako to dekake-ru toki] itsumo ame ga Taro SUB Hanako with go out when always rain SUB fur-u] (plain form) fall ‘When Taro goes out with Hanako, it always rains.’

In sentences (2a) and (2b), the verb of the main clause (i.e., furimasu or furu ‘to fall’) indexes social meaning. 2. Makino (2002) states that a verbal that appears in a subordinate clause can be marked in the masu form, which indicates “hyper-politeness.” The example given below is identical with (2a) above except that the verbal in the subordinate clause is changed to the masu form (dekakemasu). Makino (2002) refers to this type of sentence as “hyper-polite.” (2a’) [[Taro ga Hanako to dekake-masu toki] itsumo ame Taro sub Hanako with go-out when always rain ga furi-masu] sub fall ‘When Taro goes out with Hanako, it always rains. In the present data, however, there was no instance of a “hyper-polite” style (i.e., a verbal in the subordinate clause marked in the masu form).



Style shifts in academic consultations

In the present data, there are number of utterances ending in a connective particle kedo ‘though’ (or its variants keredo and kedomo). In Japanese grammar, kedo is classified as a connective particle marking a subordinate clause. However, often in conversation, kedo occurs utterance-finally without the main clause as shown in (3). (3) Ashita wa nichiyoobi da/desu kedo tomorrow top Sunday COP though ‘Tomorrow is Sunday but..’ In such an instance, the function of kedo is not a grammatical connective particle that forms a subordinate clause, but it functions as a pragmatic particle that contributes to various conversational moves (Kamio 1994; Maynard 1989; Mori 1999). Thus, when kedo is used utterance-finally as a pragmatic particle, a kedo-marked utterance is not a subordinate clause. Either the plain or masu form can occur in a kedo-marked utterance, contrasting in social meaning.3

2.2

Is the masu form a polite form?

The masu form is analyzed as an honorific form that shows politeness to the addressee (Comrie 1976; Martin 1964). Thus, it has been called a polite or formal speech marker (e.g., Hinds 1978; Ide 1982, 1989; Martin 1964; Neustupny 1978; Niyekawa 1991). In contrast, the plain form is referred to as a non-polite, or informal, form. Most studies of Japanese honorifics claim that a shift from non-honorifics to honorifics is triggered by external social factors. For example, a number of studies on Japanese honorifics state that the honorific forms are used to a higher-status addressee and in a formal situation. The underlying assumptions of these previous studies are: (1) social rules dictate language use, and (2) there is a one-to-one correspondence between linguistic form and social category. I will elaborate on these assumptions and argue against them below. First, there is a view that the external social factors (e.g., social rules) define speech form. For example, Ide’s (1982, 1989) notion of wakimae ‘discernment’ represents this view. Wakimae is social etiquette that the members of Japanese society must observe. According to this view, Japanese social rules make reference to social position, power, age and formality, and the use of the formal forms is “socio-pragmatically obligatory” when talking to a person who is in a higher position, more powerful, or older, as well as when talking in a formal social situation. In Ide’s words: …the social rules of Japanese society require one to be polite to a higher status person like a professor. This use of an honorific verb form is the socio-pragmatic equivalent of grammatical concord, and may thus be termed socio-pragmatic concord (1989: 227). 3. However, since a kedo-marked utterance such as the one shown in (3) is not the subordinate clause, it is not considered “hyper-polite” even if the masu form is used.





Haruko Minegishi Cook

Ide and Yoshida (1999: 454) further state, “In general, people who are objectively in a lower position cannot shorten the psychological distance by not using honorifics, because that behavior is considered rude according to the politeness of wakimae.” In this view, the use of honorifics, including addressee honorifics, is defined by social rules. External social factors dictate style shifts. Style shifts occur due to the shifts of the social situation and/or the addressee since the speaker is observing the social rules. According to this view, a student must use honorifics when talking to or referring to a professor. The student’s shift to the non-honorific form indicates a change of the addressee to a lower-status person. Such a shift when talking to a professor is considered a violation of a social rule. In this view, the participants of a social activity are seen as passive observers of socially defined behavioral codes. This view still underlies many studies in Japanese sociolinguistics, in particular those on honorifics. The ways in which the masu and plain forms are used in natural discourse is, however, much more complex. Scholars examining natural discourse data have found that Japanese style shifts do not necessarily correspond to a shift of social situation and/or addressee. In other words, there is no one-to-one correspondence between the use of the masu form and a particular category of addressee or social situation. Niyekawa (1977), who studied a TV home drama, has found that the hero in the drama speaks most of the time in the plain form to his subordinates but sometimes shifts to the masu form when he shifts the topic to a more serious one, requests assistance, or plays the social role of an expert. Some of the shifts (e.g., playing the role of an expert) cannot be adequately explained by the external social factors (i.e. a change of the addressee or social situation) alone. A shift may occur when the speaker has a change of feeling. Ikuta (1983), who examined a TV interview program, has also found that style shifts occur when the speaker changes how she feels without any change in the social situation or the addressee. She analyzes the masu form as a marker of distance and claims that what triggers a shift is a shift in psychological distance between the interlocutors. In the TV interview program, both the interviewer and interviewee speak in the masu form to maintain a certain distance, but a shift to the plain form may occur when the interviewer shows empathy for the interviewee. Furthermore, directly linking the masu form to a higher-status addressee poses a theoretical problem (see also Okamoto 1998). Cross-linguistically, social categories such as social status are rarely directly encoded in language (Irvine 1992; Ochs 1993; Woolard 1992). Ochs (1993) states: …the relation of language to social identity is not direct but rather mediated by the interlocutor’s understandings of conventions for doing particular social acts and stances and the interlocutors’ understandings of how acts and stances are resources for structuring particular social identity (p. 289)

In this view, social identities are jointly negotiated and constructed through stance markers in interaction by the participants rather than being pre-determined, non-negotiable and fixed categories. In this sense, linguistic forms are resources for



Style shifts in academic consultations

constructing the social identities of the participants as well as social activities. Typically, linguistic forms are direct indexes of epistemic and affective stances. A particular quality of knowledge and/or a particular quality and intensity of feelings or mood directly indexed by linguistic forms constitute the speaker’s social identity (Ochs 1996). For example, deletion of the copula in American Standard English indexes different social identities of the addresses, such as that of a child, a foreigner, a patient, or an elderly person (Ochs 1988). The social identity of the addressee is narrowed down by other co-occurring linguistic features. More recently, scholars working on Japanese honorifics have explored the complexities involved in the use of honorifics in different social contexts and have begun to see that shifts between the honorific and non-honorific forms are strategic choices the speaker makes for “persona management” (Coupland 2001). The speaker is not rigidly constrained by the social rules but strategically chooses a particular linguistic form to display a particular social persona. Okamoto (1998) researched the speech of salespersons and customers in marketplaces and department stores in Kyoto and Osaka and proposed: …it is not simply the social distance or the setting that dictates the use and nonuse of honorifics. Rather, the variation reflects the differences in the speaker’s speech-style strategies vis-à-vis their evaluation of multiple social aspects of the context (p.154).

Okamoto (1999) further examines the mixing of honorific and non-honorific forms in several different social contexts. In almost all the conversations she examined, the participants shift between the masu and plain forms while talking to the same addressee. For example, during a conversation between a professor and a college student in the professor’s office, even the student sometimes shifts to the plain form when talking to the professor. Okamoto demonstrates that style shifts are not simply caused by the change of the addressee or social situation but are motivated by the speakers’ display of the right degree of formality, deference, and/or friendliness. From the constructivist point of view, I have argued elsewhere that the masu and plain forms are linguistic tools that are used to construct various social personae and contexts (Cook 1996a, 1996b, 1997, 1998). In my analysis, masu and plain forms are affective stance markers. By “affective stance” I mean the stances that indicate “moods, attitudes, feelings, and dispositions as well as degrees of emotional intensity” (Ochs 1996: 410). The masu form directly indexes the stance of self-presentation. In contrast, the plain form is what Maynard (1993: 180) calls the “morphologically less committed form.” It directly foregrounds the innate self, a stance that is the spontaneous expression of the speaker. These affective stances further index different social personae. The speaker speaking in the masu form is acting on-stage, either literary or figuratively, showing his or her presentational persona. In contrast, speaking in the plain form indexes a lack of such a stance. The speaker speaking in the plain form is acting in his or her natural, spontaneous way without showing the presentational side. Speaking in the





Haruko Minegishi Cook

presentational stance is interpreted differently in different social contexts. Thus, the masu form indexes various social meanings in different social contexts and does not necessarily index formality and politeness in every context. Style shifts are resources for participants to index a particular social persona or to define the social activity. While most of the time parents speak to their child at home using the plain form, sometimes they shift to the masu form when they engage in parental responsibilities such as correcting the child’s behavior or serving food, or when they quote others’ voices or say fixed phrases. The masu form in these instances nonreferentially conveys the message that they are talking as someone in charge of the household matters, such as a “mother” or “father.” Or the family members’ use of the masu form may define the activity they engage in as play. In an elementary school classroom the teacher tends to speak to the students using the masu form when he or she talks to the entire class but tends to shift to the plain form when he or she talks to an individual student. The students too shift between the two forms. They use the masu form to perform happyoo (presentation) activities but shift to the plain form once they have finished. The students’ style shifts create the boundary between the classroom activity of the happyoo and the non-happyoo portion of the class activities. In sum, the recent studies on Japanese style shifts (Cook 1996a, 1996b, 1997, 1998; Okamoto 1998, 1999) have demonstrated that the shifts are motivated by, in Okamoto’s terms, “multiple social aspects of the context.” Despite these recent contributions to our understanding of Japanese style shifts, the assumption that the masu form is used when the situation is formal or when the speaker shows politeness to a higher-status person still prevails among scholars, teachers, and students of Japanese. This is partly because the masu form is called the polite or formal style. For this reason, it is important to further investigate how the masu form is used in a wider range of social situations. This paper examines the style shifts in academic consultation sessions between professors and students in Japanese universities. For two reasons, an academic consultation session is an ideal social setting to examine whether style shifts occur due to external social factors. First, both the topic and the participants are constant. The topic is focused on academic matters, and the participants are the professor and the student. If style shift is triggered by external factors such as a change of the addressee, the topic, or the setting, it should be constant during the session. Secondly, the relationship between a professor and a student has been characterized as a prototypical power relationship. According to the notion of wakimae, the student uses honorifics to the professor whereas the professor does not reciprocate. The present study examines style shifts in academic consultation sessions and shows the following: (1) the professor and student shift speech styles not to respond to changes in external social factors, but to display different facets of their social identities, and (2) the social meanings of the masu form are not tied to the linguistic form but are emergent in interaction mediated by the ideology of a particular social context.



Style shifts in academic consultations 

3. Data The data of this study come from academic consultation sessions conducted by three male professors from two different universities in the Tokyo area. Professor A is in his 50s and a professor of sociology at a famous private university. Professor Y, who is in his 40s, is a professor of linguistics at the same university. The consultation sessions of Professors A and Y with their male students are each an hour long, and the entire sessions were video- and audio-recorded. Professor A consults with his M.A. thesis advisee, Student K, who is writing a thesis on the religions of Korea. They discuss these religions from sociological perspectives. Professor Y has a directed reading course in linguistics with his student, F. Professor Y asks Student F questions concerning the assigned reading on theoretical linguistic models. Professor T, who is in his 60s, is a professor of sociolinguistics at a women’s college. Professor T’s consultation sessions consist of 6 separate segments in which he consulted with 6 different female students who are all preparing their B.A. theses. Professor T helps the students develop their thesis topics. Each session lasts about 20 minutes. All the students are in their early 20s. Whether the students come to see the professor to develop their thesis topic or discuss the content of an assigned book, the professors provide directions to the students by discussing the main points, clarifying the difficult points, and asking clarification questions. All the consultation sessions were held in the professors’ offices at the university. To help preserve the naturalness of talk, the recording was done without the researcher present. All the data were transcribed.

4. Analysis of academic consultation sessions Just as in any other institutional setting, in the academic consultation sessions, the participants are expected to act according to the institutionally predefined social roles, namely the professor as an expert and the student as a novice (Bardovi-Harlig and Hartford 1990). In this sense, the unequal power relationship between the two participants is a priori distributed in this institutional setting. However, research on academic advising sessions (He 1995; He and Keating 1991) demonstrates that each participant displays multiple social identities in the course of the session. In other words, although their institutional identities are advisor and student, the participants display multifaceted aspects of social identities through ongoing interaction. For example, the student may display her gender, major, or ethnicity (see He 1995). This is what Bakhtin (1981) refers to as “heteroglossia.” In other words, utterances are filled with multiple “voices.” In what follows, I show that the participants in the academic consultations do not always exchange masu and plain forms in a nonreciprocal manner, but jointly construct multiple social identities. I also show how the masu and plain forms as linguistic resources index these social identities.



Haruko Minegishi Cook

4.1

Presentation of “professional self ”: “Professional” professor and “professional” student

In the academic consultation sessions, often both the professor and student reciprocally exchange the masu form regardless of the status difference. Example (4) illustrates the reciprocal use of the masu form. In all the data excerpts, the masu form is in bold, and the plain form in the main clause is underlined. Here the student is planning to go to Korea to do a survey. The professor is asking the student about how his trip is connected to his proposed thesis. (4) (Professor A and Student K) 1→ S: soo desu ne, ima no tokoro wa. ‘That’s right, right now.’ 2→ P: shite iru wake desu ne= ‘You are doing so, right?’ 3 S: =hai, ima no tokoro wa= ‘Yes, right now.’ 4 P: =de ano:: ja: kankoku de yaru choosa tte iu no wa sotsuron [no ichibu ni naru to ‘And we::ll as for the survey in Korea, it will be part of your thesis 5 S [ichioo ichibu sono= ‘part of it uh’ 6 S: =shita shirabe mo kanete:= ‘as part of the preliminary study.’ 7 P: =hai ‘Yes.’ 8 S: dekireba kankoku no shuukyoo no koto o= ‘If possible, the religions of Korea 9 P: =hai ‘yes.’ 10→ S: yori fukaku [shiritai] na: tte omottemasu= ‘I’d like to investigate them more deeply, I think.’ 11 P: [hai hai] ‘huh uh huh uh’ 12 P: =hai, sorede ano sotsuron: ja nai ya kono choosahoo no baai ni wa ‘Yes, then well in the case of uh not the thesis, uh this survey’ 13 S: hai ‘Yes’



Style shifts in academic consultations 

14 P:

ano: kako no bunken wa shirabete de sore ni yotte atarashii nanraka no kasetsu o tsukuru to ‘We:ll, studying the previous literature and based on that, coming up with some sort of hypothesis’

15→ S: soo desu= ‘That’s right.’ 16 P: =de sono kasetsu o kenshoo suru koto ga kanoo na shitsumon jikoo o shitsumon hihan no naka ni ireru → to iu wake desu ne= ‘And it is the case that you will include the criticisms of the questions, the question items that make it possible to verify the hypothesis, right?’ 17 S: =hai ‘Yes.’ The professor asked the student if the topic he is talking about is related to the theme of his M.A. thesis. Speaking in the masu form indexes that the speaker is displaying the self-presentational stance, which foregrounds his professional role in the academic setting. By “the professional relationship,” I mean that the professor speaks as a “professor” and the student as a “student.” In other words, the masu exchange indexes that their relationship is official and public, and not personal and private. It co-constructs a mutually professional and official relationship between the professor and student. The reciprocal exchange of the masu form interactionally foregrounds an equal relationship between the two parties in that they both display their “professional” side, and it plays down the institutional hierarchy between the professor and the student.

4.2

Hierarchical relationship

Sometimes the professor and student co-construct a hierarchical relationship with the forms that they use. When the professor shows his spontaneous side by shifting to the plain form, often the student responds by maintaining the professional stance marked in the masu form. This non-reciprocal exchange of the masu and plain forms indexes a hierarchical relationship in the academic consultation session just as it does in many social situations.4 Consider example (5). (5) (Professor Y and Student F) 1 P: nani ka shitsumon arimasu ka ‘Do you have any questions?’ 4. Not all non-reciprocal exchanges of the masu and plain forms index power relationships. The caregiver’s use of the masu form when teaching children, for example, does not index a lower status even when the child talks to her in the plain form.



Haruko Minegishi Cook

2 S: eeto ((turning the pages of the book)) ‘We:ll’ 3 P: koko made de: [wakannakatta tokoro toka soo iu no nakatta? ‘Up to this point, isn’t there anything that you did not understand?’ 4

S: [((cough))

5 P: eego wa muzukashiku nakatta? ‘Wasn’t the English difficult?’ 6→ S: atta- ((laugh)) atta to omou n desu kedo chotto ‘There was- I think there were a few (questions).’ 7 P:

iya, eego de muzukashii no kikarete mo wakannai kedomo ‘Well I myself do not know the answer if you ask me something difficult in/about English. ‘5

In line 1 the professor asks the student in the masu form if he has questions regarding the assigned reading. In line 2, the student cannot respond to him, so he turns pages in the book. In lines 3 and 5, the professor elaborates on his first question by shifting to the plain form. In line 6 the student laughs as he says that he thinks he has a few questions, which suggests embarrassment because of his inability to come up with a specific question. He maintains the masu form here. It is the student’s choice of not shifting to the plain form that co-constructs a hierarchical relationship. The higher status of the professor indexed by his use of the plain form does not necessarily implicate the authoritarian image of the professor. In both examples (5) and (6), the professors’ use of the plain form indexes their personal involvement with the student’s situation. In line 7 in example (5), the professor states that he himself would not know the answer if the student asked him something in/about difficult English. Since the student still cannot come up with a question in line 6, the professor’s self-effacing remark in line 7 is his positive politeness strategy, which attempts to minimize the student’s possible embarrassment. In this statement, the professor uses the plain form, which indexes that he makes his self-effacing utterance not in a professional stance but in a personal stance. Shifts to the plain form are typically initiated by the professor. Such shifts tend to occur when he elaborates on his initial question or statement said in the masu form. Example (6) illustrates this point. In (6), Professor T is asking his Student N about her grandparents’ hometown because the student is going to study the dialects of the grandparents. The professor shifts to the plain form but the student maintains the masu form.

5. It is not clear whether the professor means “something difficult in English” or “something difficult about English.”



Style shifts in academic consultations 

(6) (Professor T and Student N) 1 P: ojiisan obaasan wa irassharu n desu ka? ‘Do you have grandparents?’ 2 S: aa imasu. moo aichi-ken desu. ‘Uh, yes. They are in Aichi Prefecture.’ 3→ P: hu:n nagoya de wa nai wake ne? ‘Um they are not in Nagoya.’ 4 S: soo desu. nagoya [de wa nai n desu. ‘That’s right. They are not in Nagoya.’ 5→ P: [a soo da. sono mondai da ne. ‘Oh yes. It’s that problem.’ 6 S: soo soo sore mo omotta n desu yo= ‘Yes, yes. I also thought about that.’ 7→ P:

= h so- ano: anata no yatteru sono oniisan toka kookookookoosee zembu nagoya to wa kagiranai wake da ne. ‘Uh so- uh it’s not the case that the boys you are researching, the high school, high school students are all from Nagoya.’

8 S: soo na n desu yo ne ‘That’s true.’ Here in line 1 the professor asks the student if her grandparents are alive. In this utterance he uses not only the masu form but also the referent honorific form irassharu ‘to be’ showing respect to the student’s grandparents. In line 2, the student responds positively to the professor’s question and adds that they are in Aichi Prefecture. In line 3 the professor asks for a confirmation that the grandparents are not in Nagoya, for the student’s thesis title contains Nagoya, the main city in Aichi Prefecture. In this utterance, he shifts to the plain form and continues to use it in the subsequent turns. In lines 3, 5, and 7, he confirms and clarifies the student’s answer to his initial question, which is a sign of his personal interest in the student’s thesis project. The student consistently uses the masu form in answering his questions. Excerpts (5) and (6) are typical examples of a hierarchical relationship created by the masu and plain forms. The higher-status person uses the plain form, and the lowerstatus person, the masu form. This hierarchical relationship is co-constructed by the stances the participants display toward one another. The use of the plain form, which indexes the self who is spontaneous and not concerned about self-presentation, constitutes the professor’s higher status in relation to the student, who continues to display his or her professional persona or cultivated side toward the professor. In other words, the student’s use of the masu form here indexes his or her humble demeanor or lower status in relation to the professor. Thus, the exchange of the masu and plain forms



Haruko Minegishi Cook

foregrounds a hierarchical relationship. At the same time, the professor’s plain form indexes his personal involvement of interest in the student’s academic project, like a friendly personal coach.

4.3

Two attributes of the professor’s social identity

The masu and plain forms are resources for the speakers to create desired social identities and relationships in context. For example, Okamoto (1999) demonstrates that speakers in different social situations shift back and forth between the two forms to create their desired social identities. Similarly, in the present data, all three professors shift back and forth between the two forms to index different social identities, one that is professional and egalitarian (indexed by the use of the masu form) and one that is higher in status and personal (indexed by the use of the plain form). A shift to the plain form is not necessarily triggered by an external contextual factor such as a topic shift. It tends to occur when the professor elaborates on an initial question or statement said in the masu form and shows momentarily his personal side to the student. Frequent shifts between the two forms suggest that the professors are responding to two social demands, to be higher in status and personal on the one hand, and to be “professional” and egalitarian on the other. For example, as shown in (7), Professor T, who has been showing his personal side, shifts back to the masu form in line 10 to display his “professional” social identity. (7) (Professor T and Student N, continued from (6)) 9 P: ja chotto taitoru ni nagoya tte iu no o motte kuru no wa chotto mazui kamo shirenai ne. ‘Well, then it may be a little difficult to include Nagoya in the title.’ 10 →

[tookai chihoo aichi- aichi-ken aichi-ken wa aichi-ken na n desu ka? ‘Tookai region, Aichi, Aichi Prefecture, Aichi Prefecture, is it Aichi Prefecture?’

11 S: [a::: ‘Uh:::’ 12 S: aichi-ken desu. ‘It’s Aichi Prefecture.’ 13 P: hu::n ‘hm’ Sometimes the professor momentarily shifts back to the plain form to show his personal side, but often he then immediately makes a variant repetition to keep the professional relationship. Consider (8) and (9).



Style shifts in academic consultations 

(8) (Professor A, and Student K) 1 P: kasetsu o tsukutta sono katee de koo itta hensuu no eekyooryoku o doo kangaeteta ka tte koto desu yo ne. ‘In the process of making a hypothesis, the thing is how you thought of the influence of these sorts of variables.’ 2 S: ha: ‘Yes’ 3 P: honnin no seebetsu, gakureki, oya no gakureki ‘The person’s gender, education, the parents’ education’ 4 S: un ‘um’ 5→ P: nenree haitteru ne? [haittemasu ne? nenree hai ‘The ages are included, right? they are included right? ages yes.’ 6 S: [soo desu ne. haittemasu. ‘Right. They are included.’ 7

tte iu ka kankoku dewa: nanka nenree o yappa zettai: sore wa mochiron ireru beki na n desu kedo. ‘Or in Korea the age, that should be, of course, absolutely included.’

In line 5, the professor asks if the ages of the subjects of the study are included in the student’s research. In saying this, he shifts to the plain form but immediately repeats this utterance in the masu form. In line 6, the student starts to display an acknowledgment immediately after the professor completes the verb haitteru ‘included’, said in the plain form, which suggests that the student had no trouble understanding the professor’s utterance in the plain form. Therefore, the professor’s repetition in the masu form here is not motivated by difficulty of understanding but seems to be motivated by the professor’s choice to maintain his professional stance. Similarly, talking to his female student M in (9), Professor T shifts to the plain form but immediately reiterates his utterance in the masu form. Here the professor and the student are talking about using elementary school students’ compositions as research data. (9) (Professor T and Student M) 1 P: chotto anata jishin shirabete mite [kuremasen ka? sakubunshuu donna mono ga aru ka ‘Won’t you investigate what kind of compositions there are yourself?’



Haruko Minegishi Cook

2 S: [hai wakarimashita ‘Yes, I will.’ 3 P: chotto toshokan ni itte ne [un ‘Just go to the library, yeah’ 4 S: [hai ‘Yes’ 5 S:

ato ano setagaya-ku no imooto ga shoogakusee na n de ku no- ku kara dondon deru n desu ne gakkoo de ‘Cause my younger sister is in an elementary school in Setagaya Ward, the ward publishes a lot of compositions at school.’

6

demo soo iu fuu ni naru to chiiki ga gentee [sareru to mazui] n desu ka ne ‘but if the situation is like this, and the area is restricted, it will be bad, won’t it?’

7 P: [dakara gakkoo] ‘therefore school’ 8→ P:

sore wa kamawanai kedo gakunenbetsu ni tonikaku nenreebetsu ni aruteedo deeta ga sorotteru hoo ga ii yo ne. ‘That’s all right, but it is better to have data according to grade and age to a certain extent.’

9 S: hai ‘yes’ 10→ P: dakara nenreebe- gakunenbetsu tte iu no ga ichiban ii desu yo ne. ‘So it is best to arrange by age- grade.’ In (9) both parties are speaking in the masu form, and in line 8, Professor T shifts to the plain form when he gives his opinion about a preferred type of data. But immediately he reiterates what he said using the masu form in line 10. In both examples (8) and (9), the professor’s utterance in the plain form is immediately repeated in the masu form. Note that when the professor shifts to the plain form, he is giving academic advice to the student about his or her thesis. The plain form here indexes that the professor is playing the role of a personal coach. Perhaps the professor’s immediate repetition in the masu form indicates that his preferred choice is to maintain a professional status with the student when talking about academic matters. In this way, the professor is able to achieve a fine balancing act in fulfilling the two ‘social demands’ – as a professor and as a personal coach.



Style shifts in academic consultations 

As seen above, in the academic consultation sessions, the professor shifts between the masu and plain forms to talk to the student. These shifts do not necessarily mark the shift of a topic. Rather the shift occurs within the same topic with the same addressee. Thus, style shifts index shifts in two different attributes of his social identity or “voice.” Both the professor and the student often maintain a professional stance by means of exchanges of the masu form. But at times, the professor shifts to the plain form, which indexes the more relaxed and informal side to create a more personal context. While the professor’s plain form indexes his personal stance, the student’s masu form in the immediately preceding turn co-constructs a hierarchical relationship between the two. Thus, the professor and the student co-create the social identities of the professor in the on-going interaction, namely someone who is personal and of higher status, and someone who is “professional” and egalitarian. These two social identities of the professor are both socially expected attributes of university professors in Japanese society. Perhaps the professor is under two conflicting social demands, one that requires that he be socially higher in status and personally take care of the student, who is meshita (lower in status), and another that requires that he be “professional” and egalitarian. To balance these two conflicting socially expected images, the professor shifts between the two speech styles.

4.4

Students’ shifts to the plain form

Whereas the professor often shifts back and forth between the plain and masu forms as seen above, the student’s shifts to the plain form are infrequent and limited to two particular speech acts. One is when the student makes a soliloquy-like remark or an exclamatory comment (see also Okamoto 1999), and the other is when he or she talks about the content of the academic subject in a list-like fashion. The professors also shift to the plain form in these speech acts. However, since the students’ shifts to the plain form are limited to these speech acts, they are more noteworthy. 4.4.1 Soliloquy-like remarks Consider (10), in which the student shifts to the plain form to make a soliloquy-like remark. (10) (Professor T, and Student M) 1 S: u::nto sotsuron ni tsuite na n desu kedo: ‘U::h it’s about my graduation thesis.’ 2 → S: a nani o iu ka wasurechatta. ‘I forgot what to say.’ 3 S: e::to sotsuron ni tsuite na n desu kedo ‘Well, it’s about my graduation thesis.’



Haruko Minegishi Cook

4 P: hai ‘yes.’ ((a few lines omitted)) 5 P: keeyoo hyoogen no henka? ‘Changes of modifying expressions?’ 6 S: no henka ga ichiban: ‘its changes are the mo:st’ 7 → S: nante iu no kana ‘What shall I say?’ 8 S: kencho ni mirareru n desu yo. ‘clearly observed.’ In this segment, Student M has just come into the professor’s office and started to talk about her B.A. thesis. In line 1 she brings up the topic of her thesis, speaking in the masu form. But in line 2 when she mentions that she forgot what to say, she shifts to the plain form. This is a meta-comment that she is giving about her own psychological state. She starts again in line 3 by repeating what she said in line 1. In line 7 she cannot come up with an appropriate expression to describe the topic of her thesis so she says nante iu no kana ‘What can I say?’ In so doing, she again shifts to the plain form. Both utterances in lines 2 and 7 are soliloquy-like utterances and not necessarily directed to the professor. This is when the student’s shift to the plain form occurs. 4.4.2 The content of academic subjects Another type of shift is observed in the academic consultation session. It is a shift from the masu to the naked plain form to discuss the content of the academic subject matter. The naked plain form does not co-occur with the sentence particles or other affect markers such as coalescence and it is said in a flat intonation. The naked plain form used in this manner indexes a stance detached from the addressee and focuses on the content of talk (see Cook 2002).6 Makino (1983, 2002) and Maynard (1993) point out that the naked plain form tends to be used when the speaker has a low awareness of the addressee.7 Maynard (1993) explains that one of the contexts in which the naked plain form occurs is when the speaker/writer speaks/writes from the text-internal perspective without much conscious attention paid to the addressee. This is the style typically used in newspaper articles, scientific writing, and textbooks, which all focus on content. Therefore, the shift to the naked plain form in example (11) does not index an informal and personal relation between the professor and the student. Rather, it indexes that both participants are oriented to the content of the academic subject and 6. In Cook (2002), I used the term “impersonal style” to describe the stance indexed by the naked plain form. 7.

Makino (1983, 2003) does not distinguish the naked and non-naked plain forms.



Style shifts in academic consultations 

momentarily detached from a particular type of social relationship. Some naked plain forms are in list-like remarks, and others are in acknowledgments of a receipt. In example (11), the professor is asking the student questions about Buddhist sects in Korea. (11) (Professor A, and Student K) 1 P: TAIKOSHUU kore wa shuuha desu ka? ‘Taikoshuu, is this a sect?’ 2 S: soo desu. ‘That’s right.’ 3 S: kankoku: ni wa [nanka kono mae mo: ii- iimashita kedo ‘In Korea uh as I said before but’ 4: P: [hai ‘Yes’ 5 S: kankoku ni wa: ‘In Korea’ 6→ P: futatsu aru. ‘there are two.’ 7 S: soo, motomoto hitotsu shika nakute: ‘Right, there was only one and’ 8 P: un ‘um’ 9 S: sore ga tsuma- ano oboosan ga ‘wife- uh priests’ 10 P: un ‘um’ 11 S: tsuma o motte ii ka ‘whether they can have a wife’ 12 P: un ‘um’ 13 S: yokunai ka de futatsu ni wareta n desu. ‘or not divided the sect into two.’ 14 P: hai hai hai hai ‘yes yes yes yes.’ 15 S: de kotchi no Taikoshuu tte iu hoo no ga= ‘and this one called Taiko sect’ 16→ P: =motcha ikenai. ‘does not allow marriage.’



Haruko Minegishi Cook

17→ S: kotchi no ga motte[ii ‘This one allows marriage.’ 18 P: [motte ii. hoo hoo hoo. ‘It allows it. Yeah yeah yeah’ 19 S: de ‘and’ 20 P: moo hitotsu wa? ‘The other one?’ 21 S: saigo wa sookeeshuu= ‘The last is Sookee sect’ 22 P: sookeeshuu ‘Sookee sect’ 23→ S:=tte iu no wa [motte wa ikenai. ‘the one called (sookeeshuu) does not allow marriage.’ 24→ P: [motcha ikenai. ‘not allow marriage.’ Both professor and student exchange the masu form up to line 3. They are taking a professional role and maintaining the professional stance up to this point. The student is talking to the professor about characteristics of the Buddhist sects in Korea. In line 6 the professor co-constructs a sentence with the student, and in so doing he shifts to the plain form. The student states the locative in line 5 (kankoku ni wa) and the professor supplies the predicate in the plain form in line 6 (futatsu aru), stating that there are two sects in Korea. In lines 15 and 16, they again co-construct a sentence. The student provides the subject (de kotchi no taikoshuu tte iu hoo no ga), and the professor, the predicate in the plain form (motcha ikenai). However, in line 17, using the plain form, the student repairs the predicate supplied by the professor. He asserts that this sect allows priests to have a wife (kotchi no ga motte ii). The professor in line 18 gives an acknowledgement token by repeating the student’s utterance in the plain form (motte ii). Responding to the professor’s question in line 20, the student makes a statement in the plain form in line 21. This utterance introduces a sub-topic (Sookee sect), so in this sense, this utterance is like a list. After the professor’s partial repetition (sookeeshuu), the student co-constructs with the professor in the plain form. He treats the noun uttered by the professor in line 22 as a part of the noun phrase (sookeeshuu tte iu no ‘the one called Sookeeshuu’) and topicalizes it by adding the topic marker wa. Then he provides the predicate in the plain form. The professor also co-constructs with the student by providing the predicate in the plain form in line 24. Note that in these instances, the plain forms are all naked plain forms, which are used to list the characteristics of Korean Buddhist sects, the academic topic under discussion as well as acknowledgement tokens.



Style shifts in academic consultations 

4.5

Students’ strategies for avoiding a power relationship

The students do not shift to the plain form except for the two speech acts discussed above. Typically, they keep using the masu form when their professor shifts to the plain form. This practice creates unequal statuses between the participants in the academic setting. However, the students use two strategies to avoid creating unequal statuses: (1) the use of an incomplete sentence so that they avoid marking either the masu or plain form and (2) embedding the professor’s plain form utterance by co-construction. These strategies are possible due to the morpho-syntactic structure of Japanese as well as the sequential organization of talk. 4.5.1 Avoidance of marking The speech style is indicated by the masu or the plain form, which is a morphological marking on the tense-carrying verbal of the main clause. By the use of an incomplete sentence, however, the speaker can avoid marking a particular speech style. Since Japanese is a SOV language, in the canonical word order, the predicate comes at the end of a sentence. Therefore, it is easy to say everything in the sentence except for the verb. It is also possible to use non-final verbal forms, such as a gerundive form. In either case, the sentence is not marked with a particular speech style. For this reason, the use of an incomplete sentence can be interpreted as the student’s avoidance strategy, in particular when it occurs when the professor shifts to the plain form. In other words, by not marking the masu or plain form, the student keeps the relationship with his professor obscure. Thus the student can avoid creating a hierarchical relationship. Consider example (12). In this example Professor Y is asking Student F about the reading assignment. (12) (Professor Y, and Student F) 1 P: kore mata yominaoshite ‘Read this again.’ 2 S: hai ((sniffling)) ‘Yes.’ 3 P:

de kyoo wa pinkaasu no tokoro na n desu ga kore wa doo deshita? ‘And today it’s about Pinkers but how was this (reading)?’

4 S: ((turning the pages of the book)) (13.00) soo desu ne: ‘We::ll’ 5→ P: hyaku nijuu hachi peeji made wa yonda? ‘Have you read up to page 128?’



Haruko Minegishi Cook

6 S: hai ‘Yes.’ 7→ P: hu:n. muzukashikatta desho. [muzukashiku nakatta? ‘Um: it was difficult, wasn’t it? Wasn’t it difficult? 8 S: [muzukashii iya: kotchi yori wa ‘Difficult. No it seems more 9 →

yomiyasukatta yoo na readable than this.’

10 P: u::n un un ‘Um um um.’ In line 3, using the masu form, the professor asks the student how the reading went. The student is not able to answer the question right away and turns the pages of the book. He says soo desu ne ‘well’, reciprocating the masu form. But when the professor shifts to the plain form elaborating on his question in line 5, the student responds by using an incomplete sentence in line 9. The professor shifts to the plain form asking the student if he read up to page 128. He then suggests that the reading was difficult in line 7. He does so first using the desho (tentative form of the copula in the masu form) and then the plain form. In line 8, as an acknowledgement token, the student repeats the professor’s suggestion muzukashii ‘difficult’ in the plain form, and continues to say that the assigned reading seems easier than the other book. In this utterance, the student does not complete the predicate (i.e., yomiyasukatta yoo na). The complete expression might be something like yomiyasukatta yoo na ki ga shimasu ‘I feel that it is easier to read’, in which the verb suru/shimasu ‘do’ is the final tense-carrying verb and needs to be either in the plain form (suru) or the masu form (shimasu). By not completing the predicate, the student does not define the relationship with the professor as either a hierarchical or an equal relationship. In example (13), the student uses the gerundive form as an avoidance strategy. Here Professor T and Student B are talking about the student’s dialect. The professor asks the student where she lived in the past. (13) (Professor T and Student B) 1 P: doko ni sundeta n desu ka.’ ‘Where did you live? 2 S: e::to umare wa oosaka no sakai-shi na n desu kedo: ‘Uh I was born in Sakai-city, Osaka.’ 3 P: u::n= ‘uh huh’ 4 S: =sore kara sugu Okinawa ni tenkin ni na-[narima ‘Then soon (my father) was transferred to Okinawa.’



Style shifts in academic consultations

5→ P: sore ikutsu gurai no toki? ‘How old were you?’ 6→ S: sore moo honto ni issai ni natta ka [naranai ka gurai de ‘That was about just when I was almost one or so.’ 7 P: [hu::n okinawa ‘Um Okinawa.’ 8 S: hai ‘yes.’ In this example, the participants are exchanging the masu form up to line 4. Then in line 5 the professor shifts to the plain form when he asks the student at what age she moved to Okinawa. In line 6, she ends the turn with de, the gerundive form of the copula da. The student could complete the sentence either in the masu form (deshita ‘was’) or in the plain form (datta ‘was’), but does not choose to do so. These examples show that when the professor shifts to the plain form, the student does not necessarily reciprocate by shifting to the same form. Rather he or she often chooses an incomplete sentence or a phrase in order not to commit to a particular social relationship. When the current social event is an academic consultation, a move of this sort by the student obscures his or her institutional role as a student and at the same time avoids creating an equal and personal relationship with the professor. In other words, incomplete sentences and phrases serve as resources for strategies that obscure the social relationship between professor and student. 4.5.2 Co-construction Sometimes the student changes a potential hierarchical relation to that of a mutual professional relation by co-constructing a sentence with the professor. As noted above, the plain form indexes social meaning only when it occurs in the main clause, in which it contrasts with the masu form. In the subordinate clause, it normally does not contrast with the masu form. Thus, the current speaker’s utterance ending with the plain form can be embedded in a clause ending in the masu form said by the next speaker. Consider example (14), in which Professor T is talking with Student B about her proposed B.A. thesis on Japanese accent patterns. (14) (Professor T, and Student B) 1 S: de ano sotsuron dasu toki wa komakaku settee shinaide= ‘And uh when I submit the B.A. thesis, without setting details’ 2→ P: =narubeku oomaka ni shitoita hoo ga ii yo ne ‘It is better to leave it in a rough frame, isn’t it?’ 3→ S: =desu yo ne ‘isn’t it?’





Haruko Minegishi Cook

4 P:

un o- o- oosakaben to suru no ka oosaka hoogen ni suru ka kansai hoogen ni suru ka tte iu koto arimasu keredo ‘uh there is a problem of whether it is Osaka dialect, Osaka regional dialect, or Kansai regional dialect.’

5

maa sono taitoru o kimeru: ni attatte dakara nani o gutaiteki ni yaru, ‘well, when deciding the title’

6

doo iu hookoo de yaru ka tte iu koto ni tsuite no mitooshi da yo ne ‘it’s a matter of perspective on what to do and how to approach it.’

7 S: ha:i ‘Ye:s’ 8 P: doo omoimasu ka. ‘What do you think?’ 9 S: u::n watashi mo akusento ga yappari omoshirosoo da na: to omotte iru n desu kedo ‘I also think that accent seems to be interesting’ 10 P: u::n ‘um’ 11 S: nanka ‘somehow’ 12 P:

shibotta hoo ga ne [hontoo wa ii n da yo ne zentai bakuzen to yaru yori wa ne ‘It’s really better to focus rather than studying without focus.’

13 S: [soo desu ne ‘That’s right.’ 14→ P: anata wa akusento mo tsukai wakete iru mitai? ‘Do you use your accent properly yourself?’ 15→ S: na n desu kedo: ‘(that) is the case.’ In this segment we see co-constructions of a sentence in lines 1 through 3 as well as in lines 14 and 15. In line 1, the student states, “when I submit the B.A. thesis, without setting details,” and in line 2 the professor completes her sentence by saying narubeku oomaka ni shitoita hoo ga ii yo ne ‘it is better to leave it in a larger frame.’ He uses the plain form ii yo ne ‘is good, isn’t it?’, but the student’s co-construction turn in line 3



Style shifts in academic consultations 

embeds the professor’s utterance in line 2. Because of this unfolding, the professor’s utterance becomes the subordinate clause in line 3 as shown in (14a). (14a) [narubeku oomaka ni shitoita hoo ga ii]prof desu yo ne]student Possibly roughly do sub good cop prt prt ‘It is better to leave it in a rough frame, isn’t it?’ In other words, the student’s desu yo ne ‘isn’t it?’ frames the professor’s plain-form utterance, and as a result, the co-constructed utterance ends with the masu form. Since the plain form in the non-final position in a sentence does not index any social meaning, the student’s desu yo ne changes a potentially hierarchical relationship to a mutually professional one. Similarly, in line 15, the student co-constructs with the professor’s question said in the plain form in line 14 and transforms his question to a statement ending in the masu form as shown in (14b). Thus, again what could be a hierarchical relationship is avoided, and the mutually professional relationship is maintained. (14b) [[akusento mo tsukai wakete iru mitai]prof Accent also use separate seem na n desu kedo:]student cop nom cop but ‘It is the case that (you)/I seem to use (your)/my accent properly’8 The above examples (14a and b) show that the voices of the professor and the student are merged in the co-constructed utterance. Because in Japanese the subordinate clause comes before the main clause and the predicate verb in the subordinate clause is in the plain form, this syntactic structure is taken advantage of in the unfolding sequential organization of talk. By so doing, the student changes the potentially hierarchical relationship with the professor to a mutually professional one and co-creates a unified voice.

5. Discussion The professor and the student jointly construct multiple social personae in the course of the academic consultation session. The professor’s higher status is not singly indexed by his use of the plain form. Similarly the student’s lower status is not singly indexed by his or her use of the masu form. When the professor shifts to the plain form, the stu8. The original Japanese utterance does not encode the personal pronouns. The English translation includes both you and your (the professor’s point of view) and I and my (the student’s point of view). The professor is implicating ‘you seem to use your accent properly’ but the student increment completely switches to the student’s point of view from that of the professor. Thus, the interpretation ‘I seem to use my accent properly’ arises at the end of (14b).



Haruko Minegishi Cook

dent’s move in the next turn defines the social relationship between the two. From this point of view, what social relationship the participants hold is displayed in the moment-by-moment unfolding in interaction. When the professor shifts to the plain form, it is the student’s move to maintain the masu form that co-constructs the hierarchical relation. The student is not a passive recipient of the professor’s plain form. To avoid any social marking, the student sometimes chooses to use an incomplete sentence or phrase. Such a move obscures the institutional hierarchy and does not subject the student to being interactionally lower in status. The student also co-constructs with the professor by embedding the professor’s plain form with an increment ending in the masu form. This move is a type of repetition that shifts the professor’s utterance back to the masu form. By so doing, the student maintains the professional exchange with the professor. These strategies keep the student from constantly playing the subordinate role and obscure the institutional hierarchy expected in the academic setting. The participants of the academic consultation sessions jointly create the ideals of being a professor and student in Japanese society. By exchanging the masu form, both participants define the situation as “official” and “professional.” We can say that both participants are speaking in the voice of “professor” and “student” in a professional capacity. Although the institutional roles of the professor and the student are not equal in power, the reciprocal exchange of the masu form minimizes their status difference. The professor and student are on an equal footing in the professional capacity. The professor initiates the shift to the plain form, in particular when he elaborates on his questions or statements. Since the plain form indexes the spontaneous self, the use of the plain form shifts the attribute of the professor to a more personal one if it occurs with affect keys (i.e., if it is not a naked plain form). This shift pattern is similar to that of teachers in elementary school classrooms (Cook 1996a; Rounds, Falsgraf and Seya 1997). Elementary school teachers tend to talk in the masu form to the whole class and while introducing a new topic but tend to talk in the plain form to an individual student or when elaborating on a topic already introduced. In other words, the teacher shows his or her professional side to the whole class but speaks in the personal voice to an individual student. By so doing, the teacher is conforming to two different roles of being a teacher, the official role as a teacher and the role of “personal coach” to the students. We see that the professors in the academic consultation session at the university level are also conforming to the same dual roles through the use of the masu and plain forms. On the one hand, the professor is professional and his voice is an official voice, but on the other hand, he is a “personal coach” to his student. When the professor shifts to the plain form, most of the time the student maintains the masu form to co-construct a lower status, which is in line with the institutional expectation of the student role in the academic setting. The student’s reciprocal use of the plain form would co-create an informal and equal relationship with the professor. However, perhaps because an informal and equal relationship between the professor and student is not part of the dominant institutional ideology of the academic setting, the students



Style shifts in academic consultations 

in the present data rarely use the plain form except for soliloquy-like remarks and listing items of academic content.

6. Conclusion This paper examines the style shifts in academic consultation sessions between professors and students. The findings contribute to our knowledge of style shifts and of the construction of social identities. Style shifts are not solely triggered by external social factors such as addressee and topic. Therefore, even when the speaker speaks with the same addressee and within a single topic, style shifts occur. The participants shift their speech styles to display different facets of the social personae in a given social context. This article also shows the emergent nature of social identities. Duranti (1997) writes: Social structure is an emergent product of interaction in which social actors produce culture by applying native (typically implicit) methods of understanding and communicating what they are and what they care about (p. 10).

The social meanings of a linguistic form in part depend upon the sequential organization of talk. The social meaning of the masu form in part is evoked by the use or nonuse of the masu form in the subsequent turn. Thus, the assumptions that the masu form is directly tied to the social status and has a one-to-one correspondence between the form and social meaning do not hold. The social meanings are also mediated by the ideology of a particular social context or activity. The recent literature on the Japanese honorifics reports variations in the masu and plain form exchanges in different social contexts (Cook 1996a, 1996b, 1997, 1998; Megumi 2002; Okamoto 1998, 1999). For example, Okamoto (1998) reports that department store clerks use the masu form to the customer but street venders do not. In Cook (1996), I discuss how elementary school students use the masu form in the activity of happyoo but shift to the plain form outside of happyoo even when they talk to the teacher. The previous proposal that speakers observe social rules such as wakimae cannot account for these diverse variations of the speech styles in Japanese society. This article proposes that in university academic consultation sessions, the hierarchical structure of the academic institutions evokes the hierarchical relationship in the case of non-reciprocal exchange of the two forms. This way of looking at the phenomena can explain diverse variations found in the masu and plain form exchanges in different social contexts. In sum, style shifts occur with the same addressee and within the same topic because linguistic forms are resources for the collaborative construction of multiple social personae of the participants in a given social context.



Haruko Minegishi Cook

References Bakhtin, M. 1981. The Dialogic Imagination: Four Essays. Ed. By M. Holquist, trans. C. Emmerson and M. Holquist. Austin: University of Texas Press. Bardovi-Harlig, K. and Hartford, B. 1990. “Congruence in native and nonnative conversations: Status balance in the academic advising session.” Language Learning 40(4): 467–501. Bell, A. 1984. “Language style as audience design.” Language in Society 13: 145–204. Bell, A. 1997. “Language style as audience design.” In Sociolinguistics: A Reader, N. Coupland and A. Jaworski (eds.), 240–250. New York: St. Martin’s Press. Blom, J-P. and Gumperz, J. 1972. “Social meaning in linguistic structures.” In Directions in Sociolinguistics: The Ethnography of Communication, J. J. Gumperz and D. Hymes (eds.), 407–434. New York: Holt, Rinehart and Winston. Bucholtz, M. 1999. “’Why be normal?’: Language and identity practices in a community of nerd girls.” Language in Society 28: 203–223. Comrie, B. 1976. “Linguistic politeness axes: Speaker-addressee, speaker-referent, speaker-bystander.” Pragmatics Microfiche 1.7:A3. Dept. of Linguistics, University of Cambridge. Cook, H. M. 1996a. “The use of addressee honorifics in Japanese elementary school classroom.” In Japanese/Korean Linguistics Vol 5, N. Akatsuka, S. Iwasaki and S. Strauss (eds.), 67–81. Stanford, CA: Center for the Study of Language and Information. Cook, H. M. 1996b. “Japanese language socialization: Indexing the modes of self.” Discourse Processes 22: 171–197. Cook, H. M. 1997. “The role of the Japanese masu form in caregiver-child conversation.” Journal of Pragmatics 28: 695–718. Cook, H. M. 1998. “Situational meaning of the Japanese social deixis: The mixed use of the masu and plain forms.” Journal of Linguistic Anthropology 8(1): 87–110. Cook, H. M. 2002. “The social meanings of the Japanese plain form.” In Japanese/Korean Linguistics Vol. 10, N. Akatsuka and S. Strauss (eds.), 150–163. Stanford, CA: Center for the Study of Language and Information. Coupland, N. 2001. “Language, situation, and the relational self: Theorizing dialect-style in sociolinguistics.” In Style and Sociolinguistic Variation, P. Eckert and J. Rickford (eds.), 185–210. Cambridge: Cambridge University Press. Duranti, A. 1997. Linguistic Anthropology. Cambridge: Cambridge University Press. Eckert, P. and Rickford, J. (ed.). 2001. Style and Sociolinguistic Variation. Cambridge: Cambridge University Press. Giles, H. and Powesland, P. 1975. Speech Style and Social Evaluation. London: Academic Press. Gumperz, J. 1982. Discourse Strategies. Cambridge: Cambridge University Press. He, A. 1995. “Co-constructing institutional identities: The case of student counselees.” Research on Language and Social Interaction 28(3): 213–231. He, A. and Keating, E. 1991. “Counselor and student at talk: A case study.” Issues in Applied Linguistics 2(2): 183–210. Hinds, J. 1978. “Conversational structure: An investigation based on Japanese interview discourse.” In Problems in Japanese Syntax and Semantics, J. Hinds and I. Howard (eds.), 79–121. Tokyo: Kaitakusha. Ide, S. 1982. “Japanese sociolinguistics: Politeness and women’s language.” Lingua 57: 357–385. Ide, S. 1989. “Formal forms and discernment: Two neglected aspects of universals of linguistic politeness.” Multilingua 8–2/3: 223–248.



Style shifts in academic consultations  Ide, S. and Yoshida, M. 1999. “Sociolinguistics: Honorifics and gender differences.” In The Handbook of Japanese Linguistics, N. Tsujimura (ed.).444–480. Balden: Blackwell. Ikuta, S. 1983. “Speech level shift and conversational strategy in Japanese discourse.” Language Science 5: 37–53. Irvine, J. 1992. “Ideologies of honorific language.” Pragmatics 2: 251–262. Jorden, E. and Noda, M. 1987. Japanese: The Spoken Language. New Havens and London: Yale University Press. Kamio, A. 1994. “Theory of territory of information.” Journal of Pragmatics 21: 67–100. Labov, W. 1966. Social Stratification of English in New York City. Washington, DC.: Center for Applied Linguistics. Makino, S. 1983. “Speaker/listener-orientation and formality marking in Japanese.” Gengo Kenkyuu, 84: 126–145. Makino, S. 2002. “When does communication turn mentally inward?: A case study of Japanese formal-to-informal switching.” In Japanese/Korean Linguistics Vol. 10, N. Akatsuka and S. Strauss (eds.), 121–135. Stanford, CA: Center for the Study of Language and Information. Martin, S. 1964. “Speech levels in Japan and Korea.” In Language in Culture and Society, D. Hymes (ed.), 407–415. New York: Harper and Row. Maynard, S. 1993. Discourse Modality: Subjectivity, Emotion and Voice in the Japanese Language. Amsterdam: John Benjamins. Maynard, S. 1989. Japanese Conversation: Self-Contextualization through Structure and Interactional Management. Norwood: Ablex. Megumi, M. 2002. “The switching between desu/masu form and plain form: From the perspective of turn construction.” In Japanese/Korean Linguistics Vol. 10, N. Akatsuka and S. Strauss (eds.), 206–219. Stanford, CA: CSLI Publications. Mori, J. 1999. Negotiating Agreement and Disagreement in Japanese: Connective Expressions and Turn Construction. Amsterdam: John Benjamins. Neustupny, J. V. 1978. Post-Structural Approaches to Language: Language Theory in Japanese Context. Tokyo: University of Tokyo Press. Niyekawa, A. 1977. “Code switching in a stable relationship: An analysis of Japanese TV drama.” Paper presented in the Symposium of Japanese Sociolinguistics at the Summer Institute of the Linguistic Society of America. University of Hawaii. Niyekawa, A. 1991. Minimum Essential Politeness. Tokyo: Kodansha International. Ochs, E. 1988. Culture and Language Development. Cambridge: Cambridge University Press. Ochs, E. 1993. “Constructing social identity: A Language socialization perspective.” Research on Language and Social Interaction 26(3): 287–306. Ochs, E. 1996. “Linguistic resources for socializing humanity.” In Rethinking Linguistic Relativity, J. Gumperz and S. Levinsion (eds.), 407–437. Cambridge: Cambridge University Press. Okamoto, S. 1998. “The use and non-use of honorifics in sales talk in Kyoto and Osaka: Are they rude or friendly?” In Japanese/Korean Linguistics Vol. 7, N. Akatsuka, H. Hoji, S. Iwasaki, S-O. Sohn and S. Strauss, (eds.), 141–157. Stanford, CA: Center for the Study of Language and Information. Okamoto, S. 1999. “Situated politeness: Manipulating honorific and non-honorific expressions in Japanese conversation.” Pragmatics 9: 51–74. Rounds, P., Falsgraf, C., and Seya, R. 1997. “Acquisition of sociolinguistic competence in a Japanese immersion school.” Journal of the Association of Teachers of Japanese 31: 25–51. Sacks, H. 1992. Lectures on Conversation. Edited by G. Jefferson with an introduction by E. Schegloff. Oxford: Basil Blackwell.



Haruko Minegishi Cook Schegloff, E. 1991. “Reflections on talk and social structure.” In Talk and Social Structure: Studies in Ethnomethodology and Conversation Analysis, D. Boden and D. E. Zimmerman (eds.), 44–70. Cambridge: Polity Press. Woolard, K. 1992. “Language ideology: Issue and approaches.” Pragmatics 2: 235–349.

Appendix: Transcription conventions [ ] (0.0) CAPS :: - = (( )) bold underline

the point where overlapping talk starts the point where overlapping talk ends length of silence in tenths of a second relatively high pitch lengthened syllable cut-off latched utterances transcriber’s description of events masu form plain form

Interpersonal functions of style shift The use of plain and masu forms in faculty meetings Naomi Geyer

University of Wisconsin-Madison

1. Introduction This study attempts to examine the style shifts between the plain and masu forms in Japanese discourse within the framework of politeness studies. The plain and masu forms in Japanese are morphological markings on nouns, adjectives, and verb stems appearing in the clause-final position.1 The masu form is traditionally depicted as an addressee honorific (a type of honorific that expresses the speaker’s deference toward the addressee) and a formal style ending, while the plain form is referred to as an intimate informal style ending. These references reflect the general belief that the masu form is used in formal and polite social contexts, while the plain form is used in informal and intimate contexts, and that nonreciprocal use of the two forms indicates status difference. However, studies of the two forms based on discourse data maintain that static contextual features such as interpersonal distance and status difference cannot fully account for the substantial diversity in the use of the two forms (e.g., Cook 1996, 1997, 1999; Ikuta 1983; Maynard 1993; Okamoto 1997, 1998, 1999). These studies attempt to identify the core properties of the two forms, which trigger the style shifts in discourse. Moreover, the plain and masu forms are known to evoke numerous seemingly contradictory perceptions in discourse. Okamoto (1999) illustrates how interpretations of honorific and non-honorific expressions (which include masu and plain forms respectively) can vary individually. For example, non-honorific expressions (including the plain form) may be perceived as “friendly, sincere, warm, relaxed” or “rude, authoritative, too friendly, unrefined, etc.” (Okamoto 1999: 57). This multiplicity is partly due to the numerous interpersonal functions the two forms create in ongoing discourse. The complexity of style shift and its interpersonal effects calls for description of how the two forms’ numerous interpersonal functions are created in ongoing discourse. This study attempts to illustrate the interpersonal functions associated with the plain and masu forms in Japanese multiparty discourse. It thus focuses primarily on 1.

The masu form includes the copula desu, which follows nouns and adjectives.



Naomi Geyer

the interpersonal aspect of talk, the central concern in politeness studies. By describing how the core properties of the two forms interact with other factors in discourse to achieve diverse interpersonal or politeness functions, the study contributes to a more comprehensive understanding of the intricate mechanism of style shifts. This paper first explores various studies of politeness and their relationships to the two forms and to style shift. It also reviews studies of the two forms from linguistic and discourse perspectives. Finally, it presents a discourse-based microanalysis of the two forms to describe how their interpersonal functions are achieved.

2. Politeness studies and the social context of the two forms Linguistic politeness theories have spurred great interest in the sociocultural phenomenon of politeness, as well as in the various ways it is realized in different languages. Politeness is viewed as a set of maxims concerning the interpersonal aspects of conversation (Lakoff 1973; Leech 1983), a normative act abiding by conversational rights and obligations (Fraser 1980, 1990), or a set of strategies to counterbalance the face-threat of the performed actions (Brown & Levinson 1984). Even though these theories have inspired numerous theoretical and empirical inquiries, researchers have questioned their cross-cultural validity (Gu 1990; Hill, Ide et al 1992; Ide 1989; Mao 1994; Matsumoto 1988, 1989; Nwoye 1992; Pan 1995; Sifianou 1992, 1993, 1995; Wierzbicka 1985) and applicability to actual language data (Fraser 1990; Van De Walle 1993). These arguments inevitably involve the subtle but critical discrepancy between the theoretical notions of politeness and politeness demonstrated in actual interactions. That is, politeness theories often portray a set of general rules and maxims without paying much attention to how they are used in situated connected discourse, resulting in the theories being non-operative in actual analysis. Moreover, the theories tend to focus on an act-by-act analysis of isolated sentences without paying much attention to longer discourse and its contextual information. Several recent studies point out this limitation and advocate the use of connected discourse in politeness studies (e.g., Arundale 1999; Eelen 2001; Geyer 2008; Mills 2003; Okamoto 1999; Watts 2003). This study attempts to illustrate the style shift used as a form of politeness situated in a multiparty discourse in an institutional setting. The plain and masu forms in Japanese have been discussed in politeness studies. The use of the masu form, for instance, is introduced as a strategy of “giving deference,” one of the negative politeness strategies in Brown and Levinson’s list of politeness strategies. They maintain that every speaker desires “his actions to be unimpeded by others” (1987:  62) (negative face wants) and “his wants to be desirable to others” (62) (positive face wants). Since many communicative acts are potentially face threatening, these “face-threatening acts (FTAs)” require softening devices (politeness strategies) to counterbalance the threat. The strategies attending to negative face wants are negative politeness strategies, to which the masu form is claimed to belong. The speaker chooses



Interpersonal functions of style shift 

a strategy according to the weightiness of a FTA, which is measured by assessing contextual factors such as social distance and power difference. In fact, a number of linguistic studies of the plain and masu forms and other honorific expressions in Japanese consider them to be governed by rigorous social constraints. This observation resulted in the view of Japanese politeness as “discernment,” a system that operates according to social constraints independent of the speaker’s intentions (Hill et al. 1986; Ide 1989). In this view, the use of Japanese honorifics has been claimed to be rather automatic, indexing the speaker’s understanding of social constraints. “Discernment” politeness is claimed to be different from the concept of politeness as “volitional” strategies proposed by Brown and Levinson (1987) (Hill et al. 1986; Ide 1989). Politeness is viewed as “a set of behaviour patterns preprogrammed as social norms” (Watts et al.  1992:  5), and therefore the description of social constraints becomes particularly relevant. As a result, a number of studies have sought to describe various social constraints governing Japanese politeness. As for the plain and masu forms, studies have depicted the masu form as a marker of social distance between interlocutors, the use of which is determined by their status difference and/or degree of intimacy (e.g., Hinds 1978; Shibatani 1990). In addition to status difference and social distance, researchers have identified other social constraints influencing the use of the two forms, such as gender, genre, and, as mentioned, the formality of the speech setting. These social constraints are taken into consideration when a native speaker of Japanese chooses between the plain and masu forms, and the two forms, in turn, index various social variables.

3. Motivation for style shifts: discourse-based studies The social constraints described above undeniably affect the use of the two forms. However, studies focusing on the shift between the plain and masu forms illustrate frequent switches even within a single stretch of discourse. The above-mentioned static social constraints fail to explain their diverse use within discourse. Researchers have examined style shifts in various settings and provided accounts in terms of the core linguistic properties of the two forms. Ikuta (1983) examines style shifts in conversations from TV talk shows and states that the use of des/-mas forms (i.e., masu forms in this study) represents not only a social but also an attitudinal and/or cohesional distance, and that the use of plain forms represents a lack of such distance. In other words, style shift indicates a moment by moment change in a speaker’s attitudinal distance from an interlocutor, or indicates a change in the hierarchical positioning of an utterance in a discourse. Cook (1996, 1997, 1999) examined the use of the masu form in various settings in terms of its direct and indirect indexicality (regarding the notion of indexicality, see Ochs 1990, 1993). She claims that the masu form indexes “the presentation of public



Naomi Geyer

self ” and addressee deference. That is, the speaker uses the masu form when presenting a social persona (role) and/or when showing deference to the addressee. Cook maintains that a display of self-presentation can be interpreted as a display of deference in some contexts and that contextual features of communicative events foreground one of these values over the other. For instance, a television interview program, which “demands a display of mannered self-presentation” (1999: 104), foregrounds the masu form’s indexical value of self-presentation. Other lines of inquiry that attempt to account for the alternation between the plain and masu forms involve the notion of speaker/addressee orientation: the plain and masu forms are associated with speech addressed to oneself and to others, respectively (Kindaichi 1982; Makino 1983, 2002). This orientational dichotomy is modified and acquires a continuous quality in Maynard’s (1993) analysis, where she attributes the use of the two forms to the level of speaker’s “addressee awareness.” She claims that when a speaker is more aware of an addressee (“thou,” in Maynard’s term) as a separate and potentially opposing entity, he or she tends to use the masu form. In contrast, when a speaker is less aware of the addressee, as in self-addressed utterances, the plain form is used. In casual conversation among intimates, speakers use the same style as in self-addressed utterances (i.e., the plain form) since the psychologically close relationship among interlocutors makes less distinction between “thou” (addressee) and self (speaker). As a counter-example to Maynard’s (1993) proposal, Cook (1999) presented interactions in a neighborhood quarrel where the plain form was predominant. She claims that since the two quarrelers did not appear to be psychologically close (i.e., not intimate friends), their use of the plain forms cannot be explained by Maynard’s proposal of addressee-awareness.2 However, if “addressee” (or “thou”) in Maynard’s model refers to the individual to whom the speaker presents his or her public self, “addressee (thou) awareness” and the “presentation of public self ” indicate similar interpersonal constraints. That is, the presentation of public self presupposes an addressee to whom one’s public self is projected. For instance, in the neighborhood quarrel mentioned above, the strangers used the plain form since they did not have to present their public self, or since they did not recognize each other as someone to whom they would show their public selves. Consequently, addressee awareness and the presentation of public self can be understood 2. This observation illustrating the similarity between the linguistic behavior of intimates and strangers is reminiscent of Wolfson’s (1988, 1989) Bulge Theory, which states that the power and distance between interlocutors affect the amount of elaboration in talk and that the amount of linguistic elaboration forms a bulge-shaped curve along the scale of social distance. Wolfson’s “linguistic elaboration” does not correspond to the use of addressee honorifics in question, and attempts to account for the use of the two forms in terms of social distance have been proven insufficient. However, the underlying interpersonal motivation turns out to be similar. According to Wolfson, the need to negotiate a relationship makes talk more elaborate, while the less such need exists (due to either a secure or non-existent relationship), the less elaborate the conversation.



Interpersonal functions of style shift 

as two aspects of the same phenomenon. In a similar vein, Okamoto argues that honorifics (including the masu form) express deference and/or formality toward the addressee, where the notion of formality means “one’s restrained and ritualistic attitude toward another person” (1999: 58). Here again, formality is directed to the addressee, and the speaker’s “restrained and ritualistic” attitude is closely related to his or her public image. Thus far, the core properties of the two forms have been discussed as deference and/or formality toward the addressee and the presentation of public self (the masu form), and as lack of such deference, formality, or the need to present a public self (the plain form). In the course of analyses aiming to illustrate their core properties, various interpersonal and social functions of the two forms have been observed. These and other functions associated with the two forms create various, and sometimes contradictory, interpretations regarding their use. Since the interpretations and perceptions of the use of the two forms depend largely on their interpersonal functions, it is important to illustrate how these functions are achieved in ongoing discourse. More specifically, this study takes up the above-mentioned core properties and addresses the question of how the core properties of the forms interact with other contextual and situational factors to achieve their various interpersonal functions in discourse. It is hoped that a detailed description of the interplay between their linguistic properties and the situational and social variables will help to clarify the working of style shifts in talk.

4. Method This study attempts to depict the intricate mechanisms by which the interpersonal functions of the plain and masu forms are achieved in situated discourse by analyzing their use in a set of multiparty spoken discourse in semiformal faculty meetings at Japanese secondary schools. Six semiformal faculty meetings (five to six participants per group), totaling 290 minutes, were audiotaped and analyzed from a discourse-analytic perspective. The data set contains a large quantity of style shifts, which call attention to both macro- and micro-level variables. The interpersonal functions observed in this study do not necessarily correspond directly to the core properties of the forms. Rather, they can be accounted for as a constellation of various social, situational, interpersonal, and sequential factors.

5. Overall characteristics The data set shows a dominant use of the masu form with a considerable number of shifts to the plain form. This tendency can be explained by the characteristics of the speech setting: small group meetings at a workplace. The preferential use of the masu



Naomi Geyer

form agrees with the interpretation of the form as public presentation of self in previous studies (Cook 1996, 1997, 1999), since meetings at a workplace are considered public. This also corresponds to Makino’s (1983, 2002) claim that the masu form is predominantly used in utterances addressed to others, while the plain form is used in self-addressed utterances (e.g., when the speaker mutters to himself or herself). The relatively frequent use of the plain form can be attributed to the semiformal nature of the meetings and the interpersonal relationships among the participants. Compared with formal school-wide conferences, grade-level faculty meetings have a smaller number of participants. The power differences among participants are relatively small, reflecting an absence of managing personnel such as a principal or a vice principal. Nevertheless, minor power differences may exist, depending on such factors as participants’ age, gender, and length of employment at the school.

6. Individual differences As mentioned above, the data set exhibits the mixed use of plain and masu forms, with a tendency to use the masu forms more often. However, the distribution of the two forms across speakers is not identical: some speakers use the plain form more than others. Although individual differences are not the focus of this study, it is important to consider this variable before presenting the linguistic context in which the two forms appear. If each participant uses only one of these forms exclusively, their usage could be sufficiently explained from the constellation of these social variables, and it would no longer be necessary to analyze the sequential environment in which the two forms appear. However, an examination of all the meetings shows that all participants shift their style: no one uses either form exclusively. In other words, speakers’ social attributes cannot sufficiently account for the alternating use of the two forms. Therefore, other interactional and discourse organizational motivations are indicated for style shifts.

7. Plain form as solidarity marker Core linguistic properties of the plain form have been said to be its lack of formality, deference, or need to present the speaker’s public self. The plain form is frequently observed in monologues in casual conversations among intimates where formality, deference, and the display of a public self are not necessary. However, even within meetings of coworkers where speakers predominantly use masu, frequent shifts to the plain form occur. Shifts from the masu form to the plain form are observed in interactional solidarity-building speech activities. This section explores sequences which involve this type of



Interpersonal functions of style shift 

switch and delineates how speakers mark solidarity through their use of the plain form. Consider excerpt (1), in which the mutual use of the plain form results in the elaboration of a joke. Before the excerpt, teachers were talking about supplementary lessons during summer vacation. (Transcription conventions are given in the appendix.)

(1) 1

Sato:

2

jaa hoshuu no hoo desu kedo jaa mikka teedo de [yoroshii deshoo ka ne, [masu] ‘Then, as for the supplementary lesson, is it all right to do it for about three days?’

3 Others: [humm humm. ‘Uh-huh uh-huh.’ 4 Sato:

jaa saitee mikka tte iu koto de,= ‘Then, it will be at least three days.’

5 Ota:

=ii n janai desu ka, (masu) ‘I think it is OK.’

6→ Sato:

oota sensee wa mainichi. [plain] ‘Mr. Ota will do it every day.’

7→ Kato:

mainichi (.) yonjuunichi ne? [plain] ‘Every day, for forty days.’

8 Ota:

iyaa hhaha, ‘Oh, no!’ ((laughter))

9→

kishi: aa? (.) soo ka (.) soryaa sugoi na. [plain] ‘Oh, is that right? That’s great!’

The head teacher, Sato, summarizes the preceding interaction about the supplementary classes in masu forms (lines 1 and 2), and her summary is approved by another teacher, also in a masu form structure (line 5). Upon approval, Sato switches to the plain form in line 6 and teases one of her fellow teachers, saying “Mr. Ota will do it [a supplementary lesson] every day.” This utterance is considered to be a joke due to its casual humorous tone and unrealistic content. Here the use of the plain form, along with other elements of talk, appears to create greater solidarity among participants. One important question to consider is “How is solidarity related to the core linguistic meaning of the plain form?” As discussed earlier, core linguistic properties of the plain form are to mark a lack of formality (the need to show social persona) and/ or a lack of deference. It is prototypically used in in-group interactions, in which there is no need to be formal. From this prototypical use, the function of the plain form as a solidarity marker can be readily deduced. The use of in-group language, when used in nonintimate interactions, can signal a speaker’s lack of need to present his or her social persona at the time of interaction. Previous studies have pointed out this solidarity-building function of the plain form. For example, Ikuta (1983) claims that in a television interview in which the



Naomi Geyer

predominant form is the masu form, some of the occasional switches to the plain form are triggered by the speaker’s empathy with the addressee. Cook (1999) examined a similar speech event and maintains that plain forms with affect keys (e.g., affective final particles and animated tones) can mark intimacy. Likewise, Okamoto (1998, 1999) claims that speakers use both honorific and nonhonorific expressions to express desired interpersonal relationships. According to Okamoto, nonhonorific expressions, including plain forms, can express a speaker’s less formal and more friendly stance. Nevertheless, when a mixed use of the two forms is observed during the course of an interaction, speakers do not appear to use the plain form randomly. That is, the use of the plain form cannot always mark solidarity, and speakers appear to choose certain sequential environments for their switches to the plain form. Therefore, it is premature to assume that a style shift is motivated solely by a speaker’s momentary feeling or need to express empathy, and it becomes necessary to analyze the process of creating solidarity. The second question, then, is: “How is the pragmatic function of marking solidarity achieved?” In this light, Cook (1999) observed that in a television interview, plain forms occurred mainly in the interviewer’s assessments, which were located in the second part of an adjacency pair. What accounts for this sequential constraint? Do similar constraints exist in other speech events? In order to assess such sequential constraints in the present data set, it becomes critical to consider types of interactions. Interactions within the meetings can be characterized either as planned and official or as spontaneous and unofficial. Official and planned talk is transactional in nature (i.e., expressing “content”), and its content is likely to appear in the agenda or minutes of the meeting. Unofficial and spontaneous talk is interactional in nature (i.e., expressing social relations and personal attitudes), and its content is peripheral.3 Examples of official and unofficial talk are reports and jokes, respectively. In the data set, speakers use the plain form predominantly in “unofficial” talk. This distributional pattern is related to the core linguistic property of the plain form: speakers use the plain form in unofficial talk where they do not have to show an official social persona. One type of unofficial and spontaneous talk is joking. In joking, the nature of the talk works in conjunction with the plain form to create the pragmatic function of marking solidarity. This explanation is applicable to Cook’s (1999) observation that plain-form usage is limited to assessment or evaluation sequences in a television interview program. The evaluative turn in the interview setting is relatively spontaneous in comparison with the interviewer’s planned questions. The interviewer can switch to the plain form and mark intimacy in these sequences. The nature of utterances, however, is not the only factor interacting with the core property of the plain form. Other linguistic and paralinguistic features also contribute 3. Brown and Yule (1983) make a distinction between transactional and interpersonal language, and point out the similarities of their distinction to other functional dichotomies such as “referential vs. emotive” and “descriptive vs. social-expressive” (e.g., Lyons 1977).



Interpersonal functions of style shift 

to create an overall pragmatic effect. Cook (1999) calls these features “affect keys.” Keys observed in this example include the final particle ne and its variants, as well as the animated tone of voice employed by interlocutors. In excerpt (1), Kato’s plain-form uptake of Sato’s statement in line 6 accelerates the joke when he says in line 7 “Every day for forty days.” Ota’s subsequent response, as well as a comment by another teacher, also exhibits plain forms, indicating that both speakers are playing along with the joke. Not only does the plain form signal that they are in an off-topic joking frame, but the continued use of the plain form by various participants also lends cohesion to the joke. In excerpt (1), a joke is followed by another plain-form uptake, which leads to an escalation of the proffered joke. Excerpt (2) contains another example of plain form usage in a joke. Here, a masu form uptake of the proffered joke does not lead to an elaboration. Before the excerpt, Akita, the head teacher, who belongs to the Student Life division, raised the issue of unruly behavior by female students. Toward the end of the report, Maeda inserts a brief evaluation using the plain form (line 3).

(2) 1 Akita:

2

yappa kooiu koto wa mazui to omottemasu (.) de, [masu] ‘After all, I think this kind of behavior is bad, and..’

3→ Maeda: onnanoko wa tsuyoi nee. [plain] ‘Girls are strong, aren’t they?’ 4 Others: ((laugh)) 5 Akita:

otoko wa yowai desu ne (.) sokorahen mo kangaete

6

yaranakya ikenai kanaa tte omou n desu kedo (..) ee

7

(.) ato kyoo seekatsushidoo de, [masu] ‘Men are weak, aren’t they? That kind of matter, too I think we must consider, but… er… And today, at the Student Life division meeting...’

Maeda’s evaluation, “girls are strong,” is voiced in an animated, laughing tone, indicating that he is not making a serious comment. The plain form adds to the unofficial tone of the utterance and thus marks solidarity. The coexisting affect keys, the animated tone and the final particle ne, work jointly with the plain form to achieve a pragmatic effect. After other participants respond to the joke with laughter in line 4, Akita continues with a similar comment – “men are weak” – using the masu form. Although Akita’s utterance can be considered a continuation of Maeda’s proffered joke, it does not lead to its elaboration. In fact, the comment is followed by a brief summary of the issue, and Akita proceeds to another issue discussed in the Student Life division meeting. The contrast between excerpts (1) and (2) indicates that mutual use of the plain form leads to the elaboration of jokes, which could signal greater solidarity among participants.



Naomi Geyer

The plain form is also frequently used in another type of interactional talk, indirect complaints. Indirect complaining (Boxer 1993) is the act of making a complaint to someone other than the person responsible for the action about which the speaker is complaining. It is a type of “troubles talk” (Tannen 1990, 1994) directed to a third person. Excerpt (3) consists of a series of utterances leading to a collaborative performance of indirect complaining which develops into a joke. The participants complain about a teacher who keeps materials and records to herself and does not share them with others. At the time, this troublemaker teaches a different grade and is not present at the meeting. Therefore, the performed act is an indirect complaint or an example of troubles talk. Miki is a female teacher who is relatively new in the school. For the first time she is in charge of senior guidance sessions – sessions in which teachers have to consult with students regarding their higher education. Usually, a teacher in charge is provided with a set of relevant materials and records used over the past several years so that no new paperwork has to be prepared from scratch. However, as Miki reported prior to this segment, the only material available to her was from last year. She delivers her report in the masu form.

(3) 1 Miki:

..itsuka sano-sensee ga kite sono (.) kojinshidoo

2

suru n dattara okashi shimasu nde tte osshatta nde ( )

3

toka omotta n desu kedo, [masu] ‘Sometime ago, Mr. Sano came to me and said that he would lend it to me if I plan to offer individual guidance to students.’

4 Suzuki: ichi bu gurai wa tottokanai-to ne, ‘One should keep at least one copy, right?’ 5→ Kato:

jaa nani sore dake shika nokotte-nai no? [plain] ‘Then what? Only that is left?’

6 Miki:

un.. watashi ano (.2) mae no kateeka no hito ga

7→

sono hito ga motteru kamo[shirenai. [plain] ‘Yeah, I... the former home economics teacher, she might have it.’

8 → Kato:

[zenbu ne [plain] ‘All of them.’

9

(1)

10→ Sato:

tabun koko gonenkan gurai motteru [kana(..) hyottosuruto [plain] ‘Probably she might have the records for the last five years. It’s possible.’



11→ Honda: [aa soo [plain] ‘Oh, is that so?’



Interpersonal functions of style shift 

12→ Kato:

sennen bun gurai aru n da yo ne [plain] ‘She has about a thousand years’ worth.’

13→ Sato:

juumannen. [plain] ‘A hundred thousand years.’

14 Others: ((laugh)) At the end of her report, using both a referent honorific expression, osshatta ‘say’, and the masu form, Miki mentions that Mr. Sano, a teacher from another grade, has offered to share his own materials (lines 1–3). Upon hearing Miki’s report, Suzuki utters a truncated sentence followed by the final particle ne, stating that one should keep at least one copy of the materials. Her comment implies that whoever was previously in charge of sessions should have kept records and passed them on to Miki. Kato’s following question in line 5 is made in the plain form. Here, he expresses surprise that Miki received only a few records from the person in charge of the previous guidance session. The use of the plain form indicates a spontaneous reaction and suggests that the question is not an official interrogation. In response to Kato’s question, Miki presents her speculation that the home economics teacher may have kept all the records. She uses the plain form, maintaining the “unofficial” tone initiated by the previous turn. Upon Miki’s reference to the home economics teacher, other participants commence a collaborative griping performed entirely in the plain form (lines 8–12). The indirect complaint about the teacher continues and is accelerated into a joke in which Kato states that the teacher has kept the material for the past thousand years (line 13). Sato further elaborates upon the joke, exaggerating the number of years even more. This exchange of relatively short turns with mutually supportive content creates the feeling that the complaint was co-constructed by several participants. The mutual use of plain forms in the griping sequence marks solidarity and exaggerates the complaining so that it becomes joking. In this section I have examined the use of the plain form in “unofficial” talk to mark solidarity. Factors such as the nature of performed speech acts (e.g., jokes), other linguistic features such as affect keys (e.g., the final particle ne), and turn organization (e.g., co-construction by several participants) interact with the core property of the plain form to mark solidarity. In fact, these factors can work as solidarity markers all by themselves, making the contribution of the plain form rather small. The unofficial types of talk analyzed in this section, indirect complaints and jokes, are considered solidaritybuilding activities. According to Brown and Levinson, jokes and small talk “stress mutual shared background” (1987: 124) among participants and “convey that the speaker and hearer are cooperators” (125), thereby working as positive politeness strategies. Affect keys, especially the final particle ne and laughter, can also function as solidarity markers. In the presence of these features, the contribution of the plain form to the overall pragmatic effect is rather small. It is very likely that even the use of the masu form might not have altered the character of these sequences as solidarity building.



Naomi Geyer

Nevertheless, the plain form advances solidarity by framing the utterance as less official and more spontaneous. Recall excerpt (2), in which Akita’s uptake of a previously proffered joke was delivered in the masu form. Despite the fact that the content of his statement supported the previous joke, it did not lead to any elaboration. Conversely, the uptake in the plain form provided in excerpts (1) and (3) led to an escalation of jokes, creating greater solidarity. In sum, the use of plain form in these solidarity-building activities enhanced the pragmatic effect of the performed acts.

8. Plain form as mitigation The previous section examined the working of the plain form in solidarity-building types of talk such as joking and indirect complaining. Occasionally the plain form is also employed in face-threatening acts. This section examines the use of plain forms to create various kinds of mitigation in expressing noncompliance and in self-qualifications within face-threatening acts.

8.1

Plain form use in expressing noncompliance

Consider the next example in which the plain form functions to mitigate the face threat of a statement of noncompliance. Before this segment, the teachers discussed some forthcoming individual guidance sessions with students. To schedule sessions within the already-crowded timetable of school life is a difficult task, since each homeroom teacher conducts approximately forty of them. Unfortunately, Seki, the head teacher, reported that the teachers’ request to substitute the sessions for the afternoon classes was rejected in a meeting of the administrative committee to which she belongs. The excerpt starts with the last part of Seki’s statement, in which she announces that it is impossible to cut the number of afternoon classes (line 1). (4) 1 Seki:

..dame da to iu koto na n desu kedo ne, [masu]. ‘It has been decided that we can’t cut the afternoon classes.’

2

(1)

3→ Abe:

mokuyoobi no rokkooji wa doo suru n daro. [plain] ‘I wonder what we’d do for Thursday sixth period.’

4 5 Endo:

(.5) mokuyoobi (.) soo desu ne. [masu] ‘Thursday...that’s right.’

6→ Abe:

ima made wa (.) sore mo mendan ni ateteta n janai

7→

kana, [plain] ‘Before, we also used that time for consulting, didn’t we?’



Interpersonal functions of style shift

8 Miki:

semete ne, ‘At least, right?’

9 Abe:

semete gojikan ni suru gurai wa ne (.) shitekurenaito, ‘At least reducing the classes into five periods, without them doing that...’

10 Miki:

soo desu ne, [masu] ‘That’s right.’

11→ Abe:

tootei yarikirenai naa. [plain] ‘We can’t possibly do it.’

12 Miki:

soo desu nee, [masu] ‘That’s right.’

13 Seki:

jaa (..) jaa ichioo mokuyoo no gojikan iya

14

rokujikanme wa katto shite hoshii to iu yooboo o

15

dasu tte kotode ii desu ka? [masu] ‘Well, then, is it all right to request to cut Thursday fifth, or rather, sixth period?’

After a one-second pause following Seki’s report, Abe, a male teacher, wonders aloud, in a vague, self-contemplating manner, whether there is a possibility of arranging something for sixth period on Thursday. He uses daro (a truncated form of daroo), a plain, tentative form of the copula, which is equivalent to the English phrase “I wonder.” The suggestive nature of the statement implies an incipient noncompliance. This implication, however, neither triggers Seki to reformulate her previous statement nor other participants to elaborate on it. In fact, the uptake, which follows a half-second pause, is a minimum response – a repetition of a word from the previous utterance and a backchannel token (line 5). In lines 6–7, 9, and 11, the previous speaker, Abe, elaborates on his former indirect noncompliance. Abe elaborates by stating that in the previous year the teachers held guidance sessions in the sixth period (lines 6 and 7), and that on consultation days the number of periods should be reduced to five. These elaborations, as well as the initial statement, are performed entirely in the plain form, mitigating the force of the speaker’s resistance. Let us consider how the pragmatic function of mitigation is created through the placement of the plain form. In coworker meetings, some disagreements are less imposing when presented in the plain form since they sound less official and, therefore, off-the-record. This less official nuance is derived from the interaction between the core meaning of the plain form and the speech setting. The core meaning of the plain form, or the speaker’s assessment that he or she does not have to display formality or deference to others, is related to intimacy between interlocutors and the level of formality of the speech setting. The previous section delineated the way in which the plain form is used to build solidarity when used in conversations between noninti-





Naomi Geyer

mates. In a similar way, the plain form can invest an utterance with an “unofficial” nuance when used in relatively formal settings, such as the coworker meetings in the present data set. However, the placement of the plain form alone does not create the mitigation effect; in examples (1) through (3), the plain form functions not to mitigate face threats but to create solidarity. Thus, the sequential environment of the plain form in (4) becomes relevant in delineating the mitigation process. First, the progression from weak to strong noncompliance may contribute to the overall pragmatic effect. The initial contemplation of Thursday’s sixth period, which is offered vaguely, is supported by the subsequent report that the teachers held guidance sessions in the sixth period in the previous year. The report leads to a rather strong statement in which the speaker denies the possibility of having these sessions without cutting classes. Within these utterances, other linguistic elements contribute to the overall effect as well. The tentative form of the copula, daroo, in line 3, and the final particle naa in line 11, which is usually used in self-addressed utterances, contribute to the overall mitigating effect by creating the feeling that the speaker is musing to himself. The mitigating effect is created when the core meaning of the plain form concurs with other factors such as (a relatively formal) speech setting, the face-threatening nature of the performed action, and other linguistic elements. Another important feature of excerpt (4) is that the noncompliance presented in the plain form is regularly followed by backchannels in the masu form. After the interaction, the summary offered by the chairperson in lines 12–13 incorporates the utterances addressed in the plain form as an “official” opinion. In other words, the uptake of others in the masu form (lines 5, 10 and 12) invests Abe’s self-contemplating utterances (lines 3, 6, 9, and 11) with a more official status. Excerpt (5) contains a similar example in which a disagreement is delivered in the plain form. The teachers discuss when to schedule the next parent-teacher meeting. Miki, a new female teacher in the school, is in charge of the event. In lines 1–2, she states that a similar meeting was not held around the same period in the previous year. Two teachers, female and male, provide a repair of Miki’s statement in subsequent turns.

(5) 1

Miki:

kyonen made wa kono jiki ni mendan shitenakatta

2

yoo na n desu kedo:,[masu] ‘Until last year, it seems that we didn’t have a meeting during this period, but…’

3→ Kaga:

etto nee yatta kana?= [plain] ‘Well, did we do it?’

4→ Senda: =yatta ne, [plain] ‘We did.’ 5

Miki:

a (.) shita n desu ka. (.) jaa yappari kotoshi mo kono



Interpersonal functions of style shift 

6

jiki ni tte koto-de, [masu] ‘Oh, you did? Then, after all, this year too we are thinking of around this time, and...’

In line 3, the noncompliance is prefaced by the hesitation marker etto ‘er’ and is provided in the form of a question. The subsequent uptake is the answer to the question, also offered in the plain form. These plain forms invest the question-answer sequence with an “unofficial” nuance. As a result, the sequence can be regarded as a peripheral side-sequence that the two teachers exchange only with each other, even though the content of the sequence clearly exhibits noncompliance with a third party. Realizing that her assumption is wrong, Miki reforms her proposition accordingly in the next turns (lines 5–6). In this example, the plain form, the hesitation marker etto, and the question form in line 3 collaboratively contribute to the overall effect of mitigation.

8.2

Plain form in self-qualification

Another form of mitigation appears in qualification segments, in which speakers limit their statements. In these sequences, a qualification presented in the plain form is inserted within the overall utterance, which is usually phrased in the masu form. Consider (6), which follows a report of unruly students in Oka’s class who did not participate in any class activities. Suzuki, who belongs to the Student Life division, states that Oka, the teacher in charge of that class, should have reported the situation to the other teachers.

(6) 1 Suzuki: ...oka sensee mo tsurakatta n desu keredomo::o (.)

2→

°eto ° [plain] (..) aaiu

3

toki ni kichinto renraku shinaitoikenai to omou n

4

desu yo ne, [masu] ‘...Ms. Oka also suffered, but, let’s see, what shall I say? In that situation, one has to inform others properly, that’s what I think.’

Suzuki’s suggestion threatens Oka’s face because it includes an accusation that Oka did not inform other teachers “properly.” The face-threatening act is mitigated first by its presequence, which acknowledges the suffering of Oka, the teacher in charge of the class (line 1). The inserted segment following the presequence shows the speaker’s difficulty in choosing the right expression. A soft tone of voice and the plain form frame it as a side comment. Its content (that the speaker does not know what to say) qualifies the force of the statement since it implies the speaker’s uncertainty in performing the face threat in this manner.4 In this case, the use of the plain form makes the hedge more vivid. It frames the segment as a monologue, creating the impression that the speaker is thinking as the utterance unfolds. Other features contributing to the overall mitigation 4. The speaker might be pretending that he needs time to think about the correct wording.



Naomi Geyer

include the hesitation marker eto ‘let’s see’ and the delay of performing the target act. By inserting the qualifying sequence, the speaker gains time to think about the correct wording of his suggestion. This delay can also be interpreted as a form of mitigation. Another example of hedging takes place in excerpt (7), in which the speaker inserts a plain-form sequence, a soo da ‘oh, yes’, into a larger masu-form sequence introducing a new topic.

(7) 1 Miki:

toriaezu ja kyonen no (.) ano shiryoo o (.) kotchi de

2→

kopii shimasu [masu] <°a (.) soo da °> [plain] (.) ato

3

chotto omoidashita n desu ga [masu] (.) kesa ano

4

kitakoo no sensee ga korareta n desu [masu] kedo (.) ‘For now, then, we will copy the notes from last year. Oh, yes, and I have just remembered, a teacher from Kita High School visited our school this morning and…’

The inserted sequence in line 2 is produced swiftly in a soft tone with the interjection, a ‘oh’. These features indicate that Miki is speaking to herself because she suddenly remembered something she should report. After the insertion, she immediately resumes her normal tone and speed. Brown and Levinson consider topic change to be an imposition to the hearer’s face (1987: 168) requiring hedging expressions such as “Hey, I’ve just thought...” or “Oh, I know...,” similar to the qualification segment in (7). Thus, the content of the utterance as a whole – that the speaker has just remembered something – can be considered a type of mitigation. By delivering the initial part of the utterance in the plain form and with an interjection, the speaker stresses the immediacy of the sequence and hedges the threat caused by opening a new topic. The above two sections have examined the two distinct, yet interrelated, functions of the plain form: to mark solidarity and to mitigate the imposition caused by FTAs. In both cases, the core linguistic property of the plain form interacts with particular types of performed speech actions to create pragmatic effects. In interactional talk, such as indirect complaining and joking, the plain form heightens the unofficial tone of the discourse. The plain form conveys lack of formality and, therefore, creates solidarity. Other linguistic features, such as affect keys and collaborative turn organization, jointly contribute to the overall pragmatic effect. The plain form is also observed within face-threatening speech actions. In the case of issuing noncompliance, it functions as a mitigating device by framing the utterances as unofficial and, therefore, peripheral. The plain form is also used in qualification segments; here it adds to the immediacy of the speaker’s need to qualify utterances and, therefore, leads to mitigation. In both cases, the performed speech actions, as well as other linguistic factors, interact with the core property of the plain form to jointly create these overall functions.



Interpersonal functions of style shift 

9. Masu form as deference marker I will now move on to the politeness functions of the masu form. The core linguistic property of the masu form is the speaker’s recognition of the need to display formality or deference to others. The function of the masu form examined in this section is identical to its core property. It explores the relationship between social constraints (e.g., status difference) and sequential constraints (e.g., speech act of indirect complaint) in determining the style shifts. Prior to excerpt (8), teachers had been discussing alternatives to a campfire in case of rain during an upcoming school trip. Masaki, a young female teacher, is in charge of recreational events during the trip. After she proposes indoor games, Kato, the elderly head teacher, suggests another possibility: a candlelight ceremony. Upon receiving the suggestion, two female teachers, Nakai and Masaki, recall a similar event from the previous year. At that time, they had encountered a “mean guy” – a local man taking care of the public gym near the camp, who had been strongly opposed to the candlelight ceremony. Nakai and Masaki use the plain forms in their complaint about the “mean guy.” However, when the head teacher joins the exchange, both of them switch to the masu form in addressing the head teacher, even though the head teacher himself uses the plain form. Their switch within the same sequence appears to mark Nakai and Masaki’s recognition of the age and status differences between the head teacher and themselves. (8) 1 Nakai:

kyonen ( ) ne? ‘( ) last year, right?’

2 Masaki: kyonen yatta yo ne de ya:na ojisan ga ite, [plain] ‘We did it last year, right, and there was this mean guy and...’ 3 Nakai:

soo soo. yana ojisan ne, ‘Right, the mean guy...’

4 Kato:

e? kyonen? yattakke. [plain] ‘What? Last year? Did we do it?’

5→ Nakai:

yarimashita yo::. [masu] ‘We did it.’

6→ Masaki: yatta janai-desu-ka: taiikukan de hora minna-de 7→

roosoku motta janaidesu ka:. ano yana ojisan (.)

8→

sensee ano yana ojisan to zutto kooshoo ni

9

atatterashita n desu yo. [masu] ‘We did, at the gym, and, d’you remember, we all held the candles together. That mean guy..., Mr. Kato, you were engaged with that mean guy all the time.’



Naomi Geyer

10 Kato:

aa yana ojisan ne aa soo koo mattete ne, ‘Oh, the mean guy. Oh, right. He was waiting like this.’

11 Masaki: soo kyonen roo ga ochiteta toka itte watashitachi ga 12→

kuru no o machikamaetete taihendatta n desu

13

yo nee, [masu] ‘Right. He said that traces of wax had been left on the floor last year. He was waiting for us eagerly, and it was so much trouble.’

14 Kato:

soo ne machikamaetete ne, ‘That’s right. He was waiting eagerly for us.’

15 Masaki: moo machikamaetete sanchuu ga kitara itte-yaroo



tte ne de nakai-san nakasarechatta n dakke. [plain] ‘He was waiting eagerly, thinking “if the Third [the name of the middle school] comes, I will tell them!” And, Ms. Nakai, did he make you cry?’

17 Nakai:

iya naki wa shinai kedo, ‘No, I didn’t cry, but...’

16

When Nakai mentions the previous year’s event, Masaki’s agreement leads to the memory of the “mean guy” (lines 1–2). Masaki’s utterance is presented in the plain form. In line 3, Nakai remembers the same person and starts griping collaboratively, utilizing an emphatic agreement (soo soo), a truncated sentence and the final particle ne. The use of the plain form with affect keys in the interactional act of complaining was analyzed earlier in this paper. It adds an unofficial nuance to the utterance and enhances solidarity. In line 4, the head teacher, Kato, asks a question about the event, also in the plain form. However, when answering Kato’s question, both Nakai and Masaki use the masu form. At this point, there are two possible explanations for the shift. First, they may simply have stopped complaining and gone back to a more official mode of speaking. Kato’s previous question may have triggered this change; in spite of his plain-form usage, his question does not elaborate on their complaint. Second, they may have switched to show addressee deference to Kato, even though they were still in “complaining mode.” Kato’s role as head teacher may have triggered the change, but the switch could also have been triggered by his greater age and by the fact that he is a man and the other two are women. Given these two possible interpretations, the second one – marking addressee deference – seems more tenable for several reasons. First, both Nakai’s and Masaki’s utterances employ an animated tone and elongated final particles, which indicate that they are still in the casual mode. Second, and more important, even after Kato remembers the person under discussion and joins in the complaint in line 10, Masaki keeps complaining in the masu form. Subsequently, in the same complaint sequence, she switches back to the



Interpersonal functions of style shift 

plain form in addressing Nakai (lines 15–16). In other words, within a single griping sequence, Masaki switches between the plain and masu forms according to the addressee. This supports the second explanation for the masu form usage – to mark addressee deference. Along with the masu form, other deferential expressions in this sequence include the referent honorific atatterashita ‘engaged’ in line 9. We should also consider the relationship between the function of the masu form and its sequential environment. The performed speech act is an indirect complaint, a solidarity-building interactional act. Furthermore, the speakers’ use of an animated tone, as well as solidarity-building final particles, creates a casual atmosphere. As discussed earlier, such a sequential environment usually encourages the use of the plain form. The masu form usage in this sequence calls for special consideration of aspects of the social context (e.g., age and status difference) in which the sequence is embedded. This segment shows that social constraints can override sequential constraints in affecting style shift. Thus, the sequential characteristics of this segment highlight the deferential function of the masu form.

10. Use of masu form to impersonalize the speaker As mentioned, the overall data show a predominant use of masu forms. Therefore, it is rather difficult to tease out the role the masu form plays in a particular sequential environment. In view of this issue, this section focuses on one of the participants, Fujii, a male teacher in his fifties who delivers his utterances predominantly in the plain form. His personal style of talk (i.e., the predominant use of the plain form, as well as of other informal expressions such as ore ‘I’) might be related to the fact that he is the oldest and most experienced male teacher in the group. The rare occurrences of the masu form in Fujii’s utterances highlight its particular functions. As we will see, Fujii’s use of the masu form serves to impersonalize the speaker by indicating the “official” (versus the “personal”) nature of the imposition. In other words, the masu form indicates that the speaker is not directly responsible for the performed offense by emphasizing his official role. In English, various ways to avoid the pronoun “I” (e.g., using impersonal verbs and passive voice) serve a similar interpersonal end (Brown & Levinson 1987). In the following request sequence, Fujii asks other teachers to write their names on a calendar indicating their participation in a patrol during a local Bon-dance festival. Patrolling local festivals is one of the miscellaneous services public-school teachers provide in Japan in order to maintain a strong relationship with the local community. The festival this particular year will last for nine days, and since the grade consists of eight teachers, Fujii (acting as a mediator between the local community and the school) asks everyone to volunteer at least once.



Naomi Geyer



(9) 1→ Fujii:

..bon-odori no hoo wa kokuban ni hattokimasu nde

2

(.) eeto hitori ikkai teido o mokuhyoo ni (.) ano (.)

3→

kaite kudasai kyuuko desu kara (.) uchi no gakunen

4→

wa hitotsu amarimasu kedo (..) ee (.) kaite kudasai (.)

5→

onegaishimaasu [masu] ‘As for the Bon-dance, I’ll put a calendar on the blackboard, so please write down your names. Everyone should sign up at least once. There are nine slots, so for this grade there will be one left, but anyway, please write your name. Please.’

All clause-final positions, including the ones preceding kara and kedo, are in the masu form. They include several formulaic requestive expressions: the te-form of a verb plus kudasai (“please do”) and onegaishimasu ‘please; I beg you’, the latter delivered with stress and elongated vowels. The speech act of request is considered face threatening. However, Fujii does not mitigate his utterance except with the masu form and formulaic expressions such as -te kudasai and onegaishimasu. Here, the content of the request is not concerned with the speaker’s personal imposition on the addressee. Rather, the speaker plays the role of mediator, conveying the request from the local community. The stressed masu forms function to impersonalize the speaker by framing the request as official and, therefore, not personal. A similar function of the masu form is observed in excerpt (10). In lines 1 to 3, the head teacher, Arai, asks Fujii to present the results of the meeting of the administrative committee to which Fujii belongs. The topic of the discussion is tookoobi ‘going-toschool day’, a day during summer vacation when students and their homeroom teacher get together in school. Fujii’s report of the administrative meeting is delivered in the masu form (lines 4 to 11), which highlights his official role as meeting representative. (10) 1 Arai:

hai jaa (.) konaida horyuu ni natteta tookoobi

2

no ken wa uneeiinkai de itte doo natta ka

3

tte yuu no wa, ‘OK then, about the matter of the “going-to-school day” that we didn’t discuss fully last time, what did they say at the administrative meeting?’

4 Fujii:

tokuni ee kore to itte mookenasai to yuu kimari wa

5

nai kedo maa natsuyasumi juu seeto to nanraka no

6

katachi de kontakuto ga toreru yooni shite hoshii

7→

(.) atta hoo ga ii daroo to yuu koto desu yo ne (.) de

8→

(.) zenkai mo tsukekuwaemashita kedo (.) hagaki



Interpersonal functions of style shift 

9

nitsuite wa anoo (.) gakunen de hitsuyoona maisuu ga

10→

areba (.) itte kurereba yooi shimasu n de tte koto de

11→

(.) kyootoo- sensee kara teean ga arimashita (.) maa

12

tookoobi tte no wa maa motomoto chuukan de (.)

13

are ne mannaka de kodomo ni atte natsuyasumi

14

buji ni sugoshiteiru ka tte no o miru tekina mono

15

dakara ne? dakara (.) chuukan tte yuu to= ‘There isn’t any particular rule that requires us to have it, but somehow we’d better keep contact with our students during summer vacation. As I mentioned in the last meeting, the vice principal told us that he would prepare the postcards if he knew how many he should prepare. Well, originally “going-to-school day” was arranged in the middle of summer vacation to see if students were spending it safely, right? So the middle (of summer vacation) means…’

16 Haga:

=obon ‘Bon-Festival.’

17 Fujii:

soo obon mae ka na? tooka gurai ka na? (.) daa ima

18

wa ne, moo sorezore no yotee ga atte… sore ni

19 →

kawaru katachi no mono de ii n janai ka tte ki mo suru

20

kedo ne, [plain] ‘Yes, a little before the Bon Festival, isn’t it? Around the 10th? So now, everybody has their own plan, and I have a feeling it might be OK to have an alternative form.’

After the report, he adds his personal opinion concerning tookoobi ‘going-to-school day’ – that it is difficult to set up a date that fits everyone’s schedule and that some alternative form would be desirable (lines 12–15 and 17–20). This portion of Fujii’s utterance is delivered in the plain form and includes informal elements of talk such as the sentence-final expression ka na. In other words, Fujii goes back to his normal speech style after he finishes his “official” report as representative of the administrative meeting. Thus far, the use of the masu form to indicate the official role of the speaker has been exemplified in requests and reports. In coworker meetings, these sequences themselves are considered official; requests and reports are “official” and “planned” actions that are included in the meeting minutes or agenda. The use of the masu form in these sequences distances the speaker from potential imposition by emphasizing the official role of the speaker.

 Naomi Geyer

11. Masu form as official frame5 This section examines the use of the masu form as an official frame of discourse and utterance. First, let us observe its use within the topic boundary.6 Generally, the chairperson of a meeting explicitly states a major topic shift utilizing the masu form. Even when the participants are engaged in a speech activity conducted entirely in the plain form, the chairperson switches to the masu form and initiates a topic shift. The next excerpt, which directly follows the joking segment introduced in excerpt (1), demonstrates this shift. Kishi’s last statement in the joking segment (1) is repeated here. (11) 1 Kishi:

aa? (.) soo ka (.) soryaa sugoi na [plain] ‘Oh, is that right? That’s great!’

2 Others: ((laugh)) 3→ Seki:

jaa kyoo hoshuu wa goka (..) to iu

4

koto desu ne (1) hai jaa shinro kankee

5

chotto owatte,= [masu] ‘Then today we decided that supplementary lessons should be given for five subjects, right? OK, then, we finish with matters related to senior guidance and...’

6 Miki

=hai shinro jikan torasechatte. ‘Yes, the senior guidance matters took a lot of time.’

7 Seki

ato (.) hitotsu chotto seekatsushidoo kankei ga

8→

sugoku konogoro kininattemashite::, (..) ‘And another thing is, I am quite worried about student-liferelated issues recently, and...’

5. The term “frame” has been defined by researchers in a variety of ways (e.g., Bateson 1972; Goffman 1974; Gumperz 1982; Tannen 1984, 1993). Following Tannen’s (1984) definition of frame as “a superordinate message about how the communication is intended” (p. 23), the “official frame” in this study refers to the superordinate message about the “official” nuance of the utterance. An official nuance is similar to formality, which is “commonly associated with negative politeness” (O’Driscoll 1996: 26). The formal/informal distinction, however, tends to be directly related to conversational setting (i.e., formal vs. informal setting). In order to emphasize that the style shift occurs within a single conversational setting and that the setting, therefore, is not the only determinant variable, the term “official frame” is used here. 6. The concept of topic in discourse has been examined in different traditions, as “paragraph” (Hinds 1977, 1978), “theme” (Maynard 1986), “episode” (Van Dijk 1982), “discourse topic” (Keenan & Schieffelin 1975), and “conversational topic” (Shuy 1986, Yamada 1992). The definition of topic in this study follows that of “conversational topic,” which corresponds to a rather “intuitive gauge that might respond to a nonparticipant’s question: ‘What are you talking about?’” (Yamada 1992: 23)



Interpersonal functions of style shift 

As mentioned previously, initiating a topic shift can be considered face threatening (Brown & Levinson 1987). When participants are engaged in interactional acts such as joking, a topic shift can be imposing. Here, the speaker uses the masu form to indicate that her utterance initiates a shift to the main topic on the agenda. In other words, her utterance provides an official framework for the meeting. A metacommunicative expression,...kankei ‘issue/topic related to...’, is used to mark an explicit topic shift, enhancing the official nuance of the utterance. Ikuta (1983) observes a similar use of the masu form in her TV talk show data. After expressing a positive remark in the plain form, the talk show host asks another question in the masu form. Ikuta maintains that the masu form often signals “utterances which are in a higher position of the hierarchy of utterances in a discourse.” (1983: 39) The notion of discourse hierarchy can be applied to the present analysis of multi-party institutional discourse. Within the discourse hierarchy of faculty meetings, the official framework of the faculty meeting can be considered higher and other peripheral talk can be considered lower. This notion, in conjunction with other social factors, can explain seemingly contradictory phenomena involved in style shift at topic boundaries. Previously, excerpt (7) illustrated a situation in which the imposition of topic shift was mitigated by the use of the plain form occurring within a sequence in which the speaker abruptly remembered something she wanted to say. On the other hand, excerpt (11) demonstrates the use of masu forms at a topic boundary. To answer the question why two opposite forms are employed to achieve the same effect, we have to examine the nature of the topics and the role of the person who proposes the topic shifts. In excerpt (7), the topic to be introduced was a report related to the previous talk, and a participant without any particular institutional role initiated the new topic. In other words, it was a minor topic shift. In contrast, the topic shift initiated in excerpt (11) is a major one issued by the chairperson, a shift for which an official tone is more suitable. Excerpt (2) contains another example of the masu form used at the topic boundary. When Maeda presents a joke in the plain form, “onnanoko wa tsuyoi nee (Girls are strong, aren’t they?),” the head teacher Akita responds, otoko wa yowai desu ne ‘men are weak, aren’t they?’, in the masu form before he introduces a new topic. Even though the content of Akita’s response continues Maeda’s previous joke, it does not lead to an elaboration of the joke. The use of the masu form in this sequence marks the shift to an official mode, and thus acts as a stepping stone for the topic change. Next we will examine various combinations of the plain and masu forms within an utterance. A plain-form sequence is frequently embedded within a larger masu-form structure. There are two types of plain/masu combinations. The first type is an inserted sequence in the plain form, which appears within an overall statement in the masu form. These inserts exhibit monologic characteristics, as observed in excerpts (6) and (7). The inserted plain-form sequence is produced swiftly in a soft tone and the speaker immediately resumes a normal tone and speed after the insertion. In these examples, a



Naomi Geyer

plain-form segment is inserted as a subsidiary comment in the overall “official” talk, which is conducted in the masu form. The second type of combination, more frequently observed in the data, is quotation, in which plain-form utterances are embedded in a masu-form structure. The quoted sequences in the data set usually express the speaker’s affect and opinions. By quoting their own feelings and opinions, speakers can express their sentiments and opinions in a less threatening way and adjust the level of formality and mitigation toward the end of their utterances. In the quoted sequence, a speaker first expresses candid personal feelings and opinions. The plain form adds to the unofficial and offthe-record nuance of the utterance. Then, toward the end of the utterance, the speaker can frame the whole utterance as an official one by employing a masu form ending. Moreover, the speaker can adjust the force of the statement by a variety of other utterance-final hedging options. In the simplest form of such a combination, one plain-form sequence is embedded in an overall masu-form structure. Consider the next excerpt, which occurred after a discussion of a difficult situation – a situation in which a group of troubled students had become uncontrollable. Egawa, a young male teacher, expresses his exasperation and states that the only thing the teachers can depend on is “outside forces” – i.e., the police. (12) 1 Egawa: moo gaibu ni tayoru shika nai no kanaa [plain] 2

tte iu kanji desu nee, [masu] ‘I wonder if there is no other way but depending on outside forces.’

This type of quotation structure exhibits stylistic differences from the above-mentioned monologic insertions. First, the monologic insertions were produced at a high speed and in a low tone, while the plain-form sequence in excerpt (12) is not. Second, the plain-form sequence in this example is syntactically embedded within the larger structure by the quotation marker tte iu ‘saying that’, whereas the monologic insertions did not follow this pattern. In excerpt (12), the use of a plain form and an affective final particle together enables the speaker to express his vivid feelings, even though the utterance is framed in an overall masu-form structure. In her analysis of style mixtures in dialogues of fiction, Maynard maintains that the use of “naked” plain forms (i.e., plain forms without accompanying final particles) frames the utterance as background information for the other foregrounded utterances phrased in the masu form and that it produces “the impression that the speaker was present there and then at the scene of the event and has the immediacy of spontaneous expression” (1993: 168). Maynard’s analysis is applicable in the present analysis – the plain form adds a vivid immediacy to the utterance.7 7. Maynard examined “naked” plain forms without any accompanying final particles, but most of the plain-form sequences under discussion here occur with some final particles in the present data.



Interpersonal functions of style shift 

Brown and Levinson (1987) make a similar observation about direct quotations in English narratives, maintaining that direct quotation is a device to make a story vivid and to pull the listener deeper into the narrative flow. In her examination of direct reported speech (i.e., reproducing someone else’s or a speaker’s former locution or thought), Holt (1996) maintains that direct reported speech, in addition to making the story more vivid and involving, functions to provide evidence by enabling the recipient to witness it for himself or herself. Likewise, by presenting his or her sentiments in a direct quotation form, the speaker draws the hearer closer to the speaker’s own feelings. More than one quotation in plain form can be embedded into an overall masuform structure. Before excerpt (13), teachers had been discussing how they might help students organize the seetokai, a student organization equivalent to a student council. At the time of the discussion, the student council was a passive organization in which student activities were limited to bureaucratic routines. Kaji, a middle-aged male teacher, thinks that the teachers have to show clear intentions of revitalizing the council in order to attract talented and ambitious students. He states his opinion in two direct quotes that are embedded in a masu-form structure. (13) 1 Kaji:

...senseetachi no ikoo ga denaito nakanaka

2→

ugokenai tte iu no ga aru kanaa toka (. )[plain]

3→

sokorahen o kangaenaito:: umaku hikitsukerarenai

4→

no kanaa [plain] tte iu ki ga suru n desu kedo. [masu] ‘I have a feeling that there may be an issue like, that the students will not easily move unless the teachers’ intentions are shown.., if we do not think about this kind of thing, we cannot attract the students very well.’

He first states that the students would not move easily if the teachers do not show their intentions (line 1–2). Then, in lines 3 and 4, he rephrases the same opinion, saying that if the teachers do not “think about this kind of thing,” they cannot attract the students to apply to become members of the council. Both statements utilize the plain form and the affective final particles ka and naa. At the end of the turn, he frames the two abovementioned statements as quoted clauses modifying ki ‘feeling’ by using the quotation markers, and closes the turn with a masu form of the copula. Structurally, the above-mentioned sequence resembles a direct quotation. However, pragmatically, the moment-by-moment analysis of the same sequence reveals different characteristics. In this type of structure, the speaker can express vivid emotions or candid opinions as unofficial and off-the-record statements by using the plain form. Toward the end of the utterance, the speaker can add a masu-form frame to the utterance, so that the whole utterance is framed as an official statement. In fact, a variety of hedging forms and structures are reserved for utterance-final positions. Therefore, as the utterance unfolds, the speaker can observe the addressees’ reaction and modify the strength of the assertion by utilizing one or more of these forms and



Naomi Geyer

structures. In this way, the combination of the two forms allows the speaker to express vivid sentiments and candid opinions in a less official manner, and to adjust the level of officialness as well as the amount of hedge toward the ends of his or her utterances. In this section, we have examined the combined use of the plain and masu forms. By utilizing a structure similar to direct quotations, the speakers were able to express their sentiments and opinions in a less threatening way and to adjust the level of officialness and mitigation toward the end of their utterances. In the quoted sequence, the plain form adds to the unofficial and off-the-record nuance of the utterance. It may also function to create solidarity by expressing candid personal feelings and opinions. Toward the end of the utterance, a speaker can frame the whole utterance as an official one by employing a masu-form ending. At the same time, the speaker can adjust the force of the statement by means of a variety of other utterance-final hedging options.

12. Masu form as solidarity marker So far, the masu form has only been associated with different negative politeness strategies. However, in the next example, use of the masu form, combined with other speech elements, appears to enhance the solidarity between participants. The topic of the conversation segment in excerpt (14) is the ongoing guidance sessions in which teachers consult with students individually concerning higher education and employment opportunities. Such guidance is a major task for the teachers who are in charge of ninth graders since choosing a suitable high school for the students is of primary concern not only for the students but also for their parents and school administrators. In performing this important duty, teachers are expected to follow a number of principles enforced by the local Board of Education. These principles concern the materials teachers should, or should not, use in the guidance sessions and the amount of information they should reveal to students and parents. They also prescribe the positive attitude that teachers must display in the guidance sessions. Earlier in the meeting, the teachers talked about the guidance sessions casually. Ooki, one of the homeroom teachers, stated that he frankly told one of his students that the high school the student wished to enter was too difficult for him. In excerpt (14), which occurs later in the same meeting, the head teacher Kato reminds Ooki of the positive attitude required by the Board of Education’s policies.8 (14) 1 Kato:

dame desu yo tte iu no wa chotto ima no shinro-

2→

shidoo de wa ºitchaº-ikenai kotoni haha nattemasu

3

yo ne::, ooki sensee= [masu] ‘It has been decided that we mustn’t say “it’s hopeless” to the students at the guidance session now, right, Mr. Ooki?’

8. See Geyer (2008) for more detailed analysis of this excerpt.



Interpersonal functions of style shift 

4 Ooki:

=ha:i [hai ((laughing tone)) ºganba[roo:: tteº‘Yes, yes, we should say “do your best.”’

5 Kato:

[ne, hehehehe. sensee? [takada ga hachiooji

6→

toka ittara ganbare tte iu shika-nai n desu yo ne?

7→

hehe (.) sokontoko yoroshiku onegai[shimasu. [masu] ‘Right, hehehehe Mr.? If Takada says something like Hachiooji High School, all we can say is “do your best,” right? Please go along with that.’

8 → Ooki:

[ha::i. wakarimashita ((laughing tone)) [masu] ‘Yeees. I understand.’

In lines 1–3, Kato refers to the rule that the teachers are not supposed to dismiss as impossible the wishes of students concerning high schools. This statement is delivered in a masu-form structure, indicating the official nature of the utterance. The expression itcha ikenai ‘mustn’t say’ is delivered in a mocking tone, with special emphasis on ikenai ‘mustn’t’. The final particle ne is stressed and delivered in a mocking tone as well. The speaker employs the final particle to solicit agreement from Ooki, whose name was mentioned in the turn-final position. The stressed expressions and the laughter observed in this sequence appear to mark an implicit shared knowledge. In the subsequent turn in line 4, Ooki presents his compliance by using the repeated agreement token hai ‘yes’, and an elaboration that the teachers should use ganbaroo ‘do your best’, instead of the discouraging dame desu yo ‘it is hopeless’, which was presented in the previous turn. The laughing tone, as well as the agreement token with an elongated vowel, display the speaker’s reception of the metamessage. Overlapping the previous turn, Kato confirms the understanding with the particle ne, laughter, and an address term sensee ‘teacher’. In the following part of the same turn, Kato refers to a specific incident in which Ooki diverged from the prescribed norm (lines 5–6). Takada is one of the students in Ooki’s class who wishes to enter Hachiooji High School, a fairly prestigious high school that is beyond his reach. Given the student’s grades and test results, he has only a slim chance of being able to enter the high school of his choice. Kato reminds Ooki that even in hopeless cases, all teachers are supposed to say ganbare ‘do your best’. Kato accuses Ooki of giving remarks that were too straightforward and therefore “not positive.” Her accusation is framed as an official one through the use of the masu form. An accusation, especially a formal one within a meeting, is a highly weighted face-threatening act, which invites mitigation. In this example, however, the accusation is not mitigated by any linguistic means. In line 8, Ooki’s response of compliance is proffered with the masu form of a verb, the agreement marker hai with an elongated vowel, and a laughing tone. In earlier examples, the masu form invested the utterance with an official tone and expressed deference toward the interlocutor. The utterances in this example demonstrate



Naomi Geyer

a different function. The masu-form statements are regularly accompanied by other elements of speech, such as the final particle ne, laughter, and a laughing or mocking tone. The mutual use of the masu form in combination with these elements suggests a metamessage, i.e., a shared sentiment toward the enforced principle. Not dismissing a student’s wants is an idealistic policy, not a realistic guideline. Teachers, including Kato and Ooki, know that students’ wishes are sometimes unrealistically high, and not being able to say “impossible” is too idealistic. In this context, Kato’s formal accusation is a reminder of what the enforced norm is. By employing the masu form with the other elements discussed above, both Kato and Ooki display their awareness that the principle is to be followed officially, while they simultaneously send the metamessage that they both feel the norm is sometimes unrealistic. Thus, the mutual use of the masu form, along with other strategies, creates in this example a theatrical effect in which both interactants act out the roles of the accusing and accused.

13. Conclusion This study has described the ways in which the interpersonal functions of the plain and masu forms are created in interactions. In the present data set, the plain form was used to create two distinct but interrelated functions: to mark solidarity among interlocutors and to mitigate the face threat of performed actions. The core linguistic property of the plain form works jointly with other sequential factors to achieve these pragmatic functions. These factors include speech actions, linguistic elements such as affect keys, and turn organizational features. Even though the two functions of the plain form were analyzed separately for the sake of clarity, the two functions are not mutually exclusive categories. A single placement of the plain form can work both as a solidarity marker and as a mitigation marker. The “unofficial” nuance created by the plain form can be related to both of these functions simultaneously. In the same vein, the functions of the masu form can also be related to each other. The examples in this study show how the use of the masu form impersonalizes the speaker’s role in various types of talk. By emphasizing the official nature of the talk, it distances the speaker from the imposition caused by the performed face-threatening acts. When the masu form was used in interactions between unequals, the “official framing” led to another function – addressee deference. Even in a solidarity-building activity, where interlocutors tend to use the plain form, the young female teachers continually employed the masu form when addressing the older head teacher. The mutual use of the masu form observed throughout the data can also be seen as an indicator of addressee deference. The speech setting of the present data – coworker meetings – accounts for the predominant use of the masu form. In these meetings, speakers constantly show respect to others as coworkers, utilizing various linguistic means, one of which is the masu form.



Interpersonal functions of style shift 

The study also examined cases in which speakers used a combination of plain and masu forms. By employing a structure similar to that of direct quotation, speakers utilized both forms to express their feelings and opinions. The content of the utterance is presented first in the plain form, which invests the utterance with an unofficial, friendly, and/or vivid nuance. As the utterance unfolds, speakers may eventually frame the utterance with the masu form, using an appropriate “dose” of hedging. This structure enables speakers to adjust their speech according to the verbal and nonverbal reactions of the addressee. Finally, the study described cases in which a mutual use of the masu form with other linguistic and paralinguistic elements marks the shared understanding of certain metamessages and, in turn, enhances solidarity. The mutual use of the masu form creates a theatrical effect in which both interactants jointly act out their respective roles. This study situated the style shifts within the framework of politeness studies and described how the core properties of the plain and masu forms interact with other factors in discourse to achieve diverse functions. I hope that these insights will shed new light on the intricate mechanisms of style shifts.

References Arundale, R. 1999. “An alternative model and ideology of communication for an alternative to politeness theory.” Pragmatics 9 (1), 119–153. Bateson, G. 1972. Steps to an Ecology of Mind. New York: Ballantine. Boxer, D. 1993. “Social distance and speech behavior: The case of indirect complaints.” Journal of Pragmatics 19 (2), 103–125. Brown, G. and Yule, G. 1983. Discourse Analysis. Cambridge: Cambridge University Press. Brown, P., and Levinson, S. 1987. Politeness: Some universals in Language Usage. Cambridge: Cambridge University Press. Cook, H. M. 1996. “Japanese language socialization: Indexing the modes of self.” Discourse Processes 22, 171–197. Cook, H. M. 1997. “The role of the Japanese masu form in caregiver-child conversation.” Journal of Pragmatics 28, 695–718. Cook, H. M. 1999. “Situational meanings of Japanese social deixis: The mixed use of the masu and plain forms.” Journal of Linguistic Anthropology 8 (1), 87–110. Eelen, G. 1999. “Politeness and ideology: A critical review.” Pragmatics 9 (1), 163–173. Eelen, G. 2001. A Critique of Politeness Theories. Manchester: St. Jerome Publishing. Fraser, B. 1980. “Conversational mitigation.” Journal of Pragmatics 4, 341–350. Fraser, B. 1990. “Perspectives on politeness.” Journal of Pragmatics 14 (2), 219–236. Geyer, N. 2008. Discourse and Politeness: Ambivalent Face in Japanese. London: Continuum. Goffman, E. 1974. Frame Analysis. New York: Harper & Row. Gu, Y. 1990. “Politeness phenomena in modern Chinese.” Journal of Pragmatics 14 (2): 237–257. Gumperz, J. 1982. Discourse Strategies. Cambridge: Cambridge University Press.



Naomi Geyer Hill, B., Ide, S., Ikuta, S., Kawasaki, A. and Ogino, T. 1986. “Universals of linguistic politeness: Quantitative evidence from Japanese and American English.” Journal of Pragmatics 10, 347–371. Hinds, J. 1977. “Paragraph structure and pronominalization.” Papers in Linguistics 10, 77–99. Hinds, J. 1978. “Conversational structure: An investigation based on Japanese interview discourse.” In Problems in Japanese Syntax and Semantics, J. Hinds, & I. Howard (eds), 79–121. Tokyo: Kaitakusha. Holt, E. 1996. “Reporting on talk: The use of direct reported speech in conversation.” Research on Language and Social Interaction 29 (3), 219–245. Ide, S. 1989. “Formal forms and discernment: Two neglected aspects of linguistic politeness.” Multilingua 8, 223–248. Ide, S., Hill, B., Carnes, Y., Ogino, T. and Kawasaki, A. 1992. “The concept of politeness: An empirical study of American English and Japanese.” In Politeness in Language: Studies in Its History, Theory and Practice, R. Watts, S. Ide, & K. Ehlich (eds), 21–41. New York: Mouton de Gruyter. Ikuta, S. 1983. “Speech level shift and conversational strategy in Japanese discourse.” Language Sciences 5, 37–53. Keenan, E. O., and Schieffelin, B. B. 1975. “Topic as a discourse notion: A study of topic in the conversations of children and adults.” In Subject and Topic, C. N. Li (ed), 335–384. New York: Academic Press. Kindaichi, H. 1982. Nihongo Seminaa [A Seminar on the Japanese Language]. Tokyo: Kenkyuusha. Lakoff, R. T. 1973. “The logic of politeness, or minding your p’s and q’s.” Chicago Linguistics Society 9, 292–305. Leech, G. 1983. Principles of Pragmatics. London: Longman. Lyons, J. 1977. Semantics. Cambridge: Cambridge University Press. Makino, S. 1983. “Speaker/listener-orientation and formality marking in Japanese.” Gengo Kenkyuu 84, 126–145. Makino, S. 2002. “When does communication turn mentally inward?: A case study of Japanese formal-to-informal switching.” In Japanese/Korean Linguistics, 10, N. Akatsuka and S. Strauss (eds), 121–135. Stanford: CSLI Publications. Mao, L. 1994. “Beyond politeness theory: ‘Face’ revisited and renewed.” Journal of Pragmatics 21, 451–486. Matsumoto, Y. 1988. “Reexamination of the universality of face: Politeness phenomena in Japanese.” Journal of Pragmatics 12, 403–426. Matsumoto, Y. 1989. “Politeness and conversational universals: Observations from Japanese.” Multilingua 8, 207–221. Maynard, S. 1986. “Interactional aspects of thematic progression in English casual conversation.” Text 6 (1), 73–105. Maynard, S. 1993. Discourse Modality: Subjectivity, Emotion and Voice in the Japanese Language. Amsterdam: John Benjamins. Mills, S. 2003. Gender and Politeness. Cambridge: Cambridge University Press. Nwoye, O. G. 1992. “Linguistic politeness and socio-cultural variations of the notion of face.” Journal of Pragmatics 18, 309–328.



Interpersonal functions of style shift  Ochs, E. 1990. “Indexicality and socialization.” In Cultural Psychology: Essays on Comparative Human Development, J. W. Stigler, R. A. Schwerder and G. Herdt (eds), 283–308. Cambridge: Cambridge University Press. Ochs, E. 1993. “Indexing gender.” In Sex and gender hierarchies, B. D. Miller (ed), 146–169. Cambridge: Cambridge University Press. Okamoto, S. 1997. “Social context, linguistic ideology, and indexical expressions in Japanese.” Journal of Pragmatics 28 (6), 795–817. Okamoto, S. 1998. “The use and non-use of honorifics in sales talk in Kyoto and Osaka: Are they rude or friendly?” In Japanese/Korean linguistics, 7, N. Akatsuka, H. Hoji, S. Iwasaki, S. Sohn and S. Strauss (eds), 141–157. Stanford: CSLI Publications. Okamoto, S. 1999. “Situated politeness: Coordinating honorific and non-honorific expressions.” Pragmatics 9 (1), 51–74. Pan, Y. 1995. “Power behind linguistic behavior: Analysis of politeness phenomena in Chinese official settings.” Journal of Language and Social Psychology 14 (4), 462–481. Shibatani, M. 1990. The Languages of Japan. Cambridge: Cambridge University Press. Shuy, W. 1982. “Topic as a unit of analysis in a criminal law case.” In Analyzing Discourse: Text and Talk, D. Tannen (ed), 113–126. Washington, DC: Georgetown University Press. Sifianou, M. 1992. “The use of diminutives in expressing politeness: Modern Greek versus English.” Journal of Pragmatics 17, 155–173. Sifianou, M. 1993. “Off-record indirectness and the notion of imposition.” Multilingua 12 (1), 69–79. Sifianou, M. 1995. “Do we need to be silent to be extremely polite?: Silence and FTAs.” International Journal of Applied Linguistics 5 (1), 95–110. Tannen, D. 1984. Conversational Style: Analyzing Talk among Friends. Norwood, NJ: Ablex. Tannen, D. 1990. You Just Don’t Understand: Women and Men in Conversational Discourse. Cambridge: Cambridge University Press. Tannen, D. 1993. Framing in Discourse. New York: Oxford University Press. Tannen, D. 1994. Gender and Discourse. New York: Oxford University Press. Van De Walle, L. 1993. Pragmatics and Classical Sanskrit: A Pilot Study in Linguistic Politeness. Amsterdam: John Benjamins. Van Dijk, T. A. 1982. “Episodes as units of discourse analysis.” In Analyzing Discourse: Text and Talk, D. Tannen (ed), 177–95. Washington, DC: Georgetown University Press. Watts, R. J. 2003. Politeness. Cambridge: Cambridge University Press. Watts, R. J., Ide, S. and Ehlich, K. (eds.). 1992. Politeness in Language: Studies in Its History, Theory and Practice. New York: Mouton de Gruyter. Wierzbicka, A. 1985. “Different cultures, different languages, different speech acts.” Journal of Pragmatics 9, 145–161. Wolfson, N. 1988. “The bulge: A theory of speech behavior and social distance.” In Second Language Discourse: A Textbook of Current Research, J. Fine (ed), 21–38. Norwood, NJ: Ablex. Wolfson, N. 1989. Perspectives: Sociolinguistics and TESOL. Cambridge, MA: Newbury House. Yamada, H. 1992. American and Japanese Business Discourse: A Comparison of Interactional Styles. Norwood, NJ: Ablex.



Naomi Geyer

Appendix Transcription Conventions (Adapted from Atkinson & Heritage 1984) [ = (0) (.) . , ? underline : °° hhh ˙hhh (( )) <> (segment) ()

the point where overlap begins latched utterance intervals within and between utterances a short untimed pause within an utterance a stopping fall in tone a continuing intonation a rising intonation a halting or an abrupt cutoff of sound an emphatic stress lengthened vowel sound (extra colons indicate greater lengthening) spoken softly aspirations inhalations comments on quality of speech and context spoken quickly uncertain transcription transcription impossible

Speech style shift as an interactional discourse strategy The use and non-use of desu/-masu in Japanese conversational interviews* Shoko Ikuta

Meiji Gakuin University

1. Introduction This study is an attempt to reveal part of the multi-functionality of the so-called Japanese “addressee honorifics,” desu and -masu, and the shift between the use and nonuse of these forms in interactional discourse. Rather than the social meaning of these forms, this study will focus on the discourse function of the shift between the two styles and hopefully demonstrate how such a shift contributes to construct a collaborative discourse. Through the analysis of conversational data, it has been revealed that desu/-masu are not simply politeness or formality markers, and what the use and non-use of desu/masu indexes in particular social or psychological contexts has been explored in detail (Cook 1999, Ikuta 1983, Maynard 1991, Sukle 1994). The shifting between the use and non-use of desu/-masu has also been analyzed in relation to discourse construction (Ikuta 1983, Maynard 1991). In Ikuta (1983), such style shifting in interviews is analyzed, along with its social and attitudinal/psychological meanings, as introducing a supportive/illustrative “context space,” adopting the study of coherence relations among context spaces by Reichman (1978). Context space refers to “a series of utterances that taken together constitute a unit or whole, such as the event-recounting portion of a narrative” (McLaughlin

* An earlier version of this paper was read at the Japanese Speech Style Shift Symposium held at the University of Arizona, March 7-10, 2002. I am grateful to the participants in the symposium for their comments and criticisms. Especially, my thanks go to Kimberly Jones and Yoshi Ono who read the draft and provided valuable comments and editing. I am also very grateful to an anonymous reviewer for his or her helpful comments and to John A. Bisazza for correcting my English. All errors and shortcomings remain my own.



Shoko Ikuta

1984: 271). The shifts in such capacity, however, are limited to a single turn—i.e., a single speaker’s narrative within a conversation. In this chapter, I will examine the shift between the use and non-use of desu/-masu forms, employed primarily as an interactional discourse strategy, in interview-style speech. More specifically, this study will focus on the discourse function of shifting between the two styles, examining those occurring not only within a single turn or a single speaker’s narrative but also those occurring across exchanges. It will discuss the function of such shifts as a marker of embedding a subspace in a larger context space, which may be regarded as a type of discourse marker, and will demonstrate how such shifting corresponds to the shifting of the interactional roles of the participants at the given moment of an utterance and thus operates as a “floor management” strategy, contributing toward jointly constructing a discourse.

2. Data The segments of conversation analyzed in this chapter were extracted from five separate interviews, each of which was carried out between two participants, a single interviewer (female) and each of the five interviewees (male), who all worked in traditional Japanese occupations.1 They were a rakugo-ka (sit-down comedian), a fishmonger at Tsukiji (the major fish market in Tokyo), a greengrocer, a bathhouse owner, and a chanko (cuisine for sumo wrestlers) restaurant owner. Their ages ranged from the late 20s to the 40s. Table 1 shows the number of occurrences of desu/-masu and non-desu/masu forms observed in the segments of data analyzed, which comprised approximately thirty minutes in total, ranging from six to seven minutes for each conversation. The opening and closing parts of each interview were not included in those segments because of their possible unnaturalness. In the analysis, the desu/-masu or non-desu/-masu forms in quotations, embedded clauses, and before evident ellipses, have been excluded because they do not always reflect the current speaker’s choice of style. Only those that are clearly identified as occurring in predicate-final position are analyzed. (The forms analyzed are underlined in the examples.2) Reflecting the social relationship and situation involved, the unmarked style throughout the conversation is desu/-masu. However, the interviews are done in 1. In the transcription, consonants are based on the Hepburn system, while a long vowel is indicated by doubling the vowel. Word boundaries are marked by a space, following Jorden (1987). An annotated transcription in the Japanese writing system, different from my version, can be found in Mikami and Yamagata (2000). The audio recording tapes are available from the publisher. 2. However, in conversation, there are instances which are hard to distinguish. For example, there are cases which could be interpreted either as occurring in a subordinate clause followed by ellipsis, or occurring in final position with some grammaticalized forms such as ‘kedo,’ as in



Speech style shift in conversational interviews 

Table 1  Number of desu/-masu and non-desu/-masu endings Desu/-masu endings

Non-desu/-masu endings

(a) Rakugo Interviewer Rakugo comedian S



10 (13)* 27 (46)



2 (2) 2 (4)

(b) Fishmonger Interviewer Fishmonger T



10 (10) 32 (39)



6 (7) 6 (19)

(c) Greengrocer Interviewer Greengrocer S



19 (20) 17 (23)



3 (5) 8 (13)

(d) Bathhouse Interviewer Bathhouse owner K



31 (36) 32 (35)



5 (10) 4 (7)

(e) Chanko Interviewer Restaurateur K



24 (27) 21 (43)



4 (6) 3 (6)

Total Interviewers Interviewees



94 (106) 29 (186)



20 (30) 23 (49)

*

The number in parentheses includes all the cases observed, including those which are not strictly predicate-final position.

a casual conversation style, and frequent shifting between the two styles can be observed in each of the interviews (see Table 1). Unlike small-talk type conversation among status equals, these interviews have some significant features that are helpful for the present study, with its focus on discourse organization. For one thing, there are significantly fewer overlaps than in everyday conversation. Also, in spite of the frequent transitions among subtopics, in each the exchange, ‘Onamae wa?’ and ‘Yamada desu kedo.’ The numbers in parentheses in Table 1 include such cases. In the table, we can see some differences among interviewees. For example, the rakugo comedian and the chanko restaurant owner seem to use fewer desu/-masu forms in predicate-final position. As the numbers in parentheses show, they actually use more desu/-masu forms but tend not to mark their utterances as clearly in final form, by often saying '…shimashite’ or ‘… desu kedo’ and the like. The fishmonger used, though only a couple of times, the form ‘ssu,’ a variant of ‘desu,’ as in ‘soo ssu yo’ instead of ‘soo desu yo.’ This is also a different style in a strict sense, but I treated it as a ‘desu’ form here, focusing on the shift between the desu/-masu and non-desu/-masu forms.



Shoko Ikuta

case there is one overall topic—the interviewee’s occupation. Finally, it is the interviewer who initiates the subtopic changes. All of these features help identify the discourse transitions in terms of content, thus facilitating the analysis of the correspondence between the discourse construction and style shifting.

3. Interviews as collaborative discourse In this study, style shift is seen mostly as a strategy to collaboratively construct a discourse. In the interviews examined, various examples of collaboration were observed. The simplest form of such collaboration was syntactic co-construction, as seen in example (1) below. (See Appendix for symbols used in the examples.) (1) (c-1: greengrocer) 1 R: aa soo yuu no ga yappari uresuji na n desu ka ne? ‘Uh-huh, those are the ones, as expected, that sell well.’ 2 S: soo desu ne ato wa kyabetsu to ka ‘Yeah, the rest, cabbages and the like’ 3→ R: aa daitai yoku ureru ‘Ah, (they) mostly sell well.’ 4 S: soo desu ne jibun ga uritai ‘Yeah. I want to sell (them).’ Example (1) illustrates the only occurrence of syntactic co-construction found in the data. In line 3, the interviewer R supplies an ending to S’s preceding utterance, which S did not quite complete in line 2. It is significant that R’s utterance in line 3 has a nondesu/-masu ending within the context of unmarked desu/-masu endings. It is clear that this is a turn that does not hold the “floor.” “Floor,” not an equivalent of “turn,” is defined as “the acknowledged what’s-goingon within a psychological time/space,” and it is possible to take a turn without having the floor as well as to have the floor while one is not talking (Edelsky 1981: 405–6). According to Hayashi (1996: 31), the floor is “a cognitive entity that the interactants jointly create during the course of a conversation.” Other more commonly observed examples of jointly constructed discourse in the interviews are those in which non-floor-holding subspaces are embedded in another’s floor. These correspond to a shift to non-desu/-masu style in a context that is otherwise predominantly desu/-masu. Observed in such cases is an indication of a momentary change in the interactional roles of the participants. Especially on the side of the interviewer, there is a change in “footing” (Goffman 1981) or “participant roles” (Levinson 1988) in accordance with the shift between the desu/-masu and non-desu/-masu styles. “Footing” signals how to hear what is said, i.e., the momentary identity or interactional role of the interlocutors. According to Goffman (1981: 128), “a change in footing



Speech style shift in conversational interviews 

implies a change in the alignment we take up to ourselves and the others present as expressed in the way we manage the production or reception of an utterance.” It has to be taken into account that, unlike small-talk type conversations among close friends or collaborative storytelling among colleagues, interview talk of this kind has a clear-cut role relationship between the interviewer and the interviewee that is understood by both participants. It presupposes the interactional roles of both participants, and the conversation is carried out with each following the appropriate role. However, there is frequent momentary shifting in their roles as discourse proceeds. The interviewer is the seeker of information. She has a clear understanding of the goal of the conversation. She is the one who introduces new topics. She controls the topics and topic changes. However, she does not claim the floor. By introducing a new topic, she yields the floor. The role of the interviewer is basically that of an “emcee.” However, the interviewer occasionally acts as if she were a “prompter.” As an emcee, she opens a new topic, typically by floor-yielding questions in the desu/-masu style; and as a prompter, she takes non-floor-holding and subspace-embedding turns marked by the non-desu/-masu style. On the other hand, the interviewee is the provider of information and the “leading actor,” holding the floor throughout the conversation. He does not even need to negotiate the floor, as the interviewer always yields to him and supports his floor. Most of the interviewees’ predicate endings have unmarked desu/-masu forms, except in some cases in which their utterances end with non-desu/-masu forms, showing affect such as surprise or a momentary change of their attitude toward their interlocutor (cf. Ikuta 1983, Maynard 1991). In what follows, we will see how the interviewee’s floor is given and supported in line with the shifting between the styles. The interviewer supports his floor, because the goal of conversation is having the interviewee talk about the topics that she introduces. She also makes clear what the interviewee is talking about for the potential audience members, who are not present at the site of the conversation. (Whether the speakers are conscious of the audience or not, there is an audience that is not present, since these are interviews recorded for later broadcast.) In so doing, she occasionally “stands in” or “prompts” the leading actor, who is a “lay” actor. However, in the given setting, she is not a mentor or colleague but the host of the interview, so she prompts or stands in without claiming the floor, shifting from the desu/-masu style to the nondesu/-masu style. (These non-desu/-masu endings are not followed by any final particles, which may further support the claim that the utterance does not presuppose the addressee but is to be taken as a “prompt” from backstage.) The following discussion demonstrates how such dual roles of an interviewer are managed by manipulating the shift between the desu/-masu and non-desu/-masu styles.



Shoko Ikuta

4. Style shift as a device for mitigating a face-threat In this section I will examine specifically how the shift to the non-desu/-masu from the desu/-masu style corresponds to the shift to a subspace within a larger interactional space. The use of such style shift as a marker of embedding a subspace also functions as a part of interactional politeness strategy, mitigating a possible face-threat, as discussed below.

4.1

Undertaking a listening/understanding check in a subspace

The utterances with the shift to the non-desu/-masu from the desu/-masu style in the interviews often serve as a “next-turn repair initiator” (Levinson 1983). In example (2), speaker R paraphrases T’s preceding utterance with a non-desu/-masu ending in line 5. (Speaker T, on the other hand, continues using the desu/-masu style when he reacts to R in line 6.) (2) (b-1: fishmonger) 1 R: asa tte yuu yori ka yonaka desu ne ‘It’s in the middle of the night rather than morning, isn’t it?’ 2 T: yonaka desu ne maa ano uchi no kaisha ni iru hitotachi wa minna yappari soo yuu seikatsu shiteru kara ‘It’s in the middle of the night. Well, um, the people in my company have, y’know, that kind of lifestyle, so,’ 3 R: ee ‘Uh-huh.’ T: n::n da kara nakagai-san nanka demo yappari hayai hito wa niji toka sono gurai ni wa kite masu yo ‘mmm, so, even the brokers, the early ones are also, there at two or sometime around then.’ 4

5→ R: a moo ichiba ni niji ni wa tsuite ru ‘Oh, (they’ve) arrived at the market by two o’clock.’ 6 T: un tsuite masu yo ‘Yeah, (they’ve) arrived.’ 7 R: he:: ‘Oh dear!’ 8

sore de sono ato wa doo yuu ichinichi no ugoki ni na(ru) n desu ka ‘And then, what’s your schedule for the day?’

In this particular case, it is possible that the non-desu/-masu form is selected because of R’s surprise about the fact that the workers have arrived at the market by two o’clock in the morning. However, it seems more likely that R’s utterance in line 5 is functioning



Speech style shift in conversational interviews 

to embed a subspace (lines 5 and 6) in a larger context space—T’s description about how early in the morning their day starts at the fish market. (The utterance in line 5 is not a simple repetition of a single word or two, but a paraphrase and summary of T’s preceding utterances, and is followed by T’s reply of confirmation in line 6.) Of course, it is possible that both of these factors are involved here. Example (3) is another example of a confirming utterance with the non-desu/masu ending within a desu/-masu context. In this example, the interviewee T first reacts with non-desu/-masu but immediately closes his space with desu/-masu in the following turn. (3) (b-2: fishmonger) 1 R: anoo yappa ichiba no kata tte yuu to kitto nanka koe toka okkii n deshoo ne ‘Um, y’know, speaking of the market staff, I guess, like, they have loud voices don’t they?’ 2 T: n ma seriba de watashi wa hoete masu ne itsumo ne ‘Um, well, I’m roaring at the auction site, always.’ 3→ R: hoete ru? ‘Roaring?’ 4 T: hoete ru ‘Roaring.’ 5

R: @

6 T: urenai to toku ni hoete masu yo ‘I’m roaring especially when (the fish) aren’t selling well.’ R’s utterance in line 3 is a simple repetition of the word used in the immediately preceding utterance of her interlocutor T (an echo question) and may also be interpreted as an expression of R’s surprise about T’s use of the word ‘hoeru’ (roar). In this case, T repeats her utterance in the non-desu/-masu form and jointly creates a subspace, which emphasizes the issue in the ongoing space, but T, who keeps holding the floor, subsequently shifts back to the desu/-masu style and closes the space by providing more detail. These cases illustrate that shifting to the non-desu/-masu style is seen not only in a single turn (self-embedding) but also in exchanges between the interlocutors. These two-part sequences generally do not construct a side sequence, in which the speakers detour from the ongoing context space to rectify or clarify some tangential issue raised by that space. Rather, they work to construct a subspace embedded in the ongoing context space. Thus, the relationship between these sequences is not linear but hierarchical. The utterance containing the shift to the non-desu/-masu style contributes directly to the discourse construction of the co-participant’s ongoing floor. Without interrupting the ongoing flow of talk, it functions to make the discourse more comprehensible and more coherent. Also, such a shift does not seem to change the social or psychological



Shoko Ikuta

relationship between the interlocutors, even momentarily. This function as a discourse transition cue to signal context space changes may be seen as similar to that of a number of verbal cues called “discourse markers” (Fraser 1999, Schiffrin 1987). Also, style shift may be considered similar to a tense shift, which may at times coincide with a transition from one context space to another (Reichman 1978). These markers ease the co-participant’s processing task. As do discourse markers and tense shifts, the shift between the desu/-masu and non-desu/-masu styles also works as a marker of a context space shift that contributes to construct a coherent discourse as a whole. Each of the interviewer’s utterances with non-desu/-masu discussed in this section is considered to serve as a simple repair-initiator undertaking a listening or understanding check. The interviewee could simply react to it, and complete his space. Lines 3 and 4 in example 3 are an example of a joint subspace, which consists of a style shift by the interviewer and a following reaction by the interviewee; however, line 3 by the interviewer may also be regarded as an instigator for more information, since the interviewee subsequently shifts back to the desu/-masu style and provides more detail in line 6. The following section discusses such shifts in the utterances instigating more explanation from the interviewee, rather than undertaking a simple listening or understanding check.

4.2

Instigating more information in a subspace

One type of non-desu/-masu utterance commonly observed in the data is a back-channel-like utterance, in which the interviewer repeats a whole or a part of what the interviewee said in the preceding turn, trying to clarify its meaning, or sometimes, as is common in conversation, as a sign of surprise or involvement. Expressing such reaction to the other with the indication of closeness carried by the non-desu/-masu style is also regarded as a sign of positive politeness toward the interviewee. Many such repetitions (other-repetition) observed in the interviews end with non-desu/-masu forms. This type of style shifting functions as an interactional discourse strategy for drawing out more information from the co-participant, rather than just signaling a change of the speaker’s emotional or psychological attitude toward the addressee or the ongoing topic, or indicating the utterance’s subordination to the larger discourse unit in terms of discourse organization. Thus, the non-use of desu/-masu can also be regarded as serving as a marker of a repair initiator. Back-channel utterances are “brief arguments, repetitions, or mirror responses by a listener that are believed to occur primarily during pauses in the turn of the speaker who has the floor; usually characterized by a reduced set of the normal speaker-state signals” (McLaughlin 1984: 270). Back-channel utterances used by the interviewer in the data, typically with non-desu/-masu endings, are considered to function as a “next turn repair-initiator” (Levinson 1983) as well, which encourages the interviewee to elaborate his explanation in the next turn.



Speech style shift in conversational interviews 

Examples (4) and (5) demonstrate non-desu/-masu is used in such utterances, especially with this function. (4) (a-1: rakugo) 1 R: rakugo no miryoku tte yuu to ‘(What is) the attraction of rakugo,’ 2 S: hai ‘Yeah.’ 3 R: doo yuu tokoro desu ka ne? ‘What is it?’ 4 S: soo desu nee ee muda ga nai tokoro da to omoimasu ‘Well, er, I think (the attraction) is that there’s no waste.’ 5→ R: haa muda ga nai ‘Uh-huh, there’s no waste.’ S: ano enjiru gawa ni koo: muda ga nai tokoro ga rakugo no ii tokoro da to omoimasu ‘Um, there’s no useless, er, performance, that’s the charm of rakugo, I think.’ 6

(5) (d-5: bathhouse) 1 K: iya ima wa sukunai desu yo ne ‘No, (such customers) are few now.’ 2→ R: aa sukunai ‘ah, few.’ 3 K: yappa minna ne ofuro ga tsuite masu n de ne ‘Y’know, every house has its bathtub (now), so.’ 4 R: soo desu ne ‘Right.’ The utterances with the non-desu/-masu ending in examples (4) and (5) are not regarded as just a “minimal response” (Fishman 1978), such as “yeah” or “uh-huh.” They seem to be soliciting or instigating the other speaker to elaborate on his talk or to confirm his previous remark(s), as well as showing the interviewer’s interest/involvement in the other’s talk, which can be a form of positive politeness. The former function as a pre-repair-initiator to encourage the other’s elaboration of his prior remarks is performed in a subspace, which is labeled with the non-desu/-masu ending, and not in the main interactional space. Line 7 in example (6), where the chanko restaurant owner is talking about how his customers misunderstand the way to eat chanko soup, is another example of a backchannel utterance with a similar function. The utterance, however, is not a simple repetition but a paraphrase of the interviewee’s utterance in line 6.



Shoko Ikuta

(6) (e-2: chanko) 1 R: aa omise no kata ga suupu o totte oite kudasai tte yuu no wa ato de ojiya ni shitari suru kara ‘Ah, when the restaurant staff says “please put the broth aside,” (it means) it’ll be used for rice gruel later, so’ 2 K: ee nokoshite oite kudasai [to] ‘yeah, “please save it,”’ 3 R: [ee] ‘Uh-huh.’ 4 K: yuu imi na n desu ga ‘(that’s what) it means but,’ 5 R: ee ‘Uh-huh’ 6 K: kanchigai nasatte, zenbu sukutchau n desu yo ‘(the customers) misunderstand it, and they ladle all (of the broth) in.’ 7→ R: aa torinozokete shimau. ‘Ah, (they) take it out (of the pan).’ 8 K: sore de ojiya nante iwarete mo ne are suupu ga nai n desu kedo tte ‘and then, even if they ask for rice gruel, (I’d say) “oh, here’s no broth.”’ 9 R: @ soo [ka]3 ‘@ A-ha.’ 10 K: [ee] ‘yeah.’ 11 R: suupu zenbu nonjattari shitara dame na n desu ne ‘It’s no good to drink up all the broth or anything, huh?’ The speaker R uses a non-desu/-masu ending in line 7, as she paraphrases what K said and clarifies it. This utterance could be considered to be her just musing to herself, but considering other examples such as those examined earlier, it seems more likely that the interviewer is doing her job by making explicit what her interlocutor meant. She may be doing this for the sake of the imagined audience, using a verb which is more clearly understandable by the not-on-the-scene audience than the verb K used in the preceding turn. This can be regarded as a repair by itself (other repair), rather than as an initiator of a repair (self repair). In so doing she uses the non-desu/-masu style, which is significant. She uses non-desu/-masu here because the utterance is creating 3. This use of non-desu/-masu shows that the utterance is not directed toward the addressee but to the speaker herself.



Speech style shift in conversational interviews 

not a new context space (there is no topic change), but rather, a subspace embedded in the ongoing space, in which speaker K has the floor. This is also considered to be a kind of back-channel utterance, but speaker K continues his talk without overtly reacting to it. This is differentiated from examples (4) and (5), in which R’s non-desu/-masu utterances (first pair part) are responded to by the interviewee (second pair part, which is a reaction to the R’s first pair part, and not a continuation of his previous utterances). In the three cases discussed above, the turn containing repetition or paraphrase begins with either ‘aa’ or ‘haa,’ indicating the speaker’s acknowledgement and involvement. Each of these three utterances consists of a brief repetition or paraphrase with the non-desu/-masu ending, and none of them has a sentence-final particle, such as ka or ne. Thus, they are not turn-giving utterances. In other words, the other interlocutor is not forced to overtly react to them. They are similar to back-channel utterances in that they are non-floor-claiming, but, especially in that they are paraphrasing, i.e., by providing the speaker’s interpretation, they help to clarify the preceding utterance by the co-participant, and to facilitate the completion of his episode. The following discussion demonstrates how such style shift works to avoid or mitigate a possible face-threat when the interviewer makes a challenge by supplementing a new point to the interviewee’s talk, or occasionally stands in for the interviewee’s part.

4.3

Making a challenge in a subspace

In example (7), where a bathhouse owner is being interviewed, two cases of shifting to non-desu/-masu are observed. (7) (d-3:bathhouse) 1 R: nani ka ano ofuroya-san o yatte ite taihen na koto kuroo banashi to ka arimasu ka. ‘Have you had anything, er, terrible or any tough experience while you’ve run your bathhouse?’ 2 K: iya arimashita ne ikkai ne nanka shoogakusee ga ‘Well, I had, once. Well, a grade-schooler,’ 3 R: hai ‘Yes.’ 4 K: ofuro no naka ni shanpuu irechatte ‘poured shampoo into the bathtub,’ 5 R: ee: ‘Oh, really.’ 6 K: ee de awa darake ni natchatte ‘yeah, and, (the bathtub) got filled with bubbles, and,’



Shoko Ikuta

7 R: ee:. ‘Oh.’ K: moo doo shiyoo mo nai mizu tashite mo awa kienai desu moo yappa tochuu de chotto yamete zenbu nuite mata irenaoshimashita yo sono toki wa ‘(there’s) nothing I could do. Even if you add water, bubbles don’t go away. At that time, y’know, (I) quit (adding water) in the middle and drained all (the water from the bathtub) and refilled (it).’ 8

9 R: aa soo desu ka ‘Oh, I see. 10→

e sore wa kodomo wa warugi ga nakute yatta ‘Well, that, that child didn’t mean any harm.’

11 K: sore wa soo desu yo ne ee ‘No, of course not.’ The unmarked style in this context is again desu/-masu, reflecting the social situation. The phrase ‘moo doo shiyoo mo nai ‘(there’s) nothing I could do’ in line 8 is a shift within a single speaker narrative. It occurs when the owner is talking about a problem he once experienced. This can be interpreted either as being motivated by his perplexity, or, in terms of discourse organization, indicating the subordination of this part to the whole of his narrative. In fact it is likely to be a result of both of these factors reinforcing each other. The shift in line 10, on the other hand, marks the opening of a new subspace. In line 10, the interviewer R raises a new topic--the issue of the child’s intentionality-which is not contained in K’s prior explanation on the trouble he had to go due to the child. The utterance may even be interpreted as a minor challenge to the other interlocutor, and may threaten K’s face. (Challenges are categorized as face-threatening acts, which intrinsically threaten the addressee’s face [Brown & Levinson 1987: 66].) R’s utterance, however, is marked with the non-desu/-masu ending as different from a topic-opening utterance by the interviewer, which would end with a desu/-masu form (as seen in line 8 of example 2). Instead of opening a new space/topic, line 10 is disguised as an embedded subspace that is simply adding some more detail to the ongoing main interactional space, and subsequently her interlocutor concedes to the point she made. In this manner, the interviewer R mitigates a possible face-threat and successfully manages her role as a floor supporter while, for the sake of the imaginary audience, supplementing her interviewee’s talk with a new point.

4.4

Standing in for the interviewee in a subspace

In example (8), the speaker K, a bathhouse owner, talks about the ‘furonto keeshiki’ (front desk reception style) of his bathhouse.



Speech style shift in conversational interviews 

(8) (d-1: bathhouse) 1 K: uchi wa anoo furonto keeshiki de ‘We have, er, a front desk style, so,’ 2 R: hai ‘Yes.’ K: machiaishitsu ga atte de furonto de dakara hadaka no tsukiai wa shite nai n desu ‘there’s a waiting room, and a front desk, so, we don’t have any interaction with unclothed people.’ 3

4 R: a omise no hito ni wa hadaka wa mirarenai tte yuu ka ‘Ah, well, the bathhouse staff don’t see (the customers) unclothed,’ 5 K: soo na n desu yo ‘That’s right.’ 6 R: a yoofuku o kita jootai de ano ‘er, having (their) clothes on, um,’ 7 K: soo ‘Yeah.’ 8 R: okane o haratte ‘(they) pay the fee, and’ 9 K: hai ‘Yes.’ 10 R: naka ni nugu tokoro wa moo anoo inai wake desu ne ‘there aren’t any (staff) inside, um, where (they) get undressed, are there?’ 11 K: inai wake desu ‘No, there aren’t any.’ 12 R: aa soo na n desu ka nanka ano ‘Oh, is that so. Well, um,’ 13 K: chotto zannen da kedo ne ‘It’s a bit disappointing, though.’ 14 R: @ soo desu ne ‘@ Yeah, it is.’ 15→

ja ano bandai ni tatte anoo otokoyu to onnayu o miwatasu tte kanji de wa nai ‘Then, er, it’s not like (you) stand at the attendant’s counter and look out over men’s and women’s bath areas.’

16 K: arimasen ‘It is not.’



Shoko Ikuta

17 R: aa soo na n desu ka he:: furonto tte yuu n desu ka he:: ‘Oh, I see. Oh, dear! It’s called a front desk, oh dear!’ In line 14, R ends her expression of agreement with the desu/-masu form in response to K’s preceding turn. The utterance by R in line 15 ends with the non-desu/-masu style. (The shift to non-desu/-masu by the speaker K in line 13, on the other hand, is motivated by a different factor, which is not the focus of this study.) Lines 15 and 16 are thus considered a subspace, which adds more detail, embedded within the space from lines 1 to 17, as are lines 4 to 12. Throughout this segment, K keeps the floor, and line 17 is R’s response to the floor as a whole. In this example, we can see how R actively supports K’s floor. In lines 1 and 3, K mentions the ‘furonto keeshiki’ of his bathhouse. What follows then is a more detailed explanation of the system, which is mostly provided by R. However, R’s utterances do not threaten K’s “face” as the leading actor, because they are carried out in a subspace by the use of the non-desu/-masu style.4 In lines 5, 11, and 16, K only replies to R’s utterances, but with the use of desu/-masu it is clear that K is the authority on this topic and that he holds the floor throughout. In such a manner R keeps checking whether her understanding or inferences are accurate, while, as the interviewer, she successfully provides a more detailed explanation for the not-on-the-scene audience. In line 10, R ends her utterance with a desu/-masu ending, but in her following turn in line 12, she makes clear K’s authority (his status as the first-hand information provider) by saying ‘aa, soo na n desu ka.’ Line 15, which R ends with the non-desu/masu form, along with her reaction ‘aa soo na n desu ka’ or ‘he::’ in line 17, makes even clearer R’s role as a prompter or stand-in for the leading actor, i.e., the speaker K. Example (9) is a segment following K’s lengthy episode about some trouble he experienced at his bathhouse, when a child once emptied a bottle of shampoo into the bathtub, described in example (7). (9) (d-4: bathhouse) 1 R: aa nanka ano baburubasu mitai na yoo ni ‘Oh, well, like a bubble bath,’ 2 K: soo soo soo yuu kanji ni natchatta n desu ‘Yeah, yeah, it got like that.’ 3 R: demo katei no ofuro to chigatte ookii ofuro desu mon ne ‘But, unlike bathtubs for homes, (yours) is a huge bathtub, isn’t it, and’

4. Goffman (1967: 5) defines “face” as “the positive social value a person effectively claims for himself by the line others assume he has taken during a particular contact” and claims that an interactant tries to maintain both his or her own face and the face of other participants during an interaction.



Speech style shift in conversational interviews 

4→

sore ga zenbu awa ni natchatta ‘that got filled up with bubbles.’

5 K: soo desu ‘Right.’ 6 R: oyu ga nukeru made tte kanari jikan kakaru n ja nai desu ka ‘It takes pretty long to drain the water, right?’ 7 K: u::n jippun gurai kakari masu ka ne ‘Mmm, maybe it takes about ten minutes.’ In her turn in lines 3 and 4, R emphasizes and reinforces the impact of the trouble by adding more information (about the size of the bathtub) and repeating the expression uttered by K in the preceding turn. R uses a non-desu/-masu ending in line 4, again marking the turn and its reply as a subspace of the ongoing context space. In the last two examples (8) and (9), the utterance with a non-desu/-masu ending (lines 15 and 4, respectively) is preceded by an utterance with a desu/-masu ending in the same turn. This manipulation of style shift within a turn can be regarded as a strategy to minimize any abruptness or any possible threatening effect toward both speakers’ faces, while at the same time making clear the discourse organization by marking the non-floor-holding turn with the non-desu/-masu ending. By replying (with the usual assertive affirmation) using the desu/-masu-ending for such a turn, K keeps the floor and manifests his role as the primary information provider and the leading actor. In these instances, we can see the participant role of R as a floor supporter and prompter. Without claiming the floor, she manages to add more detail to her interactant’s story and to co-construct a discourse without threatening the other’s face. In line 6, however, R shifts back to her role as an “emcee,” using a desu/-masu-ending. Example (10) demonstrates another variation of such collaborative discourse. It is a segment excerpted from the interview with the ‘chanko’ restaurant owner. The owner K talks about why he has not become a sumo wrestler despite the fact that he is from a family of sumo wrestlers. The interviewer elaborates her interlocutor’s talk, acting like a prompter for the leading actor. (10) (e-1: chanko) 1 R: zenzen sooyuu kiboo toka wa nakatta n desu ka ‘Didn’t (you) want (to be a sumo wrestler) at all?’ 2 K: anoo watashi wa nakatta n desu kedo ‘Well, I didn’t but,’ 3 R: ee ‘uh-huh.’ 4 K: otooto no hoo ga puro no maa kakkai ni haitte ru n desu ga ‘my younger brother is in the professional, er, sumo world, but,’



Shoko Ikuta

5 R: aa soo desu ka ‘Oh, I see.’ K: boku wa moo shoojiki itte moo nan te yuu n desu ka shinchoo mo taijuu mo kitei ika da kara hairenai n desu yo ‘honestly speaking, well, what to say, I was unable to enter (that sport), because my height and weight are below the requirements.’ 6

7 R: aa ja zenzen ‘Oh, so, not at all,’ 8 K: ee ‘Yeah.’ 9 R: moo chiisai koro kara ‘since childhood even,’ 10 K: ee ‘Uh-huh.’ 11 R: osumoo-san ni naroo tte yuu koto wa ‘becoming a sumo wrestler’ 12 K: ee ‘Uh-huh.’ 13→ R: amari nakatta ‘was not (your choice).’ 14 K: ee mushiro keeba no kishu tte iwarete mashita kedo ne @ ‘No, rather, (I) was being told to be a jockey @’ 15 R: aa soo desu ka ‘Oh, really.’ 16 K: chiisakatta kara ee ‘’cause (I) was small, yeah.’ R opens the topic and yields the floor to K in line 1, and in lines 7 through 13 with non-desu/-masu ending, R modifies K’s preceding utterances to accommodate them to the topic of the ongoing context space. The use of the non-desu/-masu ending has the effect of indicating that lines 7 through 13 are in a subspace within K’s ongoing space. For example, had line 13 of example (10) ended with desu/-masu (‘amari nakatta n desu ka’), the interviewer would have sounded like the “anchor,” or a co-anchor who was also holding the floor. This might threaten the other’s face, and, as a result, could sound rude or impolite as an interviewer’s utterance. On the other hand, the use of the non-desu/-masu style explicitly indicates that the utterance is a subspace that complements the larger context space constructed by the interviewee K, and consequently does not hold the floor but actively supports K’s floor, elaborating his prior remarks. By



Speech style shift in conversational interviews 

the use of such a style shift, the interviewer successfully plays the role of prompter for K, and avoids or minimizes a possible face-threat toward him. These examples of collaborative discourse in interview talk are different from stories jointly created by people who share the same experience, such as those created by co-workers, as studied in Richards (1999). As seen in example (10), the role of the interviewer is not regarded as that of a co-author but is more of an editor or a prompter. However, the interviewer works in the same way as a “co-author” in collaborating in discourse construction in the sense that “the form and content of talk is continuously reshaped by the co-participants, through their ability to create certain alignments and suggest or impose certain interpretations” (Duranti 1986: 242).

5. Conclusion We have seen aspects of the shift between the desu/-masu and non-desu/-masu styles used as a discourse organization strategy, one of the several possible functions of such shifts. Shifting to non-desu/-masu within the context of a desu/-masu-dominant interaction can mark the embedding of a subspace to an ongoing context space and can be employed for managing the floor and participant role relationships. The desu/-masu form not only assumes the presence of an addressee, but its user also becomes an “actor” in interviews. On the other hand, the non-desu/-masu style does not imply that its speaker is an actor. As we have seen, by manipulating the style shift, albeit not necessarily consciously, the interviewer may occasionally act as a “prompter,” while the interviewee continues to be the “leading actor.” The overall roles of the two interlocutors in the data are fixed throughout; i.e., the interviewer is the “emcee” of the discourse, and the interviewee the leading actor and the provider of information. The interviewer may act as a prompter to support the role of the interviewee, but it is made clear by the use of the non-desu/-masu style that the leading actor continues to be the interviewee. This role distinction has been related to the notion of floor. We have seen examples of collaborative discourse, where the interviewer complements the interviewee’s talk, while the exchange is considered to be part of the interviewee’s on-going floor. In other words, the interviewer is supporting the other’s floor, even when she takes a turn and complements or modifies the other’s utterance. What makes this possible is the shift to non-desu/-masu within the context of the predominant (unmarked) desu/masu style. The use of the non-desu/-masu style quite explicitly indicates that the utterance is a subspace that complements the larger context space, and consequently does not hold the floor but supports the other’s floor. The implication is that the manipulation of such a shift may also work as an interactional politeness strategy. We have seen that the shift to the non-use of desu/-masu style in a reciprocal desu/-masu style setting often works to save the other’s face as a consequence of shifting in interactional roles and floor management in a collaborative



Shoko Ikuta

discourse situation. It is not the style itself (such as the use or non-use of desu/-masu) that carries a certain politeness value, but rather, it is the shift between the two styles that is manipulated to achieve politeness (i.e., face-saving) goals, which in turn enables the interlocutors to collaborate to create an effective discourse.

References Brown P. and S.C. Levinson. 1987. Politeness: Some universals in language usage. Cambridge: Cambridge University Press. Cook, H. M. 1999. “Situational meanings of Japanese social deixis: The mixed use of the masu and plain forms.” Journal of Linguistic Anthropology 8 (1): 87–110. Duranti, A. 1996. “The audience as co-author: An introduction.” Text 6 (3): 239–47. Edelsky, C. 1981. “Who’s got the floor?” Language in Society 10: 383–421. Fishman, P.M. 1978. “Interaction: The work women do.” Social Problems 25: 397–406. Fraser, B. 1999. “What are discourse markers?” Journal of Pragmatics 31: 931–52. Goffman, E. 1967. Interaction ritual. New York: Doubleday. Goffman, E. 1981. Forms of talk. Philadelphia: University of Pennsylvania Press. Hayashi, R. 1996. Cognition, empathy, and interaction: floor management of English and Japanese conversation. Norwood, NJ: Ablex. Ikuta, S. 1983. “Speech level shift and conversational strategy in Japanese discourse.” Language Sciences 5: 37–53. Jorden, E. 1987. Japanese: The spoken language. Part I. New Haven: Yale University Press. Levinson, S.C. 1983. Pragmatics. Cambridge: Cambridge University Press. Levinson, S.C. 1988. “Putting linguistics on a proper footing: Explorations in Goffman’s concepts of participation.” In Erving Goffman: Exploring the interaction order, P. Drew and A. Wooton (eds), 161–227. Boston: Northeastern University Press. Maynard, S. 1991. “Pragmatics of discourse modality: A case of da and desu/masu forms in Japanese.” Journal of Pragmatics 15: 551–582. McLaughlin, M. L. 1984. Conversation: How talk is organized. Beverly Hills, CA: Sage Publications. Mikami, K. and M. Yamagata. 2000. “Nippon no oshigoto.” Nihongo Jaanaru, April, June, September, November, December issues. Tokyo: Aruku. Richards, K. 1999. “Working towards common understandings: Collaborative interaction in staffroom stories.” Text 19 (1): 143–74. Schiffrin, D. 1987. Discourse markers. Cambridge: Cambridge University Press. Sukle, R. 1994. “Uchi/soto: Choice in directive speech acts in Japanese.” In Situated Meaning: Inside and outside in Japanese self, society, and language, J. M. Bachnik and C. J. Quinn Jr. (eds), 113–142. NJ: Princeton University Press.



Speech style shift in conversational interviews 

Appendix Transcription symbols @ Laughter :: Syllable lengthening (without change in pitch) ? Final rise [ ] Overlap ' ' Translation ( ) Optional

Playing with multiple voices Emotivity and creativity in Japanese style mixture* Senko K. Maynard Rutgers University

1. Introduction Traditionally, speech styles in Japanese, especially da and desu/masu verb endings, have been characterized as being informal/casual/abrupt and formal/polite, respectively, and these forms are chosen primarily based on the particular genre and social situation. However, it is also known that styles are mixed not only within one genre, but within a single person’s spoken and written discourse segment as well. This study focuses on this phenomenon of style mixture, particularly on those cases where the same writer mixes different styles in a given discourse. In addition, in this study, I discuss the clause-noun combination with to yuu, especially when it takes the structure of [D(irect) Q(uotation) + to yuu + N]. By “direct quotation” I mean forms that bear spoken language features and give an impression of being the writer’s speech. This structure illustrates a style mixture within a single sentence in that the direct-quotation-like speech is embedded as a subordinate clause associated with the noun phrase, resulting in a case of mixing spoken and written styles. For example, sugoi kodomotachi na n desu yo, to yuu ishiki ‘the awareness such that “they are really amazing children”’ illustrates a spoken style incorporated into a noun phrase. Why does the writer mix different styles in Japanese, when he or she addresses the same reader in a stretch of text or within a sentence? This study attempts to answer this question by exploring Japanese style mixture from the perspective of emotivity and creativity. Based on examples drawn from contemporary Japanese written discourse, I make the case that style mixture is a strategy in which the writer manipulates his or her * This chapter was written and submitted to the editors in December 2003, and the volume was tentatively accepted for publication in May 2007. I would like to point out that during this interval, I have written a book (Linguistic Creativity in Japanese Discourse: Exploring the Multiplicity of Self, Perspective, and Voice, John Benjamins, 2007) which contains a part of this chapter. I would like to express my gratitude to the anonymous reviewer of the manuscript for offering constructive suggestions and directions. And my special thanks go to Kimberly Jones and Tsu­ yoshi Ono for making this project come to fruition.



Senko K. Maynard

multiple voices, a strategy that ultimately adds to the rich expressivity of language. I also interpret this strategy of style mixture from the Place of Negotiation theory, which I explored in my earlier studies (Maynard 1998a, 2000, 2001a, 2002). Although it is a common practice for a writer to mix different styles in essays, varieties and degrees of mixture seem to have intensified in recent years. For the analysis of style mixture, this study examines limited data chosen from contemporary written Japanese, i.e., certain types of essays, internet bulletin board entries, fiction, and nonfiction. It is beyond the scope of this paper to empirically investigate style mixture in terms of variables and frequencies, but by focusing on those cases where curious mixtures are observed, I hope to reveal the motivations for style mixture in certain contemporary Japanese written discourse. The essays to be examined are the kind that incorporates spoken style. Internet BBS (bulletin board service) entries are also known to incorporate a speech-like writing style. Also discussed are examples taken from fiction and nonfiction which incorporate direct-quotation-like speech in writing. The genre under investigation is (or at least it incorporates) what Satake (1995) identifies as shin genbun itchitai ‘new speaking-and-writing-agreement style’. According to Satake (1995), shin genbun itchitai is primarily used (1) among youth in communication through the internet, (2) in magazines targeted at youth, and (3) in many of the romance novels for girls. It is a written language but is written as if talking to a friend. It frequently uses interjections, particles, and sound-changing forms (e.g., graphologically marked vowel elongation), among others. As will be made explicit in this study, shin genbun itchitai appears mixed in certain writings written by mature adult writers targeted to a mature audience as well.

2. Background As background to the present study, three areas of research are relevant, i.e., the concept of mutivoicedness, style and style mixture, and the to yuu clause-noun combination. The term “voice” is not unfamiliar in Western literary theory; it is frequently used, as in “narrative voice,” “dialogized voice,” “heteroglossia” (Bakhtin 1971, 1981, 1986) and “multivoicedness” (Wertsch 1991). One useful definition of voice is how Bakhtin (1981) conceives it. According to Holquist and Emerson (in Bakhtin 1981: 434), Bakhtin’s concept of “voice” refers to “the speaking personality, the speaking consciousness” in the context of the Bakhtinian “dialogized voice.” Regarding the concept of heteroglossia, Bakhtin (1971, 1981, 1986) and Vološinov (1973 [1929]) offer guidance.1 Bakhtin (1971, 1981, 1986) takes the position that language is essentially interactional and dialogic, and language cannot simply escape from simultaneity of heteroglossia (multivoicedness). The meaning of a word is shaped and interpreted dialogically with the addressee, and with the society that endorses its very 1.

It is believed by some scholars that Bakhtin wrote under the name Vološinov as well.



Playing with multiple voices 

existence. Multiple voices echoing in one’s words (including prior as well as future speakings) are supported by social heterogeneity. Voices representing various registers, classes, cultures and sub-cultures reverberate in a speech, as they co-exist and interanimate among them. In applying this fundamentally psychological and literary concept of voice in my analysis, I mean by “voice” the writer’s personal attitude and feelings toward the content of the expression and toward the speech act itself. Different emotions echo in the writer’s voices. Moreover, voices echoing in various expressions realize the writer’s multiple presentations of self, and manifest how these selves are constructed and presented as multiple identities. Different identities speak in different voices. Further, the voice includes the writer’s creative voice. For example, the use of unexpected and excessive supra-polite style encourages an ironical reading. The use of imagined (or obsolete) styles also invites different voices into discourse. In other words, as I discuss in what follows, I take the position that addressive, emotive, and creative aspects associated with style mixture are best understood as the many voices a single writer uses in relation to the assumed reader. In this regard, I should perhaps mention the significance the concept of “voice” has played in Japanese language. Traditional as well as contemporary Japanese language studies have for a long time grappled with the distinction between two types of linguistic signs, “objective” shi, and “subjective” te-ni-o-ha or ji. The distinction between shi (roughly translated as “content words,” and therefore more “objective”) and te-ni-o-ha or ji (function words which express “subjective” voice and emotion) has been suggested by the Edo period Japanese grammarians such as Fujitani (1960 [1778]) and Suzuki (1979 [1824]), and later reintroduced to the field of Japanese language study by Tokieda (1941, 1950) and others. Specifically, it was Suzuki (1979 [1824]) who recognized the speaker’s personal kokoro no koe ‘voices from the heart’ which in Japanese are expressed by modal features such as particles and auxiliary verbs. For Suzuki, language’s essential function lies in its expressiveness realized not by shi alone but by the combination of shi and ji, the latter of which controls the voice of the speaker/writer. The literature on Japanese styles is extensive, and perhaps I should limit the review to my earlier studies on Japanese style mixture.2 However, before that, I should mention Haga’s work which offers general guidance to the study of Japanese styles. Haga (1962: 62), citing danwatai ‘spoken style’ and bunshootai ‘written style’ suggests that mixture of these styles without reason should be avoided. Haga (1962) calls this rule bunmatsu ikkan no gensoku ‘rule of consistency in sentence-final forms’. Haga, however, points out several situations in which language users may purposefully mix da, dearu, and desu/masu endings. First, in a discourse where da endings dominate with occurrences of sporadic desu/masu endings, the desu/masu endings express (1) for2. For the review of other studies on speech styles and style shifts, see Maynard (1991b, 2001b).



Senko K. Maynard

mality, (2) humor and an insertion of personal comment, (3) sarcasm, (4) vocative expression and (5) directly addressing the listener. In a discourse segment where the desu/masu style dominates but where sporadic da endings appear, da endings express an interpersonal familiarity and closeness to the listener. In addition, Haga notes that sometimes the da and desu/masu mixture results from sociolinguistically uncertain circumstances, especially when the speaker fails to clearly locate the addressee’s relative social status. Going beyond Haga’s comments, I have examined styles and style shifts in actual Japanese discourse in a series of studies. In my earlier works (Maynard 1991a, 1991b, 1993) I focused on two dominant forms in verb morphology, da (abrupt) and desu/ masu (formal) verb endings, and explored their respective functions. I examined casual conversation, dialogues from fiction, and literary essays in which abrupt and formal forms are mixed. I concluded that the choice of abrupt versus formal verb-ending forms when they are mixed may be predicted according to a low versus high awareness of “thou” or “you.”3 The low awareness situation which encourages the abrupt style occurs when: (1) the speaker is emotionally excited, (2) the speaker is involved in the event almost as if it were occurring right then and there, (3) the speaker expresses internal feelings in an almost self-addressed utterance, (4) the speaker jointly creates utterances, (5) semantically subordinate information is presented, and (6) the speaker expresses social familiarity and closeness. On the other hand, the high awareness situation which promotes the formal style occurs: (1) when the speaker expresses thoughts addressed to “you” with expressions appropriate in terms of sociolinguistic variables, and (2) when the speaker communicates information directly addressed to “you,” especially when the formal ending appears within abrupt style discourse. More specifically, when the da-style appears in the predominantly desu/masustyle text, the writer expresses surprise, abrupt remembrance, or a sudden surge of emotion. For example, in the narrative text, the writer is in the narrative world right then and there, i.e., the writer takes a perspective internal to the narrative world. In short, the choice of abrupt style within predominantly formal style achieves (1) immediacy and directness in expression, and (2) a narrative internal perspective. Conversely, when the desu/masu style appears in the context of the da-style, it marks the writer’s awareness of the writer’s own speech level. When the writer is more conscious of the reader and thus more socially aware, the writer organizes words and thoughts as social convention requires. Accordingly, the formal style is chosen, if appropriate. In a narrative text, the desu/masu-style is used to express a narrative-external voice, a voice that allows the narrator to direct commentary toward the reader. This gives the impression that the writer directly addresses the reader, that is, the addressiv3. Although I used the term “thou” in my earlier studies, I use “you” in more recent studies. See Maynard (1991a, 1991b, 1993) for the model developed with the concept of thou, attributed to Mori (1979).



Playing with multiple voices 

ity is foregrounded. The desu/masu style adds to the impression that the writer is making a conscientious effort to address the reader. Maynard (1994a) examined the style mixture used in political discourse and discussed how style mixture helps realize the juxtaposition of voices echoing in multiple semiotic contexts. I also discussed, in Maynard (1997b), the varieties of styles and style shifts practiced in Japanese discourse, and explained that style is chosen not only on the basis of social conventions but also for personal expressive reasons. Avoiding the prescriptive tone associated with style in traditional studies, I pointed out that styles are frequently mixed in Japanese and different styles have identifiable functions. In Maynard (1999), I discussed how style mixture in Japanese fiction is (and is not) translated into English. The desu/masu style in Kitchin and Tokage (by Banana Yoshimoto) signals that the narrator appeals to the reader by stepping outside of the narrated world. However, the English translations of these works fail to reflect the clear sense of the narrator’s shift of positioning. In more recent studies on style shifts (Maynard 2001a, 2001b, 2002), I discussed how emotion plays a part in the selection of speech styles. By analyzing examples drawn from Japanese television dramas, I pointed out that, in addition to the awareness of “you,” the shifting of style involves how emotions are negotiated and shared between the speaker and the partner. In Maynard (2001a), based on examples taken from television dramas, I discussed various motivations where speakers shift styles. For example, when a speaker feels vulnerable and is hesitant, the speech shifts to a softer, gentler, and often more polite style. In Maynard (2001b, 2002), I examined a television drama series (Majo no Jooken) in which two people fall in love, and how they shift styles depending on their emotions and their relationship. I pointed out that da verb endings the lovers use allow direct and forceful expression of emotion and that this behavior is predicated upon mutual intimacy. Da style indexes a sense of familiarity, tolerance, and indulgence. The analysis of the drama series has revealed that stylistic shift also occurs according to the chronologically documented emotional development as enacted in a television drama series. In Maynard (2001b, 2002) I interpreted stylistic shifts in terms of the presentation of different selves. The presentation of selves assumes certain kinds of “you,” i.e., the kind of “you” the speaker addresses. When the interactional aspect of communication is foregrounded, the speaker is highly aware of socially bound “you.” When emotive aspects become primary, the speaker becomes less sensitive to the socially-bound “you.” Instead, when the speaker uses expressions aiming for direct emotive appeal, the “you”-reaching inner self is foregrounded and this self directly addresses the intimate “you.” Likewise, when the speaker uses expressions for indirect emotive appeal, the “you”-reaching inner self indirectly addresses the intimate “you.” For example, as I pointed out, the choice of da style, through the direct emotive appeal to “you,” facilitates the presentation of self that desires to reach the inner self of the intimate “you.” In Maynard (2002), I also examined a television drama series (Long Vacation), and discussed how different kinds of feelings and attitudes influence style mixture. Using



Senko K. Maynard

the concepts of different aspects of self (gendered self, interactional self, as well as girlish, boyish, womanly, manly, subordinate, and equal selves) I analyzed the fluid and shifting speech styles of the main characters, which are in part dependent on which aspects of the partner they address. I also analyzed the playful use of styles as observed in creative role-playing, as well as the use of emotive desu. Above all, in a series of studies (Maynard 1991a, 1991b, 1993, 1997a, 1997b, 1999, 2001a, 2001b, 2002), I have emphasized that stylistic choices are motivated not only by social factors and constraints but also, and more relevant to my position, by personal emotions and desires. In fact, stylistic choice and style mixture result from the compromise between two forces, social norms and individual expressivity. The same person may mix when addressing the same partner depending on the situation, emotion, and desire. As noted by Haga (1962), and as I explored based on a variety of actual discourse examples, fluidity and variability of style are the norm, and styles, as many other linguistic and interactional strategies, can be manipulated for self-expression. Among recent studies on style in discourse, particularly relevant is Noda (1998). Noda focuses on how discourse structure and politeness are related and divides sentences into five types based on their discourse functions. They include: shinjoobun ‘inner sentences’ (which describe the writer’s inner thoughts and feelings), juuzokubun ‘subordinate sentences’ (sentences subordinate to other sentences), jijitsubun ‘fact sentences’ (which state facts), shuchoobun ‘opinion/position sentences’ (which offer judgment and explanation), and dentatsubun ‘communication sentences’ (expressions such as questions and commands). Based on these sentence types, Noda makes the following observations. First, primarily teineitai (polite) style is chosen in opinion/position sentences and communication sentences, since they address the partner. However, inner sentences and subordinate sentences take futsuutai (less polite ordinary) style, because they do not show concern toward the partner. Primarily futsuutai style is chosen in fact sentences. However, within this discourse, opinion/position sentences and communication sentences may take teineitai. Although I do not approach styles in discourse in terms of sentence types, Noda’s observations offer some guidance, and I will touch upon them where relevant. Style mixture becomes an important issue when social conventions recommend polite forms, while personally the speaker wishes to use less polite forms to show intimacy. Fukao (1998), using questionnaires, examines how college students in Japan feel about choosing politeness levels. Specifically, Fukao examines how students express shitashisa ‘familiarity, closeness, intimacy’ when politeness is socially expected, for example, when they address their seniors. It was found that juniors begin to feel uncomfortable using polite expression as shitashisa increases toward their seniors. They gradually use less polite expressions, mixing respectful forms and simple desu/masu forms, or, mixing desu/masu and da forms. At the same time, it was found that seniors allow and/or expect less polite forms coming from juniors to whom they feel close. As I have maintained in my earlier studies, being “polite” requires more than using polite expressions. This reminds us of Usami’s (2001, 2002) work in which she dis-



Playing with multiple voices 

cusses discourse politeness. Usami points out the importance of “positive politeness” in Japanese, where desire for empathy and wishes to share feelings are expressed by linguistic means that shorten the psychological distance (e.g., particles, jokes). The style mixture to be examined in this study illustrates how some of this positive politeness is expressed in discourse, in relation to the reader. Another area associated with style in Japanese is different varieties of language. While many varieties can be mixed in written discourse, in this study I touch upon the variation due to speaker’s age. Additionally, as Kinsui (2003) points out, imagined styles (or what Kinsui calls “virtual” language) are used in Japanese, and discussed in association with creativity. Now a short literature review on the [DQ + to yuu + N] structure is in order. To yuu is a quotative marker, and studies on quotation abound.4 I myself have written on this topic (Maynard 1984, 1986, 1994b, 1996, 1997a, 1997c, 1998b). Especially relevant to the present study is self-quotation, as explored in Maynard (1996), and the multivoicedness associated with quotation. In Maynard (1996) I argued that self-quotation allows for the manipulation of multiple voices in Japanese discourse and that such functions are supported by the juxtaposition of semiotic contexts, activated in part by the quotation itself. The study directly connected to the [DQ + to yuu + N] structure under discussion is Maynard (1992, 1993), in which I discussed the to yuu clause-noun combination appearing in both dialogue and narrative portions of contemporary fiction. I analyzed the clause-noun structure with to yuu, i.e., [X to yuu Y], in contrast with the clause noun structure without to yuu, i.e., [ X Y]. [X to yuu Y] includes expressions such as kaimono ni iku to yuu yotei ‘a plan (such as) to go shopping’, while [XY] refers to expressions such as kaimono ni iku yotei ‘a plan to go shopping’. I emphasized that the [X to yuu Y] structure, which I called “quotative explanation,” offers the writer an opportunity to express multiple voices and foregrounds the writer, while “nominal modification” does not. The [X to yuu Y] structure also functions as a strategy for connecting two fundamentally different but closely related activities of “saying” and “describing” in Japanese discourse. I pointed out that the distinction between these two structures may be best understood in the context of reporting and reported speech. It is often said that the (authorial) reporting of what is reported by someone else involves shifts in points of view and that the process reveals varying degrees of multiple voices representing various sources. Bakhtin (1981) describes this phenomenon in the following way. The words of the author that represent and frame another’s speech create a perspective for it; they separate light from shadow, create the situation and conditions necessary for it to sound; finally they penetrate into the interior of the other’s

4. For the review of other studies on quotation and related issues, see Maynard (1996, 1997a, 1997c, 1998b).



Senko K. Maynard

speech, carrying into it their own accents and their own expressions, creating for it a dialogized background. (358)

The [X Y] structure offers a construction in which the writer’s voice consumes the voice of the producer of X itself. Thus the effect of choosing the [X Y] structure is the kind that Bakhtin (1981) characterizes as above. On the other hand, by marking X with to yuu, the [X to yuu Y] structure offers an opportunity for the writer to explicitly qualify the voice of X’s producer. Consequently, some of the direct speech features are retained. These two different ways of merging the voice represent what Vološinov (1973 [1929]) calls “linear style” and “pictorial style.” The linear style “construct(s) clear-cut, external contours for reported speech, whose own internal individuality is minimized” (1973 [1929]: 129), while the pictorial style “obliterate(s) the precise, external contours of reported speech” and it includes “not only the referential meaning of utterance, the statement it makes, but also all the linguistic peculiarities of its verbal implementation” (1973 [1929]: 121). The [X to yuu Y] structure is perhaps best understood as quasi-pictorial style, since X can contain only limited aspects of verbal implementation. At any rate, we find here the phenomenon where linguistic devices and rhetorical strategies contribute to the manipulation in varying degrees of merging voices attributable to the writer’s different expressive manipulation. Again, let me borrow Bakhtin’s (1986) words. The word used in quotation marks, that is, felt and used as something alien, and the same word (or some other word) without quotation marks. The infinite gradations in the degree of foreignness (or assimilation) of words, their various distances from the speaker. Words are distributed on various planes and at various distances from the plane of the authorial word. (1986: 120–121)

The [DQ to yuu N] structure, as in the case of the [X to yuu Y] structure, foregrounds how the information in DQ is presented. As a result, it makes more aware the existence of the evaluator, whose attitude, opinion, and/or reaction toward the content of X become relatively more important than otherwise. The [DQ to yuu N] structure echoes at least two different voices, one expressed in DQ and the other that integrates DQ into the [DQ to yuu N] concept. Ultimately, it is a structure where the writer’s speaking (and saying) and writing (and describing) voices meet each other, echoing multiple voices in the process.

3. Styles in written Japanese In Japanese linguistics and language studies, the term “style” has primarily been used for verb morphology, i.e., da and desu/masu verb endings. In reality, style is a complex discourse phenomenon where multiple factors interact. For example, phonological features (e.g., graphologiclly marked vowel elongation), vocabulary, sentence structure, topic selection, discourse structure, and the media, among other elements, contribute



Playing with multiple voices 

to the overall style mixture. Obviously, it is beyond the scope of this paper to discuss these features in detail, but I will touch upon some of them where relevant. Based on earlier works including mine (Maynard 1991a, 1991b, 1993, 1997a, 1997b, 1999, 2001a, 2001b, 2002), and especially on those studies that discuss the variations of styles associated with da and desu/masu forms, I propose the following eight styles in written Japanese, with two base-line styles, da and desu/masu. The two baseline styles are the over-all basic styles recognizable in a piece of writing. The writer may mix and match four different styles within each of the base-line styles. These four styles are foregrounded against the context of the base-line style, and it is the total mixture of these that realizes the overall stylistic effect. base-line da: chosen in writings with general broad audience often used when the writer takes on the descriptive mode, or in sentences subordinate within the text structure use of da form emotive da foregrounds emotivity use of da form, accompanied by particles, spoken-style vocabulary, sound change, etc. (extreme variety uses blunt expressions, slang-like interjections, etc.) emotive desu/masu foregrounds addressivity with emotivity use of desu/masu form, accompanied by particles and other emotive features addressive desu/masu foregrounds addressivity use of desu/masu form supra-polite special writing genre use of desu/masu and (extensive) honorifics and polite vocabulary base-line desu/masu chosen in the writing with a tone of addressing to the audience use of desu/masu form emotive da (same as above) non-addressive da often used in sentences with subordinate information use of da form emotive desu/masu (same as above) supra-polite (same as above)

 Senko K. Maynard

Note that addressive desu/masu under base-line desu/masu and non-addressive da under base-line da are missing from the list. This is because in principle the base-line desu/masu style is addresive and the base-line da style is non-addressive. Consequently, the addressive desu/masu is labeled base-line desu/masu, and the non-addressive da, base-line da. A note of caution is in order. In addition to the two base-line styles, some writings take the emotive desu/masu or the emotive da as the base-line style. Although in most cases it is possible to identify the base-line style, in some writings the styles are so mixed it is difficult to determine the underlying base-line style. Regarding the suprapolite form listed above, I should mention that when socially unexpected, it is so deviant that an ironical reading results. In other words, a supra-polite expression calls attention to itself and prompts the reader to look back and puzzle out the reason for it. I will touch upon this phenomenon regarding examples (10) and (11) later. As I point out later, a writer may use multiple styles stereotypically associated with age. A writer may even use an archaic or obsolete style. Contrary to the commonly held view that style responds to the specific social and situational context, style may actually be used in a reverse manner, not in conformity but to create the world the writer desires. By mixing styles, the writer expresses different, often overlapping, voices. Emotive styles foreground addressive and emotive voices. Styles also identify the speaker, and furthermore, styles help realize the writer’s creativity. In what follows, then, I discuss sentences with different styles in discourse taken from select data in terms of emotivity and creativity. I will also examine one BBS entry which mixes many of the styles mentioned above.

4. Emotivity Once the base-line style in Japanese writings is determined, deviation from that style catches attention, and accordingly, the reader searches for its motivation. These deviations are often associated with emotivity in different ways as observed in examples (1) through (9). In what follows, I address emotivity in terms of (1) addressing emotivity, (2) revealing emotivity in discourse, and (3) incorporating emotivity within sentences.

4.1

Addressing emotivity

When expressing one’s emotivity, the writer may take the emotive desu/masu style within the base-line da style. By this choice, the writer conscientiously addresses the reader in a socially expected style, but still, emotivity is foregrounded.



Playing with multiple voices 

(1) is taken from a series of essays titled “This and that, I want to eat” written by Sadao Shooji (male, b. 1937), appearing in the weekly magazine Shuukan Asahi.5 The particular essay titled “Going to eat a 100-yen udon noodle dish” begins as follows.6 The base-line for this essay is da, but the text is mixed with other styles.7 Of particular interest is the emotive desu/masu style. (1) 1 P Hyakuen no udon o tabe ni itta. (base-line da) ‘I went out to eat a 100-yen udon noodle dish.’ 2 P Ii desu ka hyakuen desu yo. (emotive desu/masu) ‘Are you paying attention? It is 100 yen.’ 3 P Ikkai no shokuji ga hyakuen de sumu wake desu. (addressive desu/masu) ‘It’s that one meal costs only 100 yen.’ 4

“Fuun, sorya yokatta ne.” ‘”I see, that’s nice.”’

5 P Datte? ‘You say?’ 6 P Doomo jijoo ga yoku nomikomete-nai yoo da na. (emotive da) ‘It seems that you don’t quite understand the situation.’ 7 P Tsumari desu, hirumeshi o kui ni iku. (emotive desu/masu, base-line da) ‘In other words, I go out for lunch. 8 P Hyakuen no udon o taberu. (base-line da) ‘I eat a 100-yen udon noodle dish.’

5. Shuukan Asahi, is published by Asahi Shinbunsha, a major Japanese news media company that publishes a variety of products (newspapers, magazines, books, and so on). The magazine’s target audience consists of college-educated (primarily male) adults. 6. Japanese transliteration is given in phonetic orthography referred to as the Hepburn style, with the following additional conventions. In presenting double consonants, before cha, chi, cho and chu, t is added, thus instead of icchi ‘agreement’, itchi is used. Syllabic n is written n unless it immediately precedes a vowel, in which case it is written n’. The glottal stop, wtitten as small tsu in Japanese, is spelled out as tt- For long vowels, unless conventionalized otherwise, double vowels are used. Morphological divisions are made for the purpose of convenience only. 7. For the presentation of data, P indicates a new paragraph. For each verb-ending sentence, the style designation is given in parentheses. The sentence without the predicate (e.g., nominal sentence) is usually interpreted as a part of the base-line style, but no designation is given in the data. Also, quotations are excluded from the style designation because they do not directly represent the writer’s comments. In data presentation, the linguistic form and style designation that are under discussion appear in bold.

 Senko K. Maynard

9 P Doodoo to yooji de shiiha shinagara mise o deteiku. (base-line da) ‘I come out of the shop proudly, picking my teeth with a toothpick.’ P Hyakuen de koo yuu koto ga dekiru wake desu. (addressive desu/ masu) ‘It’s that you can do this for 100 yen.’ 10

11

“Fuun, sorya yokatta ne.” ‘”I see, that’s nice.”’

P Ano nee......imadoki nee, tatoeba nee, teishokuya no raisu datte nee, nihyakuen na wake desu. (emotive desu/masu) ‘Listen, this day and age, for example, a plateful of cooked rice costs 200 yen at restaurants where Japanese set entrees are served.’ 12

P Teishokuya e itte raisu dake tabete, doodoo to yooji de shiiha shinagara mise o deteiku koto ga dekimasu ka. (addressive desu/masu) ‘Can you go to that kind of restaurant and eat rice only and come out proudly picking your teeth with your toothpick?’ Shooji (2003: 54) 13

Shooji uses emotive desu/masu in lines 2, 7, and 12. The desu/masu style within a discourse that follows the base-line da style signals the writer’s awareness of one’s own speech level (Maynard 1991a, 1991b, 1993). As I explored in Maynard (1999), in written discourse the writer is aware of the reader and directs his commentary straight to the reader in the style that social conventions require. In line 1, that style is desu/masu. As soon as the desu/masu style appears in the base-line da discourse, the reader senses the writer’s appearance as a writer, external to the essay’s content. In lines 2, 7, and 12, the writer reveals a formal, somewhat official posture, as if he were appealing to the reader in a formal conversation. Use of the desu/masu style allows for the presentation of a different self, echoing a more formal, public voice amidst the writer’s primarily casual style. It is important to note that lines 2, 7, and 12 contain emotive expressions as well. Line 2 includes the particle yo, and line 12 repeats the particle nee as an appeal to the reader. In line 7, desu is used, like a particle, adding to the sense of strong appeal to the reader. These particles enhance the feeling that the writer is repetitiously breaking news of great importance with an urgent concern that the reader “gets it.” Also noteworthy is the use of da form, in lines 7, 8, and 9 (i.e., tabe ni iku, taberu, deteiku). These sentences are subordinate in that they provide information subordinately associated with significant points (made in other sentences) (Maynard 1991a, 1991b, 1993, Noda 1998). Lines 7- 9 are summarized in line 10, which more directly addresses the reader. This observation is further supported by the demonstrative koo yuu ‘this kind’ in line 10, which refers to the content presented in the preceding three



Playing with multiple voices 

sentences, lines 7–9. Line 6 is an example of the emotive da style in that it represents a self-revealing spoken style (partly realized by the particle na). (2) is taken from a series of essays, “Ai Iijima’s life in Kinshichoo living on royalty income,” written by Ai Iijima (female, b. 1972) appearing in Shuukan Asahi. The particular essay titled “The bay area” follows the base-line da style, and it starts with the following. (2) 1 P Hajimete Shiodome ni itta. (base-line da) ‘I went to Shiodome for the first time.’ 2

Odoroita. (base-line da) ‘I was astounded.’

3

Kyuu kokutetsu (moto JR) atochi de, ima saikaihatsu shiteiru tokoro da kedo, kanben shite yo, JR. (emotive da) ‘It is the land owned by the old national railway company (today’s JR) and is now being developed. Please, do something, JR.’

4

Nan daa? (emotive da, extreme) ‘What is it?’

5

Ano bakudaina tochi wa. ‘That vast land.’

P Koo yatte, aru hi totsuzen shutsugen suru machi tte yuu no ga aru. (base-line da) ‘In this way, one day all of a sudden a town is born.’ 6

7

Hajimari wa Ebisu Gaaden Pureisu. ‘First it was Ebisu Garden Place.’

8

Are wa biiru gaisha no tochi datta. (base-line da) ‘That land was owned by a beer company.’

9

Sorekara Azabu Juuban no ookina chika chuushajoo mo, nannen ka mae, ittara totsuzen atta ki ga suru. (base-line da) ‘And the humongous underground parking at Azabu Juuban. It was there all of a sudden when I went there a few years ago.’

10 (...)

Kore wa kaku sherutaa mo kaneteru to yuu uwasa. ‘There is a rumor that it was built so that it can be used also as a nuclear shelter.’

P Hakkiri itte watashi wa seiji ni mukanshin desu. (addressive desu/ masu) ‘To put it frankly, I am indifferent to politics.’ 11

 Senko K. Maynard

12

Sanjussai no otona toshite hazukashii kurai yononaka no koto shirimasen. (addressive desu/masu) ‘I don’t know what’s happening in society, to the extent that I should be embarrassed as a thirty-year-old adult.’

13

To yuu ka shin’yoo shite-nai no ne. (emotive da) ‘Or, more truthfully, I don’t trust things.’

14

Kodomo no toki kara, otona o shin’yoo shi-nai hanshakaiha datta n da kedo. (emotive da) ‘Since I was a child, I was anti-social and did not trust adults.’

15

“Anna otona ni wa naritaku-nai” ‘”I don’t want to grow up to be that kind of adult.”’

P To omotte, keisatsu ni tsukamari nagara sodachimashita. (addressive desu/masu) ‘Thinking so, I grew up, (occasionally) being caught by the police.’ 16

17 P Tonikaku kussetsu shitemashita. (addressive desu/masu) ‘At any rate, I had a twisted past.’ 18

Demo ima no hoo ga motto kussetsu shiteru kamo. (base-line da) ‘But, maybe now I am experiencing even more twisted ways of thinking.’

19

Moo yononaka, uso darake to sae omou. (base-line da) ‘I can only think that this society is filled with lies.’

20

Dakara jibun no me de mite, jibun de kanjite, jibun de omowa-nai kagiri shin’yoo shi-nai. (base-line da) ‘So, I don’t trust anything unless I see things with my eyes, I feel them directly, and think about them by myself.’



P Moshi terebi no konsento mo nukete, NTT ga dame ni nattara… (base-line da) ‘If the television is disconnected, and if NTT becomes dysfunctional…’ 21

22 23

Soshitara watashitachi, nannimo shira-nai de ikiteru wake yo. (emotive da) ‘Then, we would live without knowing about anything.’ Yamaoku ni sunderu, nantokazoku no hitotachi wa, ima mo sensoo ga hajimatta koto nante shira-zu ni ikiteru, kitto. (base-line da) ‘But those people of those tribes (whose names I don’t know) living deep in the mountains, even today they live without knowing that a war has begun, I’m sure.’



Playing with multiple voices 

24

Nantokazoku wa, jibun de mita koto shika shin’yoo shi-nai hazu da. (base-line da) ‘Those tribes don’t believe in things unless they witness things themselves.’

25 P Sonna koto kangaenagara Shiodome desu yo. (emotive desu/masu) ‘Thinking about those things, it’s Shiodome.’ 26 27

Atarashii machi ga dekite, nihon no kuni ga hattten shiteiku no wa suteki dashi, ii koto da to omou no. (emotive da) ‘I do think it’s fine and dandy to see new cities rise up in Japan and to see Japan develop further.’ Demo saikin no Tookyoo de wa, atarashii mono ga dekiru no wa wangan dake. ‘But recently in Tokyo, it is only in the bay area where new things are built.’ Iijima (2003: 144)

The discourse segment (2) shows a greater degree of style shift than observed in (1). Line 4 is an example of the extreme emotive da style. Nan daa? is notably blunt and straightforward. Here the writer’s feelings are foregrounded, revealing the writer’s strong emotivity. Sentences 6 through 9 are the basic-line da style. I should add that there are four paragraphs deleted here and they take primarily the base-line da style. Then in the three paragraphs of lines 11 through 17, the writer introduces formal style. The desu/masu style chosen in 11, 12, 16, and 17 shifts the writer’s position to that of formally addressing the reader, i.e., the addressive desu/masu style. The writer shows a more serious tone, in part corresponding to the seriousness of the topic. Lines 11 and 12 reveal her situation with a sense of addressing the reader. In lines 16 and 17, the writer explains about her past to the reader, as if making an official statement. These sentences fall into the case of Noda’s (1998) shuchoobun ‘opinion/position sentences’. Particularly noteworthy is that lines 16 and 17 appear in the past tense; they echo narrative voice (as if the writer were telling a story), and they add to the shift from the present situation to the past. Overall, the formal style chosen here foregrounds the writer’s addressive attitude. The reader is made aware that the writer takes to heart social constraints and expectations. Lines 18–24 basically mix the emotive da style with the base-line da style. Line 25 shifts to the emotive desu/masu style and instantly takes the “talking posture.” The writer is now outside the essay content, and emotively and intimately (although with formal attitude) appeals to the reader. Given the stylistic choice, it is not impossible to interpret this sentence as a shift within the discourse, that is, to signal the return to the topic of this essay. In other words, the stylistic choice here is motivated, in part, by the discourse organizational need (of topic maintenance) as well. Line 26 takes on the emotive da style. These lines reveal her inner feelings, foregrounding her emotivity.

 Senko K. Maynard

The overall stylistic effect observed in (2), and in other Iijima’s essays appearing in the series, is the writer’s contemporaneous casual spoken style. Being an obvious case of shin genbun itchitai, the writer presents her character (as the writer) as being contemporary, straight-forward, and openly self-revealing.

4.2

Revealing emotivity in discourse

This section examines emotivity expressed in both da and desu/masu base-line styles. In both cases, casual style reveals direct emotivity. In other words, as we see in examples (3) and (4), against the base-line da style, emotive da (and its extreme style) foregrounds emotivity, and against the base-line desu/masu style, as we see in example (5), emotive da style also foregrounds emotivity. In terms of the presentation of self, as I touched upon earlier, when the speaker uses expressions aiming for direct emotive appeal such as (extreme) casual style, the “you”-reaching inner self is foregrounded and this self directly addresses the intimate “you.” (3) is taken from a series of short commentary essays (accompanying photographs) titled “Don Konishi’s cool fashion check” in Shuukan Asahi. The fashion commentator, Don Konishi (or, Yoshiaki Konishi, male, b. 1950), critiques the fashion of celebrities, and in this particular essay the French foreign minister, Dominique de Villepin, is the subject. The essay is written in the base-line da style, and the portion to follow consists of the 3rd and 4th paragraphs of the six-paragraph commentary. Although Konishi’s writing style is generally casual and straightforward, some extremely casual and blunt styles are mixed in, as in lines 6 and 8. P Shikashi, sonna shuchoo to kaodachi to wa urahara ni, suutsu sugata wa kanari ikete-nai. (base-line da) ‘But in contrast to his assertive statements and facial features, the way he looks in his suit is definitely un-cool.’



(3) 1

2

Mazu suutsu no saizu ga dabodabo. ‘First, the size is too big.’

3

Kata no rain to ii eri no katachi to ii, furukusasa mo manten. ‘Whether it is the shoulder line or the shape of the collar, it’s completely outdated.’

4

Koryaa, nannen ka mae ni katta tsurushi da yo, kitto. (emotive da) ‘This is a ready-made suit he bought several years ago, for sure.’

P Chooshin de hansamu da kara, ikken chiteki de sutoikku ni mieteru ga, marude fasshon wa tonchinkan to mita. (base-line da) ‘Because he is tall and handsome, at first glance he looks intellectual and stoic, but his sense of fashion is completely off track.’ 5



Playing with multiple voices 

6

7 8

Sonna kare ga sekai no butai de hatsugen suru n da kara, Furansu mo oodaa no suutsu kurai zeikin de katteyare ttsuu no. (emotive da, extreme) ‘He is clad like that and speaks on the world stage. So I’m telling you, is it asking too much for the French government to use taxpayer’s money to buy him a tailor-made suit?’ Omake ni shatsu no eri wa uchigawa ni makikonderu wa, nekutai wa hinmagatteru wa. (emotive da) ‘On top of that, his shirt collars are bent inward, and his tie is crooked. (emotive da)’ Irozukai mo Furansujin rashikara-nu jimisa de, tende dame jan. (emotive da, extreme) ‘The color combination is, unlike Frenchmen, too sedate, and it’s totally disastrous.’ Konishi (2003: 16)

Ttsuu no ‘I’m telling you’ in line 6 is used only in very casual speech, and mixing this in with the writing adds an element of surprise. The use of jan in line 8 is also restricted to very casual speech. Jan is a colloquial (somewhat vulgar) version of ja-nai ka, which is used when the speaker is seeking agreement and/or confirmation. In other words, jan directly appeals to the partner’s response in a friendly and casual manner. Both ttsuu no and jan are typically associated with the speech of young male speakers. In line 8 jan appears in the expression tende dame jan ‘it’s a total disaster’, with the colloquial negative adverb tende ‘(not) at all’. These phrases intensify emotivity further. By mixing in this extremely casual style, Konishi communicates a flippant spontaneity, an irreverence strongly associated with youth. This extreme casual style captures the writer’s emotive voice in the most straightforward and revealing way. The writer directly tries to reach intimate “you” to maximize the emotive appeal. Obviously, in this series of essays written by Konishi, the topic is not serious. This may factor in as part of the reason why his writing is witty, severe, and straightforward, bordering on blatant and shocking. In addition to being provided information about fashion, the reader is given insight into the persona (real or manufactured) of the writer as well. The fictive identity created in Konishi’s writings on fashion is all part of the effect style mixture can create. (4) is a case where a direct, blunt, and strongly assertive voice is heard in an essay written by a female writer. (4) is taken from an essay written by Minako Saitoo (female, b. 1956). This writing takes the base-line da style (but it borders on base-line emotive da). The article is titled “Nursing specialist and maternity nursing specialist” and appeared in the journal Gengo. (4) 1 P A, soo soo. ‘Oh, I remembered.’

 Senko K. Maynard

2

3

4

Hookaisei de meishoo koso kawatta monono, mada dansei ni monko ga hirakaretei-nai shokushu ni “josanshi” ga aru. (base-line da) ‘The name itself has changed due to the change in the law, but there is an occupation from which men are excluded, and that is “maternity nursing specialist.”’ “Dansei josanshi ii ja-nai no. Otoko no sanfujinkai mo iru n da shi” to ittara, hantaironsha ni isshuu sareta. (base-line da) ‘I said “There are male gynecologists, so male maternity nursing specialist is fine, right?” and those who are against my position attacked me by saying,’ “Sukebe gokoro de josanshi o kokorozasu futei no yakara ga itara, doo suru no yo.” ‘”What are you going to do if there were men who want to be male maternity nursing specialists and they are perverts?”’

5 P Mata moo, sugu sooyuu moosoo o hatarakaseru. (emotive da) ‘You know, you are imagining those peculiar things again.’ 6 7

Sore tte “kangofusan” no kosupure ni seiteki moosoo o kakitaterareten no to onaji reberu ja-nee no? (emotive da, extreme) ‘Isn’t such an attitude on the same level as being aroused into a sexual fantasy when dressed like a nurse?’ A, chigaimasu ka. (addressive desu/masu) ‘Oh, that’s wrong, is it?’

Saitoo (2003: 13)

Ja-nee no? ‘Isn’t it?’ in line 6 is in extreme casual style, and it demonstrates a stark bluntness. Although the writer’s strong assertive attitude is sensed elsewhere in this writing, ja-nee no? catches attention because of its naked force. The writer’s emotion shows through, and the reader cannot help but feel being on the receiving end of sharp-tongued rebuke. Curiously, ja-nee no? is more readily associated with the socalled “masculine” speech than it is with “feminine” speech. Traces of a feminine voice are largely absent and the reader is likely to visualize being addressed by an unconstrained, bold, and somewhat controversial character. The author’s actual character is beside the point. The character presented through the writings, however, is sharply defined as frank and opinionated. Perhaps by purposefully making use of the ideology of gendered speech, and by incorporating “masculine” speech, the writer most effectively expresses her views. Also to be noted is the stylistic juncture observed between ja-nee no (emotive da, extreme style) in line 6 and chigaimasu ka (addressive desu/masu) in line 7. It appears that the writer compensates for the blunt expression in line 6 by choosing a style that formally addresses the reader. This juxtaposition makes the writing curious, exciting, and humorous.



Playing with multiple voices 

(5) illustrates a case where the emotive da is mixed in the base-line desu/masu style. (5) is taken from internet BBS entries commenting on the television drama series Bijo ka Yajuu ‘Beauty or Beast’. P Ichiya akete, mada yoin ni hitatteiru watashi desu ga, mottainai desu yo nee. (emotive desu/masu) ‘Though the morning came, I’m still enjoying the lingering effect, but it’s too much of a waste, isn’t it?’

(5) 1

2

Kono mama owatteshimau no. ‘For the drama to end like this.’

3

Fukuyama-san to Matsushima-san no sutekina kappuru ya, gooka kyooensha no katagata no subarashii chiimuwaaku, konnani sutekina dorama o konomama owari ni suru no wa mottainai!! (emotive da) ‘The nice couple played by Mr. Fukuyama and Ms. Matsushima, and the wonderful teamwork performed by the supporting actors and actresses. What a waste to end such a wonderful drama at this point!!’



4

Subete ga kakko yokatta shi, kono gojisei de, konnani takusan no hito no kimochi o wakuwaku sasetekureru mono tte, sukunai shi. (baseline da) ‘Everything was cool; and in this day and age, it is rare to excite so many people’s hearts.’



Sonna dorama no seisaku ni kakawatta subete no hito o sonkei shimasu. (base-line desu/masu) ‘I respect everyone who was engaged in producing such a drama.’

6

Zokuhen kitai shiteimasu. (base-line desu/masu) ‘I have high expectations for the sequel.’

7

Arigatoo gozaimashita. (base-line desu/masu) ‘Thank you very much.’ BBS (March 21, 2003)

5

Contrast the use of mottainai desu yo nee ‘it’s too much of a waste’ in line 1 and mottainai!! ‘what a waste!!’ in line 3. Line 1 takes the emotive desu/masu (it appears in the desu/masu style with the particles yo and nee), but line 3 takes the emotive da style (it appears in the da style with exclamation marks). The writer starts and ends this BBS entry with the formal desu/masu style, giving the impression that as a whole, it purposefully maintains addressivity. But within this commentary, the writer foregrounds emotion by using mottainai with two exclamation marks. As I already touched upon, when the da style is used in the predominantly desu/ masu style writings, at least two kinds of related effects result, (1) immediacy and directness of expression, and (2) an essay internal perspective (Maynard 1991a, 1991b, 1993,

 Senko K. Maynard

1999). In (3), mottainai!! ‘what a waste!!’ expresses a spontaneous surge of emotion, giving the impression that the writer is right then and there. The immediate emotive expression also minimizes the distance between the writer and the content of the essay. As I reviewed earlier (Fukao 1998), style shifts occur as the sense of intimacy is mutually recognized, encouraged, and endorsed. Although (5) is a written discourse, the style mixture illustrates that the writer adjusts the level of familiarity and intimacy between the writer and potential BBS readers to the extent that the writer thinks reasonable. In this way, although the written discourse does not involve interaction in a physical sense, it nonetheless remains sensitive to a kind of self-presentation and consequently to a kind of addressed “you.”

4.3

Incorporating emotivity within a sentence

Style mixture is observed not only among sentences within discourse, but also within a single sentence. One such case is the [DQ + to yuu + N] structure under discussion, where N may be a regular noun or the indefinite noun koto. (6) is taken from fiction by Ryuu Murakami (male, b. 1952), (7) is taken from a mystery novel written by Yasuo Uchida (male, b. 1934), and (8) is taken from a nonfiction work by Kootaroo Sawaki (male, b. 1947). Examples (6) through (8) take the structure of a direct quotation-like expression and a regular noun or a noun phrase connected by to yuu. The base-line for (6) through (8) is da. And yet, within the clausenoun combination, direct quotation-like expressions appear.

(6) Shichoosha no minasan, owakari desu ka, karera wa sugoi gijutsu o motta sugoi kodomotachi na n desu yo, to yuu futari no kyasutaa no ishiki ga tsutawattekuru intabyuu datta. (emotive desu/masu, base-line da)

‘Dear viewers, do you understand it? They are amazing children who possess amazing skills! It was the kind of interview in which thoughts like these held by two news anchors were expressed.’ Murakami (2002: 331) In (6), the awareness of the two news anchors is described as Shichoosha no minasan, owakari desu ka, karera wa sugoi gijutsu o motta sugoi kodomotachi na n desu yo ‘Dear viewers, do you understand it? They are amazing children who possess amazing skills!’ in the emotive desu/masu style. Obviously the news anchors did not utter these words. But by putting their thoughts into conversational expressions, the writer makes his sentence alive and immediate. The emotive desu/masu style used here evokes a situation where such a style is being used, adding to the expression of the real-life-like voice. At the same time, the main predicate takes the base-line da style, which brings the discourse into the current narrative discourse. (7) 1 Harasawa wa yaya hiita yoosu ni natta. (base-line da) ‘Harasawa hesitated a little.’



Playing with multiple voices 

2 Joodan hanbunni itte wa mita monono, kaichoo ni kono teepu o kikaseta toki no hannoo o omoeba, shimatta......to yuu ki ni natteiru no kamoshirenai. (emotive da, base-line da)

‘He responded half-jokingly, but when he thought of the president’s response to this tape, he might have been beginning to feel, woops, this is a mistake.’ Uchida (1998: 195)

(7) contains an interjectional phrase shimatta ‘woops’, a word that cannot be used for descriptive purposes. The insertion of this phrase allows access to Harasawa’s immediate conversational voice. Yet, that voice is incorporated into the narrative voice. In line 2, the conversational and narrative voices simultaneously echo and create a multivoiced discourse where speaking and writing are integrated into one. (8) 1 P Watashi wa katsute nai hodo shi no chikaku ni ita no kamoshirenakatta. (base-line da) ‘It might have been that I was really close to my death, closer than I ever was before.’

2 3 4 5

Shikashi, dakara to itte, aratamete kyoofu o oboeru to yuu koto mo nakatta. (base-line da) ‘But, I didn’t feel particularly fearful just because I recognized that.’ Naruhodo, soo datta no ka, to yuu teido datta. (emotive da, base-line da) ‘It was just like, I see, I was (close to death).’ Aruiwa, watashi wa tanni shi ni taishite donkanna dake na no daroo ka. (base-line da) ‘Or, I wonder if I simply am not in tune with the idea of death.’ Iya, soo de wa nai yoona ki ga suru. (base-line da) ‘No, I do feel that is not so.’

Sawaki (2002: 197)

In (8), the writer expresses his own response by saying that it was to the extent that he would simply say, Naruhodo, soo datta no ka ‘It was just like, I see, I was (close to death)’ Here again, instead of describing his feelings, inserting what the writer thought as if in a conversation in the emotive da style not only adds to the authenticity of his thought but also helps to bring the discourse to a real-life situation. It should be added that it is difficult to find appropriate descriptive words in place of the direct quotation-like expressions used in the [DQ + to yuu + N] structure. For example, what descriptive phrase would be appropriate to take the place of naruhodo, soo datta no ka in line 3? This phenomenon further illustrates the usefulness of mixing speaking and writing styles and incorporating multiple voices.

 Senko K. Maynard

Now, the [DQ + to yuu + N] structure may consist of DQ and the nominalizer koto, connected with to yuu. (9) is such an example. (9) is taken from a collection of essays by Hiroyuki Itsuki (male, b. 1932), and the base-line is desu/masu. (9) 1 Tsumari, ganbara-nakutemo ii, ganbare-nai hito ni ganbare nante yuu koto wa nai yo, to yuu koto desu. (emotive da, base-line desu/masu) ‘That is to say, it is all right not to make every effort, and one should not tell those people who simply cannot make every effort to do just that.’

2 Shinkokuni ochikomu koto mo mata, hito ni wa daijina no desu. (base-line desu/masu) ‘Seriously experiencing the state of the psychological low is also important for a person.’ Itsuki (1999: 278) In line 1, in two locations koto is preceded by the writer’s speech in the emotive da style. The reader has access to how the writer felt in his own words. Here, the reader is allowed a glimpse into the writer’s personal voice in which the writer talks directly to the reader. In general, the [DQ + to yuu + N] structure incorporates speaking (represented by DQ) and describing by integrating the speaking voice into a descriptive sentence. When the writer’s speech appears in emotive styles, the reader is made more aware of the writer who speaks in a particular style. It is as if the writer as a speaker intrudes into the discourse space and speaks as if carrying on a conversation, and still that is immediately incorporated into the writer’s writing voice. Recall Vološinov’s (1973 [1929]) concepts of “linear style” and “pictorial style.” The [DQ + to yuu + N] structure makes it possible to call in the pictorial style as it includes “not only the referential meaning of utterance, the statement it makes, but also all the linguistic peculiarities of its verbal implementation” (1973 [1929]: 121). It is a structure that makes possible the mixture of styles within a sentence.

5. Creativity 5.1

Echoing ironical voices

Style mixture also enhances the creative use of language. For example, a supra-polite style, when socially unnecessary and unexpected, guides the reader to interpret it as irony. Such an example appears in (10), an essay titled “Graduate school entrance ceremony” written by Makiko Uchidate (female, b. 1949). This essay appeared in Shuukan Asahi as part of the series “Jabbing your elbow against the noren (cloth entrance curtains for shops).” It takes the base-line da style.



Playing with multiple voices 

(10) 1 P Sugu ni haha ni tsukisoi o tanonda tokoro, ‘Right away I asked my mother to come with me, and then, (she said,)’ 2

“Sendai ni wa ikitai kedo nyuugakushiki ni shusseki nante yaa yo. Kono toshi ni natte, kono toshi no musume no nyuugakushiki nante” ‘”I do want to visit Sendai, but I don’t want to attend your entrance ceremony. Such an entrance ceremony for this old a daughter when I’ve become this old.”’

3 P To hana de warau. (base-line da) ‘She chuckles it away.’ 4

Daga, watashi ga otooto no kotoba o tsutaeru to, ‘But when I told her what my younger brother said, (she said,)’

5

“Sorya soo ne. Demo, hogosha datte Makiko yori wakai wa yo. Makiko no toshi da to, futsuu wa kodomo mo daigaku sotsugyoo shiteru wa” ‘”That makes sense. But I don’t want to be your guardian, because other students’ parents are younger than you are. At your age, ordinarily your children would have graduated from college.”’



6 P To, jitsubo to wa omoe-nai shibiana okotoba. (supra-polite) ‘In this way, she uses an expression so harsh, it’s hard to believe that it comes from ones’ own mother.’ 7

Soshite, ‘Then,’

8

“Ii wa, iku. Demo hitori wa hazukashii kara Setsuko mo sasou wa” ‘”All right, I’m going. But I am embarrassed to go alone, so I’ll invite Setsuko to come along.”’

9 P To yuu. (base-line da) ‘She says.’ 10

Setsuko to wa haha no imooto de Sendai ni iru. (base-line da) ‘Setsuko is my mother’s sister who lives in Sendai.’

11

Shimai de hazukashisa o wakeaeba ii to omotta rashii. (base-line da) ‘My mother seems to have thought that two sisters sharing the embarrassment would be better.’ Uchidate (2003: 66)



The writer maintains the base-line da style virtually throughout. Given this context, the deviation observed in line 6 stands out sharply. First, the convention is that when the writer refers to her family members, supra-polite forms are avoided. Okotoba (a combination of the polite prefix o- and the noun kotoba ‘phrase’) in line 6 is obviously excessive. This excessiveness clues the reader that something is askew. The reader realizes the mismatch of the style mixture, and using this as a cue, interprets the phrase ironically.

 Senko K. Maynard

Irony is basically the rejection of a literal reading of an expression. It is an “intended infelicitous speech” (Tsuji 1997: 117), in that it violates the expected speech act. To let the reader know that an expression is meant as irony, an “echo marker” (Sperber and Wilson 1981, 1988) or an “irony signal” (Tsuji 1997) is necessary. One of the signals is excessive politeness. For example, Seto (1997: 139) cites expressions such as gorippa ‘extremely fine, great’ and oeragata ‘extremely respected people’ as expressions that are typically ironically interpreted. The phrase okotoba is purposefully presented with overly exaggerated politeness, but unless that politeness is called for, it demands the opposite reading. The prefix o- appearing in line 6 is a reminder that this phrase should be interpreted not in a literal sense, but in an emphatically opposite way. Okotoba expresses the writer’s disbelief that her own mother would deliver such words (as those depicted in line 5). Let me cite another similar example. (11) is taken from an essay written by the rakugo (comic storytelling) artist Sadanji Tatekawa (male, b. 1950) in which he offers critical comments about producing certain television programs. The base-line da is the primary style, although in this essay, the writer mixes casual styles extensively. This series appears in Shuukan Asahi. (Note that sentences ending with nominals are not marked with style designation. The RoboCop mentioned in line 1 is a nickname given to Takamisakari, a sumo wrestler.) (11) 1 P Konkai wa robo koppu no jimoto Aomori e manmen ni emi o ukabeta ripootaa ga tobu. (base-line da) ‘This time, the reporter with a smiling face flies to Aomori where the RoboCop is from.’ 2 P Aimokawarazu sesuji no samuku naru yoona tsukuri. ‘The program is produced, as usual, so terribly that I feel a chill down my spine.’ 3

Wazato rashiku guuzen o yosoou, enshutsu to yuu no mo ozomashii munoo de yoochina shuzai. ‘The reporter gathers information pretending to be coincidental, and the reporting is too silly and childish to be recognized as being produced.’

4 P “Ringo nooen ni yattekimashita. Achira ni sagyoo o shiteiru kata ga imasu. Chotto ohanashi o......” ‘”I am visiting an apple orchard. There is someone working over there. I think I will interview him.”’ 5 P Oyakusoku de, kamera ga yotteiku. (supra-polite, base-line da) ‘As promised, the camera approaches.’ 6

Kao no appu, ‘A close-up shot of the face.’



Playing with multiple voices 

7

“A-tt! Yoku niteimasu ne! Moshikashite Takamisakari-zeki no......wa-tt, otoosan kamo?” ‘”Oh, you look like him! Could it be that you are Takamisakari’s,... wow, could it be, his father?”’ (Tatekawa 2003: 76)

The supra-polite phrase oyakusoku ‘promise’ in line 5 instead of yakusoku requires an ironical reading. In other words, “as promised” (with the ironical meaning of “as shamelessly prearranged”) the camera crew shows up and reports as if things are accidentally happening, although everything has been already set up! As I explored in Maynard (2000), the use of irony invites two different voices into discourse. The writer maintains the base-line da style, but the other self introduces a second, humorous voice. The irony works as irony because the second self is there speaking in place of the writer. The writer creates a double-voiced discourse, and in the process the writer’s creative voice is foregrounded.

5.2

Borrowing others’ voices

Style mixture may also be creatively used when the writer borrows someone else’s voice. When the writer speaks consistently in a certain voice elsewhere, the deviation evokes a different character. It is as if the writer assumed a different identity. Obviously, identity is something that permeates every writing with varying degrees of intensity. As I touched upon already in the preceding section, emotivity and identity interact. However, it is also possible to observe the different characters that style mixture evokes. In this section I discuss generation-based styles and regional styles. The first case involves the writer’s character expressed through a borrowed identity, particularly in terms of age and gender as shown in (12). (12) is a case where a speech style stereotypical of someone else is inserted to create a distancing effect. (12), a BBS entry commenting on the television drama series Taiyoo no Kisetsu ‘The Season of the Sun’, is written in the base-line desu/masu style. (The phrase NG appearing in (12) is a Japanese expression for “no good,” and it refers to an out-take, i.e., an unused segment of film when making a movie.) (12) 1 P 2 shiin mimashita. (base-line desu/masu) ‘I saw those two scenes.’ 2

Shoojiki, Chii-chan no NG ga mirareru nante omoi mo se-zu nanka ureshikatta. (emotive da) ‘Honestly speaking, I didn’t think I would be able to see any of Chii’s out-takes, so I was pleased.’

 Senko K. Maynard

3 P Takaoka-san no riakushon mo yokatta shi, Yuki no papa no NG go no waraigao mo saishuukai naradewa de nanka hotto shiteshimaimashita. (base-line desu/masu) ‘Takaoka’s reaction was great, and the smiling face of Yuki’s father after his out-take was something expected for the final episode, and it was comforting to see that.’ 4 P Takkii wa 2 shiin tomo NG ni warau dake de, Takkii jishin no NG wa mirare-zu chotto zannen. ‘All Takkii did was to laugh when he appeared twice, and it was a bit disappointing not to be able to see Takkii’s out-takes.’ 5 P Demo kono mada yoin ni hitatteiru toki ni mi-nakute yokatta no ka na. (emotive da) ‘But maybe it was a good thing that I didn’t see them while I was still in the middle of the lingering bliss.’ 6 P Tamiyo-san ya Koohei nanka mo mitakatta naa. (emotive da) ‘I wanted to see Tamiyo’s and Koohei’s out-takes as well.’ 7 P Fan to shite wa takusan mitakatta to yuu koto desu na. Ha ha ha. (emotive desu/masu, middle-aged male speech) ‘I guess it is that as a fan, I wanted to see as many as possible. Heh heh heh.’ 8 P Demo hoosoo ga owatta ato sabishisa o minasan kanjiteiru yoo desu shi, mochiron kono watashi mo dooyoo na node, sotchokuni ureshikatta desu. (base-line desu/masu) ‘But after the broadcast was over, it seems that everyone is feeling a bit lonesome, and of course I was like that, so honestly I was pleased (to see those out-takes).’ BBS (September 25, 2002) Line 7 takes the style desu na, which is known to be a style used primarily by middleaged males (Ozaki 2001). Based on the survey conducted by Kokuritsu Kokugo Kenkyuujo (2000), Ozaki (2001) refers to this particular style (ending with desu na(a) and masu na(a)) as a style connected to a sub-group of male speakers in their 40s and above (i.e., ojisan kotoba ‘middle-aged male speech’). The survey results show that 18.4% of men in their 60s and 7.5% of men in their 40s use the desu/masu na(a) form, although almost none of them used it in their 20s. In other words, the style observed in line 7 is linked to a generational style used by a group of speakers at a certain stage in their lives. The writer creatively uses a speech style typically connected to a social sub-group. Note that other sentences in this BBS take the base-line desu/masu and the emotive da. Line 7 is an unexpected deviation. The stylistic contrast between other sentences and line 7 reminds the reader that something else is going on. I should also add that mixing



Playing with multiple voices 

in a style that feels as if it were borrowed from someone else’s style creates a distancing effect. At the point of line 7, the writer’s perspective becomes that of an observer in the distance, partly because a different self (speaking in a borrowed voice) comes to the fore. The writer creates a discourse where different selves speak in multiple voices. The reader may raise the question of how it is possible to understand that the writer of (12) is other than a middle-aged man. Of course, it is impossible to know who the actual (or pretended) writer is when it comes to postings on the Internet. However, it is known that these BBS entries are primarily written by young people and housewives. Because the writer refers to the main actor as “Takkii” (intimate vocative of the main actor) it is difficult, to the point of impossibility, to assume that the writer is a middle-aged man. Another example of borrowing someone else’s voice is observed when a male writer uses a style that is stereotypically associated with feminine speech. (13) is a short commentary essay about a new product titled “Would you enjoy being like a rich person with fake jewelry?” and it is written by Akio Shiga (male, b. 1973). It appeared in the men’s fashion magazine Popeye, and it is written with the base-line da style. (13) 1 P Honmono ijoo ni hikari kagayaku feiku no daiya ya shinju, kinmekki no akusesarii ga kokon toko hayatteru kedo, kimo wa sono “nanchatte kan” da. (base-line da) ‘Recently accessories made of fake diamonds, pearls, and plated gold that glitter more than genuine accessories are in fashion, but the main benefit is the sense of “feeling like.”’ 2



Honmono dattara ittai oikuraman’en suru no kashira na dekasa to kagayaki no akuse wa, bakabakashii mono hodo betaa. (supra-polite, “feminine” style) ‘Regarding fake accessories, those that are huge and glittering, i.e., the kind that someone may comment – I wonder how many tens of thousand yen this would cost if it were genuine – , the sillier the better, I think.’ (Shiga 2002: 4)

The expression oikuraman’en suru no kashira ‘I wonder how many tens of thousand yen this would cost’ in line 2 is a mock stereotypical “feminine” voice. The nearly gender-neutral counterpart is nanman’en suru no ka, which conforms to the base-line da style. The choice of supra-polite “feminine” style is, in part, motivated by the fact that jewelry is stereotypically something women wear, and the stereotype is that it involves women who speak pretentiously. Although some would say the traditional “feminine” and “masculine” styles are perhaps no longer viable (Endoo 1998, 2000; Takasaki 2002), here the writer effectively mimics a stereotypical “feminine” voice to poetically capture an attitude associated with a jewelry-purchasing woman. By exploiting the stereotype (to the level of

 Senko K. Maynard

caricature) and by mixing in this style, the writer brings out another voice, the voice of a “feminine” fetish for jewelry. The writer expressively presents his view through an interanimation of different voices. I should add that a stylistic gap is observed in line 2 of example 13 between the oikuraman’en suru no kashira and the phrase dekasa ‘largeness’. Dekasa is used in blunt style echoing the “masculine” voice, and therefore, the sentence incorporates two voices simultaneously. The gap between these styles helps the interpretation of its ironical reading. Recall Bakhtin’s (1971, 1981, 1986) point that multivoicedness is supported by social heterogeneity. By foregrounding voices associated with certain social groups, the writer takes advantage of the mixture of multiple voices only to express his or her emotivity.

5.3

Introducing imagined voices

The creativity associated with style mixture takes an interesting turn when the style chosen is imagined and/or obsolete. The imagined “virtual” language (Kinsui 2003) is purposefully used for creating rhetorical effects. The imagined style observed in the following four examples reflects the kind of speech directly linked to certain characters set in certain historical periods. One frequently used imagined style is the old-man style (roojingo), featured by ja and the negative n. Ja was used as a modal verb primarily in the Western provinces in pre-modern Japan. The negative morpheme n (instead of nai) is also used mainly in Western Japan. According to Kinsui (2003), once the capital was moved from Kyoto to Tokyo at the beginning of the Edo period (1603–1868), the Edo speech (Eastern variety) was chosen by the younger generation while the Kansai dialect was retained by the older generation. As a result, ja and n have been stereotypically identified as being oldman style. We find the use of these features in (14). (14) is taken from an essay titled “Movies and I” written by Keiko Nobumoto (female, b. 1964) that appeared in the movie magazine Kinema Junpoo. This particular essay takes the base-line da in discussing the movie Thelma and Louise. (14) 1 P Burapi-chan wa totemo yokatta. (base-line da) ‘Brad Pitt was really nice.’ 2

Kare ni wa nanno tsumi mo nai. (base-line da) ‘He has nothing to be blamed for.’

3

Daga, ano onnadomo wa nan da? (emotive da, extreme) ‘But, what are those women?’

4

Jiina Deivisu wa, sukina taipu no joyuu da. (base-line da) ‘Geena Davis is the kind of actress I like.’

5

Suuzan Sarandon mo, yakusha toshite wa, sugoi to omou. (base-line da). ‘I think Susan Sarandon is great as an actress, too.’



Playing with multiple voices 

6 7

Shikashi, sutoorii joo no onnadomo wa, mukatsuite shaanai. (emotive da, extreme) ‘But I am totally disgusted with women in the story.’ Are o kakko ii onna toka itteru yatsu mo iru ga, doko ga ja! (emotive da, imagined) ‘There are some people who say those women are cool, but how in the world (can they think so)!’



Uzai baka onna ni shika watashi ni wa mie-n! (emotive da, imagined) ‘To me they look like nothing but bothersome stupid women!’

9

Aa, aa, katteni shitara ii sa. (emotive da) ‘Ah, well, let them do whatever they want to do.’

10

Kuruma goto doko ni demo tsukkondekure. (base-line da) ‘Why don’t they get in a car altogether and crash into something?’ (Nobumoto 2002: 124)

8

The expression doko ga ja ‘how in the world (can they think so)’ in line 7 and mie-n ‘they look like nothing but’ in line 8 echo the voice of the old-man style, although it appears in the female writer’s discourse. Mixing this imagined style helps make the writing more interesting, humorous, and entertaining. The writer, by using doko ga ja and mie-n, borrows the authoritative and adamant voice associated with the old-man style, and, if only for a moment, speaks in his voice. Another example of using imagined style is observed in (15). (15) appeared in the series “Did you pull the plug?,” written by Kootaroo Arashiyama (male, b. 1942). This particular article, titled “Chatting about Mr. Beckham at a barbershop,” ends as follows. Again, this article appeared in Shuukan Asahi, and the base-line is da. (15) 1 P Tokoya no musume ga, mimimoto de, ‘The daughter at the barbershop, close to my ear,’ 2

“Bekkamu-sama wa kamusama da kara bekkamisama na no yo. Kinpatsu ga sutekina Bekkamisama. Bekkamusama no naka ni wa kamisama ga haitteru no” ‘”Mr. Beckham (Bekkamu-sama) is like a God (because kamusama sounds like kamisama ‘God’), so he is Bek-kami-sama. Oh, Mr. God, Beckham (Bek-kami-sama), with beautiful blond hair. There is God inside Mr. Beckham.”’

3 P To yuu. (base-line da) ‘she says so.’ 4

A, soo de gozaimasu ne. (supra-polite) ‘Oh I see.’

 Senko K. Maynard

5

Ke-tt. (emotive da) ‘Ni-’

6

Kekkoona koto de gozansu. (emotive desu/masu, imagined) ‘Nice, how very nice it is!’ (Arashiyama 2003: 129)

The use of kekkoona koto de gozansu in line 6 clearly catches attention. Gozansu, which is obsolete today, was used primarily by female prostitutes in red light districts during pre-modern Japan (Tanaka 1999). Along with the supra-polite form a, soo de gozai­ masu ne in line 4, the writer introduces another voice humorously. In this essay, the writer is at a barbershop and is describing the conversation taking place there. The barber’s wife speaks in supra-polite form, and their daughter speaks in the emotive da style. Both try to persuade the writer why David Beckham, the British soccer player, is so popular in Japan and is treated like a “prince.” Although the writer makes comments in the base-line da style throughout the essay, toward the end he chooses supra-polite and imagined styles. The writer engages in a playful virtual roleplaying in the way he responds to the barber’s wife and daughter. The daughter explains, by using the pun, that there is a God (kami) inside Mr. Bekkami (i.e., Beckham). Obviously (while enjoying the silliness of the whole conversation) the writer is not completely convinced by the daughter’s explanation, but responds by saying “Oh, I see” anyway. His response is spiteful, an attitude that is expressed through the borrowed voice of a women from a pre-modern period (who can supercede the two women the writer faces). It is as if the writer escaped from the current discourse to another imagined space, and made one final comment, “Nice, how very nice it is!” The ironical humor comes from the “wrong” and excessive use of the supra-polite and imagined styles. Imagined styles are creatively used for expressive purposes as they represent different registers, sub-cultures, and classes. In the examples above, the Bakhtinian heterogeneity of language is creatively manipulated for expressive purposes. In the process, the writer’s voices representing multiple variations reverberate in the writings as they co-exist and interanimate among them.

6. Playing with multiple voices Having discussed style mixture in essays and BBS entries, I now turn to one BBS example to discuss five different styles proposed earlier and explore how multiple voices reverberate throughout the BBS. Again, (16) is a BBS entry for the television drama Taiyoo no Kisetsu ‘The Season of the Sun’. (16) 1 P Annani taiyoo ga giragira shiteita noni… (base-line da) ‘Although the sun was so bright…’



Playing with multiple voices 

2 P Itsuno manika, aki ga fukamatte orimasu. (supra-polite) Before we know it, we are already well into fall.’ 3 P Mi-chan koto, watashi kara, TBS-sama ni saigo no onegai ni ukagatte orimasu. My name is Mi-chan, and I am writing this to make one final plea to the honorable TBS.’ 4 P Kono watashi, moo sugata, katachi wa shoomi kigen gire to otto wa notamoote orimasu. (supra-polite) ‘This me, my husband says that my figure and form have well passed beyond the use-by date.’ 5

(Bijin zuma wa doko?) ‘(Where has the beautiful wife gone?)’

6 P Zuibun to kuroo shita haato wa, totemo furekishiburu. ‘My heart went through a lot, and as a result, it is surprisingly flexible.’ 7 P Mori Shin’ichi (furu-tt) ja-nai kedo, atatakaku tsutsumu yo. (emotive da) ‘Like Shin’ichi Mori’s song (it’s ancient, isn’t it?), my heart is warm and tolerant.’ 8 P (Ofukurosan) ‘(Dear mother)’ 9 P Ima no watashi no kadai wa “seijuku” no nimoji. ‘The theme of myself at this moment is represented by two characters, sei and juku, i.e., sei-juku ‘maturity’.’ 10

Gutto, ningentekini seijuku shitai to hibi negau watashi dearimasu. (addressive desu/masu) ‘I am wishing everyday that I would mature into a kind and humane person.’

11 P Gaaaa!! ‘Ghaaa!!’ 12 P Koko kara ga hondai desu. (addressive desu/masu) ‘From here, I’m getting to the main point.’ 13 P Takizawa-kun no koto to naru to, juuhassai, iya chuugakusei, shoo­ gakusei kamo? (base-line da) ‘When it comes to Takizawa, I become an eighteen year old; no more like a junior high student, or maybe an elementary school kid.’ 14

Henshin shiteshimau, watashi. ‘I go through a transformation.’

 Senko K. Maynard

15 P Omoi ga ichizu de mawari ga mie-nai!? (emotive da) ‘My passion is so strong that I cannot see things around me (I lose control)!?’ 16 P Kokoro de wa, Takizawa-kun, takusan ren’ai shite jibun o migaite ne. (emotive da) ‘In my heart, Takizawa, you fall in love and be a fine man,’ 17 P To iitai tokoro da ga, naze ka dorama no karami demo, jerashii ni moeru, watashi. ‘I want to say, but even when he falls in love in the drama, for some reason, I burn with jealousy.’ 18 P (Aho mitai!) (emotive da) ‘Like a fool! (Foolish me!)’ 19 P Taki: “Ren’ai yori, shigoto da yo” ‘Taki: “More than love, I concentrate on work” 20 P (Yoshiyoshi, sono chooshi) ‘(Right, that’s good.)’ 21 P Taki: “Itsumo yaroo bakari to asonderu yo” ‘Taki: ”I always hang around with guys.”’ 22 P (Hontoo????) (emotive da) ‘(Really????)’ 23 P Taki: “Sukina taipu wa Matsushita Yuki” ‘Taki: “I like the type of girls like Yuki Matsushita.”’ 24 P (Yurushite ageru ze!) (emotive da, extreme) ‘(I can allow that, for sure!)’ 25 P Nante kanji de, totemo makaineko-chama mitaku, kanojo-san ni amaete nee nante, kesshite ie-nai. (base-line da) ‘This is how I go, and I can never say, as Makaineko says, “Enjoy sweet time with your lover.”’ 26 P (Nasake naai) (emotive da) ‘(Really pitiful, aren’t I?)’ 27 P Sonna ori, Takizawa-kun jikai shuen saku, minasama no kitai ga takamatteiru mitai. (base-line da) ‘At this point, it seems that everyone’s expectations are high about the next drama starring Takizawa.’ 28 P Konkai suutsu sugata de hashitteta kara, keiji mono nante ii kamoo. (emotive da) ‘Because he was running in his suit (in the drama) this time, a detective drama may be nice.’



Playing with multiple voices 

29 P (Un, un) ‘(Yes, yes) 30 P Koko de, TBS-sama ni onegai desuu. (emotive desu/masu) ‘So, now, I have a favor to ask you, dear honorable TBS.’ 31 P Dooka, jikai shuensaku desu ga. (addressive desu/masu) ‘Please, about the next drama starring him.’ 32 P “Kindan” mono dake wa, gokanben kudasareeee odaikan-samaaa. (supra-polite, imagined) ‘Please, dear officer, please do not make a drama that features a forbidden love relationship.’ 33 P (Shufu wa kindan no nimoji ni yowai, bangumi ran de sugu me ni tsuku, desho?) (addressive desu/masu) ‘(Housewives go for the word “forbidden love”; it pops out in the list of programs, right? )’ 34 P Tatoeba, shin’nin taiiku kyooshi to joshi kookoosei no koi toka hitozuma furin mono toka. ‘For example, a love relationship between a newly appointed gym teacher and a female high school student, or an illicit affair with a housewife.’ 35 P Dame, dame, zettaini dame desu. (addressive desu/masu) ‘No, no, absolutely no way.’ 36 P Nan demo, ari no TBS-sama na node sakini kugi sashiteokimasu. (addressive desu/masu) ‘I’m warning them about this beforehand, because TBS broadcasts just about anything.’ 37 P Dekireba, moo sukoshi watashi ga seijuku? shite kara ni shite ne! (emotive da) ‘Please wait until I mature (?) a bit more, will you?’ BBS (October 8, 2002) Although this BBS contains many styles and expresses many voices, the base-line is da. Admittedly, formal desu/masu does appear frequently, but overall, when it is used, it directly addresses the television station, and its use is closely associated with addressive desu/masu rather than base-line desu/masu. BBS contributors are expected to use fictitious names, and their identities and demographic information are unverifiable. Although it is impossible to know the gender and age of the actual contributor of this BBS, the character presented in this BBS is a middle-aged housewife. The reader can reasonably be sure about this interpretation, given the fact that the writer states in line 4 that she is a wife who is no longer young.

 Senko K. Maynard

The BBS starts with supra-polite forms, as in aki ga fukamatte orimasu ‘we are already well into fall’ in line 2, onegai ni ukagatte orimasu ‘I am writing this to make one final plea’ in line 3, and notamoote orimasu ‘he says’ in line 4. In line 4 the writer describes the husband’s action as notamoote orimasu. This is peculiar, because notamou is not usually used in regard to the husband.8 The supra-polite style is excessive, and notamou is out of place. This triggers an ironical and humorous reading. By using supra-polite forms the writer creates a playful discourse. A number of sentences appear in parentheses. These reflect the writer’s inner voice in base-line da and mostly in the emotive da style. For example, in line 18, the writer yells out at herself by saying Aho mitai! ‘Like a fool! (Foolish me!)’. This expression carries with it a tone of Kansai speech, because aho is a phrase directly linked to the Kansai dialect. The reader is likely to hear this voice as linked to a specific regional variety. Within this BBS, the writer creates a conversation as shown in lines 19 through 24. Lines 19, 21, and 23 are presumably spoken by the popular actor Hideaki Takizawa (referred to as Taki or Takkii). Through this conversation between borrowed voice and constructed voice, the writer expresses emotivity that only conversation can bring. Inner feelings are expressed by the emotive da styles, often with humor. Noteworthy is the use of particle ze in line 24. Ze marks a strong and raw emotion of the writer. Although ze is not limited to masculine speech, it is stereotypically considered a part of masculine speech. Because the writer is characterized as being a housewife, it brings to the discourse the playful mixing of voices across genders. Line 30 is interesting in that it takes emotive desu/masu. The phonological change of suu (instead of su) adds to the speech-like atmosphere, and emotivity is further foregrounded. Line 32 is written in the imagined style. The writer’s expression, gokanben kudasare, odaikan-samaa ‘please, Dear officer, please do not do that’ creates a virtual world in which TBS (the name of a television station) becomes a feudal officer. The analogy here is that TBS is like a feudal officer who mercilessly enforces feudal laws and rules with absolute power. All that writer can do is to beg! The writer’s out-of-place respectful request adds to the humorous effect. As observed above, in this BBS entry, the writer mixes base-line desu/masu style (foregrounding addressivity), emotive desu/masu and emotive da styles (foregrounding emotivity), supra-polite style (foregrounding creativity), and imagined styles (for creativity). The extensive mixture of styles observed in this and other shin genbun itchitai ‘new speaking-and-writing-agreement style’ illustrates the extent to which styles are chosen not only by social factors, but by personal desires as well. In fact in these examples, the socially motivated choice is overwhelmed by the writer’s expression of emotivity and creativity. 8. Notamau ‘to say’ is an expression in classical Japanese, used in two ways, when a higher ranking person addresses toward the lower ranking person, and in reference to someone superior saying something, as a respectful form.



Playing with multiple voices 

7. Reflections In this study, based on examples drawn from contemporary Japanese written discourse, I argued that style mixture is a poetic strategy through which the writer presents multiple voices and a strategy that ultimately adds to the rich expressivity of language. I have also shown that style mixture manifests the writer’s emotivity and is used for creative purposes as well. Above all, style is not something that exists a priori to the situation and to which the writer simply conforms. Rather, style is the writer’s expression of self, a combination of many strategies that together express the writer’s emotivity and creativity. I should, at this point, discuss how this observation can be understood from a theoretical perspective. In Maynard (1998a, 2000, 2002), I have proposed a negotiative theory of linguistic meaning, i.e., the Place of Negotiation theory. From the perspective of this theory, style shift is a strategy that projects onto the emotive place as well as the interactional place. The emotive place foregrounds the speaker’s broad emotional attitudes. This is the space primarily concerned with the psychological and emotional aspects of communication. What is foregrounded here is the speaker who engages in self-expression as he or she incorporates social as well as personal emotions. Attitudes and feelings are expressed through multiple strategies, including lexicon, grammar, rhetoric, and discourse structure. The interactional place is where the partner comes into sharp focus. Within this interactional place, an interactional social atmosphere is created, coordinated, and managed while incorporating personal interests. In the interactional place, special attention is paid toward partners as well as participants in speech events. Here the main concern lies with how speaker, partner, and other participants (if any) express, understand, and manage interpersonal relations among themselves. Associated with these two dimensions of place are four related, distinguishable, but not mutually exclusive functions. These include (1) expression of emotional attitude, (2) communication of attitudes toward others (related to the emotive place), (3) management of participatory action, and (4) coordination of joint utterances (related to the interactional place). Particularly relevant to the style mixture examined in this study are the first two categories. Style mixture foregrounds expression of emotional attitude because it communicates different voices in the base-line style mixed in with other styles. It also functions in terms of communication of attitudes because the writer shifts styles as required by addressivity. Characters realized through style mixture also express emotional attitude in that the negotiative meaning is interpreted in association with those characters. Creativity realized through style mixture is also expressive, particularly in terms of emotivity. Given that the data are limited to written discourse, style mixture does not directly function in terms of management of participatory action and coordination of joint utterances. However, if the data included conversation, it is likely that style

 Senko K. Maynard

mixture would function in these areas as well. It should be added that in the mixture of style within a sentence through the [DQ + to yuu + N] structure, conversation is brought in, and the discourse evokes both interactional and emotive places. Although shin genbun itchitai is often associated with youth and with limited genres, this study has shown that it is more widespread than generally thought. It is used in essays written by writers ranging from 30 to 66 years old, and by both male and female writers. The [DQ + to yuu + N] structure, which is less associated with youth, is used by writers up to 67 years old. It is beyond the scope of this study to investigate how frequently, by whom, and under what circumstances style mixtures are or are not used. I have focused on limited data and analyzed those cases where the kind of style mixtures discussed in the present study are observed. Inquiries into variation in style and style mixtures must await future studies. The findings in this study lead to the view of style as being expressive, personal, creative, fluid, and varied. This contrasts with the traditional view of style, that is, as being primarily social, stable, and prescriptive. Given this context, it is probably not totally out of line to touch upon some pedagogical implications. As I discussed in many of my earlier works, and particularly in Maynard (1997b), style in Japanese discourse is varied and flexible, and appreciation of such variability is important when teaching Japanese as a foreign language. One concrete way to encourage such an appreciation is to encourage students to discuss cases where speech style diverts from what is conventionally expected (see Maynard [1997d]). And, it is also advisable to show that style can be used not only to fulfill social conventions but also as a tool to create effects such as parody, irony, and humor. Overall, as proposed in Maynard (1997a, 1997d), a “critical pedagogy” approach is necessary where the knowledge based on analysis of authentic discourse is incorporated in teaching. Traditional explanations of the Japanese language and its use must always be critically examined. This is particularly so in the area of style mixing, where new styles and innovative style mixtures are observed. This study has focused on two expressive aspects of style mixture, i.e., emotivity and creativity. Although I have discussed them separately, the two aspects are closely related and often overlap. The creative use of language is likely to include emotivity, and emotivity is often the source for creative adventure. One may still argue that writers may use style and style mixture primarily for social reasons. When writing a formal letter, for example, the stylistic choice is often based on social conventions. Still, as discussed in the present study, style mixture is also a strategy that the writer manipulates for the expression of multiple voices and a strategy that ultimately enriches the expressivity of language.



Playing with multiple voices 

References Bakhtin, M. M. 1971. “Discourse typology in prose.” In Readings in Russian Poetics, L. Mateika and K. Pomoroka (eds.), 176–196. Cambridge, MA: The MIT Press. Bakhtin, M. M. 1981. The Dialogic Imagination. Ed. by M. Holquist, trans. by C. Emerson and M. Holquist. Austin: The University of Texas Press. Bakhtin, M. M. 1986. Speech Genres and Other Late Essays. Ed. by C. Emerson and M. Holquist, trans. by V. W. McGee. Austin: The University of Texas Press. Endoo, O. 1998. “Tookyoo no josei no kotoba no ima.” Gengo 27 (January): 76–81. Endoo, O. 2000. “Ninki dorama no hanashi kotoba ni miru seisa, TV dorama ‘Byuutifuru Raifu’ no mojika shiryoo kara.” Kotoba 21: 13–23. Fujitani, N. 1960 [1778]. Ayuishoo. Ed. by N. Nakada and M. Takeoka. Tokyo: Kazama Shoboo. Fukao, M. 1998. “Daigakusei no keigo ishiki: Teineisa to shitashisa no choosetsu ni tsuite.” Nihongogaku Ronsetsu Shiryoo 35, 5: 110–122. Haga, Y.1962. Kokugo Hyoogen Kyooshitsu. Tokyo: Tookyoodoo. Kinsui, S. 2003. Vaacharu Nihongo Yakuwarigo no Nazo. Tokyo: Iwanami Shoten. Kokuritsu Kokugo Kenkyuujo (ed.). 2000. Shin Kotoba Shiriizu, Vol. 12, Kotoba ni Kansuru Mondooshuu–Kotoba no Tsukaiwake. Tokyo: Kokuritsu Kokugo Kenkyuujo. Maynard, S. K. 1984. “Functions of to and koto-o in speech and thought representation in Japanese written discourse.” Lingua 64: 1–24. Maynard, S. K. 1986. “The particle -o and content-oriented indirect speech in Japanese written discourse.” In Direct and Indirect Speech, F. Coulmas (ed.), 179–200. Berlin: Mouton de Gruyter. Maynard, S. K. 1991a. “Buntai no imi, da-tai to desu/masu-tai no kon’yoo ni tsuite.” Gengo 20 (February): 75–80. Maynard, S. K. 1991b. “Pragmatics of Discourse Modality: A case of da and desu/masu forms in Japanese.” Journal of Pragmatics 15: 551–582. Maynard, S. K. 1992. “Where textual voices proliferate: The to yuu clause-noun combination in Japanese.” Poetics 21: 169–189. Maynard, S. K. 1993. Discourse Modality: Subjectivity, Emotion and Voice in the Japanese Language. Amsterdam: John Benjamins. Maynard, S. K. 1994a. “Images of involvement and integrity: Rhetorical style of a Japanese politician.” Discourse & Society 5: 233–261. Maynard, S. K. 1994b. “To yuu hyoogen no kinoo—washa no hassoo, hatsuwa taido no hyooshiki toshite.” Gengo 23 (November): 80–85. Maynard, S. K. 1996. “Multivoicedness in speech and thought representation: The case of selfquotation in Japanese.” Journal of Pragmatics 25: 207–226. Maynard, S. K. 1997a. Danwa bunseki no Kanoosei: Riron, Hoohoo, Nihongo no Hyoogensei. Tokyo: Kuroshio Shuppan. Maynard, S. K. 1997b. Japanese Communication: Language and Thought in Context. Honolulu: University of Hawai‘i Press. Maynard, S. K. 1997c. “Textual ventriloquism: Quotation and the assumed community voice in Japanese newspaper columns.” Poetics 24: 379–392. Maynard, S. K. 1997d. “Manipulating speech styles in Japanese: Context, genre, and ideology.” Proceedings of the Fifth Princeton Japanese Pedagogy Workshop: 1–24. Maynard, S. K. 1998a. “Patosu toshite no gengo.” Gengo 27 (June): 34–41.

 Senko K. Maynard Maynard, S. K. 1998b. “Ventriloquism in text and talk: Functions of self- and other-quotation in Japanese discourse.” In Japanese/Korean Linguistics 7: 17–37. Stanford, CA: Center for the Study of Language and Information. Maynard, S. K. 1999. “A poetics of grammar: Playing with narrative perspectives and voices in Japanese translation texts.” Poetics 26: 115–141. Maynard, S. K. 2000. Jooi no Gengogaku: Bakooshooron to Nihongo Hyoogen no Patosu. Tokyo: Kuroshio Shuppan. Maynard, S. K. 2001a. Koisuru Futari no Kanjoo Kotoba: Dorama Hyoogen no Bunseki to Nihongoron. Tokyo: Kuroshio Shuppan. Maynard, S. K. 2001b. “Falling in love with style: Expressive functions of stylistic shifts in a Japanese television drama series.” Functions of Language 8: 1–39. Maynard, S. K. 2002. Linguistic Emotivity: Centrality of Place, the Topic-comment Dynamic, and an Ideology of Pathos in Japanese Discourse. Amsterdam: John Benjamins. Mori, A. 1979. Mori Arimasa Zenshuu, 12. Tokyo: Chikuma Shoboo. Noda, H. 1998. “‘Teineisa’ kara mita bunshoo danwa no koozoo.” Kokugogaku 194: 1–15. Ozaki, Y. 2001. “Nihongo no sedaisa wa nakunaru ka.” Gengo 30 (January): 66–72. Satake, H. 1995. “Wakamono kotoba to retorikku.” Nihongogaku 14 (November): 53–60. Seto, K. 1997. “Imi no retorikku.” In Bunka to Hassoo to Retorikku, M. Nakau (ed.), 93–183. Tokyo: Kenkyuusha. Sperber, D. and Wilson, D. 1981. “Irony and the use-mention distinction.” In Radical Pragmatics, P. Cole (ed.), 295–318. New York: Academic Press. Sperber, D. and Wilson, D. 1988. Relevance: Communication and Cognition. Cambridge, MA: Harvard University Press. Suzuki, A. 1979 [1824]. Gengyo Shishuron. Tokyo: Bunseidoo. Takasaki, M. 2002. “‘Onna kotoba’ o tsukurikaeru josei no tayoona gengo koodoo.” Gengo 31 (February): 41–47. Tanaka, A. 1999. Nihongo no Isoo to Isoosa. Tokyo: Meiji Shoin. Tokieda, M. 1941. Kokugogaku Genron. Tokyo: Iwanami Shoten. Tokieda, M. 1950. Nihon Bunpoo Koogohen. Tokyo: Iwanami Shoten. Tsuji, D. 1997. “Aironii no komyunikeeshonron.” Tookyoo Daigaku Shakai Joohoo Kenkyuujo Kiyoo 55: 91–126. Usami, M. 2001. “Poraitonesu riron kara mita ‘keii hyoogen’.” Gengo 30 (December): 18–25. Usami, M. 2002. Discourse Politeness in Japanese Conversation: Some Implications for a Universal Theory of Politeness. Tokyo: Hituzi Syobo. Vološinov, V. N. 1973 [1929]. Marxism and the Philosophy of Language. Trans. by L. Matejka and I. R. Titunik. New York: Seminar Press. Wertsch, J. V. 1991. Voices of the Mind: A Sociocultural Approach to Mediated Action. Cambridge, MA: Harvard University Press.



Playing with multiple voices 

Data references Arashiyama, K. 2003. “Tokoya dangi Bekkamu-sama.” Shuukan Asahi (July 11): 128–129. BBS of Taiyoo no Kisetsu, a television drama series, September 21 to October 10, 2002. TBS: http://.www.tbs.co.jp BBS of Bijo to Yajuu, a television drama series, March 19 to March 23, 2003. Fuji Television: http://www.fujitv.co.jp Iijima, A. 2003. “Wangan.” Shuukan Asahi (February 18): 144 Itsuki, H. 1999. Taiga no Itteki. Tokyo: Gentoosha. Konishi, D. 2003. “Dobiruban Furansu gaisoo.” Shuukan Asahi (April 4): 16. Murakami, R. 2002. Kiboo no Kuni no Ekusodasu. Tokyo: Bungenshunjuu. Nobumoto, K. 2002. “Eiga to watashi.” Kinema Junpoo, late March, 124–125. Saitoo, M. 2003. “Kangoshi to josanshi.” Gengo 32 (May): 12–13. Sawaki, K. 2002. Iruka to Tsuiraku. Tokyo: Bungeishunjuu. Shiga, A. 2002. “Feiku mono de okanemochi o yosootte mimasu?” Popeye (September 9): 4. Shooji, S. 2003. “Hyakuen udon o tabe ni iku.”.Shuukan Asahi (February 14): 54–55. Tatekawa, S. 2003. “Sadanji no terebi shadanki. Shuukan Asahi (June 13): 76. Uchida, Y. 1998. Aiiro Kairoo Satsujin Jiken. Tokyo: Koodansha. Uchidate, M 2003. “Daigakuin no nyuugakushiki.” Shuukan Asahi (March 2/9): 66–67.

Riyuu ‘reason’ for nai desu and other semi-polite forms* Mutsuko Endo Hudson Michigan State University

1. Introduction All Japanese predicates (copulas, adjectives and verbs) display the affirmative-negative distinction and may appear in the ‘plain’ -da/-ru or ‘polite’ -desu/-masu style. Located in between these two is yet another style, whose forms are basically derived by adding the polite marker -des-u to the plain counterparts.1 Let us call this third style ‘semipolite.’ See Table 1 for the inflectional endings of Japanese predicates in polite, semipolite, and plain styles.2 The present study examines the ‘semi-polite’ style in Japanese, with a particular focus on verbal negatives; that is, verb+nai desu (nonpast) and verb+nakatta desu (past).3 The use of such forms is reported to be on the rise in contemporary Japanese, but has not hitherto received much attention in the literature. An investigation of nai desu and other semi-polite style forms is important as it bears on a broader aspect of the language, namely, keigo ‘honorifics,’ as will be shown in subsequent sections. Hereafter, ‘nai desu form’ and ‘masen form’ will be used as cover terms referring to both their nonpast and past tense variants, along with the terms ‘semi-polite’ and ‘polite’ * My sincere appreciation goes to the organizers of the 2002 Japanese Speech Style Shift Symposium, K. Jones, T. Ono, and T. Vance, and to the participants, especially S. Okamoto, whose comments benefited me greatly. I am also grateful to T. Tanomura, S. Uehara and E. Fukushima for their assistance. Any error that may be contained herein is of course my own. 1. An exception is with the plain nonpast affirmative of the copula -da. Des-u replaces it in semi-polite style, instead of being added to it. 2. The ja appearing as part of the negative copulas in Table 1 is informal, and may be substituted for by its formal variant de or dewa. Certain verbs end in -da, instead of -ta in the past affirmative (cf. Tsujimura 1996). The negative morpheme -na-i, which appears with all plain predicates, is inflected the same way as an -i adjective. The nonpast affirmative copula and the nonpast and past affirmative adjective endings in semi-polite are identical to those of their respective polite counterparts, details of which will be discussed in subsequent sections. 3.

For ease of reading, I hereafter omit hyphens between morphemes, except when relevant.

Tense

Nonpast Past

Nonpast Past

Nonpast Past

Category

Copula

Adjective

Verb

-ku-na-i -ku-na-kat-ta

-(a)-na-i -(a)-na-kat-ta

-i -kat-ta

-u -ta

-u-des-u -ta-des-u

-i-des-u -kat-ta-des-u

-des-u -dat-ta-des-u

-ja-na-i -ja-na-kat-ta

-da -dat-ta

-(a)-na-i-des-u -(a)-na-kat-ta-des-u

-ku-na-i-des-u -ku-na-kat-ta-des-u

-ja-na-i-des-u -ja-na-kat-ta-des-u

Semi-polite Affirmative Negative

Plain Affirmative Negative

Table 1.  Inflectional endings of Japanese predicates

-(i)-mas-u -(i)-mashi-ta

-i-des-u -kat-ta-des-u

-des-u -deshi-ta

Affirmative

Polite

-(i)-mas-en -(i)-mas-en-deshi-ta

-ku-arimas-en -ku-arimas-en-deshi-ta

-ja-arimas-en -ja-arimas-en-deshi-ta

Negative

 Mutsuko Endo Hudson

Riyuu ‘reason’ for nai desu and other semi-polite forms 



negatives, respectively. Table 2 illustrates verbal inflections in the three styles with ik-u ‘go’ as an example. My interest in the present topic arose in 1999, when I read Miyabe Miyuki’s murder mystery called Riyuu ‘Reason’ (1998, 572 pages).4 Nai desu forms were ubiquitous in this novel. I was aware of the increasing usage of nai desu at the time, most typically in sentences with ja nai desu ka ‘isn’t it?; you see?,’ as discussed in Jinnouchi (1998: 16-20), Inoue (1998: 150-154), and elsewhere. I had also heard it used in real life. One day in Japan, for example, a high-school-age stranger said to me on a train mooshi wake nai desu! ‘excuse me!’ when he realized that he was blocking my way. The phrase caught my attention because it sounded brusque compared to the more usual mooshi wake arimasen. Table 2.  Inflections of the verb iku ‘go’ in three styles

Nonpast Past

Aff. Neg. Aff. Neg.

Plain

Semi-polite

Polite

Meaning

iku ikanai itta ikanakatta

iku desu ikanai desu itta desu ikanakatta desu

ikimasu ikimasen ikimashita ikimasendeshita

‘do/will go’ ‘do/will not go’ ‘did go, went’ ‘did not go’

The large bulk of the data for the present study are taken from the aforementioned novel Riyuu. See excerpts (1)-(4), for example.5 (1) Plain and Semi-polite styles –Ano kazoku ga jitsu wa “kazoku” dewa nakute, Sunagawa Nobuo igai wa namae mo chigau to iu koto o, anata wa gozonji datta n desu ka? “Ie, ano koro wa mada zenzen shiranakatta desu. Ki ga tsukanakatta ne.” (Ishida Naozumi: male, middle-aged) [p. 542] −−あの家族が実は「家族」ではなくて、砂川信夫以外は 名前も違う ということを、あなたはご存じだったん ですか？ 「いえ、あのころはまだ全然知らなかったです。気が つかなかった ね。」 (石田直澄：男、壮年） 4. This is a hardcover version, which is different in terms of the page numbers from the paperback version published in 2002 (619 pages). 5. In all the examples from the novel Riyuu ‘reason,’ the interviewer’s remarks are set off with short lines (–), and those of the interviewees’ are surrounded by double quotation marks, honoring the original format. The quotes are provided both in Romanization and Japanese script, in which polite style predicates are marked by double underlines, semi-polite style predicates, by bold single underlines, and plain style predicates, by single underlines. The interviewee’s name, gender, and age group are given at the end of each quote. The translations are all mine.

 Mutsuko Endo Hudson

‘–Did you know that that family was in fact not a “family” and that, other than Sunagawa Nobuo, they were all using pseudonyms? “No, I didn’t know that at all in those days. I didn’t realize it.”’ (2) Semi-polite and Polite styles “Sagamihara to Arakawaku wa tochigara mo mattaku chigau shi, kyori mo nishi to higashi ni hanarete iru noni, yahari achira no koonyuu kiboosha no naka ni dooyoo ga umareru n desu yo. Maa, muri mo nai desu ga ne. Yasui kaimono ja arimasen kara.” (Ide Yasufumi: male, middle-aged) [pp. 127-128] 「相模原と荒川区は土地柄もまったく違うし、距離も西と 東に離れ ているのに、やはりあちらの購入希望者の なかに動揺が生まれるん ですよ。まあ、無理もない ですがね。安い買物じゃありませんか ら」 （井出康文：男、壮年） “‘Sagamihara and the Arakawa Ward are completely different in terms of their character and are very far apart, but people who were interested in buying a condo there still felt uneasy. Well, I don’t blame them (It’s not unreasonable). It’s not like buying something cheap, you know.”’ (3) Plain and Polite styles “Dakedo sonotoki, Itsuko Obasan wa, Shizuchan ga kaette kiteru nante koto, hitokoto mo iwanakatta. Okashii ja arimasen ka. Watashi wa Shizuchan no osana-tomodachi na n desu noni.” (Sakata Naoko: female, age 44) [p. 170] 「だけどそのとき、逸子おばさんは、シズちゃんが帰って きてるな んてこと、ひと言も言わなかった。おかしい じゃありませんか。わ たしはシズちゃんの幼友達なん ですのに」 （坂田尚子：44歳） ‘But at that time Aunt Itsuko didn’t say a word about Shizu-chan’s being back home. That’s strange, isn’t it? I’ve been Shizu-chan’s friend ever since we were kids, you know.’ (4) Plain, Semi-polite and Polite “Sore o neta ni shite shaberichirasu nante koto wa shitaku nakatta. Shujin ni mo tomeraremashita. … Watashi mo sore made kabaoo to wa omoimasen. Demo, marude Shizuchan ga hitogoroshi o shita mitai ni iwarete ne. Soredake ja nai, Gakusei jidai no dansei kankei no uwasa nanka made hipparidashite kite torizata shitari suru no wa, anmari ja nai desu ka.” (Sakata Naoko: female, age 44) [p. 184] 「それをネタにしてしゃべり散らすなんてことはしたくな かった。 主人にも止められました。...わたしもそれまで かばおうとは思い ません。でも、まるでシズちゃんが 人殺しをしたみたいに言われて

Riyuu ‘reason’ for nai desu and other semi-polite forms 





ね。それだけじゃない、 学生時代の男性関係の噂なんかまで引っぱ り出してき て取りざたしたりするのは、あんまりじゃないです か。」 （坂田尚子：44歳）

“‘I didn’t want to gossip around about that. My husband told me not to, also. … I don’t intend to deny that. But people were talking like Shizu-chan had committed a murder. That’s not all, they were even bringing up rumors about her relationships with men during her school days, which is just too much, isn’t it?”’ Notice that, in the above excerpts, style shifts occur frequently within a speaker’s turn, just as in natural conversation. The motivations for style shift are admittedly complex, for which the reader is referred to works such as Cook (1999, 2002), Ikuta (1983, 2002), and Makino (2002), in addition to the articles in this volume. There are four main questions to be investigated in the present study: (1) at what rate various predicates appear in the nai desu form, as compared to the masen form; (2) what types of predicates commonly occur with nai desu; and (3) what functions the nai desu form serves; and (4) who uses nai desu. The remainder of the present article is organized as follows. Section 2 chronologically outlines the characterizations of desu, as well as of nai desu, offered by various linguists. Section 3 reviews two major studies that have examined the actual usage of nai desu and masen forms, and the results of two recent public polls. Section 4 presents Riyuu ‘Reason’ data, along with discussions of the results. Section 5 offers my analysis, and section 6, my conclusions and future directions.

2. Desu and nai desu6 2.1

Historical overview

The history of desu and nai desu usage is a partial reflection of that of keigo ‘honorifics.’ According to Tsujimura (1992: 449-500), keigo has undergone gradual simplification, especially during the Meiji period (1868-1912), when the strict class system declined and the idea of democracy was introduced to Japan. The nature of keigo changed from the traditional ‘class-based’ type (kaikyuu keigo) to the newer ‘social’ type (shakoo keigo). Furthermore, “the focus of keigo is shifting from sozai keigo ‘referent honorification’ to taisha keigo ‘addressee honorification’” (Tsujimura 1964). In other words, the speaker’s deference toward the addressee has come to control keigo usage more prominently than does the relative status of an intra-sentential referent vis-à-vis the speaker. 6. All the works cited except for Martin (1975) and Jorden (1987) are written in Japanese, and the translations are mine.

 Mutsuko Endo Hudson

Inoue (1998) echoes this observation, stating that teineigo ‘polite expressions,’ a newer concept, are becoming more and more dominant relative to the traditional sonkeigo ‘exalting/respect word’ and kenjoogo ‘humbling/modesty word’. The 1993 report by Kokugo Shingikai [National Japanese Language Council] lists common types of keigo errors. Among them is the use of kenjoogo for sonkeigo; e.g. Tsugi wa donata ga dete mairimasu deshoo ka ‘I wonder who [exalting] will come [humbling] out next’, using the humbling form mairimasu instead of irasshaimasu or koraremasu ‘come [exalting]’ (underlines mine).7 As discussed in section 1, semi-polite forms are generally derived by simply adding desu to their plain style counterparts. The semi-polite style then may be viewed as ‘doubly simplified’ keigo since it is simpler not only in its conceptual framework, but also in its formation.

2.2

Kindaichi

Kindaichi (1942; cited in Endo 1994) regards affirmative verb+desu forms (e.g. kuru desu ‘do/will come’, iku desu ‘do/will go’) as ‘errors,’ and their masu form equivalents (e.g. kimasu, ikimasu) as the standard. In the case of adjectives, he contends that we have no other choice but to use forms such as nai desu ‘is nonexistent’ and atsui desu ‘is hot’ since the verbal suffix masu cannot be attached to them. He further notes that these adjectival forms are maa yoi ‘sort of okay’ for men, but are bukkiraboo ‘brusque’ and tekitoo de nai ‘inappropriate’ for women.8 It seems safe to assume that Kindaichi's negative view on adjective+desu applies to verb+nai desu as well.

2.3

Tokieda

Tokieda (1950: 179) analyzes the desu in Ano kata wa watashi no sensei desu ‘That (honorable) person is my teacher’ as a keijoogo, expressing “the speaker’s respect and modesty toward the hearer.” He coined the term keijoogo ‘respect-modesty word,’ objecting to the traditional distinction of sonkeigo ‘respect word’ and kenjoogo ‘modesty word,’ since in his view ‘it is not possible for the speaker to feel one way and not the other’ (Tsujimura 1992: 194). Tokieda (1950: 179) also adds that the desu in the above sentence corresponds to da, the non-keijoogo auxiliary for ‘designation’ (shitei). This suggests that he does not distinguish the copulative desu from what is usually called

7. Tsujimura (1992: 479-484) claims that the change in usage from humbling (kenjoogo) to polite (teineigo) occurred in the old days as well, such as with haberi ‘be,’ and that the polite ending -masu is probably a result of the combined use of the humbling expressions moosu ‘say’ and mairasu ‘come.’ 8. Jorden (1987: 41) explains the two types of adjectival negatives along a similar line: as compared to -ku nai desu, -ku arimasen forms are “a bit more formal, and in the view of some speakers, are slightly more elegant.”

Riyuu ‘reason’ for nai desu and other semi-polite forms 



the politeness marker desu. Such a view is justifiable from a historical perspective since there was only one desu originally, as will be discussed in Section 2.6. Tokieda (1950: 182) states that, corresponding to the negative plain forms of verb (e.g. furanai ‘does/will not fall’), adjective (e.g. samuku nai ‘is not cold’), and the copula (e.g. tenki de nai ‘is not sunny’), their respect-modesty forms (e.g. furanai desu, samuku nai desu, tenki de nai desu) “also exist,” and that the latter forms are derived from “reversing the order of the honorification and negation.” That is, instead of the usual order of honorification (-mas-) → negation (-n), which would have produced -mase-n, (furimasen, samuku arimasen, tenki de arimasen), negation (na-i) has applied before honorification (-des-). It is interesting that, only eight years after Kindaichi’s (1942) reluctant acknowledgment of adjective +desu (cf. Section 2.2), Tokieda is treating the use of nai desu with all types of predicates as if they had already been established.9 He (1950: 181-182) further notes that nonpast affirmative verb+desu (e.g. furu desu ‘does/will rain’) is “used sometimes, but not generally,” and that past affirmative verb/adjective+desu forms (e.g. futta desu ‘rained,’ samukatta desu ‘was cold’) are “also found.” These forms are undoubtedly the result of the keigo simplification, discussed in Section 2.1.

2.4

Martin

Based on Tsujimura (1967), Martin (1975: 1029) explains that what are herein referred to as semi-polite forms “have been spreading since Meiji times, but …the verb forms such as kaeru desu = kaerimasu have been slower to spread than the adjective forms, including verbal negatives.” According to Martin (1975: 1029-1031), verb+desu (e.g. yobu desu ‘call’) “seems to be gaining popularity among younger Tookyoo speakers” as an alternative to the typical polite stylization, verb+masu (e.g. yobimasu ‘call’), and “you will also hear datta desu for desita.” He (p. 1030) states, “[s]ometimes one gets the feeling that desu can be added to just about anything to lend a touch of politeness,” adding that its abuse is “ingratiating.” He (1975: 1029-30) cites two scholars’ characterizations: Yoshida (1971) claims that the plain and polite affirmative verb+desu forms (i.e. -u desu, -masu desu) are “not very Tookyoo-like – at most, hama talk,” and Tsujimura (1967), that the -masu desu form is “used by merchants.”10 One can surmise that the forms were not highly regarded in the 1960s and 1970s, at least by scholars. As for the origin of verbal/adjectival+desu forms, Martin (1975: 1029) says that they “are often considered to be abbreviations of verbal/adjectival+no desu,” and that “this may be the historical origin, but synchronically the two should be kept distinct.” 9. In fact, the adjective+desu form was “recommended for general use” in a 1952 report, Korekara no Keigo [Polite Language in the Future], prepared by Kokugo Shingikai [National Japanese Language Council]. 10. It is unclear whether the word hama is to be interpreted as ‘beach’ or an abbreviation of Yokohama.

 Mutsuko Endo Hudson

2.5. Teramura Teramura’s (1984: 53) reference to the nai desu form is brief, appearing only in a footnote on the conjugation of masu. The common polite counterpart of yomanai ‘does/will not read’ is yomimasen, and yomanai desu is not often used....[T]he past tense forms yomimasendeshita and yomanakatta desu ‘did not read’ both seem to be in use, [but] some people consider the latter unnatural.

These remarks suggest that there was still a stigma attached to the semi-polite style even in the 1980s, despite the fact that the forms had been “spreading since Meiji period” (Martin 1975: 1029). As for the morpheme desu, Teramura (1984: 54) distinguishes two types: one is the copula, or a hanteishi ‘evaluative/assessing word,’ and the other, a jodooshi ‘auxiliary word.’

2.6

Inoue

Inoue (1998: 154-157) gives a historical account, summarized below. While the morpheme masu has been in use for hundreds of years since the Muromachi Period (1392-1573), the appearance of desu dates back only to the mid-1800s.11 Desu was originally used as a copula, as in noun+desu, while adjective+gozaimasu and verb+masu were other standard endings. It was not until just before the World War II that adjective+desu began to be used. Although the degree of its acceptability still varies according to speakers’ age and regions, adjective+desu is regarded as natural today, especially when sentence-final particles such as ne ‘tag question’ and yo ‘assertion’ are attached. 12 Desu then started to be used with verbs, though mainly in the tentative form deshoo. Such usage of desu was prompted by the need to distinguish ishi ‘volition’ from suiryoo ‘conjecture,’ both of which were expressed by mashoo in the old days; e.g. ikimashoo ‘I will go, let’s go’ and ‘(they) (will) probably go.’ Mashoo would then be used mostly to express volition, and deshoo, only conjecture. This deshoo usage in turn set the stage for the appearance of verb+desu so that iku desu yo ‘go’ and iu desu yo ‘say,’ for example, started to appear in the transcriptions of discussions in magazines. The desu usage thus developed ‘sporadically,’ reflecting the fact that language change never occurs all at once. In a survey of 600 junior high schools around the country, verb+desu

11. According to Martin (1975: 1028), da, the plain counterpart of desu, goes back to the Muromachi period. 12. The functions and meanings of the sentence-final particles ne and yo are much more complex than are indicated here. For details, see, for example, Cook (1990), Hudson (2003), Kamio (1990, 1994, 1998), Lee-Wong (2001), Maynard (1989), Ohta (1994), Takubo and Kinsui (1996), Tanaka (2000) and Uyeno (1971).

Riyuu ‘reason’ for nai desu and other semi-polite forms 



was found to be quite common in Kyushu.13 Inoue (1998) predicts that at some future time desu will be used with nouns, adjectives and verbs, thereby simplifying the keigo system even further.

2.7

Summary

The morpheme desu first appeared in the mid-1800s as a copula used with nouns. Its use with adjectives began in the mid-1900s, and soon after that, it began to be used with verbs. The verbal negative (e.g. ikanai desu ‘not go’) spread faster than did its affirmative counterpart (e.g. iku desu ‘go’). As the negative morpheme nai is conjugated in the same way an adjective is, the emergence of nai desu owes much to the establishment of the adjective+desu (especially followed by sentence-final particles such as ne ‘tag question’ and yo ‘assertion’), which was considered substandard when it first appeared. The negative verb+desu form is gaining popularity today, especially among young people. Although the affirmative version still remains ‘marked,’ it is possible that that, too, will be more prevalent in the future as part of keigo simplification.

3. Previous studies 3.1

Tanomura

Previous research on nai desu and masen usage has been scarce, and, as seen in the previous section, most of the claims have yet to be substantiated with quantitative data. Tanomura (1994: 52) attributes the traditional neglect of the nai desu form in research to two major factors: the tsuujiteki jijitsu ‘diachronic fact’ that the form developed later than did the masen form and, stemming from this fact, kihan ishiki ‘prejudice toward the canonical forms.’ To remedy the situation, he conducted a large-scale corpus-based study using newspaper articles dating from 1989 to 1992 as his data, totaling approximately 230 million characters in length–the equivalent of several hundred books. Tanomura’s (1994) study made many important discoveries, seven of which are listed below. 1. For the negative of the existential verb aru ‘exist, be,’ arimasen was an overwhelming favorite over nai desu, at 82.7% vs. 17.3% overall. The number of tokens according to the tense is given in Table 3.

13. The year of the survey is not provided; Inoue (1998:156) simply says katsute no choosa ‘survey in the past.’ See also Inoue (1995) and Inoue and Yarimizu (2002: 153).

 Mutsuko Endo Hudson

Table 3.  Tokens of arimasen and nai desu ‘not exist, copula’ in Tanomura (1994)

Nonpast Negative Past Negative Total

arimasen

nai desu

Total

7295 (82.2%) 523 (90.3%) 7818 (82.7%)

1584 (17.8%) 56 (9.7%) 1640 (17.3%)

8879 (100%) 579 (100%) 9458 (100%)

Note that the tokens include not only arimasen and nai desu that function as the main verb ‘not exist,’ but also those appearing as part of the negative inflections of the copula (e.g. hon dewa arimasen/nai desu ‘is not a book’), auxiliaries (e.g. …shite arimasen/ nai desu ‘have not done…’), and modals (e.g. …suru made mo arimasen/nai desu ‘there is no need to do…’). 2. With regular (i.e. non-existential) verbs, the rate of verb+nai desu occurrences, compared to those of verb+masen, was even lower, at less than 2% in both nonpast and past tenses. This is shown in Table 4. 3. A majority (88.8%) of nai desu tokens were immediately followed by one type of sentence-final particle or another. Tanomura (1994: 58) speculates that this may be related to the fact that adjective+desu, too, sounds more natural with sentencefinal particles (cf. Section 2.6). 4. In the case of existential verbs, arimasen occurred much more frequently than did nai desu when there was no sentence-final particle attached to it (98.7%), but it was a different story when there was one. With ka ‘question,’ yo ‘assertion,’ and ne ‘tag question,’ nai desu was preferred over arimasen in 58.0% of the non-past sentences, and similar results obtained with past sentences. With non-existential verbs, in contrast, the occurrence of verb+nai desu was very low even with these sentence-final particles attached to them. 5. Ja nai desu ka ‘isn’t it…?, you see?’ occurred almost twice as frequently as did ja arimasen ka. 6. In ‘volition’ (ishi) / ‘invitation’ (kanyuu) sentences meaning ‘why not do…?’, verb(y)oo ja nai desu ka (2 tokens) occurred conspicuously less often than did verb-(y) oo ja arimasen ka (73 tokens). Table 4.  (Non-existential) verb+masen and verb+nai desu tokens in Tanomura (1994)

Nonpast Negative Past Negative Total

verb-masen

verb-nai desu

Total

16631 (98.4%) 1676 (98.1%) 18307 (98.4%)

268 (1.6%) 32 (1.9%) 300 (1.6%)

16899 (100%) 1708 (100%) 18607 (100%)

Riyuu ‘reason’ for nai desu and other semi-polite forms 



7. Nai desu was found to be more informal than masen, as expected. Two pieces of evidence were offered: nai desu was preferred over masen when ja (as opposed to the more formal dewa) preceded it and also when a main clause was missing, both of which are common features in informal style. A majority of arimasen tokens co-occurred with dewa (85.3%), while a majority of nai desu tokens co-occurred with ja(a) (77.6%).

3.2

Public opinion polls, 1997 and 199914

As did Tanomura (1994), the 1997 and 1999 public opinion polls both found that, overall, masen was more commonly used than was nai desu. The poll results also showed, however, that the use of nai desu was on the rise. Regarding ja nai desu ka ‘isn’t it…?, you see?,’ the 1997 poll results were that the youngest group, ages 16-29, had the highest rate of users at 37-38%, and that the percentage decreased dramatically to 12% with people in their 30s, then gradually to 9% with those in their 40s, 8% with those in their 50s, and so on. In the 1999 poll, too, more than 10-30% of the 2200 respondents acknowledged using nai desu instead of, or in addition to, masen in one situation or another. Their choice of form depended on the functions of the sentences, such as statement, question, or invitation. When asked which form they would use to say ‘isn’t it the case that...?,’ quite an amazing figure, 70%, chose...n ja nai desu ka over...n ja arimasen ka. Recall that in Tanomura’s 1994 study, too, ja nai desu ka ‘isn’t it…?, you see?’ was almost twice as common as ja arimasen ka. At this writing in 2004, the 1997 survey participants classified in the ages 16-29 group would be in their twenties and thirties. Assuming that they have continued to use (n) ja nai desu ka, and that today’s ages 16-29 group have grown up hearing that form almost exclusively and not much (n) ja arimasen ka, it seems reasonable to expect that the usage of the former will steadily rise each year and that other uses of nai desu, such as verb+nai desu, will likely rise with it.

3.3

Fukushima and Uehara

Fukushima and Uehara (2001) is an important study as it is to my knowledge the first to compare nai desu and masen usage using naturally occurring data. They used two types of data: TV commercials totaling 48 hours in length and twelve dyadic interviews between graduate and undergraduate students who had never met before. The interviews lasted about 100 minutes and were all recorded. One of their major findings was that there was no occurrence of nai desu among the 24 tokens in the commercials. The authors speculate that masen forms are preferred because the producers probably opt for a form that is “more widely accepted” in order to appeal to as wide a range of 14. Both surveys were conducted by Bunkacho Bunkabu Kokugoka (the Japanese Language Division of the Culture Section of the Cultural Agency). 1997 results are from Inoue (1998) and Jinnouchi (1998), and 1998 results are from Fukushima and Uehara (2001).

 Mutsuko Endo Hudson

the general public as possible. In the interview data, on the other hand, nai desu (92 tokens) occurred almost 3 times more frequently than did masen (33 tokens). The authors classified the predicates preceding masen and nai desu into existential verbs, (other) verbs, adjectives, copulas, idioms, and modality expressions. The last category includes expressions such as kamo shiremasen/shirenai desu ‘may’ and wake ja arimasen/nai desu ‘it does not follow that…’ With non-existential verbs, the total number of tokens for nai desu (26 tokens) was higher than those for masen (23 tokens), but the numbers were not significantly different. With existential verbs, in contrast, the difference was striking: 21 nai desu tokens vs. just 2 arimasen tokens. The results of Fukushima and Uehara (2001) are shown in Table 5.15 Fukushima's and Uehara's (2001) findings concerning nai desu usage with both existential and non-existential verbs are surprising, considering Tanomura’s (1994) results and also historical facts, discussed in section 2. It turns out that the numbers are actually skewed; almost a quarter (8 tokens; 24%) of the total 33 masen tokens and nearly half (41 tokens; 44.6%) of the total of 92 nai desu tokens are attributed to a single (male) individual. The outlier problem aside, Table 6 shows gender differences (or a lack thereof), which are supplied here based on the figures in the original article. Figures according to tenses are not available. Two of the tendencies found in Fukushima and Uehara (2001) are as follows. First, the masen form is more common when starting a conversation and/or a new topic; five out of seven speakers who used both types of negative forms used masen in such cases. Second, the masen form is preferred at the point when speakers conclude their turn, having expressed their main point, while the nai desu form is preferred when they give background information and continue.16 Table 5.  Tokens in Fukushima and Uehara (2001) by predicate types masen Verbs Existential verbs Adjectives Copula Idioms Modality Expressions Total



23 2 0 0 4 4 33

(46.9%) (8.7%) (0%) (0%) (80.0%) (9.5%) (26.4%)

nai desu

26 21 1 5 1 38 92

(53.1%) (91.3%) (100%) (100%) (20.0%) (90.5%) (73.6%)

Total 49 (100%) 23 (100%) 1 (100%) 5 (100%) 5 (100%) 42 (100%) 125 (100%)

15. As the percentages were not included in the original article, I added them. 16. Fukushima and Uehara (2001) cite the following as evidence for their claim that nai desu is less formal than masen: informal contracted variants occurred only with nai desu, and not with masen; e.g. su [← desu] in suru n ja nai su ka ne ‘I guess I do’; n [← ra] in waka-n-nai desu ‘I don’t know.’ Actually, this is to be expected because (nai)-su and waka-n-(nai) are only compatible with nai desu and not masen (suru n ja arimasen ka ne; wakarimasen).

Riyuu ‘reason’ for nai desu and other semi-polite forms 



Table 6.  Tokens in Fukushima and Uehara (2001) by gender

Male Female Total

masen

nai desu

Total

25 (26.9%) 8 (25.0%) 33 (26.4%)

68 (73.1%) 24 (75.0%) 92 (73.6%)

93 (100%) 32 (100%) 125 (100%)

4. The novel Riyuu ‘Reason’: the data and results 4.1

Data collection and examples

The data for the present study were collected from Miyabe Miyuki’s Riyuu ‘Reason’ (1998: 572 pages), a murder mystery that takes place in Tokyo and its vicinity.17 It is a work of ‘fiction’ (Miyabe 1998: 571), almost entirely comprised of ‘interviews’ with various characters, such as the suspect, his and the victims’ families and acquaintances, neighbors, and a lawyer.18 The interviewer remains the same throughout, and, though not clearly stated, seems to be a woman gathering information to write a book about the crime. The interviews are conducted predominantly in the desu/masu style, as the interviewees are mostly adults, and are strangers to the interviewer, and both parties seem to be keeping their psychological distance from each other. As such, the context is conducive to semi-polite and polite style. Plain style ‘speech’ does occasionally occur, such as when the interviewer talks to children, in rare conversations among family members, and also in some desu/masu style conversations (see excerpts [1], [3], and [4] in Section 1.2). From all the “spoken” portions, every negative phrase in plain, semi-polite, and polite styles was extracted along with several clauses before and after it, noting each speaker’s identification. The total number of tokens collected was 308. For reference, affirmative semi-polite (i.e. plain affirmative predicate+desu) tokens were also extracted, and these tokens totaled 21 (e.g. omotta desu ‘thought,’ datta desu ‘was’). In total, 34 of the characters who appear in the story use one negative form or another. They include men, women and children with various backgrounds, ranging in age from 14 to 75. The age groups were divided into three, 29 or younger (see excerpts [5] and [6]), 30s-50s (excerpts [7] and [8]), and 60+ (excerpts [9] and [10]).

17. Although the language used in a novel may not be considered completely natural, it does constitute authentic language, written by and for native speakers of Japanese. This may be especially true for this novel because the author is popular in Japan and this particular book is a bestseller that has won the prestigious Naoki Prize. 18. There is some narration as well, but it appears in plain style, as is typical.

 Mutsuko Endo Hudson

(5) –Sono koro kimi wa, shiranai hito tachi ga 2025-gooshitsu ni iru koto o, donna fuu ni kaishaku shite ita no? Kimi nari ni nattoku dekiru setsu wa kangaetsuite ita? “Kaishaku nante, shiyoo ga nakatta desu” Koito Takahiro: male, age 14) [p. 356]

−−そのころ君は、知らない人たちが二○二五号室にいるこ とを、ど んなふうに解釈していたの？ 君なりに納得 できる説は考えついて いた？ 「解釈なんて、しようがなかったです」 （小糸孝弘：男、14歳） ‘–In those days how did you interpret the fact that strangers were living in room 2025? Did you have some convincing theory of your own? “There wasn’t any way to interpret it.”’

(6) “Kenka shita toki koofun shite itta kotoba ni sonnna ni kodawattara, Oniichan kawaisoo ja nai desu ka.” (Ishida Yukari, female, high school age) [p. 298]

「喧嘩したとき興奮して言った言葉にそんなにこだわった ら、お兄 ちゃん可哀想じゃないですか。」 （石田由香利：女、高校生）[p. 298] '"If you keep harping on something Naoki (older brother) said in a huff during the fight, that’s unfair, isn’t it.”’



(7) “Tookei o totta wake dewa nai desu shi, mata konna koto wa honrai tookei no toriyoo mo nai koto nande, watashino keekensoku kara kuru inshoo de shika arimasen ga, doomo shuugoo juutaku to iu no wa, juutaku to shite no kuoriti ga agareba agaru hodo, nyuukyosha soogo no kooryuu no doai ga sagaru keikoo ga aru mitai na n desu yo.” (Ide Yasufumi: male, middle-aged) [p. 113]

「統計をとったわけではないですし、またこんなことは本 来統計の とりようもないことなんで、私の経験則から くる印象でしかありま せんが、どうも集合住宅という のは、住宅としてのクオリティがあ がればあがるほど、 入居者相互の交流の度合いが下がる傾向がある みたい なんですよ。」 （井出康文：男、壮年） “‘It’s not like I've gotten statistics, and, besides, this is not the kind of thing you can compile statistics for, so it’s not anything more than a theory based on my experience, but there seems to be a tendency that, the higher the quality of the multiple dwelling, the less the residents interact with each other.’”

(8) “Doa ga hiraku toki ni, naka ni dare ka notte iru kamo shirenai to omotte, waki ni yokemashita. Dooshite ka wakaranai desu ga, dareka notte isoo na kanji ga shita n desu” (Kasai Mieko: female, middle-aged) [p. 21]

Riyuu ‘reason’ for nai desu and other semi-polite forms 



「ドアが開くときに、なかに誰か乗っているかもしれない と思っ て、脇に避けました。どうしてかわからないで すが、誰か乗ってい そうな感じがしたんです」 (葛西美枝子：女、壮年) “‘When the door opened, I stepped aside thinking that someone might be in [the elevator]. I don’t know why, but I just felt like someone would be there.”’

(9) “Watashi wa shinde anoyo ni itta toki oyaji ni kaomuke dekinai desu yo. Datte, watashi wa oyaji no tsukutta mono o nakushite shimau dake no jinsee datta to iu koto ni natte shimaimasu wa naa, sore jaa ne” (A: male, age 75) [pp. 234-235]

「わたしは死んであの世に行ったとき親父に顔向けできな いです よ。だって、わたしは親父の作ったものをなく してしまうだけの人 生だったということになってしま いますわなあ、それじゃあね」 （Ａ：男、75歳） “‘When I die and go to the other world, I won’t be able to face my father. Because my life would amount to be nothing but one in which I lost things my father had created. If that’s the case...”’ (10) “Mise o tsugu n ja nakatta no ka tte, awatete kikimashita. Murijii suru tsumori wa nakatta desu keredomo, nanishiro Naozumi ga kodomo no koro kara zutto sono tsumori de orimashita kara, ittai doo shita n daroo to, sore bakkari odoroite shimaimashite ne” (Ishida Kinue: female, age 60+) [p. 270] 「店を継ぐんじゃなかったのかって、あわてて訊きました。 無理強 いするつもりはなかったですけれども、なにしろ直澄が子供のころ からずっとそのつもりでおりましたから、いったいどうしたんだろ うと、そればっかり 驚いてしまいましてね」 （石田キヌ江：女、60歳＋） “‘Weren’t you going to take over the store? I asked him in a fluster. I didn’t have any intention of forcing it on him, but for all I knew he was set for that ever since he was little, so I was shocked, not knowing what on earth had happened.”’

4.2

Results

The predicates occurring with masen and nai desu were classified into the following categories: existential verbs, (other) verbs, auxiliary verbs, modality expressions, idioms, adjectives and copulas. Table 7 presents examples of semi-polite forms appearing with various predicates.

 Mutsuko Endo Hudson

Table 7.  Semi-polite style examples in Riyuu classified by predicate Predicate

Tense

Aff./Neg Examples

Existential verbs

Nonpast Neg. Past Neg. Aff.

(akibeya wa) nai desu ‘there isn’t (a vacancy)’ (jikkan ga) nakatta desu kara ne ‘there wasn’t (a feeling that it was real)’ (kikime) atta desu nee ‘there was (an effect)’

(Other) verbs

Nonpast Neg. Past Neg. Aff.

wakaranai desu (ga) ‘don’t know, (but…)’ shiranakatta desu (yo) ‘didn’t know’ yonda desu (yo) ‘read’

Auxiliary Verbs

Nonpast Neg. Past Neg. Aff.

(shite) nai desu (yo) ‘didn’t do it’ (kangaete) inakatta desu (yo) ‘had not (thought about it)’ (tte koto ni) natteta desu (kara ne) ‘(that) was assumed, (so…)’

Modality expressions

Nonpast Neg.

(chigatteta) kamo shirenai desu (yo) ‘may have been (wrong)’ (wakara) nai demo nakatta desu (kara) ‘it wasn’t that I didn’t (understand) (so)’

Past

Neg.

Idioms

Nonpast Neg. Past Neg.

machigai nai desu (yo ne?) ‘no mistake about it, (right?)’ mooshi wake nakatta desu ‘felt bad’

Adjectives

Past

(yo) ku nakatta desu (yo) ‘wasn’t (good)’

Copulas

Nonpast Neg. Neg. Past

Neg..

Aff.

(hito-chigai) ja nai desu (yo) ‘isn’t (a mistaken identity)’ (warui kaisha) ja nakatta desu (yo) ‘wasn’t (a bad company)’ (taihen) datta desu (yo) ‘was (tough)’

Of the 308 total negative tokens in the present data, the overall rate of masen and nai desu favored masen at a rate of about 2 to 1, with 211 tokens (68.5%) vs. 97 tokens (31.5%). A summary of Riyuu tokens is given in Table 8. Table 8.  Summary of masen and nai desu tokens in Riyuu* Masen Nonpast Neg. Past Neg. Total

masen masendeshita

Nai desu 129 (66.5%) 82 (71.9%) 211 (68.5%)

Total

nai desu

65 (33.5%)

194 (100%)

nakatta desu

32 (28.1%)

114 (100%)

97 (31.5%)

308 (100%)

*Tokens in which n is used instead of nai are excluded (with 4 verbs and 1 auxiliary verb); e.g. wakaran desu ‘don’t know/understand’, kiitoran desu ‘haven’t heard.’

Riyuu ‘reason’ for nai desu and other semi-polite forms 



Table 9.  Riyuu tokens by predicate type Existential Verbs Forms Nonpast Neg. Past Neg. Total

Masen

Nai desu Masen

8 8 (50.0%) (50.0%) 6 4 (60.0%) (40.0%) 14 12 (53.8%) (46.2%) Idioms

Forms Nonpast Neg. Past Neg. Total

(Other) Verbs

Masen Nai desu 3 9 (25%) (75%) 4 1 (80%) (20%) 7 10 (41.2%) (58.8%)

Auxiliary Verbs

Nai desu Masen

38 (76%) 42 (84%) 80 (80%)

12 (24%) 8 (16%) 20 (20%)

14 (93.3%) 11 (73.3%) 25 (83.3%)

Nai desu Masen 1 (6.7%) 4 (26.7%) 5 (16.7%)

Copula

Adjectives

Masen Nai desu

Masen Nai desu

9 (56.3%) 2 (33.3%) 11 (50.0%)

7 (43.7%) 4 (66.7%) 11 (50.0%)

1 (100%) 1 (50%) 2 (66.7%)

Modalities

0 (0%) 1 (50%) 1 (33.3%)

Nai desu

56 28 (66.7%) (33.4%) 16 10 (61.5%) (38.5%) 72 38 (65.5%) (34.5%) Total 194 114 308

With existential verbs, the rate of masen (14 tokens, 53.8%) and nai desu (12 tokens, 46.2%) occurrences was not significantly different. On the other hand, masen was four times more frequent than nai desu (80 % vs. 20%) with non-existential verbs, five times more frequent with auxiliary verbs (83.3% vs. 16.7%), and almost twice as frequent with modals (65.5% vs. 34.5%). Interestingly, however, nai desu (58.8%) is somewhat preferred to masen (41.2%) with idioms in general, and especially in nonpast tense (75% vs. 25%). In the case of the copula, the rate of masen and nai desu occurrences is even, with 11 tokens each. This is undoubtedly related to the roughly even situation with existential verbs, since the negative forms of the copula, ja arimasen and ja nai desu, contain the existential predicates arimasen and nai desu, respectively. Nothing significant can be said of adjectives with such a small number of tokens–2 for masen and 1 for nai desu. Table 9 presents the number of tokens classified by predicate type in Riyuu. Tables 8 and 9 show that the tense does not affect the choice of the forms significantly. An exception is with idioms: nai desu is preferred over masen in nonpast (3 masen vs. 9 nai desu tokens), but masen is preferred in past (4 masen vs. 1 nai desu tokens). The idiom+nai desu numbers also show a big gap between nonpast (9 tokens) and past (1 token). It may well be an accident, but one possible reason is that negative idioms such as machigai nai desu ‘no mistake about it’ (4 tokens) and muri wa/mo nai desu ‘not unreasonable’ (3 tokens) are interpreted as ‘adjectives’ containing nai, just like tondemonai ‘no way!’ and tsumaranai ‘boring.’ If so, there would be only one way to make a polite form; i.e. by adding desu to it, as with all adjectives. See Table 10 for all the idioms that end in a negative form.

 Mutsuko Endo Hudson

Table 10.  Idioms in the negative in Riyuu Tense

Masen

Nai desu

Non-past

sumimasen ‘I’m sorry’ nan demo arimasen ‘it’s nothing’ mooshi wake arimasen ‘I’m sorry (polite)’

tamaranai desu yo ‘it’s unbearable’ machigai nai desu {∅ / yo ne? / ne } ‘there’s no mistake about it’ muri wa/mo nai desu {∅ /ga ne/kedo ne} ‘it’s not unreasonable’ mooshi wake nai desu ga ‘I’m sorry (polite)’

Past

tamarimasendeshita yo ‘it was unbearable’ toritsuku shima mo arimasen deshita yo ‘There was no way to talk to (him)’ doo to iu koto mo arimasendeshita wa ne ‘it wasn’t anything to speak of ’ aita kuchi ga husagarimasendeshita ‘I was so shocked I was speechless’

môshi wake nakatta desu ‘I was sorry (polite)’

Table 11 lists the predicates co-occurring with nai desu in Riyuu, in order of frequency. Fukushima’s and Uehara’s (2001) ranking (Table 5, section 3.3) is repeated here on the right. The ranking of the present study using a novel and that of Fukushima and Uehara (2001) using natural conversation data coincide to an amazing degree. In the present study, ‘auxiliary verbs,’ such as verb+te iru ‘progressive, stative’ and verb+te kureru ‘benefactive’ are counted separately from ‘verbs,’ while these categories are classified as just ‘verbs’ in the latter study. If those two categories in Riyuu are collapsed, the Table 11.  Nai desu tokens in Riyuu and Fukushima and Uehara (2001) by predicate types Riyuu

Predicate type Nai desu

Rank

Masen

Fukushima and Uehara (2001) Total

Rank

Nai desu

Masen

Total

1

Modal. exps.

38 (34.5%)

72 (65.5%)

110

1

38 (90.5%)

4 (9.5%)

42

2

Verbs

20 (20.0%)

80 (80.0%)

100

2

26 (53.1%)

23 (46.9%)

49

3

Exist. Verbs

12 (46.2%)

14 (53.8%)

26

3

21 (91.3%)

2 (8.7%)

23

4

Copula

11 (50.0%)

11 (50.0%)

22

4

5 (100%)

0 (0%)

5

5

Idioms

10 (58.8%)

7 (41.2%)

17

5

1 (20.0%)

4 (80.0%)

5

6

Aux. verbs

5 (16.7%)

25 (83.3%)

30

7

Adjectives Total

1 (33.3%)

2 (66.7%)

3

97 (31.5%)

211 (68.5%)

308

6

1 (100%)

0 (0%)

1

92 (73.6%)

33 (26.4%)

125

Riyuu ‘reason’ for nai desu and other semi-polite forms 



Table 12.  Most frequent nai desu expressions and their masen counterparts in Riyuu Nai desu (#)

Masen (#)

Meaning

ja / dewa / demo nai desu (31 total) ja nai desu ka (8) n ja nai desu ka (8) wake ja nai desu (8) n ja nai desu (7) nai desu (12) shiranai desu (5) wakaranai desu (5)

ja / deshika arimasen (26 total) ja arimasen ka (7) n ja arimasen ka (2) wake ja arimasen (8) n ja arimasen (9) arimasen (11) shirimasen (13) wakarimasen (7)

‘is not…’ ‘…, isn’t it?’ ‘isn’t it that …?’ ‘is not the case that…’ ‘is not that…’ ‘not exist’ ‘not know’ ‘not understand/know’

newly formed category of ‘verbs’ will still be in the second place, thereby making the ranking completely identical in the two studies. The nai desu form that most frequently appeared in the Riyuu data was the copulative ja nai ‘is not’ (31 tokens). Then came the negative of aru ‘exist’ (12 tokens), shiru ‘know’ (5 tokens), and wakaru ‘know/, understand’ (5 tokens). It would be interesting to find out if this finding is particular to this novel or if it is because these verbs generally enjoy high frequency in actual language use. Table 12 lists the nai desu expressions occurring five times or more, along with their masen counterparts and their figures. Note that the total numbers of tokens are different for masen (211 total) and nai desu (97 total), as was shown in Table 8.

4.3

Summary

In this section, we examined nai desu and masen tokens in Riyuu data, in terms of the co-occurring predicates, idiomatic usage, the most frequent nai desu phrases, and overall figures. Overall, masen is favored over nai desu, which is not surprising. With existential verbs, masen is only slightly favored over nai desu, and with copulas, masen and nai desu occurred at exactly the same rate. In contrast, masen is more frequent than is nai desu with non-existential verbs, auxiliary verbs, and modality expressions. Idioms are the “black sheep of the family,” with nai desu being more frequent than masen. The frequency ranking of the predicate types co-occurring with nai desu in Riyuu is as follows: modality expressions, (non-existential) verbs, auxiliary verbs, existential verbs, copulas, and idioms. This ranking is exactly the same as that in Fukushima and Uehara (2001), which analyzes natural data. This seems to show that, although a novel represents a make-believe world, it can reflect language use in the real world, at least in the ways perceived by the writer, and can prove useful in linguistic analysis.

 Mutsuko Endo Hudson

5. Analysis 5.1

Nature and functions of semi-polite style

As Jorden (1987: 32-33) aptly points out, “...there is no neutral style in the Japanese language.... Every use of the language requires a stylistic choice.” Furthermore, there are many cases of mixed style usage within the same speaker’s turn, such as seen in excerpts (1)-(4) in section 1.2. One thing for certain is that we cannot entirely attribute the use of a certain style to the formality level of the context, the interlocutors’ relative rank, and/or their familiarity with each other, since these factors often remain constant throughout the discourse segment. With regard to ‘distal’ (i.e. desu/masu) style, Jorden (1987: 32) observes: [It] indicates that the speaker is showing solicitude toward, and maintaining some linguistic distance from, the addressee, i.e., s/he is being less direct and more formal as a sign of deference to the person addressed (and/or the topic of discussion), rather than talking directly, intimately, familiarly, abruptly, or carelessly.

Jorden’s characterization of distal style seems to apply to semi-polite style, though partly, and not entirely. There is no question that a nai desu sounds more informal than does masen, for which two pieces of evidence are offered in Tanomura (1994), as discussed in section 3.1. At the same time, it clearly sounds more polite and less “intimate, familiar, abrupt or careless” than does its plain counterpart. Plain predicate+desu is herein called “semi-polite” style precisely because it is located between plain and polite styles in terms of politeness levels. Consider constructed examples (11)-(13), in which the speaker has received an unreadable Japanese message, illustrating the differences. (11) Polite style (Sumimasen ga,) kore yomemasen. ‘(I’m sorry, but) I can’t read this.’ (12) Semi-polite style (Sumimasen ga / Warui kedo,) kore yomenai desu. ‘(I’m sorry / Sorry, but) I can’t read this.’ (13) Plain style (Warui kedo,) kore yomenai. ‘(Sorry, but) I can’t read this.’ Note that, rather than a tertiary distinction, politeness is a matter of degree on a continuum, just as are other situational categories, such as formality, closeness, and rank. They all can and often do change within a conversational exchange and even within one speaker’s turn, as the speaker “self-contextualizes” (Maynard 1989). Turning to the functions, I analyze that desu in semi-polite style functions in two ways. One is to signal the speaker’s psychological distance from the hearer, and the



Riyuu ‘reason’ for nai desu and other semi-polite forms 

other is to impart the meaning that the speaker is explaining or assessing a situation. The term “psychological distance” is understood here as the speaker’s assessment that the hearer is not a member of his/her in-group at the time of the utterance and so sounding familiar would be inappropriate. In other words, the major effect of adding desu is to avoid being too familiar or rude. This distance is responsible for an utterance in semi-polite style sounding more indirect and, as a consequence, politer than its plain style counterpart. In my view, desu does not express solicitude toward the addressee, which may be why semi-polite style sounds less polite than polite style. A piece of support comes from the following. A young man whom I supervised daily several years ago would say X to mooshimasu ‘My name [humbly] is X’ every time he telephoned me. This phrase, however, can only be used at the first encounter, and from the second time on, a caller is supposed to say X desu ‘I’m X’ or X de gozaimasu ‘I am [hyperpolite] X’ to identify him/herself. It would have been odd, however, for him to use de gozaimasu, as it would sound excessively polite, especially for a young man. It is not hard to imagine that he felt X desu was too short and not duly polite talking to his supervisor, and so resorted to the longer, politer X to mooshimasu. Since then I have noticed other young people saying X to mooshimasu in the same infelicitous situations as well. The other function of desu, to impart the meaning that the speaker is explaining or assessing a situation, is perhaps the vestige of its dantei ‘assertion’ function (hantei ‘evaluative, assessing’ in Teramura [1984]). The attachment of desu, either to an affirmative or negative verb, turns (or keeps) a sentence a stative one, rather than an active one expressing actions and events; that is, rather than ‘X will/does/did (not) do…’ it has nuances of ‘X will/does/did (not) do…, and that’s how it is.19 The underlined part of the meaning conveyed by desu can make a speaker appear as a “know-itall.” It is also responsible for the brusqueness of the style; e.g., Mooshiwake nai desu ‘(lit.) There’s no excuse, and that’s how it is’ (cf. section 1). In sum, semi-polite style uniquely seems to convey to the addressee the metalinguistic message: “This is how things are as I see it, and also because I don’t want to sound too familiar, I’m adding desu to maintain my distance, but it does not necessarily mean that I regard you as my superior or a big shot (erai hito).” The nai desu form should then be less likely to occur when such an attitude would be scorned, and speakers must display their utmost politeness to the addressee, projecting themselves as educated and well-versed in traditional keigo usage. This is an empirical question. I posit two major motivations for using semi-polite style. One is that a speaker adds desu at the end of a plain style sentence in order to sound politer, as if an afterthought addition. This is analogous to adding please at the end of a sentence in English;

19. Da, the plain counterpart of desu, can be added for a similar effect. For example, there was a popular song called Ora wa Shinjimatta-da ‘I’ve died+da’ in the early 1970s in Japan. The da makes it sound as though the songwriter is asserting or evaluating the proposition that he has died.

 Mutsuko Endo Hudson

e.g. Come here, please.20 The other is that, in the case of verbal negatives, speakers may say verb-nai desu simply because these are the only or the main ‘polite’ options they have in their active use repertoire, with the masen form relegated to passive knowledge. See Hudson (forthcoming) for further details on the motivations and also for grammatical constraints on nai desu.

5.2

Effects of social variables on semi-polite style

Let us consider the possible effects of social variables, such as gender, age, sociocultural background, and the communication medium, on the use of nai desu and other semipolite style forms. Obviously, the information about fictitious characters that can be gleaned in a novel and its utility is limited. Although a reliable theory must await research based on naturally occurring data, Riyuu data do raise some interesting issues. First, gender does not seem to be a factor. Despite Kindaichi’s (1942) characterization that the nai desu form is “inappropriate (tekitoo de nai) for women,” no gender differences were found in Riyuu.21 The total number of tokens is close for males and females, 156 vs. 152, and so is the rate of masen to nai desu, 66% to 34% for males vs. 71.1% to 28.9% for females, as shown in Table 13. The second variable to be considered is age. Although the nai desu form is used by speakers of all ages, it is especially popular among young people, according to the 1997 and 1999 public poll results, discussed in section 3.2.22 Also, Martin (1975: 1029) had remarked in 1975 that the verb+desu form seemed “to be gaining popularity among younger Tokyo speakers.” To examine the age factor from a different angle, I compared the occurrences of nai desu forms in two murder mysteries by major authors representing older generations than that of Miyabe Miyuki (who was born in 1960 in Tokyo) and written earlier than Riyuu (1998).23 One is Aoi Byooten ‘Blue Point’ (1960, 380 pages) by the late Matsumoto Seicho (1909-1992), a male writer who was born in 1909 in Fukuoka. The other is Joohatsu ‘Evaporation’ (1977, 421 pages) by Natsuki Shizuko, a female writer born in 1938. As expected, Aoi Byooten and Joohatsu include far fewer desu/masu-style exchanges than does Riyuu. Riyuu is unique in consisting mostly of interviews between strangers. Be that as it may, the difference was striking; there was no token of nai desu form in either of the older stories. This further supports the idea that its use is indeed a contemporary phenomenon.

20. The two kinds of motivations for nai desu usage find parallels in postposed sentences, discussed in Simon (1989). 21. Regrettably, the findings regarding gender differences in Fukushima and Uehara (2001) are tainted by the outlier problem. 22. For Riyuu examples by different age groups, see Section 4.1. 23. Riyuu first appeared in an Asahi Shimbun Newspaper series in 1996 to 1997.

Riyuu ‘reason’ for nai desu and other semi-polite forms 



Table 13.  Tokens by speakers’ gender in Riyuu

Male Female Total

Masen

Nai desu

103 (66.0%) 108 (71.1%) 211 (68.5%)

53 (34.0%) 44 (28.9%) 97 (31.5%)

Total 156 152 308

Excerpts (14) and (15) illustrate possible contexts for nai desu. In (14), taken from Matsumoto’s Aoi Byooten, the heroine Noriko, a college graduate in her early twenties who works for a publisher, is talking to strangers, first to a middle-aged woman, then to a landlord. In (15), from Natsuki’s Joohatsu, the main character Goroo, a male reporter who is 33 years old, is talking to Hisako, a 29-year-old flight attendant, with whom he is not well acquainted.24 (14) “Akita no kyoori no kata ja nai deshoo ka?” to (Noriko ga) kiitemo, “Shirimasen ne.” to (chuunen josei wa) kubi o kashigeta. “Ano, Okusan wa ichido, koko ni, kaerareta no deshoo ka?” to Noriko ga tou to, “Yoku wakarimasen. Nan deshitara, Yanushi-san ni otazune ni nattara?” to senaka o yusuri nagara itta. Sono yanushi to iu nomo, “Iya, shirimasen ne. … Ikisaki no koto wa, nan nimo iwanai kara wakarimasen naa.”to buaiso ni kotaeta. (Matsumoto Seichoo, Aoi Byooten 1960:202) 「秋田の郷里の方じゃないでしょうか？」と（典子が）き いても、 「知りませんね。」と（中年女性は）首を傾 げた。「あの、奥さん は一度、ここに、帰られたので しょうか？」典子が問うと、「よく 分かりません。何 でしたら家主さんにおたずねになったら？」と（ 中年 女性は）背中を揺すりながら言った。その家主という のも、「 いや、知りませんね。… 行先のことは、何に も言わないから分か りませんなあ。」と無愛想に答え た。 (松本清張『蒼い描点』1960) ‘(Noriko) asked, “Isn’t she from Akita?,” but (the middle-aged woman) said, “I don’t know,” with her head tilted to one side. When Noriko asked, “Um, did the wife happen to come by here?,” she responded, “I’m not sure. Why don’t you ask the landlord, if you’d like,” rocking the baby on her back. The landlord, too, would only give a curt reply, “Gee, I don’t know.... He didn’t say anything about where he was going, so I have no idea.”’ (15) “Sono anata to moo hitori kauntingu o okonatta suchuwaadesu to iu no wa, Kikuhata Toshie-san dewa arimasen ka.” (Goroo) … “Iie, ano toki wa, watashi to Shigematsu-san to iu wakai kata ga kauntingu itashimashita. …” (Hisako) 24. In data sets (14) and (15), the speaker’s name is added in parentheses, when unclear. The translations are mine.

 Mutsuko Endo Hudson



“Machigai arimasen ka?” (Goroo) “Hai, …” (Hisako) [22 lines skipped] “… anata-gata wa ki ga tsukanai no dewa arimasen ka.” (Goroo) “Soo desu wa ne. Sore wa tabun ki ga tsukimasen deshoo.” (Hisako) (Natsuki Shizuko, Joohatsu 1977:353-355)

「そのあなたともう一人カウンティングを行ったスチュワ ーデスと いうのは、菊畑敏江さんではありませんか」 （悟郎）... 「いいえ、 あのときは、私と重松さんとい う若い方がカウンティングいたしま した。...」（久子） 「まちがいありませんか」（悟郎）「はい、... 」（久 子）[22行省略]　「... あなた方は気がつかないのでは ありま せんか」（悟郎）「そうですわね。それは多分 気がつきせんでしょ う。」（久子） (夏木静子『蒸発』1977) “‘Wasn’t the stewardess who did the counting with you Toshie Kikuhata?” (Goroo) … “No, at that time, I did the counting with a young woman named Shigematsu. …” (Hisako) “There’s no mistake about it?” (Goroo) “No,…” (Hisako) [22 lines skipped] “Isn’t it the case that you wouldn’t notice it?” (Goroo) “That’s true. We probably wouldn’t notice.” (Hisako)’ Recall that shiranai ‘not know,’ wakaranai ‘not know/understand,’ dewa nai ‘not be,’ and machigai nai ‘there is no mistake about it’ are among the phrases that most commonly appear in the nai desu form in Riyuu, as discussed in Section 4.2. Had exchanges (14) and (15) been written by Miyabe or other contemporary writers, there would almost certainly be at least one instance of nai desu in each. Third, let us consider the effects of speakers’ sociocultural backgrounds, such as economic class and education level. The following characterization of ja nai desu ka ‘…, isn’t it?, you see?’ by Koyano (1995, cited in Jinnouchi 1998: 20) seems to apply to nai desu in general. [Ja nai desu ka] is a convenient phrase for youngsters whose level of sophistication is not yet high enough (shakaiteki kunren no mijuku na wakamono) to use without being overly strained when talking to someone with whom they must watch their language.

According to this, speakers would be less likely to use ja nai desu ka if their level of sophistication were high or if they aspired to project themselves to others as such. As mentioned in Section 5.1, desu expresses psychological distance and, by extension, indirectness and politeness as well. As a politeness device, the desu attachment sounds less “sophisticated” than the regular masu usage, however. This is nowhere more evident than when desu is attached to verbs in the affirmative and those already in polite style. See excerpts (16) and (17), both taken from Riyuu.



Riyuu ‘reason’ for nai desu and other semi-polite forms 

(16) “Soo desu ka. Maa, Yashiro Yuuji o omoitodomaraseru koto mo, aitsu o kaeru koto mo dekinakatta kedomo, watashi ni wa kikime atta desu nee.” (Ishida Naozumi: male, middle-aged) [p. 560] 「そうですか。まあ、八代祐司を思いとどまらせることも、 あいつ を変えることもできなかったけども、私には効 き目、あったですね え」 （石田直澄：男、壮年） '"Is that right. Well, it couldn't stop or change Yashiro Yuuji, but it had an effect on me.”’ (17) “Tantoo no hito wa ‘Oyajisan, warui nee’ tte, sumanagatte kuremashita desu.” (A: male, age 75)　[p. 234] 「担当の人は『親父さん、悪いねえ』って、済まながって くれまし たです。」 （Ａ：男、75歳） 'The person in charge was nice enough to apologize to me, saying, “I’m sorry, sir.”’ Compared to their respective regular polite counterparts, arimashita ‘had’ and kuremashita (or, especially, kudasaimashita) ‘was nice enough to,’ the forms in (16) and (17) sound less polite and less refined. The fourth and the last variable to be examined is the medium of communication. This factor does not constitute a binary opposition, but rather, a continuum. Although we have so far concentrated on nai desu style in spoken language (or in written depictions of spoken language), it may also be used in writing depending on to what extent the language is conversational. That is, the more formal and presentational the text is, the less likely it will be used.25 Consider excerpt (18), taken from Miyabe's postscript (atogaki) to Riyuu (Miyabe 1998: 572-573).26 (18) Hon-sakuhin wa fikushon de ari, toojoo suru jinbutsu, chimei, dantaimei nado wa, jitsuzai suru mono dewa arimasen. … Choohen shoosetsu wa, sakusha no chikara dake de seiritsu suru mono dewa arimasen. 本作品はフィクションであり、登場する人物、地名、団体 名等は、 実在するものではありません。...　長編小説は、 作者の力だけで成 立するものではありません。

25. For example, letters and e-mails written to one’s own family and close friends are informal and conversational, while scientific expositions, signs and other writings for the general public will be more presentational. 26. No postscript is included in the paperback version of Riyuu (2002).

 Mutsuko Endo Hudson

‘This is a work of fiction, and the names of the characters, places, organizations, etc. are not real. … A long story is not possible by the author’s efforts alone.’ The 2-page postscript mixes polite and plain style sentences quite freely depending on the author’s relationship to each person addressed, but, interestingly, does not include a single instance of semi-polite style. By way of experiment, let us change both instances of the polite negative copula (dewa) arimasen ‘is/are not’ in (18) to semi-polite nai desu. (19) Hon-sakuhin wa fikushon de ari, toojoo suru jinbutsu, chimei, dantaimei nado wa, jitsuzai suru mono {??dewa/ja nai desu }. … Choohen shoosetsu wa, sakusha no chikara dake de seiritsu suru mono {??dewa/ja nai desu }. 本作品はフィクションであり、登場する人物、地名、団体 名等は、 実在するもの　{??では／じゃないです。}　... 長編小説は、作者の力 だけで成立するもの {??では／ じゃないです。} Clearly, the nai desu sentences would sound overly informal and even “pushy,” as if to correct the reader’s wrong assumptions. It is evident then that the nai desu form has yet to be integrated into the presentational written language and that it is basically a spoken language phenomenon, at least currently.

5.3. Summary We discussed above the nature and major functions of semi-polite style, and also the possible effects of social variables on its use. Politeness levels cannot be discretely divided, but are a matter of degree on a continuum, and are changeable within an exchange and even within a speaker turn. One thing that is clear is that semi-polite style (e.g. yomenai desu ‘can’t read’) should be placed between the plain (e.g. yomenai) and the polite style (e.g. yomemasen).27 One function of desu in semi-polite style is to signal the speaker’s psychological distance toward the hearer at the time of the utterance. The distance is responsible for the utterance sounding more indirect and, as a consequence, politer, than does the plain style. The addition of desu does not necessarily express deference toward the addressee, however. It mainly helps the speaker not to sound overly familiar, thereby avoiding sounding rude. The other function is to impart the meaning that the speaker is explaining or evaluating a situation (e.g., yomenai desu ‘I can’t read it, and that’s how it is’). It sounds less ‘refined’ than does polite style (e.g. yomemasen), and it may even strike others as brusque. The nai desu form is typically used in conversational (i.e. nonpresentational) situations, especially by young men and women. 27. It is not clear where exactly polite form+desu (e.g. yomemasen desu) should be located on the politeness continuum. I leave this question to future research.

Riyuu ‘reason’ for nai desu and other semi-polite forms 



6. Conclusions and future directions There is no denying that the nai desu form is here to stay. It is a contemporary and growing phenomenon, used by people of all ages and both sexes to almost anyone to whom the speaker does not want to sound too familiar. The form sounds politer than does its plain counterpart nai, but not as polite as its polite counterpart masen, hence the name ‘semi-polite.’ In Riyuu data, masen and nai desu occurred at about a 2 to 1 ratio, favoring masen. The nai desu form most frequently occurred with verbs aru ‘exist, be,’ shiru ‘know’ and wakaru ‘know, understand’ and with some idiomatic expressions, such as machigai nai ‘there is no mistake about it.’ Although the present study used data from a novel for the investigation, it is hoped that some of the results will be tested against naturally occurring data, and also that it will provide a start on what to explore in the future regarding nai desu and other semi-polite forms. There are many tasks to be undertaken regarding nai desu. The most urgent among them is sociolinguistic research on the interaction between the semi-polite style usage and social variables. An examination of situational factors on the choice between nai desu and masen with speakers who use both forms will also be fruitful. Regarding discourse factors, Fukushima’s and Uehara’s (2001) findings seem worth confirming by additional data: (1) masen forms are preferred when starting a conversation and/or a new topic; and (2) they are preferred at the point when the speaker concludes his/her turn, as opposed to nai desu forms, which are preferred when the speaker gives background information and is continuing. Testing Tanomura’s (1994) results against natural data may also offer new insights. A comparative analysis of desu and no desu ‘it is that…’ from synchronic (syntactic, pragmatic, textual) and diachronic perspectives should also be interesting.

References Bunkacho Bunkabu Kokugoka. 1997. Yoron Choosa Hookokusho, Kokugo ni kansuru Yoron Choosa [Report on the Public Poll regarding the Japanese Language]. Japanese Government. Bunkacho Bunkabu Kokugoka. 1999. Yoron Choosa Hookokusho, Kokugo ni kansuru Yoron Choosa [Report on the Public Poll regarding the Japanese Language]. Japanese Government. Cook, H. M. 1990. “The sentence-final particle ne as a tool for cooperation in Japanese conversation.” In Japanese/Korean Linguistics 1, H. Hoji (ed.), 29-44. Stanford: Csli. Cook, H. M. 1999. “Situational meanings of Japanese social deixis: The mixed use of the masu and plain forms.” Journal of Linguistic Anthropology 8 (1): 87-110. Cook, H. M. 2002. “Construction of speech styles in a Japanese elementary school classroom.” A paper presented at Japanese Speech Style Shift Symposium, University of Arizona. Endo, O. 1994. “Wakai josei no kotoba – Ronpyou de tuzuru sono Showa-shi.” Nihongogaku 13: 19-32. Fukushima, E. and Uehara, S. 2001. “Gendai Nihongo ni okeru teineitai hitei keishiki.” Toohoku Daigaku Ryuugakusei Sentaa Kiyoo 5: 11-17.

 Mutsuko Endo Hudson Hudson, M. E. 1999. “Teinei hyooshiki to shite no kenjoo dooshi.” In Gengogaku to Nihongo Kyooiku 1, Y. Alam (ed.), 259-274. Tokyo: Kuroshio. Hudson, M. E. Forthcoming. “On verb+nai desu.” Hudson, M. E. and Lu, W.-Y. 2003. “Japanese ne and Chinese ba.” In Meaning through Language Contrast 2 [Pragmatics and Beyond New Series 99], K. M. Jaszczolt and K. Turner (eds.), 197-212. Amsterdam/Philadelphia: John Benjamins. Ikuta, S. 1983. “Speech level shift and conversational strategy in Japanese discourse.” Language Sciences 5: 37-53. Ikuta, S. 2002. “Speech style shift as an interactional discourse strategy: A study of the shift between the use and non-use of -des/-mas in Japanese conversational interviews.” A paper presented at Japanese Speech Style Shift Symposium, University of Arizona. Inoue, F. 1995. “Teinei hyoogen no genzai – Desu/masu no yukue [The current state of polite expressions–The future of desu/masu].” Kokugogaku 40 (14). Inoue, F. 1998. Nihongo Wocchingu [Japanese Language Watching]. Tokyo: Iwanami Shoten. Inoue, F. and Yarimizu, K. (eds.). 2002. Jiten Atarashii Nihongo. Tokyo: Toyo Shorin. Jinnnouchi, M. 1998. Nihongo no ima [The current state of Japanese]. Tokyo: Alc Press. Jorden, E. with Noda, M. 1987. Japanese: The spoken language. New Haven: Yale University Press. Kamio, A. 1990. Joohoo no nawabari riron [The Theory of Territory of Information]. Tokyo: Taishukan. Kamio, A. 1994. “The theory of territory of information: The case of Japanese.” Journal of Pragmatics 21, 67-100. Kamio, A. 1998. “An analysis of Japanese ne in terms of the theory of territory of information.” In Japanese/Korean Linguistics 7, N. Akatsuka, et al. (eds.), 231-242. Kindaichi, K. 1942. “Joseigo to keigo [Women’s language and keigo].” Kokugo Kenkyuu. Tokyo: Yakumo Shorin. Kokugo Shingikai [National Japanese Language Council]. 1993. Gendai Kokugo no Shomondai nitsuite [Issues in Modern Japanese]. Japanese Government. Koyano, T. 1995. “Terebi to wakamono kotoba [Tv and youth language].” Nihongogaku 15: 10. Lee-Wong, S. M. 2001. “Coherence, focus and structure: The role of discourse particle ne.” Pragmatics 11, 139-153. Makino, S. 2002. “When does communication turn mentally inward?: A case study of Japanese formal-to-informal switching.” In Japanese/ Korean Linguistics 10, N. Akatsuka and S. Strauss (eds.), 121-135. Stanford: Csli. Martin, S. A. 1975. Reference Grammar of Japanese. New Haven: Yale University Press. Matsumoto, S. 1960. Aoi Byooten [Blue Point]. Tokyo: Kobunsha. Maynard, S. K. 1989. Japanese Conversation: Self-contextualization through Structure and Interactional Management. Norwood: Ablex. Miyabe, M. 1998. Riyuu [Reason]. Tokyo: Asahi Shinbunsha. Miyabe, M. 2002. Riyuu [Reason]. Tokyo: Asahi Shinbunsha. Natsuki, S. 1977. Joohatsu [Evaporation]. Tokyo: Kadokawa Shoten. Ohta, A. S. 1994. “Socializing the expression of affect: An overview of affective particle use in the Japanese as a foreign language classroom.: Issues in Applied Linguistics 5 (2): 303-325. Simon, M. E. 1989. An analysis of postposing construction in Japanese. Ph.D. dissertation, University of Michigan.



Riyuu ‘reason’ for nai desu and other semi-polite forms  Takubo, Y. and Kinsui, S. 1996. “Fukusuu no shinteki ryooiki niyoru danwa kanri [Discourse management in terms of multiple mental domains].” Ninchi Kagaku [Cognitive Science] 3 (3): 59-74. Tanaka, H. 2000. “The particle ne as a turn-management device in Japanese conversation.” Journal of Pragmatics 32 (8), 1135-1176. Tanomura, T. 1994. “Teineitai no jutsugo hiteikei no sentaku ni kansuru keiryooteki choosa–’ ‘-masen’ to ‘-nai desu’ [A corpus-based study of the polite negative predicate forms in Japanese]. Osaka Gaikokugo Daigaku Ronshuu 11: 51-66. Teramura, H. 1984. Nihongo no Shintakusu to Imi 2. Tokyo: Kuroshio. Tokieda, M. 1950. Nihon Bunpoo Koogohen. Tokyo: Iwanami Zensho. Tsujimura, N. 1996. An Introduction to Japanese Linguistics. Cambridge: Balckwell Publishers. Tsujimura, T. 1964. “Keigo no keikoo” [Tendencies of keigo]. Kooza Gendaigo Dai 6-kan Koogo Bunpoo no Mondaiten, 248-263. Tokyo: Meiji Shoten. Tsujimura, T. 1967. Gendai no Keigo [Modern Keigo]. Tokyo: Kyoobunsha. Tsujimura, T. 1992. Keigo Ronkoo [On Keigo]. Tokyo: Meiji Shoten. Uyeno, T. 1971. A Study of Japanese Modality: A Perfomative Analysis of Sentence Particles. Ph.D. Dissertation. University of Michigan.

Masen or Nai Desu – That is the question A case study into Japanese conversational discourse* Satoshi Uehara and Etsuko Fukushima Tohoku University

1. Introduction Traditionally, Japanese possesses two distinct sets of predicate forms, one for the polite level of speech and the other for the plain level. Thus, Jorden and Noda (1987), in their textbook, Japanese: The Spoken Language, list the following inflectional paradigms for Verbs, Adjectives and Copulas (Nouns) in Table 1. It should be noted that Adjectives and Copulas (but not Verbs) have two structural variants for their polite style negative forms. For differences between the two polite-style negative variants, however, Jorden and Noda only note, in referring to those forms in the Adjective paradigm, that A-ku arimasen forms, in comparison to A-ku nai desu, are “a bit more formal, and in the view of some speakers, are slightly more elegant” (1987: 41). Two new forms seem to have recently joined, or be emerging in, the Verbal paradigm above, namely, V-nai desu as another polite negative nonpast form, and V-nakatta desu as another polite negative past form, thus yielding for the Verbs the following paradigm in Table 2, analogous to those for Adjectives and Copulas in Table 1.

* Earlier Japanese versions of this paper were presented at the 7th Annual Meeting of the Association of Natural Language Processing of Japan at University of Tokyo in March 2001, in Bulletin of Tohoku University International Student Center, No. 5 (Fukushima and Uehara 2001), and at the 3rd International Conference of Practical Linguistics of Japanese at San Francisco State University in March 2002 (as Fukushima and Uehara 2003 in the conference proceedings). Our sincere thanks go to those who gave us their insightful and critical comments on the earlier versions. We are especially indebted to Mutsuko Endo Hudson, Masahiko Minami, Tsuyoshi Ono, Kimberly Jones, and an anonymous reviewer for this volume. Our gratitude also goes to Bob Alpert and Bob Sanders for textual improvements. Remaining errors are, of course, our own.

 Satoshi Uehara and Etsuko Fukushima

Table 1.  Basic inflectional paradigms in Japanese nonpast affirmative negative VERB Plain Polite ADJECTIVE Plain Polite

COPULA Plain Polite

past affirmative

negative

V-ru V-masu

V-nai V-masen

V-ta V-mashita

V-nakatta V-masen deshita

A-i A-i desu

A-ku nai A-ku nai desu A-ku arimasen

A-katta A-katta desu

A-ku nakatta A-ku nakatta desu A-ku arimasen deshita

da desu

ja nai ja nai desu ja arimasen

datta deshita

ja nakatta ja nakatta desu ja arimasen deshita

Notice here that, for all word classes now, the two variant forms in the polite negative cells in Tables 1 and 2 share the same morphemes: one of the variant pair has nai desu for nonpast and nakatta desu for past (collectively called “nai desu” forms, hereafter), while the other has masen for nonpast and masen deshita for past (collectively, “masen” forms). This set of polite-style negative variants is the topic of this paper. Examining naturally occurring nai desu and masen forms in Japanese conversational discourse, the paper addresses the following issues: 1) to what extent the nai desu forms and the masen forms are used, especially for Verbal polite negative forms, and 2) what discourse contexts, if any, favor the use of one over the other. We have organized the paper in the following manner. Section 2 examines some previous approaches to these two variant, polite-style negative forms. Section 3 introduces the approach adopted in our research and explains the nature of the data used in it. Section 4 presents the formal organization of the data, and Section 5 examines the results, presents our analysis, and discusses its theoretical implications. Section 6 concludes the paper. Table 2.  New inflectional paradigms for Verbs in Japanese nonpast VERB Plain Polite

affirmative V-ru V-masu

negative V-nai V-nai desu V-masen

past affirmative V-ta V-mashita

negative V-nakatta V-nakatta desu V-masen deshita

Masen or nai desu: That is the question 



2. Previous approaches to the problem and the current study In the previous chapter, Hudson presents an excellent summary of previous approaches to nai desu forms and what she calls more generally “semi-polite” forms,1 so we recommend readers read her chapter and the summary in it. In this chapter, our focus will be on the two polite-style negative forms. In this section we examine previous analyses of the two negative forms to show how our approach departs from them.

2.1

Grammar descriptions

Reference grammars of Japanese we surveyed list only one of the two formal variants of polite negative predicate forms, or at best list the two forms without discussing any usage difference between them. Thus, Tokieda (1950: 212) lists masen forms for Verbs, Adjectives and Copulas in the “polite negative” cell of his table of polite auxiliaries and notes in his supplementary explanations to the table that there also exist nai desu forms for them (215). In a similar manner, Teramura (1984: 53), in a note to his discussion of the inflectional paradigm of masu, adds that for Verbs, masen forms are more common than nai desu forms, which are seldom used, but that for the past tense forms of Verbs both masen deshita and nakatta desu appear to be used, although the latter is considered unnatural by some.

2.2

Japanese language textbooks

In addition to Jorden and Noda (1987) cited above, we examined the treatment of the two polite negative variant forms in the following five popular Japanese language textbooks used at college-level institutions in Japan and the United States: a. An Introduction to Modern Japanese. 1977. Mizutani, Osamu and Nobuko Mizutani. b. Nihongo Shoho. 1981. The Japan Foundation. c. A Course in Modern Japanese, Vol. I. 1983. Staff of the Japanese Section, Center for Linguistic and Cultural Research, The University of Nagoya. d. Shin Nihongo no Kiso. 1990. The Association for Overseas Technical Scholarship. e. Minna no Nihongo. 1998. 3A Corporation. This examination was intended to find how these two variant forms are believed to differ and how they are taught. The results are perplexing. As we noted above, Jorden and Noda (1987) list both masen and nai desu forms for Adjectives and Copulas, but list only the masen form for Verbs. The five textbooks examined, however, list no 1. Hudson’s “semi-polite” forms include polite-style affirmative forms as well, such as V-ru desu (e.g. taberu desu ‘eat’), which are not necessarily readily accepted as correct by most native speakers.

 Satoshi Uehara and Etsuko Fukushima

variant forms at all. Thus, for Copulas as well as for Verbs they list masen forms only, and for Adjectives, two (a and c) list masen forms only while the other three (b, d, and e) list nai desu forms only. The result is summarized in Table 3. Table 3.  Polite-style negative forms listed in the textbooks Textbooks

Verbs

Copulas

Adjectives

a b c d e

masen masen masen masen masen

masen masen masen masen masen

masen nai desu masen nai desu nai desu

2.3

Public opinion poll

In 1999 the Japanese Language Division of the Japanese government’s Agency for Cultural Affairs (Bunkachoo Bunkabu Kokugoka) conducted a public opinion poll regarding many aspects of the Japanese language. The questionnaire included the following four pairs of expressions. 2200 people were asked to indicate which of the pair they used, or whether they used both. a. Isshoni [ikimasen ka / ikanai desu ka]. ‘Won’t you come along?’ b. Kyoo wa tokuni yotee ga [arimasen / nai desu]. ‘I have no particular plans for today.’ c. [Mikakemasen deshita ka / Mikakenakatta desu ka]. ‘Didn’t you see (him)?’ d. (Kono hon,) muzukashii n ja [arimasen ka / nai desu ka]. ‘Isn’t it that (this book) is difficult?’ It is extremely difficult, if not impossible, to find any reasonable patterns in the use of the two variant forms in the survey results. These pairs differ drastically in their speech act functions (a=invitation; b=assertion; c and d=question). According to the results, however, masen forms were strongly preferred for items a, b, and c with only 10.8% to 32.8% of the respondents reporting that they would use nai desu forms instead of or in addition to masen forms. On the other hand, for item d, 70.6% of the respondents reported that they would use the nai desu form over the masen form.

Masen or nai desu: That is the question 



2.4

Approach in the current analysis

As we just noted, previous approaches to the two variants of polite negative predicate forms are all intuitive and/or conjectural in nature. That is, they are based on native speakers’ out-of-context grammaticality judgments of the two forms in isolation or at best in a sentence.2 In other words, these approaches are not based on natural speech discourse data. In dealing with language use, it is crucial to examine data with a rich context surrounding the linguistic forms in question–especially so when we are dealing with the use of markers of social deixis or indexicality. Aware of this necessity, we took up the current analysis. Our analysis is, therefore, based on data from natural spoken discourse and examines syntactic and discourse contexts in which the two forms are used.3

3. Data As we have observed, empirical research that investigates ways in which these two negative forms are used in the natural flow of conversational discourse is almost nonexistent. It is important to examine in detail the discourse contexts in which the two forms are used (i.e., exactly what contextual factors distinguish their uses) in actual interactions.4 Thus we took extreme care to collect natural conversational data. We asked participants to have one-to-one discussions with persons with whom they were unacquainted, having made sure that neither knew the purpose of our work or the focus of data collection. We circulated a flyer around the university community and asked for volunteers to join a session to discuss some topics concerning cross-cultural communication after viewing some video clips on them. Out of those volunteers who agreed to cooperate 2. The two polite-style negative forms are used not only in natural interactions but also in some written texts, such as newspaper reports and novels. Just before we finished writing this paper, Hudson brought to our attention two written discourse studies on closely related topics. One is Tanomura (1994) on the two polite negative predicate forms in newspaper reports, and the other is Hudson, who describes in this volume what she calls “semi-polite” forms (the two polite negative predicate forms and, as she argues, polite affirmative predicate forms as well) in a novel. Readers of this paper are encouraged to examine these two papers as well. Although their results differ from those of our research, one must note that their research covered primarily written Japanese, which tends to be more formal. 3. To the best of our knowledge, two spoken discourse studies, Noda 2004 and Kobayashi 2005, both on the two polite-style negative forms in Japanese, were published after we finished writing this paper. Their findings basically support and/or are compatible with our findings in this paper and its earlier versions. Readers are encouraged to read them as well. 4. One can easily see it would be extremely unusual, for example, to ask people to “speak politely using negative forms” to get the kind of data needed for the current research.

 Satoshi Uehara and Etsuko Fukushima

with our project, we paired up people who had never seen or spoken to each other. One of each pair was shown how to operate the video player beforehand, and was asked to assume the role of what we call the “interviewer” during the session, starting and stopping a maximum of three video clips and initiating a discussion after each video clip viewing. Having received the informants’ permission to make a recording of their discussion ahead of time, we would set up a session, introduce the two participants to each other, and leave them by themselves in the room. In this way we avoided the awkward situation where informants feel aware of being observed by bystanders. Each pair was given 15–30 minutes, after which the investigators returned to the room. If the conversation ended before the given time had elapsed, the interviewer phoned the investigators and told them that the conversation was over. Only recordings made when the two participants alone were present constitute our data, and thus, for instance, the conversations recorded before the investigators left the scene were excluded. The three video clips were all from a TV program where Japanese and non-Japanese discuss their opinions about some aspects of Japanese culture and behaviors of Japanese people. The topics used for the recording sessions were: 1) everyday greetings (e.g. Ii tenki desu ne ‘It is a nice/sunny day, isn’t it!’), 2) parent-child relationships (e.g. whether or not children are spoiled in Japan), and 3) husbands and wives having different surnames. Recording sessions for the current research were conducted at two different times. The first sessions were conducted in January and February of 1999. All 14 informants were undergraduate or graduate students of Tohoku University (9 in their teens and 5 in their twenties). Two out of the 14 served as interviewers and were paired with the other 12, so a total of 12 recording sessions were conducted with 12 pairs of people who were meeting for the first time. The total conversation time for this set of data, referred to as Data 1 in this paper, amounts to about 100 minutes. All the informants were originally from, and/or had spent most of their lives in, some part of either the Tohoku or Hokkaido region, but as is usually the case with people meeting for the first time, they spoke in standard Japanese. The second set of recording sessions was conducted in July and August of 2001 to collect more data. This time a total of 10 recording sessions were conducted with 10 pairs of people who were teachers, undergraduates, and graduate students at Tohoku University (none from the first sessions) and who were meeting for the first time. However, for reasons which will be made clear later, only the three recordings with instances of both of the polite negative forms of verbs in one or both informants’ discourse were selected for the current research. These three recordings, #13, #18, and #22, together total approximately 75 minutes and are together referred to as Data II in this paper. In these recordings, all three interviewers and the interviewee in #13 were female, in their thirties, and originally from Tokyo. The interviewee in #18 was a woman in her twenties, and the interviewee in #22 was a male teenager. Both were originally from the Tohoku region.

Masen or nai desu: That is the question 



4. Analysis of Data I It is difficult to find many instances of negative predicate forms in uncontrolled, natural conversational discourse. We reviewed the transcriptions in Data I and found 124 instances of the polite-style negative forms in question. All 124 instances found were in the present tense, that is, no past tense negative forms are attested in the data.5 Both variants occurred; we found 32 masen forms and 92 nai desu forms.6 Out of 124 instances of polite-style negative forms, 44 were Verbs, which were divided into 20 masen forms and 24 nai desu forms.7 The following section presents more detailed results of the data analysis.

4.1

Results

Table 4 below shows the number of the two negative forms used by each speaker. Table 4.  Use of the two negative forms according to speaker masen

Speaker IM01 M04 M06 M07 M08 M10 M11 M12



8 (6) 0 (0) 3 (1) 6 (3) 1 (1) 1 (1) 1 (1) 4 (2)

nai desu

40 4 1 2 9 2 1 9

(7) (1) (0) (0) (4) (1) (1) (1)

Speaker IF01 F01 F02 F03 F05 F09 Total



masen

nai desu

3 2 0 0 3 0

11 2 2 5 2 2

(0) (2) (0) (0) (3) (0)

32 (20)

(3) (0) (0) (4) (1) (1)

92 (24)

I = interviewer, M = male, F = female, ( ) = verb forms only. Numbers following interviewee M or F correspond to their recording numbers. 5. This is rather natural if we consider the topics of conversations in the data. To get data on past tense negative forms in future research, one could, for example, set up a situation in which participants were supposed to describe past experiences to each other (i.e., narrate past events). We are thankful to Kim Jones for pointing this out for us. 6. The result here contrasts intriguingly with those of written text analyses by Tanomura (1994) and Hudson (this volume), both of which show that, unlike our current analysis, the use of masen forms predominates over the use of nai desu forms both overall and for Verbs alone. 7. The inanimate existential verb, aru ‘exist’, is excluded from the count of Japanese Verbs here because its negative forms are formally irregular and distinct from those of Verbs in general (see section 4.1 for details and its separate count). We have also excluded the three verbal aspect marker forms V-te (i)masen and V-te (i)nai desu (1 and 2 tokens, respectively), from the count of Verbs, although they were included in the count in earlier versions of this paper. This applies to all the tables and discussions in this paper. We believe this exclusion facilitates a more transparent comparison of our analysis of spoken discourse to Hudson’s analysis of written discourse.

 Satoshi Uehara and Etsuko Fukushima

Table 4 clearly demonstrates that there was some individual variation in the usage of the two negative forms in question, but at the same time it shows certain overall tendencies as well. One tendency observable in Table 4 is that, as the total numbers of usage tokens show, overall, nai desu forms were more frequently used than masen forms. Nine out of 14 informants used the former more frequently than the latter, two used the two forms evenly, and only three preferred the latter over the former. We can confirm this tendency by observing that while some (four people) did not use the masen form at all, every informant used the nai desu form at least once in their discourse. If we limit our scope to Verbs (see parenthetical results in Table 4) alone however, this tendency is weakened; all things considered, the masen form is preferred for Verbs more than for predicates in general. Thus, three people used only the masen form (i.e. no nai desu form) for Verbs, whereas no one did so for predicates in general. The difference in behavior between Verbs and other predicates leads us to consider all 124 instances of polite negative forms with respect to their word class membership. Table 5 shows this. Table 5.  Use of the two negative forms according to word class Verbs V-masen V-nai desu Inanimate Existential Verb** arimasen nai desu Adjectives A-ku arimasen A-ku nai desu Copula (Nominal predicates) N de wa/ja arimasen N de wa/ja nai desu Idiomatic expressions ‘Pardon me/I am sorry.’ sumimasen/suimasen sumanai desu ‘It cannot be helped/No use’ shoo ga arimasen shoo ga nai desu

Verbal (aspectual) auxiliaries 20 24 3 21 0 1 0 5

4 0 0 1

‘be V-ing/have V-ed’ V-te (i)masen* V-te (i)nai desu Modality expressions ‘may/might’ kamoshiremasen kamoshirenai desu ‘Isn’t it the case that...?’ (n/wake) ja arimasen ka (n/wake) ja nai desu ka ‘must/have to’ nai to ikemasen nai to ikenai desu TOTAL masen nai desu

1 2

4 21 0 16 0 1 124 32 92

* The /i/ sound which is in parentheses indicates that it is frequently dropped in conversation, and it was indeed missing in all three instances in the data. ** This type also includes the composite expression with the Inanimate Existential Verbs, … koto (ga/wa/mo) + arimasen/nai desu, ‘(I) have no experience/occasion’ (lit., ‘Experiences/Occasions do not exist (with me)’).

Masen or nai desu: That is the question 



Some additional explanation of language-specific aspects of Japanese shown in Table 5 above is in order. First, the Inanimate Existential Verb (Existential Verb with inanimate subjects) is separated from other Japanese Verbs in the table, because the Inanimate Existential Verb has a distinctive negative inflectional pattern and its two negative forms do not share the same lexical root (arimasen vs. nai desu), whereas other Verbs always do (V-masen vs. V-nai desu). Second, idiomatic expressions listed here derive originally from the negative form of a Verb (sumimasen/sumanai desu ‘Pardon me.’) or a phrasal Verb (shoo ga arimasen/shoo ga nai desu ‘No way to avoid it.’). They are lexicalized to function almost as one word, but they are listed here because their original status as a Verb gives these lexicalized expressions the two variant forms as well. Finally, verbal auxiliaries and modality expressions listed in Table 5 also originate in phrasal verbs/verbal compounds, and that is why the same two variant forms are available for them. In that sense, they are similar to idiomatic expressions above, but while the latter are functioning as lexical elements, the former are functioning as grammatical elements. They are attached to preceding predicate forms and express those forms’ grammatical functions. These grammaticalized forms of Verbs are listed separately from Verbs because the former exhibit grammatical behavior different from the latter, such as contraction of verbal auxiliary /i/ in Table 5.

4.2

Discussion of Data I

As we have noted in the previous section, the two variant forms for Verbs appear with almost the same degree of frequency. However, for all other categories, either the nai desu form or the masen form dominates, i.e., one form is used exclusively or one form is used far more frequently than the other. What factors might lead to such a skewed usage? Let us examine those categories with a statistically significant number of occurrences in a discussion of the results of Data I shown in Table 5, and ignore forms such as Adjectives, which occurred rarely. Consider those categories for which only one of the two variant forms is used exclusively. There are two types, one for which only the masen form is used and the other for which only the nai desu form is used. The former is represented solely by the idiomatic expression sumimasen ‘Pardon me/Excuse me’ and the latter by the modality expression (n/wake) ja nai desu ka ‘isn’t it (the case) that...?’). Sumimasen is the only expression for which the masen form is used exclusively in the whole list of word items.8 What is it about this expression that so differentiates it from the rest of the expressions? Here is an example from the data:

8. This expression sumimasen ‘pardon me’ occurs four times–not many, to be sure, but taking into account its lexical (non-grammatical, i.e., non-recurring) status, we consider the four tokens (all in the masen form) to be sufficiently significant.

 Satoshi Uehara and Etsuko Fukushima

(1) (IF brings a new topic into the conversation.) IF: Konaida no nanka, dokoda keishichoo no nanka, enjokoosai no bideo no sakusei no koto ni, nanka mondai, saikin nattemasu yo ne. ‘Adolescent prostitution is a big problem, these days, isn’t it?... uh...a while ago.. -who was it?- The Police Agency or someone, was making a promotional film against it.. you know.’ M: Shirimasen, sumimasen. ‘I don’t know... Sorry.’ An inquiry into the discourse contexts where the expression was used helps us account for the unique behavior of this expression. Japanese speakers understand almost reflexively that the nai desu version of the expression, sumanai desu does not sound polite enough in contexts which commonly call for sumimasen in our data. Thus, to use sumanai desu in place of sumimasen in (1) above makes M’s utterance inappropriate. Sumanai desu in the above conversation would sound as if the speaker M did not really mean what he is saying, i.e. he is not sorry, just bored. In that context his response appears flippant, almost rude – as if M did not care much about the topic that his interlocutor had been discussing. In other contexts as well, sumanai desu would be inappropriate. Thus, we conclude that sumimasen, because of its semantic and semiotic functions, is used in very polite contexts where the speaker needs to apologize to the interlocutor for whatever blunder she has made,9 and, because sumanai desu is not as polite as sumimasen, it is not available for use. Of course, we have not explained why sumanai desu is less polite than sumimasen even though the two have the same semantic function. We will return to this question later. Now let us look at expressions for which the nai desu form is used exclusively. The modality marker (n/wake) ja nai desu ka ‘Isn’t it (the case) that...?’ represents such expressions. This modality marker was used 16 times in total and the nai desu form was used in every case. Consider the following instance in (2):10

(2) (M is talking about his experience of being surrounded by foreigners and spoken to in their language.)

M: Nani sareru n daroo tte{laugh}omot[te], sore de yappari hora, gaikokujin tte ‘I wondered what they were gonna do to me, and then...you know, you see... foreigners...’

9. The predominance of the sumimasen form suggests that it may be the default version in such contexts, whereas the use of sumanai desu may be a more marked activity, going beyond the more formulaic or semi-automatic employment of sumimasen (Wray 2002). We are thankful to an anonymous reviewer for pointing this out to us. 10. The following transcription conventions are used: ___ : polite-style negative form; { }: nonverbal action; [ ] : unsure hearings; // : the point where overlapping talk begins or latching occurs; ? : rising intonation.

Masen or nai desu: That is the question 



IF: Sono mama [tachisatta no{laugh}] ‘so you just left them without even...’ M: Honto ni shaberu kikai tte nai ja nai // desu kaa. ‘we hardly ever get a chance to talk with them, right?’ IF: //Soo, kamoshirenai desu nee. ‘Yeah, I suppose maybe you’re right.’ We need not render an account for this specific modality expression but say that it follows the general pattern (which will be discussed below) because as we saw above most other expressions we observed follow the same pattern and use the nai desu form exclusively or predominantly. In fact, the nai desu form above in (2) could be replaced with its masen version (ja arimasen ka) without rendering the sentence ungrammatical or strange (cf. Sumimasen ‘Pardon me.’ above). One possible explanation, however, of the most skewed occurrence (16 nai desu vs. 0 masen) of this modality expression is once again usage-based. Uehara and others (1998) call the ja nai part of the total expression (n/wake) ja nai desu ka “modality janai”, which they contrast to the “propositional/negative janai”. According to them, the former contrasts both structurally and functionally with the latter, and has lost its negative (and interrogative) sense in the process of developing into a grammatical (modality) marker of its own.11 If they are correct, they have developed a partial account of why, given the two composite structures, ja nai desu ka and ja arimasen ka, the former tends to be the default: in a discourse context where either composite expression is possible but where janai, a component of ja nai desu ka, is already lexicalized as the grammatical marker specific to that discourse function, that form is likely to be activated by default. In that case the compositional process of the composite expression simply continues, with no need to repeat the whole process from scratch.12 Of course, in cases where some other discourse factor must be highlighted, the default compositional process is not selected. This seems to be exactly when the other form, ja arimasen ka, can become appropriate and selected, and why it becomes possible to use it today along with ja nai desu ka. We will return to this last point later. In certain curious cases both nai desu and masen forms are found in our data: the Inanimate Existential Verb (3 arimasen vs. 21 nai desu), ‘maybe’ modality markers (4 kamoshiremasen vs. 21 kamoshirenai desu), and Verbs (20 V-masen vs. 24 V-nai desu). Verbs are particularly intriguing because their two forms are used with almost equal frequency and because even though some explanations similar to ja nai desu ka in 11. It is interesting in this regard that the “modality marker janai” (‘you know?’) has further developed into jan in the colloquial language of (younger generation) speakers of Tokyo Japanese as well as some other dialects. Note that jan has been specifically grammaticalized and occurs only in the plain style (*… jan desu ka.). 12. In other words, janai can be said to have attained some degree of “unit status” (Langacker 1987) or become a “prefab” (Erman and Warren 2000).

 Satoshi Uehara and Etsuko Fukushima

terms of lexicalization or gramaticalization patterns can apply to the nai desu forms of the Inanimate Existential Verb and ‘maybe’ modality markers, no such explanation seems to apply to Verbs. In other words, the two variant forms for Verbs do not seem to differ in terms of lexicalization or grammaticalization patterns, so if the use of the two forms is differentiated, it must be because of something else, i.e., some discourse contextual factors. To find such contextual factors, we must examine the contexts where these two forms are used for Verbs as well as other predicate forms. Another totally different, but logically possible, conclusion from the data is that the variation in the use of these two forms of the three types of predicates above may simply be due to individual preference. That is, although both of the two variant forms show up in the data, it is possible that each participant uses only one of the two forms in a given conversation. We need to focus, therefore, on those cases where a speaker uses both forms in the course of his/her conversation with the same interlocutor. This use of a single interlocutor is critical. Otherwise one speaker might use only one form in speaking with someone and use only the other form in speaking with someone else,13 in which no discourse contextual factors could be identified as responsible for the kind of variation in use that we are interested in here. Do such cases of using both forms for a single interlocutor exist in our data? They do, as can be seen in Table 6, an expanded version of Table 4, which lists the uses of the two negative forms for Verbs for each conversation and each individual speaker. There are seven cases where a speaker uses both negative forms for Verbs in a conversation with the same interlocutor. They are in boldface in Table 6. In the next section we will carefully examine these seven cases from Table 6 and similar cases for other predicates from Data I together with those in Data II as well. Table 6.  Use of the two negative forms for Verbs for each dialogue and speaker Dialogue 1 2 3 4 5 6 7 8 9 10 11 12

Speaker

masen

nai desu

Speaker

masen

nai desu

IM01 IM01 IM01 IM01 IM01 IM01 IM01 IM01 IM01 IF01 IF01 IF01

1 0 0 0 1 0 2 1 1 0 0 0

2 1 1 1 0 0 0 0 2 0 3 0

F01 F02 F03 M04 F05 M06 M07 M08 F09 M10 M11 M12

2 0 0 0 3 1 3 1 0 1 1 2

0 0 4 1 1 0 0 4 1 1 1 1

13. Such cases, if attested, are interesting by themselves as well. We found a few cases in our data, but they did not exhibit any statistically significant pattern.

Masen or nai desu: That is the question 



Table 7.  Dialogues where both negative forms are used by one or both of the speakers Dialogue

Speaker

Verbs (V-masen vs. V-nai desu) 1 IM01 5 IM01 8 IM01 9 IM01 10 IF01 11 IF01 12 IF01 13 IF02 18 IF03 22 IF03

masen nai desu 1 1 1 1 0 0 0 0 1 2

2 0 0 2 0 3 0 6 3 2

Speaker

masen nai desu

F01 F05 M08 F09 M10 M11 M12 F13 F18 M22

2 3 1 0 1 1 2 1 0 1

0 1 4 1 1 1 1 6 0 9

M04 M11

0 0

2 0

“maybe” modality expressions (kamoshiremasen vs. kamoshirenai desu) 7 IM01 0 3 M07

3

1

Inanimate Existential Verbs (arimasen vs. nai desu) 4 IM01 1 1 11 IF01 1 2

5. Analysis of Data I + Data II Using the new, more restrictive criteria introduced above, we will now broaden our analysis to cover a new set of data (Data II). As noted in Section 3, this new set of data was collected in the same manner as Data I, but after examining all the dialogues, we found only three dialogues where a speaker used both negative forms for Verbs. The seven dialogues in Table 6 and the three dialogues (#13, #18, and #22) for Verbs from Data II, together with two for Inanimate Existential Verbs and one for “maybe” modality expressions, constitute our data in this section, which is summarized in Table 7 (the relevant cases are in boldface). A total of 14 cases are available for our examination in terms of discourse contextual factors, to which we now turn.

5.1

Analyses of contextual factors

5.1.1 Appearance orders When we examined the discourse contexts where both of the two forms were used in one dialogue by a single speaker, a key pattern emerged. To understand its significance, let us look at Table 8, where all the instances of polite negative forms in each of

 Satoshi Uehara and Etsuko Fukushima

Table 8.  Appearance orders of the two negative forms Dialogue

Speaker

Verbs (V-masen vs. V-nai desu) 1 IM01

Appearance order /Xx/x

5 8 9

F05 M08 IM01

XXx/X/ Xxx/xx/ x/xX/

10 11 12 13 18 22 22

M10 M11 M12 F13 IF03 IF03 M22

/Xx/ X/x/ /X/xX X/xxxxxx xX/xx X//xXx xxxxx/Xx/xxx

X = X = / =

masen nai desu a new video viewing break

Inanimate Existential Verbs (arimasen vs. nai desu) 4 IM01 Xx// 11 IF01 /X/xx “maybe” modality expressions (kamoshiremasen vs. kamoshirenai desu) 7 M07 /X/xXX

the cases in boldface in Table 7 above are indicated in order of appearance in each conversation. In Table 8, an uppercase X indicates the use of masen forms and a lowercase x indicates the use of nai desu forms. Table 8 clearly demonstrates that there is somewhat of a tendency for the masen forms, if they do occur, to do so before the nai desu forms do, and that once a nai desu form is used, it keeps being used for the rest of the discourse unless sometimes when the masen form (re)appears after the conversation comes to a halt for viewing another video clip and then resumes on a new topic. Consider the following examples. Speaker F13 in Dialogue #13 uses the masen Verbal form only at the beginning of her discourse and uses the nai desu form of Verbs thereafter. Speaker M22 in Dialogue #22 uses the nai desu form of Verbs most of the time but uses the masen form once right after a discussion break has finished and the conversation has resumed with a new topic. Table 9 below lists the actual forms used by Speakers F13 and M22 in the order of their appearance, each with their affirmative basic form and its English gloss in parentheses. “/” is again used to indicate a discussion break during video viewing.

Masen or nai desu: That is the question 



Table 9.  Appearance orders of all polite-style negative forms Speaker

Polite negative forms of verbs used

F13

shiMASEN (suru ‘do’) – – / wakanNAI DESU (wakaru ‘know/understand’) – – wakanNAI DESU – – dekiNAI DESU (dekiru ‘can do’) – – wakanNAI DESU (wakaru ‘know/ understand’) – – wakanNAI DESU – – tonNAI DESU (toru ‘take’) iNAI DESU (iru ‘exist [animate being]’) – – iwaNAI DESU (iu ‘say’) – – iNAI DESU (iru ‘exist [animate being]’) – – shin’yoo dekiNAI DESU (shin’yoo dekiru ‘can trust’) – – shiraNAI DESU (shiru ‘know’) – – / iiMASEN (iu ‘say’) – – soodan wa shiNAI DESU (soodan suru ‘consult’) – – / tsukawaNAI DESU (tsukau ‘use’) – – wakanNAI DESU ‘wakaru ‘know/understand’) – – shiNAI DESU (suru ‘do’)

M22

An interesting case is M22’s use of both negative forms of the same verb iu ‘say’ in the same conversation with the same interlocutor. The nai desu form, iwanai desu (his second Verb in Table 9), is used where both the preceding and following Verbs are in the nai desu form as well. The masen form, iimasen (his sixth Verb in Table 9), is also preceded and followed by nai desu forms of Verbs, but the difference here is that this Verbal negative expression is used right after a break in the discussion. The relevant parts of Dialogue #22, one discourse segment where M uses a nai desu form and another segment right after a video viewing break, where he uses the masen form and then a nai desu form, are shown in (3) and (4), respectively.

(3) (talking about their classmates from abroad as to whether they say “Gohan tabemashita ka” (Have you eaten?) in Japanese for a greeting)

IF: Huun, huun. Chuugokukei no hito warito iu to omou n desu kedomo … ‘Hum, Hum. Chinese people often say so, I think...’ M: E? Soo desu ka? // ‘Well, is that so?’ IF: //Un. Iwanai? ‘Yeah. Doesn’t he?’

Ja, // moo narete n da na, kotchi no seikatsu ni ne? // Un. ‘Well, then, he is already accustomed to the life here, perhaps?’

M: //Iwanai desu ne.// Hai. ‘He doesn’t make such a greeting. Yes.’

Chuugokukei no amerika-jin mitai na hito …. desu kara. Hai. ‘He looks like a Chinese-American … That’s why… You know.’

(4) (after viewing another video clip of a new topic)

IF: Sa, o. oya o zettai-teki ni shin’yoo shiteru kodomo ….no yoo desu keredomo... ‘Now, he has a complete trust in his parents,… it seems, you know.’

 Satoshi Uehara and Etsuko Fukushima

M: Hai. ‘Yes.’ IF: Gojishin wa oya ni taishite, sonna shin’yoo wa, shiteru hoo desu ka? ‘How about yourself? Do you have such trust in your parents?’ M: Shin’yoo shitemasu nee, // ima wa. ‘I do have trust in them, now, at least.’ IF: //Hooo. Yappari oya no iu koto wa tadashii tteyuu kanji? ‘Oh! After all, parents are right, you mean?’ M: Iya, zenmen-teki ni tadashii to made wa iimasen kedo[mo], // ‘Well, I wouldn’t say they are ALWAYS right, but..’ IF: //Un un ‘Hum Hum’ M: Aru, aru teido no koto wa honto ni tadashii n ja nai ka to. ‘To, to the most part, what they say is really true, I suppose…’ IF: Huun. Ja, komatta toki wa oya ni soodan shimasu? Yappari? ‘Hum. Well, then, you go to your parents when in trouble? Of course?’ M: Ee....// ‘Well…’ IF: //Doo shiyokka na mitai na? ‘Like when you are at a loss what to do?’ M: Soo desu ne, anma soodan wa shinai desu ne, oya ni wa… ‘Well…, for advice I seldom go to my parents…’ It should be noted that non-contracted forms (e.g. keredomo ‘but’) and even an honorific form, go-jishin ‘honorific-oneself ’, are used right after the video viewing break in (4), while later, contracted forms (e.g. anma ‘(not) much’) are observed. We will get back to this point later in 5.2. The observation concerning appearance order can be rephrased as follows. The masen forms are used at the beginning of one’s speech and/or when resuming one’s talk with the same interlocutor after a break in the conversation, provided a new topic is introduced. The nai desu forms are used otherwise. Once the speaker starts using the latter, therefore, she keeps using it and does not switch (back) to the former except for some instances when the conversation resumes with a new topic after a break. This generalization seems to reasonably account for the prevalence of nai desu forms in our conversational data: conversational discourse in our data (and in our daily lives) does not always start and come to an end right away, and the longer a conversation continues, the more nai desu forms end up being used. In fact, this generalization accounts for 9 out of the 14 cases of discourse in question (64%) and for 16 out of the 21 cases of shift in form in Table 6 (76%).

Masen or nai desu: That is the question 



hanashi ('speak') MASEN X

IM01 F01

X ii ('say') MASEN

X ii ('say') MASEN

shi ('do') NAI DESU x

kamo shire NAI DESU x

x x wake ja koto mo NAI DESU NAI DESU

Two consecutive utterances by one speaker Two consecutive utterances by two different speakers One or more other utterances in between English gloss for a Verbal stem

Figure 1.  Use of all negative forms in the beginning part of Dialogue 1

Furthermore, this pattern of usage of the two polite-negative forms observed in those cases of discourse where a speaker uses both negative forms of the same word categories partially correlates with the usage patterns of two variant forms in non-Verbal predicates and in the interlocutor’s conversation. In other words, even though the use of the two negative forms in some other word categories is not distinguished, the two forms are used in a manner similar to those in the cases discussed above, where their contrast is evident. This is illustrated by Figure 1, which lists in order of appearance all the polite-style negative forms used in the first part of Dialogue 1. As shown in Figure 1, not only IM01’s two Verbal negative forms (the first is a masen form and the second a nai desu form), but all others as well, including those in the interlocutor’s discourse, follow the pattern just described. The usage pattern of the two polite-style negative forms in our conversational data clearly shows that, although the two forms both belong to the polite-style level, they have slightly different functions. It suggests that the two forms represent two sub-levels in the polite-style level of the language and that interlocutors shift “styles” from one sublevel to the other in the middle of an ongoing discourse. We will get back to this point later after we examine exceptions to the general usage pattern in the next section. 5.1.2 Other related factors We have seen that the usage pattern of the two negative forms in terms of their appearance order is largely predictable and can account for most of the cases in Table 8. There are five cases, however, that constitute exceptions to the generalization, namely, Dialogues #9, #12, #18, #22 (IF03), and #7. Let us closely examine these five exceptional cases in this section. Examination of the contextual occurrence of the five cases reveals three factors, which are mutually related and are also related to the appearance order paradigm described above. First, when the topic of conversation turns toward something with which the hearer, but not the speaker, feels comfortable or familiar, the speaker “retreats” to

 Satoshi Uehara and Etsuko Fukushima

have some distance from the hearer, and the masen form is selected even in a non-initial stage of discourse, and even if it is preceded by a nai desu form. This context can be identified for two of the five exceptional cases in question, namely, #9 and #22 (IF03), the topics of which were sex and the koseki ‘Japanese family registry’ system, respectively. The relevant part of Dialogue #9 is reproduced in (5) as an illustration:

(5) (talking about differences in parent-child relationships between the US and Japan)

F: Nandemo hanasu, tatoeba sono //, sei, ni tai // suru koto tte ‘(Parents in the US) talk about anything, say, uh-, about sex...’ IM: //Un. //Ee. Ee. ‘Hm, Hm Hum’ F: Kooo nihon no oya tte nanka, ocha o nigosu tte{laugh}iu kaa, ‘You know, Japanese parents, like, just give evasive replies (laugh) or something,’ IM: Iyaa, oyaa izen ni watashi mo, oya ni taishitee soo iu wadai wa kotchi kara mo mochikakemasen kedo nee, // mazu. ‘Not only parents, but children as well, like me, would not bring up such a topic, in the first place.’ F: //Nn. Demo nanka, amerika wa sugoi oopun nan desu yo. ‘Yes. But anyway, people in America are very frank, you know.’ IM: Un, dakara sakki mo, kodomo no hoo kara mo hanashiteta mitai desu yo nee. ‘Yes, I suppose that’s why before (in the video) it seemed children brought up such topics.’ In (5), IM and F were talking about parent-child relationships in the United States. Incidentally, F, who had spent some time in the U.S., mentioned that sex was one of those topics that American parents feel free to talk about with their children. IM, whose tone of voice sounded relaxed up until that time, suddenly became “distant” perhaps due to finding himself talking about sex with a woman he had just met. In fact, he had been using nai desu forms for Verbs and all other predicates as well up till this time, when he suddenly switched to the masen form in trying to preserve a little distance between himself and his interlocutor due to the sensitivity of the topic. He used the masen form only once there, after which he switched back to the nai desu form. In a similar manner, the other exceptional case of masen in Dialogue #22 occurred when the speaker confirmed with the hearer, an undergraduate student in the Department of Law, about systems of family registry in other countries. Second, we observed in our exceptional cases that one interrogative sentence without the question particle ka used the masen forms. This is the case in Dialogue #18, the relevant part of which is reproduced in (6) below for illustration:

Masen or nai desu: That is the question 



(6) (talking about F’s classmates from abroad) IF: E, ryuugakusei wa ajia no hoo wa // ookatta no? ‘uh.. foreign students from Asia were numerous?’ F:

//Un. Anoo, ‘Hm, uh-..’ chuugoku to, a, ima wa chuugoku to // kankoku? ‘mostly from China, well, from China and Korea, now, you know?’

IF: //Un. Huun, ‘Hm. Hum,’ de gohan tabemashita ka tte iimasen? ‘and don’t they say “Have you eaten?” (for a greeting)?’ F: [Kiita koto nai.] ‘Never heard of that.’ Questions in Japanese are generally formed by adding the so-called question particle ka at the end of a sentence, but in colloquial language it is possible to form a question by just ending the sentence with a raising intonation but without adding ka. This pattern gives a casual tone to the utterance in question. In polite speech, the masen form can be used with this idiomatic question pattern, but the nai desu form can not (yet). Thus, all native speakers we consulted with observed that the use of iwanai desu (‘say. NEG’) with a question intonation is not felicitous when used in place of iimasen (‘say. NEG’) in (6) above, and some added that iwanai desu with a question intonation does not sound right in any context. In other words, the interrogative sentence without the question particle ka appears to grammatically avoid the nai desu form and thus strongly invites the use of the masen form, which accounts for the occurrence of the latter in (6). In fact, the speaker IF used the masen form here in (6) and an idiomatic expression sumimasen in the beginning of the conversation. Her six other uses of the polite negative expressions are all in the nai desu form in Dialogue #18. The final contextual factor commonly identified in the two other exceptional cases in Dialogues #12 and #7 is one that invites the use of a nai desu form over a masen form, rather than the opposite, as did the other two contextual factors. In these two cases the speaker slows down a little as if thinking what to say, and speaks as if he “forgets” to be polite. He then throws in desu after nai as if he suddenly remembered his manners. Thus, in (7), which reproduces the relevant part of Dialogue #7, Speaker M uses kamoshinnai, a contracted (i.e., more casual) form of kamoshirenai ‘maybe’, as though the speaker had temporarily forgotten to be polite, and adds desu after a short pause (indicated by a comma): (7) (talking about a married couple having different surnames) M: {laugh}Kaero toka, {laugh} iyada toka iware tara, ‘If my wife says, (laugh) “YOU change yours” or (laugh) “I don’t wanna change mine”...’

 Satoshi Uehara and Etsuko Fukushima

IM: Un. ‘Humhum.’ M: [demo] suru ka to kikare tara, nn, dotchi de mo ii ki ga shimasu kedo, ‘if asked whether or not I will do so, well, I feel as if either way is ok, but …’ IM: Ah. ‘Humhum.’ M: [...] soo iware tara chotto [...] mayou [no] kamoshinnai, desu kedo, ‘if I was in that situation, you know, I might.. be a bit at a loss, though.’ It should be noted that Speaker M in Dialogue 7 is one of the few speakers who used more masen forms than nai desu forms in his conversation, and that for “maybe” modality expressions he used the nai desu form only once while the three other such expressions were all in the masen form.

5.2

Summary and discussion

The following three observations may stand as a summary of our findings from the preceding sections. First, the two negative forms, masen and nai desu, are both polite style predicates, but nai desu forms occur in less formal/more casual contexts. We base this conclusion first, and most crucially, on the differing contexts in which these two forms appear. Both are used in conversations between two people who have just met. When a speaker alternates between these two forms, masen forms are typically used in the beginning of the conversation and/or after a resumption of the conversation (following a break) about a new subject. Nai desu forms, however, are used at other times— for example, as the speaker’s familiarity, and possibly comfort, with the interlocutor increases in the course of their “first-encounter” conversation. The exclusive use of the masen form in the apologetic expression sumimasen ‘pardon me; I’m sorry’ also supports the above point. Sincere apologetic expressions (as those cases in our data) are perfectly compatible with courteous language but are incompatible with casual speech. Note also the much greater frequency of the nai desu forms used with contractions in our data as well as their own ability to contract. This flexibility of the nai desu forms lends support to our argument that they mark less formal contexts than masen forms do. The nai desu form in (7) above shows such contracted forms. (8)-(10) are some other examples from our data: (8) [Wak]kan’nai, desu kedo (Dialogue #8) ‘I don’t understand, though.’ (9) Doo na no ka wakaranai[de]su ne (Dialogue #4) ‘I don’t know what it is.’

Masen or nai desu: That is the question 



(10) Maa, suru n ja nai su ka ne (Dialogue #8) ‘Well, (he) would probably do that.’ Notice here that there are two types of contraction for nai desu polite negative forms: one contracts the morpheme preceding nai desu as in (7) and (8), and the other applies to the nai desu form itself, i.e. the weakening or deletion of de of desu as in (9) and (10). The former type of contracted forms is used quite frequently without the following desu, i.e. in plain speech. Such contracted forms signify the speaker’s casual mode of speech, more typical of plain style than polite style. In fact the profusion of nai desu polite negative forms composed of such contracted negative (nai) forms plus desu suggests that the casualness associated with the plain style carries over to the contracted nai desu forms. The contracted nai desu forms also demonstrate that such forms belong to a less formal level due to the corruption of the politeness marker itself. These two cases strongly suggest that the overall politeness level of nai desu forms is less than completely formal. Our second observation is closely related to the first one. The masen form is used when “politeness” is foregrounded, i.e., when the speaker becomes keenly aware of the necessity of using polite speech toward the listener. The nai desu form, on the other hand, is used when the conveyance of something else takes precedence over that necessity, which then becomes secondary and supplementary. This nature of nai desu forms becomes clear when we consider multiple examples, such as (7) above, in which the speaker slows down, seemingly thinking of what to say while speaking, and pauses before he adds desu at the end. We also observe nai desu forms frequently in overlapped utterances, as in (11) below taken from Dialogue #3. In these cases, before a speaker finishes an utterance with desu, that is, right after saying nai, her interlocutor interrupts and starts to speak: (11) (after viewing a video clip on Japanese people’s greetings) IM: Anoo, ii tenki desu ne toka iimasu [ka]. ‘Uh- do you say, “It’s nice weather, isn’t it?” (for a greeting)?’ F: Sore wa iwanai // desu ne. ‘I wouldn’t say that.’ IM: //Iwanai desu ka, [yappari]. ‘You wouldn’t say that, that’s what I thought.’ In other words, in such contexts, the interlocutor ignores desu since he is more interested in content than formality. Finally, the difference between masen and nai desu forms follows the principle of “iconicity” (Haiman 1980, 1983), i.e., there is a parallelism between their form and function, and the difference in their use is consonant with their morpheme order. Masen is the combination of mase, a politeness morpheme (see masu for Verbs in Table 1), and n, a negative morpheme (still prominent in Kansai dialect), while nai desu consists of nai, a negative morpheme, followed by desu, a

 Satoshi Uehara and Etsuko Fukushima

politeness morpheme. Overlapped utterance patterns of the kind shown in (11) are made possible because of the morpheme order of nai desu forms. A final observation, inferred from the above two: at least for polite-style negative predicates, a more casual, or “semi-polite” (Hudson, this volume) level, which can dictate the use of nai desu forms, is now commonly used in present-day conversational Japanese. Thus, with these two negative forms, speakers can shift styles slightly in the middle of ongoing polite discourse. As historical studies (and the lack of mention in Japanese language textbooks) show, the nai desu form itself is a rather recent addition to the Verbal paradigm. This becomes clear when we note that in our data Verbs are the only word class which still privileges masen forms to some extent (see Table 5). Nai desu forms’ entrance into the Verbal paradigm, consequently, creates a complete wordclass paradigm, parallel to that of masen forms. Thus we conjecture that the subtle stylistic contrast between the two polite-level negative forms can be fully and systematically represented. However, only the masen form has an established use in the question pattern without ka. This suggests that the nai desu form has not completely attained the same status as the masen form.14 Our data demonstrate that this “semi-polite” level, established for polite negative predicates and associated with nai desu forms, does not (as yet) exist for polite affirmative predicates (but see Hudson, this volume). No instance of variant forms of polite affirmative predicates appeared in our data. In other words, all the polite affirmative predicates in our data were used with their traditional inflectional forms (listed in the polite-affirmative cells of Table 1). This in turn means that the subtle distinction between the two sub-levels of the polite level, separately denoted by two polite negative forms, has not been established in the case of affirmative predicates, where one affirmative form (masu for Verbs and desu for others) covers both sub-level contexts. (Note the affirmative masu forms coexisting with the negative nai desu forms in (7) and (11) above.) Thus, our data show that no variant forms exist for polite affirmative predicates, or that, if variant affirmative forms do exist, they do not occur in the same politeness level as the two sub-levels covered by their negative counterparts. In fact, in the case of affirmative counterparts to the polite negative forms, no word classes are yet equipped with the two polite forms with full-fledged inflectional paradigms. Adjectives, for example, have one polite-style affirmative (non-past) form A-i desu but no parallel form for the sublevels we have described in negative usage. Therefore, the polite sub-level distinction in affirmative predicates, if any, is no match for the flexibility and subtlety of negative predicates in present-day spoken Japanese. In this regard consider the curious paradigm of the recently emerging (s)su form, as in Aitsu mo iku-ssu yo (‘That guy will also go, you know’). The (s)su form is presumably 14. This is true not just for Verbs but for other word classes as well. This seems to be partly due to the fact that the nai desu form itself was established much later than the masen form in the history of the language. See Kokugoshingikai 1952, for example, for the establishment of the Adjective + desu pattern, of which the nai desu form is an instantiation.

Masen or nai desu: That is the question 



a shortened form of desu (as in nai desu) and indicates some degree of politeness, although it is considered too colloquial for general use and gender-marked because it is typically used by male university students when speaking to their senior/fellow students. It is thus unimaginable to find, for example, middle-aged women using the (s)su form in speaking to anyone. This form, however, can be attached to any plain-style predicate forms, non-past or past, affirmative or negative to create a degree of politeness (appropriate for the contexts just described), so its inflectional paradigm covers all word classes. The “semi-polite” nai desu form we have examined in this paper resembles the (s)su form both in form and function. Recall that desu of nai desu forms often get contracted to su, and like the (s)su forms, functions in a discourse situated between the polite and plain levels of speech. If these two patterns, nai desu (used generally but for negatives only) and (s)su (used for both affirmatives and negatives but marked in use), somehow merge in form and function in the future, then the “semipolite” level will be a bona fide style category comparable to polite and plain styles.

6. Conclusion This paper has examined two different forms for polite-style negative predicates, namely, masen and nai desu forms, in Japanese conversational discourse. Our analysis has demonstrated that these two forms are used systematically in two slightly different contexts although both are used in conversations between two people who have met for the first time, and can therefore be considered to belong to the same, polite style. We have argued that the two variant negative forms, masen and nai desu, are not synonymous, but mark two slightly different sub-levels of polite-style speech, with the latter marking the less formal, “semi-polite” sub-level. Furthermore we contend that this subtle speechlevel distinction has been established for polite-style negative predicate forms. The current research thus poses some problems for the more simplistic polite/ plain style model we tend to have of Japanese. Japanese speakers now have another resource available to them when they want to shift styles slightly in the middle of polite discourse. Speakers can use this masen/nai desu negative distinction to shift the tone of the conversation to a slightly more casual level as the talk proceeds, or to shift in response to a topic that seems to require a little more caution or distance. What will become of this “semi-polite” level? This is an interesting question. As of now, it creates a minor distinction largely specific to negative predicates. Possibly this minor distinction for negative predicates will be extended to affirmative predicates so that the two-way polite/plain style distinction of present-day Japanese will eventually become a three-way polite/semi-polite/plain distinction. The recently emerging, (s)su form, which resembles the nai desu form both in form and function, seems to suggest this direction. Korean, for example, a language which shares many typological characteristics with Japanese, divides addressee honorifics into not just two, but four or more levels (e.g. Martin 1964). It is also possible that the subtle distinction in polite negative

 Satoshi Uehara and Etsuko Fukushima

predicates will eventually disappear, with nai desu forms covering both sub-levels by effectively eliminating masen forms, so that the language will revert to a two-way style distinction. This seems possible because the negative morpheme n of masen is a remnant of an older negative inflection and may well become pleonastic in modern Japanese. Our data show that except for Verbs and the idiomatic expression sumimasen ‘Pardon me’, the use of nai desu forms predominates in all word classes in modern spoken Japanese. Time, of course, will determine which scenario prevails.

References Bunkachoo Bunkabu Kokugoka. 1999. Seronchoosa Hookokusho: Heisei 10-Nendo Kokugo ni kansuru Seronchoosa. Bunkachoo Bunkabu Kokugoka. Erman, B. and Warren, B. 2000. “The idiom principle and the open choice principle.” Text 20 (1): 29–62. Fukushima, E. and Uehara, S. 2001. “Gendai nihongo ni okeru teineitai hiteikeishiki.” Bulletin of Tohoku University International Student Center 5: 11–17. Fukushima, E. and Uehara, S. 2003. “’Iimasen’ to sika boku wa iwanai desu: 2 polite-style negative forms in Japanese conversational discourse.” In Linguistics and Japanese Language Education III, M. Minami and M. Asano (eds.), 269–286. Tokyo: Kurosio Publishers. Haiman, J. 1980. “The iconicity of grammar: Isomorphism and motivation.” Language 54: 565–89. Haiman, J. 1983. “Iconic and economic motivation.” Language 59: 781–819. Hudson, M. E. This volume. “Riyuu ‘Reason’ for nai desu and other semi-polite forms.” Jorden, E. H. and Noda, M. 1987. Japanese: The Spoken Language. New Haven: Yale University Press. Kobayashi, M. 2005. “On the use of ‘–masen’ and ‘–nai desu’ in Japanese conversation.” Nihongo Kyooiku 125, 9–17. Kokugoshingikai. 1952. Korekara no Keigo. Monbushoo. Langacker, R. 1987. Foundations of Cognitive Grammar. Vol. I: Theoretical Prerequisites. Stanford: Stanford University Press. Martin, S. E. 1964. “Speech levels in Japan and Korea.” In Language in Culture and Society: A Reader in Linguistics and Anthropology, D. Hymes (ed.), 407–415. New York: Harper and Row. Noda, H. 2004. “Factors relating to the use of -masen and -nai desu as polite negative form: Based on a usage investigation and on a questionnaire investigation of young generation speakers.” The Mathematical Linguistic Society of Japan 24 (5), 228–244. Tanomura, T. 1994. “A corpus-based study on the polite negative predicate forms in Japanese.” In Journal of Osaka University of Foreign Studies 11: 51–66. Teramura, H. 1984. Nihongo no Shintakusu to Imi. Tokyo: Kurosio Publishers. Tokieda, M. 1950. Nihonbunpoo Koogohen. Tokyo: Iwanami Shoten. Uehara, S, Fukushima, E., and Kitano, H. 1998. “Soo janai n janai n desu ka?: A conversationalbased study of janai in Japanese.” In Linguistics and Japanese Language Education, Y. S. Alam (ed.), 197–212. Tokyo: Kurosio Publishers. Wray, A. 2002. Formulaic Language and the Lexicon. Cambridge: Cambridge University Press.

The power of femininity Can Japanese gender variation signify contradictory social meanings?* Yuka Matsugu

1. Introduction Gender-related variation has long been an attractive topic for scholars from various fields. However, despite the fact that various approaches propose different relationships between gender and language, the concept of gender as a binary opposition has often been taken for granted. Furthermore, in these binary analyses, gender dualism seems to correlate with the relative power of women and men as expressed through their language use.1 That is, if a language has gender variations, “women’s language” is said to show speakers’ softness, weakness, powerlessness, and politeness, while “men’s language” is said to show speakers’ assertiveness, strength, power, and directness. There may be forms which are preferred by one gender group (Bodine 1975, Crosby and Nyquist 1977, Labov 1966, Trudgill 1972), and in some languages, it may be safe to say that a group of variations, “women’s language,” has nuances of softness and

* This paper has benefited greatly from the help of many people. Kimberly Jones and Tsu­ yoshi Ono provided comments on numerous drafts, and I wish to offer them my sincere gratitude and appreciation for their helpful suggestions. Many thanks also to Jane Hill, Susan Philips, Timothy J. Vance, Natasha Warner, and Linda Waugh for their useful comments on a previous draft of this paper. In addition, I would like to thank my colleagues and classmates while I was a graduate student at the University of Arizona who, on various occasions, provided very thoughtful questions and suggestions with respect to the issues under consideration here. I also wish to thank to Maggie Camp, Peter Norquest, Heidi Orcutt-Gachiri, and Sean Winship for assistance with the manuscript. Any errors are, of course, my own. 1. Scholars agree that binary gender ideologies are ideological constructs, and there are cultures with three or more named gender categories. Nevertheless, gender dualism is very strongly developed and salient in gender ideologies around the world and through time (Philips 2001: 7). I believe that this tendency also exists with respect to research into ideas about gender variations in language and to the understanding of their normative characteristics.

 Yuka Matsugu

powerlessness.2 Yet, even in those languages, such as Japanese, it is important to ask whether the supposed correlation between “women’s language” and powerlessness is always true in actual usage. Ochs (1992: 336–337) states that concepts of gender are socialized, sustained, and transformed through talk, and as a result, in certain contexts, images of women can be linked to other images (such as mothering) that are constituted and mediated by other social constructs. In this vein, I propose that a particular sub-group of Japanese “women’s language,” “super-feminine” forms, can index not only images of “womanhood/ femininity,” but also “power” in certain situations.3 In other words, the use of the “super-feminine” form as a style – a variation to control and identify intra-speaker identities4 – does not only signify a fixed identity or a set of identities. Rather, depending on the context, it can signify different identities that may seem somewhat contradictory (e.g., softness and powerfulness).

2. The relationship between the status of women and the politeness, formality, and powerlessness of their speech is by no means as simple and straightforward as has been assumed (Brown 1980: 112). For instance, Keenan (1974) reports that women are less polite than men in a Malagasy village. Likewise, Kulick (1992) reports that in a Papua New Guinean community, women are considered to be the “emotional gender,” who explicitly show their anger in public. These studies, however, analyze these phenomena as results of the fact that women do not (or “cannot”) obey sociocultural norms as responsible members of society, which indirectly shows their inferiority to men. In other words, a society may put pressure on women to behave as subordinates (e.g., soft, weak, and powerless), but at the same time, women’s misbehavior may be expected and accepted because of their lower status in a society. I am aware of the danger of simplifying this complex relationship between gender and language. However, it has been said that the cultural ideal for Japanese women is to be powerless and obedient (Bernstein 1991, Horii 1993, Lebra 1985), and their language usage has been characterized as weak and polite (Ide 1982, 1990; McConnell-Ginet 1988; Mizutani and Mizutani 1987; Okamoto 1996: 290; Reynolds 1985; Smith 1992). Thus, I focus on this particular relationship between gender and language – women and their lack of power as expressed in their language use – as one to be questioned in this paper (see footnote 3 for a definition of “power”). 3. Power is a complex term and is the subject of a great deal of debate as regards its nature and forms (Wetzel 1988: 392). Lakoff states that the expression of sorrow is an expression of powerlessness and helplessness; anger, of potency (2003: 163). Roughly following her statement and Galbraith’s (1983) definition, I define power as the possibility of imposing one’s emotions and thoughts upon other persons. In other words, if one can directly express disagreement and/or displeasure by using a particular linguistic feature, I submit that the linguistic feature can express a speaker’s power. 4. In an article called “‘Style’ as Distinctiveness” (2001), Irvine defines style as a system of intra-speaker distinction, in which a style contrasts with other possible styles. According to her, style is a linguistic sign which signifies contrasting social meanings, and the social meanings signified by the style have to be interpretable for other members in society. I adapt her definition of style in this paper.



The power of femininity 

After summarizing traditional understandings of Japanese gender variations, I analyze three conversations individually. In the analysis, I argue that Japanese “superfeminine” forms as a style are multidimensional and locally granted. The argument also covers the factors that determine when a specific dimension of a “super-feminine” form becomes more salient than the others. I hope to show clearly the significance of style, as well as the phenomenon called style shift, in interactions.

2. Japanese gender variations Variations in the Japanese language are notably rich at the morpheme level (Matsumoto 2001). Consequently, early attempts to study gender and language issues in Japanese focused on establishing a list of morphemes representative of the speech of each gender. These morphemes are commonly categorized into five sub-groups, “super-feminine” forms, “feminine” forms, “neutral” forms, “masculine” forms, and “super-masculine” forms, depending upon how frequently one gender uses the forms (e.g., Okamoto and Sato 1992: 478–488; Shibamoto 1985: 58–64, 1987: 25–49; Tsujimura 1996: 372–377).5 Although there are some forms that different scholars categorize in different groups, there is general agreement on most of the categorizations. For this particular study, I follow the categorization that is recognized in most research.6 Figure 1 shows a conventional description of Japanese gender variations: the functions in balloons are followed by categorizations, users’ gender for each category, and some example forms. Additionally, the forms that are the focus of this paper, “super-feminine” forms, are listed in Table 1. Some linguistic research suggests that conscious “misuse” of gender forms by young, innovative female speakers often causes a change in the classification of those forms (e.g., Wetzel 1988). That is, the categorizations of particular forms are not permanent, but rather, they can switch categories over time. Nevertheless, this “misuse” is not eradicating the basic idea that the forms can be categorized (Takasaki 2002), and to my knowledge there is no report of “super-feminine” forms becoming “neutral” or “masculine.” In consideration of these factors, it seems fair to say that the “super-feminine” forms are understood as a gender variation that exclusively indexes prototypical femininity, such as powerlessness, softness, and politeness. Despite this fact, I have 5. In my analysis, I continue to use the term “feminine/super-feminine” forms, “neutral” forms, and “masculine/super-masculine” forms, although I am aware of the danger of perpetuating the stereotypes which these terms may represent. Also, it should be noted that in some earlier studies such as Okamoto and Sato (1992), the terms “strongly/moderately feminine/masculine” forms, are used instead. 6. The categorization for this paper is based on the following studies: Mizutani and Mizutani 1987; Okamoto and Sato 1992: 478–488; Shibamoto 1985: 58–64, 1987: 25–49; Terao and Zimmerman 2000: 41–57; Tsujimura 1996: 372–377; and Uchida 1993.

 Yuka Matsugu

Hegemonic Femininity: softness, powerlessness, weakness, politeness, etc.

Hegemonic Masculinity: assertiveness, directness, strength, power, etc.

Women’s Language SuperFeminine feminine Users

Forms

Almost exclusively women Copula+wa Copula+wa yo Copula+wa yone kashira

Neutral

Men’s Language SuperMasculine masculine

Mostly women

Both genders

Mostly men

Verb(te-form) Verb(te-form)+ne Verb(plain)+no desho(o)

Verb(plain) Noun mon kana

Verb(plain)+yo Copula+ne Copula+yo daro(o)

Almost exclusively men Verb+na ze zo ka yo

Figure 1.  Japanese Gender Variations

Table 1.  List of “Super-Feminine” Forms Forms

Examples in Japanese

English Translation *

wa / ** wa ne wa yo wa yone Copula + wa Copula + wa ne Copula + wa yo Copula + wa yone Noun + yo / Na-adjective(stem) + yo / Noun + yone Noun + no Na-adjective + no no ne no yo no yone kashira

Iku wa (mild emphasis) Iku wa ne Iku wa yo Iku wa yone Ashita da wa Ashita da wa ne Ashita da wa yo Ashita da wa yone Ashita yo Kiree yo Ashita yone Ashita na no Kireena no Iku no ne Iku no yo Iku no yone Iku kashira

Go wa Go wa ne Go wa yo Go wa yone It is tomorrow wa It is tomorrow wa ne It is tomorrow wa yo It is tomorrow wa yone Tomorrow yo Beautiful yo Tomorrow yone Tomorrow na no Beautiful no Go no ne Go no yo Go no yone (I wonder) Go kashira

Many particles in the Japanese language express various modalities (Kitahara 1998; Ko­ noshima 1966; Sato 2000) and are not easily translated into English. Therefore, they are left in their original forms in this table. *

/ = Rising intonation; No intonation marker means that the form is categorized as a “superfeminine” form regardless of its intonation. **



The power of femininity 

occasionally observed Japanese female speakers shifting to “super-feminine” forms in order to complain explicitly and/or to show authoritative power. In the following sections, I discuss how this impressionistically counter-normative use of “super-feminine” forms can take place. Three naturally occurring conversations – two single-sex female conversations and one mixed-sex conversation – are closely examined. The nature of each conversation (e.g., type of situation and participants) is explained, along with the analysis.

3. “Super-feminine” forms, womanhood, and power 3.1

A mother-daughter conversation

Excerpt (1) in section 3.1. and excerpt (2) in section 3.2. show that the “super-feminine” forms can index not only values normatively associated with femininity, such as softness, weakness, and politeness, but also women’s maturity compared with that of girls. Consequently, the use of such forms can raise speakers’ status as adult members of society and help them to express their power in interactions. The first conversation I examine here is a 70-minute, mother-daughter conversation which was videotaped in Tokyo in 1999. The mother in this data set (who is 52 years old) and the daughter (who is 24 years old) have a lot in common. They are both speakers of so-called Standard Japanese, were born and grew up in Tokyo, and graduated from junior college. They are both working and supporting themselves. The 70-minute data set is the last part of a 2-hour conversation that was held during and after dinner in their living room from nine o’clock to eleven o’clock. This study focuses on the only time in the 70-minute conversation that the daughter used a “super-feminine” form. The transcription follows the Du Bois et al. system (Du Bois et al. 1992)7 and contains close to 1600 intonation units (roughly 710 clauses),8 excluding backchannels and quotations. Throughout the conversation, the daughter mainly ends her utterances with the bare form of a noun, verb, or adjective (198 times, 55%), or a bare form followed by a neutral sentence-final particle (91 times, 25%). These 289 cases (80%) are categorized as “neutral” forms. “Masculine” forms and “feminine” forms are observed only 46 times (13%) and 27 times (7%), respectively, and there are no “supermasculine” forms or “super-feminine” forms except the one that is the focus of this section. Because of the distribution of the daughter’s gender form usage described above, her unmarked gender style in the conversation with her mother is considered

7.

See the appendix for transcription conventions.

8. Following Jorden’s (1987) classification, I count three types of clauses: verbal, nominal, and adjectival.

 Yuka Matsugu

“neutral,” and the instance with the “super-feminine” form (excerpt [1], line 36) raises the question of why she chose to use that particular form. Five minutes before the excerpt, the mother had looked at the newspaper, seeming surprised, and said, kono hito shinda no? ‘did this guy die?’ Despite the fact that her daughter quickly replied to her by saying dare? ‘who?’, the mother continued to read the article without interacting with her daughter. Later, the article caught the daughter’s eye. She showed surprise and started the conversation as shown in excerpt (1).

(1) (Mother M and Daughter D, following more than half a minute in which no talk occurred) 1 D: ((looking at a newspaper)) a! ‘Oh!’ 2

Yamada san? ‘Mr. Yamada?’

3 M: ... un. ‘... Yes.’ 4

shitteru desho? ‘You know who he is, don’t you?’

5 D:

un. ‘Yes.’

6 M: ne:. ‘Right.’ 7

bikkuri shichatta. ‘I was so surprised.’

8 D:

.. e:, ‘.. Wow’

9

bikkuri. ‘what a surprise.’

10 M: ne:. ‘Isn’t it.’ 11 D:

nansa:i? ‘How old was he?’

12 M: rokujuusan datte. ‘It says 63.’ 13 D:

= wakai jan. = ‘Young, isn’t he.’

14 M: = dakara sa:. = ‘That’s why,’ 15

... (2.0) [XXX],



The power of femininity 

16 D:

[kyuushi] jan ne:. ‘[Sudden death], isn’t it.’

17

[2 kyuushi 2]. ‘[2 Sudden death 2].’

18 M: [2 u:n 2]. ‘[2 Yes 2].’ 19 D:

= a, = ‘Oh,’

20

kyuusei shinfuzen da. ‘it says “a heart attack.”’

21

... fu:n. ‘... Well.’

22

... (3.0) ima, ... (3.0) ‘These days,’

23

nani shiteta n dakke? ‘what had he been doing?’

24 M: ... <WH nani shita n daroo WH>. ‘... <WH I wonder what he had been doing WH>.’ 25 D:

... (4.0) fu:[:n]. ... (4.0) ‘[Well].’

26 M: [ne:]. ‘[Yes].’ 27 D:

... ara ara, ‘... My, my,’

28

bikkuri. ‘what a surprise.’

29 M: ... ma:, ‘... Well,’ 30

anata mo shitteru nante omowanakatta wa. ‘I didn’t think that you would know him, too.’

31 D:

e? ‘What?’

32

kono hito tte, ‘This guy,’

33

deteta jan. ‘was on TV, wasn’t he.’

 Yuka Matsugu

34 M: deteta yo ne:. ‘Yes, he was, you know.’ 35 D:

un. ‘Yes.’

36→

... (3.0) shitteru wa yo: sore gura:i. ... (3.0) ‘Of course I know that much.’

37 M: iya wakannai wa, ‘Well, who knows,’ 38

[Yatchan] XXX, ‘[Yatchan] (daughter’s nickname) XXX,’

39 D:

[shitsurei] shicha:u! ‘[You have no manners]!’

40 M: ... shitteru to omowanakatta wa. ‘... I didn’t think that you knew him.’ 41 D: ... (3.0) kanji waru:i. ‘... (3.0) Rude.’ ((After a one-second pause, M initiates a new topic.)) Up until line 28, mother and daughter seem to have a nice conversation by confirming that they both know the man who died and they both share the same sense of surprise. This is started by the mother’s utterance in line 4, shitteru desho? ‘you know who he is, don’t you?’. Unlike a question sentence such as shitteru? ‘do you know who he is?’, the tag question in line 4 shows that the mother presumes that she and her daughter share common knowledge about the man. This assumption is clearly accepted by the daughter’s reply, un ‘yes’, in line 5. Then, the mother instantly reconfirms the agreement by using an agreement-seeking particle ne: ‘right’ in line 6.9 The agreement-seeking particle ne is repeatedly used by both speakers (in lines 10 and 26 by the mother, and line 16 by the daughter), emphasizing the sharedness of their surprise. An agreement marker un ‘yes’ in line 3 by the mother and in line 5 by the daughter also helps to create a smooth conversation in this short segment. With these linguistic features, they have created a shared atmosphere – they both know the man (it would be more than his name and/or face – e.g., “how young he was,” “how healthy he looked,” etc.) and they both agree on how surprising it is to learn of his death. But in line 30, the mother suddenly changes the atmosphere by saying, anata mo shitteru nante omowanakatta wa ‘I didn’t think that you would know him, too’. This statement in line 30 is quite a big shift

9. Lebra (1976) defines the sentence-final particle ne as a marker that solicits the addressee’s agreement. Cook (1992) fundamentally supports this perspective and defines ne as an index that requires the addressee’s cooperation (e.g., requesting confirmation). Either way, I argue that ne helps the participants in this excerpt to create a shared atmosphere.



The power of femininity 

after the sharedness of the conversation; therefore, it leads to the daughter’s surprise marker e? ‘what?’ in line 31, said with a rising tone. Line 30 can be interpreted in both positive and negative ways for the daughter. A positive interpretation would be something like, “I didn’t think that he was famous enough that a young person like you would know of him.” A negative interpretation would be something like, “I didn’t think that you were sophisticated enough to know about a man like him.” Although we cannot confirm the mother’s intention based on the language used, the daughter seems to have taken it negatively. She recognizes the man as a popular person on TV (see lines 32 and 33); therefore, not knowing who he is could be taken as a sign of ignorance. After recovering from her surprise, the daughter raises her voice and claims that she knows that much (shitteru wa yo: sore gura:i ‘of course I know that much’) in line 36. The “super-feminine” form wa yo is used by the daughter to make a claim for the first and the last time in the data. According to her unmarked neutral gender style, we might expect this utterance to instead be “neutral” (e.g., shitteru sore gura:i; shitteru mon sore gura:i), or at least “masculine” (e.g., shitteru yo sore gura:i) or “feminine” (e.g., shitte­ rutte sore gura:i). It is hard to believe that the daughter specifically chooses the “superfeminine” form for the utterance merely to express a very feminine gender identity. The interpretation I suggest here is that the daughter shows her extreme displeasure by switching style to the “super-feminine” form. During the 70-minute conversation, a few disagreements or small conflicts are observed. In all instances except this one, the daughter started the conflicts by making various claims, such as that her mother is careless, lacks accurate knowledge of the topic under discussion, or makes contradictory statements. In other words, the daughter threatens the mother’s face (Brown and Levinson 1987), but not vice versa. In these cases, the daughter keeps her unmarked “neutral” form, while the mother switches her speech style from casual forms to polite forms.10 This time, however, the daughter’s face is threatened for the first time, and she finds it necessary to defend herself. Linguistically, there are two pieces of evidence that suggest why the mother’s statement in line 30, ‘I didn’t think that you would know him, too,’ may sound threatening and trigger an attempt by the daughter to defend herself. First, the mother uses the “super-feminine” sentence-final particle wa, which can be understood as a mild emphasis/assertion marker,11 to present a statement in line 30. In the 70-minute 10. This is indeed an interesting speech style shift, but one which is beyond the scope of this paper. 11. The sentence-final particle wa with a rising tone is always categorized as a “super-feminine” form. However, researchers have not yet come to an agreement about its pragmatic interpretation. For instance, Matsumoto (2001: 342) states that wa conveys a delicate and attenuated assertion, while Kitagawa (1977: 287) claims that wa denotes a strong sense of insistence and a “gentle-question intonation” that is a source of femininity. McGloin (1991: 31–33) disagrees with Kitagawa and argues that wa expresses strong emotional feelings and can be used as “positive politeness” (Brown and Levinson 1987). On the other hand, Okamoto and Sato (1992) de-

 Yuka Matsugu

conversation with her daughter, the mother also uses “neutral” forms as her unmarked gender style, but at the same time, it is clear that the mother tends to use more “feminine” or “super-feminine” forms than does her daughter.12 Consequently, the mother’s “super-feminine” forms might not stand out the way her daughter’s usage does. However, many of her “super-feminine” forms are used when she is expressing opinions. The mother uses “super-feminine” forms 18 times besides the 3 instances in excerpt (1). Within the 18 cases, the mother uses “super-feminine” forms 8 times to defend herself in conflict situations. Adding in two other cases in which the mother tries to persuade her daughter of something, over half of the “super-feminine” forms used by the mother present somewhat strong assertions. Considering this tendency, it is conceivable that the daughter takes the mother’s statement with a “super-feminine” particle in line 30 as too opinionated, or even as an attack on her. Along with the mother’s “super-feminine” form usage, the mother’s use of a second-person reference form, anata ‘you’, in line 30 may make the daughter believe that her face is in danger. It has been reported that Japanese people use more 1st-person and 2nd-person pronouns in the subject position in conflict conversations (Jones 1993).13 The daughter may understand the use of these linguistic features as a face-threatening act. If so, the daughter might feel that after a smooth interaction (from lines 1 to 28), her mother has suddenly and strongly asserted a contradictory perspective in line 30. This may motivate her to defend herself by showing her strong displeasure. The daughter’s defense against the mother’s statement in line 30 is rather aggressive. Prosodically, the daughter shows her displeasure in lines 36, 39, and 41 with higher pitch and lengthening. High pitch and lengthening may at first appear to be cues that the daughter is whining and acting like a weak, sweet girl. However, her stolid facial fine wa as a mild emphasis marker, and Uyeno (1971: 69) defines it as a mild insistence marker. The only common ground among these scholars seems to be that wa is a feminine way of expressing one’s feelings, thoughts and/or assertions explicitly. 12. The following is the distribution of gender form usage by the mother: “super-feminine” forms, 29 times (8%); “feminine” forms, 48 times (14%); “neutral” forms, 230 times (67%); and “masculine” forms, 38 times (11%). There are no “super-masculine” forms used by the mother in the data set. Within 29 cases of the “super-feminine” forms usage, kashira (I wonder), a self-/ other-addressed question marker, is used 8 times. Linguistically, the mother has an alternative choice, kana (neutral). However, she exclusively uses kana to question matters that only she may know about, and kashira to question about the daughter or general matters. Therefore, I consider that the mother personally does not have alternative forms to ask her daughter questions besides kashira. As a result, I regard the remaining 21 cases as the cases in which the mother chooses to use the “super-feminine” forms. 13. In the 70 minutes of data, the mother uses 2nd-person reference forms 6 times in subject position: anata ‘you’, 2 times; Yatchan ‘the daughter’s nickname’, 3 times; and oneesan ‘older sister’, 1 time. Two out of the six cases are used in the short conflict segment in excerpt (1): in line 30 (anata: ‘you’) and in line 38 ‘(Yatchan): nickname’. It should be noted that in Japanese a kinship term such as oneesan ‘older sister’ is often used to refer to a person in that position.



The power of femininity 

expression and lack of laughter do not support this interpretation. Moreover, she states her displeasure explicitly in line 39, shitsurei shicha:u! ‘you have no manners!’,14 and in line 41, kanji waru:i. ‘rude’. What is most interesting here is that the daughter chooses to use “women’s language” to show her displeasure and strike back at her mother. In this section, we have sought an explanation for the daughter’s sole use of a “super-feminine” form in the 70-minute data set. This excerpt has distinct characteristics compared with other parts of the same data – this is the only instance in which the daughter’s face is threatened by her mother. The daughter could protect her face either by fighting back against her mother or conversely by showing her subordinate position, thus convincing her mother not to attack her.15 If we follow the traditional understanding of “women’s language,” we would take the latter interpretation and say that the daughter mitigates her expression of displeasure. However, the daughter’s displeasure is made clear not only covertly but also overtly, as discussed above. From her overt expressions such as ‘you have no manners’ (line 39) and ‘rude’ (line 41), it is hard to believe that the daughter anticipates appeasing her mother by being soft and polite. Therefore, defining “super-feminine” forms as simple softness/politeness/powerlessness markers for a woman to present herself as subordinate seems to be a misinterpretation. Far from that, it seems as if, by showing her “womanhood” – adulthood as a woman – with the “super-feminine” forms, the daughter can shift her status from that of a daughter to that of a woman – of equal status to her mother – and express her displeasure towards her mother. At any rate, the “super-feminine” form in this excerpt somehow functions as a sort of power marker and helps the daughter to effectively express and assert herself in this mother-daughter conversation. The topic is dropped shortly after the use of the “super-feminine” form in line 36, and after a one-second pause, the mother initiates a new topic about the foods on their dinner table.

3.2

A mixed-sex friends’ conversation

This same relationship among “super-feminine” forms, “womanhood,” and “power” can also observed in a mixed-sex conversation. The next excerpt shows a “super-feminine” form usage in a quoted sentence that presents a female speaker’s voice in a hy14. On various occasions, native speakers of Japanese have commented to me that the phrase shitsurei shichau is exclusively used by female speakers; therefore, it should be categorized as a “super-feminine” form. Intuitively, I agree with this suggestion. However, to be as objective as possible, I only count forms which have been proposed as such in previous research. Nevertheless, the idea that shitsurei shichau counts as a “super-feminine” form further supports my proposition. 15. Rosenberger (1994) proposes that in Japan, a person of superior status has a responsibility to protect and take care of his/her subordinates. Thus, by showing one’s subordinate position linguistically, one may expect one’s superior to accept one’s imperfect and/or unintelligent behavior. That is, in this case, by switching to the “super-feminine” form, which may express one’s powerlessness, the daughter may expect her mother to accept her ignorance and to stop threatening her face.

 Yuka Matsugu

pothetical situation. It is from a conversation in which a female speaker, Speaker A (aged 27), was talking to two male friends (aged 30 and 28, respectively) about an experience she had with some high school girls. Speaker A and a female friend went to a store that sells photographs of popular singers. While they were waiting for the cashier, a high school girl behind them looked at them and whispered to another high school girl, nenpai no hito mo kiteru n da ‘older people are here, too’. Speaker A and her friend heard it, but did not say anything to the girls. The way in which Speaker A delivers the story is not serious, but is rather quite comical. Throughout the story, all the speakers laugh a lot. After the story, as excerpt (2) shows, Speaker B, a male speaker, suggests that Speaker A should have reacted to the high school girls. His suggestion for her was to show her displeasure rather rudely and quite directly to the girls by turning around and glaring at them (lines 1 to 3). The action might be taken as childish if Speaker A had actually gone through with it. However, in the humorous atmosphere, she favors the suggestion. Responding to it, she says a phrase in quoted form as if she would utter it to put the girls in their place. The phrase, ‘We have been living longer than you guys though, do you have a problem with that?’ (lines 5, 6, and 10–12),16 ends with the “super-feminine” form no yo. (2) (three friends, quoting what a female speaker could have said) 1 B: ... soko de nani.. dame da yo. ‘... Then what?.. no way.' 2

.. furikaette, ‘.. (You should do something like) turn around and’

3

tsukisasu yoona me <@ de niramitsukeru toka [sa @>]. ‘with sharp eyes <@ glare at them @>.’

4 A:

[soo] soo soo. ‘Yes, yes, yes.’

5

... ? ‘... ?’

6

.. ttoka. ‘.. something like.’

7

@[@@@@]

8 B:

[@@@@]

9 C:

[@@@@]

16. Bernstein (1991: 3) proposes that Confucian teachings about filial piety give older women a full measure of respect. Therefore, it may be safe to believe that speaker A links being older with unquestionable authority in the Japanese sociocultural hierarchy, which gives her the right to do what she wants to do (e.g., in this case, to be at a store that primarily targets young customers). This point will be revisited in the next section.



The power of femininity 

10 A:

...
11

... atashitachi no hoo ga, ‘... we have been,’

12→ ... nagaku ikiten no yo Q>.17 ‘... living longer (nevertheless, do you have a problem with that?).’ 13 B:

@[@]

14 C:

[@]@

Does Speaker A use a “super-feminine” form to soften and/or weaken her imagined statement? Is it necessary for her to present herself as a polite woman in this conversation? There is some evidence to suggest that Speaker A does not need to be polite to either the supposed addressees of the quoted sentence or to her addressees in the ongoing conversation. Obviously, the addressees of the quoted sentence, the high school girls, are not present in the ongoing conversation, and there will not be a second chance for Speaker A to say anything to those girls in the future. That is, Speaker A does not need to put too much effort into mitigating her statement or being nice to the girls in the quoted phrase. In addition, there is no need for Speaker A to worry about her present addressees’ reactions, either. From the previous part of the conversation, it is clear that all of these speakers consider themselves to be in the same generation and feel the same generation gap with high school students. In addition, as we see in excerpt (2), Speaker B actually encourages Speaker A to be explicitly rude. These points suggest that Speaker A does not need to present herself as a weak, passive, polite woman by using “women’s language” in this excerpt. Despite the fact that Speaker A does not need to be soft and powerless, a “superfeminine” form no yo is used here. This instance even looks as if the “super-feminine” form no yo exaggerates her “age” – as an adult woman – and therefore exaggerates her power, which allows her to be explicit and authoritative. It should be noted that Speaker A’s unmarked gender form is “neutral,” and she uses “super-feminine” forms only three times in the conversation.18 Two of the three instances, including the case in excerpt (2), are used to present her as an older woman and to support her right to do what she wants to do. Both of those cases of a “super-feminine” form are followed by laughter from the participants. The conversations then move on to a different topic just as in excerpt (1), as if there is nothing to negotiate on the statement anymore.

17. Neutral counterparts of this utterance would be nagaku ikiteru mon and nagaku ikiteru. 18. In 7 minutes, Speaker A uses “super-feminine” forms 3 times (6%), “feminine” forms 13 times (27%), “neutral” forms 31 times (63%), and “masculine” forms 2 times (4%).

 Yuka Matsugu

3.3

The power of womanhood

In both excerpts (1) and (2), speakers do not have any particular reason to soften their assertions but, rather, have good reasons to be more aggressive and assertive. Moreover, they have the choice not to use “super-feminine” forms. However, they use those forms in a particular situation – a case in which being seen to be older might give the speaker the power to speak as she wishes. In this section, I explore a sociocultural interpretation of the connection between “super-feminine” forms and power. As I mentioned in the introduction, various approaches have been proposed in order to understand gender variation in language. The differences among these theories are explicitly taken up by other scholars, yet there are also similarities between the theories that are often overlooked. One of the similarities is the idea of dualism, in which “women and men” are categorically contrasted to each other. In this binary opposing view of gender, “women’s language” is understood to signify a set of fixed meanings: femininity, softness, and powerlessness. To some extent, this tendency is true for the study of Japanese gender variations, and Japanese “women’s language” is understood to express hegemonic feminine identities (e.g., Ide 1982). However, Japanese society has an additional social meaning for “being women,” which overlaps the binary opposition of women and men, and that is “being adult.” Bernstein (1991) elucidates that in a Confucian society like Japan, both women and men have to go through several sociocultural stages in their lives. These stages include education, work, marriage, and so on. Each stage gives a person a different identity within his/her respective gender group, as well as in society. Yet, the most important stage for both women and men is “to be an adult.” In other words, “being a woman” is not only contrasted with “being a man,” but is also contrasted with “being a child/immature.” This latter dimension connects “being women” with different meanings, such as authority and power. This bi-dimensional view of Japanese gender is illustrated in Figure 2. (low)

Adulthood/ Maturity

Childhood/ Immaturity

Sociocultural Power

(high)

F emale

Mal e

Woman (Becoming one of the oldest members of society) (Having a child and a house) (Getting married) (Getting a job) Girl (Getting an education) (Before getting an education)

M an (Becoming one of the oldest members of society) (Having a child and a house) (Getting married) (Getting a job) B oy (Getting an education) (Before getting an education)

Figure 2.  Bi-Dimensional View of Japanese Gender

(low)



The power of femininity 

In the case of excerpt (1), despite the fact that the daughter is in her 20s, she has a definite lower social status as the child in a mother-daughter relationship. However, by using a “super-feminine” form, she can implicitly identify herself as an adult and ask her mother to treat her appropriately. On the contrary, Speaker A in excerpt (2) is given the role of an older person by the high school girls. Speaker A performs the role by using the “super-feminine” form, which is not her unmarked gender style. By doing so, she can claim a higher social status, along with her power to assert her right to speak as she wishes. The binary opposition of adulthood and childhood is not mutually exclusive of the binary opposition of women and men. In this sense, gender style inclusively signifies multiple social meanings. This is how “super-feminine” forms can signify femininity as well as womanhood and can express somewhat contradictory meanings: a powerless identity and a powerful one.

4. “Super-feminine” forms, femininities, and power 4.1

Tea ceremony lessons

In some special situations, the authoritative power of being women may be culturally supported. In this section, I examine some excerpts from sadoo (tea ceremony) lessons, where normative images of femininity are highly valued. As a result, I propose that the style “women’s language” is not simply opposed to “men’s and/or girls’ languages,” but rather is highly context-dependent and complex. The conversation I discuss below captures a sequence of tea ceremony lessons. The conversation is 110 minutes long and was videotaped in Tokyo in 1999. For this study, the first 45 minutes (before a male student arrives) is examined. There is a female instructor with four female students who take turns serving tea. The instructor is a 61-year-old female, who spent her childhood in Fukushima (the northern part of Japan) and moved to Tokyo at age 18. She studied tea ceremony and became a professional instructor. The four students are aged 70, 68, 60, and 55, respectively. Two were born and grew up in Tokyo. The other two were born in different prefectures (Fukushima and Ibaraki) but have now lived in Tokyo a fairly long time (28 years and 39 years). All of the participants in this data set are speaking so-called Standard Japanese. Although their ages vary, the social hierarchy between the instructor and her students is significant and stable. Therefore, theoretically, as a person of a higher social status, the instructor does not have to use “polite” forms and/or “feminine” forms in order to be less assertive. Despite this fact, the instructor shows a distinct and interesting style shift from “neutral” gender forms to “super-feminine” forms when she expresses displeasure to her students. During the tea ceremony lesson, the students take turns being the host, who serves tea, and the guest, who drinks tea. Both the host and the guest, but especially the host,

 Yuka Matsugu

have to follow strict rules that are highly ritualistic. Throughout the lesson, the instructor delivers short oral instructions that hint at the next step. Typical instructional phrases end with nouns (“neutral” forms) or the te-form of verbs (“feminine” forms), as can be seen in excerpt (3). (3) (Instructor I and Student S) 1 I: ... kiwa ni, ‘... On the edge,’ 2

... soo soo soo, ‘... Right, right, right,’

3

... hai omote, ‘... Yes, front,’

4

de ochawan hiite, ‘Then pull the bowl,’

5

... hai. ‘... Yes.’

6

... karioki migi. ‘... Put it to the right.’

These instructions are not given because the instructor doubts the student’s competence. The student’s task in the tea ceremony lesson is to perform smooth, beautiful body movements with no hesitation or unnecessary pauses. Thus, it is common for the instructor to provide minimal instruction so that the students need not worry about recalling the next step, but can instead focus on various aspects of their movements, such as the angle of arms to fingers, the position of eye gaze, the speed of each motion, and the timing. However, the students were experiencing difficulties on the day of the recording. Because they had been busy attending several real tea ceremonies, they had not had lessons for over two weeks. In other words, they had been guests at several tea ceremonies but had not been hosts for a while, and they had forgotten how to properly make and serve tea. Because of this fact, the instructor had to give additional instructions to the students in the host role. In such cases, the instructor often showed her frustration by raising her voice and using “super-feminine” forms.19 In excerpt (4), one student is having a hard time whipping the tea powder with hot water. She grasps the tea bowl from the side instead of pushing the edge of the bowl (where you put your mouth to drink) from the top; therefore the bowl moves around 19. It is impossible to measure the level of the instructor’s frustration linguistically; consequently, it is impossible to count the occurrences of her frustration. As a result, I cannot provide numbers to show a strong relationship between her frustration and the “super-feminine” form. However, it should be noted that although not all repetitive instructions are given with the “super-feminine” forms, all “super-feminine” forms by the instructor are used when she needs to give additional guidance to her students.



The power of femininity 

along with her whipping motion. This is a common mistake that can ruin both an expensive tea bowl and a lacquered desk by allowing them to rub against each other. The instructor gives the student instructions with a relatively loud voice. (4) (Instructor I and Student S) ((I has watched S’s movement carefully.)) 1→ I: ... ue kara shikkari osaeru no yo! ‘... Push (it) down from the top!’ 2

chawan o, ‘the bowl,’

3

... gu tto osaete, ‘... Push hard,’

4

... soo suru to ugokanai kara, ‘... if you do so, it won’t move,’

5

... omoikitte osaet- -‘... Push harder --’

6→

chikara iru no yo! ‘You need strength!’

7

kekkoo. ‘quite.’

8

hirajawan tte iu no wa. ‘especially with a hira-bowl (a type of bowl).’

9 S:

... soo desu ne. ‘... Yes, you’re right.’

10 I:

motsu to yurayura shichau kara, ‘If you grasp it, it moves, so,’

11→

ue kara gyu tto yatta tsumori de ya n no yo. ‘you have to do it so as to push it down from the top.’

12 S:

... muzukashii desu ne. ‘... It’s difficult.’

13 I:

... (2.0) tekubi tsukai nagara, ... (2.0) ‘Use your wrist (when you whip tea),’

14

... tekubi, ‘... Wrist,’

15

... ya- -‘... A- --’

 Yuka Matsugu

16

... yukkuri de ii kara, ‘... It’s O.K. to slow down,’

17

... tekubi koo, ‘... Wrist, like this,’

18

.. koo, ‘.. Like this,’

19

... koo, ‘... Like this,’

20

koo koo, ‘Like this, like this,’

Until the time when another step starts in line 13, the instructor repeatedly stresses that the student should press the tea bowl from the top toward the desk firmly, so the bowl will not move (from lines 1 to 11, except line 9). In this short segment, the instructor uses a “super-feminine” form, no yo, extensively (in lines 1, 6, and 11), even though she could have used her unmarked instruction form (e.g., ue kara shikkari osaete) for all cases. Differences in the instructor’s attitude between the first half of this excerpt (from lines 1 to 12) and the second half (from lines 13 to 20) are significant. She speaks louder and faster in the first half compared with the second half. She is also leaning toward the student until the short pause in line 13. Thus, although there is no overt expression of frustration other than prosodic indications, the emotional difference is obvious before and after the short pause in line 13. The instructor seems to be frustrated in the first half. With the utterances she makes using “super-feminine” forms, she strongly asserts what the student is supposed to do. The instructor uses “super-feminine” forms extensively in the same sense: when she really wants a student to get a point, she uses a “super-feminine” form in the instruction. Excerpt (5) is another good example of this usage. (5) (Instructor I and Student S) 1 I: ... hai yoko, ‘... Then, side,' 2

... hai tochuu made. ‘... Then, to the middle of it.’

3

... zu:tto nesete, ‘... Gradually slide,’

4

.. ikki ni, ‘.. In one movement,’

((S fails to perform the movement correctly.)) 5

.. hitokokyuu de yatte goran. ‘.. Make it in one movement.’



The power of femininity 

6

.. chotto, ‘.. Wait,’

7

moo ikkai yatte. ‘Do it once again.’

8 S:

.. hai, ‘.. Yes,’

9 I:

.. hai. ‘.. yes.’

10

tairana toko kara, ‘From the flat,’

11

.. ano saisho kara, ‘.. Well, from the beginning,’

((S cannot figure out from what point she should start again.)) 12 S:

a koo? ‘Like this?’

13 I:

.. saisho kkara hai, ‘.. From the beginning, yes,’

((S repeats the movement but still does not succeed in doing it correctly.)) 14→ hitokokyuu yo! ‘In one movement!’ 15

ikkai ga. ‘One sequence.’

16 S:

hai, ‘Yes,’

17→I:

i:chi tte iku no yo! ‘Move like “one”!’

18

.. soo. ‘.. Yes,’

19

.. hai yoko, ‘.. Then side,’

20

.. zu: tto, ‘.. Smoothly,’

21

... hai nigitte, ‘... Then grab it,’

22

.. zu: tto itte sa: tto nesechau. ‘.. Smoothly move and slide.’

 Yuka Matsugu

23→

hitokokyuu yo! ‘In one movement!’

((S repeats the movement again.)) 24 S:

.. hai. ‘.. Yes.’

25 I:

moo ikkai. ‘Once again.’

In this excerpt, the instructor is first giving instructions as usual by using nouns (sometimes followed by particles) and the te-form of verbs (from lines 1 to 4). However, noticing the student’s non-proficient movement, she requests that the student redo that part of the tea ceremony. A clear instruction is given as a “neutral” form in line 5, hitokokyuu de yatte goran ‘make it in one movement’, to let the student know what was wrong with her technique. Although the student restarts the same part in line 12, the student again fails to do the movement properly. The instructor then switches from “neutral” forms to “super-feminine” forms (in lines 14, 17, and 23). It should be noted that all three “super-feminine” forms mark similar instructions for how to carry out a movement that the student repeatedly fails to perform successfully. In this section, we observed a typical pattern in which the instructor switches her gender variation from “neutral/feminine” forms to “super-feminine” forms and vice versa, and discussed the idea that “super-feminine” forms are used to assert a point which a student repeatedly fails to comprehend. When the instructor is frustrated with her students’ continued failures, she raises her voice, talks faster, and uses “super-feminine” forms repeatedly.

4.2

The power of femininity

Sadoo, the practice of tea ceremony, is not practiced in daily life in modern Japanese society. Since it is an expensive hobby, however, it is identified as a sign of wealth, cultural sophistication, and even mental or moral sophistication. Especially for women, the tea ceremony suggests good family background and a pleasant personality.20 This understanding of sadoo in society most likely comes from an expected relationship 20. My understanding of tea ceremony in modern Japanese society comes from my experience in the personnel section of a Japanese company – in particular, how we saw the personality of female applicants based on information in their curriculum vitae. For example, many male administrators in this company would explicitly state that female workers who had practiced tea ceremony should not be assigned positions in a construction department, where they would have to deal with subcontractors who are normatively believed to be rough. Rather, the administrators suggested that those female workers should be assigned positions in the personnel section or the sales section, where they might interact with people in higher positions, such as company presidents and department managers.



The power of femininity 

between the ideal woman and the image of sadoo as soft, smooth, calm, and polite. People who practice sadoo are taught to make “perfect movements” – soft, smooth, calm, and polite movements – in tea ceremony. This does not go along with the normative image of masculinity, but does fit the normative image of femininity. After all, it places high value on a particular aspect of “womanhood”: being feminine. Bourdieu (1977) claims that the functions of “women’s language” are influenced by the local organization of gender roles, rights, and expectations. Hence, it may be rather natural to think that “women’s language” can have a sense of power in a local situation like sadoo, in which “being feminine” is highly valued. This is why a woman in an authoritative position, who is not powerless and does not need to appear powerless, may use “super-feminine” forms to express her assertions to her subordinates in naturally occurring conversations. In these cases, “women’s language” does not represent weakness and subordinate status, but rather represents power and superiority.

5. Contextualization for another dimension I have discussed the multidimensionality of “super-feminine” forms as a style and proposed that “super-feminine” forms can signify somewhat contradictory realities: powerless identity and powerful identity. However, each of these social meanings signified by the “super-feminine” form has to be interpretable not only by a speaker, but also by the other participants in the conversation. In this section, I discuss the contextualization cues in discourse through which the powerful identity seems to be signified as the most salient identity in a given context. One of the key points is that the speakers’ unmarked gender variation is not feminine. Therefore, what appears to be a shift of gender style to “super-feminine” forms might actually represent a subtle shift in meaning. In other words, style shift to “superfeminine” forms allows a speaker to preserve the social gender persona implied by her unmarked gender variation and suggests that the social meaning signified by the “super-feminine” form is not her gender persona, but rather represents another social persona, such as that of an adult. Another important point is that the use of “superfeminine” forms as a power marker is only observed among female participants. This does not mean that male participants as members of Japanese society do not recognize this linguistic system. For instance, excerpt (2) shows that the female speaker uses a “super-feminine” form to express her authoritative power to the imagined female addressees in a hypothetical situation. However, male participants in the ongoing conversation seem to have no problem understanding this non-normative use of “women’s language.” They laugh and collaboratively create the female speaker’s powerful persona in the hypothetical situation by using a “super-feminine” form. One may also see how situations of conflict or opposition seem to bring out the elevation, so to speak, of the power of the younger woman by invoking an older age identity. Situations of conflict also seem to bring out the elevation of the power of the woman in a superior position

 Yuka Matsugu

by invoking a socioculturally valued identity in the particular speech community. These points suggest that the multidimensionality of Japanese “super-feminine” forms is tacitly recognized by society and is used fairly systematically to identify a speaker’s multiple identities.

6. Conclusion The binary oppositions traditionally associated with masculinity and femininity are salient and highly pertinent to any discussion of gendered behavior (Johnson 1997: 23). However, they fail to sufficiently capture every aspect of gender and language use. Gender identities are socially constructed and highly contextualized, hence fluid and variable (Johnson 1997: 19). Therefore, gendered styles, which may signify speakers’ gender identities, can also express other social meanings that index each gender group in a given society or a situation. It has been said that various so-called “feminine” forms in the Japanese language reflect women’s position in Japanese society (Mizutani and Mizutani 1987, Reynolds 1985, Shibamoto 1985); unlike their masculine counterpart forms, which can present strong assertion and will, “feminine” forms are believed to represent softness, weakness, politeness, and powerlessness. However, this study argues that by shifting to “super-feminine” forms, female speakers can present a powerful persona, displaying emotions such as strong displeasure or authoritarian frustration. In these cases, “womanhood” is no longer a symbol of powerlessness, softness, and/or politeness, but the powerfulness, strength, and authority that an adult woman can hold. Ide and Inoue (1992) state that women present their authority by displaying their abilities to use “polite/formal language.” Along the same line, Takekuro (2002) compares two sets of conversations by the same speakers – one recorded when the speakers were university students, and the other recorded after the same speakers each established a certain status in society – and proposes that women use more feminine forms and honorifics to present their adulthood. These statements support my proposal: women may be able to show their adulthood, and consequently their authority, by showing the ability to use “women’s language” properly. The association between gender variations and their relative power has been one of the essential questions in the study of gender and language. Much work has claimed that women’s language and powerless language are different, but are correlated because most women are in powerless positions in society (Eckert and McConnell-Ginet 1992, Inoue 1994, Maltz and Borker 1982, O’Barr and Atkins 1998, Okamoto 1995, Okamoto and Sato 1992, Quinn 1977, Reynolds 1998: 303, Tannen 1990, Uchida 1993, West 1984: 18). In any case, the language which is predominantly used by women has been recognized as a symbol of powerlessness in many languages, including Japanese. While this paper continues the dialogue regarding gender variation and the expression of power, it raises an important question. Particular linguistic features may be used for certain purposes and effects in conversation, but it cannot be said that one feature al-



The power of femininity 

ways and unambiguously has the same function (Graddol and Swann 1989: 74). That is, the meaning of “being women” will vary at any given time. Although it may at times be related to weakness and softness, this is not always the case. The power of “femininity,” which we have observed in this paper, suggests that each linguistic form, as well as sociolinguistic variation, might covertly convey multiple social meanings. With the results in this paper, I urge a different perspective on the study of gender and language – a multidimensional, context-based examination of the function of gendered forms.

References Aoki, H., Miyashita, M., Ono, T. and Sadler, M. To appear. A Corpus of Conversational Japanese: Arizona Papers in Japanese Linguistics. Tucson, AZ: University of Arizona Press. Bernstein, G. L. 1991. “Introduction.” In Recreating Japanese Women, 1600–1945, G. L. Bernstein (ed.), 1–14. Berkeley, CA: University of California Press. Bodine, A. 1975. “Sex differentiation in language.” In Language and Sex: Difference and Dominance, B. Thorne and N. Henley (eds.), 130–151. Rowley, MA: Newbury House Publishers. Bourdieu, P. 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press. Brown, P. 1980. “How and why are women more polite: Some evidence from a Mayan community.” In Women and Language in Literature and Society, S. McConnell-Ginet, R. Borker and N. Furman (eds.), 111–136. Praeger; NY: Greenwood Publishing Group. Brown, P. and Levinson, S. C. 1987. Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Cook, H. M. 1992. “Meaning of non-referential indexes: A case study of the Japanese sentencefinal particle ne.” Text 12(4): 507–539. Crosby, F. and Nyquist, L. 1977. “The female register: An empirical story of Lakoff ’s hypotheses.” Language in Society 6: 313–322. Du Bois, J. W., Cumming, S., Schuetze-Coburn, S. and Paolino, D. 1992. Discourse Transcription: Santa Barbara Papers in Linguistics, Vol. 4. Santa Barbara, CA: Department of Linguistics at University of California. Eckert, P. and McConnell-Ginet, S. 1992. “Communities of practice: Where language, gender, and power all live.” In Locating Power: Proceedings of the Second Berkeley Women and Language Conference, K. Hall, M. Bucholtz and B. Moonwomon (eds.), 89–99. Berkeley, CA: Berkeley Women and Language Group, University of California. Galbraith, J. K. 1983. The Anatomy of Power. Boston, MA: Houghton Mifflin. Graddol, D. and Swann, J. 1989. Gender Voices. New York, NY: Blackwell Publishers. Horii, R. 1993. “Joseigo no seiritsu [Establishment of women’s language].” Nihongogaku 12(5): 100–108. Ide, S. 1982. “Japanese sociolinguistics: Politeness and women’s language.” Lingua 57: 387–385. Ide, S. 1990. “How and why do women speak more politely in Japanese?” In Aspects of Japanese Women’s Language, S. Ide and N. H. McGloin (eds.), 63–79. Tokyo, Japan: Kuroshio. Ide, S. and Inoue, M. 1992. “Josei kotoba ni miru aidentitī – shakai no josei no baai – [Identities in women’s language – A case study of women’s language in society – ].” Gengo 21(10): 46–47.

 Yuka Matsugu Inoue, M. 1994. “Gender and linguistic modernization: Historicizing Japanese women’s language.” In Cultural Performances: Proceedings of the Third Berkeley Women and Language Conference, M. Bucholtz, A. C. Liang, L. A. Sutton and C. Hines (eds.), 322–333. Berkeley, CA: Berkeley Women and Language Group, University of California. Irvine, J. T. 2001. “‘Style’ as distinctiveness: The culture and ideology of linguistic differentiation.” In Style and Sociolinguistic Variation, P. Eckert and J. R. Rickford (eds.), 21–43. Cambridge: Cambridge University Press. Johnson, S. 1997. “Theorizing language and masculinity: A feminist perspective.” In Language and Masculinity, S. Johnson and U. H. Meinhof (eds.), 8–26. Oxford, United Kingdom: Blackwell Publishers. Jones, K. 1993. “Nihongo no danwa ni okeru iken no shōtotsu to ninshōshijishi no yōhō [Conflict and use of personal pronouns in Japanese discourse].” Paper given at the National Language Research Institute, 3rd International Symposium: On the Methods of Discourse Analysis. Jorden, E. with Noda, M. 1987. Japanese: The Spoken Language Part 1. New Haven, CT: Yale University Press. Keenan, E. O. 1974. “Norm-makers, norm-breakers: Uses of speech by men and women in a Malagasy community.” In Explorations in the Ethnography of Speaking, R. Bauman and J. Sherzer (eds.), 125–143. Cambridge: Cambridge University Press. Kitagawa, C. 1977. “A source of femininity in Japanese: In defense of Robin Lakoff ’s ‘Language and Woman’s Place’.” Papers in Linguistics 10(3–4): 275–298. Konoshima, M. 1966. Kokugo Joshi no Kenkyū – Joshishi no Sobyō [Study of Japanese Particles – Outline of the History of Particles]. Tokyo, Japan: Ourinsha. Kulick, D. 1992. “Anger, gender, language shift and the politics of revelation in a Papua New Guinean village.” Pragmatics 2(3): 281–296. Labov, W. 1966. The Social Stratification of English in New York City. Washington, D.C.: Center for Applied Linguistics. Lakoff, R. 2003. “Language, gender, and politics: Putting “women” and “power” in the same sentence.” In The Handbook of Language and Gender, J. Holmes and M. Meyerhoff (eds.), 161–178. Oxford, United Kingdom: Blackwell Publishers. Lebra, T. S. 1976. Japanese Patterns of Behavior. Honolulu, Hawaii: University of Hawaii Press. Lebra, T. S. 1985. Japanese Women: Constraint and Fulfillment. Honolulu, Hawaii: University of Hawaii Press. Maltz, D. N. and Borker, R. A. 1982. “A cultural approach to male-female miscommunication.” In Language and Social Identity, J. J. Gumperz (ed.), 196–216. New York, NY: Cambridge University Press. Matsumoto, Y. 2001. “Gender identity and the presentation of self in Japanese.” In Gendered Practices in Language, S. Benor, M. Rose, D. Sharma, J. Sweetland and Q. Zhang (eds.), 339–354. Palo Alto, CA: Center for the Study of Language and Information Publications. McConnell-Ginet, S. 1988. “Language and gender.” In Linguistics: The Cambridge Survey; Vol. 4., Language: The Socio-Cultural Context, F. J. Newmeyer (ed.), 75–99. Cambridge: Cambridge University Press. McGloin, N. H. 1991. “Sex difference and sentence-final particles.” In Aspects of Japanese Women’s Language, S. Ide and N. H. McGloin (eds.), 23–41. Tokyo, Japan: Kuroshio. Mizutani, O. and Mizutani, N. 1987. How To Be Polite in Japanese. Tokyo, Japan: The Japan Times.



The power of femininity  O’Barr, W. M. and Atkins, B. K. 1998. “‘Women’s language’ or ‘powerless language’?” In Language and Gender: A Reader, J. Coates (ed.), 377–387. Malden, MA: Blackwell Publishers. Ochs, E. 1992. “Indexing Gender.” In Rethinking Context: Language as an Interactive Phenomenon, A. Duranti and C. Goodwin (eds.), 335–358. Cambridge: Cambridge University Press. Okamoto, S. 1995. “‘Tasteless’ Japanese: Less ‘feminine’ speech among young Japanese women.” In Gender Articulated: Language and the Socially Constructed Self, K. Hall and M. Bucholtz (eds.), 297–325. New York, NY: Routledge. Okamoto, S. 1996. “Indexical meaning, linguistic ideology, and Japanese women’s speech.” Berkeley Linguistics Society 22: 290–301. Okamoto, S. and Sato, S. 1992. “Less feminine speech among young Japanese females.” In Locating Power: Proceedings of the Second Berkeley Women and Language Conference, K. Hall, M. Bucholtz and B. Moonwomon (eds.), 478–488. Berkeley, CA: Berkeley Women and Language Group, University of California. Philips, S. U. 2001. “Gender ideology, cross cultural aspects.” In International Encyclopedia of the Social and Behavioral Sciences Volume 9, N. J. Smelser and P. B. Baltes (eds.), 6016–6020. Oxford, United Kingdom: Elsevier Science Ltd. Quinn, N. 1977. “Anthropological studies of women’s status.” Annual Review of Anthropology 6: 181–225. Reynolds, K. A. 1985. “Female speakers of Japanese.” Feminist Issues 5(Fall): 13–46. Reynolds, K. A. 1998. “Female speakers of Japanese in transition.” In Language and Gender: A Reader, J. Coates (ed.), 299–307. Malden, MA: Blackwell Publishers. Rosenberger, N. R. 1994. “Indexing hierarchy through Japanese gender relations.” In Situated Meaning: Inside and Outside in Japanese Self, Society, and Language, J. M. Bachnik and C. J. Quinn, Jr. (eds.), 88–112. Princeton, NJ: Princeton University Press. Sato, T. 2000. Tenbō: Gendai no Nihongo [Outlook: Today’s Japanese Language]. Tokyo, Japan: Hakuteisha. Shibamoto, J. S. 1985. Japanese Women’s Language, 58–65. Orlando, FL: Academic Press, Inc. Shibamoto, J. S. 1987. “The womanly woman: Manipulation of stereotypical and nonstereotypical features of Japanese female speech.” In Language, Gender, and Sex in Comparative Perspective, S. U. Philips, S. Steele and C. Tanz (eds.), 26–49. Cambridge: Cambridge University Press. Smith, J. S. 1992. “Women in charge: Politeness and directives in the speech of Japanese women.” Language in Society 21: 59–82. Takasaki, M. 2002. “‘Onna kotoba’ o tsukurikaeru josei no tayōna gengo kōdō [Women’s various language usages which recreate ‘women’s language’].” Gengo 31(2): 40–47. Takekuro, M. 2002. “Indexicality and socialization: Age-graded changes of young Japanese women’s speech.” In Bunka-Intāakushon-Gengo [Culture-Interaction-Language], K. Kataoka and S. Ide (eds.), 195–214. Tokyo, Japan: Hitsujishobō. Tannen, D. 1990. You Just Don’t Understand: Women and Men in Conversation. New York, NY: William Morrow. Terao, R. and Zimmerman, E. 2000. “Converging toward the interlocutor: Sentence-final forms in Japanese conversations.” Word 51(1): 41–57. Trudgill, P. 1972. “Sex, covert prestige, and linguistic change in the urban British English of Norwich.” Linguistic Society 1: 179–195. Tsujimura, N. 1996. An Introduction to Japanese Linguistics, 372–377. Cambridge, MA: Blackwell Publishers.

 Yuka Matsugu Uchida, N. 1993. “Kaiwa-kōdō ni mirareru seisa [Gender differences in conversation].” Nihongogaku 12(6): 156–168. Umehara, Y. 1989. “Joshi no kōbunteki kinō [Structural functions of particles].” In Kōza Nihongo to Nihongokyōiku 4: Nihongo no Bunpō/Buntai (Jō) [Lecture: The Japanese Language and Education 4: Japanese Grammar and Syntax I], Y. Kitahara (ed.), 321–324. Tokyo, Japan: Meijishoin. Uyeno, T. Y. 1971. A Study of Japanese Modality – A Performative Analysis of Sentence Particles. Ann Arbor, MI: University of Michigan Doctoral Dissertation. West, C. 1984. “When the doctor is a lady.” Symbolic Interaction 7(1): 87–106. Wetzel, P. 1988. “Are “powerless” communication strategies the Japanese norm?” In Language and Gender: A Reader, J. Coates (ed.), 388–395. Malden, MA: Blackwell Publishers.

Appendix Transcription Conventions (Based on Du Bois et al. 1992) Units: Intonation unit {carriage return} Truncated intonation unit - Truncated word Speakers: Speech overlap [ ]; [2 2]; [3 3] Transitional Continuity: Final . Continuing , Appeal (question/confirmation) ? Appeal (exclamation) ! Lengthening: Lengthening : Pauses: Long ...(N) Medium ... Short .. Latching = Vocal Noises: Laughter @ Quality: Quotation quality Laugh quality <@ @> Whispered <WH WH> Transcriber’s Perspective: Transcriber’s comment (( )) Indecipherable syllable X



The power of femininity 

Others: Parentheses

Bold

material in the translation not explicitly mentioned in the original utterance. “super-feminine” form

Tuning speech style and persona* Yoshiko Matsumoto Stanford University

1. Introduction This paper considers the motivations and functions of shifts between expressions indexing speakers’ varied stances that are also gendered. Examining informal conversations among Japanese middle-aged middle-class female peers, I first illustrate that, contrary to a common expectation that such speakers (especially stay-at-home mothers) speak uniformly in the normative female speech style, their speech presents stylistic variations not only across speakers but also within one speaker, even when external speech conditions do not change. The observed style variations and shifts in these conversations, therefore, are not conditioned by external factors such as change of audience and situation (e.g. Bell 1984), but they are instead internally motivated. I argue that the observed variations in style offer insights into how ordinary speakers use available resources to express their own personae and to create their own personal styles in everyday settings. In doing so, I emphasize the advantage of distinguishing (at least) two dimensions of styles that speakers may choose to take rather than analytically conflating them, although they are related to each other, for the purpose of examining the multiplicity of meanings that shape the individual styles of speakers. One dimension of style is stance-oriented, which is based on the pragmatic meanings of expressions conveying the speaker’s stance and attitude, and another dimension of style indexes social group identities such as gender. A similar distinction between gender-oriented and stance-oriented styles is suggested by Coupland (2001: 189f), namely the distinction between ‘dialect style’ (associated with social groups and identity) and ‘expressive’ or ‘attitudinal’ style. I will discuss this issue further in the following sections.

* I am thankful to the organizers of the symposium, Kimberly Jones and Tsuyoshi Ono, and their associates at the University of Arizona for the opportunity to present the paper.

 Yoshiko Matsumoto

2. Gender, stance and speech styles In the past decade, studies of language and gender that are informed by the notion of indexicality (e.g. Eckert and McConnell-Ginet 1992, Gal 1995, Ochs 1993) have questioned the assumption that the use of certain linguistic forms is a direct marking of the speaker’s gender. They have argued that the gender of the speaker is indexed by expressions through the mediation of socio-cultural beliefs about women’s and men’s appropriate linguistic behavior, and that such beliefs are affected by cultural images of femininity and masculinity. Ochs (1993) noted, for example, that the Japanese pragmatic particle wa, which directly indexes the stance of ‘delicate intensity’, is conventionally associated with, or indirectly indexes, a ‘female voice’ by virtue of an ideology in the culture that links delicacy to femininity. In other words, in a culture where an affective disposition of delicacy is viewed as part of the preferred image of women, delicate expressions become associated with female voice and can motivate women’s differential use of such forms. Under different cultural assumptions, the indexing relation could change. The fact that forms that index female voice do so indirectly through gendered assumptions about affective dispositions means that the use of such forms can at the same time be described in terms of the stance that they directly index. In other words, the ‘feminine style’ or ‘masculine style’ has a parallel description in terms of stance-oriented style (e.g. ‘delicate style’ or ‘forceful style’). Such stance-oriented styles may often be subsumed in gender-oriented styles, but the recognition of the two dimensions of styles, as with the two types of indexicality (direct indexicality and indirect indexicality), is a useful tool in investigating style shifting, as it allows us to consider the multiplicity of the meanings conveyed by expressions that also have gendered interpretations. A comparable distinction is suggested by Coupland (2001: 189) between ‘dialect style’ and ‘expressive’ or ‘attitudinal style’. Referring to the sociolinguistic tradition in which the study of style has been embedded within the study of dialect differentiation, Coupland points out the need to recognize a different type of style from a ‘dialect style’, namely, ‘expressive’ or ‘attitudinal style’. He notes that the consideration of styles inevitably extends beyond the concerns of ‘dialect style’, whose variables are semiotically associated with a social group and identity. “Variation in forms of address, politeness, conversational dominance, lexical formality, communicative key, self-disclosiveness”, he states, are in ‘expressive’ or ‘attitudinal style’. The gender-oriented and stance-oriented styles that I am discussing in this paper are generally comparable to his ‘dialect’ and ‘expressive’/attitudinal’ styles, respectively, except in that Coupland does not address the indirect indexing of a social group identity through the use of the expressive/attitudinal style that may be stereotypically associated with such a group. In considering such an indirect indexing, I am extending his observation and suggesting that the two styles (gender-oriented style and stance-oriented style) are not independent entities, just as direct and indirect indexicality are not independent phenomena. Some groups of speakers may have a tendency to use forms and expressions that pragmatically indicate direct and forceful attitudes, while some others may be inclined to choose



Tuning speech style and persona 

expressions that convey less direct and more delicate stances. At the level of the individual speaker, and, even more narrowly at the level of the individual utterance, one can analyze at the level of the variables of ‘forceful’ or ‘delicate’ style, which show the speaker’s attitudes, or as ‘male’ or ‘female style’, because of stereotypical cultural associations. There is no single preferred level of interpretation, since either or both may best explain the speaker’s choice of expression and the listener’s interpretation. It is such layers of interpretations available in expressions that provide speakers with resources for creating their own personae through shifts in styles in conversation. The focus of the discussion in this paper is mostly on the stylistic choices of sentence-final expressions in Japanese, with some attention to a few other expressions that are stylistically consonant. The sentence-final position is a prominent place where the speaker’s attitude is expressed in Japanese and therefore has been the focus of many previous studies investigating speakers’ subjective attitudes toward the content of their utterances and their relationships with each other. For example, the plain form copula da, which is used following a noun or an adjective, conveys a relatively definite and forceful stance. The plain form copula da conveys as its modal function the speaker’s certainty and definiteness toward the propositional content of the sentence (Konomi 1994, Teramura 1984). In contrast, the sentence-final pragmatic particle wa is understood as conveying the speaker’s reserved and delicate stance. Wa was used historically as an exclamation, expressing as its core pragmatic meaning the speaker’s light self-addressed confirmation of the propositional content of the sentence (Nihon Kokugo Daijiten ‘The Comprehensive Dictionary of the Japanese Language’ 1978, Uyeno 1971, 1972). Since it marks an utterance as addressed to oneself, the pragmatic force of an assertion with wa – in particular, its claim about the addressee’s belief – is less intense than is the force conveyed by the absence of a particle or by the presence of more assertive particles such as yo, ze, and zo. When wa is uttered with a rising or sustained intonation, the combination of the intonation contour and the pragmatic meaning of wa conveys an open-endedness that attenuates the speaker’s assertion. Therefore, the use and non-use of the copula da and the sentence-final particle wa can indicate the forceful style and delicate style that are in use by the speaker. The use of forceful vs. delicate styles has also been associated closely with gender by virtue of the traditional beliefs of normative male and female behavior, as illustrated in studies of language and gender (e.g. Ochs 1993). The following table (Table 1), based on a language textbook (Mizutani and Mizutani 1977), shows this sort of association to a striking extent. In this table, sentence-final expressions are presented as variables of gender styles. As mentioned above, the locus of the discussion will be examples from informal conversations among middle-aged, (upper-) middle-class women (mostly stay-athome mothers) from Tokyo. Their informal conversations provide unexpectedly interesting shifts. These speakers might be expected to present a uniform pattern of normative female speech, as they represent privileged Tokyo speakers who may be thought most likely to embody the traditional ‘good wife, wise mother’ ideal for women (e.g.

 Yoshiko Matsumoto

Table 1.  Gender differences presented by a Japanese textbook (Mizutani and Mizutani 1977: 150–151). Men usually say

Women usually say

English glosses

yokatta yo, yokatta ne

yokatta wa yo, yokatta wa ne

‘(it) was good’

iku yo, iku ne

iku wa yo, iku wa ne

‘will go’

ikanai yo, ikanai ne

ikanai wa yo, ikanai wa ne

‘won’t go’

soo da yo, soo da ne

soo yo, soo ne

‘that’s right’

soo da, soo datta

soo da wa, so datta wa

‘(it) is so, (it) was so’

asita da yo, asita da ne

asita yo, asita ne

‘(it) is tomorrow’

Okamoto and Sato 1992). However, the data present mixed styles of speech, as I illustrated elsewhere (e.g. Matsumoto 2002, 2004), even when there are no external conditions, such as a change of audience, topic or setting, which have been said to motivate style shifts in some past studies (e.g. Bell 1984, Labov 1972). Therefore, the informal conversations among these middle-aged middle-class women provide a good site for considering explanations and internal motivations for shifting styles. As mentioned above, the locus of the discussion will be examples from informal conversations among middle-aged, (upper-) middle-class women (mostly stay-athome mothers) from Tokyo. Their informal conversations provide unexpectedly interesting shifts. These speakers might be expected to present a uniform pattern of normative female speech, as they represent privileged Tokyo speakers who may be thought most likely to embody the traditional ‘good wife, wise mother’ ideal for women (e.g. Okamoto and Sato 1992). However, the data present mixed styles of speech, as I illustrated elsewhere (e.g. Matsumoto 2002, 2004), even when there are no external conditions, such as a change of audience, topic or setting, which have been said to motivate style shifts in some past studies (e.g. Bell 1984, Labov 1972). Therefore, the informal conversations among these middle-aged middle-class women provide a good site for considering explanations and internal motivations for shifting styles. In the following, I will discuss four segments of informal conversations among middle-aged women from the perspective of how stance-oriented styles and social identity-oriented (in this case, gender-oriented) styles are used. The first extract is from the collected conversations discussed at the Japanese Speech Style Shift Symposium, and the second to fourth examples are from my collection. In these two different groups of data, a similar pattern in choices of styles is observed.



Tuning speech style and persona 

3. Style shifting in conversations The first example is an excerpt from a casual conversation among four middle-aged female friends who went to the same junior high school.1 J and T are 52-year-old homemakers, and R and S (who does not speak during this segment) are 52-year-old kindergarten teachers. They have been chatting about how good fresh vegetables taste, and just before (1), R had started to talk about her favorite rice recipe with fresh ginger. She mentions that she tries to restrain herself from eating too much ginger rice since it is very delicious. The single underlines indicate expressions that are in a forceful style, and the double underlines indicate expressions in a delicate style. The pattern that we observe below between the two speakers J and R is interesting for two reasons: first, neither of them speaks uniformly in the stereotypical female speech style represented in the textbook (Table 1), and second, there is variation between J and R, who share the social background of class, school, and geographical region. J chooses a forceful or a delicate style depending on the situation, while R mixes the two styles even within the same string of utterances to the same hearer. Example (1) is an extract from their conversations. (1) 1 R:

[1 un 1]. [1 ‘yeah’ 1].

2

(.1) [2 oishii mon 2]. (.1) [2 ‘it’s delicious’ 2].

3 J:

[2 soo da yo ne= 2]. [2’ right, isn’t it= ‘2].

4

anata, ‘you,’

5

fukuyokasa kawannai mon [ne=]. ‘have not changed your well-roundedness [have you=’].

6 R:

[soo]. [‘right’].

7 J:

daietto suru dano, ‘you will be on a diet,’

8

nanda no kanda no, ‘or some such thing,’

9

tte, ‘you said,’

1. I am grateful to Tsuyoshi Ono for making this data available to the participants in the Japanese Speech Style Shift Symposium and for allowing me to use it in this paper. The speakers J and R were originally presented as ‘I’ and ‘K’, but because of confusion that these acronyms might cause in the paper, I changed them to J and R. Otherwise the transcript is as in the original.

 Yoshiko Matsumoto

10

[zutto itteru keredomo] [‘you have been saying this for a long time’]

11 R:

[shiten da yo, [‘I’ve been on a diet,’

12

kore demo]. ‘although you may not think so’].

13

<@ uhuhuhuhuhuhuhu @>.

14

<@ demo ne, <@ ‘but,’

15

kawaranai no yo @>. ‘nothing changes’ @>.

16 T:

@@@

17 J:

sayoo desu ka. ‘is that right.’

R switches between forceful and delicate styles in one utterance, as shown in lines 11 and 15. She asserts forcefully in line 11 that she has been on a diet, countering J’s utterance that insinuates otherwise by pointing out that R’s full figure has not changed. In line 15, on the other hand, she resigns herself to accepting that her weight has not changed, using a less forceful expression without the copula da. J in this segment uses a forceful style, indicated by da in line 3, agreeing with R’s assertion that the ginger rice is tasty, but this is the only use of forceful style in the transcribed part of the conversation (1’ 40”). Out of 54 intonation units in the transcription, J exhibits 6 instances of speaking in delicate style and 1 instance of forceful style, namely the one in line 3. (In comparison, R shows, among 56 intonation units, 5 instances of forceful style and 2 instances of delicate style.) When J speaks in delicate style in other parts of the transcribed conversation, her utterances contain her statements and opinions, such as watashi soo iu ten de wa shiawase da wa (rising intonation) ‘I’m lucky in that regard’, taberare nai no ‘I cannot eat it’, mazui no yo ne ‘it tastes bad, doesn’t it?’ It seems that there is a division in the use of forceful and delicate styles depending on the purpose of the utterances; namely, J uses a delicate style to assert her point of view while she uses the forceful style to give her agreement with another person. R, on the other hand, does not seem to show such a division of use, at least within the transcribed segment of the conversation. A similar observation is made among speakers whom I examined in my earlier study (e.g. Matsumoto 2002, 2004).2 In example (2) below, the speaker E (Eri) presents a similar pattern to J, and M (Minako) is somewhat reminiscent of R, in that her shifting between the two styles is not apparently motivated by different speech acts. 2. The discussion here builds on the basic analyses presented in Matsumoto 2002, 2004.



Tuning speech style and persona 

The speakers are four women in their late 30’s to early 40’s whose children attended the same public middle school in a middle- to upper-middle-class residential neighborhood in Tokyo. The women were members of the local PTA Public Relations Committee. The main speaker, whom I refer to as Minako (M), was the oldest in this group (43 years old at the time of the recording). The other speakers are referred to as Chie (C), Eri (E), and Fumiko (F). (2) 1 M: chotto kotchi iketakke ‘I wonder if we can go this way’ : 2 C:

... ki no oto ga sugoo(ku) ‘...the sound of trees was really terri(ble)’

3 M: a, yabbee, kore migi shika ikenai wa. ‘Oh, damn! I can only go right here.’ 4 E:

a, honto da ‘Oh, yeah, that’s true’

5 M: arararara ‘Dear dear dear’ 6 E:

minna ippoo-tsuukoo ni natchatte ‘They all became one-way’

7 M: kotchi hairitakunai n da yo na ‘I don’t want to go in here’ 8 E:

ara, zutto da wa ‘Oh, all of them are’

9 C:

koko n toko zutto soo nan desu yo, michi ga ‘It’s all like that around here, the roads are’

10 E:

a soo nan da yo ne ‘Yeah, that’s right, isn’t it?’

11 M: ura ikitai no yo ne, ura o ‘I want to take a back road, back road’ Eri’s utterance in lines 4 (honto da ‘oh, yeah, that’s true’) and 10 (soo nan da yo ne ‘that’s right, isn’t it?’) with the use of da seem to express positive attitudes of agreement with or support for the other speakers. With the use of da, Eri’s assertions are more definitive and convey a more forceful stance toward her expressions of agreement. In contrast, Eri uses an attenuated expression in line 8, ending with wa – zutto da wa ‘all of them are’ – observing and describing (seemingly to herself) the fact that the roads they are crossing are all one-way. Eri’s choice between forceful and attenuated expressions can be explained according to the purposes of her speech act, as in the case of Speaker

 Yoshiko Matsumoto

J in example (1): she uses forceful expressions when agreeing with other’s opinions and delicate expressions when expressing her own view. As mentioned earlier, the style expressing forceful attitude has been commonly associated with the gender-oriented style of men in informal settings. The attitudinal style of being delicate, on the other hand, has been typically associated with adult women as a choice or as a social constraint on their linguistic behavior. However, apart from these gender-oriented interpretations of styles, we can also say that these two stances or attitudinal styles are strategically used by Eri, depending on the goals of her speech acts, to convey a friendly persona and concern for others. One way is to strongly support her interlocutor by making assertions (expressions of agreement) straightforward and forceful. Another strategy is to show her concern for others by not pressing her own views (when she expresses her opinions) and by conveying friendliness in a more reserved manner. It could be argued that these effects of being forceful and reserved were achieved by using or appropriating prerogatives of the associated gender. That would be one possible analysis, but the expressions at issue carry pragmatic meanings (index stances) that form the basis for the given gender indexes, and it is accordingly more plausible to consider the attitudinal styles directly in terms of such pragmatic meanings when discussing the effect of style shifts rather than indirectly in terms of the gender that they index. In contrast to Eri, Minako’s use of forceful and delicate expressions is not as clearly demarcated by speech acts as Eri’s. This is similar to R in example (1). To express her feelings about the unwanted situations that she was put in, Minako uses forceful expressions in lines 3 (yabbee ‘damn’) and 7 (hairitakunai n da yo na ‘I don’t want to go in here’), but more delicate expressions in lines 5 (arararara ‘dear, dear’) and 11 (ikitai no yo ne ‘I want to go there’). We cannot, therefore, explain Minako’s choices of expressions simply by linking them to different speech acts as we did for Eri. A possible explanation for her choice is that Minako’s speech pattern represents a more complex persona of being direct but delicate in stance and of recognizing the social constraints on a ‘traditional’ middle-class adult woman while noticing what I call the ‘covert stigma’ attached to such an anachronistic gender ideology. Her utterances can be seen as an outcome of exploiting the two dimensions of style. By using both forceful and delicate styles, she adjusts her stances and expressions of her relationship to others, and balances and negotiates between traditional and innovative images of being a woman. In the above discussion, we have considered two types of speakers whose speech presents shifts in style that created complexity in self-presentation -- Eri’s utterances show one forceful and one delicate expression, and Minako uses 6 forceful and 2 delicate expressions. There are others, however, whose styles do not exhibit an obvious shift. For example, Fumiko (F), in the stretch of discourse given in (3) below, uses 5 forceful expressions with no delicate expressions regardless of the content of her utterances. The non-use of delicate expressions is most often offset by adding the pragmatic sentence-final particle ne(e), which gently invites the addressee’s involvement. The personae that these expressions project may vary depending on the intonation



Tuning speech style and persona 

contour and on the social background of the interpreter, but they generally give an image of someone who is wholesome and straightforward. Example (3) provides a further illustration of the variations of verbal interaction among the speakers who were observed in example (2). The excerpt is from the last part of the same 120-minute conversation we saw in (2). In this example the mothers are looking for a place to rest and have something to drink before their expedition at the park, Lake Sagami Picnic Land. (3) 1 M:

are o hidari ni iku no? ‘Do we go left over there?’

2 E:

un, 412, un ‘uh-huh, 412, uh-huh’

3 F:

a! pikunikku rando tte kaitearu. ‘Ah! It says ‘Picnic Land’.’

4 E:

sono saki o sa-setsu ‘Make a left turn after that’

5 F:

migi ga sagamiko eki da ne. ‘Lake Sagami Station is on our right, isn’t it?’

6 M:

oi oi! [pause] ‘Hey, hey!’ [pause]

7

nanika o-mise mo rokuni aitemasen koto yo [laughter] ‘It seems that stores are scarcely open’ [laughter]

8 E:

[laughter]

9 F:

nee ‘yeah, right’

10 M:

nanka shinto hora atchikotchi shimatteru yo. hora mada. [laughter] ‘It looks quiet. Look, stores are closed here and there. Look, not yet open.’ [laughter]

11

koohii nomitaku temo nome-nai janai desu ka, kore ja ‘Even if we want to, we can’t have coffee, can we, if it’s like this’

12 C, E, F: [laughter] 13 M:

nanika attara nomoo ne, toriaezu, sagami p pikunikku rando ni hairu mae ni ‘Let’s have coffee, if there’s any, one way or the other, before we get into Sagami P Picnic Land’

14 E:

mada mada ‘not yet open, not yet’

15 F:

mada 10-ji dakara aite nai ne. ‘Since it’s still 10 o’clock, they are not open, are they?’

 Yoshiko Matsumoto

16 E:

aite nai ne. makudonarudo mo nai shi tomerareru toko mo nai’ ‘They are not open, are they? There isn’t even McDonald’s, and there is no place to stop, either’

17 M:

nai yo. ‘No.’

18

aa waratchau yo ne. ‘Oh, this is something to laugh at, isn’t it?’

19 E:

resuto hausu atte, yasumu toko nai no kashira. komatchau wa ne. ‘There is a rest area, I wonder if there is any place to rest. I don’t know what to do.’

20 M:

demo yokatta ne, hayaku tsuite ne. ‘But, it’s good that we got here quickly, isn’t it?’

21 F:

nee. suiteta ne. ‘Yeah, the traffic was smooth, wasn’t it?’

22 M:

suiteta suiteta. ‘It was, it was.’

23 F:

suitetara densha yorimo zutto hayai n da ne. ‘When the traffic is smooth, driving a car is much faster than a train, isn’t it?’

24 M:

hayai wa yoo. ‘It is faster’

In this example Eri again uses the same pattern of alternating expressions as we saw in example (2): when she agrees with someone, she uses a more forceful expression, as in line 16 (which is a repetition of what Fumiko just said), and when she expresses her thoughts and concern, she uses less forceful expressions, as in line 19. Fumiko employs forceful expressions consistently through the extract. Similarly, Minako’s utterances again display a complex distribution of forceful and delicate expressions. As I also discussed elsewhere (Matsumoto 2002, 2004), Minako’s utterances from line 6 to 13 are of particular interest in terms of style shifting. Oi oi ‘hey, hey!’ in line 6 is a more forceful exclamatory attention-getter than, for example, chotto or ara (example 2, lines 1 and 8) and can be directed to oneself as well as to others. After a pause, Minako introduces a new topic in line 7, commenting that hardly any shops are open. She repeats this observation in line 10, using the style of expressions that indicate a different attitude. The linguistic form in line 10 shimatteru yo ‘stores are closed’ lacks the particle wa between the predicate and the assertive particle yo, and it is therefore relatively forceful. In contrast, the main predicate aitemasen ‘is not open’ (rather than aitenai) in line 7 is what is commonly termed the ‘polite form’, unlike the majority of the utterances in this recording. What follows the polite form predicate is the formal noun, or nominalizer, koto (literally meaning ‘fact’), which



Tuning speech style and persona 

functions to make the content of the preceding clause an established fact and has a distancing effect. Through this combination of polite and formal forms, the emotional impact of the statement is lessened, and therefore its affective force is attenuated. This form [predicate + koto + yo] can fit the traditional cultural norm of femininity and has been known as one of the stereotypical old-fashioned forms of upper-middle-class female speech. It is likely that the speaker is aware of the connotations of the form and is using it ironically for humorous effect. The following stretch of conversation in example (4) includes an interesting contrast of styles by two speakers neither of whom shifts styles, although the styles they maintain are disparate. The two speakers in question, Yuriko (61 years old) and Kazuko (52 years old), both belong to a group of amateur singers composed of alumnae of a university chanson club, of which Minako is a member. Although Kazuko is much older than Fumiko, her utterances in this conversation are reminiscent of Fumiko’s in example (3), in that she uses the forceful style and does not shift to the delicate style. Yuriko, on the other hand, uses exclusively the delicate style without shifting to the style of forceful stance. Kazuko and Yukiko are similar, however, in that their styles do not shift. According to Minako, who knows them well as fellow members, what we observe in this segment of the conversation is typical of these speakers’ speech styles. The example is an excerpt from a conversation recorded in a gathering of the group after a concert, in which Minako sang solo pieces. Yuriko and Kazuko, who are both senior to Minako, are congratulating Minako on her great performance. Here they started talking about how Minako’s children responded to Yuriko’s compliments to their mother. (4) 1 Y: Okaasama suteki datta deshoo tte ittara, anoo. ‘When I said ‘your mother was fabulous, wasn’t she?’, umm.’ 2 → K: soo itteta yo. ‘They said so.’ 3 → Y: soo ittemashita yo, soo itte irashita wa ee. suteki datta tte nee. ‘They said so, they said (honorific) so, yes. It was fabulous, right.’ 4 N: aa soo. ‘Is that so.’ 5 M: demo chotto utai-kata ga chigatta ne tte iwarechatta ‘But they told me that the way I sang was a bit different’ 6 Y: aa soo ‘Is that so?’ 7 K: waaa ‘Wow’ 8 N: heee ‘Gee’

 Yoshiko Matsumoto

9 R: hontoo ‘Really?’ 10 → Y: yokuu ja kiite ite kekkoo ne kiiterashita no nee. ‘Well then they were listening fairly well, they listened (honorific).’ 11 → K: yoku kiiteru n daa ‘They listened well’ 12 M: kashi machigaeru na tte iwareru n desu yo, oboeteru n desu yo, kekkoo ‘Don’t make mistakes in the lyrics’ I’m told, they remember the lyrics fairly well’ 13 K: sugoi nee ‘I’m amazed’ On two occasions, lines 2 and 3, and lines 10 and 11, Yuriko and Kazuko say the same thing repeating the content of each other’s utterances almost overlappingly, but their styles are contrastive. The contrast is reminiscent of the two different strategies that we discussed in the previous examples. Kazuko uses straightforward expressions (soo itteta yo ‘they said so’, yoku kiiteru n daa ‘they were listening well’) in this excerpt, conveying camaraderie in the same direct style used by Fumiko in the previous example. Yuriko, on the other hand, uses more delicate expressions. These examples illustrate the contrast of two personae presented by Kazuko and Yuriko. Even middle-aged women who share several aspects in life, such as educational background, general socio-economic class, and artistic interests do not speak in one style, but in different styles depending on their personal choices in self presentation. Another interesting point to note in this segment of conversation is Yuriko’s adjustment of her own speech style. In both lines 3 and 10, she quickly rephrases her speech – she first uses less- or non-honorific forms of expressions, ittemashita ‘they said’ and kiiteite ‘was listening to’, but she changes them to expressions of delicate stance and of referent honorifics, i.e. itte irashita wa and kiiterashita no nee. I should therefore qualify what I said earlier about lack of style shift in Yuriko’s speech. Yuriko’s speech does not shift between the contrastive styles focused in this paper, forceful and delicate styles, as Minako’s and Eri’s did, but indicates intentional adjustments to a more refined style with honorifics and delicate expressions. It is noteworthy that Yuriko uses such delicate and honorific expressions very consciously. In both instances of rephrasing to an honorific expression, the target of honorification is Minako’s children, even though adults do not generally use honorifics in referring to children. Various interpretations can be offered for the rephrasing from expressions of more or less neutral stance to more delicate and ideologically laden expressions. For example, Yuriko may have thought that she should express her recognition of Minako’s performance by showing that she is worthy of respect (via use of honorifics to her children) in a traditional delicate way, or she may have been trying to maintain the “ideal” image of herself suited to her upper-middle background and age. As the only participant who was



Tuning speech style and persona 

exposed to the pre-war value system and education (and teaching home economics at a girl’s high school), she may still uphold in principle the traditional “good wife and wise mother” ideology and its associated linguistic behavior. Yuriko’s choice of expressions further creates a stereotypical image of a traditional upper-middle class female speaker. It is interesting to note, however, that although she is the only speaker in my data who possibly can represent the stereotype of the traditional female speech, she is the only one who is not a stay-at-home mother, the (imagined) paragon of stereotypical female speech. Whether the explanation suggested above reflects the speaker’s intention or the listener’s interpretation, it is important to recognize that these layers of meanings are available in the context.

4. Conclusion We have discussed the motivations and functions of different patterns of style shifts, including one case in which shifts did not occur, in peer conversations among three different groups of middle-class middle-aged women in Tokyo. Our focus was on shifts in stance-oriented styles (forceful and delicate) and on gender identity-oriented styles (male and female). A key point of departure for the study was the observation that, despite the common belief that the style used by speakers of the sociological profile of those in my study should be exemplary of normative female speech style, the styles of their speech in fact vary even when external speech conditions are kept constant (Matsumoto 2002, 2004). The utterances in the examples show shifts in the speakers’ styles in casual conversations among friends. Some speakers used different styles depending on the purposes of the speech, i.e. supporting others’ assertions or giving one’s own; others did not make such distinctions but used both forceful and delicate styles to express similar content; and some did not shift styles regardless of the speech purposes, but maintained one particular style. A most likely explanation for these observations is that the style shifts are not dependent on external factors or simple social identities, but on each speaker’s personal choice in presenting herself and negotiating her persona in the context of the utterance. One of the implications of this study is the importance of emphasizing the individual speaker’s personal style and choice, beyond her membership in a pre-defined social group, i.e. women, middle-class or middle-age individuals, or even a self-selected community such as a club. There is no guarantee that one speaker’s choice of style at a specific point in a speech situation will be completely predictable from her identity in terms of social groups. Johnstone (1996) argues for the importance of considering individual styles that are distinctive among people within the same socially identified group, and that the specific style of one speaker is observable across different genres such as written and spoken language. The findings in this paper provide ample evidence in support of at least the first point of Johnstone’s observations.

 Yoshiko Matsumoto

This study also shows that there are various interpretations that can be associated with the use of certain expressions and styles rather than one particular meaning being associated with one expression. In order to understand the multiplicity of conveyed meanings, I argue that it is useful to maintain the distinction between stance-oriented styles, such as forceful and delicate styles, and gender-oriented styles, such as male and female styles, rather than subsuming one under another. A speaker may be choosing a certain expression consonant with one style rather than another to express her attitude toward the situation (i.e. ‘attitudinal style’) rather than to express her identity with a social group known to speak in such a style (i.e. ‘dialect style’). Certain attitudes can always be interpreted as emblematic of one social group, and the opposite can also be true, i.e. the style associated with a gender group may suggest certain stereotypical attitudes. Accordingly, the use of a particular expression can be interpreted as an instance of either or both types of styles. The distinction is analogous to the useful idea of attitudinal and dialect styles, but I emphasize that the two levels of description can both apply to a single utterance in a context. Through negotiation of these layers of meanings available in expressions, the individual speaker tunes her style and persona in interaction.

References Bell, A. 1984. “Language style as audience design.” Language in Society 13: 145–204. Coupland, N. 2001. “Language, situation, and the relational self: Theorizing dialect-style in sociolinguistics.” In Style and Sociolinguistic Variation, P. Eckert and J. Rickford (eds.), 185–210. Cambridge: Cambridge University Press. Eckert, P. and Rickford, J. R. 2001. Style and Sociolinguistic Variation. Cambridge: Cambridge University Press. Eckert, P. and McConnell-Ginet, S. 1992. “Communities of practice: Where language, gender, and power all live.” In Locating Power: Proceedings of the Second Berkeley Women and Language Conference 1 & 2, K. Hall, M. Bucholtz and B. Moonwomon (eds.), 89–99. Berkeley: Berkeley Women and Language Group. Gal, S. 1995. “Language, gender, and power.” In Gender Articulated: Language and the Socially Constructed Self, K. Hall and M. Bucholtz (eds.), 169–181. New York: Routledge. Johnstone, B. 1996. The Linguistic Individual. New York and Oxford: Oxford University Press. Konomi, E. 1994. The Structure of the Nominal Predication in Japanese. Unpublished Ph. D. dissertation. Cornell University. Labov, W. 1972. Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. Matsumoto, Y. 2002. “Gender identity and the presentation of self in Japanese.” In Gendered Practices in Language, S. Benor, M. Rose, D. Sharma, J. Sweetland, and Q. Zhang (eds.), 339–354. Stanford, CA: CSLI Publications. Matsumoto, Y. 2004. “Alternative femininity and the presentation of self in Japanese.” In Japanese Language, Gender, and Ideology, J. Shibamoto Smith and S. Okamoto (eds.), 240–255. New York: Oxford University Press.



Tuning speech style and persona  Mizutani, O. and Mizutani, N. 1977. An Introduction to Modern Japanese. Tokyo: The Japan Times, Ltd. Nihon Kokugo Dai-ziten (The Comprehensive Dictionary of the Japanese Language). 1978. Tokyo: Syoogakkan. Ochs, E. 1993. “Indexing gender.” In Sex and Gender Hierarchies, B. D. Miller (ed.), 146–169. Cambridge: Cambridge University Press. Okamoto, S. and Sato, S. 1992. “Less feminine speech among young Japanese females.” In Locating Power: Proceedings of the Second Berkeley Women and Language Conference 1 & 2, K. Hall, M. Bucholtz and B. Moonwomon (eds.), 478–488. Berkeley: Berkeley Women and Language Group. Teramura, H. 1984. Nihongo no Shintakusu to Imi (Japanese Syntax and Meaning) Vol. II. Tokyo: Kurosio Syuppan. Uyeno, T. Y. 1971. A Study of Japanese Modality -- A Performative Analysis of Sentence Particles. Unpublished Ph. D. dissertation. University of Michigan. Uyeno, T. Y. 1972. “Shuujosi to sono shuuhen.” (Sentence-final particles and related issues). Nihongo Kyooiku (Journal of Japanese Language Teaching) 17: 62–77.

Speech style and the use of regional (Yamaguchi) and Standard Japanese in conversations* Shigeko Okamoto

University of California, Santa Cruz

1. Introduction Regional variation in speech is one of the most widely studied topics in sociolinguistic research on diversity in Japanese. Extensive research in modern Japanese dialectology, whose history dates back about 100 years, has greatly enhanced our understanding of the nature of this variation, which can be sometimes so great that mutual intelligibility may be hindered. At the same time, it is often noted that regional variation has diminished considerably in modern Japanese. That is, language standardization is said to have been steadily advancing due to a variety of factors associated with the nation’s modernization, such as the government’s language policy in pre- and post-War Japan and the development of mass communication and modern transportation systems (Inoue 1995; Kobayashi and Shinozaki 2003; Sanada 1991, 1996, 2002; Shibata 1988; 1999 [1965]; Yoneda 1997). On the other hand, it has also been pointed out that regional varieties of Japanese have not lost their vitality and that many speakers are now bi-dialectal and code-switch between the regional and Standard varieties according to the situation (e.g. Carroll 2001; Miyake 1995; Shibata 1988). For example, Inoue (1988: 19–20) states that while standardization is proceeding, dialectal forms are still used vigorously (in informal situations), showing “a kind of bi-dialectalism.” Inoue further points out that new dialect forms have even been born, indicating that the direction of language change is not only toward standardization. Recognizing some regional differences in the manner of standardization, Long (1996: 122) notes that whereas the typical trend in EasternJapanese dialect changes has been a shift from the local dialect to the Standard, “the * This study is partly supported by a California State University Research Grant. I would like to thank those who participated in the data collection and analyses, in particular, Kimiko Izumi, Takae Izumi, and Reiko Ano. I also thank the following people for their valuable comments and discussion: Kimberly Jones, Ritva Laury, Karen Mistry, Tsuyoshi Ono, Sumiyuki Yukawa, Raymond Weitzman, and an anonymous reviewer.

 Shigeko Okamoto

trend in Western-Japanese dialects has been towards bi- (or multi-) dialectalism” with “situational code-switching.” Sanada (1996, 2000) uses the term neo-hoogen, or neodialect, to refer to a koodo ‘code’, or a linguistic system, that lies in between hoogen ‘dialect’ and hyoojun-go ‘Standard language’. Speakers are then said to shift between these three varieties depending on the situation, which is considered a matter of stylistic variation concerning formality, with hyoojungo corresponding to a formal style, neo-hoogen to a neutral style, and hoogen to an informal style. These previous studies provide us with some general ideas about the use of regional and Standard Japanese (SJ, hereafter). However, many of them are based on self-report survey data or researchers’ informal field observations, and it is still quite unclear exactly how speakers use regional and Standard Japanese in actual conversations. It is only recently that researchers have started examining actual speech data. For example, a group of researchers at Osaka University examined the use of variables (mainly grammatical and lexical) related to formal and informal styles by dialect speakers in conversations carried out in two kinds of situations: one in which the interlocutor is a familiar person and the other in which the interlocutor is a stranger (i.e. a researcher). As part of this project, Abe and Sakaguchi (2002), Takagi (2002), and Hosotani (2004), for instance, report many cases in which speakers differentiated the use of “regional” and “Standard” variants as style switching (sutairu kirikae). In particular, categorical switching according to the situation was observed more among younger speakers than among older speakers (see also Shingu 1997). Against this background, the present study attempts to further investigate the question of how regional and Standard Japanese are used in conversations carried out by speakers interacting in different social situations. I am particularly interested in examining whether the notion of code-switching that is often used to refer to the language practice of dialect speakers is in fact adequate to describe such practice. That is, if speakers are bi-dialectal, do they use one variety exclusively depending on the situation, which may be referred to as “situational code-switching”? (See also Blom and Gumperz [1972] for a discussion of “situational code-switching” in a rural community in Norway; cf. Maelum [1996].) My analysis will demonstrate that the notion of codeswitching is not adequate to describe the linguistic practice examined in the present study, because the speakers do not use SJ and their regional dialect as discrete codes. I then employ the notion of variant choice and examine the use of SJ and regional variants in relation to the nature of conversational context and the notion of formality of style. As mentioned above, the use of regional and Standard Japanese is sometimes linked to formal or informal style. I would like to examine this link further in terms of the notion of formality as involving the speaker’s stance toward the given conversational context. Lastly, the results of my analyses will be discussed in relation to the macro-level sociopolitical and historical contexts. The regional dialect involved in this study is Yamaguchi Dialect (YD, hereafter). The data were collected in Shuuhoo-choo, a small rural town in Yamaguchi prefecture in Western Japan. As of 2001, its population was 6,374 and the number of households



Regional and standard Japanese in conversations 

was 2,166. Its main industry is agriculture with a focus on such products as rice and pears (Fureai Yamaguchi 2002: 6–7). As is the case in many other rural areas in Japan, the population of Shuuhoo-choo has been steadily declining, because young people increasingly move to more urban areas. The main means of transportation are buses and cars; no train goes to Shuuhoo-choo. Overall, the town can be regarded as relatively isolated. Before turning to the next section, explanations of certain key notions used in this study are in order. I use the term “variable” as an abstract linguistic item that has multiple realizations, or variants. For example, the verb past tense form is a variable that has variant forms such as V-oota (in “YD”) and V-tta (in “SJ”) (e.g. koota vs. katta for ‘bought’). The notions of variable and variant thus concern individual linguistic items. On the other hand, the notion of code concerns a grammatical system (Gumperz 1982), or variety, consisting of a set of linguistic properties. Note also that I regard the notion of authentic or pure variety, such as “authentic YD” or “authentic SJ” as an ideal. As will be discussed in Section 3, the difference between varieties is on a continuum both diachronically and synchronically and does not warrant any variety as a distinct, or pure, category. In this respect, terms such YD and SJ used in this study should be understood as “YD” and “SJ” – as referring to broad differences. However, to avoid clumsiness, I do not use quotes for them except when it is important to use them. I use the term “SJ,” or “hyoojungo,” as a variety presumed to be used in the Tokyo Metropolitan area and to be understood by most Japanese.1 The term “context” will be used in the sense of conversational context, consisting of a variety of aspects, including the participants and their relationship (e.g. status difference, degree of intimacy), the domain (e.g. home, work), genre (e.g. small talk, sales talk), and speech acts (e.g. apology, request). This is not a large-scale study involving many speakers and I do not intend to claim that its findings can be generalized to all speakers of regional dialects in Japan. But I hope my study based on discourse analysis will serve as a step toward the development of a new perspective on the use of Standard and non-Standard, or regional, varieties of the same language – a language which is said to have been undergoing substantial standardization.

1. The term kyootsuugo ‘common language’, which was introduced after World War II, is often used instead of hyoojungo. Originally, it was meant to be understood as a variety that might not be a “perfect” form of hyoojungo, but could be understood in any place in Japan (Sanada 1991; Shibata 1997, 1999 [1965]). I use the term “SJ” to include both hyoojungo and kyootsuugo, because the latter has come to be used simply as an alternative to hyoojungo, making it difficult to distinguish the two in terms of linguistic properties (although the use of kyootsuugo may appear to be more “democratic” than that of hyoojungo – see Section 3.4 for further discussion).

 Shigeko Okamoto

2. Method The data used for this study were collected in 1997 and 1998 in Shuuhoo-choo in Yamaguchi prefecture. The main body of the data consists of nine audio-taped dyadic conversations. Additionally, a questionnaire was distributed a few weeks after recording the conversations (see below). Altogether, ten speakers participated in recording conversations (Table 1; note that Conversations A-5 and B-4 are the same). The speakers recorded their own conversations, sitting face to face; conversational topics were self-selected; and the researcher was not present at the time of the recording. All the participants have lived in Shuuhoo-choo or in its vicinity most of their lives. As shown in Table 1, participant A spoke with 6 different persons, and participant B spoke with four different persons. In this study I mainly focus on the speech of these two participants. Speaker A was born in Fukuoka, Kyushu, but has lived in Shuuhoo-choo and its vicinity most of the time since childhood. She is a high school graduate and works for the family business. Speaker B was born and lived in Shuuhoochoo throughout her life. She has graduated from a junior college and is a tour guide. The first eight minutes of all conversations were transcribed. I examine these conversations with regard to the use of morphological and lexical variables by Speakers A and B. Phonological differences between YD and SJ are small and not examined in this study. For example, the pitch-accent patterns of YD are basically the same as those of SJ (Nakayama 1982; Shibatani 1990), as illustrated in Table 2, which shows the pitchaccent patterns for three-mora phrases. (The patterns in two other dialects are also included in Table 2 to show that they can differ considerably.) Table 1.  Conversations analyzed in this study Conversations

Speaker A (female, age 58), talking with…

A-1 A-2 A-3 A-4 A-5 A-6

Speaker

C D E F B G

(female, age 78, A’s mother) (female, age 60, A’s older sister) (female, age 23, A’s daughter) (female, age 59, A’s close friend) (female, age 25, A’s daughter’s friend) (female, age 69, A’s daughter’s former high school teacher)

Speaker B (female, age 25), talking with… B-1 B-2 B-3 B-4

Speaker

H I J A

(female, age 44, B’s mother) (male, age 58, B’s father) (female, age 23, B’s close friend) (female, age 58, B’s friend’s mother)



Regional and standard Japanese in conversations 

Table 2.  Examples of pitch-accent patterns in YD and SJ

Tokyo Yamaguchi Osaka Kagoshima

Kaze-ga* wind

yama-ga mountain

sora-ga sky

ame-ga rain

LHH LHH HHH LHL

LHL LHL HLL LLH

HLL HLL LLH LLH

HLL HLL LHL LLH

* The particle ga is a nominative case marker; L stands for a low pitch height and H for a high pitch height. (based on Shibatani 1990: 190)

Table 3.  Examples of morphological differences between YD and SJ YD form Inflectional endings

SJ form

English gloss

Example in YD

*Adv-oo *Adv-uu

Adv-aku/oku Adv-uku

adverbial adverbial

hayoo ‘quickly’ samuu ‘cold’

*AN/N-ja/ya

AN/N-da

‘be’, copulative AUX

toshi ja ‘it’s the age’

*A/AN/N/V -jaro(o)/yaro(o)

A/AN/N/V -daro(o)

‘isn’t it?’, ‘probably’

soo jaro ‘probably so’

V-charu

V-te aru

‘is done’

kaicharu ‘it’s written’

*V-choru

V-te(i)ru

V progressive/ stative

shitchoru ‘(I) know.’

*V-n

V-nai

‘not V’

shiran ‘(I) don’t know.’

*V-oota

V-tta

V past

koota ‘bought’

*V-oote

V-tte

V continuative

koote ‘buy (and)’

V-san

V-nasai

V imperative

miisan ‘look’

S so

S no

‘you know’

nai so ‘there’s none, you know’

S ho

S no

‘you know’

moroota ho ‘I received, you know.’

Interactional particles

 Shigeko Okamoto

As will be seen in the following section, there are many morphological and lexical forms that are shared by both YD and SJ – i.e. forms that are not variables. In my analysis I focus only on certain variables, or forms that have regional and Standard variants. Two native speakers of YD identified variables and their variants. Note that it is not always clear whether a particular form is a dialectal or Standard form for reasons I discuss in the following section. Accordingly, terms such as “SJ variant” and “YD variant” are not meant to represent absolute categories. Tables 3 and 4 illustrate respectively the morphological and lexical differences2 between YD and SJ. (The examples shown are taken from the current data. The following abbreviations are used in Table 3: AN [adjectival noun]; AUX [auxiliary]; N [noun]; S [sentence]; and V [verb].) Table 4.  Examples of lexical differences between YD and SJ YD form SJ form Verbs, Adjectives, Adverbs

English gloss

Example in YD

buchi

totemo

‘very, really’

*ee *donee *sonee yooke

ii doo, donna soo, sonna takusan

‘good, fine, OK’ ‘how’ ‘such, so’ ‘many, much’

buchi kimochi ekatta ‘it felt really good’ ee nee ‘it’s good, isn’t it?’ donee jatta ‘how was it?’ sonee omoote ‘thought so’ yooke oru ‘there are many’

hiruma mainichi jibun

‘day time’ ‘every day’ ‘oneself ’

hinaka itte ‘go during the day time’ maihi ni ‘every day’ meeme ga shitara ‘if (he) himself did’

soshitara sorede sorede, dakara dakedo, demo dakara

‘then’ ‘then, so’ ‘so, therefore’ ‘but’ ‘so, therefore’

hetara ne ‘then, you know’ hede ima made wa ‘so up until now …’ hode anta mo ‘so you, too …’ jakedo yappari ‘but as expected’ heyakara ma hashitte ‘so by running …’

Nouns hinaka maihi meeme Conjunctions *hetara *hede, hete *honde, hode *(he)jakedo *(he)yakara

2. Some of the morphological and lexical variables may appear to be phonological variables. However, I treated a variable as morphological if it involves a grammatical feature (e.g. verb conjugation) rather than all instances of the sound in question in any words. I treated a variable as lexical if it involves only certain lexical items. Thus, for example, the adverbial variable with the variants Adv-oo and Adv-aku/oku is treated as morphological because it concerns the formation of the grammatical category ‘adverb’; the variable ii ‘good’ (with ee and ii variants) is treated as lexical, because this phonological difference does not apply to all instances of the sound ii, but rather, is restricted to this lexical item.



Regional and standard Japanese in conversations 

Among the variables listed in Tables 3 and 4, for the quantitative analyses I focus on those that occurred relatively frequently in most conversations – those marked with an asterisk. I examined the distribution of the YD and SJ variants of each of these variables used by each of the two speakers (Speakers A and B) in each conversation. For the qualitative analyses, I consider those marked with an asterisk as well as others, including some not listed in Table 4. As mentioned earlier, a questionnaire survey was conducted in order to identify the speakers’ subjective views toward YD and SJ. Nine out of the ten people who participated in recording conversations responded. The questionnaire included two questions: (1) What kind of language do you think Standard Japanese (hyoojungo) is?, and (2) What kind of language do you think Yamaguchi Dialect is? For each question, the same thirteen descriptors (and one additional one for the first question) were given, and the respondents were asked to choose all the descriptors that they thought were applicable to each question. They were also asked to write any other characteristics, if any. The thirteen descriptors were: (1) it has feeling (aji ga aru); (2) it has no feeling; (3) it is relatively polite (teenee); (4) it is relatively rough (zonzai); (5) it is used in writing; (6) it is correct (tadashii) language; (7) it feels relatively intellectual (chiteki); (8) it is used in formal (aratamatta) occasions; (9) it is used in informal (kudaketa) occasions such as when talking with family members or close friends; (10) it has a relatively cold (tusmetai) feeling to it; (11) it has a relatively warm (atatakai) feeling it; (12) I feel attachment (aichaku) to it; and (13) relatively speaking, it is a language I dislike. The one descriptor added to the first question was: it is a language used by people in the Kanto region, especially those in Tokyo.

3. Results and discussion 3.1

Inter- and intra-speaker variation in the use of YD and SJ variants

The result of my analysis shows wide inter- and intra-speaker variation in the use of YD and SJ morphological and lexical variants. Figure 1 shows this variation in term of the percentage of the SJ variants used in each conversation. Table 5 provides more detailed information by showing the distribution of morphological and lexical variables separately. (Note that the total number of tokens in some of the conversations is relatively small, because the amount of conversational floor the speaker – A or B – occupied was small.)

 Shigeko Okamoto

Conversations

(SJ% )

A-1 (mother) A-2 (older sister) A-3 (daughter) A-4 (close friend) A-5 (daughter’s friend) A-6 (daughter's teacher)

B-1 (mother) B-2 (father) B-3 (close friend) B-4 (friend’s mother) 0

10

20

30

40

50

60

70

80

90

Figure 1.  Use of SJ morphological and lexical variants by Speakers A and B

Table 5.  Use of SJ morphological and lexical variants by Speakers A and B Conversations

Morphological SJ/Total # of tokens (%)

Lexical SJ/Total # of tokens (%)

Total SJ/Total # of tokens (%)

Speaker A A-1 (mother) A-2 (older sister) A-3 (daughter) A-4 (close friend) A-5 (daughter’s friend) A-6 (daugher’s teacher)



Total



15/173

(9)



B-1 (mother) B-2 (father) B-3 (close friend) B-4 (friend’s mother)



9/27 11/52 10/35 20/25

(33) (21) (29) (80)



3/3 (100) 5/8 (63) 20/26 (77) 15/16 (94)

12/30 16/60 30/61 35/41

Total



50/139 (36)



43/53

93/192 (48)

0/13 (0%) 0/40 (0) 2/35 (6) 0/41 (0) 7/30 (23) 6/14 (43)



1/8 (13%) 6/27 (22) 4/12 (33) 4/41 (10) 8/12 (67) 7/7 (100) 30/107

(28)

1/21 (5%) 6/67 (9) 6/47 (13) 4/82 (5) 15/42 (36) 13/21 (62) 45/280 (16)

Speaker B

(81)

(40) (27) (49) (85)



Regional and standard Japanese in conversations 

Both Figure 1 and Table 5 show clear inter-speaker variation. Speaker A used SJ variants much less frequently than Speaker B, even though both have lived in the same town most of their lives and even when they spoke in similar situations. For example, when the interlocutor was the speaker’s mother, Speaker A used SJ (morphological and lexical) variants only 5% of the time (in Conversation A-1), while Speaker B used them 40% of the time (in Conversation B-1); when the interlocutor was the speaker’s superior, Speaker A used SJ variants 62% of the time (in Conversation A-6), while Speaker B used them 85% of the time (in Conversation B-4). The greater use of SJ variants by Speaker B may derive from a variety of reasons. For example, it may suggest that standardization is more advanced among younger generations than older generations. It may also be related to the fact that Speaker B is a tour guide and meets people from different parts of Japan on a daily basis, while Speaker A is working for the family business and meets mainly with local people. Figure 1 and Table 5 also show wide intra-speaker variation in the use of YD and SJ variants. In the case of Speaker A, her use of SJ (morphological and lexical) variants ranged from 5% to 62%, with varying percentages (9%, 13%, and 36%) between the two poles. Speaker B’s use of SJ variants also varied widely from 27% to 85%, with varying percentages (40% and 49%) between the two poles. A close examination of this wide intra-speaker variation suggests that the use of YD and SJ forms is not random but is functionally motivated. I discuss this issue in the following subsections. The results of the questionnaire survey will be discussed in Section 3.4.

3.2

Code-switching or variant choice?

The use of SJ-YD variants by Speakers A and B observed in the preceding subsection indicates that, while the influence of SJ on YD is evident, speakers do not code-switch between YD and SJ as distinct varieties, or linguistic systems (cf. Blom and Gumperz 1972; Giacalone Ramat 1995). Rather, the speakers use YD and SJ variants in variously differing proportions in different conversations. In none of the conversations do the speakers use YD or SJ variants exclusively. Further, a close examination of the data shows that speakers mix YD and SJ forms in a quite complex way. As shown in Table 5, YD and SJ variants were not used in the same ratio for morphological and lexical variables in the same conversation. Generally, in each conversation both speakers used SJ morphological variants less frequently than SJ lexical variants. Although more data need to be examined, this difference may suggest that as far as the variables examined here are concerned, morphological variables are standardized to a lesser degree as compared to lexical variables. The data also show that the mixing of YD and SJ variants occurs both inter- and intra-sententially in the same conversation. This is illustrated by the following three examples. I discuss only the speech of Speakers A and B in these examples. (In the following examples, YD variants are in boldface and SJ variants are underlined. Forms

 Shigeko Okamoto

that are not marked at all are thus regarded as those that do not have YD-SJ variants, or are shared by both YD and SJ.) Example (1) illustrates the inter-sentential mixings of YD and SJ forms. This example is taken from Conversation A-6. Among the six conversations examined, Speaker A used SJ variants most frequently in this conversation (Table 5) but still used YD variants occasionally.3

(1) (From Conversation A-6; Speaker A, talking with Speaker G, her daughter’s former high school teacher, at G’s home; A is expressing her gratitude to G, who agreed to record her conversation with A.) 1 A: ma::, watashi:: ano:: honto (.) = komatchotta= n desu i ne. ‘Oh, I was really at a loss.’ 2 G: =doozo= doozo doozo. ‘Please, please, please.’ 3 A: = =sensee ni honto (.) tasukete moratte kara (.) konna ni:: ureshii koto wa arimasen. [laugh] ‘I’m so happy that you are helping me.’ 4 G: maa, sonna ni, ano:: (.) yappari nante iimasu ka nee, (1) watashi mo (1) kodomo-san-tachi to sessuru koto ga ne, ima made ni saisai arimashita kara. ‘Oh, such, uh, what should I say, I have often had occasions to get in touch with the children, so...’ 5 A: =soo desu ne.= ‘That’s right.’ 6 G: =nani ka go-en= ga aru yoo na ne, kanji ga. ‘I felt we are somehow connected.’ 7 A: kitto kyooryoku-shite moraeru to omotte, ‘I thought you would certainly cooperate with me.’ 3. The following transcription conventions are used in this study. . falling intonation , slight fall indicating continuity ? rising intonation - truncation :: lengthened segment = xxx = “xxx” overlapping with “yyy” = yyy = = = xxx latching (.) pause of less than 1 second (n) pause of n seconds (xxx) a good guess at an unclear expression (e.g. (ashita)) [xxx] nonlexical phenomena (e.g. [laugh])



Regional and standard Japanese in conversations 

In line 1 in example (1), two YD morphological variants (-cho instead of -te, and i, an emphatic morpheme) are used; in line 3 one SJ morphological variant ((mor)atte instead of (mor)oote) and one SJ lexical variant (konna ni instead of konee ni) are used; and in line 7 one SJ morphological variant ((omo)tte instead of (omo)ote) is used. The use of YD and SJ variants in example (1) may suggest that the two kinds of variants cannot be used simultaneously in the same sentence, but that is not the case, as illustrated by the intra-sentential mixings in example (2).

(2) (From Conversation B-1; Speaker B, talking with Speaker H, her mother; B is talking about a television program about pearl oysters that she watched recently.) 1 B: mi,= minna mi ni= tsukete:: ippai tsukechoru kedo are wa kai ga ‘Everyone is wearing them (pearl accessories), wearing them a lot, but that is because the shellfish (pearl oyster) …’ 2 H: =ha:: ne,= ‘yeah, right’ 3 B: itakute itakute (gan-) koo nokeyoo to shite,= mawari= ni tsukete sa, ‘had a lot of pain, and so tried to get rid of (the pain) and formed (the protection) around (it),’ 4 H: =un= u::n. ‘yeah, yeah’ 5 B: yareyare (.) yatto itami ga osamatta to omottara, kuchi akete korosareru no wa kawaisoo ya= na::= to omotte, ha, ohanashi o kiichotta. ‘when (the shellfish) thought the pain is finally gone, its mouth gets opened and it gets killed, and I was listening to the story feeling sorry.’ Lines 1, 3, and 5 in example (2) constitute one long sentence and contain four YD variant tokens – i.e. three morphological tokens (two tokens of -cho instead of -te and one token of -ya instead of -da) and one YD lexical variant (noke ‘get rid of ’ instead of torinozok(u)) – and four SJ variant tokens – i.e. four morphological variant tokens (two tokens of (it)akute ‘painful’ instead of (it)oote, one token of (omo)tta ‘thought’ instead of (omo)ota, and one token of (omo)tte ‘think (and)’ instead of (omo)ote).4 Example (3) is an excerpt from the conversation between Speakers A and B. It shows that both speakers mixed YD and SJ variants inter- and intra-sententially. What

4. The interactional particle sa ‘you know’ in line 3 may be treated as an SJ morphological variant, but it is not so considered here, because its YD variant form is unclear. Further, some scholars define SJ as a relatively formal variety of Japanese (e.g. Sanada 1996). Such a definition would not treat sa as part of SJ because it is used only in very informal conversations. This kind of problem is one of the reasons why it is difficult to specify exactly what constitutes SJ (or YD or any other regional variety).

 Shigeko Okamoto

is particularly noteworthy is the fact that for the copulative auxiliary the same speaker (B) mixed the YD variant ya and SJ variant da in the same sentence (line 2).

(3) (From Conversation A-5/B-4; Speaker A, talking with Speaker B, a friend of A’s daughter, at B’s house.) 1 A: Kojima-san nara koso:: (2) kyooryoku shite moratte honto kansha shichoru yo? ‘I’m grateful that you are helping me (as I thought you would).’ 2 B: ie ie ie, T-san mo sugoi benkyoo gambatchotte mitai ya kara, ma watashi ga kore gurai dattara, watashi mo honto o-tetsudai dekiru n de, ‘No, no. T seems to be studying very hard, and I can help her with (small) things like this, so...’ 3 A: =ureshii= ‘I’m glad.’ 4 B: =itsu demo= itte kudasai. ‘Please ask me any time.’ 5 A: ureshii.(.) honto. (.) Iya, sasuga dookyuu-see ja nee. ‘Glad to (hear that). Really. Well, that’s (how) classmates (are).’ The foregoing analysis demonstrates that it is difficult to identify YD and SJ as two distinct varieties because the use of YD and SJ variants is a matter of degree, and because there are numerous expressions that are shared by both varieties,5 as noted earlier. Although I have analyzed speech from only two speakers here, a glance at the rest of the data set indicates that the other speakers also mixed YD and SJ variants in varying proportions. My previous study of Osaka dialect (OD) also revealed a finely graded use of OD and SJ variants in different social situations (Okamoto 2007, 2008). While I do not mean to suggest that these speakers’ practices can be generalized to all dialect speakers in Japan, they well illustrate the non-discrete nature of speech varieties. In light of these observations, I employ the term variant choice instead of code-switching to refer to the use of regional and Standard variants as observed in this study (as well as in my previous study of OD). A similar observation is made by Maelum (1996) in relation to Norwegian. In her critique of Blom and Gumperz (1972), she argues that the use of “discrete code-switching strategies” is rather unusual among speakers of regional dialects in Norway. She then states that their speech is mostly characterized by “a mixture of dialect and standard forms,” and that “context-conditioned variation” is “manifested for the most part as a matter of degree, with relative proportions of dialect and standard variants changing

5.

These expressions probably constitute the majority, although no statistical data are available.



Regional and standard Japanese in conversations 

from context to context, with marked individual differences” (Maelum 1996: 753).6 This Norwegian situation seems similar to the Japanese case examined in this study. The notion of variant choice employed here seems to enable us to capture the gradient use of variants more accurately than the notion of code-switching, which must assume the existence of two distinct codes as linguistic systems. As we saw earlier, neither Speaker A nor B used YD or SJ variants exclusively in any of the conversations examined. On the basis of this fact, one may say that in all conversations the two speakers used a hybrid variety, or neohoogen, that is, a code that lies in between hoogen ‘dialect’ and hyoojun-go ‘Standard language’ (Sanada 1996, 2000). Such an account, however, cannot fully capture the considerable differences in the proportions of YD and SJ variants used in different conversations, such as those observed above – differences that seem to be functionally motivated (see below for further discussion). Moreover, the notion of hybrid variety must assume the existence of two pure varieties (i.e. pure YD and pure SJ), which is difficult to do because of the presence of many forms shared by both varieties (i.e. not variables) and because of the ongoing process of standardization. As mentioned earlier, diachronic change toward standardization has been recognized in many regional dialects. Such change (real-time variation) then would manifest as synchronic inter-speaker or inter-generational variation (apparent-time variation). Due to this ongoing process, it is not always clear whether a particular linguistic expression is a YD or SJ form. For example, among the speakers in the present data set, the older generations, including Speaker A, used YD variants much more frequently than the younger generations, such as Speaker B. Some of the YD expressions are used only by oldest generations. Speaker A, for instance, used conjunctions such as (he)jakedo ‘but’ and hete ‘and, so’ and adverbs such as donee ‘how’ and sonee ‘such’ frequently, but Speaker B did not use them at all. In such cases, the new, or SJ, forms (e.g. kedo ‘but’, doo ‘how’) may be supplanting the old, or YD, forms. When most people in the community have come to use the new (or “SJ”) forms regularly even in informal conversations, they may no longer be perceived as SJ forms (i.e. as something foreign). This perception seems to depend on the degree of standardization of a particular form and may also vary among individuals. Moreover, even those who use both YD and SJ variants of a particular variable may not necessarily perceive them as YD and SJ forms if both are used regularly for different purposes (see below for further discussion). In sum, what is considered YD or SJ forms may vary across individuals and across time, which also makes it difficult to identify YD and SJ as pure varieties. Terms such as hoogen and neo-hoogen may be useful, and perhaps even necessary, to refer to broad differences between different “varieties.” However, they presuppose 6. Maelum (1996:758) uses the term “variant-switching,” by which some expressions or grammatical forms from the standard are introduced into otherwise dialectal utterances. I do not employ this term, because my data indicate that it is often difficult to identify one variety as the basic variety, as observed in Maelum’s data.

 Shigeko Okamoto

the existence of distinct speech varieties, or linguistic systems. The foregoing discussion suggests they are, rather, idealizations of linguistic categories vis-à-vis the synchronic and diachronic continuum of differences. (In this respect, my own use of such terms as YD and SJ is also problematic, or self-contradictory, but as mentioned earlier, they are used as convenient labels for referring to broad differences.)

3.3

Style management: Interface between regionality and formality

The YD-SJ variant choice seen above can be said to contribute to indexing the speaker’s regionality to a greater or lesser extent. However, indexing regionality does not seem to be the only function of variant choice. One possible factor that may affect variant choice is the interlocutor’s dialect. If the interlocutor is an SJ speaker, it may affect the speech of speakers of a regional dialect (e.g. Hosotani 2004), although the extent of this effect is likely to depend on the speaker’s attitudes toward his or her dialect and toward SJ. In the present data, however, the interlocutors of Speakers A and B in all conversations are from the same region and are not speakers of SJ. Thus the interlocutor’s dialect does not seem to be the major factor. Consideration of each conversational context indicates that the variant choice in question is also related to formality of speech style (cf. Sanada 1996, 2000). I regard formality of speech style as a sign of the speaker’s stance regarding the question of to what extent it is appropriate to appear restrained/ controlled or relaxed/spontaneous in a given situation. This assessment then is made based on the consideration of various aspects of the context, including the relationship between the speaker and the interlocutor (e.g. degree of intimacy, status difference), domain (e.g. work, home), speech act type (e.g. apology, greeting), topic, and emotion (e.g. anger, surprise). A variety of linguistic devices can contribute to making speech style more or less formal (or informal), such as honorific or non-honorific expressions and particular lexical items (e.g. honjitsu vs. kyoo for ‘today’). It seems regional and Standard variants can also be employed for this purpose. In the present data, when Speaker A talked with her family members or friend, she used SJ variants only infrequently, but she increased their use considerably when talking with her daughter’s friend, who is not close to her, and she increased their use even more when talking with someone who could be considered to be her superior (her daughter’s former high school teacher, who is older than she is). Likewise, Speaker B varied the proportion of YD and SJ variants according to her interlocutor. Further, Speakers A and B mostly used SJ variants when they were greeting their interlocutors at the beginning of some of the conversations (Conversations A-6, B-3). Other aspects of the contexts such as domain and genre may also be relevant. For example, in my OD data, SJ variants were used more frequently at work than at home (Okamoto 2007, 2008). The above observation is partly supported by the fact that the use of SJ variants is related to the use of honorific expressions, which also contributes to bringing about different degrees of formality (Okamoto 1999). Table 6 shows the use of SJ variants and addressee honorifics (the copulative auxiliary -desu and the verb ending -masu and



Regional and standard Japanese in conversations 

their variants) by the two speakers in each conversation. The percentage for addressee honorifics is based on their use against the total number of tokens where addressee honorifics could potentially be used. As can be seen in Table 6, both Speakers A and B used addressee honorifics frequently in the conversations in which they also used SJ variants most frequently. That is, when honorifics are used frequently, SJ variants are also used frequently. This, however, does not mean that the amount of SJ variants exactly parallels the amount of honorific use. Even when SJ forms are used relatively frequently, as in Conversations B-3 and A-5, honorifics may not be used equally frequently.7 Note also that it is not the case that SJ variants always index formal speech and YD (or any regional) variants always index informal speech. Within SJ, for example, different degrees of formality can be created by differential uses of honorific and non-honorific expressions. Further, it does not mean that Speaker A’s speech is more informal than Speaker B’s simply because she uses more YD variants in like circumstances. It seems that when a speaker uses both YD and SJ forms as variants of a certain variable (e.g. V-oote vs. V-tte), the YD variant tends to be used more in informal speech and the SJ variant in formal speech.8 Moreover, for such speakers the distinction between standardness (or regionality) and formality may not be clearly made, as discussed above. Table 6.  Use of SJ variants and honorifics (in percentage) Conversations

SJ variants

Addressee honorifics

Speaker A A-1 (mother) A-2 (older sister) A-3 (daughter) A-4 (close friend) A-5 (daughter’s friend) A-6 (daughter’s teacher)

5% 9 13 5 36 62

0% 0 0 0 8 83

Speaker B B-1 (mother) B-2 (father) B-3 (close friend) B-4 (friend’s mother)

40 27 49 85

0 0 1 88

7. The fact that the use of honorifics does not exactly coincide with that of SJ variants may be related to a variety of factors, including the degree of standardization of one’s speech. It is an interesting issue that requires further research. 8. This is only a tendency. There are SJ expressions that are used only in informal conversations. The interactional particle sa ‘you know’ is such an example. However, these expressions may or may not be considered as part of SJ, depending on how one defines SJ, as discussed in Note 5.

 Shigeko Okamoto

Characterizing the use of Standard and regional variants as contributing to creating different degrees of formality as the speaker’s stance toward a given context implies that variant forms are only indirectly related to specific contextual features rather than each variant directly indexing a particular contextual feature, such as the relationship between the conversational participants, domain, or speech act type (see Ochs 1993 for a discussion of indirect indexes). One may, for example, characterize a YD variant as a direct index of intimacy, or uchi ‘in-group’ relationship between the participants, and an SJ variant as a direct index of non-intimacy, or soto ‘out-group’ relationship, as analogous to ‘we-code’ and ‘they-code’, the terms used by Gumperz (1982).9 Or, a YD variant may be regarded as a direct index of a certain domain, such a familial context. According to this approach, one’s linguistic choice must be an automatic response to a particular contextual feature. Thus, if one is talking to an out-group person, one must consistently use SJ variants. Such a characterization cannot adequately account for the simultaneous use of YD and SJ forms in the same conversation or even in the same sentence, as observed in this study. Nor can it account for the creative or strategic use of variants to define, or bring about, a particular context. Further, the notion of indirect indexing allows one to take into consideration the role of ideology (Silverstein 1979; Woolard 1992) in language practice. While individual speakers’ language ideologies, or beliefs and attitudes toward language use, may be greatly influenced by societal norms, they may vary to a greater or lesser extent and lead to different language practices (see also Okamoto 1999, 2004). Thus, in the case of variant choice, even in the same social situation, individuals may take different stances with regard to formality because of different ideologies, and hence use different linguistic forms. Further, I argue that while specific contextual features may not directly constitute pragmatic meanings of YD and SJ variants, they do contribute to shaping such meanings. If, for example, YD variants are repeatedly used in certain kinds of contexts – contexts in which the speaker thinks it is acceptable to be relaxed/spontaneous, or informal, such as when talking with a family member or close friend – this usage, or the meaning of informality, would come to be associated with YD variants as their pragmatic meaning. Once the standardization of a particular variable becomes complete and the SJ variant form replaces the YD variant form, then the two forms will cease to have different meanings. In other words, the meanings of variants are inseparable from their use in interaction, and they seem to be in a dialectic relationship with the context of use. Note that there may be cases in which a particular variant choice appears to directly signal a certain discourse-related function, such as joking, emphasizing, and expressing a strong emotion, as in the case of bilingual code-switching (see Auer 2000, Gardner-Chloros et al. 2000, and Gumperz 1982, among others). For example, regional variants may be used to express one’s anger. But I would argue that even in such a case, the meaning of anger is indirectly indexed by the use of regional variants in that 9. Note that the term ‘code’ in ‘we-code’ and ‘they-code’ refers to a variety as a linguistic system, and not a single variant form.



Regional and standard Japanese in conversations 

the informality/spontaneity associated with such variants is thought to be suitable for expressing anger in a given context. In other contexts, the use of SJ variants, or the sense of formality associated with them, may be considered more effective for the same function.

3.4

Variant choice and style management: Negotiating competing motivations

The foregoing analysis demonstrates that YD and SJ variants can be used as a resource for style management, as it relates to formality, although individuals may differ considerably in their usage. The question remains as to why the two kinds of forms are used in the way they are. I would like to consider this issue below in relation to the sociopolitical and historical context and the role of language ideology (Gal 1992; Hill 1992; Woolard 1985, 1992). As mentioned earlier, the standardization of Japanese has been enhanced by a number of factors associated with Japan’s modernization (Sanada 1991, Shibata 1988, 1999 [1965], Yasuda 1999). In particular, the government’s language policy seems to have played an important role. Considering language standardization essential to the development of the modern nation-state, the Meiji government (1868–1912) attempted to establish hyoojun-go, or Standard Japanese, on the basis of the variety (presumably) spoken by “educated Tokyoites” (Lee 1996, Sanada 1991, Shibata 1999 [1965]), while at the same time promoting the hoogen bokumetsu ‘eradication of regional dialects’ movement. During this time, regional dialects were regarded as “incorrect,” “bad,” “vulgar,” etc., while SJ was considered “correct,” “beautiful,” “decent,” etc. Speakers of regional dialects were punished or ridiculed, and even the speakers themselves often evaluated their own dialect negatively (Lee 1996, Sanada 1991, Yasuda 1999). The hoogen-bokumetsu movement was most active until the end of World War II, but as Yasuda (1999) argues, the ideology underlying this language policy persisted even after World War II. After the war, the term kyootsuu-go ‘common language’ came to be preferred to the term hyoojun-go ‘Standard language’, which was thought to project a strong regulatory image (Shibata 1997, 1999 [1965]). However, Yasuda contends that the notion of kyootsuu-go still assumes some kind of standard, and that the “democratization” of language after the war meant the “democratization” or “popularization” of the Standard, not regional dialects. Thus, although it is said that today Japanese have become more tolerant toward regional dialects (Shibata 1988), as seen in survey results (Inoue 1988, 1995, Kanazawa 1991), SJ (or Tokyo dialect) continues to be evaluated more highly than regional dialects, particularly on what Inoue (1988, 1995) calls “intellectual image.” According to my own survey of nine (out of ten) participants in the conversations examined in this study (see Section 2), the following four descriptors were the most commonly chosen answers to the question, “What kind of language do you think Standard Japanese is?”: It is a relatively polite language (7 respondents); it is used on formal occasions (7); it is used in writing (7); and it is correct language (6). All other descriptors had three or fewer responses; interestingly, only 2

 Shigeko Okamoto

respondents chose the descriptor “it is a language used by people in the Kanto region, especially those in Tokyo.” People’s evaluations such as these seem to reflect the longlasting effect of the government language policy in the history of modern Japan, including the fact that it is used in the nation’s capital, education, media, public speech, etc. Yet, regional dialects do not seem to be disappearing. One important factor for this seems to be speakers’ provincialism, or their attachment to their region (Inoue 1988, Shibata 1988). Further, the general public are said to have become more tolerant of or positive toward the use of regional dialects as part of the resurgence of regionalism in Japan since the 1980s. Consistent with this change, government language policy has also shifted since the late 1980s from the promotion of dialect eradication to the recommendation of differential use of SJ and regional dialects according to the situation (Carroll 2001, Yasuda 1999). Surveys report that regional dialects tend to be viewed positively on what Inoue calls “nostalgic image” (e.g. Inoue 1995, Kanazawa 1991). According to the aforementioned survey I conducted, the following four descriptors were the most commonly chosen answers to the question, “What kind of language do you think Yamaguchi Dialect is?”: it has feeling (9 respondents); I use it in informal situations such as when talking with family members or close friends (7); it has a relatively warm feeling to it (6); and I feel attachment to it (6). All other descriptors had one or no respondents except for the descriptor “it is relatively rough,” which had 4 respondents. Although the responses to surveys cannot be considered to accurately correspond to actual language use, the differential use of YD and SJ variants observed in this study may in part reflect people’s attitudes toward the two “varieties” as discussed above. Thus, for example, the use of regional forms may serve to (indirectly) index the speaker’s perception of the nature of interpersonal relationships (e.g. intimacy/solidarity or lack of it). Accordingly, the encroachment of Standard forms in informal conversations meets more resistance. That is, regional forms continue to be used in certain domains as a force competing with the dominant Japanese language ideology. In this respect, YD may be considered a ‘we code’ and SJ a ‘they-code’ (Gumperz 1982). But it is to be emphasized that this association between a code/variety and a social group is only a symbolic one, as pointed out by Gumperz (1982: 66). As discussed earlier, it cannot predict each specific variant choice. The use of regional and Standard variants seen in this study suggests that SJ is neither accepted nor rejected fully by speakers of YD. This kind of language practice then cannot be adequately explained in terms of either the consensus model of society, whose members are assumed to share the same norms and values, or the conflict model of society, which is characterized as consisting of two (or more) groups with fundamentally different norms and values. (See Milroy and Milroy 1998, Rickford 1986, and Woolard 1985 for discussions of these models.) These models might be applicable to stable communities, but not suitable for more fluid communities. The linguistic behavior observed in this study may be regarded as a result of speakers’ creative responses to cope with the tension between competing social values. That is, they are a result of



Regional and standard Japanese in conversations 

their negotiation between global and local linguistic norms (see Gal 1992 and Woolard 1985, 1992 for relevant discussions).

4. Conclusion The analysis presented in this chapter has demonstrated that the speakers used regional and SJ variants in a complex manner by mixing them in differing proportions depending on the context. I have argued that such use of variants is not random but is functionally motivated, and that it can be a resource for style management as it relates to formality of style. Further, I hope to have shown that the kind of linguistic practice observed in this study cannot be fully accounted for by static descriptions based on the notion of code-switching, which must assume that varieties such as YD and SJ are discrete codes. Rather, I have argued that the notion of variant choice can better capture such linguistic practice because it allows us to describe the interface between regional and stylistic variation in terms of the notion of a dynamic continuum of (synchronic and diachronic) differences. As noted earlier, I am not claiming that the speech behavior described in this study is applicable to all speakers of regional Japanese, although it does not seem to be an uncommon one. There may also be circumstances in which speakers of regional dialects apply different strategies. For example, in the present study the interlocutors of Speakers A and B are all from the same region; it would be interesting to see how these two speakers use YD and SJ variants when talking with speakers of SJ or other dialects. Speakers A and B have lived in the same town most of their lives, but speakers who have lived elsewhere may show different speech patterns. Further, it is often noted that speakers of different dialects have different attitudes toward their own dialect and SJ, and this difference may affect the use of variants. Further research is called for, involving more speakers of different dialects interacting in different social situations.

References Abe, T. and Sakaguchi, N. 2002. “Tsugaru-hoogen washa no sutairu kirikae (Style switching of Tsugaru dialect speakers).” Handai Shakai-gengogaku Kenkyuu-nooto 4, Tokushuu, Sutairu Kirikae 1 (Osaka University sociolinguistics research notes, no. 4, Special issue on style switching 1): 12 -32. Auer, P. 2000. “A conversation analytic approach to code-switching and transfer.” In The Bilingualism Reader, L. Wei (ed.), 166–187. London: Routledge. Blom, J.-P. and Gumperz, J. J. 1972. “Social meaning in linguistic structure: Code-switching in Norway.” In Directions in Sociolinguistics: The Ethnography of Communication, J. J. Gumperz and D. Hymes (eds.), 407–434. New York: Holt, Rinehart and Winston. Carroll, T. 2001. Language Planning and Language Change in Japan. Surrey: Curzon Press.

 Shigeko Okamoto Fureai Yamaguchi. 2002. Newsletter of Yamaguchi Prefecture, 6–7. Gal, S. (1992). “Multiplicity and contention among ideologies: A commentary.” Pragmatics 2: 445–449. Gardner-Chloros, P. Charles, R. and Cheshire, J. 2000. “Parallel patterns?: A comparison of monolingual speech and bilingual codeswitching discourse.” Journal of Pragmatics 32: 1305–1341. Giacalone Ramat, A. 1995. “Code-switching in the context of dialect/standard language relations.” In One Speaker, Two Languages: Cross-Disciplinary Perspectives on Code-Switching, L. Milroy and P. Muysken (eds.), 45–67. Cambridge: Cambridge University Press. Gumperz, J. J. 1982. “The sociolinguistics of interpersonal communication.” In Discourse Strategies, J. J. Gumperz (ed.), 9–37. Cambridge: Cambridge University Press. Hill, J. H. 1992. “Today there is no respect: Nostalgia, ‘respect’ and opposition discourse in Mexicano (Nahuatl) language ideology.” Pragmatics 2: 263–280. Hosotani, F. 2004. “Osaka-shi-hoogen washa no sutairu kirikae (Style switching of speakers of Osaka City dialect).” Handai Shakai-Gengogaku Kenkyuu-Nooto 6, Tokushuu, Sutairu Kirikae 3 (Osaka University sociolinguistics research notes, no. 4, Special issue on style switching 3): 12 -32. Inoue, F. 1988. “Dialect image and new dialect forms.” Tokyo University of Foreign Languages, Area and Culture Studies 38: 13–22. Inoue, F. 1995. “Classification of dialects by image – English and Japanese.” ZDL-Beiheft 77: 355–368. Kanazawa, H. 1991. “Gengo-ishiki to hoogen (Language attitudes and dialects).” In Shin-Hoogen-Gaku o Manabu Hito no tame ni (For Those Who Learn the New Study of Dialects), M. Tokunaga and S. Sanada (eds.), 117–131. Kyoto: Sekai-shisoo-sha. Kobayashi, T. and Shinozaki, K. (eds.). 2003. Gaidobukku: Hoogen Kenkyuu (A Guidebook: Dialect Research). Tokyo: Hitsuji-shobo. Lee, Y. 1996. Kokugo to yuu Shisoo (The ideology of “the National language”). Tokyo: Iwanami Shoten. Long, D. 1996. “Quasi-standard as a linguistic concept.” American Speech 71: 118–135. Maehlum, B. 1996. “Codeswitching in Hemnesberget: Myth or reality?” Journal of Pragmatics 25: 749–761. Milroy, J. and Milroy, L. 1998. “Mechanisms of change in urban dialects: The role of class, social network and gender.” In The Sociolinguistic Reader: Vol. 1, P. Trudgill and J. Cheshire (eds.), 179–195. London: Arnold. Miyake, Y. 1995. “A dialect in the face of the standard: A Japanese case study.” Proceedings of the 21st Annual Meeting of the Berkeley Linguistics Society, 217–225. Nakayama, K. 1982. “Yamaguchi-ken no hoogen (The dialect of Yamaguchi prefecture).” In Kooza Hoogen-Gaku: Chuugoku, Shikoku-Chihoo no Hoogen (Seminar in Dialectology: Dialects of the Chuugoku and Shikoku Areas), K. Iitoyo, S. Hino and R. Sato (eds.), 142–173. Tokyo: Toshokankokai. Ochs, E. 1993. “Indexing gender.” In Sex and Gender Hierarchies, B. D. Miller (ed.), 146–169. Cambridge: Cambridge University Press. Okamoto, S. 1999. “Situated politeness: Manipulating honorific and non-honorific expressions in Japanese conversations.” Pragmatics 8(2): 51–74. Okamoto, S. 2004. “Ideology in linguistic practice and analysis: Gender and politeness in Japanese revisited.” In Japanese Language, Gender, and Ideology: Cultural Models and Real Peo-



Regional and standard Japanese in conversations  ple, S. Okamoto and J. S. Shibamoto Smith (eds), 38–56. Oxford/New York: Oxford University Press. Okamoto, S. 2007. “Kaiwa ni okeru hoogen to hyoojungo (Regional and Standard Japanese in conversations).” Gengo (Language) 36(3): 53–61. Okamoto, S. 2008. “The use of ‘regional’ and ‘Standard’ Japanese in conversations: A case study from Osaka.” In Japanese Applied Linguistics: Discourse and Social Perspectives, J. Mori and A.S. Ohta (eds), 132–159. New York: Continuum. Rickford, J. R. 1986. “The need for new approaches to social analysis in sociolinguistics.” Language and Communication 6(3): 215–221. Sanada, S. 1991. Hyoojun-go wa ika ni shite seeritsu shita ka: Kindai Nihon-go no hatten no rekishi (How the standard language was established: A history of the development of modern Japanese). Tokyo: Sotakusha. Sanada, S. 1996. Chiiki-Go no Dainamizumu: Kansai-Hen (Dynamism of regional languages: The case of Kansai). Tokyo: Ofu. Sanada, S. 2000. Henyoo-suru Osaka-kotoba (Changing Osaka language). Gengo (Language) 29(1): 49–53. Sanada, S. 2002. Hoogen no Nihon Chizu (A dialect map of Japan). Tokyo: Kodansha. Shibata, T. 1988. Ikite Iru Nihon-Go: Hoogen Tansaku (Living Japanese: Exploring dialects). Tokyo: Kodansha. Shibata, T. 1997. “Hoogen, kyootsuu-go, hyoojun-go (Dialects, common language, standard language).” Nihongo-Gaku (Japanese linguistics) 16(4): 100–108. Shibata, T. 1999 [1965]. “The rise and fall of dialects.” In Takeshi Shibata: Sociolinguistics in Japanese Contexts, T. Kunihiro, F. Inoue, and D. Long (eds.), 183–206. Berlin: Mouton. [Translated from the Japanese original published in 1965]. Shibatani, M. 1990. The Languages of Japan. Cambridge: Cambridge University Press. Shingu, I. 1997. Dynamics of Code-Switching Behavior in the Context of the Kansai Dialect and the Standard Variety in Japanese. Unpublished M.A thesis. University of Massachusetts, Amherst. Takagi, C. 2002. “Koochi-ken Hata-hoogen washa no sutairu kirikae (Style switching of speakers of Hata Dialect in Kochi prefecture).” Handai Shakai-Gengogaku Kenkyuu-Nooto 4, Tokushuu, Sutairu Kirikae 1 (Osaka University sociolinguistics research notes, no. 4, Special issue on style switching 1): 55–67. Woolard, K. A. 1985. “Language variation and cultural hegemony: Toward an integration of sociolinguistic and social theory.” American Ethnologist 12: 738–748. Woolard, K. A. 1992. “Language ideology: Issues and approaches.” Pragmatics 2: 235–249. Yasuda, T. 1999. Kokugo to Hoogen no Aida: Gengo-Koochiku no Seeji-Gaku (Between the national language and dialects: Political science of language construction). Kyoto: Jinbun Shoin. Yoneda, M. 1997. “Survey of standardization in Tsuruoka, Japan: Comparison of results from three surveys conducted at 20-year intervals.” Nihongo Kagaku (Japanese Linguistics) 2: 24–39.

chapter 11

“Involved” speech style and deictic management of spatiotemporal and textual reference A case of ko/so-deictics in Japanese* Kuniyoshi Kataoka

Aichi University, Japan

1. Introduction One way of defining speech style shift, though a broad one, would be that it is an alternative way of saying the same thing with different indexical values based on different contexts and domains of use. Speech style shift has usually been regarded as an individual form of speech management directed toward interactive endeavors, and previous studies of Japanese speech style shift have focused, for example, on a choice between des-/mas- and non-des-/mas-forms (Cook 1996, 1998; Ikuta 1983; Maynard 1991; N. Okamoto 1996; Tokuchi 2002), male and female language styles (Matsumoto, this volume; S. Okamoto and Sato 1992), and regional and standard language varieties (S. Okamoto, this volume). These studies have clearly shown that the shift of such forms is motivated by social and psychological evaluations arising at the time of interaction, rather than being invariably based on a pre-given dichotomy with which such forms are traditionally associated – e.g., formal vs. informal, feminine vs. masculine, or prestigious vs. stigmatized. The choice between these forms has also been discussed in terms of the “discourse modality” associated with certain speech styles (Maynard 1993), as well as in terms of discourse functions that serve as floor-management and/ or context-creating strategies (Ikuta 1983). The causes and effects of style shifting are essentially social in nature. To the extent that a shifting of styles mediates one’s immediate stance vis-à-vis the interlocutor, it * I would like to thank the two editors, Kim Jones and Tsuyoshi Ono, for their patience and insightful comments on previous versions of this paper, and to an anonymous reviewer for pointing out the remaining problems for the final revision. My gratitude extends to Ray Hardesty, who carefully proofread the manuscript. The preparation of the paper was supported by a Japanese Grant-in-Aid for Scientific Research (C) (2) (No. 12610566) and (C) (1) (No. 13610667).

 Kuniyoshi Kataoka

counts as social deixis (see, for example, the framework of deixis proposed by Levinson (1983) as consisting of five categories – person, time, space, discourse [or textual], and social). Within this framework, Japanese social deixis is often accompanied by obligatory lexical agreement with person deixis – Anata/*Temee no onamae wa nan desu ka? (‘What’s your name?’), where the term with the asterisk is a vulgar form of personal reference that is incongruent with the formal des/mas-form.1 In this sense, the social and person aspects of deixis are mutually dependent and facilitative and can be highly strategic communicative resources that carry complex social indices and emotional imports. However, other aspects of deixis (i.e., time, space, and textual) have been unduly slighted as being ineffective resources for speech style shifts. Few studies – except perhaps for those on the Conversational Historical Present (e.g., Johnstone 1988, Sakita 2002, Schiffrin 1981, Wolfson 1978) – have investigated the function of these types of deixis in terms of the embodiment of speech styles. Here, by employing a new type of data and a specific criterion, I consider these neglected aspects of deixis from a viewpoint of “involvement” (Tannen 1989), and I examine how they are implemented in monologic narratives about highly physical and emotional experiences such as rock climbing. So far, the question of speech style shift has mostly been addressed as pertaining to interpersonal transactions, despite the fact that we spend much more time talking to ourselves than to others. This paper thus attempts to supplement such a dialogic orientation with a look at intrapersonal, rather than interpersonal, involvement, which is embodied through the use of “indexicals” (Hanks 1992; Silverstein 1976, 1995; see also Kataoka 2002, 2004 for Japanese deictics).2

1.1

Involvement strategies in narrative

It has been assumed that the “involved” speech style can be differentiated from an ordinary speech style in terms of prosodic, lexical, and syntactic features. Building on Gumperz (1982) and Chafe (1985), Tannen defines it as “an internal, even emotional connection individuals feel which binds them to other people as well as to places, things, activities, ideas, memories, and words (1989: 12).” She regards (conversational) “involvement” as an achievement – not a given – that consists of various “strategies” rather than “features.” Although involvement strategies are assumed to build upon sys1. It is arguable, however, that these second-person pronouns in Japanese are really “pronouns.” Many linguists have claimed that they are instead “ordinary nouns” because they have different social values and context-dependent uses, unlike a more neutral personal pronoun like you. 2. In this study, I intended to exclude such social factors as power and status, gender, regional background, etc., by looking at the monologic recalls (where there are no direct interlocutors) and to try to stay as close as possible to the shifts of intrapersonal involvement strategies. This is the primary reason that I focus on the spatio-temporal and textual aspects of deixis and their discourse functions related to style shifting.



Deictic management of speech styles 

tematic interactional control, they do not necessarily rely upon conscious planning or plotting (1989: 15). Involvement is achieved by two major strategies, “sound” and “meaning,” (Tannen 1989: 17), and one or more subcategories of these two strategies will be observed in any conversation. Specifically, these subcategories are as follows. Sound includes (1) rhythm, (2) patterns based on repetition and variation of phonemes/morphemes/words/collocations of words/longer sequences of discourse, and (3) style figures of speech. Meaning includes (1) indirectness, (2) ellipses, (3) tropes, (4) dialogues, (5) imagery and detail, and (6) narratives.

Many of these would actually overlap, but I will mainly deal with “patterns based on variation of morphemes” and “imagery and detail” created in “narratives,” which contribute to “an emergent sense of coherence” and “emotional identification.”3 Involvement is usually seen as an interpersonal endeavor, but it could be purely personal. For example, Chafe (1985: 116) classified involvement into three types – (1) self involvement of the speaker, (2) interpersonal involvement between the speaker and the hearer, and (3) the speaker’s involvement with what is being talked about – although it is quite likely that they occur intertwined to some extent. Due to the nature of monologic discourse examined here (not in the sense of Bakhtinian monologue), this study mostly concerns the first and third types. In spite of these seemingly clear definitions, a rationale for naturalizing the concept of “involvement” often comes from an intuitive feeling, such as “the speaker is ‘involved’ because he or she employs such-and-such strategies,” or vice versa. As Besnier (1994) pointed out, we obviously cannot escape from a certain circularity: is involvement a cause of the use of particular strategies, or does the use of these strategies lead to the representation of involvement? When we look at the empirical side of style shifting, we face the same question: do we style-shift to encode different social perceptions, or are different social perceptions achieved by using different styles? Again, it is not very meaningful to ask which precedes which. Even if we admit that these relationships are inevitably circular and mutually dependent, the major criteria for appreciating an involved speech style have rested upon the researcher’s own interpretations of discourse features manifested through sound and meaning. Here, instead of taking some discourse strategies as pre-givens and letting the notion of involvement run circularly, I focus on an empirical base that will engender and achieve natural involvement. I thus employ an alternative, objective scale for involvement – a “difficulty” factor for a bodily performance – under the hypothesis that psychological/emotional involvement co-varies with bodily involvement at the time of 3. Specifically, Tannen (1989) regards involvement as consisting of several properties such as “a joint production,” “intertextuality,” “collusion,” “an emergent sense of coherence,” and “emotional identification.”

 Kuniyoshi Kataoka

peak performance, if not vice versa. As a representative activity of such occasions, I chose rock climbing. Conveniently enough, rock climbers refer to an internationally established grading system to evaluate the difficulty of routes. Below, I look at the degrees of co-variation between route grades and language use, instead of taking it for granted that an involved speech style is spontaneously realized by using certain strategies. I then recapitulate that some strategies actually originate in, and emanate from physical/psychological involvement.

1.2

Japanese deixis

In cross-linguistic terms, Japanese deixis has a widely acknowledged three-term system, representing the “proximal,” “medial,” and “distal” domains, which are respectively realized by the ko-, so-, and a- prefixes (Anderson and Keenan 1985, Sakuma 1951). Specifically, functions of Japanese deixis have been classified into two major types: genba-shiji (mainly space deixis) and bunmyaku-shiji (mainly textual deixis).4 Genba-shiji fundamentally (but not exclusively) relies upon the immediate spatial configuration. Among the languages with a three-region system (e.g., Turkish, Spanish, Latin, Southern Sotho), Japanese is classified as a “person-oriented” deictic system such that the “medial” region (the so-series in Japanese) is identified in relation to the locus of the addressee and, in this respect, differs from, say, Spanish, which has a distance-oriented, “speaker-centered” deixis system (Anderson and Keenan 1985). Therefore, in this classic treatment, an object located near the speaker is most likely to be referred to by the ko-series, while one nearer to the hearer will generally be referred to by the so-series. When it is relatively remote from both participants, it will be identified by the a-series.5 In addition, in a single-participant monologic discourse, Japanese speakers are reported to rely on the distance-oriented deictic system based on their on-site perspective (Nishide 1993). I posit that this formation mostly, though not exclusively, applies to the analysis of the monologic narratives that follow. The hitch, however, is that establishing the territories according to distance turned out to be inexorably indecisive because each domain could be conceptually enlarged or shrunk, depending upon tasks, contexts, and/or types of (inter-)subjective alignment with the audience/objects in communication. Because of this, the focus of research has 4. There are, however, other types of deixis identified that do not clearly fit into the distinction. See, for example, Horiguchi’s (1978) zettai-shiji ‘absolute reference,’ Yoshimoto’s (1992) kansetsu-shiji ‘indirect reference,’ and Morita’s (2001) taishoo-shiji ‘objective reference,’ all of which generally point to the surrounding context or environment of the utterance (based on, arguably, long-term memory (Yoshimoto 1992)). 5. Sakuma (1951) was the first linguist who revealed the intricate deictic system in Japanese (Table i). He formalized it according to the functions they serve in reference, and he identified seven such categories into which these stems are prefixed: (1) thing/object, (2) direction, (3) place, (4) thing/person, (5) state, (6) determiner/demonstrative, and (7) manner.



Deictic management of speech styles 

shifted from the previous notion of Japanese deixis based on the speaker’s and hearer’s relative distances, to the types of alignment between them and to epistemological differences reflected upon the choice of deictic forms (see Kinsui and Takubo 1992b for an overview of the historical development of these ideas; see also Horiguchi 1978, Kamio 1990, Kinsui and Takubo 1992a, Morita 2001, Shooho 1980, and Yoshimoto 1992 for specific theories). However, questions still remain as to the fluidity of referents in such referential practices. On the other hand, the functions of bunmyaku-shiji may be anaphors (items referring to what an NP in the preceding discourse refers to), cataphors (items referring to what an NP in the following discourse refers to), or any exophoras (items bringing the speaker’s attention to (imaginary) objects or locations in the outside world) (Diessel 1999). When the ko- and so-series serve as anaphors and refer to an antecedent, they are referentially equivalent, whereas only the ko-series has a cataphoric function. As to the anaphors, the so-series is usually an unmarked choice, and the ko-series a marked one that exhibits emotional vividness in talking about the object as if it were “visible” and “at the speaker’s side” (Kuno 1973).6 This type of emotion-laden usage of the ko(and a-) series is functionally equivalent to “emotional deixis” (Lakoff 1974) or a “recognitional” use of deixis (Diessel 1999), as seen in a statement such as “That Henry Kissinger sure knows his way around in Hollywood” (Lakoff 1974).

Thing

Direction

Place

State

Demon./ det.

Manner

ko-ko ‘here’

Thing/ person (derog.) ko-itsu ‘this guy’

Proximal ko-

ko-re ‘this’

Medial so-

so-re ‘that’

Distal a-

a-re ‘that (over there)

ko-chira/ ko-tchi ‘this way’ so-chira/ so-tchi ‘to you’ a-chira/ a-tchi ‘that way’

ko-nna ‘like this’

ko-no ‘this ~’

ko-o ‘this way’

so-ko ‘there’

so-itsu ‘that guy’

so-nna ‘like that’

so-no ‘that ~’

so-o ‘that way’

a-soko ‘over there’

a-itsu ‘that guy’

a-nna ‘like that’

a-no ‘that ~ over there’

a-a ‘that way’

Table i.  Japanese deixis and its functions. 6. Kuno (1973: 290) defined inherent properties of ko/so/a deictics, which later developed into discourse-based studies of deictics, as follows: (i) The a-series is used for referring to something at a distance, either in time or space, that the speaker knows both he and the hearer know personally or with which they have a shared experience. (ii) The so-series is used for referring to something that is not known personally to either the speaker or the hearer or that has not been a shared experience between them. (iii) The ko-series is used semi-anaphorically, as if the object being talked about were visible and were at the speaker’s side.

 Kuniyoshi Kataoka

1.3

Discourse functions of deixis

The context dependency of deixis has been widely recognized as “a pain in the neck” by many (formal) linguists. The serious research into the discourse function of deixis is a relatively recent endeavor, and essentially had to wait until the appearance of Brown and Gilman’s (1960) seminal study on the use of European T/V pronouns. In addition, compared with discourse studies on personal pronouns (e.g., Brown and Ford 1964; Duranti 1984; Johnson, Roepcke, and Kataoka 1997; and Mühlhäusler and Harré 1990), those on demonstrative uses have been relatively few in number and are much later developments, with the main focus placed on the linguistic analysis of the information statuses (e.g., Ariel 1988; Gundel, Hedberg, and Zacharski 1993). One seminal study on discourse uses of demonstratives is that of Linde (1979), in which she examines the use of it and that in apartment descriptions. She found that, although being chosen to refer to items of the same sort (typically, “rooms”), it was used to refer to objects within the discourse in focus, while that was used to provide contrast to previous apartments and to preceding out-of-focus referents. (See also a different formation of this/that/it by Strauss [1993], who proposed a gradient schema in terms of high, medium, and low “focus,” which are represented, respectively, by this, that, and it on the referential continuum.) Duranti (1984) also claimed that we need to take into account the interactive meanings associated with discourse participants for the study of pronouns and demonstratives. He found that, in Italian, demonstratives tended to be used for expressing negative feelings, lack of empathy, and emotional distance. The same type of pejorative or distancing function of certain demonstratives is very often observed in political discourse for “other referencing,” as was pointed out with regard to former British Prime Minister Margaret Thatcher’s frequent use of “those” for the projection of negative connotations (Wilson 1990). Glover’s (2000) study of English demonstratives also deals with their semantic extensions in discourse. She examined the choice between proximal and distal deixis in English in negotiation talk and found that the proximals this/these and distals that/ those are not exactly used deictically but rather, attitudinally, paralleling a negotiable vs. received orientation in discourse. Thus, she suggests that new-speaker distal references tend to appear at the start of a turn, prior to invoking a negotiable orientation with proximals. Rubba (1996) also carefully showed that the interpretation of English deictics depends upon various features such as shared norms for behavior, shared cultural maps of an environment, and “scripts” or “scenarios.” For example, when people mention ‘that part of the city,’ it refers to various locations and areas by relying on “space-builders” that map the original ground onto alternate ones, extending the scope of meaning and possible referents. Overall, deictic reference is highly versatile and socio-centric. Hanks (1990, 1992) is the strongest proponent of this line of thinking. His study in a Mayan community clearly demonstrates that the social asymmetries among discourse participants constrain



Deictic management of speech styles 

the range of deictic language use, which responds to the indigenous norms and interpretations of a figure-ground structure created in the mind of co-participants. When we turn to the discourse study of Japanese demonstratives, there are not many precursors. Studies of Japanese deixis have mainly been conducted on the linguistic level, by calculating “acceptability” or “degree of fit” in an idealized context, and only rarely from an empirical perspective that encompasses actual discourse. Exceptions are, for example, Mayes and Ono (1991), who found that ano hito ‘that person’ can often be used negatively, especially when explicitly mentioned. Explicit mention of such demonstrative pronouns, even a socially neutral one such as anata ‘you,’ exhibits specific pragmatic connotations and is highly correlated with conflict use (Jones, unpublished ms.). Hayashi (2004) also investigates the use of a distal demonstrative are ‘that one, that thing’ in Japanese, and claims that, despite delayed projectability in shaping turns due to some typological features of Japanese grammar, speakers can nonetheless achieve early projection of forthcoming action through the use of this “action-projecting” demonstrative pronoun. Another recent study of Japanese demonstratives in discourse is that of Naruoka (2006), who examined some variations of ko/so/a deictics in natural conversation, including kooyuu/sooyuu/aayuu ‘like this/that’ and ko-nna/so-nna/a-nna ‘this/that kind of.’ She showed that these forms represented not only referential functions but also other interactional functions such as hedging, negative evaluation, and/or the expression of surprise. Finding, however, that certain forms are used flexibly across these functions, she also suggests that Hanks’ notion of the “indexical ground” (Hanks 1990, 1992), from which I also gained valuable insights for this study, would better explain the versatile nature of (demonstrative) deictics. The above studies confirm a recent awareness that the emotional and contextual meanings of deixis are constructed, maintained, and extended in ongoing discourse and that we need to consider the specific context of use of deictics.

1.4

Purpose of the study

Although previous studies in Japanese deixis did try to encompass psychological and emotional factors (e.g., Horiguchi 1978, Kinsui and Takubo 1992, Yoshimoto 1992), their analyses depended entirely upon the researcher’s judgments of constructed examples about the degree of fit with the context. The current study, the scope of which is somewhat limited, employs an outer criterion for evaluating discourse participants’ psychological and emotional factors. This criterion, I hypothesize, mediates their physical experience and deictic usage and co-varies with the degree of bodily and psychological involvement. The deictics under discussion are obtained in monologic narratives in which space/time and textual deixis play major roles. This kind of shift in speech style, often subtle and accumulative, was indexed by the systematic choice of deictics, echoing the degree of involvement in the ongoing context.

 Kuniyoshi Kataoka

In this respect, examining monologues has the advantage of minimizing social milieus, which are often confounding factors in the study of deixis, and it offers a primordial site where the neglected aspects of deixis can still serve as a resource for styleshifting. In others words, style-shifting is not necessarily interpersonal nor is it necessarily dependent upon the sorts of social motivations that are oriented toward politeness, gender, regional identity issues, and so forth. It will be worthwhile to show that different degrees of involvement in monologue reflect corresponding linguistic realities in the choice of deictics even in this socially impoverished situation. Specifically, I confirm that the so-series seems to be the preferred choice for most of the narrators and that it contributes to tighter cohesion of text, while the ko-series is preferred for manifesting evaluation, focalization, or the relative peak of involvement, which in rock climbing is often associated with the primary “crux” (the most strenuous or technical part[s]), or with salient “holds” and “moves” on the climbing route. In our case, the speaker covertly managed the ko/so-switching in accordance with the actual physical involvement (crux or non-crux sections) through referentially equivalent but emotionally differentiated space/time/textual deixis. I also claim that style shifting should not be discussed only in terms of purely formal features but also in terms of the relative divergence from the contextual normalcy or unmarkedness in the figure-ground relationship evaluated in situ (Hanks 1992). Put another way, the switch between ko- and so-deictics is an “indexical” means (Hanks 1992, Silverstein 1976) of mediating the degrees of involvement in speech, whether or not the speaker is aware of the effect. This suggests that a more inclusive notion – “situational weight of salience” – will more adequately cover a range of behaviors of deictics, such as ko/so-switching.

1.5

Participants and data

I collected the data on several occasions from 1999 to 2002, based on friendly relationships in my own climbing community. The participants were mostly beginning- and middle-level climbers living in the central part of Japan and were frequent visitors to the same rock-climbing gym. Their ages ranged from the mid-20s to the early 50s. The data were collected on three different occasions, with each occasion dealing with a particular route. The first set of analyses examines the relationship between ko/so-deictics and the physical features characteristic of the four sections in the wall (Figure 1). For this research, 10 climbers (8 men and 2 women) were asked to describe as precisely as possible the hardest route they had climbed at the gym over the past six months, with special attention to the shapes and locations of holds, features of the surface, and movements of



Deictic management of speech styles 

the body.7 First, six climbers (A to F) were asked to describe Route I. Later, two more climbers (G, H) were recruited to describe Route II, and another two (I, J) to describe Route III, in order to confirm the observations obtained from the first set of analyses. During the narration, each participant was separated from the other climbers and placed in a room with a tape recorder. The collected expositions of routes lasted about 3 to 10 minutes and consisted of approximately 40 to 130 clauses for each. On this artificial climbing wall, there are various types of inner and outer holds attached in/on the surface for supporting the body’s weight, in addition to the original texture inscribed and carved on the climbing panels. These are the major physical features referred to in the narrations. For this analysis, the wall is tentatively separated into four sections, with each section demarcated by a series of required moves and different angles from the vertical axis. The beginning and the end of these sections constitute boundaries because climbers quite often take rests at these points to restore energy or prepare for subsequent moves. Sections 2 and 4 are distinct from the other sections because of their characteristic overhangs. My hypothesis is that the occurrence of ko/so-deictics may be closely correlated with the “difficulty” grades of a climbing route. Here I must add that, while a route grade is determined by the hardest “move” on the route, one route usually consists of more than one “crux” with corresponding grades, as indicated by the degrees of shading in Figure 1. The shaded areas in Figure 1 indicate the primary and secondary cruxes on the route, differentiated by the density. The blank area roughly corresponds to an easy section, rated approximately 5.8 to 5.9 on the decimal scale.8 The lightest shade (a) on Routes I and III is equivalent to 5.10- to 5.10+, the medium shade (b) on Routes I, II, and III, to 5.10+ to 5.11-, and the darkest shade (c) on Route II, to 5.11+ to 5.12-. The focus items were those expressions that had explicit (and sometimes implicit) spatio-temporal and textual referents in the narration, which correspond to bunmyaku-shiji and genba-shiji. I also included other deictics that refer to vague and assumed referents such as the general environment or background knowledge – i.e., items called zettai-shiji ‘absolute reference’ (Horiguchi 1978) or taishoo-shiji ‘objective reference’ (Morita 2001) – as far as I was able to make a relatively good guess about them. Specifically, they were ko/so-pronouns (e.g., ko-re/so-re ‘this one/that one’), demonstratives (e.g., ko-no N/so-no N ‘this N/that N,’ ko-nna ‘this kind of ’), and 7. Initially, twelve climbers were recruited, but two of them were later excluded from the analysis because one of them had not completely climbed the route by the time of the narration and the other preferred making the descriptions right in front of the wall, unlike the others who agreed to be secluded in a different room. 8. The grading system used here is usually called the decimal system, which is widely used in the U.S. and elsewhere. The scale ranges from around 5.6 to 5.15, which is currently the hardest grade. There are four finer ratings differentiated for each grade after 5.10 and up, as in 5.10a, 5.10b, 5.10c, and 5.10d, going onto 5.11a, 5.11b, and so forth; no such distinctions exist below 5.10. However, these fine-grained distinctions are not always made by all climbers – some may just prefer, as I basically did here, to divide the scale into three ratings of 5.10-, 5.10, and 5.10+.

 Kuniyoshi Kataoka



Section 4

Section 4

Section 4

Section 3

Section 3

Section 3

Section 2

Section 2

Section 2

Section 1

.

<Side view>

Route I (Climbers A to F)

Section 1

Route II Climbers G, H)

Section 1

130 degrees

90–105 degrees

120 degrees

90–105 degrees

3m

3m

3m

4m

Route III Climbers I, J)

Figure 1.  Climbing wall and described routes Notes: No shade = grade 5.8 ~ 5.9; light shade = 5.10- ~ 5.10+; Medium shade = 5.10+ ~ 5.11-; darkest shade = 5.11+ ~ 5.12-

adverbials (e.g., ko-ko ‘here,’ so-ko ‘there,’ and so-re kara ‘then’), which were employed to explicate recalled spatio-temporal experiences. The a-deictics, which are typically used to refer to an entity conceptually shared with the hearer (e.g., Iwasaki 2002), rarely appeared in the monologues, and thus they were excluded from analysis.9

9. In fact, some narrators used a-deictics, but there were only 18 tokens in all. As far as I could tell from the intonational contours, most of them were the filler anooo ‘we=ll/a=nd,’ and only four or five tokens were genuine deictics.



Deictic management of speech styles 

Also, I decided to exclude the following items: those that were originally deictics but that are grammaticalized and semantically bleached, such as fillers, hesitation markers, or interjections (e.g., anooo ‘well,’ sonooo ‘well,’ are! ‘oh!,’ sonna! ‘what!’); aizuchi ‘back channels’ (e.g., soo soo ‘yeah, yeah,’ aa soo ‘oh, ok’); deictic collocations (e.g., achi kochi ‘here and there,’ are kore ‘this and that’); and restarts or false starts due to truncation (e.g., the first element in sono-, sono hito ‘that–, that person’) – although some of these never occurred.10 Example (1) is an excerpt from the data, and the underlined words/phrases are the cases in point for the analysis.11 (1) 1 Kondo wa, u=, hidari, ichiban hidarigawa ni guree no autaa ga arunde,12 ‘Next, ahh, left, there is a grey outer hold to the left-most side,’ 2 sore ni hidariashi o kake masu. ‘hook (your) left foot on that.’ 3 De, migiashi wa, sono shita ni i= asai innaa ga arunde, ‘And, for the right foot, there is a shallow inner hold below that,’ 4 soko ni migiashi o kakete, ‘(you) put the right foot there,’ 5 e= migi o, migiashi o kyon suru yoona katachi de, ‘a=nd, like (you) “drop-knee” (your) right leg,’ 6 kondo sono saigo no pointo o tsukamini iki masu. ‘and next, (you) go grab that final point.’ 7 Demo kore wa kanari, sonnani mutsukashiku nai desu ne. ‘But this (move) is, quite–, not very hard, you know.’ 8 Migiashi ga kichinto kiite ireba, ‘If (you) get (your) right foot securely locked,’ 9 kiite– kyon shite-reba, ‘locked with the drop-knee move,’ 10 ato wa eiya-tto koo ikuto, ‘and if (you) go swoosh, like this,’ 10. However, I included ko-o ‘this way’ and so-o ‘that way’ because these deictics cannot often be separated from instrumental expressions like kore de ‘with this/this way.’ 11. The notations used for transcription are restricted to the minimum. I only used the following: = lengthening – truncated word , continuous contour . falling contour (( )) author’s comment. 12. There are two types of “hold” (artificial attachments for climbing bolted into/onto the wall) – inner and outer ones. The inner hold is designed to fit into a cup-like concave on the wall, while the outer hold is simply bolted onto the face of the wall.

 Kuniyoshi Kataoka

11 daitai tore masu. ‘(you) could usually grab (it).’ 12 Soko de oshimai desu ne. ‘(The route) ends there.’ As far as we can see from the excerpt, there does not seem to be a regular pattern for the occurrence of these ko/so-deictics. In fact, however, when I looked at a longer stretch of data from a (con)textual perspective, highly principled choices of deictics and of their alignments with particular factors surfaced. Specifically, if we segment the whole text according to the four sections of the wall (Figure 1), which were demarcated by the physical features and their “grades,” particular deictics were preferred at particular locations of the narrated scenes. To confirm the validity of the segmentation of sections and route grades, I also conducted interviews with some of the participants. As Yoshimoto (1992) postulated about the “emotional” use of the ko-deictics in narration, exophoric and endophoric frames of reference become merged through projection onto the navigating agent (who is usually the speaker him/herself) – as in kore on line 7 in example (1), which may be seen as endophorically referring to an antecedent in the preceding utterance or exophorically referring to a physical object (e.g., a hold or a bodily motion recalled in his/her memory). These types of reference are made possible in the envisioned “here-and-now,” where the narrator views the virtual world from the imaginary origo. There, the narrator creates an immediate deictic field based on a relative perceptual distance with regard to referents, whether imaginatively envisioned (in genba-shiji) or textually activated (in bunmyaku-shiji). Therefore, the most important aspect of such usage, I posit, is not the finer classification of deictics into subtypes but the analysis of switches among referentially equivalent but non-referentially valorized choices of ko/so-deictics. For this reason, I exclusively focus on the vicissitudes and ratios of these deictics used by the narrators and do not further subcategorize them beyond the surface ko/so-types (except for a proximal equivalent of ko-deictic ima ‘now’ in Section 2.6).

2. Analysis of monologues – Describing rock-climbing routes 2.1

Route I

The graphs in Figures 2 and 3 show the frequencies of the ko/so deictics used in each section of Route I. The six narratives by Climbers A to F were, respectively, divided into four segments according to the sections of Route 1. That is, when a climber talked about the physical environment and/or bodily movements in Section 1, the ko/so-tokens used there were assigned to Section 1 in the graph. Although the database is rather limited, two general patterns – threshold and incremental – emerged for the occurrence of these deictics, although I would not claim that



Deictic management of speech styles 

these are the only possible patterns of use. In the following analyses, I first examine the frequency factor of the ko/so-deictics across sections, and then I look at the occurrence ratio of these deictics as compared with the number of clauses in each section.13 As for the overall patterns of occurrence, I tentatively call the first pattern a threshold type (Figure 2), in the sense that the first crux (Section 2, Grade 5.10) serves as an initial obstruction that impedes access to the upper region. One could easily imagine that it is frustrating and disappointing indeed to be denied access at such a low level, and I expect that such frustration and disappointment would have motivated the narrators to pay more focused attention to the ways in which they inched through that section than to the other sections. We also see in Figure 2 that, once the narrators managed to get through Section 2, they generally came to use fewer so-deictics to describe the remaining (and even the hardest) sections. For these participants, the medial so-deictics seem to serve as a referential cohesion device that encodes the medium salience (in terms of difficulty) adequate for the 16 14 12 10 8 6 4 2 0

10

15 11 8 5

8 So Ko

1 0 0 0 1 (25) 2 (35) 3 (20) 4 (33) Section

5

4 2 0

(a) Climber A 12 10 8 6 4 2 0

6

1

2

3

So Ko

1

0 0 0 1 (10) 2 (13) 3 (10) 4 (12) Section

(b) Climber B

10 6

7 4

So Ko

0 0 0 0 1 (13) 2 (24) 3 (12) 4 (15) Section

(c) Climber C(

Figure 2.  Ko/so-frequencies in each section for Route I (the threshold type) Note: The number in the parenthesis on the horizontal axis indicates the number of clauses used in each section

13. Here, a clause is defined as a phrase that consists of a verb and the accompanying elements.

 Kuniyoshi Kataoka

secondary crux. Example (2) is an excerpt of Climber C’s description of Section 2, in which he explicated the continuous “moves” (body movements) with a temporal connective (sore de ‘then’), a locative (soko ‘there’), an adnominal demonstrative (sono ‘that ~’), and an anaphoric pronoun (sore ‘that [one]’). All of these expressions (underlined in text) refer to some aspects of the time and space evoked in the narration, augmenting the spatio-temporal cohesion. (2) 1 Ee, eee, sore de, ashi wa sono hangu shita no, hando hoorudo no, e=, e=, e=tto, shita ni migiashi o nosete, ‘Ah, ahhn, then, as for (your) foot, ah, ah, we=ll, put (your) right foot on the handhold below, under that overhang,’

2 soko de mata nosekaete, ‘and change (your) feet again there,’ 3 hidarite agete, ‘and raise (your) left hand,’ 4 de hida–, ee=tto migite no innaa totte, ‘and, take the inner with (your) lef–, right hand’ 5 sore de tachiagarinagara, ‘and as (you) stand up with that (one),’ 6 hidarite de sono ue no innaa o totte, ‘grab the inner above that with (your) left hand,’ 7 soko de mata mochikaete, ‘and switch (your) hands there again,’ 8 e=tto migi ni mochikaete, ‘a=nd, switch to (your) right hand,’ 9 hidari naname ue no innaa totte,… ‘and grab the inner on (your) upper left ….’ Noticeably, Figures 2 (a) and (b) reveal an interesting tendency for the ko-series to occur in Section 4 (the primary crux: Grade 5.11-), but not in the other sections. The relative dearth of ko-deictics may be partially explained by the nature of the current activity, in which the narration relied solely on the auditory channel and not on the visual, which could mean that some ko-deictic uses accompanied by pointing gestures (e.g., kore ‘this’ and konna ‘like this’) were suppressed. Despite this possibility, it is still notable that ko-deictic uses were limited to this section, which implies that the occurrence of ko-deictics may be somehow related to the primary crux, where the moves are the hardest of the route and where physical and emotional involvement is accordingly maximized. This is still a hypothesis, or a speculation at best, because Climber C does not exhibit this pattern. The second pattern might be called an incremental type (Figure 3) because the frequencies of so-/ko-deictics generally increased toward the primary crux in Section 4. I



Deictic management of speech styles 

10 8 6

So Ko

4 2 0

1

2

0 0 0 0 1 (3) 2 (11) 3 (10) 4 (23)

Section

(a) Climber D 10 8 6 4 2 0

24 20 16 12 8 4 0

23

11

So Ko

11

6 1

5

0 0 1 (22) 2 (28) 3 (22) 4 (60)

Section

(b) Climber E

8 So Ko

4 1

1

0 0 0 1 (2) 2 (9) 3 (13) 4 (40)

Section

(c) Climber F

Figure 3.  Ko/so-frequencies in each section for Route I (incremental type) Note: The numbers in parenthesis indicate the number of clauses used in each section

call this pattern incremental because it differs from the threshold pattern, in which the highest prominence is at the initial (but “secondary”) crux (Section 2). Instead, the incremental elaboration concurs with the heavier use of the ko/so-deictics and the higher numbers of clauses used in Section 4 (except by Climber D, who used no ko-deictics). In a sense, it is not particularly surprising that a higher frequency of ko/so-deictics was observed for heavily elaborated sections. With this assumption in mind, I will show in Section 2.1.1 that the proportion of the ko/so-reference actually increased as the focus of description moved into Sections 2 and 4. In addition, the incremental pattern appears most conspicuous for Climber F, who used the ko-deictics most extensively as the preferred choice for the account of Section 4. This reversal was unexpected and contrary to expectation, given the current dominance of so-deictics in this monologic discourse. It would be revealing to compare this result with the assumptions by Horiguchi (1978) and Kinsui and Takubo (1992a), who claimed that in monologues the ko- and a-deictics are the norm and that the so-deictics are not favored – which is obviously not the case here. I offer a tentative explanation of these apparently confounding tendencies in later sections. What is still common between the threshold and incremental types is the ko-deictics’ higher sensitivity to the primary crux (Section 4). For Route I, most of the climbers explicitly mentioned in follow-up interviews that a few moves in Section 4 were definitely the hardest on the route. At the same time, they elaborated the peak experience by switching from so- to ko-deictics. In such a case, ko-deictics were observed to

 Kuniyoshi Kataoka

mainly offer their “evaluative commentary” on the crucial moves or holds, as typically seen in example (3), from Section 4 of Climber E’s narrative: (3) 1 Kakari ga asakute. ‘(It) is shallow,’ 2 tsukaminikui’n desu kedo mo, ‘and hard to grab.’ 3 boku ni totte, koko ga sootoo kuroo shita, ‘Here (it) is the part (I) had the most trouble,’ 4 kono kabe no ichiban no kakushinbu desu ne. ‘and the main crux of this wall, you know.’ 5 Sore o shikkari totte kara, ‘After (you) grab it firmly,’ 6 e=, tsugi, hidari ashi o sono piramiddo no bubun ni aru innaa ni ire masu. ‘a=hn, next, (you) put (your) left foot in the inner hold on the pyramid.’ 7 Koko de kono hidariashi ni shikkari tachikomu koto ga ‘This move, standing on the left foot (hold) here,’ 8 ichiban kakushin da to ‘was the crux move,’ 9 omottan desu kedo mo. ‘(I) thought.’ Similar explicit references to the crux with ko-deictics can be easily found in other narratives (ex. [4] and [5]). In these cases, however, their function seems to be more “focalization” than “evaluative commentary.” In narratives, the narrator may become sympathized and synchronized with the agent in the reminisced world, and the mode of reference could be equated with spatial deictic referencing, or genba-shiji. Thus, the sodeictics here may establish the spatial arrangement that evokes medial distance projected from the agent’s origo in the narrated scene. The so-referred objects then get focalized with ko-deictics at the peak performance, which are equivalent to what is termed “recognitional” use of deixis (Diesel 1999). (4) 1 E= sono ichiban ue, sono hida–, mannaka no ichiban ue ni aru innaa, ‘We=ll, that topmost, that lef–, the uppermost inner hold in the middle is,’ 2 kanari ii–, hikkakareba kikiyasui innaa ga ari masu kara, ‘quite good, a nice one, if (you) can get (it),’ 3 sore o torini iki masu. ‘so (you) go get it.’ 4 Koko ga ichiban no kakushin desu ne. ‘Here (this) is the crux, you know.’



Deictic management of speech styles 

(5) 1 Sore o torini itte, ‘(You) grab it,’ 2 de mata, a=, sayuu no ashi o irekaete, ‘and again, switch (your) right and left feet,’ 3 hidarite de e= kondo mannaka no rain no inna= o torini iki masu.’ ‘and next, (you) get the inner hold in the center with the left hand.’ 4 Kore ga e=tto,… jyuu, jyuusandanme ni narun desu ne ‘This is, well, the thirteenth panel ((= the final crux section)).’ The accumulative frequencies of the ko/so-deictics are summarized in Figure 4 (a) and (b). Even if we accept the complexity of individual differences and different depictive styles, it seems fairly correct to say that higher frequencies of ko/so-deictics are consistently observed in Sections 2 and/or 4 (Figure 4 [a]). However, the obvious difference is the function of ko, the frequencies of which evidently stand out in Section 4 (Figure 4 [b]). Nevertheless, we need to be cautious about making a blanket statement at the moment, considering the fact that the ko-deictics’ outstanding frequency in Section 4 was mainly caused by Climbers E and F’s usages. 60

Climber F Climber E Climber D Climber C Climber B Climber A

50 40 30 20 10 0

Section 1 Section 2 Section 3 Section 4

(a) Accumulative frequencies of so -deicitcs 16

Climber F Climber E Climber D Climber C Climber B Climber A

12 8 4 0 Section 1 Section 2 Section 3 Section 4

(b) Accumulative frequencies of ko-deictics Figure 4.  Accumulative frequencies of so- and ko-deictics

 Kuniyoshi Kataoka

On the other hand, in less challenging sections, where climbers can easily carry out moves, the need for such sequential imagery seems to be less activated. In fact, some climbers could barely remember the features of Section 1 (the non-crux section: example [6]), or often skipped references to unimportant holds (example [7]). Despite the sparse references to these objects, the climber in example (6) was actually voluble and highly elaborate in describing the crux section later in example (8), in which the kodeictics focalize the crucial “facing” posture toward the wall. These examples suggest that ko-deictics are often preserved for upcoming, experientially salient, and situationally important referents (moves and holds), with so-deictics laying the referential groundwork before the narrator switches to ko-deictics. (6) 1 E= daiichi hangu koeta tokoro ni ookina kinokojoo no hoorudo ga aru. ‘Well, there’s a big hold like a mushroom. After the first overhang.’ 2 Daiichi hangu made–, koeru made ni tsuite wa, ‘Up until (you) get over the first hang,’ 3 amari ishiki sezuni nobotte imasu kara, ‘(I) don’t care much, and continue climbing,’ 4 e=, maikai sono karada no ugokashikata ni tsuite wa, ‘and, about the moves of the body,’ 5 e= chigatteiru to omoi masu. ‘(I) climb differently every time, (I) think.’ 6 Mata, ruuto mo sootoo chigatte irunde nai ka to jibun de wa omotte, omotte ori masu. ‘Also, the courses (I) take may be quite different, too, (I) think.’ (7) 1 E=tto,…okutte itte, ‘Well, …keep moving your limbs,’ 2 ichi, nii, san, yon, go-ko, rok-ko me ni marui totte no aru mono o tsukande, ‘and after one, two, three, four, five, and six panels, (you) grab the one (= hold) like a round jug there,’ 3 sukoshi migi no hoo no tookuni aru mono o tsukamu. ‘and grab the one not far away and a little bit on (your) right.’ (8) 1 Sono tsugi wa hidarite ga torereba, ‘After that, if (you) can take the left-hand (hold),’ 2 tashika kore seetai de sono ue no hoorudo o totte, ‘(I) believe, for this move, (you) grab the hold above that with a “facing” posture,’ 3 sono mama seetai de, anoo, saishuu no baa o tsukamu to yuu katachi ni ‘and maintaining that facing posture, ahh, (you) grab the final bar,’



Deictic management of speech styles 

4 naru to omoi masu. ‘which (I) believe is the correct move.’ In the following, I pick up where I left off about the cohesive ties of referents. As mentioned in the previous section, the frequency factor is only part of the story. Although the frequency of ko/so-usage noticeably increased in Sections 2 and 4, this may be simply showing that some sections happened to be more extensively elaborated than others – if one has more to say about some sections, the number of clauses used there would become larger, and accordingly, more ko/so-deictics would result as a tracking device of referents (because one of their main functions is endophoric usage). In other words, the heavier use of ko/so-deictics may serve as a discourse strategy for achieving tighter referential cohesion (Halliday and Hasan 1976), made available as a device for higher involvement. The following represents the tabulated occurrences of the ko/so-deictics and the number of clauses for the six climbers (Table 1). Based on Table 1, we get Table 2, which shows that the ratios of referential cohesion for the secondary and primary cruxes are evidently higher than for the non-crux sections. This observation confirms that the crux is more densely elaborated by spatio-temporal and textual deixis and that not only the frequency but also the proportion of these deictics could firmly ground the “here-and-now” properties at the experience time, simultaneously corroborating the cohesive ties at the narration time through ko/so-deictics. Table 1.  Overall frequencies of ko/so across sections and climbers * cl. = # of clauses Section 1 (non-crux)

Section 2 (secondary crux)

Section 3 (non-crux)

Section 4 (primary crux)

Total

ko

so

cl.*

ko

so

ko

so

ko

so

ko

so

cl.

Climber A

0

5

25

0

15

35

0

8

20

1

11

33

1

39

113

Climber B

0

1

10

0

5

13

0

2

10

1

3

12

1

11

45

Climber C

0

6

13

0

10

24

0

4

12

0

7

15

0

27

64

Climber D

0

0

3

0

0

11

0

1

10

0

2

23

0

3

47

Climber E

0

6

22

1

11

28

0

11

22

5

23

60

6

51

132

Climber F

0

0

2

0

1

9

1

0

13

8

4

40

9

5

64

Subtotal

0

18

–

1

42

–

1

26

–

15

50

–

17 136

–

18

75

27

87

65 183

153

465

Total

cl.

43 120

cl.

cl.

Table 2.  Ratios of ko/so-deictics for crux vs. non-crux sections (Route I)

Ratio

Non-crux (Section 1+3):

Secondary crux (Section 2):

Primary crux (Section 4):

(18+27)/(75+87) = 27.8%

43/120 = 35.8%

65/183 = 35.5%

 Kuniyoshi Kataoka

Cohesive ties should be differentiated from “topic continuity.” As many have claimed (e.g., Givón 1983, Hinds 1983), topic continuity in Japanese is more likely to be achieved by such means as zero anaphora or ellipsis. While the topic continuity is calculated by the range of activation and the subsequent robustness of a single referent over clauses, the main “topics” here are bodily moves and climbing holds, which are referentially rapidly changing. That is, there is little empirical base for one topic to be continually activated for a long stretch of discourse. Instead, we find a rapid shift between chunks of cohesive ties established by ko/so-deictics, which could translate into some involvement strategies (patterns based on repetition of words/phrases and the “crowded stage” assumed by Longacre [1996]). If we want to confirm that ko/so-deictics were actually an involvement strategy and were preferred for achieving tighter cohesion in the crux (Sections 2 and 4), we need to consider, as shown above, the ratio of ko/so-deictics as compared with the total number of clauses in each section. What, then, would motivate the inclusion of various referents instead of the maintenance of topic continuity? I believe it is the type of activity under discussion. All of the climbers would agree on the importance of learning the sequence of moves in the crux section because there is often only one way to go through it, especially when they are performing at their physical limits. (However, climbers will be able to climb more freely and in various manners as they become more proficient.) This fact necessitates that they correctly differentiate usable holds and remember the continuous flow of movements. Consequently, the ko/so-deictics will be employed in more technical or strenuous parts of the route – i.e., in Sections 2 and 4. Put another way, narrators were required to invoke procedural memory with sequential details for tighter cohesion, while also differentiating the recalled experiences with differently evaluated ko/sodeictics according to the degree of involvement with each event. These deictics thus serve as an effective means of adequately managing the representation of involvement through “imagery and detail” (Tannen 1989) and “heightened vividness” of the scene (Longacre 1996).

2.2

Route II

An aspect not fully addressed yet is whether ko-deictics should be contextually sensitive to the primary crux or to something else. Here, I present other accounts of another route (Route II) and show that ko-deictics are in effect sensitive to such entities that are experientially salient as well as to the difficulty factor. I recruited two more climbers to have them narrate their climbs under the same conditions as the first group (Figure 5). In this case, however, the route grade was much higher (5.12- at the hardest point), with the primary crux in Section 4 and the secondary cruxes in Sections 2 and 3 (see also Figure 1). Accordingly, the proficiency levels of the climbers were also higher.



Deictic management of speech styles 

10 8 6 4 2 0

9 7 4

So Ko

5 3 1

1

0 1 (14) 2 (28) 3 (27) 4 (43)

Section

(a) Climber G

10 8 6 4 2 0

6 3 2 2 1 0 0 0 1 (13) 2 (16) 3 (26) 4 (38)

So Ko

Section

(b) Climber H

Figure 5.  Ko/so frequencies in each section for Route II Note: The number in parentheses indicates the number of clauses used in each section

After their monologues, I interviewed each of the climbers about their impressions of the route. The hardest section was unanimously (including my own impression) agreed to be Section 4, followed by Sections 2 and 3. For sections 2 and 3, I received comments that they were equally hard but that the required moves were qualitatively different. As is evident in Figure 5, both participants exhibited roughly the same incremental pattern, which meant that the overall frequencies of ko/so-deictics gradually increased from Section 1 through Sections 2 and 3 and eventually reached the highest levels in Section 4. Another interesting fact is that, as was observed for Climber F (Figure 3), Climber H relied upon the ko-series as the preferred choice for describing the overall performance, while the so-tokens were only used in the primary crux section. This pattern is contradictory to a previous observation of ko-deictics as the marker of peak involvement. (I will also look into this question in Section 2.5 in terms of the preferred conceptual styles in ko/so-alternations.) In the previous analysis, I regarded the switch into the ko-deictics as an indexical device for contextually marking off the peak involvement. In fact, despite the fact that the primary crux occupies only one section out of four on each route, more than twothirds of total ko-deictics occurred there (67.4%: 31 out of 46 tokens of ko on Routes I and II).14 Under closer examination, however, we see that the ko-series deictics are not exclusively attached to crux moves or holds – they were also used to depict a bucket (meaning “big and easy-to-grab”) hold (examples [9] and [10]), a rest point (example [11]), or any entities that were situationally important or highly usable in that particular context. That is, these referents emerge into awareness as experientially salient entities only against the backdrop of continuous difficulty and would not have deserved to be mentioned with ko-deictics in other “easy” sections. The occurrence of such noncrux-referring ko-deictics accounts for 32.6% of the total ko occurrences (15/46). (9) 1 Karada ga antee shita tokoro de, ‘When (your) body gets stable,’ 14. This figure includes the tokens for Climbers I and J, to be introduced in Section 2.3.

 Kuniyoshi Kataoka

2 detto-ppoku hidari no poketto o tori ni iku. ‘go get the left pocket with a dead-point move.’ 3 E=, koko wa gaba nande, ‘A=nd, here (this) is a bucket hold,’ 4 e=, migite o juubun ni yasumete, ‘so=, (you) rest (your) right arm well,’ 5 tsugi wa kyon-ppoku hidari, ‘and the next is a drop-knee-like move to the left,’ 6 e=, migite de poketto o toru. ‘a=nd, grab a pocket with (your) right hand.’ 7 Koko mo gaba nande, ‘It is a bucket hold here, so’ 8 e=, tsugi wa hidarite o resuto. ‘next, (you) rest (your) left arm.’ (10) 1 De, hangu no dedashi ni aru innaa hoorudo?, ‘And, the inner hold at the beginning of the overhang?’ 2 Kore wa gaba desu kedo, ‘This is a bucket hold, so’ 3 kore o migite de toru. ‘grab this with (your) right hand.’ (11) 1 Koko de yasumanaito ‘If (you) don’t take a rest here,’ 2 ue no hoo de panpan ni hacchau kara ‘(you) will get pumped up around there,’ 3 omoikkiri yasunde kara, ‘so after (you) take a very good rest,’ 4 u=tto,…hidari, hidarite chigau, hidarite o kondo mannaka no kabe no hoo ni utsushite itte, ‘we=ll, left-, left hand, no, move your left hand toward the center of the wall,’ 5 u=to, kokode ashi o chotto daijyoobu na tokoro ni oitara, ‘a=nd, if (you) put (your) legs on a secure hold here,’ 6 koko de mo chotto yasunde kara,… ‘(you) also take a short rest here, …’ Given the relatively high ratio of such ko-deictics scattered throughout the description, the previous characterization of ko-deictics is still meager. It would make more sense, instead, to reconsider the observed functions of ko-deictics in a broader concept such



Deictic management of speech styles 

as a gauge for “situational weight of salience” by temporarily crediting ko-attached referents with the “figure” status against the conceptual “ground” in situ. In specific terms, ko-deictics are psychologically/emotionally evaluated and perceptually/conceptually focalized, as seen earlier, and situationally emergent, as seen here – as well as contributing to overall involvement.

2.3

Route III

Given the above observations, however, we still cannot rule out the possibility that the higher frequencies of ko/so-deictics could have been induced by a particular physical feature – the overhangs (in Sections 2 and 4) – which typically impel climbers to perceive more gravity than on the vertical wall and require them to cope with greater physical stress. The following data indicate that the higher frequency of ko/so-occurrences in such sections is not straightforwardly associated with overhangs but necessarily with the “difficulty” scale, according to which they were required to make variable commitments to crucial holds and moves. For this analysis, I asked Climbers I and J to describe Route III (Figure 6), which has the crux in Section 3, not in Section 2 or 4. Because artificial holds are removable and can be attached anywhere as long as there are holes for the bolts, climbers sometimes get a crux on a vertical section instead of an overhang. In this case, Section 3 required of the climbers delicate balancing and technical sophistication rather than the muscular strength expected on overhangs. The overhangs in Sections 2 and 4 were judged to be roughly the same difficulty as, or slightly easier than, the challenges in Section 3. Again, we find the same pattern as before – the most difficult section (Section 3 here) led to the inclusion of the highest frequency of ko/so-deictics, indicating that the difficulty scale more precisely reflects the occurrences of those deictics. One may wonder whether the climber’s proficiency level may have affected the linguistic performance in such a way that the more expert one becomes, the more 14 12 10 8 6 4 2 0

4 2

6

5

1 0 0 0 1(17) 2(27) 3(28) 4(18)

(a) Climber I

So Ko

Section

14 12 10 8 6 4 2 0

13 11 So Ko

7 5 2 2 1 0 1(11) 2(28) 3(45) 4(36)

Section

(b) Climber J

Figure 6.  Ko/so frequencies in each section for Route III Note: The number in the parentheses indicates the number of clauses used in each section

 Kuniyoshi Kataoka

relaxed and objective an attitude one can assume in route description, which might accordingly lead to less involvement and lower frequencies of ko/so-deictics. As probable as it sounds, that possibility may be ruled out for the current population for two reasons. First, as the basic requirement, the climbers were asked to describe the hardest route they were able to climb at the time of narration. Thus, they were talking about their peak performance, where maximum physical and psychological commitment was expected. Second, this particular pattern, in which the frequencies of ko/so-deictics typically increase at the cruxes, was also observed for both beginners (Climbers A to F on Route I, and Climbers I and J on Route III) and skilled climbers (Climbers G and H on Route II). Given these outcomes, proficiency levels would seem to be a negligible factor.

2.4

Overall tendency of ko/so-occurrences

We have already seen for Route I that higher frequencies of ko/so-deictics consequently lead to a higher degree of referential cohesion (Table 2 above). Now, let us take a closer look at the summary ratios of ko/so-deictics for all of the routes. Table 3 shows again that the primary crux section generally displays the highest ratio of ko/so-occurrences per clause, indicating that not only the elaboration but also the cohesion of discourse were most intensively achieved there. When compared with Route I (reproduced here) and Route III, Route II exhibits a noticeable difference in that the noncrux section attains roughly the same (actually a slightly lower) ratio of ko/so-deictics as the secondary crux. A possible cause for the high convergence would be that, for these skilled climbers (G and H), neither the non-crux nor the secondary cruxes were hard enough to require of them involvement as committed as was required in the primary crux. Whether or not this conjecture is correct, a common feature was that the density of ko/so-deictics for the primary crux section was generally the highest for all the routes described (except for the roughly equal ratio for Route I). In sum, the difference between involved and non-involved styles may be indexed by these covert switches in the relative frequency and the ratio of these deictics, as well as by the distributional patterns of deictics, to which we now turn. Table 3.  Ratios of ko/so-deictics for crux vs. non-crux sections (Route I to III) Route I Climbers A-F

Non-crux (1 Section +3): 45/162 = 27.8%

Secondary crux (Section 2): 43/120 = 35.8%

Route II Climber G, H

Non-crux (Section 1): 5/27 = 18.5%

Route III Climber I, J

Non-crux (Section 1): 3/28 = 10.7%

Secondary crux (Section 2+3): 17/97 = 17.5% Secondary crux (Section 2+4): 32/109 = 29.4%

Primary crux (Section 4): 65/183 = 35.5% Primary crux (Section 4): 22/81 = 27.2% Primary crux (Section 3): 24/73 = 32.9%



Deictic management of speech styles 

2.5

Ko-so conversion and the “figure-ground” relationship

In this section, I resume the topic left unattended in Section 2.2 (on Route II), and examine the inversion patterns of ko/so-deictics for Climbers F and H, who counterintuitively used ko-deictics as the referential matrix. Table 4 shows the distributions of ko/so-deictics by those climbers and evidently indicates that, if the climbers’ preferred deictic is the ko-series, it is conversely the so-series that come into play to mark off the peak involvement in the crux section. Climber F used five so-tokens in total during the account: one in Section 2 (the secondary crux), and the other four in Section 4 (the primary crux). Among the four tokens in Section 4, the first so-token was used as a “temporal,” sore de ‘then,’ but was functionally a conjunctive – this is qualitatively different from the other tokens, and may be negligible. All the rest of the so-deictics appeared after the narrator’s explicit reference to the crux move at (10). In contrast, Climber H used the so-series only twice, both at the final stage of Section 4. Here again, the crux-referring so-series all occurred in, and elaborated on, the move sequence in Section 4 after Climber H explicitly mentioned that he was now moving into the crux moves at (9). Table 4.  (Anomalous) ko-to-so switch at the crux Climber F (Route I)

Climber H (Route II)

Section 1 Section 2

N.A. (1) sono ugokashikata ‘that motion’

Section 3

(2)

‘this (section)’

Section 4 (crux section)

(3) (4) (5) (6) (7) (8) (9) (10)

kore ‘this (hold)’ sore de ‘then’(= temporal) kore ‘this (move)’ kore ‘this (hold)’ kore ‘this (hold)’ ‘this (hold)’ ‘here’ kore ‘this (move)’ <= reference to crux sono tsugi ‘(next move) after that’ kore ‘this (move)’ ue ‘(move) above that’ sono mama ‘that way’ <=end of crux

(1) (2) (3) (4) (5) (6) (7) (8) (9)

(11) (12) (13) (14)

(10) (11) (12) (13) (14)

koko ‘here’ koko ‘here/this move’ koko ‘here/this move’ kara ‘from here’ koko ‘here’ ‘here’ koko ‘here’ koko ‘here’ kok-kara ‘from here’ <= reference to crux koko ‘here’ koko ‘here’ sono ato ‘after that (move)’ sono ue ‘(move) above that’ koko ‘here’ <= end of crux

 Kuniyoshi Kataoka

Although the distribution of ko/so-deictics is reversed – like the famous figure-ground conversion of a “face-base” pattern – the resultant “effect” of the ko-to-so switch is essentially the same as the so-to-ko switch observed above – i.e., the marking of peak involvement. Notice that the centrifugal ko-to-so (proximal-to-medial) switch is actually de-focalization, and the previous focalizing function does not apply here. In addition, so-deictics were always realized in the two-part, “demonstrative + nominal” construction (e.g. sono ugokashikata ‘that motion’ and sono tsugi ‘(next move) after that’), in contrast to the mono-phrasal ko-deictics (e.g. koko ‘here’ and kore ‘this one’). This periphrastic elaboration for the ko-to-so switch may seem like a mode of evaluation typically observed at the peak of narrative in terms of “excessive details,” but it is actually not. As seen in Table 5 below, so-deictics consistently exhibit this construction even in the so-to-ko (medial-to-proximal) switch, indicating that this elaboration process could be based on a formulaic motivation, and is not the exclusive property of the ko-to-so switch. This observation suggests that the shifts between ko and so do not represent a predetermined, one-way mechanism but rather a highly flexible and adaptive one, managed in situ to make the most effective choice from the given stock of discursive resources. That is, the involvement effect of ko-deictics is not absolute but only relative to the surrounding environment, both physical and textual.

2.6

Ima ‘now’ as temporal ko-deictic

Further, temporal and spatial deictics largely share conceptually equivalent values. In some cases, the ko-deictic emerged as a temporal equivalent, ima ‘now,’ which is also a proximal deictic (like the ko-series) that indexes the temporal origo in the narrated scene. (Since it does not have an overt ko-stem, it was not included in the frequency count. However, Halliday and Hasan (1976: 57) classified ‘now’ as a demonstrative reference.) As briefly discussed below, Climber G was the one who extensively used ima ‘now’ during the narration. In effect, Climber G used ima extensively along with other ko-marked deictics such as koko ‘here’ and kore ‘this (one)’ (see Table 5). Although the ko-marked temporal expressions kono toki ‘at this time’ and kore kara ‘from this (= from now on)’ are plausible expressions, they are not usually used to indicate ‘now.’ At least, no participants used them as such in my data. When the temporal deictics are differentiated from other deictics and are arranged according to the order of occurrence, a systematic configuration of ima ‘now’ and the ko-marked deictics becomes evident (Table 5). Interestingly enough, occurrences of ima and ko-deictics respectively converge on Section 2 through Section 4 (the primary and secondary cruxes), but not on Section 1 (non-crux). Also, the use of ima only occurred immediately before and after Climber G grumbled that (9) kore ga warui ‘this (move) is tough/hard,’ which was echoed with more emphasis in (27) kore ga hijoo ni warui ‘this (move) is extremely tough/hard’. And these tokens of kore ‘this’ are the initial and final occurrences of the ko-deictics, which serve, together with ima, as a



Deictic management of speech styles 

Table 5.  Climber G’s use of a temporal deictic ima ‘now’ and others on Route Temporals Section 1 (4) sore kara ‘then/after that’ Section 2 (secondary crux)

Section 3 (secondary crux)

Section 4 (primary crux)

(6 ) ima ‘now’ (7) sore kara ‘then/after that’ (8) ima ‘now’

(11) (12) (13) (14)

sore kara ‘then/after that’ sore kara ‘then/after that’ sore kara ‘then/after that’ sore kara ‘then/after that’

(21) sore kara ‘then/after that’

(25) sono toki ‘at that time’ =>mention of primary crux

(28) sore kara ‘then/after that’ (31) ima ‘now’ (33) ima ‘now’ (34) Ima ‘now’

Others (mostly spatial) (1) sono ue ‘above that’ (2) sono ue ‘above that’ (3) sono ue ‘above that’ (5) sono ue ‘above that’

(9 ) kore ‘this one’ (kore ga warui ‘this is tough’) (10) sono ue ‘above that’

(15) (16) (17) (18) (19) (20)

kore sore soko sono kore kore

‘this one’ ‘that one’ ‘there’ hangu ‘that overhang’ ‘this one’ ‘this one’

(22) sono innaa ‘that inner hold’ (23) kore ‘this one’ (24) sore ‘that one’ (26) koko ‘here’ (27) kore ‘this one’ (kore ga hijoo ni warui ‘this is extremely tough.’) (29) sono ue ‘above that’ (30) sono kabe ‘that wall’ (32) sono yoko ‘beside’ that (35) sono kakusinnbu ‘that crux’

 Kuniyoshi Kataoka

“contextualization cue” into and out of the main stage of the cruxes. In other words, ima serves as a proximal temporal surrogate that fore-/after-shadows the “danger zone” on the route, whereas the ordinary ko-marked deictics are preserved as the default for spatial referencing of the crux. Further, in Sections 1, 2, and 3, temporal and other deictics tended to appear in chunks of threes and fours, while Section 4 exhibited frequent switches and a higher degree of nexus among the depicted objects and moves – or what Longacre (1996) calls the “crowded stage.” Accordingly, the stage of involvement is covertly indexed by these choices and distributions of ko/so-deictics, which may be customized and adapted as an overt (if not obvious) means of speech style shift. To borrow Ochs’ term (Ochs 1990), ko/so-deictics are both a direct index of speech styles (through the referential meaning of evaluation and focalization) and an indirect index (through non-referential patterns of involvement). These deictics constitute a systematic configuration in situ by marking off involved and less-involved performances in the crux and non-crux sections, while at the same time reflecting, through their use and distribution, the epistemological interpretation of experience. Here again, linguistic encoding of ko/sodeictics builds upon a “bricolage-like” management of the non-referential potential of deictics rather than upon the purely referential function. A similar phenomenon has been examined in studies of tense switching for the conversational historical present (CHP) (cf. Johnstone 1987, Sakita 2002, Schiffrin 1981, Wolfson 1978). As these studies showed, one of the most important functions may be the switching itself (such as that between simple past and historical present), which serves to highlight the text-structural and intersubjective understanding of the ongoing discourse. Because deictic management could surface in multiple guises, the ko/so-switching could be another case of such linguistic and non-linguistic, referential and non-referential, organizations of discourse.

3. Discussion and conclusion In the current climbing monologues, a higher frequency of ko-deictics and proportion of ko/so-deictics were mainly observed in “crux” sections. There, the dominant and unmarked choice of deictics to refer to holds or moves was the so-deictics, whereas a higher frequency of ko-deicitcs was found to be associated with entities that were salient in some ways, rather than simply associated with physical features such as strenuous moves or overhangs. So-deictics were dispersed over the narratives and served as the conceptual “ground” with respect to ko-deictics. They were also found to be a fundamental contributor to the tighter referential cohesion, serving as a major facilitator of narrative involvement in terms of “imagery and detail” and “longer sequence of discourse” (Tannen 1989), and “rhetorical underlining” (Longacre 1996). On the other hand, ko-deictics offered a resource for the narrator’s (and the imaginarily identified agent’s) valorization of saliency, which is associated with evaluation,



Deictic management of speech styles  (a) Genba-shiji (Space deixis) (1) Unmarked construal for involvement (Centripetal, medial-to-proximal: so-to-ko switch)

so ko A

N

= figure = ground A = Agent N = Narrator

projection

Narrative world (2) Marked construal for involvement (Centrifugal, proximal-to-medial: ko-to-so switch)

so ko A

N

(b) Bunmyaku-shiji (Discourse deixis) Utterance

Utterance

antecedent

so

~ ko

anaphoric

N

ko cataphoric

Figure 7.  Style shifting of different construals for genba-shiji and bunmyaku-shiji

focalization, urgency, and importance conceived at the indexical center (the ko-domain). They typically index a conceptual “figure” (Figure 7), appearing sparingly to underscore transient and immediate involvement, whether emotional, perceptual, or physical. The association of these marked features with ko-deictics (and unmarked features with so-deictics) are never prescriptive and stationary but are relative to the ongoing context and may also be inverted. It is true that this focusing construal by the kodeictics is unmarked and more plausible in other cases, because the conceptual figure tends to be smaller and more moveable, like the agent’s ko-domain recalled via genbashiji (Figure 7 [a-1]). I believe that it was not accidental that this type of construal – so as the ground and ko as the figure (or the so-to-ko switch) – was observed more frequently in 8 out of 10 cases. However, our data (by Climbers F and H) also showed that the reversed construal – ko as the ground and so as the figure (or the ko-to-so switch)

 Kuniyoshi Kataoka

– was also possible, though rare, at the ratio of 2 out of 10 cases, probably due to the marked construal where the more expansive, medial so-domain needed to be focalized after the introduction of the proximal ko-domain (Figure 7 [a-2]). Also, these contradictory orientations may have been derived from different “starting points” (McWhinney 1977), which respectively motivated these different preferences and construals. The same process would equally apply to bunmyaku-shiji (Figure 7 [b]), where perceived distance (or degree of salience) from the textual antecedent would be differentially indexed by the choice of ko/so-deictics. It was possibly due to the status of monologic narration that the cataphoric ko-usage was not observed in our data. Because it is often an audience-directed strategy and is unavailable in this case, it is represented here by broken lines and a broken square. Also, because the so-deictic has no cataphoric function, the management of indexical values through the ko/so-switching, which was exploited for anaphoric reference, could not be activated for cataphoric usage. In previous studies, Horiguchi (1978) and Kinsui and Takubo (1992a) maintained that a speech environment lacking a hearer (as in monologue) favors the use of ko- and a-deictics. For example, Kinsui and Takubo (1992a) claimed that there is a trigger hierarchy for the use of deictics in terms of “genba ‘actual space’ (ko-reference) > keiken ‘experience’ (a-reference) >> sonota ‘others’ (so-reference).” Because the objects in genba ‘actual space’ are registered in direct experience, they are preferably referred to in recall by ko-, next most frequently by a-, and least often by so-deictics. This formalization is intuitively correct but is also a purely theoretical construct since the “inner” monologue only survives in the speaker’s mind. Our results serve as a clear counter-example to this generalization. Although some tokens show a vague status between genba-shiji and bunmyaku-shiji, so-deictics were the primary choice for referring to physical objects and moves envisioned through memory. And for most of my participants, the ko-deictic is the secondary choice that serves special discourse functions of evaluation, focalization, or involvement.15 Given all of this, we could at least confirm that a universal, one-way treatment of a “monologue” is ill-defined. It is obviously necessary to subcategorize monologue into at least two modes of narration (Table 6): the “inner” (i.e., silent or non-verbalized) and the “outer” (i.e., verbalized) monologues, with the former intrinsically favoring a silent reminiscer’s perspective – which will consequently lead to the dominant use of ko/adeictics, as was previously assumed by some Japanese linguists – and with the latter, by the very action of uttering words, inevitably evoking a virtual space with imaginary “participants,” with whom/which the narrator identifies objects and moves through spatial deixis (genba-shiji). In the latter case, the existence of vocalization and the absence of an interlocutor seems to be the key to the dearth of a-deictics. Likewise, the physical reality of utterance (“word”) could facilitate the textual reality of deixis (bunmyaku-shi 15. Also, in bunmyaku-shiji (or textual deixis), the so-series expressions are the preferred choice (see Kuno 1973, Yoshimoto 1992) because endophoras essentially serve a textual function and are normally irrelevant to the speaker’s experiential basis that ko-/a-deictics imply.



Deictic management of speech styles 

Table 6.  Preferred use of deictics vis-à-vis different modes of verbalization Multi-party ~ conversation

Ko-deicitcs So- deicitcs A – deicitcs Studies:

Dialogue

Monologue Outer

Inner

(+vocalization, +interlocutor)

(+vocalization, -interlocutor)

(-vocalization, -interlocutor)

√ √ √ Mayes and Ono 1991, Hayashi 2004, Naruoka 2006, etc.

√ √ (√?) This study

√ (√?) √ Horiguchi 1978, Kinsui and Takubo 1992, etc.

ji), for which so-deictics may be employed by default as anaphors. These characterizations suggest that the “inner” and “outer” monologues may represent distinct levels of verbalization, for which the ko/a- and ko/so-pairs are respectively preferred. These different modes of verbalization may have been an empirical motivation for what is called a “double binary” (ko-a vs. ko-so), rather than a tripartite (ko-so-a), formation of Japanese deixis, typically seen for deictic collocations such as are-kore (kangaeru) ‘(think) this and that’ and soo-koo (suru) ‘(do) this way and that’ (Mikami 1970). More complication will result if we include a situation with a-deictics more easily evoked in interaction. We have yet to know how those three types of deictics will fare in holistic accounts of the deictic usage. In addition, there still remains a question as to the levels of integration. We simply posited that spoken dialogue would be like multi-party conversation with regard to the use of deictics, but we may ultimately find different deictic behaviors for these modes of narration. In conclusion, studies of Japanese deixis have largely neglected the actual behaviors of deictics in discourse (especially in monologues). This is in part inevitable, given the extremely complex ways in which deictics are used and interpreted. However, it is undoubtedly necessary to start incorporating naturalistic data in order to construct more realistic accounts of deictic usage because monologue and dialogue (and also conversation) seem to employ different modes of (inter)subjective construals. In this very sense, I hope to have shown that deictic management in monologue is also closely related to such issues as speech style shift and involvement in general.

References Anderson, S. and Keenan, E. 1985. “Deixis.” In Language Typology and Syntactic Description, Vol. III, T. Shopen (ed.), 259–308. Cambridge: Cambridge University Press. Ariel, M. 1988. “Referring and accessibility.” Journal of Linguistics 24: 65–87.

 Kuniyoshi Kataoka Besnier, N. 1994. “Involvement in linguistic practice: An ethnographic appraisal.” Journal of Pragmatics 22: 279–299. Brown, R. and Ford, M. 1964. “Address in American English.” In Language in Culture and Society, D. Hymes (ed.), 233–244. New York: Harper and Row. Brown, R. and Gilman, A. 1960. “The pronouns of power and solidarity.” In Language and Social Context, P. Giglioli (ed.), 252–282. London: Penguin Books. Chafe, W. L. 1985. “Linguistic differences produced by differences between speaking and writing.” In Literacy, Language, and Learning: The Nature and Consequences of Reading and Writing, D.R. Olson, N. Torrance, and A. Hildyard (eds.), 105–123. Cambridge: Cambridge University Press. Cook, H. M. 1996. “The use of addressee honorifics in the Japanese elementary school classroom.” In Japanese/Korean Linguistics, N. Akatsuka, S. Iwasaki, and S. Strauss (eds.), 5: 67–82. Cook, H. M. 1998. “Situational meanings of Japanese social deixis: The mixed use of the masu and plain forms.” Journal of Linguistic Anthropology 8: 87–110. Diessel, H. 1999. Demonstratives. Amsterdam: Benjamins. Duranti, A. 1984. “The social meaning of subject pronouns in Italian conversation.” Text 4(4): 277–311. Givón, T. (ed.). 1982. Topic Continuity in Discourse: A Quantitative Cross-Language Study. Amsterdam: John Benjamins. Glover, K. D. 2000. “Proximal and distal deixis in negotiation talk.” Journal of Pragmatics 32: 915–926. Gumperz, J. J. 1982. Discourse Strategies. Cambridge: Cambridge University Press. Gundel, J. K., Hedberg, N., and Zacharski, R. 1993. “Cognitive status and the form of referring expressions in discourse.” Language 69: 274–307. Halliday, M. A. K., and Hasan, R. 1976. Cohesion in English. London: Longman. Hanks, W. F. 1990. Referential Practice: Language and Lived Space Among the Maya. Chicago: University of Chicago Press. Hanks, W. F. 1992. “The indexical ground of deictic reference.” In Rethinking Context, A. Duranti and C. Goodwin (eds.), 43–76. Cambridge: Cambridge University Press. Hayashi, M. 2004. “Projection and grammar: Notes on the ‘action-projecting’ use of the distal demonstrative are in Japanese.” Journal of Pragmatics 36 (8): 1337–1374. Hinds, J. 1983. Topic continuity in Japanese [1]. In Topic Continuity in Discourse: A Quantitative Cross-Language Study, T. Givón (ed.), 43–93. Amsterdam: John Benjamins. Horiguchi, K. 1978. “Shiji-go no hyoogensei.” In Shiji-Shi, S. Kinsui and Y. Takubo (eds.), 74–90. Tokyo: Hituji Syobo. Ikuta, S. 1983. “Speech level shift and conversational strategy in Japanese discourse.” Language Sciences 5: 37–53. Iwasaki, S. 2002. Japanese. Amsterdam: John Benjamins Johnson, D. M., Roepcke, Y., and Kataoka, K. 1997. “Constructing social groups in discourse.” Pragmatics and Language Learning 8: 37–53. Johnstone, B. 1987. “’He says … so I said’: Verb tense alternation and narrative depictions of authority in American English.” Linguistics 25: 33–52. Jones, K. “Nihongo no danwa ni okeru iken no shoototsu to ninshoo shiji-shi no yoohoo. (Conflict and use of person reference in Japanese conversation.” Unpublished manuscript.



Deictic management of speech styles  Kamio, A. 1990. Joohoo no Nawabari Riron: Gengo no Kinooteki Bunseki (A Theory of Territory of Information: Functional Analysis of Language.) Tokyo: Taishukan. Kataoka, K. 2002. “Referential vs. non-referential meaning and cultural practice: ‘Indexicality’ in language use.” The Japanese Journal of Language in Society 4(2): 21–41. Kataoka, K. 2004. “Co-construction of a mental map in spatial discourse: A case study of Japanese rock climbers’ use of deictic verbs of motion.” Pragmatics 14(4): 409–438. Kinsui, S. and Takubo, Y. 1992a. “Daiwa kanri riron kara mita Nihongo no shiji-shi.” In Shiji-Shi, S. Kinsui and Y. Takubo (eds.), 123–149. Tokyo: Hituji Syobo. Kinsui, S. and Takubo, Y. 1992b. “Nihongo shiji-shi kenkyuu-shi kara/e.” In Shiji-Shi, S. Kinsui and Y. Takubo (eds.), 151–192. Tokyo: Hituji Syobo. Kuno, S. 1973. Structure of the Japanese Language. Cambridge, MA: MIT Press. Lakoff, R. 1974. “Remarks on this and that.” In Papers from the Tenth Regional Meeting: Chicago Linguistic Society, M. La Galy, R. Fox, and A. Bruck (eds.), 345–56. Chicago: Chicago Linguistic Society. Levinson, S. C. 1983. Pragmatics. Cambridge: Cambridge University Press. Linde, C. 1979. “Focus of attention and the choice of pronouns in discourse.” In Syntax and Semantics 12, T. Givón (ed.), 337–354. New York: Academic Press. Longacre, R. E. 1996. The Grammar of Discourse (2nd ed.). New York: Plenum Press. MacWhinney, B. 1977. “Starting points.” Language, 53(1): 152–168. Mayes, P. and Ono, T. 1991. “Social factors influencing reference in Japanese: With a special emphasis on ano hito.” Santa Barbara Working Papers in Linguistics, 84–93. Santa Barbara, CA: University of California at Santa Barbara. Maynard, S. 1991. “Pragmatics of discourse modality: A case of da and desu/masu forms in Japanese.” Journal of Pragmatics 15: 551–582. Maynard, S. 1993. Discourse Modality: Subjectivity, Emotion and Voice in the Japanese Language. Amsterdam: John Benjamins. Mikami, A. 1970. Bunpoo Shoo-Ronshu. Tokyo: Kurosio. Morita, Y. 2002. Nihongo Bunpoo no Hassoo. Tokyo: Hituji Syobo. Mühlhäusler, P., and Harré, R. 1990. Pronouns and People: The Linguistic Construction of Social and Personal Identity. London: Basil Blackwell. Naruoka, K. 2006. “The interactional function of the Japanese demonstratives in conversation.” Pragmatics 16(4): 475–512. Nishide, K. 1993. “Kuukan no gengogaku.” Gengo 22: 36–43. Ochs, E. 1990. “Indexicality and socialization.” In Cultural Psychology, J. Stigler, R. Shweder, and G. Herdt (eds.), 289–308. Cambridge: Cambridge University Press. Okamoto, N. 1997. “Kyoositsu danwa ni okeru buntai shifuto no shihyooteki kinoo.” Nihongogaku 16(3): 39–51. Okamoto, S. and Sato, S. 1992. “Less feminine speech among young Japanese females.” In Locating Power: Proceedings of the 2nd Berkeley Women and Language Conference, K. Hall, M. Bucholtz, and B. Moonwomon (eds.), 478–488. Rubba, J. 1996. “Alternate grounds in the interpretation of deictic expressions.” In Spaces, Worlds, and Grammar, G. Fauconnier and E. Sweetser (eds.), 227–261. Chicago, IL: The University of Chicago Press. Sakita, T. I. 2002. Reporting Discourse, Tense, and Cognition. Amsterdam: Elsevier. Sakuma, K. 1951. Gendai Nihongo no Hyoogen to Gohoo (2nd ed.). Tokyo: Kooseekaku. Sakuma, M. 2002. “Setsuzoku-shi, shiji-shi to bun rensa.” In Fukubun to Danwa, H. Noda, T. Masuoka, M. Sakuma, and Y. Takubo, 117–189. Tokyo: Iwanami.

 Kuniyoshi Kataoka Schiffrin, D. 1981. “Tense variation in narrative.” Language 57(1): 45–62. Shooho, I. 1980. “Ko, so, a no taikei.” (The system of ko, so, a). Nihongo no Shijishi, 51–122. Tokyo: Kokuritsu Kokugo Kenkyuujo. Silverstein, M. 1976. “Shifters, linguistic categories, and cultural description.” In Meaning in Anthropology, K. Basso and H. Selby (eds.), 11–55. Albuquerque, NM: University of New Mexico Press. Silverstein, M. 1995. “Indexical order and the dialectics of sociolinguistic life.” SALSA III: 266–295. Strauss, S. 1993. “Why ‘this’ and ‘that’ are not complete without ‘it’.” Chicago Linguistics Society 29: 403–417. Tannen, D. 1989. Talking Voices. Cambridge: Cambridge University Press. Tokuchi, S. 2001. “Shakai to gengo o tsunagu imi ni tsuite.” The Japanese Journal of Language in Society 4(1): 68–80. Wilson, J. 1990. Politically Speaking. London: Basil Blackwell. Wolfson, N. 1978. “A feature of performed narrative: The conversational historical present.” Language in Society 7: 215–237. Yoshimoto, K. 1992. “Nihongo no shiji-shi ko/so/a no taikei.” In Shiji-Shi, S. Kinsui and Y. Takubo (eds.), 105–122. Tokyo: Hituji Syobo.

Variation in prosodic focus of the Japanese negative nai Issues of language specificity, interactive style, and social situations* Shoji Takano

Hokusei Gakuen University

1. Introduction While social situations and related norms for linguistic performance influence which style or register should be chosen (i.e., the unmarked choice), it is a well-known linguistic fact that speakers’ stylistic (or register) choices vary even within identical communicative situations. Beginning with such seminal work as Ferguson (1964) and Blom and Gumperz (1972), linguists have become increasingly interested in investigating this dynamic aspect of code or style alternations in actual linguistic practice – whatever may be the implicit or “metaphorical” social meanings speakers try to convey beyond what is said through their situationally marked choices of the code or style in the immediate context of use (Gumperz 1982). Prosody is one of the most efficient means to deliver these “paralinguistic” messages, which typically consist of such interactive dimensions as solidarity and social distance, power distribution, deference and confrontation, and the speaker’s current emotional state (Ladd 1996: 33). So far, the great majority of studies of stylistic variation in Japanese have been restricted to such overt morpho-syntactic characteristics as plain vs. polite (-desu/masu) alternations of the predicate structure and their correlation with the situational context of use. The present study is an attempt to explore the largely neglected area of prosody by shedding light on situationally marked but * This paper is heavily indebted to inspiration from Malcah Yaeger-Dror. Her insightful comments on an early version of the paper and her encouragement are particularly appreciated. I would like to express my gratitude to the two editors, Kimberly Jones and Tsuyoshi Ono, for this opportunity to contribute to the volume and their valuable comments on previous versions of the paper. I am also very grateful to an anonymous reviewer for providing me with helpful comments and suggestions, and to Ray Hardesty for carefully proofreading the manuscript. All remaining errors are my own.

 Shoji Takano

systematically variable uses of focal prominence in close linkage to interactive dimensions of style and sociolinguistic properties of social situations (i.e., register) in Japanese language practice. While prosodic prominence involves different acoustic parameters such as amplitude, duration, and pitch variation, the present analysis is mainly concerned with pitch phenomena because fundamental frequency (F0) plays the primary role in both production and perception of focal prominence in Japanese (Azuma 1992a,b; Koori 1989a,b; Sugitou 1982). Surprisingly little attention has been paid to the analysis of focal prominence in intonational phonology. It follows that in the current state of the discipline researchers have not yet drawn any satisfactory conclusion as to why and how a speaker places prosodic focus on certain elements of the utterance (Ladd 1996). A major universalist viewpoint, which is concerned with information structure in discourse, proposes that new significant information in a discourse tends to be given prosodic focus (Brown 1983, Grosz and Sidner 1986, Nooteboom and Kruyt 1987, Prince 1981, and see particularly Cutler et al. 1997). This generalization, however, may be refuted on at least three grounds, which the present study aims to verify based on the analysis of everyday natural interactions from the Japanese speech community. These are: 1. Prior studies neglect a considerable amount of cross-linguistic variability involved in focal prominence phenomena and suffer from being English-centric (Ladd 1996; Yaeger-Dror 2002a,b). 2. Prior studies are mainly based on the reading of constructed sentences taken out of context or on monologue readings in a laboratory setting, where the dynamic interactive roles prosody plays in face-to-face exchanges are not taken into account (Couper-Kuhlen and Selting 1996). 3. Any conclusions derived predominantly from a single, non-interactive kind of social situation lead to an impoverished picture of what prosody can really do in everyday interactions that are inherently rich in both style and register (YaegerDror 1996, 1997, 2001). To investigate how dynamically (and systematically) focal prominence operates in interactive language use, the present study focuses on an aspect of face-to-face exchanges that requires highly interactive work: negation. The particular locus of analysis is the variable manifestation of focal prominence placed on the Japanese negative nai. It has been observed in the present data that the speaker seems to place differential degrees of focal prominence on nai depending upon the social meanings of negation at every moment of talk-in-interaction, while morpho-syntactic style in its conventional sense is highly consistent within the same social situation. In other words, the speaker prosodically style-shifts between the situationally unmarked or marked degrees of focal prominence placed on the potential locus of negation in order to convey certain paralinguistic messages. The Japanese-specific mechanism of conflict management is also directly relevant to issues of register variation. There has been a cultural stereotype that interpersonal har-



Variation in prosodic focus of nai 

mony and collective unity are highly esteemed as rigid social norms in Japanese society. In reality, however, the social norms are more flexible than the cultural stereo­type, in that they are dependent upon the dimensions of interpersonal relations within the society (Befu 1980, Ishida 1984, Krauss et al. 1984).1 Research on non-harmonious types of interactions remains relatively rare in Japanese linguistics, perhaps due to this “myth of harmony” (Jones 1990). While a few pioneering studies have dealt with a single social situation (either casual conversation or workplace discourse) (Jones 1990, 1995; Niyekawa 1984), the present study will incorporate a comparative perspective across different communicative settings, taking into account the differential effects of social situations (i.e., register) as a variable in expressions of disagreement in Japanese interactions. In order to reach a better understanding of how variable uses of focal prominence in everyday interactions are governed by sociolinguistic grammar,2 we need to identify the matrix of potential constraints on variability and their co-occurrence restrictions (Terken 1997), which are likely to be unique to each individual language (Ladd 1996). The present study adopts the variationist framework of analysis (i.e., the variable rule approach) to detect potential constraints that simultaneously influence observed variability in focal prominence and to account for the relative significance of the constraints responsible for the variation (i.e., the hierarchy of constraints) (Preston 1991, Sankoff 1986). The types of constraints to be investigated include the structural environment (i.e., the structural principles of Japanese prosody) in which the negative nai is embedded, the status of information conveyed by the negative in a discourse, particular interactive work in interpersonal exchanges (Schegloff et al.  1977), the speaker’s stance or footing of negation (Goffman 1981), and the informational-interactive properties of different social situations (Biber 1988, 1995). Two diametrically distinctive types of interactive registers and one non-interactive register serve as the data. One of the interactive registers consists of four televised one-on-one tooron ‘debates’ on political issues,3 in which the formal desu/masu-style predicate was the sole option used, without any apparent style shifting in its conventional sense. The other interactive register consists of five casual same-sex conversations between close friends, in which the plain style predicate or fragmental utterances were predominantly used. The non-interactive register, which acts as the control, con1. Ishida (1984), for example, argues that interpersonal conflicts in Japanese society are accommodated or resolved efficiently by the interaction of these four socio-cultural dimensions: uchi (‘in-group’), soto (‘out-group’), omote (‘surface or formal arena’), and ura (‘backstage or informal arena’). 2. “Sociolinguistic grammar” is equivalent to a performance grammar involving structured, rule-governed variability in language use. It is covariate with a composite of linguistic/discoursal constraints and extra-linguistic factors (Cedergren and Sankoff 1974). 3. Japanese tooron is broader in its concept than “debates” in Western cultures. It includes various formats of confrontational talk such as one-on-one debates and group, panel, and roundtable discussions. See Section 4 for a detailed description of the data.

 Shoji Takano

sists of news read by newscasters on televised news programs. The three types of data were transcribed, and all of the utterances involving the negative nai were analyzed prosodically, based roughly on the Japanese ToBI labeling scheme of prosodic transcriptions (Venditti 1995, 2005).

2. Previous work on variables in focal prominence According to Ladd (1996), there have been two major theoretical stances that attempt to account for phenomena of focal prominence in prosody. The first stance resorts to “highlighting-based” accounts (Ladd 1996: 163), in which focal prominence plays a pragmatic role, being typically given to words or phrases of relatively heavier semantic weight than the others in an utterance. Any salient entity that has new information status in the flow of discourse tends to be pronounced with focal prominence, whereas entities that provide old, previously mentioned, or insignificant information are likely to be de-emphasized prosodically. Though this generalization is meant to establish the “universalist” principles that can apply to any human language, it can be criticized as “English-centric” in that there is abundant evidence of cross-linguistic variability (Ladd 1996: 168–97). This stance takes into serious account such pragmatic notions as discourse salience and speaker intentions, but it lacks attention to potential structural constraints (i.e., languagespecific patterns of accents and intonation) on the realization of focal prominence. The other stance centers on what Ladd (1996: 163) calls “structure-based” accounts, in which focal prominence is subject to the prosodic structure unique to each individual language, and thus is a “non-universal” phenomenon. It emphasizes the rule-governed nature or autonomy of patterns of focal prominence in natural speech. Once the focused element of the utterance is specified, the prosodic pattern of the rest is predictable by language-specific rules or structural principles. However, a variety of contextual incentives that may cause the speaker to choose certain elements of the utterance on which to place focal prominence are “at best poorly understood,” and more research from interactional perspectives is needed (Ladd 1996: 164, 197–99). In the Japanese language context, research on focal prominence has always been heavily concerned with its relationships with the phonological properties of the utterance (especially with lexical accents) (e.g., Hattori 1933; Kawakami 1957, 1965; Kindaichi 1951; Oishi 1959; Wada 1975). Recent studies have been most active in the field of laboratory phonology and have proposed non-universal principles specific to Japanese. Sugitou (1985, 1986), for example, provides empirical evidence that disproves highlighting-based accounts like Cutler’s (Cutler et al.  1997 and papers cited there) and the studies summarized in Hirst and Di Cristo (1998). Sugitou found that the word providing new information in a discourse is unlikely to be produced with pitch (F0 or fundamental frequency) prominence by native Japanese speakers, whereas speakers of Western languages emphasize such words with much higher pitch than the words providing old information. Instead, in Japanese, a significant correlation has



Variation in prosodic focus of nai 

been found between focus and its syntactic position. Pitch prominence is placed typically on whichever content words are located in the utterance-/phrase-initial position, regardless of the information status they represent in a discourse (Sugitou 1985, 1986; see also similar claims in Koori 1989a,b).4 In connection with this positional constraint on the realization of focal prominence in Japanese, there are also other language-specific characteristics of Japanese prosody that may be relevant to variable phenomena of focal prominence: downstep (or catathesis) (Beckman and Pierrehumbert 1986, Kubozono 1989, Pierrehumbert and Beckman 1988) and degeneration of pitch accents (Koori 1989b, Maekawa 1994). Downstep is a phonological process involving an iterative decline and narrowing of pitch range typically found in Tokyo Japanese.5 Within an intonation phrase (IP hereinafter), the pitch range of the succeeding accentual phrase(s) (AP hereinafter) becomes narrower when preceded by an accented AP.6 It is claimed that focus blocks this propagation of downstep and functions to reset the pitch range, introducing a new IP. Degeneration of lexical accents is commonly observed toward the end of an utterance (or the IP). Toward the right edge of an IP, lexical accents are likely to be weakened due to creaky phonation or amplitude lowering. Consequently, pitch movement is highly leveled, often without salient pitch accent realizations, as an IP proceeds. In fact, the IP-final degeneration of pitch accents is so prevalent and characteristic of Japanese that it has led Japanese ToBI researchers to establish an independent tier for “finality contours” (Venditti 1995: 17). It is then quite likely that prevalent downstepping of pitch and degeneration of pitch accents create an antagonistic phonetic environment for the realization of focal prominence in Japanese speech production. Examining realizations of focal prominence in connected speech, other researchers have demonstrated that factors such as combinations of the accentual patterns of the word in focus as well as of adjacent words, and focus on postpositional particles and auxiliary verbs, interact with one another and together affect the intonational pat4. It should be noted that there is a study focusing on information structure at the sentence level. Equating focal prominence with such a paralinguistic dimension as the speaker’s “focus of appeals” (“uttaekake no shooten”), Koori (1997b: 140) argues that focal prominence is likely to be placed on the word that carries significant information relative to the others in a sentence. 5. The phonological process of “downstep” is the key element of a general downtrend in Japanese intonation. The phonetic process of “declination,” which is a gradual decrease in pitch toward the end of an utterance, is also identified as another element that contributes to this downtrend. The two processes differ most distinctively in the prosodic domains to which they apply (Pierrehumbert and Beckman 1988: 57-91). 6. The accentual phrase in Japanese is the smallest prosodic unit, based on lexical accents. It is tonally defined as having particular phrase-initial and terminal tones and one phrasal tone and/or pitch accent (Pierrehumbert and Beckman 1988). Every accentual phrase is either accented or unaccented. The intonation phrase in Japanese is the largest prosodic unit, typically accompanied by pitch reset, pause, decreased amplitude, pitch lowering, and segment lengthening (Venditti 1995).

 Shoji Takano

terns of the utterance as a whole (Fujisaki et al. 1984, Kawakami 1965, Oishi 1959). Focal prominence not only is constrained by the local prosodic environment in which focal entities are embedded but also strongly affects the accentual patterns of the words preceding and succeeding the word in focus (Koori 1989b). There are also rhythmic issues to consider in accounting for the systematicity of focal prominence. The manipulation of pauses is found to be closely linked with focal prominence in Japanese (and perhaps in other languages as well) (Koori 1989b, Sugitou 1982). The speaker typically places pauses either immediately before or after the entity in focus so that the focal word can attract the listener’s attention. The location of these emphatic pauses does not have to coincide with syntactic boundaries such as that between subject and predicate. They can be used independently for the sake of prosodic emphasis. In addition, slow tempo also typically co-occurs with prosodic focus (Koori 1989b). What has been reviewed so far clearly suggests that patterns of focal prominence in Japanese are constrained by the language-specific structural environment in which it takes place. As Terken (1997) states, one of the vital questions that needs to be addressed in studies of focal prominence is understanding this complex matrix of structural constraints on its systematicity. The present study attempts to accomplish this.

3. Prosodic focus on negation and issues of style and register The data analyzed in past studies of focal prominence are derived predominantly from speech produced in non-interactive contexts (i.e., the reading of sentences in isolation or monologues). In a wide variety of studies that show how dynamic the roles of prosody are in talk-in-interaction (e.g., Couper-Kuhlen and Selting 1996, Gumperz 1982), the main unsolved problems in intonational phonology are consistently seen as arising from the examination of “citation forms” (i.e., socially de-contextualized sentences) to try to determine how sentences are accented (Ladd 1996: 198). The first robust attempt to account for these interactive aspects of prosodic focus in natural speech is a series of variationist studies conducted by Yaeger-Dror (1985, 1996, 1997, 2002a,b). Studying the great variability in focal prominence of the English negative “not” in everyday language use, the author claims that two particular principles provide a useful yardstick for understanding the observed variation. Yaeger-Dror (1997) calls the first principle the “Cognitive Prominence Principle” (CPP hereinafter), which has been put forth by a number of studies (e.g., Brown 1983, Cutler et al. 1997, Hirschberg 1990, Hirst and Di Cristo 1998, Nooteboom and Kruyt 1987, O’Shaughnessy and Allen 1983, Prince 1981). This conception captures the variation in focal prominence based on the speaker’s cognitive judgment of the information structure: new information in a discourse is given more focal prominence than other information. The other principle, which Yaeger-Dror (1997) calls the “Social Agreement Principle” (SAP hereinafter) and which was originally proposed by conversational analysts (e.g., Schegloff et al.  1977), states that there is a “universal” preference for speakers engaged in conversation to em-



Variation in prosodic focus of nai 

phasize signs of their agreement with co-participants. A corresponding dispreference for disagreement is manifested by minimizing the extent of disagreement when it arises. Yaeger-Dror’s series of studies demonstrate that the variable degrees of focal prominence on English negation are highly rule-governed in terms of the extent to which speakers are subject to those principles, and that the effects of the principles on the speaker’s performance vary systematically based on the following two types of factors: (1) the informational-interactive properties of social situations, and (2) the interactive meanings of negation at every moment of talk-in-interaction even within a single social situation. As for the communicative properties of social situations, Biber’s studies (1988, 1995) show that the registers one manipulates in one’s everyday life involve differential degrees of orientation toward informativeness and interactiveness, which compose a continuum. For example, casual conversations are considered to be a typical interaction-oriented register at the most interactive end of the continuum. In this type of register, the SAP is likely to prevail over the CPP. Interpersonal rapport and solidarity are highly esteemed over the exchange of information. In highly information-oriented registers such as discussion or debate, on the other hand, the CPP is likely to prevail over the SAP in that successful transmission of information is given the first priority over affective dimensions of interaction. Yaeger-Dror’s theory predicts that the variable uses of focal prominence in expressing disagreement should quantitatively reflect the types of social situations in which the speaker is engaged, and that they will also likely be affected by the sociocultural norms the speaker is expected to meet as a successful interactant in a given social situation in a given culture (Yaeger-Dror 2002a, b). Even within a single social situation, both the CPP and the SAP have to be dynamically “tuned in” to every moment of talk-in-interaction. In expressing his or her disagreement through negation, the speaker is obliged to determine the interactional meanings of the negatives and to control the stylistic parameters of their utterance depending upon the degree of social impact on the co-participant(s) (and others not present) in conversations. To illustrate, some major types of interactional meanings of negatives (the Japanese nai) have been extracted from my dataset, based on Yaeger-Dror’s (1997: 6–8) classification. In Excerpt (1) from a political debate, one of the participants (Speaker D3b), a male congressman, uses the negative nai twice (see the lines indicated by the arrow) in providing new information useful for the other participants in further pursuing their co-activity (political debate). (1) Negatives as “informative” (neutral) interactive meaning7 Speaker D3b: sorekara amerika demo desu ne, (.)ano:: 7. Broad transcriptions are based on the Sacks et al. (1974) system with some additions. Major symbols include: [ ] (speech overlap); underlining (emphasis); underlining and CAPS (more emphasis); (.) (short interval); (2.0) (timed interval); :: (lengthening); ? (full raise); period (fall to low); comma (fall not to low); / / (slow tempo); @ (laughter); <@ @> (laugh quality). ! (appeal); x (indecipherable syllable).

 Shoji Takano

→ →

noosu dakota mitai na, hito no toora-nai yoona (0.5) to yuu to (.) okorareru keredomo, hotondo jinkoo INAI yoona chiiki no desu ne, toko ni mo migotona ano:: haiuei dekitemasu ne.



‘Besides, also in America, um, somewhere like in North Dakota, where no one passes through, though someone may be offended if I say this. In regions that have almost NO population, splendid, um, freeways are constructed.’

The theory predicts that in highly information-oriented social situations such as this televised political debate, the negatives that provide new significant information are more likely to receive focal prominence than those in other social situations (e.g., casual conversation). Notice, however, that the first token of the negatives (toora-nai) is not given prosodic emphasis. I assume this prosodic choice would be linked to the speaker’s hesitant, self-acknowledged tone of apology when uttering the remark that his words might possibly offend the inhabitants of North Dakota (noosu dakota mitai na hito no toora-nai yoona [0.5] to yuu to [.] okorareru keredomo ‘somewhere like in North Dakota, where no one passes through, though someone may be offended if I say this’). The speaker then rephrased what he had just said using a more generalized, anonymous term (“hotondo jinkoo INAI chiiki ‘in regions that have almost NO population’), but this time with the negative highly prominent (INAI). Here, the CPP came back in, as is required in this type of social situation. Excerpt (2) comes from the same debate between Speakers D3a and D3b, who are the representatives of opposing political parties. (2) Negatives as “face-threatening” interactive meaning (Brown and Levinson 1987) Speaker D3a: ironna sono:: e:: fukushi ya nanka no bubun de, kyuuchoo-en suiageta. /sore ga/ ne (.) yappari ano:: (.) keiki ga warukunatta kore ookina genin desho! [xxx] ‘Various, um, social welfare and others spent 9 trillion yen. That is, um, still the very reason why the economy went down, isn’t it?’ Speaker D3b: →

[iya aru] teido genin aru kedo, (.) ano sore ga ano:: sengo desu ne, ano nanjuunen rai to yuu yoo na, sonna fukeiki no genin dewa NAI to omoimasu. ‘Well, it is a cause to some extent, but, um, I, um, I think that is NOT the cause of this economic recession, which has been the most serious in the decades since the War.’

Speaker D3b, using the negative nai (see the line with the arrow), directly expresses his disagreement with Speaker D3a, negating D3a’s preceding claim about a potential

Variation in prosodic focus of nai 



cause of the recent economic recession. Here, he employs a highly prominent prosodic focus on the very locus of the negation (NAI) to express his opposing stance. Notice also that Speaker D3b makes a concession remark (aru teido genin aru kedo ‘it is a cause to some extent, but’) before issuing his very overt disagreement, with the negative being highly prominent. This seemingly contradictory sequence is characteristic of Japanese interactions in conflict (Honda 2002, Jones 1995). The preceding concession can be regarded as an essential constituent of Japanese disagreement, which is motivated by the interpersonal desire to save the face of the opponent, whereas the subsequent overt disagreement is driven by the institutional context of this televised debate, in which the participants are expected to meet the communicative requirement of expressing their opposing views clearly to the audience (Honda 2002). Obviously, Speaker D3b chose the CPP over the SAP when issuing the negative as his stance in order to meet the communicative requirement of this particular social situation as a competent member of the culture. Excerpt (3) comes from a social situation, in which the SAP is highly esteemed – a casual conversation between college classmates (Speaker C2a and C2b). (3) Negatives as “supportive” interactive meaning Speaker C2b: nani minna no wa kara hazusareru no? ‘What? Are you gonna kick me out?’ Speaker C2a: @@@@@[@@@] Speaker C2b: [naze?] @@@ [nanka warui <@koto shita?@>] ‘Why?’ @@@ Have I done anything wrong?’ Speaker C2a:

[@@@@@@@@] (0.5) sooyuu wake ja na::i tte sa::. @@@@@@@@ ‘That is not the case! (of course, you’ve done nothing wrong.)’

Discussing room assignments for their class trip to a hot spring town, Speaker C2b jokingly reacts to Speaker C2a’s preceding tease that Speaker C2b will have to stay in a room with other classmates with whom she is not close. When Speaker C2b playfully pursues the issue, Speaker C2a finally provides a supportive statement with the focused negative with heavy lengthening (na::i) to maintain Speaker C2b’s positive face, implying that she has not been cast out of their social circle. While Speaker C2a’s prosodic emphasis on the negative strongly denies the propositional content of Speaker C2b’s preceding utterance (nanka warui koto shita?), it reconfirms interactive meanings of in-group rapport and solidarity, with the result that the principle of social agreement (“disagreements preferred”) is satisfied (Pomerantz 1984). It has been demonstrated that the speaker systematically controls the degree of focal prominence on negation, depending upon the type of social meaning a particular token of negation creates at a specific moment in talk-in-interaction (Takano 2001,

 Shoji Takano

Yaeger-Dror 1996, 1997). Putting aside the differential properties of social situations mentioned above for the moment, the theory predicts that focal prominence on the face-threatening nai in Excerpt (2) is likely to be minimized and that the supportive nai in Excerpt (3) is likely to be maximized in accordance with the SAP. The informative (thus, neutral to the interlocutor’s face wants) nai in Excerpt (1), on the other hand, is likely to be prominent because the efficient delivery of information is a major concern for the speaker, in accordance with the CPP. Close examination, however, shows that social situations of language use intervene in the activation of these principles. As legitimate members of a culture, the participants need to activate those principles properly by evaluating their interactive impact on the other participants in a given social situation and the communicative requirements of the situation. Thus, once highly interactive data are closely examined, the status of information conveyed by negatives in a discourse is not the sole determinant of focal prominence (which is in discord with the universalist paradigm). Rather, the negatives with or without prosodic focus carry significant paralinguistic messages that are dynamically adjusted from moment to moment even within a single interaction. It is already evident from these sample exchanges that the choice mechanism of prosodic styles is indeed multi-layered, in that moment-to-moment mappings of interactive principles such as the CPP or the SAP interact in some principled ways with social situations of language use as the overall normative framework. This is why systematic accounts of social situations (i.e., style and register) should be integrated into the process of building theories about language use as one of the essential considerations.

4. Data The data for the present study come from three social situations – two diametrically contrastive situations in terms of the informational-interactive continuum and one non-interactive situation as the control. The first social situation, at the most interactive end of the spectrum, consists of five informal, dyadic, same-sex conversations between close friends, and the second social situation, more highly informational, consists of four televised tooron (‘debates’) between male politicians. Two of the tooron are one-on-one debates, and the remaining two are more like one-on-one panel discussions with the moderators being present. The control data set consists of broadcast news read by several newscasters on televised news programs in Japan. Table 1 describes the number of the negative nai tokens analyzed and each speaker’s demographic background. All of the participants in the five informal conversations (C1a through C5b) are speakers of the Hokkaido dialect, which is considered to be relatively similar to Tokyo Japanese in its accentual system. The four televised debates involved leading figures of major opposition parties (D1a through D4b).

Variation in prosodic focus of nai 



Table 1.  Number of the Japanese Negative -NAI Analyzed and Speaker’s Background REGISTER

SPEAKER

 

      News

C1a C1b C2a C2b C3a C3b C4a/C5a C4b C5b   D1a/D2a D1b D2b D3a D3b D4a D4b   6 Newscasters

 

Grand Total:

Casual Conversations             Debates

NO. OF TOKENS 72 20 51 35 51 39 190 66 53 Total: 577 58 32 46 25 22 45 59 Total: 287 161 1025

SPEAKER’S BACKGROUND female, mid 20s, office worker female, late 20s, housewife female, early 20s, college student female, early 20s, college student female, early 20s, college student female, early 20s, college student female, late 60s, housewife female, early 70s, housewife female, late 60s, housewife male, late 50s, politician male, late 50s, politician male, late 50s, politician male, late 50s, politician male, late 50s, politician male, late 60s, politician male, late 50s, politician   3 male, late 30s-late 40s, 3 female, late 20s to late 30s  

Notes: The speaker symbol [C4a/C5a] means that this particular speaker participates in Casual Conversations No.4 and No.5. The speaker symbol [D1a/D2a] means that the speaker participates in Debates No.1 and No.2.

It is unknown, however, which specific dialect each participant in the debates and the news broadcasts speaks as his or her vernacular and how variable the prosodic patterns are among the speakers of different dialects. With regard to the current database, my impression is that all of the participants spoke a common or nearly common variety of Japanese (kyootsuu-go), which is appropriate for formal communicative settings such as these nationally televised programs. A total of 1025 occurrences of the Japanese negative nai were analyzed in terms of prosodic prominence.8 8. The following specific types of utterances were excluded from analysis: frozen/idiomatic expressions (e.g., -sezaru o enai, -nakereba naranai, -kamo shirenai, etc.), contracted forms (e.g., -nakereba --> nakya), archaic forms (e.g., yomanu, kozu, etc.), interrogative utterances (e.g., nomanai no?) and imperatives (e.g., iwanaide).

 Shoji Takano

5. Analytical approach 5.1

The domain and criteria for prosodic prominence

For my analysis of focal prominence on nai to be as consistent and objective as possible, I tonally transcribed the present corpus, roughly adopting the Japanese ToBI labeling scheme of prosodic transcriptions (Venditti 1995, 2005)9. In addition to tonal events, my transcription was also focused on the identification of disjunctures between adjacent words and the assignment of an appropriate value of the Break Index (BI hereinafter) – BI 3 for the IP and BI 2 for the AP.10 An IP can be identified by the most strongly perceived disjuncture (BI 3), and the perceived strength of this disjuncture is based on a number of prosodic factors and their combinations. IP final disjunctures typically consist of F0 lowering, decreased amplitude, segmental lengthening and pause, and IP initial junctures are typically marked by pitch range reset.11 An IP in Japanese consists of one or more APs (BI 2), which is a prosodic unit above the word. The Japanese AP is identified by a medium degree of disjuncture, which is generally characterized by particular tonal events (H*+L, H-, or both) as well as falling tonal movement (L%), but it lacks such prosodic cues as the IP markers (e.g., F0 lowering, decreased amplitude, etc.). Every AP is either accented or unaccented, depending upon the words involved, and it is perceptually a larger unit than a word (BI 1) but smaller than (or equal to) an IP (BI 3). Whether the negative nai is prominent can be judged from several specific criteria. Based on past studies that regard pitch as playing the primary role and intensity and duration as playing secondary roles in phenomena of Japanese focus (Azuma 1992a,b; 9. The Japanese ToBI labeling scheme is designed to transcribe Tokyo Japanese, elicited mainly from read speech, although it is meant to be continuously revised to accommodate other styles of spontaneous speech. Therefore, I must say that due to a number of technical problems such as the difficulty in obtaining clear pitch tracks, etc., I have “roughly” taken advantage of only certain portions of the scheme to transcribe naturally occurring spontaneous speech in the current study. In addition, most of the speakers transcribed are speakers of either the Hokkaido dialect or some unknown regional variety, although all of them seem to command a nearly common variety of Japanese (kyootsuu-go). 10. In the ToBI labeling scheme, “disjuncture” means the degree of prosodic association between two sequential words, and the “break index” represents the subjective values based on the measures of perceived disjuncture between adjacent words (Venditti 1995). The Break Index in Japanese ToBI ranges from 0 to 3, but I omitted the Break Indices 0 and 1 in my transcriptions because of the analytical objectives. 11. Other criteria for determining an intonation unit come from Cruttenden’s notion of “intonation group boundaries” (1986: 37) and Chafe’s discourse notion of “intonation unit” (1994: 57). In the former, a unit is based on an identification of such markers as pause, anacrusis, final syllable lengthening, and pitch level change, whereas the latter includes a wider repertoire of devises such as pre- and post-pauses, acceleration/deceleration in syllable duration, overall pitch declination, and voice quality (e.g., creaky voice at the end).

Variation in prosodic focus of nai 



Koori 1989a,b), I directed my analyses primarily toward pitch contours (i.e., tonal events) given to the negative nai and the identification of an IP in which the negative nai was embedded, while paying attention to any marked use of intensity or duration as well. If a negative nai was recognized as being tonally a part of an AP in which it was embedded within a single IP, that particular token was coded as receiving “no focal prominence.” If a negative nai tonally initiated a brand-new IP, typically along with pitch range reset, that particular token was coded as receiving “focal prominence.” I conducted these coding processes using the speech analysis software program PitchWorks, which permits analysis of pitch movement (F0), intensity (loudness), and duration among others. To illustrate the actual coding procedures in more detail, I discuss Figures 1 through 3 below. In Figure 1,12 three IPs can be marked by BI 3 (see the 4th BI tier from the top): the 1st IP, Irena’i? (Speaker C2a asks Speaker C2b, ‘Don’t you tuck it [your shirt] in [your pants]?’), which is also a single AP; the 2nd IP, Pa’jama no ue wa irenai kedo (Speaker C2b responds to C2a’s question, saying ‘I do not tuck in my pajama shirt, but,’), consisting of three APs (/pa’jama/, /ue wa/, /irena’i kedo/); and the initial portion of the 3rd IP (pa’jama no na’ka ni kiru …) is also seen (Speaker C2b continues, ‘[what] I wear inside my pajamas ….,’). Though the negative nai is involved, the 1st IP (irena’i?) has been excluded from analysis because it is an interrogative utterance. As

Figure 1.  Sample Analysis 1 12. In the word tier in the transcriptions, /x’/ indicates the location of pitch accents.

 Shoji Takano

can be seen, C2b’s response (pa’jama no ue wa irenai kedo pa’jama no na’ka ni kiru …) is divided into two separate IPs in that at the beginning of the second clause (pa’jama no na’ka ni kiru …) the pitch contour is reset, which signals the start of a new IP. Thus, the domain of analysis of nai prominence in this particular file is the 2nd IP (Pa’jama no ue wa irenai kedo), and so this particular token was coded as non-prominent due to the fact that the negative can be regarded as tonally a part of the AP in which it is embedded (irena’i kedo ‘[I] do not tuck [it] in, but’) within the single IP. In Figure 2, C3a’s utterance (Kyooshitsu i’doo toka anma NA’I jan shooga’kkoo nande, ‘There aren’t many classroom changes, right, because it’s an elementary school’) was divided into two IPs, but a crucial difference from Figure 1 concerns the fact that it is the negative (NA’I ‘none’) that initiates a brand-new IP, along with pitch range reset derived from prosodic focus (BI 3).13 This particular token was coded as prominent. Japanese is a lexically accented language. Depending upon the accentual type of the lexical item to which nai is attached as the negative suffix, as well as the prosodic environment that immediately follows, the negative -nai itself should be either

Figure 2.  Sample Analysis 2

13. This particular BI could be identified as 2m (2 mismatch) rather than 3, in that the perceived disjuncture between the former AP (anma) and the latter AP (na’i jan) is not as strong as a normal IP boundary (BI 3), although the tonal pattern clearly indicates the beginning of a new IP.

Variation in prosodic focus of nai 



Figure 3.  Sample Analysis 3

accent-bearing or unaccented.14 Because of these restrictions, when nai is unaccented based on the prosodic environment in which it is embedded (i.e., nai itself cannot bear its accented pitch), the presence or absence of focal prominence was judged on the basis of the tonal patterns given to the lexical item to which nai is suffixed. Figure 3 illustrates this point. In this utterance by D3b (Kihonteki ni ie’ba nihon no shoo’rai wa NAORA’nai to omoimasu, ‘Basically [speaking], I think that Japan’s [economic] future will not recover.’), the negative nai is a suffix to the accented verb, nao’ru (‘recover’), the lexical accent of which is transformed to the accentual pattern naora’nai (‘recoverNEG’) – i.e., the negative nai cannot bear its own pitch accent. While it is obvious that the pitch placed on this unaccented nai itself shows a falling contour, this token was coded on the basis of the pitch contour directly assigned to the entire phrase naora’nai. As is evident in Figure 3, the AP naora’nai initiates a brand-new IP with pitch range reset due to focal prominence. Notice also that the final AP (to omoimasu, ‘I think that’) is greatly reduced in pitch and intensity because of the lowered amplitude and

14. The grammatical status of the Japanese negative nai is twofold: (1) the auxiliary verb suffixed to verbs and other auxiliaries, or (2) the negative adjective that predicates nouns, other adjectives, adjectival nouns, and nominalized phrases, often with postpositional particles (wa, de, dewa) intervening (Nihongo Kyouiku Gakkai 1993). Unaccented nai tokens are most likely to occur when nai is suffixed to accented verbs.

 Shoji Takano

creaky phonation. This prosodic weakening toward the end of the IP is characteristic of spontaneous Japanese speech, as mentioned earlier. In addition to these primary criteria, the following cases were also coded as receiving focal prominence, even if no pitch reset was observed: (1) nai itself or the lexical item to which nai is suffixed is given a highly marked degree of loudness or duration (though this was rarely observed), and (2) a preceding lexical accent is moved to an inherently unaccented nai, which contributes to its perceptual saliency (e.g., tabe’ru ‘eat’ → tabena’i rather than a standard pattern tabe’nai ‘don’t eat’). Based on the criteria discussed so far, every coded token of nai was double-checked by another native speaker of Japanese. If the judgments did not match, the token in question was discussed until agreement was reached.

5.2

Hypothesis of potential constraints and the analytical program

A predominant majority of prior studies have investigated one-on-one correlations between prosodic focus and some potential constraint (e.g., information structure). To investigate variable grammars, however, it is prerequisite to presuppose that the speaker’s decision to place prosodic prominence on the negative nai is affected by a variety of factors simultaneously (Sankoff and Labov 1979). Based on previous studies of Japanese focal prominence and on preliminary investigation of my own, Figure 4 describes my hypothetical model that accounts for systematic variability in Japanese focal prominence. The model consists of a wide variety of constraints at different levels of linguistic practice: the internal structure of the IP in which the negative nai is embedded, the status of information conveyed by the negative, interactive dimensions, and social situations. To examine the model, I adopted the variable rule approach, which is specifically designed to handle the inevitably skewed sociolinguistic data elicited from natural communicative settings (Sankoff 1985, 1986, 1988).15 Specifically, I used GoldVarb (Rand and Sankoff 1990), the Macintosh application of the variable-rule-based statistical model developed by Cedergren and Sankoff (1974). The program conducts a multivariate analysis of data using the maximum-likelihood technique, which allows one to measure the relative effectiveness among the intersecting factors as potential constraints and yields a probability estimate of the effect of each constraint on the application of the rule in question (i.e., focal prominence on nai) in relation to the other remaining constraints. Because of the skewedness of sociolinguistic data and the 15. A similar statistical program is ANOVA. Algorithms for calculating ANOVA, however, normally require balanced numbers of tokens in each cell, which would be possible only with data from controlled experimentation (Young and Bayley 1996). Therefore, VARBRUL is the only alternative for successfully handling the extremely skewed sociolinguistic data derived from naturally occurring speech (see Young and Bayley 1996 for further discussion of the validity and implementation of VARBRUL for sociolinguistic research).

Variation in prosodic focus of nai 



1. Morphosyntactic Status of –NAI: Verbal Suffix, Negative Adjective

6. Hierarchy of Information: Main Clause, Subordinate Clause, Embedded Clause

7. Information Status: Brand-new, Old, Inferable, Contrastive

2. Proximity to the End of the IP: Long Distance, Short Distance

8. Preface to –NAI: Hint, Upgrader, Absence

3. Other Accented AP Preceding –NAI: Presence, Absence

9. Postposing: Variable: FOCAL PROMINENCE

4. Other Focus Preceding –NAI: Presence, Absence 5. Rhythmic Issues: Pre-pause, Post-pause, Slow Tempo, No Pause

11. Social Situations: News, Casual Conversation Debate

Presence, Absence 10. Footing: FTA, Informative, Supportive, SelfProtective,Selfdenigrating, Selfcorrection, Agreement Seeker , FTA with Humor

Potential Effects Internal Structures of the Intonation Phrase Status of Information Conveyed by –NAI Interactive Dimensions Registers

Figure 4.  Simultaneous effects of hypothesized independent variables on focal prominence

necessity of identifying the complex intersecting relationships among a number of potential factors, probabilistic accounts of occurrences or non-occurrences of a variable in question have proven to be superior to the use of bare percentages (Sankoff 1985, 1986, 1988). The program also allows us to conduct a stepwise regression analysis, which goes through multiple permutations of factor groups and sorts out the groups of variables with statistically significant distributions of factor weights as the optimal set of predictors of the dependent variable. The results obtained can be considered to describe the sociolinguistic grammar of Japanese focal prominence, “predicting” the effect of each of the constraints for the use or non-use of focal prominence on nai based on the present dataset as a representative sample of the speech community (Preston 1989). Here is a brief description of each potential constraint that is hypothesized to be responsible for systematic variability in Japanese focal prominence. Prior research relevant to the constraints is listed in parentheses.

 Shoji Takano

Internal Structures of the Intonation Phrase 1. Morpho-syntactic status of the negative nai: verbal suffix, independent negative adjective. The Japanese negation marker nai has two distinct morpho-syntactic statuses, as follows: (1) -nai as a verbal suffix (e.g., hana’su ‘talk’ → hanasa’-nai ‘don’t talk’; iku ‘go’ → ika-na’i ‘don’t go’),16 in which -nai is prosodically part of the preceding verb. Thus, whether -nai itself carries its lexical accent depends upon the lexical accent types of preceding verbs as well as on the succeeding prosodic environment in natural speech production. (2) The word nai, as a negative adjective, that is preceded by nouns (e.g., kane ga/wa na’i ‘(I) have no money’), adjectives (e.g., oishiku [wa] na’i ‘not delicious’), or adjectival nouns (e.g., kirei de[wa] na’i ‘not beautiful’) with or without postpositional particles, carries its independent lexical accent.17 It can be assumed that the morpho-syntactically independent use of nai (Type 2) has a differential status as a constraint from the morphologically more dependent verbal suffix -nai (Type 1), in that the former -nai should be perceptually more salient than the latter in delivery of new information and paralinguistic messages. 2. Proximity of the nai token to the end of the IP (BI 3 boundary) (Sugitou 1985, 1986; Koori 1989a,b): long (6 or more morae), short (5 or fewer morae). I measured the distance of nai from the end of the IP, counting the number of morae. In Figure 3 (…NORA’nai to omoimasu), for example, the distance is eight morae including the two morae of nai itself to the end of the IP. 3. Other accented AP(s) preceding the nai token within the same IP (Fujisaki et al. 1984, Kawakami 1965, Koori 1989b, Oishi 1959, Pierrehumbert and Beckman 1988): presence, absence. Past studies of connected speech data from read materials show that preceding lexical accents dynamically affect the shape of other lexical accents in the succeeding prosodic environment. The IP in Figure 1, for example, involves an accented AP preceding the negative nai: pa’jama (‘pajama’). 4. Other focused element(s) preceding the nai token in the same IP (Fujisaki et al. 1984, Pierrehumbert and Beckman 1988): presence, absence. For example, in the single IP Zettai sensoo ni’wa nara’nai to iitsuzu’keta n desu yo. (‘I kept saying [it] will absolutely not result in war.’), the adverb zettai (‘absolutely’) is given prosodic focus preceding the negative nara’nai (‘become-Neg’). Previous literature shows that the prosody of the segments following focal prominence is heavily reduced within a single IP.

16. “x’x” indicates the location of a lexical accent. 17. It is still debatable whether the grammatical status of nai co-ocurring with adjectives (e.g., oishiku-wa-nai), for example, should be considered as a verbal suffix type (ika-nai) or as an independent negative adjective (kane wa nai) (Nihongo Kyooiku Gakkai 1993: 383).



Variation in prosodic focus of nai 

5. Rhythmic issues – pause and tempo (Koori 1989b, Sugitou 1982): pause preceding nai, pause immediately following nai, slow tempo (often along with some hedges) (e.g., anoo ‘well’), no pause. Pauses may be used at any one of the following places: before the lexical item negated by nai, right before nai itself (very rarely), or right after the nai in the same IP. Speech production involving the negative token occasionally becomes slower. Information Status 6. Hierarchy of information (Azuma 1992b, Koori 1997b): main clause, subordinate clause, embedded clause. In terms of the hierarchy of information, the main clause generally delivers the primary information of the utterance as a whole, which would be more likely to receive focal prominence according to the universalist claim. The subordinate clause generally delivers the secondary information that would be less likely to be given focal prominence. For analytical purposes, the category of main clauses includes utterances that either share some hierarchical relations with other clauses (shusetsu ‘main clauses’ in fukubun ‘complex sentences’) or stand alone without such relations (tanbun ‘simple sentences,’ including fragmental/inconclusive utterances as well as juxtaposed sentences in juubun ‘compound sentences’). The category of subordinate clauses is further divided into two groups: subordinate clauses (typically with conjunctions), which are subordinate to a main clause in a complex sentence, and embedded clauses, which are typically quotative utterances, clausal nominals, or modifying clauses. 7. Information status of nai in discourse (Prince 1992): brand-new information, contrastive information, inferable information, old information. As reviewed earlier, a number of studies (mainly of English intonation) in interactively impoverished communicative settings regard the informational content of the word or phrase in relation to the prior context of discourse as the impetus for focal prominence – i.e., new significant information in the flow of discourse tends to become prominent. The theory predicts that negation, which inherently adds new or contrastive propositional content to an ongoing discourse, will be a good candidate for focal prominence. Close examination of information status in the context of discourse in the present data shows that while many of the nai tokens provide new information, there are a number of nai occurrences in which information can be inferred by the listener from the preceding context of discourse or through interactants’ shared knowledge. Repetitive nai tokens can also be considered to represent old information.

 Shoji Takano

Interactive Dimensions 8. Preface to the nai negative (Jones 1990, Levinson 1983, Mori 1999, Pomerantz 1984, Schegloff 1980, Yamada 1992):18 discourse marker as an upgrading preface, discourse marker as a hint preface, straight negation without any preface. Three types of prefaces are identified in the present study. The first consists of the clause-initial uses of: (1) discourse markers which “upgrade” or “assert” the upcoming negatives (e.g., datte ‘because,’ dakara/desukara ‘so,’ [sore]de ‘then,’ shitagatte ‘therefore,’ or combinations of these), (2) those which “hint” that a negative is on the way (e.g., demo ‘but,’ shikashi[nagara] ‘but/however,’ dake[re]do[mo] ‘though,’ kedo ‘though,’ tokoroga ‘but,’ [i]ya ‘nah/no,’ or combinations of these), and (3) straight negation without any such discourse markers. It has been pointed out in studies of disagreeing turns in both Japanese and English that certain linguistic materials are exploited as markers of opposition to preface (or contextualize) upcoming disagreements. The present analysis explores the possibility of any meaningful interplay between those devices and prosodic parameters. 9. Concurrence with the shift of linguistic elements to the post-predicate position (i.e., dislocation/postposition) (Fujii 1995, Hinds 1982, Ono and Suzuki 1992): presence, absence. The dislocation of linguistic elements to the post-predicate position in Japanese, an SOV language, has to do with the speaker’s emphasis of information in a discourse. How prosodic emphasis interacts with this syntactic alternative will be investigated. 10. Footing (Goffman 1981; Yeager-Dror 1996, 1997): informative (or neutral to the interlocutor’s face wants) (see Excerpt [1] above), face-threatening (e.g., Excerpt [2] above), supportive (see Excerpt [3] above), self-protective (or making excuses) (e.g., Speaker C3b, making an excuse for her recent poor performance and disinterest in a school subject: Aru to omou kara benkyoo-shinaishi sa. ‘Because I know that I possess [the copies of previous exams to refer to], I don’t study hard [for the exam].’), self-denigrating (e.g., Speaker C3a, responding to Interlocutor C3b’s previous statement that she has succeeded in losing some weight: Yaserenai ‘I can’t lose weight.’), self-correction (e.g., Speaker C3a, talking about a former teacher of hers in junior high school: Kibishii desho, Naguru desho, (.) Naguri wa shinaika. ‘He WAS strict, and HIT [his students]. He didn’t hit, exactly.’), agreement seeker (e.g., Speaker C1a to Speaker C1b: Ichiichi hiyakedome nante nutterarenai ssho. ‘You cannot apply the lotion to prevent sunburn every time [you go out].’), facethreatening with humor (only in casual conversation data) (e.g., Speaker C3a, responding to Interlocutor C3b’s preceding joke that C3b’s mother, who is going to visit her, may follow her around on campus, even to the students’ cafeteria, but the

18. The scope of analysis of prefaces extends beyond the intonation phrase, which is the basic unit of analysis as discussed in Section 5.1.

Variation in prosodic focus of nai 





mother is not permitted to enter the cafeteria without a student pass: Sonna kimari naitte. ‘There is no such rule.’). In the act of negation, the speaker not only negates the propositional content but also creates a particular footing, considering the negation’s social meanings and impact on the listener at every moment during interaction. In the present dataset, the negative nai is also used to fulfill a wide variety of interactive meanings as well as to directly express the speaker’s disagreement with the listener.

Social Situations 11. Social Situations: casual conversations, political debates, news broadcast. It is assumed that prosodic variation is motivated by the speaker’s sensitivity to the interactional norms for particular communicative settings in terms of the degree of informativeness and interactiveness and the communicative requirements.

6. Sociolinguistic grammar of prosodic focus on the negative nai 6.1

Variability linked to social situations

Among the outputs of several GoldVarb runs of stepwise regression analysis,19 Table 2 shows the variability linked to the three social situations in terms of the frequencies of prosodic focus on the negative nai and their corresponding probability estimates (VR weights). As mentioned earlier with regard to Figure 4, VR weights indicate the predicted relative effectiveness among all of the intersecting factors that are hypothesized to Table 2.  Variation Specific to Three Social Situations in Prosodic Prominence on -NAI SOCIAL SITUATIONS

RATES OF PROMINENCE

Casual Conversations



33%

(190/577)

0.40

Debates



48%

(137/287)

0.64

News Broadcast



39%

(62/161)

0.59

TOTAL: 38% (389/1025)

p <.05

 

VR WEIGHTS

Input Value = 0.366, Chi-square per cell = 1.0250

19. These GoldVarb runs were conducted with all of the independent variables listed in Section 5.2 involved. I needed to perform multiple runs of the program because of possible interactions among the independent variables. See the explanation in more detail in Note 22 in the next section (6.2).

 Shoji Takano

simultaneously affect variable uses of focal prominence. A weight of 0.50 indicates that the factor has no effect on the production of the dependent variable (i.e., prosodic focus on nai). The closer the weight is to 0, the more strongly the factor inhibits the dependent variable. The closer the weight is to 1, the more strongly the factor promotes it. At the bottom row of the table, the average “chi-square per cell” indicates the degree to which the independent variables considered (i.e., the hypotheses constructed) account for the data. The smaller this figure is, the surer we can be that it is not necessary to consider additional variables. Values below 1.5 (conservatively, 1.0) indicate that the fit between the hypothesized model and the data is good (Preston 1989: 15–6), and the present value (1.0250) is within that secure range. The “input value” (0.366) indicates the likelihood value for the dependent variable (focal prominence on the negative nai) to be used regardless of the effects of independent variables. The independent variable “social situations” was identified as being statistically significant by stepwise regression analysis (p <.05). The results demonstrate that prosodic focus on Japanese negatives is sensitive to social situations. In accord with Yaeger-Dror’s (1997) predictions with regard to English negatives, Japanese focal prominence is also governed by the universalistic SAP and CPP, and the degree to which it is affected by the principles depends upon the differential orientations of social situations to the continuum of interactiveness/informativeness. In a highly interaction-oriented situation such as casual conversations (CC hereinafter), prosodic focus on the negative is generally inhibited due to the importance of the SAP over the CPP (0.40).20 In contrast, in two highly information-oriented social situations such as political debates (PD hereinafter) and news broadcasting (NB hereinafter), focal prominence on the negative is generally promoted due to the CPP at 0.64 and 0.59, respectively. The higher probability weight for PD quite likely comes from the adversary nature of the social situations, in which negation as the key element of disagreement is more likely to be emphasized for the achievement of the communicative goals required by these particular social situations. Thus, variability in Japanese prosody, at least in negative statements, is systematically linked to social situations, and we should be careful not to overgeneralize the outcome of prior investigations based on a single register (i.e., read speech, conversations) as representing the whole range of communicative settings in the Japanese speech community (cf. Takano 2001).21 20. Some readers might find this particular weight (0.40) higher than expected. Possible reasons for this will be presented in Section 6.4. 21. My prior work (Takano 2001) on the same variable (i.e., prosodic prominence on nai) reveals that this situation-dependent systematicity also holds true in workplace interactions. This workplace corpus displayed a gradient shift in uses of focal prominence, with different footings based on different properties of settings across the informativeness/interactiveness continuum. Business meetings, where informativeness is highly required, involved the highest rate of occurrences of focal prominence on face-threatening nai (52%), and business-linked interactions outside of meetings involved focal prominence on the same footing 43% of the time. Casual chats unrelated to business, where interactiveness is highly esteemed, involved it only 29% of the time,

Variation in prosodic focus of nai 



6.2

Internal structures of the intonation phrase as the constraint

Table 3 shows all of the remaining results of variable rule analysis obtained from the runs of GoldVarb with the three social situations considered separately.22 The average “chi-square per cell” shows that the hypothesized models account for the data well in each of the social situations, in that all of the values (1.0373 for CC, 0.9651 for PD, 0.8183 for NB) are within the secure range (i.e., below 1.5).23 Stepwise regression analyses have selected six out of ten factor groups as being responsible for systematic variability in prosodic focus on the Japanese negative nai in CC, four factor groups in PD, and three factor groups in NB, all at a p <.05 level. Among the factor groups (FG hereinafter) that are concerned with the internal structure of the IP (FG-1, 2, 3, 4, 5), the robust effects of the position of the negative nai within an IP (FG-2) are found in all of the social situations. When the nai token is located further than 6 or more morae from the end of the IP (including the 2 morae of nai itself), it is more likely to receive focal prominence (long distance: 0.56 for CC, 0.58 for PD, 0.64 for NB). When it is located closer than 5 or fewer morae (including the 2 morae of nai itself) to the end, focal prominence on nai is likely to be inhibited (short distance: 0.44 for CC, 0.42 for PD, 0.27 for NB). I based my decision of either “short” or “long” distance on an observation of a major break existing between 5 and 6 morae from the end of the IP in terms of the frequency of nai prominence (approximately a 15% gap). While such lexical items as final particles, the extended predicate, or utterance-final connectives are typically accommodated with the negative nai within 5 or fewer morae that follow, the cases in which nai was 6 or more morae from the end of the IP tended to involve another clause or phrase following nai.

while involving the highest rate of nai prominence on the supportive footing (75%). Also, the rates of focal prominence were linked to power relationships among the speakers – with speakers of higher occupational ranks using nai prominence with both face-threatening and supportive footings more frequently than lower-ranked speakers across the settings (cf. Pearson 1988, 1989; Takano 1997, 2005). 22. Algorithms for VARBRUL do not allow for any interaction among the independent variables. Therefore, I conducted several Goldvarb runs so as not to include the factors that appear to interact with one another in a single run. Those factors groups are FG-2 and FG-6, FG-2 and FG-9, and FG-7 and FG-8. Since I obtained different versions of the results from several Golvarb runs, the VR weight for each independent variable reported in Table 3 is the one derived from a run yielding the lowest average chi-square per cell, which indicates the goodness of fit between the present model and the data. 23. Each value represents the worst (i.e., highest) chi-square per cell value among all of the GoldVarb runs from each social situation.



















Verbal Suffix

Negative Adjective

Long (6 or more morae)

Short (5 or fewer morae)

Absence

Presence

Absence

Presence

Following Pause

None

Preceding Pause

Slow Tempo

1) Morphosyntactic Status

 

2) Proximity to the End

of the IP

3) Other Accented AP

Preceding -Nai

4) Other Focus

Preceding -Nai

5) Rhythmic Issues

 

 

(62/165)

(99/368)

(91/209)

(83/293)

(9/13)

(64/207)

NONE

(7/24)

25%

48%

38%











Brand-new

Contrastive

Old

Discourse Marker as Hint



Inferred

Discourse Marker as Upgrader

No Preface

7) Information Status

 

 

 

8) Preface

 

 

(18/48)

(23/48)

(20/81)

(16/48)

(61/193)

(93/193)

(39/131)

(34/113)

31% (149/481)

33%

32%

36%

30%



Embedded Clause

 

30%



Subordinate Clause

35% (117/333)

29%

32% (174/540)

69%

31%

34% (126/370)

27%

44%

28%

38% (107/284)

38%

31% (128/412)

% -nai Prominence

Casual Conversations

 





Factors

Factor Groups

6) Hierarchy of Information Main Clause



 

SOCIAL SITUATIONS

0.47

0.59

0.68

0.35

0.47

0.50

0.56

0.47

0.47

0.52

–

0.45

0.50

0.73

0.47

0.52

0.43

0.62

0.44

0.56

0.48

0.51

VR Weight

 

53% (92/173)

43% (84/197) 52% (85/165)

 

44% (12/27) 47% (114/243)

  *p <.05

65% (11/17)

38% (13/34)

*p <.05  

49% (18/37)

45% (50/112)

54% (56/104)

 

 

 

Not Signif. 58% (52/89)

41% (24/58)

44% (61/140)

(4/6)

67% (12/18)

Not Signif. 67%

 

42% (97/229)

71% (24/34)

Not Signif. 43% (52/122)

59% (53/90)

42% (59/142)

*p <.05  

54% (78/145)

 

*p <.05

 

 

 

0.47

0.55

0.79

0.41

0.58

0.46

0.56

0.58

0.43

0.48

0.73

0.65

0.44

0.76

0.42

0.56

0.47

0.57

0.42

0.58

0.41

0.56

*p <.05

 

*p <.05









(14/60)

(27/48)

(10/57)

(3/12)

(17/48)









Not Signif.

 

 

Not Signif.

 

 

 

Not Signif.

(7/27)

(8/24)

(3/6)

(7/13)

(10/28)

(19/60)

(26/60)

38% (59/155)

NONE

50%

54%

36%

32%

43%

43% (47/110)

26%

33%

NONE

*p <.05



NONE

40% (59/149)

25%

35%

40% (45/113)

31% (35/113)

56%

18%

50% (52/104)

23%

48% (48/101)

% -nai Prominence

News (Control Sample)  

 









 



  *p <.05

Not Signif.

 

 

% -nai Prominence VR Weight Signif.

Not Signif. 39% (45/114)

 

Signif.

Political Debates

Table 3.  Variable Rule Analysis of Potential Constraints on Focal Prominence on the Negative -NAI: Three Social Situations

0.50

–

0.63

0.59

0.51

0.45

0.53

0.49

0.48

0.57

–

–

0.50

0.51

0.47

0.51

0.43

0.66

0.27

0.64

0.36

0.58

VR Weight

Signif.

Not Signif.

 

 

Not Signif.

 

 

 

Not Signif.

Not Signif.

 

 

Not Signif.

 

*p<.05

 

*p <.05

 

*p<.05

 

 

 Shoji Takano

45%

FTA with Humor (Teasing)











Agreement Seeker

Self-Protective/Excuses

Self-denigrating

FTA

Informative/Neutral

TOTAL

 

 

 

 

 

 

 

 

70%



Supportive

10) Footing



59%



No

 

(27/49)

(74/317)

(11/33)

(18/45)

(17/45)

 

0.38

0.52

0.55

0.57

0.67

0.79

0.79

0.49

0.65

VR Weight

 

(1/3)

 

48% (137/287)

   

0.42

0.55

–

0.69

0.40

–

–

0.50

0.28

 

Not Signif.

 

 

 

 

 

Not Signif.

 

* = Factor group selected by stepwise regression analysis Input Value = 0.477 Chi-square per cell = 0.9651

45% (58/130)

48% (63/130)

NONE

63% (15/24)

33%

NONE

*p <.05

 

 

 

 

NONE

48% (135/282)

(2/5)

*p <.05  

 

% -nai Prominence VR Weight Signif. 40%

Signif.

Political Debates

 

* = Factor group selected by stepwise regression analysis Input Value = 0.309 Chi-square per cell = 1.0373

33% (190/577)

23%

33%

40%

38%

(40/88)

(7/10)

(23/39)

31% (163/528)

55%

% -nai Prominence



Factors

Yes

Casual Conversations

Factor Groups

 

9) Dislocation/Postposing

SOCIAL SITUATIONS

 

–

–

–

–

–

–

–

–

–

VR Weight

 

 

–

 

 

 

 

 

–

 

Signif.

* = Factor group selected by stepwise regression analysis Input Value = 0.353 Chi-square per cell = 0.8183

39% (62/161)

39% (62/161)

NONE

NONE

NONE

NONE

NONE

NONE

NONE

NONE

% -nai Prominence

News (Control Sample)  

Variation in prosodic focus of nai 

 Shoji Takano

As a general phonetic fact, Japanese tonal events are likely to involve “declination,” a slope-like pitch decline toward the end of an utterance. Furthermore, a Japanese-specific phonological principle of “downstep” predicts that pitch range tends to become narrower after any preceding accented APs toward the end of the IP (Azuma 1993, Beckman and Pierrehumbert 1986, Kubozono 1989, Pierrehumbert and Beckman 1988). Thus, in these prosodic environments tonal events are likely to be lower in their momentum, which is also substantiated by a general phenomenon of accent degeneration to the end of the IP (Koori 1989b, Maekawa 1994, Venditti 2005). It was assumed in the present study that these general principles are counter-productive to the realization of focal prominence on the negative nai, which occurs mostly at the predicate-final position in canonical Japanese utterances. The present results clearly demonstrate that focal prominence on the negative is also heavily subject to these structural principles in Japanese intonational phonology. Furthermore, the effects of this positional constraint are indeed greatest in read speech (i.e., in news broadcasts, long: 0.64, short: 0.27), which is equivalent to the data elicitation task generally adopted in prior studies of Japanese intonational phonology (cf. Umeda 1982). Another constraint that is found to exert rather robust effects on focal prominence on the negative nai is FG-3 (the presence or absence of other accented APs preceding the negative nai), and this constraint is directly relevant to the phonological principle of Japanese downstep mentioned above. In accord with the principle, our results also demonstrate that the absence of preceding accented APs is likely to promote prosodic focus (0.62 for CC, 0.66 for NB), whereas their presence inhibits prosodic focus (0.43 for CC, 0.43 for NB). While this directionality is also maintained in PD, the magnitude of the effects is smaller (0.57, 0.47) than the other social situations and turns out not to be statistically significant. These results demonstrate that the variable realization of focal prominence is based on rather mechanical operations relative to tonal events in the preceding prosodic environment. The issue that remains to be discussed, however, concerns the fact that these mechanical principles do not seem to apply for PD to a statistically significant extent (presence 0.47, absence 0.57; p >.05). In PD, instead of FG-3, a statistically significant degree of effects is found with respect to other focuses in the prosodic environment preceding the negative nai (FG-4 in Table 3):24 the absence of preceding focuses promotes focal prominence (0.56) and their presence inhibits it (0.42). This differentiation, however, is not meaningful for CC and NB (CC: 0.52, 0.47; NB: 0.51, 0.47, respectively). Further analysis reveals that while the overall ratios of prosodic focuses preceding the negative nai do not involve so much difference, especially between PD and CC (42.5% [122 uses of other focuses/287 total tokens of the negative nai] in PD; 35.9% [207/577] in CC; cf., 29.8% [48/161] in NB), this marked use of extra focuses in the preceding prosodic environment has a considerable amount to do with the types of 24. For an illustration, see the fourth item in the list of potential constraints in Section 5.2.

Variation in prosodic focus of nai 



footings the negative conveys to the interlocutor in the immediate context of use. In PD, 50% (61/122) of such focuses “co-occur” with a face-threatening footing relevant to the interactional goals of that social situation. A sample discourse segment is given in Excerpt (4), a short narrative by Speaker D2b who participated in a televised debate as the representative of one of the opposition parties who criticized a manifesto issued by the government (Liberal Democratic Party [LDP]).

(4) Concurrence of focal prominence on the negative and other prosodic focus. → Speaker D2b: Keiki no koto mattaku haitte-inai. KOYOO no koto mo mattaku haitte-inai. Sore ga jimintoo no seisaku nan desu.

‘(The LDP’s manifesto) does not include policies on (the recession in) the economy at all. (It) does not include policies on (the decline in) EMPLOYMENT at all. That is the LDP manifesto.’ Notice that the negative nai was used in each of the first two consecutive utterances, but neither token was judged as being prosodically prominent. Instead, the speaker placed focal prominence on another element (KOYOO ‘employment’) at the beginning of the utterance. Placing marked emphasis relatively early in the IP is a typical prosodic move for reinforcing a confrontational footing in Japanese language usage in general. Koori (1997a: 197) characterizes kanjoo shuchoo-choo (‘emotionally appealing tone’) in Japanese as the combination of emphatic pitch in the initial part of an utterance and weakened pitch in the predicate (i.e., the final part of the utterance). Furthermore, as past studies point out, the presence of focus greatly restrains or weakens the dynamics of tones in the succeeding prosodic environment (Pierrehumbert and Beckman 1988), which would even more strongly highlight such an appealing tone. It is then quite plausible to assume that because of the nature of the social situation, the participants in PD take advantage of such prosodic moves as a strategy to strengthen their confrontational stances. Thus, given the two types of potential constraints (FG-3 [proceeding accented AP] and FG-4 [proceeding extra focus]) that restrain or promote focal prominence in the succeeding prosodic environment, the present results show interestingly that the effects are differentiated between different social situations. Variability in focal prominence in CC and NB is more likely to be subject to the “unmarked” principles of Japanese intonational phonology. In PD, on the other hand, focal prominence is pragmatically either favored or disfavored by the preceding prosodic focus over the structural principle of Japanese prosody. This paralinguistic manipulation seems to be taken advantage of as a prosodic strategy by the speaker, who aims to achieve specific communicative goals required by that particular social situation. The remaining two factor groups concerning the internal structures of the intonation phrase (FG-5 and FG-1) are also differentiated among the social situations. The results for FG-5 (rhythmic issues), which is specific to PD, demonstrate that a pause immediately following nai strongly promotes focal prominence (0.76) and that slow

 Shoji Takano

tempo also promotes it to an almost equal extent (0.73). The pause occurring either before the phrase involving nai or (very rarely) right before nai itself is also found to exert quite a strong positive effect (0.65). This quantitative evidence shows that the uses of pauses and slow tempo are simply not part of CC or NB registers but can be the strategic components of highly information-oriented, interactive registers like PD, in that a predominant majority of the tokens from these social situations do not co-occur with those parameters (a total of 2.8% use [16/577] in CC; 1.9% use [3/161] in NB) as compared with their distribution in PD (a total of 13.9% use [40/287]) (p <.001). Manipulating rhythmic parameters of interactions has a great deal to do with the hearer’s perception of the element in focus (Koori 1989b, Sugitou 1982). For the sake of their communicative goals, the participants in PD seem to take advantage of these parameters as a prosodic strategy to draw the hearer’s (and the larger audience’s) attention toward the element in focus and to develop persuasive arguments.25 The last structural constraint (FG-1) is concerned with the morpho-syntactic status of the negative nai, which is also found to be differentiated among social situations (see FG-1 in Section 5.2 for illustrations). Our results demonstrate that this factor is meaningful only for highly information-oriented social situations such as PD and NB (verbal suffix: 0.56, 0.58; negative adjective: 0.41, 0.36, respectively; p <.05), whereas its effects are neutralized in an interaction-oriented social situation such as CC (0.51, 0.48). I interpret these results as stemming from the interplay between the degree of perceptual saliency conveyed by different morpho-syntactic structures and different orientations toward informativeness/interactiveness required for particular social situations. The negative nai suffixed to the verb (e.g., hana’su ‘speak’ → hanasa’-nai ‘do not speak’; iku ‘go’ → ika-na’i ‘do not go’) is perceptually less salient than nai as the negative adjective (e.g., kane wa/ga na’i ‘(I) have no money’; oishiku wa na’i ‘[it] is not delicious’), which is morpho-syntactically independent, with its own lexical accent generally given. In highly information-oriented social situations (PD and NB), where successful transmission of information is given the first priority, the present results seem to demonstrate that the speakers compensate for the scarcity of perceptual saliency of the verbal suffix -nai by exploiting prosodic focus (0.56, 0.58, respectively), whereas they tend to moderate a higher degree of saliency of the negative adjective nai by resorting to prosodic focus to a lesser extent (0.41, 0.36, respectively). In a highly interactive social situation such as CC, on the other hand, the analysis of relative effectiveness among a number of intersecting constraints indicates that this particular structural constraint is not as significant as interaction-related constraints (FG-7, FG-8, FG-9, FG-10). 25. This claim that the use of pauses is specific to certain social situations for communicative purposes is also substantiated by other studies. For example, Silverman et al. (1992), who compared the use of pauses between read speech and spontaneous speech, found that the use of pauses is both quantitatively and qualitatively different. In spontaneous speech in particular, pauses tended to be longer and occurred more often, and the authors attributed this registerlinked difference to the speaker’s lexical access delay as well as the pragmatic marker for the listener to wait for unpredictable information on its way.

Variation in prosodic focus of nai 



Table 4.  Frequencies of Focal Prominence on -NAI and Morphosyntactic Structures  

Verbal Suffix “-Nai”

Negative Adjective “Nai” without Particles

Negative Adjective “Nai” with Particles

 

e.g. ika-na’i ‘don’t go’

e.g., oishiku (wa) na’i ‘not delicious’

e.g., oishiku wa na’i ‘not delicious’

PD



53% (92/173)



43%

(12/28)



38% (33/86)

NB



48% (48/101)



20%

(1/5)



24% (13/55)

CC



31% (128/412)



45%

(24/53)



34% (38/112)

Further evidence that substantiates these interpretations comes from the ratios of focal prominence in relation to the shifts in prosodic saliency derived from different morpho-syntactic structures of the negative nai. In the above-mentioned examples, kane ga/wa na’i ([I] have no money.) and oishiku wa na’i ([it] is not delicious.), in which nai is generally granted its own independent lexical accent, the postpositional particles such as ga and wa are often omitted with the propositional meaning being unchanged in spoken Japanese. This omission causes some change in the degree of perceptual saliency, in that nai prosodically becomes more like a part of the immediately preceding nouns (kane na’i) and adjectives (oishiku nai). In Table 4, the quantitative evidence shows that there is a continuum formulated in uses of focal prominence in accord with the degree of saliency of the negative nai as determined by those morpho-syntactic makeups as far as the information-oriented social situations (PD and NB) are concerned. As the degree of morpho-syntactic independence of the negative nai increases (from left to right), the ratios of prosodic focus decrease, particularly in PD (PD: 53% → 43% → 38%).26 While particle omission is very unlikely to occur in formal communicative settings like televised news programs (NB), which yielded a skewed result (NB, 20% [1/5]), the same directionality of change is observed in the ratios of focal prominence (NB: 48% → 20% → 24%). However, such a gradient of systematicity is not observed in the highly interaction-oriented social situation of conversation (CC: 31% → 45% → 34%). Notice also in Table 4 that in CC the ratio for focal prominence on the negative adjective nai without postpositional particles (middle column) is strikingly high (45% [24/53]) as compared with those in the other morpho-syntactic structures (31% for verbal suffix; 34% for negative adjective with particles). I argue that this ratio is likely the byproduct of the series of effects of various footings favoring focal prominence to a statistically significant extent (see FG-10 in Table 3). Among the total of 24 occur26. This three-way distinction of the morpho-syntactic status of the negative nai did not produce any statistically significant results by GoldVarb analysis. Consequently, the two-way distinction (i.e., verbal suffix and negative adjective) was adopted as the potential morpho-syntactic constraint on nai prominence because of its statistical significance (FG-1 in Table 3).

 Shoji Takano

rences of prosodically prominent tokens with postpositional particles omitted, 17 tokens (70.8 %) carry specific pragmatic meanings such as agreement seeking (6 prominent out of 6 tokens, 100%), supportive (4 prominent out of 4, 100%), self-denigrating (1 prominent out of 4, 25%), self-protective (2 prominent out of 4, 50%), face-threatening with humor (teasing) (3 prominent out of 3, 100%), and face-threatening (1 prominent out of 1, 100%). Furthermore, high frequencies of concurrence of Japanese particle ellipsis and the use of pragmatic force has been documented repeatedly in past variationist studies (e.g., Matsuda 1992, Takano 1998), which demonstrate that the ellipsis of Japanese postpositional particles is promoted by the pragmatic force of the utterance. These findings seem to suggest that, as far as CC is concerned, the remarkably high ratio of focal prominence on the negative in this particular morpho-syntactic structure is likely to be motivated by interactional factors rather than by the speaker’s control of perceptual saliency. Here again, the nature of the social situation figures systematically in speakers’ decision-making processes for the mappings of potential intersecting factors into sociolinguistic grammar.

6.3

Information structure and focal prominence

The results in FG-6 and 7 (Table 3) demonstrate that variability in Japanese focal prominence is not heavily subject to information structures of discourse (e.g., newness or significance of information conveyed by the variable in question), which is discordant with the aforementioned “highlighting-based” universalist view of focal prominence based on “non-interactive” read speech in Western languages. Our statistically insignificant results even from identical “non-interactive” social situations like NB (i.e., read speech) and highly information-oriented PD demonstrate that neither information-linked factor group (FG-6 or 7) exerts critical effects on variability in Japanese prosodic focus. In fact, this outcome coincides with Sugitou’s (1985, 1986) argument for language specificity in focal prominence phenomena in Japanese. Both the finding for FG-2 (i.e., the farther nai is located from the end of the IP, the more likely it is to obtain prominence) and these results (FG-6 and 7) substantiate Sugitou’s finding that linguistic elements in the clause-/phrase-initial position tend to receive prosodic prominence, regardless of the information status they may represent in the flow of discourse. In addition, the statistically significant results of FG-3 discussed above further justify the superiority of this “structure-based” account over the “information-based” account, in that the internal prosodic structures of the IP (i.e., the presence/absence of preceding lexical accents) play a definitive role in variable uses of Japanese focal prominence. The statistically significant results (FG-7) for CC also show that the “highlightingbased” accounts of prosodic focus cannot be substantiated because of such probability weights as 0.50 (i.e., no effect) for “brand-new” information and a weak negative effect (0.47) for “contrastive” information. The statistical significance of this particular factor group stems primarily from the strong effect of “old” information status disfavoring

Variation in prosodic focus of nai 



nai prominence (0.35). Although I cannot offer any empirical interpretation regarding this particular finding at this point, I suspect that high degrees of common knowledge shared between the participants (i.e., close friends) might be a factor inhibiting the use of prosodic focus for old, already-familiar information. Note also that the results indicate a weak positive effect of “inferred information” (0.56) on nai prominence. In fact, this is closely linked to consistent positive effects of “preface” on -nai prominence (FG-8). The results of this interactive constraint, which will be discussed in detail in the next section, show that the use of discourse markers as “prefaces” to upcoming negatives consistently favors the use of focal prominence to a statistically significant extent. These interactive devices clearly help the listener to “infer” the information conveyed by the upcoming negative in the flow of discourse. Thus, once “interactive” social situations (i.e., casual conversations) are accommodated in the scope of research, this new dimension of information status comes in as another possible constraint to be considered.

6.4

Interactive parameters

As pointed out earlier, past studies on prosodic focus, which have been done predominantly with laboratory phonology, have critically neglected the impact of interactive dimensions on the phenomena because of the use of non-interactive registers as analytical data. A major thrust of the present study is to shed light on the dynamic systems of prosody that are generally highly susceptible to the interpersonal dimensions of everyday language use in various social settings. FG-10 in Table 3 (see Section 5.2 for discourse examples) shows that the types of footing of the negative nai have been found to exert robust effects on the phenomena in CC, whereas the effects are found not to be statistically significant in PD. This latter outcome is due to the mapping of independent variables sensitive to the communicative goals of this particular social situation, in which the participants choose to resort to this prosodic means to consistently maintain their confrontational stance as well as the communicative efficiency over moment-to-moment changes in interactive meanings and their impact on the listener. The two social situations are very contrastive with respect to the overall distribution of the nai tokens. While about the half of the negatives are used to create informative (neutral) type of footing both in CC and PD (54.9% [317/577]; 45.3% [130/287], respectively), face-threatening uses of nai constitute almost half of the tokens in PD (45.3% [130/287]). In contrast, only 5.7% (33/577) of the tokens in CC are used to express direct disagreement with co-participants in accordance with the SAP – i.e., the signs of disagreement are minimized for the sake of the universal preference for agreement in conversations.27 The remaining tokens of CC display a much wider range of distribution across various types of footings, as compared with those of PD that are 27. This difference between the social situations is statistically highly significant (p < .001).

 Shoji Takano

mainly the type relevant to the interactional purposes such as protecting one’s own stance (self-protection: 8.4% [24/287]). Thus, contrary to the “myth” of interpersonal harmony in Japanese culture, Japanese people openly disagree with one another as long as the social situations (such as PD) and related sociocultural expectations for communicative appropriateness allow them to do so (Honda 2002, Ishida 1984, Jones 1990). The results show that prosody can be one of the parameters useful for systematically expressing or reinforcing the speaker’s confrontational stance. Future work on how Japanese people disagree with one another should take into serious account the communicative properties of different social situations and their sociocultural functions in everyday language use. The statistically significant results of CC (FG-10) reveal that tangible principles underlie the speakers’ apparently chaotic behaviors in exploiting focal prominence – i.e., the speaker’s decision-making processes are subject to the interactive meanings of negation at every moment of its use. The participants in CC whose interactional goal is to maintain or promote interpersonal rapport and solidarity appear to resort to focal prominence for paralinguistic messages of positive politeness, as shown in those highprobability tokens for supportive interactive meaning (0.79) (see Excerpt [3] above), face threatening with humor (teasing) (0.79) (Excerpt [5] below), and agreement seeking (0.67) (Excerpt [6] below).28 In Excerpt 5, speakers C3a and C3b are college classmates who are wondering whether they can bring parents to an event at which students are supposed to pick their attire to wear for an upcoming graduation ceremony. Speaker C3b (Masako) feels uneasy about making her own decision. She wonders about the possibility of bringing her mother to help with her decision making, but she thinks that a college student being accompanied by his or her parent would look strange to other people. Speaker C3a playfully pursues C3b’s thoughts. (5) Face threatening with humor (teasing) Speaker C3a: gakusei no furi sureba <@ iin da ne.@> [@@@@@] ‘It would work if she pretends to be a student, wouldn’t it?’ Speaker C3b: [@@@@@] Speaker C3a: <@ sore ga ichiban da.@> ‘That’s the best (way).’ Speaker C3b: @@@[@@@]

28. The relatively high probability for the “self-protection” footing (0.57 [38%, 17/45]) in Table 3 is largely due to the speaker’s emotional responses (often along with humor and jokes) to the interlocutor’s challenge regarding personal topics (e.g., a former boyfriend, makeup, job hunting, etc.).

Variation in prosodic focus of nai 



Speaker C3a: [@@@] futsuuni jibun mo <@ kiruyoona kanji de.@> @@@ <@ doomo Masako tokatte.@> @@@@ ‘Just like a regular student, she also wears (an attire), and like, she says “Hello, Masako.”’ (A few more utterances of a similar kind by Speaker C3a are omitted here. After interacting with C3a’s series of jokes with agreeable laughter, Speaker C3b finally responds with humor, clearly negating the possibilities suggested by C3a with the prominent nai.) → Speaker C3b: usso:: anna gakusei ina:i ssho hansoku desho. ‘No way. There is no student like that. (That’s) against the rules.’ In Excerpt (6), Speakers C4a and C4b, female friends in their late 60s to early 70s, are talking about uncomfortable experiences they have had at the hospital. C4a tells a short story about the difficulties she experienced in receiving an injection in a blood vessel. (6) Agreement seeking Speaker C4a: iya: ano: tsukisoi-san ga ne, moo yamete kudasai tte yuu hodo ne, ‘Well, so, as much as my attendant said, “please don’t do it anymore,”’ Speaker C4b: un. ‘Yeah.’ Speaker C4a: ano: sasaretan da yo [ (0.5) hari.] ‘I was stung with needles (so many times).’ Speaker C4b: [ maa: ] ‘Really.’ → Speaker C4a: tsumetai tokoro wa kekkan denai desho. ‘The blood vessel doesn’t stand out under the cold part (of the skin), does it?’ Speaker C4b: aa tsumetai kara:. ‘Yeah, because (the skin) is cold.’ The observed patterns of variability also provide empirical evidence of discordance with another cultural stereotype, which is that Japanese language use is heavily oriented toward negative politeness and deference to others. As can be seen in this example, positive polite norms of interaction are equally stressed and heavily exploited as the core elements for building interpersonal relations in this kind of social situation, in which in-group rapport and solidarity are highly valued. This is also one of the reasons why we have found a relatively higher probability weight (0.40) for the prominent nai tokens in CC (Table 2) relative to the other social situations (see Note 20 in Section

 Shoji Takano

6.1). Future research on Japanese language usage should include work on this relatively neglected aspect of Japanese linguistic behavior. The shift of linguistic elements to the post-predicate position also favors focal prominence in CC to a statistically significant extent (FG-9 in Table 3) – i.e., postposition strongly favors focal prominence (0.65), whereas canonical word order has almost no effect (0.49) (see Excerpt [7] for an illustration).

(7) Postposing (Speakers C2a and C2b, college classmates, are talking about C2a’s past boyfriend having head lice.) Speaker C2a: mita koto nai? ‘You haven’t seen (lice)?’ → Speaker C2b: shiranai mita koto nai yo sonna no. ‘I don’t know. I’ve never seen such a thing.’ Speaker C2b’s response would be sonna no mita koto nai yo if it followed canonical word order in Japanese, but instead the noun phrase was dislocated to the post-verbal position. A striking contrast with PD and NB lies in its frequency of occurrence. While postpositions co-occurred with negation 8.5% of the time in CC (49/577), they cooccurred only 1.7% of the time in PD (5/287) and never co-occurred in NB. This difference is statistically highly significant (p <.001). Ono and Suzuki (1992) argue that, in contrast to postpositions in which a break in intonation contour or a pause intervenes between the predicate and the postposed element, postpositions expressed throughout a single intonation contour without any break evoke emotive overtones, enhance discourse cohesiveness with the preceding context, and strengthen the speaker’s stance toward the proposition, referent, or topic being discussed in the preceding context.29 Fujii (1995), though not referring to intonational characteristics, also makes a similar generalization that postpositions function to highlight the propositional content of the preceding clause. The patterns of variability in focal prominence detected here precisely represent these characteristics – i.e., postposed elements contribute to highlighting the locus of negation in the preceding clause by prosodic means (0.65). It should also be noted that postpositions are linked to the positional factor discussed in FG-2: the farther the negative -nai is located from the end of the intonation phrase, the more likely it is to receive focal prominence. It is evident that syntactic dislocation creates this favorable prosodic environment for focal prominence and that the speaker seems capable of manipulating both syntactic and prosodic means of focus in a synergistic fashion. This is one of the examples of meaningful collaboration between syntax and prosody found in the present study. The last interactive parameter to be discussed concerns another pattern of syntaxprosody collaboration – i.e., focal prominence on the negative nai is promoted by the 29. Postposed elements analyzed in the present study belong to this latter type because the intonation phrase as the domain of analysis is based on a single intonation contour without a major prosodic break.

Variation in prosodic focus of nai 



syntactic markers of preface (FG-8 in Table 3). Excerpts (8) and (9) illustrate the two types of markers under analysis. In (8), Speaker C4a, a woman in her late 60s, is talking to her friend (early 70s) about a mutual friend with whom they have not been in contact for a long time. Speaker C4a tells a short narrative about what was talked about over the phone when she finally decided to give her a call. (8) Preface as a hint (in boldface) Speaker C4a: ara soo datta no shiranaide tte soo itte, → mimai ni ikitai n dakedo ttara, iya ima chotto aitakunai tte yuu n da yone,...

‘”ah, is that so? (I) didn’t know that,” (I) said, then “(I) want to go and see you,” (I) said. But (she) said, “Nah, (I) kind of do not want to see (people).” So..’

The utterance-initial iya ([negatively implied] ‘well’) hints to the listener the negative aitakunai that is on its way. In (9), Speaker C3b (Masako) has a strong desire to live in a rural town like the one Speaker C3a is from. Speaker C3a is strongly against the idea and tries to persuade Masako to give it up. (9) Preface as an upgrader (in boldface). Speaker C3a: iya:: Masako zettai muri da wa. ‘No, it’s totally impossible for you (to live there.)’ Speaker C3b: @@@ → Speaker C3a: datte honto nanmo nai n da yo asobu mono ga. ‘Because there is really nothing to do for fun around (there).’ A connective datte (‘because’) upgrades the upcoming negative nanmo nai (‘nothing’) as a preface. The results show that the use of such preface markers exerts consistent effects favoring focal prominence on nai in all of the social situations, although the relative effectiveness is found not to be statistically significant in PD and NB. Past studies of syntactic operations in Japanese disagreement (e.g., Honda 2002, Jones 1990, Mori 1999, Watanabe 1993, Yamada 1992) commonly demonstrate that Japanese ways of delivering direct disagreement are characteristic of what precedes the locus of disagreement, taking advantage of various interactive devices (e.g., connectives, hedges, intensifiers, discourse framing) as the markers of opposition moves or as cues to contextualize such moves (Gumperz 1982). While the present study concerns not only direct disagreement (i.e., the face-threatening footing of nai) but also other types of footings as expressed by negation, our findings are parallel to those of the prior studies in that the preceding syntactic elements interact with the upcoming locus of negation, “feeding” its prosodic focus in order to convey certain paralinguistic messages to the listener.

 Shoji Takano

The degree of this interactive linkage between the preface and prosodic focus on negation, on the other hand, seems to be differentiated between the two types – i.e., hints (0.68 in CC; 0.79 in PD, 0.63 in NB) and upgraders (0.59 in CC; 0.55 in PD; no token in NB). The upgrading prefaces (e.g., datte ‘because,’ dakara/desukara ‘so,’ sorede ‘then,’ shitagatte ‘therefore,’ etc.) play the pragmatic role of conjoining the immediately preceding clause (or phrase), which sets up a context relevant to the upcoming negation, with the following locus of negation (as in Excerpt [9]) in a wider stretch of discourse. The hint prefaces (demo ‘but,’ dakedo ‘though,’ iya ‘no,’ etc.), on the other hand, seem to involve more local operations within a single clause, which contributes directly to the shift from the immediately preceding clause (or phrase) to the clause in which they are embedded in terms of propositional meanings (see Excerpt [8]). Although these interpretations are all still speculative at this point, it seems that these “localized” hint prefaces are more heavily integrated into the interactive operations of Japanese negation than previously acknowledged and reinforce the speaker’s upcoming message, as is shown by the observed patterns of systematicity in prosodic focus.30

7. Summary and conclusion Focal prominence in natural speech is such a variable phenomenon that capturing its underlying principles is extremely difficult. Adopting a variationist approach to linguistic variation, the present study has attempted to detect the “orderly heterogeneity” (Weinreich et al. 1968) involved in the phenomenon, taking into account a variety of constraints from a wide range of domains: language-specific prosodic principles, informational parameters in a discourse, interactive parameters manipulated by the speaker at every moment of talk-in-interaction, and the sociolinguistic properties of social situations in the speech community. Based on the sociolinguistic grammar that has been described so far, we can make good predictions about whether the negative nai will receive or not receive prosodic focus in a given utterance in a given social situation, and we can also account for why observed variability is obtained in particular utterances. The present analysis of the relative effects of the wide variety of intersecting constraints has empirically shown that in the case of Japanese focal prominence, languagespecific, “structure-based” accounts should contribute to the formation of a more legitimate theory than universalistic “highlighting-based” accounts. Focal prominence 30. The higher probability for hint prefaces than for upgrade prefaces coincides with the results (0.77 for hints, 0.51 for upgraders) of my preliminary work (Takano 2002), which is based on a much smaller-scale sample of casual friendly conversations (a total of 264 tokens). In this preliminary study, the clauses preceding the clause with the negative nai were also entered as a separate preface category into GoldVarb analyses. The results, however, were that the entire factor group of “preface” was statistically not significant, which urged me to focus in the present study only on those conjunctive preface markers that represented potential variables.



Variation in prosodic focus of nai 

phenomena in any of the social situations we have investigated are consistently subject to the unmarked application of prosodic principles specific to the language, regardless of the informational status that a given element in focus has in a discourse, and this aspect of language specificity has been shown most eloquently in highly “informationoriented” social situations such as NB and PD. Our results demonstrate that it is the internal prosodic environment of the intonation phrase in which the focused element is embedded that heavily governs observed variability (FG-2, 3) rather than informational parameters in a given discourse (FG-6, 7), which a number of prior studies of Western languages have postulated as the primary constraint on prosodic focus. Furthermore, the results also show that Japanese speakers in highly information-oriented social situations such as PD and NB seem to know how to resort to this prosodic means for conveying information efficiently to the hearer, based on the communicative requirements of the social situation. In so doing, they estimate the degree of perceptual saliency of the negatives in terms of their morpho-syntactic status (rather than the informational significance of the negatives in the flow of discourse) and exploit prosodic focus based on that estimate (FG-1). The present study has also attempted to fill a critical gap in prior research, which has neglected the highly interactive aspects of prosody in natural face-to-face exchanges. My goal was to obtain supporting evidence for a general sociolinguistic principle that the social situation systematically influences language use, not merely by influencing word choice or syntactic positioning but also by influencing even the prosodic realization of what is said. The results obtained from a highly interaction-oriented social situation such as CC demonstrate that variability in focal prominence on the negatives is also constrained systematically by a variety of interactive meanings negotiated between the co-participants at every moment of talk-in-interaction (FG-10). Moreover, the patterns of systematicity vary in response to the communicative properties and interactional goals of particular social situations. Japanese speakers in PD are found to take advantage of prosodic strategies to reinforce confrontational stances (FG-4) or to appeal to both the opponent and a larger audience (FG-5), according to the communicative requirements of the activity. Our results have empirically demonstrated that the speaker’s mapping of meaningful constraints into sociolinguistic grammar is also closely linked to the sociolinguistic properties of social situations. In other words, the communicative properties of social situations exert decisive effects on the sociolinguistic grammar Japanese speakers rely on in exploiting prosodic focus in particular social situations. In highly interaction-oriented social situations such as CC, the majority of powerful constraints that “promote” prosodic focus stem from interactive parameters as well as internal structures of the intonation phrase (i.e., FG-10, 8, 9, 3, 2 in descending order of effectiveness), whereas influential constraints that “inhibit” focal prominence are derived primarily from the structural principles of prosody specific to the Japanese language (i.e., FG-3, 2). As a notable dimension of the sociolinguistic grammar I have described, the results revealed the significant interplay between syntax and prosody – the participants

 Shoji Takano

in highly interactive social situations such as CC manipulate the syntactic means to feed or compensate for prosodic distinctions (FG-8, 9). In highly informative social situations such as PD, on the other hand, many of the interactive parameters become less influential in promoting focal prominence. Instead, register-specific dimensions of prosody begin to prevail in favor of focal prominence in response to the communicative requirements of the activity (FG-5, 4, 1), while the disfavoring effects of structural principles of prosody remain powerful (FG-2). In the non-interactive social situation studied (NB), both the positive and negative effects of those structural principles of Japanese prosody are more conspicuous than in the interactive social situations. These findings provide us with some new insights into which approach is truly valid for investigating language use that is inherently highly variable. Approaches to interactional dimensions alone, on which a great majority of prior studies of language use have focused as the driving force of variable linguistic behaviors, may not necessarily be successful in explicating the entire picture of such complex phenomena. The present study demonstrates that the structural principles specific to Japanese prosody play critical roles in exerting robust effects “disfavoring” prosodic focus, and it suggests the importance of paying much closer attention to rather mechanical, surface-level conditioning derived from the language-specific makeup of prosodic structures. Moreover, the mapping of constraints that governs systematic variability in natural speech has also been found to depend upon the interactive-informative orientations appropriate to given social situations. These observations signify the necessity of a “multi-stylistic (or -register)” approach to the study of language use beyond a single social situation, with casual conversations as the fundamental site. This research project was initially motivated by the perceived need to supplement the dichotomy of styles acknowledged for Japanese surface morpho-syntactic characteristics (i.e., direct-casual or distal-polite). Systematic analysis of paralinguistic aspects of language use–prosodic style shift–suggests that Japanese style shifting should be accounted for as a multi-layered mechanism involving prosodic parameters as one component. While surface morpho-syntactic styles may be overt ways for speakers to conform to the sociocultural norms for interaction in particular social settings or participant frameworks (Befu 1980, Bell 1984), prosodic styles provide the speaker with a rich repertoire of rather covert, non-transparent, strategic choices for representing his or her actual stance, thoughts, and feelings that dynamically respond to every moment of talk-in-interaction in a given social situation. It is hoped that exploring this neglected area of sociolinguistic variation further will eventually lead to a more legitimate picture of style in Japanese language use.

Variation in prosodic focus of nai 



References Azuma, J. 1992a. “F0 no shisatsu to chookaku inshoo ni motozuku nihongo intoneeshon hyooki no kanoosei ni tsuite (Observation of F0 and possible descriptions of Japanese intonation based on auditory perception).” In Nihongo Akusento to Intoneeshon, M. Sugitou (ed.), 89–97. Azuma, J. 1992b. “Nihongo no inritsu taikei (Prosodic systems of Japanese).” Proceedings of International Symposium on Japanese Prosody. 53–61 Beckman, M., and Pierrehumbert, J. 1986. “Intonation structure in Japanese and English.” Phonology Yearbook 3: 255–309. Befu, H. 1980. “A critique of the group model of Japanese society.” Social Analysis 5/6: 29–43. Biber, D. 1988. Variation Across Speech and Writing. Cambridge: University Press. Biber, D. 1995. Dimensions of Register Variation. Cambridge: University Press. Bell, A. 1984. “Language style as audience design.” Language in Society 13: 145–204. Blom, J. P., and Gumperz, J. J. 1972. “Social meaning in linguistic structure.” In Directions in Sociolinguistics, J. J. Gumperz and D. Hymes (eds.). New York: Holt, Rinehart and Wilson. Brown, G. 1983. “Prosodic structure and the given/new distinction.” In A. Cutler and D. R. Ladd (eds.), Prosody: Models and Measurements. Heidelberg: Springer. Brown, P., and Levinson, S. 1987. Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Cedergren, H. J., and Sankoff, D. 1974. “Variable rules: Performance as a statistical reflection of competence.“ Language 50: 333–55. Chafe, Wallace. 1994. Discourse, Consciousness, and Time. Chicago: University of Chicago Press. Couper-Kuhlen, E., and Selting, M. (eds.). 1996. Prosody in Conversation. Cambridge: University Press. Cruttenden, Alan. 1986. Intonation. Cambridge: University Press. Cutler, A., Dahan, D., and van Donselaar, W. 1997. “Prosody in the comprehension of spoken language: A literature review.” Language and Speech 40(2): 141–201. Ferguson, C. A. 1964. Diglossia. In Language in Culture and Society, D. Hymes (ed.). New York: Harper and Row. Fujii, Y. 1995. “Nihongo no gojun no gyakuten ni tsutite: Kaiwa no naka no joohoo no nagare o chuushin ni (On the reversal of word order in Japanese: Focusing on information flow in conversation).” In Nichi-eigo no Uhoo Idoo Koobun: Sono Koozoo to Kinoo (Right Dislocation in Japanese and English: Its Structure and Functions), K. Takami (ed.), 167–198. Tokyo: Hituji Syobo. Fujisaki, H., Hirose, K., Takahashi, N., and Yokoo, M. 1984. “Renzoku onsei chuu no akusento seibun no jitsugen (Realization of accent components in connected speech).” Nihon Onkyoo Gakkai Onsei Kenkyuu-kai Shiryoo 84–36: 279–286. Goffman, E. 1981. Forms of Talk. Oxford: Blackwell. Grosz, B., and Sidner, C. 1986. “Attention, intention, and the structure of discourse.” Journal of Computational Linguistics 12: 175–204. Gumperz, J. J. 1982. Discourse Strategies. Cambridge: University Press. Hattori, S. 1933. “Akusento to hoogen (Accents and dialects).” Kokugo Kagaku Kooza VII. Meiji Shoin. Hinds, J. 1982. Ellipsis in Japanese. Edmonton, Alberta: Linguistic Research.

 Shoji Takano Hirschberg, J. 1990. “Accent and discourse context: Assigning pitch accent in synthetic speech: The given/new distinction and deaccentability.” Proceedings of the 8th National Conference on Artificial Intelligence. Cambridge, MA: MIT Press. 952–57. Hirst, D. and Di Cristo, A. (eds.). 1998. Intonation Systems: A Survey of Twenty Languages. Dordrecht: Kluwer. Honda, A. 2002. “Conflict management in Japanese public affairs talk shows.” Journal of Pragmatics 34: 573–608. Ishida, T. 1984. “Conflict and its accommodation: Omote-ura and uchi-soto relations.” In Conflict in Japan, E. S. Krauss, T. P. Rohlen, and P. G. Steinhoff (eds.), 16–38. Honolulu: University of Hawaii Press. Jones, K. 1990. Conflict in Japanese Conversation. Ph.D. dissertation. University of Michigan. Jones, K. 1995. “Masked negotiation in a Japanese work setting.” In The Discourse of Negotiation, A. Firth (ed.), 141–158. Pergamon. Kawakami, S. 1957. “Tokyo-go no takuritsu kyoochoo no onchoo (Tones of emphatic prominence in Tokyo Japanese).” Kokugo Kenkyuu 6: 21–31. Kokugakuin University. Kawakami, S. 1965. “Kyoochoo ni yoru gokei hakai to sono tekiyoo hani (Intensity-emphasis and its range of application).” Onsei Kenkyuu 6: 61–76. Kindaichi, H. 1951. “Kotoba no senritsu (Melodies of Japanese).” In Nihongo On-in no Kenkyuu. Tokyo-do. Koori, S. 1989a. “Fookasu jitsugen ni okeru onsei no tsuyosa, jizoku jikan, Fo no yakuwari (Intensity, duration, the role of Fo in focus).” Onsei Gengo III. 29–38. Koori, S. 1989b. “Kyoochoo to intoneeshon (Emphasis and intonation).” In Kooza Nihongo to Nihongo Kyooiku Vol. II, M. Sugitou (ed.), 316–342. Tokyo: Meiji Shoin. Koori, S. 1997a. “Nihongo no intoneeshon: Kata to kinoo (Japanese intonation: Types and functions).” In Akusento, Intoneeshon, Rizumu to Poozu, T. Kunihiro et al. (eds.), 169–202. Tokyo: Sanseido. Koori, S. 1997b. “’Tooji no murayama-shushoo’” no futatsu no imi to futatsu no yomi (Two meanings and two readings of “the then Prime Minister Murayama”).” In Bunpoo to Onsei, Onsei Bunpoo Kenkyuu-kai (ed.), 123–146. Tokyo: Kuroshio. Krauss, E.S., Rohlen, T.P., and Steinhoff, P.G. (eds.). 1984. Conflict in Japan. Honolulu: University of Hawaii Press. Kubozono, H. 1989. “Syntactic and rhythmic effects on downstep in Japanese.” Phonology 6(1): 39–67. Ladd, R. D. 1996. Intonational Phonology. Cambridge: University Press. Levinson, S. C. 1983. Pragmatics. Cambridge: University Press. Matsuda, K. 1992. “Accusative case marker -o deletion in Tokyo Japanese.” The Annual Meeting of. the Chicago Linguistic Society 28: 356–366. Maekawa, K. 1994. “Is there ‘dephrasing’ of the accentual phrase in Japanese?” Ohio State University Working Papers in Linguistics from the Linguistics Laboratory 44: 146–65. Mori, J. 1999. Negotiating Agreement and Disagreement in Japanese: Connective Expressions and Turn Construction. Philadelphia: John Benjamins. Nihongo Kyouiku Gakkai (ed.). 1993. Nihongo Kyouiku Jiten (An Encyclopedia of the Teaching of Japanese). Tokyo: Taishuukan. Niyekawa, A. M. 1984. “Analysis of conflict in a television home drama.” In Conflict in Japan, E. S. Krauss, T. P. Rohlen and P. G. Steinhoff (eds.), 61–84. Honolulu: University of Hawaii Press.



Variation in prosodic focus of nai  Nooteboom, S. G., and Kruyt, J. G. 1987. „Accents, focus distribution, and the perceived distribution of given and new information: An experiment.“ Journal of the Acoustic Society of America 82: 1512–24. Oishi, H. 1959. “Purominensu ni tsuite (On prominence).” In Kotoba no Kenkyuu. Tokyo: National Language Research Institute. 87–102. Ono, T., and Suzuki, R. 1992. “Word order variability in Japanese conversation: Motivations and grammaticization.” Text, 12(3): 429–445. O’Shaughnessy, D., and Allen, J. 1983. “Linguistic modality effects on fundamental frequency.” Journal of the Acoustical Society of America 74: 1155–71. Pearson, B. 1988. “Power and politeness in conversation: Encoding of face-threatening acts at church business meetings.” Anthropological Linguistics 30(1): 68–9 Pearson, B. 1989. “’Role-ing out control’ at church business meetings: Directing and disagreeing.” Language Sciences 11(3): 289–304.3. Pierrehumbert, J. and Beckman, M. 1988. Japanese Tone Structure. Cambridge: The MIT Press. Pomerantz, A. 1984. “Agreeing and disagreeing with assessments: Some features of preferred/ dispreferred turn shapes.” In Structures of Social Action: Studies in Conversation Analysis, J.M. Atkinson and J. Heritage (eds.), 152–163. Cambridge: University Press. Preston, D. 1989. Sociolinguistics and Second Language Acquisition. New York: Basil Blackwell. Preston, D. 1991. “Sorting out the variables in sociolinguistic theory.” American Speech 66: 33–56. Prince, E. 1981. “Toward a taxonomy of given-new information.” In Radical Pragmatics, P. Cole (ed.), 223–256. New York: Academic Press. Prince, E. 1992. “The ZPG letter: Subjects, definiteness, and information status.” In Discourse Description: Diverse Linguistic Analyses of a Fund-Raising Text, W. C. Mann and S. A. Thompson (eds.), 295–325. Philadelphia: John Benjamins. Rand, D., and Sankoff, D. 1990. Goldvarb Version 2: A Variable Rule Application for the Macintosh. Montreal: Universite de Montreal. Sacks, H., Schegloff, E. A., and Jefferson, G. 1974. “A simplest systematics for the organization of turn-taking for conversation.” Language 5(4): 696–735. Sankoff, D. 1985. “Statistics in linguistics.” In Encyclopedia of Statistical Sciences, 74–81. New York: Wiley. Sankoff, D. 1986. “Variable rules.” In Sociolinguistics: An International Handbook of the Science of Language and Society, U. Ammon et al. (eds.). Berlin and New York: de Gruyter. Sankoff, D. 1988. “Sociolinguistics and syntactic variation.” In Linguistics: The Cambridge Survey, vol. 4: The Socio-cultural Context, F. J. Newmeyer (ed.), 140–161. Cambridge: University Press. Sankoff, D., and Labov, W. 1979. “On the uses of variable rules.” Language in Society 8: 189–222. Schegloff, E. A. 1980. “Preliminaries to preliminaries: ‘Can I ask you a question?’”, Sociological Inquiry 50(3/4): 104–152. Schegloff, E., Jefferson, G., and Sacks, H. 1977. “Preference for self-correction in the organization of repair in conversation.” Language 53: 361–82. Silverman, K.E.A., Blaauw, E., Spitz, J., and Pitrelli, J. 1992. “A prosodic comparison of spontaneous speech and read speech.“ The Proceedings of International Conference of Spoken Language Processing 92: 1299–1302. Sugitou, M. 1982. “Osaka hoogen ni okeru kyoochoo no onkyoo-teki tokuchoo (Acoustic characteristics of emphasis in Osaka dialect).” Shooin Kokubungaku 19: 55–63.

 Shoji Takano Sugitou, M. 1985. “Nihongo to eigo ni okeru ‘shin-joohoo’ no hatsuwa no onkyoo-teki tokuchoo (The acoustic characteristics of new information in Japanese and English).” Onsei no Kenkyuu 21: 335–348. Sugitou, M. 1986. “Akusento no toraekata (Ways of understanding accents).” In Oyoo Gengogaku Kooza Vol. II, S. Hayashi (ed.), 76–96. Tokyo: Meiji Shoin. Takano, S. 1997. The Myth of a Homogeneous Speech Community: The Speech of Japanese Women in Non-traditional Gender Roles. Ph.D. Dissertation, The University of Arizona. Takano, S. 1998. “A quantitative study of gender differences in the ellipsis of the Japanese postpositional particles ‘-wa’ and –’ga’: Gender composition as a constraint on variability.” Language Variation and Change, 10(3): 289–323. Takano, S. 2001. “Nihonjin no gengo koushou ni okeru tairitsu iken kyoumei no inritsu modaritii: Rejisutaa kenkyuu no shiten kara (Exploring prosodic modality in Japanese disagreement from a perspective of register).” Hokusei Review 38: 13–31. Takano, S. 2002. “A variationist study of prosodic focus in naturally occurring interactions: The case of the negative ‘-nai’ in Hokkaido Japanese.” Onsei Kenkyuu (Journal of the Phonetic Society of Japan) 6(3): 25–47. Takano, S. 2005. “Re-examining linguistic power: Strategic uses of directives by professional Japanese women in positions of authority and leadership.” Journal of Pragmatics 37(5): 633–666. Terken, J. 1997. “Variation of accent prominence within the phrase: Models and spontaneous speech data.” In Computing Prosody, Y. Sagisaka, N. Campbell and N. Higuchi (eds.), 95–116. New York: Springer-Verlag. Umeda, N. 1982. “’FO declination’ is situation dependent.” Journal of Phonetics 10: 279–290. Venditti, J. 1995. “Japanese ToBI Labelling Guidelines.” Unpublished manuscript, Ohio State University. (Downloadable from: ling.ohio-state.edu/phonetics/J_ToBI/jtobi_homepage. html). Venditti, J. 2005. “The J_ToBI model of Japanese intonation.” In Prosodic Typology: The Phonology of Intonation and Phrasing, S. A. Jun (ed.), 172–200. Oxford: University Press. Wada, M. 1975. “Akusento, intoneeshon, purominensu (Accents, intonation, prominence).” In Nihongo to Nihongo Kyooiku – Hatsuon–Hyoogen Hen, National Language Research Institute, 53–86. Watanabe, S. 1993. “Cultural differences in framing: American and Japanese group discussions.” In Framing in Discourse, D. Tannen (ed.), 176–209. Oxford: University Press. Weinreich, U., Labov, W., and Herzog, M. I. 1968. “Empirical foundations for a theory of language change.“ In Directions for Historical Linguistics, W. P. Lehmann and Y. Malkiel (eds.), 95–195. Austin, Texas: University of Texas Press. Yaeger-Dror, M. 1985. “Intonational prominence on negatives in English.” Language and Speech 28: 197–230. Yaeger-Dror, M. 1996. “Register as a variable in prosodic analysis: The case of the English negative.” Speech Communication 19: 39–60. Yaeger-Dror, M. 1997. “Contraction of negatives as evidence of variance in register-specific interactive roles.” Language Variation and Change 9(1): 1–36. Yaeger-Dror, M. 2001. “Primitives of a possible system for ‘style’ and ‘register.’” In Style and Sociolinguistic Variation, P. Eckert and J. Rickford (eds.), 170–185. Cambridge: University Press. Yaeger-Dror, M. 2002a. “Introduction.” Journal of Pragmatics 34: 1333–43.



Variation in prosodic focus of nai  Yaeger-Dror, M. 2002b. “Register and prosodic variation, a cross language comparison.” Journal of Pragmatics 34: 1495–536. Yamada, H. 1992. American and Japanese Business Discourse: A Comparison of Interactional Styles. Norwood, NJ: Ablex. Young, R., and Bayley, R. 1996. “VARBRUL analysis for second language acquisition research.” In Second Language Acquisition and Linguistic Variation, R. Bayley and D.R. Preston (eds.), 253–306. Philadelphia: John Benjamins.

Name index A Abe, T.  230 Allen, J.  290 Anderson, S.  254 Arashiyama, K.  119, 120 Ariel, M.  256 Arundale, R.  40 Atkins, B. K.  206 Auer, P.  244 Azuma, J.  286, 296, 303, 310 B Bachnik, J. M.  1 Bakhtin, M.  11, 17, 92, 97, 98, 118 Bardovi-Harlig, K.  17 Bateson, G.  60n Bayley, R.  300n Beckman, M.  289, 302, 310, 311 Befu, H.  287, 322 Bell, A.  10, 213, 216, 322 Bernstein, G. L.  186n, 196n, 198 Besnier, N.  253 Biber, D.  287, 291 Blaauw, E.  287, 290, 291 Blom, J.-P.  9, 230, 237, 240, 285 Bodine, A.  185 Borker, R. A.  206 Bourdieu, P.  205 Boxer, D.  48 Brown, G.  46n, 286, 290 Brown, P.  40, 41, 49, 54, 57, 61, 63, 82, 186n, 193, 292 Brown, R.  256 Bucholtz, M.  10 C Carnes, Y.  40, 41 Carroll, T.  229, 246 Cedergren, H. J.  287n, 300 Chafe, W.  252, 253, 296n Charles, R.  244 Cheshire, J.  244 Comrie, B.  13

Cook, H. M.  2, 3, 9, 15, 16, 26, 34, 35, 39, 41, 42, 44, 46, 47, 71, 135, 138n, 192n, 251 Couper-Kuhlen, E.  286, 290 Coupland, N.  10, 15, 213, 214 Crosby, F.  185 Cruttenden, A.  296n Cumming, S.  189, 210 Cutler, A.  286, 288, 290 D Dahan, D.  286, 288, 290 Di Cristo, A.  288, 290 Diessel, H.  255 Du Bois, J. W.  189, 210 Duranti, A.  1, 35, 87, 256 E Eckert, P.  9, 206, 214 Edelsky, C.  74 Eelen, G.  40 Ehlich, K.  41 Endo, O.  117, 136 Erman, B.  171n F Falsgraf, C.  34 Ferguson, C. A.  285 Fishman, P. M.  79 Ford, M.  256 Fraser, B.  40, 78 Fujii, Y.  304, 318 Fujisaki, H.  290, 302 Fujitani, N.  93 Fukao, M.  96, 110 Fukushima, E.  4, 141, 142, 143, 148, 149, 152n, 157, 161, 171, 199 G Gal, S.  214, 245, 247 Galbraith, J. K.  186n Gardner-Chloros, P.  244 Geyer, N.  3, 39, 40, 64n Giacalone Ramat, A.  237 Giles, H.  10 Gilman, A.  256

Givón, T.  270 Glover, K. D.  256 Goffman, E.  60n, 74, 84n, 287, 304 Goodwin, C.  1 Graddol, D.  207 Grosz, B.  286 Gu, Y.  40 Gumperz, J.  9, 10, 60n, 84n, 230, 231, 237, 240, 244, 246, 252, 285, 290, 319 Gundel, J. K.  256 H Haga, Y.  93, 94, 96 Haiman, J.  181 Halliday, M. A. K.  269, 276 Hanks, W. F.  252, 256, 257, 258 Harré, R.  256 Hartford, B.  17 Hasan, R.  269, 276 Hattori, S.  288 Hayashi, M.  257, 281 Hayashi, R.  74 He, A.  10, 17 Hedberg, N.  256 Herzog, M. I.  320 Hill, B.  40, 41 Hill, J. H.  245 Hinds, J.  13, 41, 60n, 270, 304 Hirose, K.  290, 302 Hirschberg, J.  290 Hirst, D.  288, 290 Holt, E.  63 Honda, A.  293, 316, 319 Horie, P. I.  2 Horiguchi, K.  254n, 255, 257, 259, 265, 280, 281 Horii, R.  186n Hosotani, F.  230, 242 Hudson, M. E.  131, 138n, 152, 163, 165n, 167n, 182

 Style Shifting in Japanese I Ide, S.  10, 13, 14, 40, 41, 186n, 198, 206 Iijima, A.  103, 105, 106 Ikuta, S.  2, 14, 39, 41, 45, 61, 71, 75, 135, 251 Inoue, F.  2, 133, 136, 138, 139, 141n, 229, 245, 246 Inoue, M.  206 Irvine, J.  14, 186n Ishida, T.  287, 316 Itsuki, H.  112 Iwasaki, S.  2, 260 J Jefferson, G.  287, 290, 291n Jinnouchi, M.  133, 141n, 154 Johnson, D. M.  256 Johnson, S.  206 Johnstone, B.  225, 252, 278 Jones, K.  194, 257, 287, 293, 304, 316, 319 Jorden, E.  12, 72n, 136n, 150, 161, 163, 189n K Kamio, A.  13, 138n, 255 Kanazawa, H.  245, 246 Kataoka, K.  251, 252, 256 Kawakami, S.  288, 290, 302 Kawasaki, A.  40, 41 Keating, E.  17 Keenan, E.  60n, 186n, 254 Kindaichi, H.  42, 136, 137, 152, 288 Kinsui, S.  97, 118, 138n, 255, 257, 265, 280, 281 Kitagawa, C.  193n Kitano, H.  171 Kobayashi, M.  229 Kobayashi, T.  165n Konishi, D.  106, 107 Konomi, E.  215 Konoshima, M.  188 Koori, S.  286, 289, 290, 297, 302, 303, 310, 311, 312 Koyano, T.  154 Krauss, E. S.  287 Kruyt, J. G.  286, 290 Kubozono, H.  289, 310 Kulick, D.  2, 186n Kuno, S.  255, 280n L Labov, W.  9, 185, 216, 300, 320

Ladd, R. D.  285, 286, 287, 288, 290 Lakoff, R.  40, 186n, 255 Langacker, R.  171n Lebra, T. S.  186n, 192n Lee, Y.  245 Leech, G.  40 Lee-Wong, S. M.  138n Levinson, S.  40, 41, 49, 54, 57, 61, 63, 74, 76, 78, 82, 193, 252, 292, 304 Linde, C.  256 Long, D.  95, 229 Longacre, R. E.  270, 278 Lyons, J.  46n M MacWhinney, B.  280 Maehlum, B.  230, 240, 241 Maekawa, K.  289, 310 Makino, S.  2, 12, 26, 42, 44,135 Maltz, D. N.  206 Mao, L.  40 Martin, S.  13, 137, 138, 152, 183 Matsuda, K.  314 Matsumoto, S.  152, 153 Matsumoto, Y.  2, 4, 5, 40, 187, 193n, 213, 216, 218, 222, 225, 251 Mayes, P.  257, 281 Maynard, S. K.  2, 3, 4, 13, 15, 26, 39, 42, 60n, 62, 71, 75, 91, 92, 93n, 94, 95, 96, 97, 99, 102, 109, 115, 125, 126, 138n, 150, 251 McConnell-Ginet, S.  186n, 206, 214 McElhinny, B.  2 McGloin, N. H.  193n McLaughlin, M. L.  71, 78 Megumi, M.  35 Mikami, K.  72n, 281 Mills, S.  40 Milroy, J.  246 Milroy, L.  246 Miyabe, M.  133, 143, 152, 154, 155 Miyake, Y.  229 Mizutani, N.  163, 186n, 187n, 206, 215, 216 Mizutani, O.  163, 186n, 187n, 206, 215, 216 Mori, A.  121 Mori, J.  13, 304, 319 Morita, Y.  254n, 255, 259 Mühlhäusler, P.  256

Murakami, R.  110 N Nakayama, K.  232 Naruoka, K.  257, 281 Natsuki, S.  152, 153, 154 Neustupny, J. V.  13 Nishide, K.  254 Niyekawa, A.  13, 14, 287 Nobumoto, K.  118, 119 Noda, H.  96, 102, 105, 161, 163, 165n Noda, M.  12 Nooteboom, S. G.  286, 290 Nwoye, O. G.  40 Nyquist, L.  185 O O’Barr, W. M.  206 Ochs, E.  10, 14, 15, 41, 186, 214, 215, 244, 278 Ogino, T.  40, 41 Ohta, A. S.  138n Oishi, H.  288, 290, 302 Okamoto, N.  252 Okamoto, S.  2, 5, 14, 15, 16, 22, 25, 35, 39, 40, 43, 46, 186n, 187, 193n, 206, 216, 229, 240, 242, 244, 251 Ono, T.  257, 281, 304, 318 O’Shaughnessy, D.  290 Ozaki, Y.  116 P Pan, Y.  40 Paolino, D.  189, 210 Pearson, B.  307n Philips, S. U.  185n Pierrehumbert, J.  289, 302, 310, 311 Pitrelli, J.  312 Pomerantz, A.  293, 304 Preston, D.  287, 301, 306 Prince, E.  286, 290, 303 Q Quinn, C. J.  1 Quinn, N.  206 R Rand, D.  300 Reynolds, K. A.  186n, 206 Richards, K.  87 Rickford, J.  9, 246 Roepcke, Y.  256



Name index  Rohlen, T. P.  287 Rosenberger, N. R.  195n Rounds, P.  34 Rubba, J.  256 S Sacks, H.  10, 287, 290, 291n Saitoo, M.  107, 108 Sakita, T. I.  252, 278 Sakuma, K.  254 Sanada, S.  229, 230, 231n, 239n, 241, 242, 245 Sankoff, D.  287, 300, 301 Satake, H.  92 Sato, S.  187, 193n, 206, 216 Sato, T.  188, 251 Sawaki, K.  110, 111 Schegloff, E.  10, 287, 290, 291n, 304 Schieffelin, B. B.  60n Schiffrin, D.  78, 252, 278 Schuetze-Coburn, S.  189, 210 Selting, M.  286, 290 Seto, K.  114 Seya, R.  34 Shibamoto, J. S.  187, 206 Shibata, T.  229, 231n, 245, 246 Shibatani, M.  41, 232, 233 Shiga, A.  117 Shingu, I.  230 Shinozaki, K.  229 Shooho, I.  255 Shooji, S.  101, 102 Shuy, W.  60n Sidner, C.  286 Sifianou, M.  40 Silverman, K. E. A.  312n Silverstein, M.  244, 252, 258 Simon, M. E.  152n Smith, J. S.  2, 186n Sperber, D.  114 Spitz, J.  312

Steinhoff, P. G.  287 Strauss, S.  256 Sugitou, S.  286, 288, 289, 290, 302, 303, 312, 314 Sukle, R.  2, 71 Suzuki, A.  93 Suzuki, R.  304, 318 Swann, J.  207 T Takagi, C.  230 Takahashi, N.  290, 302 Takano, S.  5, 285, 293, 306, 307n, 314, 320n Takasaki, M.  117, 187 Takekuro, M.  206 Takubo, Y.  138n, 255, 257, 265, 280, 281 Tanaka, A.  120 Tanaka, H.  138n Tannen, D.  48, 60n, 206, 252, 253, 270, 278 Tanomura, T.  139, 140, 141, 142, 150, 157, 165n, 167n Tatekawa, S.  114, 115 Teramura, H.  138, 151, 163, 215 Terao, R.  187n Terken, J.  287, 290 Tokieda, M.  93, 136, 137, 163 Tokuchi, S.  251 Trudgill, P.  185 Tsuji, D.  114 Tsujimura, N.  131n, 135, 136, 187 Tsujimura, T.  135, 137 U Uchida, N.  187n, 206 Uchida, Y.  110, 111 Uchidate, M.  112, 113 Uehara, S.  4, 141, 142, 143, 148, 149, 152n, 157, 161, 171 Umeda, N.  310 Usami, M.  96, 97

Uyeno, T.  138n, 194n, 215 V Van De Walle, L.  40 Van Dijk, T. A.  60n Van Donselaar, W.  286, 288, 290 Venditti, J.  288, 289, 296, 310 Vološinov, V. N.  92, 98, 112 W Wada, M.  288 Warren, B.  171n Watanabe, S.  319 Watts, R. J.  40, 41 Weinreich, U.  320 Wertsch, J. V.  92 West, C.  206 Wetzel, P.  2, 186n, 187 Wierzbicka, A.  40 Wilson, D.  114 Wilson, J.  256 Wolfson, N.  42n, 252, 278 Woolard, K. A.  14, 244, 245, 246, 247 Wray, A.  170n Y Yamada, H.  60n, 190, 304, 319 Yamagata, M.  72n Yamazaki, M.  2 Yarimizu, K.  139n Yasuda, T.  245, 246 Yokoo, M.  290, 302 Yoneda, M.  229 Yoshida, M.  14, 137 Yoshimoto, K.  254n, 255, 257, 262, 280n Young, R.  300n Yule, G.  46n Z Zacharski, R.  256 Zimmerman, E.  187n

Subject index A accent  232–233, 288–290, 298–300, 302, 310–314 accent degeneration  289, 310 affect  2, 15, 26, 62–63, 75, 214, 223, 291 affect keys  34, 46–47, 49, 54, 56, 66 affirmative forms  131–133, 136–137, 139, 143, 151, 154, 162–163, 165, 174, 182–183 age  2, 13, 23–24, 31, 44, 55–57, 115–117, 123, 138, 141, 143, 152, 157, 183, 197–199, 205, 224–225 agreement (see also Social Agreement Principle)  56, 65, 107, 192, 218–220, 304, 314–316 anaphora  255, 264, 270, 280–281 apology  170, 292 attitudinal style  213–214, 220, 226 B backchannels  51–52, 78–79, 81 base-line style  4, 99–125 bi-dialectalism  229 bunmyaku-shiji (textual deixis)  254–255, 259, 262, 279–280 C casual (style, speech, etc)  9, 56–57, 91, 106–108, 114, 179–183, 217, 322 casual conversation  42, 44–45, 73, 225, 287, 291–293, 304–309 cataphora  255, 280 co-construction  29, 32, 49, 74, 85 code  230–231, 241, 244, 246–247 code-switching  9–10, 229–230, 237, 240–241, 244, 246, 285 Cognitive Prominence Principle (CPP)  290–294, 306

cohesion  47, 258, 263–264, 269–270, 274, 278 collaborative discourse  71, 74–75, 85, 87 collocations  261, 281 conflict (see also ‘confrontation’, disagreement’)  193–194, 206, 257, 286–287, 293 confrontation (see also ‘conflict’, ‘disagreement’)  311, 315–316, 321 Confucian society  198 context space  71–72, 77–78, 81, 85–87 contextualization cues  205, 278 contraction, contracted forms  169, 176, 179–181, 183 copula  11, 15, 131–132, 136–140, 142, 146–149, 156, 161–164, 168, 215, 218 creativity  91, 100, 112–120, 124–126 critical pedagogy  126 culture  93, 186, 199, 214–215, 223, 291, 293–294, 316 D da style (see also ‘non-desu/masu forms’, ‘plain forms’, ‘plain style’)  94–96, 99–100, 102–103, 105–107, 109–113, 115, 117, 120, 124 debate  287, 291–295, 305–306 deference  3, 15, 39–40, 42–45, 51, 55–57, 65–66, 135, 150, 156, 285, 317 deixis  5, 251–252, 254–281 delicate style  5, 214–215, 217–218, 220, 222–226 desu/-masu  2–4, 10–35, 39–67, 71–88, 91, 93–96, 98–106, 108–110, 112, 115–116, 123–124, 131–157, 161–165, 167–184 addressive desu/masu  99–105, 108, 121, 123

dialect, regional (see also ‘Hokkaido dialect’, ‘neohoogen’, ‘Tokyo Japanese’, ‘Yamaguchi dialect’  2, 5, 9, 229–247 dialect eradication  245–246 dialect style  213–214, 226 disagreement (see also ‘conflict’, ‘confrontation’)  3, 51–52, 186, 193, 291–293, 304–306, 315–16, 319 discourse markers  72, 78, 304, 315 discourse organization  3, 44, 73, 78, 82, 85, 87, 105 dislocation, syntactic  304, 318 distance  14, 39, 41, 110, 150, 178, 183 social/interpersonal distance  1, 14–15, 41–42, 285 psychological/emotional/ attitudinal distance  2, 14, 41, 97, 143, 150–151, 154, 156, 256 distancing effect  115, 117, 223 downstep  289, 310 E emotivity  95–96, 99–112, 115–116, 118–126, 318 evaluation, evaluative commentary  46–47, 156, 257–258, 266, 276, 278, 280 exophora  255, 262 expressive style (see ‘attitudinal style’) F face  40, 84, 87–88, 114, 293–294, 304, 311, 314–316, 319, 321, 325 face-threat  40–41, 50, 52, 54, 58, 61, 65–66, 76, 81–82, 193–195, 292, 294, 304, 311, 314–316, 319

 Style Shifting in Japanese femininity  185–189, 198–199, 204–207, 214, 223 feminine forms  186–190, 193–202, 204–207 feminine style/voice  108, 117, 186, 214 figure (see also ‘figureground’)  273, 279 figure-ground (see also ‘figure’, ‘ground’)  257–258, 275–276 first-encounter conversations  4, 180 floor  72, 74–75, 77–78, 81–82, 84–87, 251 focal prominence  286–294, 296–297, 299–303, 306–307, 310–311, 313–316, 318–322 focalization  258, 266, 276, 278–280 footing  74, 287, 304–305, 311, 313, 315, 319 forceful style  214–215, 217–220, 223–226 formality  43–45, 54–55, 242–245 formal forms  23, 223 formal style  94, 105, 109, 123, 230, 287 frame, framing  3, 47, 50, 53–54, 58, 60–64 G genba-shiji (space deixis)  254, 259, 262, 279–280 gender  2, 4, 115–117, 124, 142–143, 152–153, 183, 185–207, 213–226 GoldVarb  300, 305, 307 grammaticalization  169, 171–172, 261 ground (see also ‘figureground’)  257, 273, 278–279 H hedging  3, 53–54, 62–64, 67, 257 hierarchy, social  3, 19–22, 25, 29–35, 150, 196, 199 hierarchy, discourse  41, 61, 77 hierarchy of information  303 Hokkaido dialect  294 honorifics  11, 13–16, 35, 39, 41–43, 46, 127, 131, 135–137, 139, 151, 176, 206, 224, 242–243 addressee honorifics  10, 39, 71, 135, 183, 242–243

referent honorifics  21, 49, 57, 135, 224 hoogen (see ‘dialect’) hoogen bokumetsu (see ‘dialect eradication’) hyoojun-go (see ‘Standard Japanese’) I iconicity  181 identity  93, 107, 115, 186, 198–199, 205–206 age identity  206 gender identity  193, 206, 225 social identity  2, 4, 9–10, 14–17, 22, 25, 213–214, 216, 225–226 ideology  34–35, 108, 185, 214, 220, 224–225, 244–246 idioms  147, 149, 157, 169, 179, 184 indexing  4–5, 10, 15–17, 19–22, 24–26, 31, 33–34, 41–42, 95, 186–187, 189, 206, 213–214, 220, 242–244, 246, 257–258, 271, 274, 276, 278–280 informality  244–245 informal style  141, 230, 242 information  62, 94, 98–99, 102, 142, 157, 291–292, 306, 312–313 information status, information structure  256, 286–290, 294, 300–304, 314–315, 321 interviews  3, 14, 42, 45–46, 71–87, 141–143, 152, 166 involvement  5, 20, 22, 63, 78–79, 81, 94, 220, 247, 252–254, 257–258, 264, 269, 271, 273, 276, 278–281 irony  112, 114–115, 126 J joking  45–50, 54, 60, 61, 244, 293, 304, 316–317 K keigo (see ‘honorifics’) kyootsuu-go  245, 295 L language policy  229, 245–246 lexicalization  169, 171–172 linear style  98, 112

M masculinity  188, 205–206, 214 masculine forms  187–191, 193–194, 206, 215 masculine style/voice  4, 108, 117–118, 124, 214 masu form, masu style (see ‘desu/-masu’) mitigation  3, 50–54, 58, 61–62, 64–66, 76, 81–82, 195, 197 monologue  44, 53, 61–62, 252–254, 257–258, 260, 265, 271, 280–281, 286, 290 N negation, negative forms  4–5, 131–157, 161–184, 285–322 neo-hoogen (neo-dialect)  230, 241 neutral forms (gendered language)  187–190, 193–194, 197–200, 204, 230 neutral style (gendered language)  4, 117, 193 neutral style (regional dialects)  230 news (see ‘television, news’) non-desu/-masu forms (see also ‘da style’, ‘plain forms’, ‘plain style’)  72–88 O overlap, overlapped utter­ ances  73, 181–182, 210 P paralinguistic  46, 67, 285–286, 294, 302, 311, 316, 319, 322 pause  78, 179, 181, 290, 296, 303, 311, 312, 318 pictorial style  98, 112 pitch  194, 286, 288–289, 296–300, 310–311 plain forms (also ‘da style’, ‘nondesu/-masu forms’, ‘plain style’)  2–3, 10–35, 39–67, 137, 150, 157, 215 naked plain form  26–28, 34, 62 plain style (also ‘da style’, ‘nondesu/-masu forms’, ‘plain forms’)  2–3, 131–136, 143, 150–151, 156, 161, 181, 183, 287



Subject index  politeness  3–4, 13–14, 16, 39–41, 55, 76, 87–88, 96–97, 114, 137, 150–151, 154, 156, 181–183, 185–189, 195, 206, 214, 248, 258 negative politeness  40, 64, 317 positive politeness  20, 40, 49, 78–79, 97, 293, 316 polite forms (also ‘polite-style forms’)  12, 96, 161–170, 177, 180, 182–183, 193, 200, 222 polite style  96, 143, 181, 183 polite-style affirmative  137, 182 polite-style negative  4, 156, 159, 161–163, 165–168, 173, 175, 177, 179, 181–184 postpositional particles  313–314 postpositions  304, 318 power  4, 13, 16–17, 29, 34, 41, 44, 185–189, 195, 197–199, 204–207 prosody  5, 285–322 Q qualification segments  53–54 question particle  178–179 R repair  28, 52, 76, 78–80 reset (pitch reset)  289, 296–300 S semi-polite forms, semi-polite style  4, 131–157, 182–183 sentence-final position  215

shin genbun itchitai  92, 106, 124, 126 Social Agreement Principle (SAP)  290–291, 293–294, 306, 315 sociolinguistic grammar  287, 301, 314, 320–321 solidarity  3, 44–52, 54, 56–57, 64–67, 246, 285, 291, 293, 316–317 soto  244 speech acts  25, 29, 49, 54–55, 57–58, 164, 218–220, 231, 242, 244 standard language, standard variety  9, 10, 240, 245 Standard Japanese (SJ)  2, 5, 166, 189, 199, 229–247 standardization  229, 231, 237, 241, 244–245 super-feminine forms  186–190, 193–200, 202, 204–206, 211 super-masculine forms  187, 189 supra-polite (form, style, etc)  93, 99, 100, 112–115, 117, 119–121, 123–124 T television  295 television debates (see also ‘debate’)  287, 292–294 television dramas  14, 95 television interviews/talk shows (see also ‘interviews’)  42, 45–46

television news 288, 294, 303, 306, 310, 313 tempo  290, 303, 312 te-ni-o-ha  93 Tokyo Japanese  289, 294 topic  9–10, 14, 16, 19, 25, 28, 35, 47, 73–75, 86, 105, 150, 170, 177–178, 183, 195, 197, 216, 242 new topic  34, 54, 61, 75, 82, 142, 157, 174, 176, 195, 222 topic change, topic shift  22, 25, 54, 60–61 topic continuity  270 U uchi  244 V variable rules  300 voice  11, 17, 25, 33–34, 92–95, 97–98, 100, 102, 105, 107–108, 110–112, 115, 117–120, 124, 195, 214 voices, multiple (also ‘multivoicedness’)  91–92, 97–98, 111, 117–118, 120–126 volitional politeness  41 W wakimae (discernment) politeness  13–14, 16, 35, 41 written style  91, 93, 98–99 Y Yamaguchi dialect (YD)  229–235, 237–248

Pragmatics & Beyond New Series A complete list of titles in this series can be found on the publishers’ website, www.benjamins.com 182 Lee, Seung-Hee: Requests and Responses in Calls for Service. Expected 2009 181 Maschler, Yael: Metalanguage in Interaction. Hebrew discourse markers. Expected 2009 180 Jones, Kimberly and Tsuyoshi Ono (eds.): Style Shifting in Japanese. 2008. vii, 335 pp. 179 Simões Lucas Freitas, Elsa: Taboo in Advertising. 2008. xix, 214 pp. 178 Schneider, Klaus P. and Anne Barron (eds.): Variational Pragmatics. A focus on regional varieties in pluricentric languages. 2008. vii, 371 pp. 177 Rue, Yong-Ju and Grace Zhang: Request Strategies. A comparative study in Mandarin Chinese and Korean. 2008. xv, 320 pp. 176 Jucker, Andreas H. and Irma Taavitsainen (eds.): Speech Acts in the History of English. 2008. viii, 318 pp. 175 Gómez González, María de los Ángeles, J. Lachlan Mackenzie and Elsa M. González Álvarez (eds.): Languages and Cultures in Contrast and Comparison. 2008. xxii, 364 pp. 174 Heyd, Theresa: Email Hoaxes. Form, function, genre ecology. 2008. vii, 239 pp. 173 Zanotto, Mara Sophia, Lynne Cameron and Marilda C. Cavalcanti (eds.): Confronting Metaphor in Use. An applied linguistic approach. 2008. vii, 315 pp. 172 Benz, Anton and Peter Kühnlein (eds.): Constraints in Discourse. 2008. vii, 292 pp. 171 Félix-Brasdefer, J. César: Politeness in Mexico and the United States. A contrastive study of the realization and perception of refusals. 2008. xiv, 195 pp. 170 Oakley, Todd and Anders Hougaard (eds.): Mental Spaces in Discourse and Interaction. 2008. vi, 262 pp. 169 Connor, Ulla, Ed Nagelhout and William Rozycki (eds.): Contrastive Rhetoric. Reaching to intercultural rhetoric. 2008. viii, 324 pp. 168 Proost, Kristel: Conceptual Structure in Lexical Items. The lexicalisation of communication concepts in English, German and Dutch. 2007. xii, 304 pp. 167 Bousfield, Derek: Impoliteness in Interaction. 2008. xiii, 281 pp. 166 Nakane, Ikuko: Silence in Intercultural Communication. Perceptions and performance. 2007. xii, 240 pp. 165 Bublitz, Wolfram and Axel Hübler (eds.): Metapragmatics in Use. 2007. viii, 301 pp. 164 Englebretson, Robert (ed.): Stancetaking in Discourse. Subjectivity, evaluation, interaction. 2007. viii, 323 pp. 163 Lytra, Vally: Play Frames and Social Identities. Contact encounters in a Greek primary school. 2007. xii, 300 pp. 162 Fetzer, Anita (ed.): Context and Appropriateness. Micro meets macro. 2007. vi, 265 pp. 161 Celle, Agnès and Ruth Huart (eds.): Connectives as Discourse Landmarks. 2007. viii, 212 pp. 160 Fetzer, Anita and Gerda Eva Lauerbach (eds.): Political Discourse in the Media. Cross-cultural perspectives. 2007. viii, 379 pp. 159 Maynard, Senko K.: Linguistic Creativity in Japanese Discourse. Exploring the multiplicity of self, perspective, and voice. 2007. xvi, 356 pp. 158 Walker, Terry: Thou and You in Early Modern English Dialogues. Trials, Depositions, and Drama Comedy. 2007. xx, 339 pp. 157 Crawford Camiciottoli, Belinda: The Language of Business Studies Lectures. A corpus-assisted analysis. 2007. xvi, 236 pp. 156 Vega Moreno, Rosa E.: Creativity and Convention. The pragmatics of everyday figurative speech. 2007. xii, 249 pp. 155 Hedberg, Nancy and Ron Zacharski (eds.): The Grammar–Pragmatics Interface. Essays in honor of Jeanette K. Gundel. 2007. viii, 345 pp. 154 Hübler, Axel: The Nonverbal Shift in Early Modern English Conversation. 2007. x, 281 pp. 153 Arnovick, Leslie K.: Written Reliquaries. The resonance of orality in medieval English texts. 2006. xii, 292 pp. 152 Warren, Martin: Features of Naturalness in Conversation. 2006. x, 272 pp. 151 Suzuki, Satoko (ed.): Emotive Communication in Japanese. 2006. x, 234 pp. 150 Busse, Beatrix: Vocative Constructions in the Language of Shakespeare. 2006. xviii, 525 pp.

149 Locher, Miriam A.: Advice Online. Advice-giving in an American Internet health column. 2006. xvi, 277 pp. 148 Fløttum, Kjersti, Trine Dahl and Torodd Kinn: Academic Voices. Across languages and disciplines. 2006. x, 309 pp. 147 Hinrichs, Lars: Codeswitching on the Web. English and Jamaican Creole in e-mail communication. 2006. x, 302 pp. 146 Tanskanen, Sanna-Kaisa: Collaborating towards Coherence. Lexical cohesion in English discourse. 2006. ix, 192 pp. 145 Kurhila, Salla: Second Language Interaction. 2006. vii, 257 pp. 144 Bührig, Kristin and Jan D. ten Thije (eds.): Beyond Misunderstanding. Linguistic analyses of intercultural communication. 2006. vi, 339 pp. 143 Baker, Carolyn, Michael Emmison and Alan Firth (eds.): Calling for Help. Language and social interaction in telephone helplines. 2005. xviii, 352 pp. 142 Sidnell, Jack: Talk and Practical Epistemology. The social life of knowledge in a Caribbean community. 2005. xvi, 255 pp. 141 Zhu, Yunxia: Written Communication across Cultures. A sociocognitive perspective on business genres. 2005. xviii, 216 pp. 140 Butler, Christopher S., María de los Ángeles Gómez González and Susana M. Doval-Suárez (eds.): The Dynamics of Language Use. Functional and contrastive perspectives. 2005. xvi, 413 pp. 139 Lakoff, Robin T. and Sachiko Ide (eds.): Broadening the Horizon of Linguistic Politeness. 2005. xii, 342 pp. 138 Müller, Simone: Discourse Markers in Native and Non-native English Discourse. 2005. xviii, 290 pp. 137 Morita, Emi: Negotiation of Contingent Talk. The Japanese interactional particles ne and sa. 2005. xvi, 240 pp. 136 Sassen, Claudia: Linguistic Dimensions of Crisis Talk. Formalising structures in a controlled language. 2005. ix, 230 pp. 135 Archer, Dawn: Questions and Answers in the English Courtroom (1640–1760). A sociopragmatic analysis. 2005. xiv, 374 pp. 134 Skaffari, Janne, Matti Peikola, Ruth Carroll, Risto Hiltunen and Brita Wårvik (eds.): Opening Windows on Texts and Discourses of the Past. 2005. x, 418 pp. 133 Marnette, Sophie: Speech and Thought Presentation in French. Concepts and strategies. 2005. xiv, 379 pp. 132 Onodera, Noriko O.: Japanese Discourse Markers. Synchronic and diachronic discourse analysis. 2004. xiv, 253 pp. 131 Janoschka, Anja: Web Advertising. New forms of communication on the Internet. 2004. xiv, 230 pp. 130 Halmari, Helena and Tuija Virtanen (eds.): Persuasion Across Genres. A linguistic approach. 2005. x, 257 pp. 129 Taboada, María Teresa: Building Coherence and Cohesion. Task-oriented dialogue in English and Spanish. 2004. xvii, 264 pp. 128 Cordella, Marisa: The Dynamic Consultation. A discourse analytical study of doctor–patient communication. 2004. xvi, 254 pp. 127 Brisard, Frank, Michael Meeuwis and Bart Vandenabeele (eds.): Seduction, Community, Speech. A Festschrift for Herman Parret. 2004. vi, 202 pp. 126 Wu, Yi’an: Spatial Demonstratives in English and Chinese. Text and Cognition. 2004. xviii, 236 pp. 125 Lerner, Gene H. (ed.): Conversation Analysis. Studies from the first generation. 2004. x, 302 pp. 124 Vine, Bernadette: Getting Things Done at Work. The discourse of power in workplace interaction. 2004. x, 278 pp. 123 Márquez Reiter, Rosina and María Elena Placencia (eds.): Current Trends in the Pragmatics of Spanish. 2004. xvi, 383 pp. 122 González, Montserrat: Pragmatic Markers in Oral Narrative. The case of English and Catalan. 2004. xvi, 410 pp. 121 Fetzer, Anita: Recontextualizing Context. Grammaticality meets appropriateness. 2004. x, 272 pp. 120 Aijmer, Karin and Anna-Brita Stenström (eds.): Discourse Patterns in Spoken and Written Corpora. 2004. viii, 279 pp.

119 Hiltunen, Risto and Janne Skaffari (eds.): Discourse Perspectives on English. Medieval to modern. 2003. viii, 243 pp. 118 Cheng, Winnie: Intercultural Conversation. 2003. xii, 279 pp. 117 Wu, Ruey-Jiuan Regina: Stance in Talk. A conversation analysis of Mandarin final particles. 2004. xvi, 260 pp. 116 Grant, Colin B. (ed.): Rethinking Communicative Interaction. New interdisciplinary horizons. 2003. viii, 330 pp. 115 Kärkkäinen, Elise: Epistemic Stance in English Conversation. A description of its interactional functions, with a focus on I think. 2003. xii, 213 pp. 114 Kühnlein, Peter, Hannes Rieser and Henk Zeevat (eds.): Perspectives on Dialogue in the New Millennium. 2003. xii, 400 pp. 113 Panther, Klaus-Uwe and Linda L. Thornburg (eds.): Metonymy and Pragmatic Inferencing. 2003. xii, 285 pp. 112 Lenz, Friedrich (ed.): Deictic Conceptualisation of Space, Time and Person. 2003. xiv, 279 pp. 111 Ensink, Titus and Christoph Sauer (eds.): Framing and Perspectivising in Discourse. 2003. viii, 227 pp. 110 Androutsopoulos, Jannis K. and Alexandra Georgakopoulou (eds.): Discourse Constructions of Youth Identities. 2003. viii, 343 pp. 109 Mayes, Patricia: Language, Social Structure, and Culture. A genre analysis of cooking classes in Japan and America. 2003. xiv, 228 pp. 108 Barron, Anne: Acquisition in Interlanguage Pragmatics. Learning how to do things with words in a study abroad context. 2003. xviii, 403 pp. 107 Taavitsainen, Irma and Andreas H. Jucker (eds.): Diachronic Perspectives on Address Term Systems. 2003. viii, 446 pp. 106 Busse, Ulrich: Linguistic Variation in the Shakespeare Corpus. Morpho-syntactic variability of second person pronouns. 2002. xiv, 344 pp. 105 Blackwell, Sarah: Implicatures in Discourse. The case of Spanish NP anaphora. 2003. xvi, 303 pp. 104 Beeching, Kate: Gender, Politeness and Pragmatic Particles in French. 2002. x, 251 pp. 103 Fetzer, Anita and Christiane Meierkord (eds.): Rethinking Sequentiality. Linguistics meets conversational interaction. 2002. vi, 300 pp. 102 Leafgren, John: Degrees of Explicitness. Information structure and the packaging of Bulgarian subjects and objects. 2002. xii, 252 pp. 101 Luke, K. K. and Theodossia-Soula Pavlidou (eds.): Telephone Calls. Unity and diversity in conversational structure across languages and cultures. 2002. x, 295 pp. 100 Jaszczolt, Katarzyna M. and Ken Turner (eds.): Meaning Through Language Contrast. Volume 2. 2003. viii, 496 pp. 99 Jaszczolt, Katarzyna M. and Ken Turner (eds.): Meaning Through Language Contrast. Volume 1. 2003. xii, 388 pp. 98 Duszak, Anna (ed.): Us and Others. Social identities across languages, discourses and cultures. 2002. viii, 522 pp. 97 Maynard, Senko K.: Linguistic Emotivity. Centrality of place, the topic-comment dynamic, and an ideology of pathos in Japanese discourse. 2002. xiv, 481 pp. 96 Haverkate, Henk: The Syntax, Semantics and Pragmatics of Spanish Mood. 2002. vi, 241 pp. 95 Fitzmaurice, Susan M.: The Familiar Letter in Early Modern English. A pragmatic approach. 2002. viii, 263 pp. 94 McIlvenny, Paul (ed.): Talking Gender and Sexuality. 2002. x, 332 pp. 93 Baron, Bettina and Helga Kotthoff (eds.): Gender in Interaction. Perspectives on femininity and masculinity in ethnography and discourse. 2002. xxiv, 357 pp. 92 Gardner, Rod: When Listeners Talk. Response tokens and listener stance. 2001. xxii, 281 pp. 91 Gross, Joan: Speaking in Other Voices. An ethnography of Walloon puppet theaters. 2001. xxviii, 341 pp. 90 Kenesei, István and Robert M. Harnish (eds.): Perspectives on Semantics, Pragmatics, and Discourse. A Festschrift for Ferenc Kiefer. 2001. xxii, 352 pp. 89 Itakura, Hiroko: Conversational Dominance and Gender. A study of Japanese speakers in first and second language contexts. 2001. xviii, 231 pp. 88 Bayraktaroğlu, Arın and Maria Sifianou (eds.): Linguistic Politeness Across Boundaries. The case of Greek and Turkish. 2001. xiv, 439 pp.

87 Mushin, Ilana: Evidentiality and Epistemological Stance. Narrative Retelling. 2001. xviii, 244 pp. 86 Ifantidou, Elly: Evidentials and Relevance. 2001. xii, 225 pp. 85 Collins, Daniel E.: Reanimated Voices. Speech reporting in a historical-pragmatic perspective. 2001. xx, 384 pp. 84 Andersen, Gisle: Pragmatic Markers and Sociolinguistic Variation. A relevance-theoretic approach to the language of adolescents. 2001. ix, 352 pp. 83 Márquez Reiter, Rosina: Linguistic Politeness in Britain and Uruguay. A contrastive study of requests and apologies. 2000. xviii, 225 pp. 82 Khalil, Esam N.: Grounding in English and Arabic News Discourse. 2000. x, 274 pp. 81 Di Luzio, Aldo, Susanne Günthner and Franca Orletti (eds.): Culture in Communication. Analyses of intercultural situations. 2001. xvi, 341 pp. 80 Ungerer, Friedrich (ed.): English Media Texts – Past and Present. Language and textual structure. 2000. xiv, 286 pp. 79 Andersen, Gisle and Thorstein Fretheim (eds.): Pragmatic Markers and Propositional Attitude. 2000. viii, 273 pp. 78 Sell, Roger D.: Literature as Communication. The foundations of mediating criticism. 2000. xiv, 348 pp. 77 Vanderveken, Daniel and Susumu Kubo (eds.): Essays in Speech Act Theory. 2002. vi, 328 pp. 76 Matsui, Tomoko: Bridging and Relevance. 2000. xii, 251 pp. 75 Pilkington, Adrian: Poetic Effects. A relevance theory perspective. 2000. xiv, 214 pp. 74 Trosborg, Anna (ed.): Analysing Professional Genres. 2000. xvi, 256 pp. 73 Hester, Stephen K. and David Francis (eds.): Local Educational Order. Ethnomethodological studies of knowledge in action. 2000. viii, 326 pp. 72 Marmaridou, Sophia S.A.: Pragmatic Meaning and Cognition. 2000. xii, 322 pp. 71 Gómez González, María de los Ángeles: The Theme–Topic Interface. Evidence from English. 2001. xxiv, 438 pp. 70 Sorjonen, Marja-Leena: Responding in Conversation. A study of response particles in Finnish. 2001. x, 330 pp. 69 Noh, Eun-Ju: Metarepresentation. A relevance-theory approach. 2000. xii, 242 pp. 68 Arnovick, Leslie K.: Diachronic Pragmatics. Seven case studies in English illocutionary development. 2000. xii, 196 pp. 67 Taavitsainen, Irma, Gunnel Melchers and Päivi Pahta (eds.): Writing in Nonstandard English. 2000. viii, 404 pp. 66 Jucker, Andreas H., Gerd Fritz and Franz Lebsanft (eds.): Historical Dialogue Analysis. 1999. viii, 478 pp. 65 Cooren, François: The Organizing Property of Communication. 2000. xvi, 272 pp. 64 Svennevig, Jan: Getting Acquainted in Conversation. A study of initial interactions. 2000. x, 384 pp. 63 Bublitz, Wolfram, Uta Lenk and Eija Ventola (eds.): Coherence in Spoken and Written Discourse. How to create it and how to describe it. Selected papers from the International Workshop on Coherence, Augsburg, 24-27 April 1997. 1999. xiv, 300 pp. 62 Tzanne, Angeliki: Talking at Cross-Purposes. The dynamics of miscommunication. 2000. xiv, 263 pp. 61 Mills, Margaret H. (ed.): Slavic Gender Linguistics. 1999. xviii, 251 pp. 60 Jacobs, Geert: Preformulating the News. An analysis of the metapragmatics of press releases. 1999. xviii, 428 pp. 59 Kamio, Akio and Ken-ichi Takami (eds.): Function and Structure. In honor of Susumu Kuno. 1999. x, 398 pp. 58 Rouchota, Villy and Andreas H. Jucker (eds.): Current Issues in Relevance Theory. 1998. xii, 368 pp. 57 Jucker, Andreas H. and Yael Ziv (eds.): Discourse Markers. Descriptions and theory. 1998. x, 363 pp. 56 Tanaka, Hiroko: Turn-Taking in Japanese Conversation. A Study in Grammar and Interaction. 2000. xiv, 242 pp. 55 Allwood, Jens and Peter Gärdenfors (eds.): Cognitive Semantics. Meaning and cognition. 1999. x, 201 pp. 54 Hyland, Ken: Hedging in Scientific Research Articles. 1998. x, 308 pp. 53 Mosegaard Hansen, Maj-Britt: The Function of Discourse Particles. A study with special reference to spoken standard French. 1998. xii, 418 pp.

52 Gillis, Steven and Annick De Houwer (eds.): The Acquisition of Dutch. With a Preface by Catherine E. Snow. 1998. xvi, 444 pp. 51 Boulima, Jamila: Negotiated Interaction in Target Language Classroom Discourse. 1999. xiv, 338 pp. 50 Grenoble, Lenore A.: Deixis and Information Packaging in Russian Discourse. 1998. xviii, 338 pp. 49 Kurzon, Dennis: Discourse of Silence. 1998. vi, 162 pp. 48 Kamio, Akio: Territory of Information. 1997. xiv, 227 pp. 47 Chesterman, Andrew: Contrastive Functional Analysis. 1998. viii, 230 pp. 46 Georgakopoulou, Alexandra: Narrative Performances. A study of Modern Greek storytelling. 1997. xvii, 282 pp. 45 Paltridge, Brian: Genre, Frames and Writing in Research Settings. 1997. x, 192 pp. 44 Bargiela-Chiappini, Francesca and Sandra J. Harris: Managing Language. The discourse of corporate meetings. 1997. ix, 295 pp. 43 Janssen, Theo and Wim van der Wurff (eds.): Reported Speech. Forms and functions of the verb. 1996. x, 312 pp. 42 Kotthoff, Helga and Ruth Wodak (eds.): Communicating Gender in Context. 1997. xxvi, 424 pp. 41 Ventola, Eija and Anna Mauranen (eds.): Academic Writing. Intercultural and textual issues. 1996. xiv, 258 pp. 40 Diamond, Julie: Status and Power in Verbal Interaction. A study of discourse in a close-knit social network. 1996. viii, 184 pp. 39 Herring, Susan C. (ed.): Computer-Mediated Communication. Linguistic, social, and cross-cultural perspectives. 1996. viii, 326 pp. 38 Fretheim, Thorstein and Jeanette K. Gundel (eds.): Reference and Referent Accessibility. 1996. xii, 312 pp. 37 Carston, Robyn and Seiji Uchida (eds.): Relevance Theory. Applications and implications. 1998. x, 300 pp. 36 Chilton, Paul, Mikhail V. Ilyin and Jacob L. Mey (eds.): Political Discourse in Transition in Europe 1989–1991. 1998. xi, 272 pp. 35 Jucker, Andreas H. (ed.): Historical Pragmatics. Pragmatic developments in the history of English. 1995. xvi, 624 pp. 34 Barbe, Katharina: Irony in Context. 1995. x, 208 pp. 33 Goossens, Louis, Paul Pauwels, Brygida Rudzka-Ostyn, Anne-Marie SimonVandenbergen and Johan Vanparys: By Word of Mouth. Metaphor, metonymy and linguistic action in a cognitive perspective. 1995. xii, 254 pp. 32 Shibatani, Masayoshi and Sandra A. Thompson (eds.): Essays in Semantics and Pragmatics. In honor of Charles J. Fillmore. 1996. x, 322 pp. 31 Wildgen, Wolfgang: Process, Image, and Meaning. A realistic model of the meaning of sentences and narrative texts. 1994. xii, 281 pp. 30 Wortham, Stanton E.F.: Acting Out Participant Examples in the Classroom. 1994. xiv, 178 pp. 29 Barsky, Robert F.: Constructing a Productive Other. Discourse theory and the Convention refugee hearing. 1994. x, 272 pp. 28 Van de Walle, Lieve: Pragmatics and Classical Sanskrit. A pilot study in linguistic politeness. 1993. xii, 454 pp. 27 Suter, Hans-Jürg: The Wedding Report. A prototypical approach to the study of traditional text types. 1993. xii, 314 pp. 26 Stygall, Gail: Trial Language. Differential discourse processing and discursive formation. 1994. xii, 226 pp. 25 Couper-Kuhlen, Elizabeth: English Speech Rhythm. Form and function in everyday verbal interaction. 1993. x, 346 pp. 24 Maynard, Senko K.: Discourse Modality. Subjectivity, Emotion and Voice in the Japanese Language. 1993. x, 315 pp. 23 Fortescue, Michael, Peter Harder and Lars Kristoffersen (eds.): Layered Structure and Reference in a Functional Perspective. Papers from the Functional Grammar Conference, Copenhagen, 1990. 1992. xiii, 444 pp. 22 Auer, Peter and Aldo Di Luzio (eds.): The Contextualization of Language. 1992. xvi, 402 pp.

21 Searle, John R., Herman Parret and Jef Verschueren: (On) Searle on Conversation. Compiled and introduced by Herman Parret and Jef Verschueren. 1992. vi, 154 pp. 20 Nuyts, Jan: Aspects of a Cognitive-Pragmatic Theory of Language. On cognition, functionalism, and grammar. 1991. xii, 399 pp. 19 Baker, Carolyn and Allan Luke (eds.): Towards a Critical Sociology of Reading Pedagogy. Papers of the XII World Congress on Reading. 1991. xxi, 287 pp. 18 Johnstone, Barbara: Repetition in Arabic Discourse. Paradigms, syntagms and the ecology of language. 1991. viii, 130 pp. 17 Piéraut-Le Bonniec, Gilberte and Marlene Dolitsky (eds.): Language Bases ... Discourse Bases. Some aspects of contemporary French-language psycholinguistics research. 1991. vi, 342 pp. 16 Mann, William C. and Sandra A. Thompson (eds.): Discourse Description. Diverse linguistic analyses of a fund-raising text. 1992. xiii, 409 pp. 15 Komter, Martha L.: Conflict and Cooperation in Job Interviews. A study of talks, tasks and ideas. 1991. viii, 252 pp. 14 Schwartz, Ursula V.: Young Children's Dyadic Pretend Play. A communication analysis of plot structure and plot generative strategies. 1991. vi, 151 pp. 13 Nuyts, Jan, A. Machtelt Bolkestein and Co Vet (eds.): Layers and Levels of Representation in Language Theory. A functional view. 1990. xii, 348 pp. 12 Abraham, Werner (ed.): Discourse Particles. Descriptive and theoretical investigations on the logical, syntactic and pragmatic properties of discourse particles in German. 1991. viii, 338 pp. 11 Luong, Hy V.: Discursive Practices and Linguistic Meanings. The Vietnamese system of person reference. 1990. x, 213 pp. 10 Murray, Denise E.: Conversation for Action. The computer terminal as medium of communication. 1991. xii, 176 pp. 9 Luke, K. K.: Utterance Particles in Cantonese Conversation. 1990. xvi, 329 pp. 8 Young, Lynne: Language as Behaviour, Language as Code. A study of academic English. 1991. ix, 304 pp. 7 Lindenfeld, Jacqueline: Speech and Sociability at French Urban Marketplaces. 1990. viii, 173 pp. 6:3 Blommaert, Jan and Jef Verschueren (eds.): The Pragmatics of International and Intercultural Communication. Selected papers from the International Pragmatics Conference, Antwerp, August 1987. Volume 3: The Pragmatics of International and Intercultural Communication. 1991. viii, 249 pp. 6:2 Verschueren, Jef (ed.): Levels of Linguistic Adaptation. Selected papers from the International Pragmatics Conference, Antwerp, August 1987. Volume 2: Levels of Linguistic Adaptation. 1991. viii, 339 pp. 6:1 Verschueren, Jef (ed.): Pragmatics at Issue. Selected papers of the International Pragmatics Conference, Antwerp, August 17–22, 1987. Volume 1: Pragmatics at Issue. 1991. viii, 314 pp. 5 Thelin, Nils B. (ed.): Verbal Aspect in Discourse. 1990. xvi, 490 pp. 4 Raffler-Engel, Walburga von (ed.): Doctor–Patient Interaction. 1989. xxxviii, 294 pp. 3 Oleksy, Wieslaw (ed.): Contrastive Pragmatics. 1988. xiv, 282 pp. 2 Barton, Ellen: Nonsentential Constituents. A theory of grammatical structure and pragmatic interpretation. 1990. xviii, 247 pp. 1 Walter, Bettyruth: The Jury Summation as Speech Genre. An ethnographic study of what it means to those who use it. 1988. xvii, 264 pp.