51.1.1.1 .1... .
3- .rﬁLﬁ. . a:
s

.3

n4... .0. dis...“

A ‘ ..

keg. nit.
. L

2 L

 

 

 

 

4 "Tag;

 

r) LIBRARIES
’1) MICHIGAN STATE UNIVERSITY
f‘ - EAST LANSING, MICH 48824-1048
w: a 3 (Si

This is to certify that the
dissertation entitled

JAPANESE WOMEN’S LISTENING BEHAVIOR
IN FACE-TO-FACE CONVERSATION:
THE USE OF REACTIVE TOKENS AND NODS

presented by

Sachie Miyazaki

has been accepted towards fulﬁllment
of the requirements for the

Ph.D. degree in Linguistics and Germanic,
Slavic, Asian, and African
Languages

 

 

 

22¢,Z3iulo 21% [(7LM/

 

Major Professor's Signature
g" / ,? / DJ,

Date

 

MSU is an Affirmative Action/Equal Opportunity Institution

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 chlRC/DateDueJndd-pJS

 

 

 

JAPANESE WOMEN’S LISTENING BEHAVIOR
IN FACE-TO-FACE CONVERSATION:
THE USE OF REACTIVE TOKENS AND NODS

By

Sachie Miyazaki

A DISERTATION

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

DOCTOR OF PHILOSOPHY

Department of Linguistics and
Germanic, Slavic, Asian, and African Languages

2005

ABSTRACT
JAPANESE WOMEN’S LISTENING BEHAVIOR
IN FACE-TO-FACE CONVERSATION:
THE USE OF REACTIVE TOKENS AND NODS
By
Sachie Miyazaki

The present study examines Japanese women's use of reactive tokens (Clancy et
al. 1996) in Japanese face-to-face conversation and perception of other's listening
behavior in relation to the social variables of age and conversation styles (formal and
informal). Reactive tokens (RTs) examined in the present study are 8 verbal RTs such as
backchannels, reactive expressions, and laughter and 1 non-verbal RTs, nods.

The investigation is based on a quantitative analysis of formal and informal
conversation data from 30 participants and qualitative data of participant interview. The
perception is examined based on the ratings of 5 different listening behaviors in videos
and interview. The ﬁndings show that the use of RTs was inﬂuenced by age and
conversation styles but that perception does not vary according to age.

Age inﬂuences the proportion of nods and verbal RTs in the formal conversation.
The younger participants use more nods and fewer verbal RTs than do older participants
when there is difference in power between a speaker and a listener. In the informal
conversation, there is not age-graded effect in the use of verbal RTs and nods but 3
common interaction styles by all age groups are identiﬁed.

The variation of listening behavior in the present study is explained by v

'multidimensional modcl' (Tannen 1993a) in which 4 social contexts are deﬁned by

power and solidarity between participants.

The follow-up interview and the results of perception study revealed that the
perception of other's listening behavior may not vary according to perceivers' age.
Participants perceived the one which is closest to their own performance in informal
conversation most natural in the present study. This conﬁrms Hymes's notion of 'Speech
community' (1974) that members of a speech community share norms.

Findings of the present study suggest that nods are used as RTs as a strategy to
mitigate face threatening acts (Brown and Levinson 1987). In Japanese women's face-to-
face conversation, social variables of age and conversation styles play important roles in

the choice of an appropriate listening behavior in the contexts.

Copyright by
SACHIE MIYAZAKI
2005

To my son, Kai

ACKNOWLEDGEMENTS

First, I would like to thank Dr. Mutsuko Endo Hudson, the chair of my
dissertation committee and a mentor, for her guidance throughout my graduate study.
Next, I wish to thank my other committee members, Dr. Dennis Preston, Dr. Susan Gass
and Dr. Barbra Abbott, for their valuable suggestions to entire research design, insightful
cements and guidance to my dissertation. I am grateful to a number of researchers, Dr.
Shigeko Okamoto, Dr. Haruko M. Cook, and Dr. Polly Szatrowski and others for their
generous comments about methodology, interpretation of data, and analyses.

My sincere thanks go to students and staff of Tokyo Woman's Christian
University and Tokyo University of Pharmacy and Life Sciences and others who gave me
their precious time to participate in my study. Special thanks go to Dr. Tazuko Uyeno, Dr.
Suzuko Nichihara, Nanae F ukutomi, Kaneko Ishihara and many others who kindly
helped me to ﬁnd participants. I thank Dr. Akiko Hagiwara for her help in data collection
and technical support to make the data usable. I also thank Stephanie Shefﬁeld at MSU
Writing Center who read my manuscript from the early stages and corrected my English.
Her patience and continuous support even when I was in Japan helped me organize my
dissertation and keep writing far away from US.

I am very grateful to my parents for their endless support and encouragement
throughout my life. My special thanks go to my husband, Yoichi, for continuous support
and understanding without which I couldn't complete the program. Lastly, my deepest
thanks go to my son, Kai, who was born in the ﬁrst year of my Ph.D. program and has
become 8 years old 2 days after my defense. His encouragement and growth have pushed

me to strive to ﬁnish this sometimes overwhelming work.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................ xi
LIST OF FIGURES ........................................................................................................ xiii
CHAPTER 1
INTRODUCTION ............................................................................................................. 1
1.1. General remarks and goals of the present study ........................................................... 1
1.2. Hypotheses and ﬁndings ............................................................................................... 3
1.3. Outline of the dissertation ............................................................................................. 7
CHAPTER 2
REVIEW OF THE LITERATURE ON LISTENING BEHAVIOR ............................. 8
2.1 . Introduction .................................................................................................................. 8
2.2. Listening behavior in English ....................................................................................... 9
2.3. Listening behavior in Japanese ................................................................................... 12
2.3. 1 .Deﬁnitions .................................................................................................. 1 3
2.3.2. Functions of backchannels and nods ......................................................... 15
2.3.2.1. Discoursal and interactive functions ........................................ 15
2.3.2.2. Social functions-Politeness and solidarity ................................ 17
2.4. Cross-linguistic studies ............................................................................................... 19
2.4.1. Cross-linguistic differences in non-verbal listener responses .................... 21
2.4.2. Culture and interactional style ................................................................... 22
2.5. Social variables ........................................................................................................... 24
2.5.1. Sex ............................................................................................................. 24
2.5.2. Age ............................................................................................................. 26
2.5.3. Styles of conversation ................................................................................ 27
2.6. Preliminary studies ...................................................................................................... 29
2.6.1. Hypotheses and method ............................................................................. 29
2.6.2. Results ........................................................................................................ 30
2.6.3. Conclusions ................................................................................................ 33
2.7. Summary ..................................................................................................................... 33
CHAPTER 3
HYPOTHESES AND METHODOLOGY ..................................................................... 34
3. 1 .Introduction .................................................................................................................. 34
3.2. Hypotheses .................................................................................................................. 34
3.3. Method ........................................................................................................................ 38
3.3.1. Data ............................................................................................................ 38
3.3.2. Participants ................................................................................................ 38
3.3.3. Procedures .................................................................................................. 40
3.3.4. Formal conversation .................................................................................. 41
3.3.5. Informal conversation ................................................................................ 41

vii

3.4. Listening behavior to be analyzed in the present study .............................................. 42

3.4.1. Verbal reactive tokens ............................................................................... 42
3.4.1.1. Backchannels, aizuchi .............................................................. 43
3.4.1.2. Reactive expressions ................................................................ 43
3.4.1 .3. Repetitions ............................................................................... 44
3.4.1.4. Collaborative ﬁnishes ............................................................... 45
3.4.1.5. Laughter .................................................................................... 46
3.4.1.6. Short comments ........................................................................ 47
3.4.1.7. Resumptive openers .................................................................. 48
3.4.1.8. Paraphrases ............................................................................... 48
3.4.2.Non-verbal reactive tokens ......................................................................... 49
3.5. Social variables ........................................................................................................... 50
3.5.1. Age ............................................................................................................. 51
3.5.2. Style of conversation ................................................................................. 52
3.6. Summary ..................................................................................................................... 53
CHAPTER 4
REACTIVE TOKENS IN FORMAL CONVERSATION ........................................... 54
4.1 . Introduction ................................................................................................................. 54
4.2. Quantitative results of total reactive tokens ................................................................ 54
4.2.1. Total reactive tokens used in 2 minutes ..................................................... 54
4.2.2. Reactive tokens by age group .................................................................... 57
4.2.3. Interrelation between age and verbal and non-verbal reactive tokens63
4.2.4. Hypotheses 1 and 2 .................................................................................... 63
4.2.5. Types of reactive tokens by age ................................................................. 64
4.2.6. Types of backchannels used by each age group ........................................ 65
4.3. Effects of interactional styles ...................................................................................... 75
4.3.1. Interactional styles ..................................................................................... 76
4.3.2. Variation of interactional styles ................................................................. 77
4.4. Summary ..................................................................................................................... 80
CHAPTER 5
REACTIVE TOKENS IN INFORMALCONVERSATION ........................................ 81
5.1 . Introduction ................................................................................................................. 81
5.2.Quantitative results of informal conversation .............................................................. 81
5.2.1. Data ............................................................................................................ 81
5.2.2. Quantitative results .................................................................................... 84
5.2.3. Reactive tokens by age group .................................................................... 85
5.2.4. Types of reactive tokens by age ................................................................. 86
5.2.5. Types of backchannels used by age group ................................................. 87
5.3. Characteristics of female friend talk ........................................................................... 89
5.3.1 Assessment strategies ................................................................................. 89
5.3.1.1. Types of assessment strategies in the present study ................. 90
5.3.1.2. 'Complex assessment sequences' .............................................. 92
5.3.2. Other Characteristics ................................................................................. 95
5.3.2.1. Non-primary speakers' questions .............................................. 95

viii

5.3.2.2. Interruptions and overlapping ................................................. 100

5.3.2.3. Use of particles, ne ................................................................. 102
5.3.2.4. Use of masculine and feminine forms .................................... 106
5.3.3. Interactional styles in informal conversation ........................................... 109
5.4. Summary ................................................................................................................... 111
CHAPTER 6
DIFFERENCES IN STRUCTURE BETWEEN FORMAL AND
INFORMAL CONVERSATION .................................................................................. 113
6.1 . Introduction ............................................................................................................... 1 13
6.2. Quantitative results of formal and informal conversation ........................................ 113
6.2.1. Frequency of reactive tokens ................................................................... 114
6.2.2. Verbal reactive tokens and nods ............................................................... 115
6.2.3. The nature of verbal reactive tokens in formal and informal
conversation ............................................................................................. l 15
6.3. Age factor and reactive tokens .................................................................................. 117
6.4. Interactional styles .................................................................................................... 120
6.4.1. Listener's strategies .................................................................................. 120
6.4.2. Three commonly observed styles in Japanese listening behavior ........... 122
6.5. The 'multidimensional model' in relation to social factors ....................................... 123
6.5.1. Formal conversation model ..................................................................... 123
6.5.2. Informal conversation model ................................................................... 124
6.5.3. Japanese listening behavior ..................................................................... 125
6.6. Hypothesis 3 .............................................................................................................. 127
6.7. Summary ................................................................................................................... 127
CHAPTER 7
RESULTS OF THE INTERVIEW AND PERCEPTION STUDY ............................ 128
7 .1 . Introduction ............................................................................................................... 128
7.2. Follow up interview about listening behavior .......................................................... 128
7.2.1. Difference between formal and informal conversation ........................... 128
7.2.2. Types of verbal reactive tokens ............................................................... 130
7.2.3. Perception of nods ................................................................................... 131
7.2.4. Interactional style of informal conversation ............................................ 135
7.2.5. Goals of communication .......................................................................... 137
7.2.6. Social contexts and interactional style ..................................................... 138
7.2.7. Perception of family discourse ................................................................ 139
7.3. Perception study ........................................................................................................ 143
7.3.1. Method and data ...................................................................................... 144
7.3.2. Results ...................................................................................................... 146
7.3.2.1. Rating ..................................................................................... 146
7.3.2.2. Age and perception ................................................................. 152
7.3.2.3. Follow up interview about video clips ................................... 155
7.3.3. Hypothesis 4 and discussion .................................................................... 158
7.4. Summary ................................................................................................................... 159

ix

CHPTER 8

CONCLUSION .............................................................................................................. 160

8.1. Findings of the present study .................................................................................... 160
8.2. Suggestions for future study ..................................................................................... 163

APPENDICES ................................................................................................................ 165
APPENDIX A. Transcription methods and conventions .............................................. 166
APPENDIX B. Script of instructions ............................................................................ 167
APPENDIX C. Transcripts of conversation of video clips ........................................... 172
APPENDIX D. Questionnaires ..................................................................................... 176
APPENDIX E. Rating sheets for video clips ................................................................ 177
BIBLIOGRAPHY .......................................................................................................... 178

LIST OF TABLES

Table 2.1. List of participants in the preliminary study ..................................................... 29
Table 2.2. Frequency of verbal backchannels and nods .................................................... 32
Table 3.1. List of participants in formal conversation ....................................................... 39
Table 3.2. Age groups of participants ................................................................................ 51
Table 4.1. Total of reactive tokens by all participants (11 = 30) ......................................... 56
Table 4.2. Verbal and non-verbal reactive tokens used in 2-minute instruction by
age groups (n=30) ............................................................................................ 57

Table 4.3. Results of Tukey post hoc test of total reactive tokens among age groups ...... 58

Table 4.4. Results of Tukey post hoc test of verbal reactive tokens among

age groups ........................................................................................................ 59
Table 4.5. Results of Tukey post hoc test of nods among age groups ............................... 59
Table 4.6. Types of reactive tokens used by age groups .................................................... 65
Table 4.7. Result of Tukey post hoc test of backchannels among age groups .................. 65
Table 4.8. Types of backchannels by age groups .............................................................. 66
Table 4.9. Results of Tukey post hoc test of ee among age groups ................................... 71
Table 5.1. Lists of participants of informal conversation .................................................. 82
Table 5.2. Results of reactive tokens by all participants (n=13) ........................................ 85
Table 5.3. Verbal and non-verbal reactive tokens used by age groups (n= 1 3) .................. 86
Table 5.4. Age groups and types of reactive tokens in informal conversation .................. 87
Table 5.5. Types of backchannels by age groups .............................................................. 88
Table 6.1. Number of nods and verbal reactive tokens used in 60 minutes

of conversation ................................................................................................. 113
Table 6.2. Results of regression analysis between verbal reactive tokens and nods ....... 115

xi

Table 6.3. Results of regression analysis between age and nods in formal
conversation ................................................................................................... l 17

Table 6.4. Results of regression analysis between age and verbal reactive tokens ......... 118
Table 6.5. Results of regression analysis between age and backchannels ....................... 118

Table 6.6. Percentage of ee ‘yeah’ and resumptive openers by backchannels
in formal and informal conversation .............................................................. 119

Table 6.7. Interactional styles and participants in formal and informal

conversation ................................................................................................... 123
Table 7.1. Number of verbal reactive tokens and nods in 50—second video clips ............ 145
Table 7.2. Results of Tukey post hoc test of naturalness among videos ......................... 147
Table 7.3. Results of Tukey post hoc test of attentiveness among videos ....................... 149
Table 7.4. Results of Tukey post hoc test of preference among videos ........................... 151

Table 7.5. Results of ANOVA in terms of naturalness, attentiveness, preference,
and similarity in each video scene ................................................................. 152

Table 7.6. Results of Tukey post hoc test of video 2 in terms of naturalness
rating among groups ...................................................................................... 154

Table 7.7. Results of Tukey post hoc test of video 3 in terms of naturalness
rating among groups ...................................................................................... 154

xii

LIST OF FIGURES

Figure 2.1. Frequency of verbal backchannels .................................................................. 31
Figure 2.2. Frequency of nods ........................................................................................... 31
Figure 2.3. Frequency of verbal backchannels and nods ................................................... 32
Figure 4.1. Age and total number of reactive tokens ......................................................... 55

Figure 4.2. Verbal and non-verbal reactive tokens used in 2-minuite instruction
by age groups (n=30) ...................................................................................... 58

Figure 4.3. Interactional styles ........................................................................................... 76

Figure 4.4. Intersecting dimensions of closeness/distance and hierarchy/equality

(modiﬁed Tannen 1993a) ............................................................................... 77
Figure 5.1. The percentage of reactive tokens used in formal

and informal conversations .............................................................................. 88
Figure 6.1. Interactional styles in formal and informal conversations ............................. 116
Figure 6.2. Contexts of formal and informal conversation .............................................. 124
Figure 6.3. A model of Japanese listening behavior ........................................................ 126

Figure 7.1. The average of rating in terms of naturalness of listening
behavior in each video scene (n=30) ............................................................. 147

Figure 7.2. The average of rating in terms of attentiveness of listening
behavior in each video scene (n=30) ............................................................. 148

Figure 7.3. The average of rating in terms of preference of listening behavior
by the behavior in each video scene (n=3 0) .................................................. 150

Figure 7.4. The average of rating in terms of similarity to their own listening
behavior in each video scene (n=30) ............................................................. 151

xiii

CHAPTER 1

INTRODUCTION
1.1. General remarks and goals of the present study

Listening behavior in face-to-face conversation is characterized by verbal
backchannels (e.g. yes and uh-hum), short utterances which do not claim the ﬂoor (e.g.
partial repetitions), and non-verbal behavior (e.g. nods, gaze, gesture, and facial
expressions) (Yngve 1970; Duncun & Fiske 1977; Clancy et al. 1996 and others).
Listening behavior is considered 'multi-functional' (White 1987); that is, one form of
backchannels can serve multiple functions in a given context. For example, uh-hum can
function to express a listener's understanding, acknowledgement, and emotion toward the
speaker's utterances, and it may inﬂuence the conversation. Among various functions, it
is agreed that the primary function of backchannels is to serve as 'a continuer' (Shegloff
1982), a signal that the speaker can keep holding the ﬂoor.

The routine of listening behavior is highly conventionalized in a speech community,
and differs from culture to culture, as each language has interactional styles that may not
be common in others (Tannen 1882, 1984). For example, the Japanese language uses
verbal backchannels and nods more frequently compared with American English (Lebra
1976, Locastro 1987, Maynard 1989). A lack of listener responses may result in
misunderstandings and an overuse may also wrongly give an unintended negative
impression (Lebra 1976, Hinds 1982). The knowledge of socially accepted listening
behavior is important for both listeners and speakers to achieve successful

communication and conversation management (Maynard 1989; Szatrouski 2000).

Listener responses are not used at random but rather are governed by conversational
rules. Thus, the occurrence of listener responses is predictable by intonational,
grammatical, semantic and pragmatic contexts (Goodwin 1979, 1986; Maynard 1989;
Clancy et a1. 1996; F uro 200; Szatrousky 2003). Speakers of a language acquire strategies
as to how to use listener responses in different contexts of face-to-face conversation.

The norms of language use sometimes promote stereotypes about people who
belong to the speech community. For example, Japanese culture is said to value harmony
between participants of conversation (Watanabe 1993; Yamada 1992). Also, it is
commonly accepted that Japanese women's language is characterized by its indirect,
gentle, non-assertive, polite, and empathetic tone (Ide 1982; Shibatani 1990; Smith 1992
and others). These stereotyped cultural views however, have been challenged since they
do not necessarily represent the actual practice or its variation (Okamoto and Sato 1992;
Okamoto 1995; Cook 1999). Even within a language, social factors such as ethnic
groups, sex, and age may inﬂuence language practice and give rise to variations from the
norm (Tannen 1993b; Eckert 1993)

The present study investigates the verbal and non-verbal listening behavior of
Japanese female speakers of different ages. Listener responses are deﬁned as: the
listener's short utterances such as non-lexical vocalic forms and lexical forms that do not
claim the ﬂoor. These and vertical head movement that does not accompany verbal
utterances are also called Reactive Tokens (Clancy et a1. 1996, henceforth RTs).

The goals of this dissertation are: (1) to analyze how native female speakers of
Japanese use verbal and non-verbal RTs, nods, in face-to-face conversation; (2) to

investigate the relationship among verbal RTs; and (3) to analyze how age and styles of

g--.

conversation affect the use of RTs and the perception of listening behavior in others.

1.2. Hypotheses and ﬁndings
The following four hypotheses were formulated based on my earlier studies
(Miyazaki 2001b) and will be tested in the present study.
Hypothesis 1: The ratio of verbal to non-verbal RTs may vary among individuals,
but the total number will be similar in similar social contexts.

Hypothesis 2: The frequency of verbal and non-verbal RTs is affected by age. Older
female speakers use more verbal RTs than do younger speakers.

Hypothesis 3: The formality level of conversation inﬂuences not only the choice of
RT types and frequency, but also the interactional styles.

Hypothesis 4: The. perception of other's listening behavior varies according to the
part1c1pant's age.

Hypothesis 1 concerns the rules of listener responses that individual speakers of
Japanese follow in face-to-face conversation. Hypotheses 2 and 3 concern the social
variables of age and style (speech genre). Hypothesis 4 concerns the relationship between
the perception of listening behavior of others and the participant's age. These hypotheses
are based on the conviction that the use of backchannels does not occur randomly but is
governed by conversational rules. Their occurrences, therefore, are predictable (Maynard
1989; Clancy et a1. 1996).

Also, my preliminary study (Miyazaki 2001b) found that the total number of
listener verbal RTs and nods that do not accompany verbal RTs does not statistically
differ when listeners hear similar stories from the same person. In other words, the

ﬁndings showed that the total amount of listener response does not vary to a signiﬁcant

degree, although the ratio of verbal RTs to nods does. Among all the studies on Japanese
listener responses, there has yet to be one on the balance of verbal and non-verbal RTs
(i.e. nods) which the present study investigates. The rationales for the four hypotheses
will be discussed in fuller detail in Chapter 3.

In the present study, the dyadic conversation data of female speakers between 19
and 61 years old were collected in two types of conversations. In one type, a participant
listened to the researcher's instructions, while in the other, she chatted with her close
friend. The participants were all monolingual native speakers of Japanese who were born
and raised in Tokyo or its surrounding prefectures, such as Kanagawa and Saitama.

Each pair of participants met the researcher on the scheduled day. First the
researcher gave instructions individually for about 2 minutes in a separate room and then
the participants talked with each other for 15 minutes. The topic of conversation was on
what was given in the instruction to each participant. Both the instructions and the
conversation were recorded on a video camera, and a tape recorder or a mini disk
recorder. After the recording, participants were asked to watch 5 video clips (50 seconds
each) of casual conversation together and responded to a questionnaire. Then they were
asked to talk freely about what they thought about the video clips or anything about
listening behavior. 'Their comments were audio recorded and analyzed as follow-up
interview data.

A total of 80 minutes and 26 seconds of the conversations were transcribed and
analyzed. The results supported Hypothesis 1: The ratio of verbal to non-verbal RTs
varies among individuals, though the total number is similar in similar social contexts.

The total number of verbal RTs and non-verbal RTs (nods) occurring during 2-minute

instructions did not vary much among participants while the combination of verbal and
non-verbal RTs varied among individuals (from verbal, 99% vs. non-verbal, 1% to verbal,
6% vs non-verbal, 94%). The total number of verbal and non-verbal RTs showed a
negative correlation.

Hypothesis 2: The frequency of verbal and non-verbal RTs varies among
different age groups. Older female speakers’ use of more verbal RTs than younger
speakers was also supported statistically. That is, age and verbal RTs had a positive
correlation, and age and non-verbal RTs had a negative correlation. Younger participants
used more non-verbal responses while older participants tended to use more verbal ones
in formal conversation.

Hypothesis 3 was also supported statistically. The results of the informal
conversation data present different listening behavior from that in formal conversation.
For example, nods were used less frequently in informal conversation than in formal
conversation, and unlike formal conversation, the total number of verbal RTs and nods
did not have a relation in informal conversation. The results showed that the non-verbal
RTs were more important linguistic strategy in formal conversation than in informal
conversation.

The style of conversation also affected the listeners' choice of RTs. In formal
conversation participants used backchannels most frequently among all the verbal RTs;
on the other hand, they used non-backchannel verbal RTs, such as short comments,
repetitions, and collaborative ﬁnishes, more frequently than in formal conversations. The
different choices of verbal RTs inﬂuenced the interaction style in informal conversation.

That is, the interaction during informal conversation included more interruptions

(overlap) and frequent speaker changes than in formal conversation.

Age and interactional style did not show any particular relationship in informal
conversation and similar interactional styles were observed in all age groups. The total
number of backchannels and age, however, did show a positive correlation, and so did the
total number of verbal RTs and age. The ﬁndings show that in informal conversation even
though speakers of all ages tend to respond verbally, older speakers respond more
verbally than younger speakers. In other words, there is a larger gap in younger people's
styles of interaction between formal and informal conversation than in older people.

The variation according to genre, formal and informal conversation found in the
present study can be explained by Tannen's (1993a) 'multidimensional model.‘ In this
model, power and solidarity form one of the dimensions and each context of conversation
is plotted according to the degree of solidarity and power between participants. Details
will be presented in Chapters 4, 5, and 6.

Three interactional styles that can account for all the formal and informal
conversation data in the present study are proposed in Chapter 6. Japanese female
speakers' listening behavior can be identiﬁed as belonging to one of the three
interactional styles depending on the use of listener strategies.

The results of the follow-up interviews with the participants provide further
evidence for the variations in listening behavior in formal and informal conversation
presented in the present study. For example, some comments explained how younger
people perceived older people's listening behavior--why they did not use particular forms
of RTs, and what kind of listener responses were expected by speakers. This will be

discussed in detail in Chapter 7.

Finally, hypothesis 4: The hypothesis that perception of other's listening
behavior varies according to the participant's age was not supported by the results in the
video clip study. All the participants rated 5 different types of listening behavior
similarly. The one that consisted of the highest frequency of verbal RTs with mid
frequency of nods was rated as the most annoying listening behavior and most distant
from their own behavior by all age groups. And the one that consisted of the lowest
frequency of verbal RTs with the lowest frequency of nods was judged to be most natural
and preferable behavior by all age groups. The results showed that native speakers of
Japanese perceive someone else's listening behavior similarly and their intuitions do not

vary according to age.

1.3. Outline of the dissertation

The organization of this dissertation is as follows. Chapter 2 reviews relevant
literature. Chapter 3 presents the methods of data collection and analyses of the present
study. Chapter 4 offers analyses of RTs in formal conversation and Chapter 5, analyses of
RTs in informal conversation. Chapter 6 contrasts the ﬁndings of formal and informal
conversation, and discusses the types of interaction. Chapter 7 presents the perception
data of listening behavior and discusses the implications of the participants' comments to

the ﬁndings of the study. Chapter 8 offers the conclusions of the present study.

CHAPTER 2

REVIEW OF THE LITERATURE ON LISTENING BEHAVIOR
2.1. Introduction

Listening behavior consists of verbal and non-verbal behavior. In general, short
messages such as uh-hum, yeah, and nods used by a listener during a speaker's speaking
turn are considered listener responses. The types of utterances, frequency and contexts of
listener responses are determined by the rules of conversation, which every speaker of a
language has. Failure to exercise the rules of conversation causes misunderstanding of the
message or results in stereotyping of people who have that particular cultural background
(Lebra 1976; Tannen 1982).

The present study investigates the rules of listener responses in Japanese and the
social factors that affect listening behavior. In order to examine these issues, this chapter
will present major ﬁndings from related research on listening behavior.

The ﬁrst section reviews the history and ﬁndings of studies of listener responses
in English since listener responses are common conversational routine and those
behaviors in English have been investigated for many years. The following sections
present deﬁnitions, functions and research ﬁndings of Japanese listening behavior. Next,
the research ﬁndings about social factors, which account for the variation in listening
behavior in the present study will be reviewed. Findings from cross-cultural studies
which illustrate how routines of Japanese listening behavior differ from those in other
Languages will follow. At the end of this chapter, the results of my preliminary study are

presented.

2.2. Listening behavior in English

The study of listening behavior did not initially attract scholars’ attention
because the study of face-to-face interaction had focused on the speaker's utterances but
not the listener's role. Fries (1952: 49) analyzed yes, unh hunh, yeah, I see, good, oh, and
other utterances by listeners as 'signal free utterances' that can signal listeners' attention
while speakers are talking. Today these vocal forms are commonly known as 'back-
channels' that are originated by Yngve (1970). Yngve (1970) expands the interpretation
of back-channels to include longer expressions such as, 'you've started writing it then,
your dissertation?‘ (p.574).

The term, 'back-channel' was also used by Duncan (1974) who analyzed verbal
and non-verbal actions in video taped doctor-patient conversations and conversations
between two doctors. He identiﬁed three signals which inﬂuenced speaker-auditor
interaction during speaking turns; that is, a speaker within-turn signal, an auditor back-
channel signal, and a speaker continuation signal. 'Auditor back-channel signals' (p.166)
analyzed in the study were vocalic sounds such as m-hm, yeah, and right, sentence
completions, requests for clariﬁcation, brief restatements, and head nods and shakes.

The major goal of earlier studies was to understand how speakers and listeners
interact in face-to-face interaction, or how they manage speaking turns. As a result, their
concern in deﬁning back-channels was whether or not the listener's short utterances occur
during the turn of the speaker's. In that discipline, scholars broadened the deﬁnition of
back-channels by identifying requests for clariﬁcations, brief restatements, completions
of sentence by a listener (Duncan and Fiske 1977), as well as non-verbal behavior such as

gaze, hand gesture, head movement and postures (Kendon 1967; Dittman and Llewellyn

1968).

The functions of backchannels are deﬁned as expressing listeners' attention
(Fries 1952), signaling listeners' attention and understanding (Kendon 1967), and keeping
the conversation going smoothly (Dittman and Llewellyn 1968). Duncan (1974) also
identiﬁes back-channel functions as a signal to show that the auditor is following the
speaker's message or the auditor's acknowledgement toward what the speaker has said.

Schegloff (1982) investigated naturally occurring conversations and claimed that
uh-huh and yeah uttered by a listener and a listener's head nods primarily serve as a
continuer. White (1987) identiﬁed two major functions of backchannels: signaling the
listener's comprehension of the message the speaker has sent, and encouraging the
speaker to keep the ﬂoor. According to her, backchannels are multifunctional so that one
backchannel uttered by a listener can possibly have more than one function in the
context.

Although the early research on listener responses was conducted by
psychologists who were interested in listeners' not only verbal but also non-verbal
behavior, later studies of listener responses caught attention of eonversationalists, who
were interested in the turn-taking system (Sacks et a1. 1974; Goodwin 1979; Schegloff
1982; Clancy 1982). A characteristic of the conversationalists' notion to identify listers'
response is that 'turn' and 'ﬂoor' are the most important concepts for explaining
backchannels.

Ford and Thompson (1996: 136) state that a 'turn' is deﬁned as 'an interactionally
validated unit of talk.’ According to them, tum-taking (speaker changes) in conversation

consists of three kinds of turns; that is, full turns, backchannel turns, and laughter turns.

10

Backchannel turns are considered non-ﬂoor-taking turns since listeners do not claim the
speakership by sending backchannels; they simply encourage speakers to keep talking. In
this respect, Clancy et a1. (1996) include backchannels in 'reactive tokens' (RTs
thereafter), as well as other expressions that are uttered by listeners. They (1996: 136)
deﬁne a RT as "a short utterance produced by an interlocutor who is playing a listener's
role during the other interlocutor's speakership."

The goal of conversationalists is to uncover the projectability or predictability of
the turn end in natural conversation and what consists of turn units. Sacks et al. (1974)
deﬁned turn construction units (TCUs) as sentential, clausal, phrasal, and lexical
constructions. F 0rd and Thompson (1996) question the validity of a syntactic completion
to predict a turn completion. Ford and Thompson (1996: 137) claim that conversation
analysts should consider not only syntactic completions, but also, intonational
completions, and pragmatic completions, which involve speaker change caused by
backchannels as local pragmatic completions.

Ford and Thompson call such units identiﬁed by syntactic, intonational, and
pragmatic completions, 'Complex transition relevance places' (CTRPs) and claim that if
CTRPs are used, the projectability of the turn-taking mechanism will be improved (Ford
and Thompson 1996). Clancy et a1. (1996), following Ford and Thompson's notion of
CTRPs, investigated the occurrence of RTs at CTRPs in Japanese, English, and Chinese
and found that the rate that RTs occur at CTRP differs among three languages (Japanese
30.8%, English 45,1% and Chinese 79.5%). The notion of CTRT can account for the
different practices of listening behavior among languages as well as improving the

projectability of the turn-taking system.

11

2.3. Listening behavior in Japanese

For Japanese people, listening behavior has been a common topic regarding the use
of the language even among ordinary people, as pointed out by Okamoto (2002).
However early studies of Japanese listening behavior were limited to investigating verbal
responses, and were not based on empirical data (Ikegami 1952; Miyaji 1959).

Studies of Japanese listening behavior have attracted researcher's attention from the
language pedagogy perspectives, as well, for over two decades. Mizutani (1984) states
that for non-native speakers of Japanese, backchanneling is a difficult strategy even
though it is a necessary device for smooth communication. Matsuda (1988) and Nishihara
(1995), among others, make similar comments about the acquisition of backchannels by
learners of Japanese. The purpose of these studies was to describe the various functions
and the use of backchannels, and to provide evidence for its frequent use in Japanese
conversation. However, much of the data used in the previous studies were not natural
conversation. Neither did they include non-verbal features, although the importance of
non-verbal features, especially nods, has been discussed widely (Mizutani 1984; 030
1988; Kita 1996; Horiguchi 1997; Uyeno 1999).

Maynard's study (1989) was the ﬁrst to use 20 dyadic natural conversations of same
sex participants to analyze both verbal and non-verbal backchannels. She found that
listeners' backchannels provide important information for speakers to "self-contextualize"
during the conversation.

Szatrowsky (1993) investigated the structure of Japanese invitation discourse using
naturally-occurring telephone conversations as her data. She examined how listeners used

backchannels when they were invited to do something on the phone in various contexts.

12

She presents qualitative analyses, which show that backchannels are a multifunctional
linguistic device for the speech act of invitation in Japanese.

Studies of non-verbal backchannels are even scarcer. Maynard (1987) analyzed the
frequency and functions of speaker and listener nods and found that speakers' and
listeners' nods are used emphasize the message, to show clause boundaries, and to signal
tum-end or turn-claim. Sugito (1989) investigated the relationship between nods and
verbal backchannels. The ﬁndings shows that the frequency of use of nods that don't
accompany verbal backchannels varies among individuals (from 3 to 49 times) but that
the ratio of nods to verbal backchannels does not vary (76% to 86%).

Szatrouski (2000, 2003) examined the relationship among nods, gaze and verbal
backchannels and found that those behaviors are interrelated. For example, the
"addressed recipient was most likely to respond with an aizuti plus a head nod(s) when
the speaker gazed directly at her and nodded" (p. 287). Studies on variations of listening
behavior associated with social factors such as age, gender and styles are also very few,
with the exception of Kogure (2003), Okamoto & Sato (1992), Okamoto (1995), and

Nakamura (1998).

2.3.1. Deﬁnitions

Japanese backchannels are often referred to as aizuchi in the related literature
(Locastro 1987; Szatrousky 2002, 2003; Okamoto to appear). Aizuchi literally consists of
ai 'partner's' and tsuchi 'hammer.'Aizuchi originally referred to two blacksmiths
hammering a sword one after the other, according to Kita (1996). It also refers to a

listener's verbal messages sent during a speaker turn without intending to take a ﬂoor.

13

Aizuchi, therefore, is often characterized as non-ﬂoor-taking utterances and nods. In the
present study, "backchannels" and "aizuchi" may be used interchangeably.

Below are the deﬁnitions of Japanese backchannels, which vary from scholar to
scholar in terms of whether they include various short utterances in backchannels.
Mizutani (1984) deﬁnes aizuchi as listeners' responses such as hai 'yes,' ee 'yeah,' haa
'yes,' un 'yeah,' hun 'uh-hum,’ naruhodo 'I see,‘ soudesuka 'Is that so,’ soudeshoone 'that is
right,’ laughter, yapparine 'that is what I thought,’ soryasoo 'that is right,’ sorewaieru 'you
can say that,‘ nee 'you know,’ sentence completions, and reinforcements. According to
Mizutani, aizuchi are used to "keep the conversation going smoothly" (p.267) [aite no
hanashi no shinko o unagasu tame m' tsukawareru].

Horiguchi (1988) classiﬁed backchannel expressions into different groups of
aizuchi-ski (backchannel group) according to the degree to which they are
conventionalized. In her classiﬁcation, aizuchi-shi includes hai 'yes,' ee 'yeah,' naruhodo
'I see,‘ sodesune 'that is right,‘ un 'yeah,' honto 'really,' repetitions, and sentence
completions, paraphrases and clariﬁcations.

Sugito (1989) deﬁnes aizuchi syntactically depending on whether or not the
utterances carry substantives such as verbs and nouns, and also semantically depending
on whether or not they function as a question, request, etc. Sugito includes repetitions,
clariﬁcations, exclamations, and laughter in aizuchi if they satisfy the above deﬁnition.

To sum up, in a narrow sense, aizuchi consists of two types of expressions: one
comprised of non-lexical vocalic expressions such as an 'yeah,' ee 'yeah,' aa 'yeah,' hee
'oh', hoo ' wow,’ and haa 'yes,' which do not have propositional meanings; the other type

are lexical items such as hai 'yes,' 300 'it is so,’ naruhodo 'I see,’ which are

14

conventionalized as aizuchi when uttered during the interlocutor's turn. Other expressions
such as repetitions, clariﬁcations, paraphrases, sentence completions, laughter, and nods
are also deﬁned as aizuchi in a broader sense from their functions in the interaction.

The problem in deﬁning listener responses in Japanese is that there is no single
criterion--except for Sugito (l989)--that deﬁnes them. In order to provide a standard of
listener responses, careful investigation of the natural conversation of people of different

ages and genders in various situations is necessary.

2.3.2. Functions of backchannels and nods
Japanese backchannels are considered multifunctional; that is, they ﬁmction in
many ways in given contexts. These functions include discoursal, interactional, and social

functions, to be explained below.

2.3.2.1. Discoursal and interactional functions

Horiguchi (1988: 14 -16) points out ﬁve functions of backchannels: to signal (1)
listeners are listening, (2) understanding, (3) agreement, (4) negation, and (5) expressing
emotion. Matsuda (1988: 62-3) subdivides Horiguchi's category (1988) further into
twelve functions to signal: (1) that listeners are listening, (2) that listeners are following
what the speaker says, (3) acknowledgement, (4) that listeners are understanding what the
speaker is feeling, (5) that listeners are recalling the listener's own experience, (6)
agreement, (7) empathy, (8) vague agreement, (9) negative feelings or doubt, (10) strong
emotion, (11) interest, (12) that listeners are ﬁlling silence. She points out that the

intonation when a backchannel is uttered--rising, falling, or ﬂat--is important in

15

differentiating the function of backchannels. Matsuda claims that Japanese backchannels
are multifunctional. That is, one form can index listeners' emotional states (e.g.,
empathetic or negative feeling etc.) toward speakers' utterances by changing the
intonation at the same time that they send the message in response to the content speakers
send (e.g. agree or understand, etc.).

Maynard (1989) classiﬁcation is similar to Horiguchi’s (1988) except for the last
two categories: (1) continucr, (2) understanding, (3) agreement, (4) strong emotional
response, (5) support toward the speakers' judgement, and (6) minor addition, correction,
or request for clariﬁcation. Function (5) support toward the speakers' judgement, could be
vocalic forms such as hai 'yes' or repetitions of part of the speaker’s utterance. Function
(6) is actually included in Horiguchi's (1988) deﬁnition of backchannels as a form of
clariﬁcation and paraphrasing. Furthermore, Kita (1996)'s category includes a rhythm
taking function.

Although the terms used for functions are not the same, the functions of
Japanese backchannels listed above overlap one another. Compared with the main
function of English backchannels as continuers (Schegloff 1982), Japanese backchannels
are characterized as part of social functions, such as 'phatic communion' (Richards 1982)
expressing emotion or attitudinal stance toward the speakers' utterance, as well as
interactional ﬁmctions.

Functions of nods were investigated by Maynards (1987). She observed head
movements of listeners and speakers and classiﬁed their ﬁmctions into 5 categories: (1)
affirmation, (2) claim for turn-end and turn-transitions, (3) pre-turn and turn-claim, (4)

turn-transition period ﬁllers, (5) backchannel, and (6) rhythm taking. Listeners' nods

16

serve the functions (1), (3), (4), (5), and (6).

As we can see, the functions of listeners' nods are the same as backchannels.
Maynard states (p. 590) that "nods serve only secondary and sometimes redundant
functions, but their pervasiveness and the prominent place in actual interaction deserves
our attention." Maynard's analysis (1987) doesn't distinguish nods that accompany verbal
backchannels or utterances and nods that don't accompany verbal utterances. According
to Maynard, nods are secondary to verbal backchannels and since their functions are the
same as verbal backchannels, nods co-occurring with verbal backchannels are redundant.
In other words, nods that don't accompany verbal backchannels function as backchannels

by themselves.

2.3.2.2. Social functions

The way people interact always carries social meanings according to the context.
Both speakers and listeners send messages providing information that help each other
know how to interpret social meanings sent to one another. Listeners are expected to use
backchannels as a linguistic strategy in order to behave appropriately in social contexts.
This section examines two social functions that various listening behaviors can serve in

face-to-face encounters: politeness and solidarity.

Politeness
Brown and Levinson (1987) explain politeness as a strategy that saves one's
positive face (i.e. want to be understood) or negative face (i.e. want not to be disturbed).

In Brown and Levinson's politeness theory, the same linguistic forms may be used as

17

either a positive or negative strategy depending on the context. According to LoCastro
(1987: 112), " back-channeling is addressing the Japanese need to preserve positive face."
Frequent use of aizuchi is a conversational routine and it fulﬁls the cultural expectations
of politeness by avoiding confrontation and accomplishing smooth conversation. As
LoCastro states, the use of backchannels serves as a positive politeness strategy that saves
one's positive face by showing the listener's supportive attitude. However, when we
consider listener responses including nods, nods can possibly work both as a positive or a
negative politeness strategy depending on the situation.

For example, listeners can choose either verbal responses or nods to show that
they are paying attention, for instance. If they choose verbal backchannels, the listener’s
behavior works as a positive strategy by saving positive face. If they choose nods, since
nods are less intrusive and indirect, their behavior saves the speaker’s negative face.
Social factors, such as age, status, sex, and psychological distance between interlocutors
inﬂuence the listener's choice of strategies.

In Japanese culture, politeness is characterized as being less intrusive and
indirect (Ide 1982). The listener's nods can save a speaker's negative face in a context
where 'discemment' (Hill et a1. 1986) is required. Establishing rapport by using verbal
backchannels, on the other hand, is a positive politeness strategy. We will return to this

issue in Chapter 4.

Lakoff (1975) claims that politeness is determined by three factors: 'distance,’
'deference,' and 'camaraderie' The three factors form the rules of politeness in a language.
Politeness issues in Japanese woman's language are often explained in terms of

'deference' (e.g. Ide 1982, 1991). However, the rules of 'camaraderie' are as strong as

18

deference in the 'intimate style' (Tannen 1993a; Lauwerenyns 2000). As Ochs (1992)
states, politeness issues should take into account a range of situational parameters, such
as the speech activity, and the speaker-hearer relationship. Variations in Japanese

listening behavior also need to be explained taking situational parameters into account.

Solidarity

Solidarity motivates certain linguistic variations in terms of the particular
linguistic forms or interaction. Lauwerenyns (2000) examined the use of hedges by
young Japanese speakers and found that some forms of hedges are used to promote
solidarity among speakers. Okamoto (1995) observed that solidarity is the major
motivation for young female speakers to speak like men using masculine forms, and that
they choose not to use honoriﬁcs to show solidarity (Okamoto 1998, 2002).

Ide (1982) explains that people can form solidarity not only by sharing major
factors such as a common cultural, social, or geographical background, but also the same
interests or responsibilities. Ide's notion about how to express solidarity has the
implication that linguistic forms (e.g. certain forms of backchannels) or styles of
interaction (e. g. listening behavior) can be the means for sharing among people who wish
to establish solidarity. People of different age groups possibly exercise different listening

behavior in in-group and out-group conversation.
2.4. Cross-linguistic studies

Cross-linguistic studies of backchannels found that backchannels are used more

frequently in Japanese than in other languages, such as Chinese, American English, and

19

Korean (Locastro 1987; Maynard 1989; Yim 1995; Clancy et a1. 1996).

Maynard (1986, 1989) investigated the use of English and Japanese backchannels
and found that backchannels in Japanese and English have similar functions but that
Japanese use backchannels and nods three times more frequently than do Americans.

Yamada (1992) examined backchannels in intracultural and intercultural
communication in English conversations. She found that Japanese use about 2 times as
frequent backchannels as do Americans in intracultural interaction. Yamada's ﬁndings
reveal how one's ﬁrst language interactional style transfers into a second language.

As reviewed in an earlier section, Clancy et a1. (1996) examined Japanese, Chinese
(Mandarin), and American English data. They found that the three languages differ in the
percentage of speaker changes that are RTs. Japanese showed the highest percentage of
RTs (3 9%) compared with American English (37%) and Mandarin (10%).

The different results between Clancy et al. and Maynard can be explained by
methodological differences between the two studies in defming RTs and backchannels. In
Clancy et al. (1996), RTs included resumptive openers, etc., which Maynard did not
include. If we only compare backchannels, the percentage of RTs that are backchannels
used by all speakers is 68.3% in Japanese, while it is 37.9% in English. This is closer to
Maynard's ﬁndings.

The most important ﬁnding of Clancy et al.'s study, however, is the difference in
the location of RTs in Japanese and English; their analysis revealed that Americans place
45% of their RTs at CTRPs and 78% at grammatical completion points, while Japanese

Place only 30% of their RTs at CTRPs, and 36% at grammatical completion points. In

other words, Japanese listeners use RTs while the primary speaker is talking rather than

20

waiting for completion points. The results show that Japanese listeners overlap or talk
along with the speaker more often than do American English or Chinese listeners. Hinds
(1982) also points out that there is a different type of overlap between Japanese and
English conversational interaction--in Japanese, overlaps occur far more frequently and
involve longer stretches.

Although Clancy, et a1. point out that this practice of RTs by Japanese may be
interpreted to be destructive to the primary speaker, it reﬂects the cultural concern for
harmony and co-operation. Clancy, et a1. state that "the non-primary speakers are
expected to show concern for the primary speaker's sense of security in holding the ﬂoor;
an empathetic conversational partner provides this sense of security by giving RTs during
the primary speaker's turn.

Although Clancy, et al. (1996) analyze only verbal RTs and not non-verbal
features, they provide a comprehensive description of differences in interactional practice
in casual conversations among Japanese, Mandarin, and English speakers. They state (p
380) that "the frequency, types, and placement of RTs are part of an interactional system
which competent language users know and which gives rise to clear cultural expectations

about what speakers and listeners are doing in ordinary talk."

2.4.1. Cross-linguistic differences in non-verbal listener responses

Hinds (1978, 1982) claims that in Japanese conversations interactants are required to
be sensitive to non-verbal cues, such as eye-contact or head nodding, which may signal
the turn completion or other messages without overt utterances. In Hind's view, nodding

has the same function as verbal backchannels. Although he does not clearly distinguish

21

nods that accompany verbal utterance from those that do not, nods are used along with
verbal utterances. Examining the co-occurrence of verbal and non-verbal backchannels of
Japanese listeners, White (1987) states that there is a strong correspondence of about 85%
between vocal responses and head nods.

Maynard (1987) found Japanese listeners use nods two times as often as do
Americans. There are few studies that have focused on cross-linguistic differences in

non-verbal listening behavior.

2.4.2. Culture and interactional style

Goffman (1986) states that people within a society share 'frames', that is,
principles of organization which govern events. However, it is known that even within a
language society, people from different ethnic groups use different kinds of linguistic
strategies which may be misunderstood by those who don't share the subculture (e.g.
Tannen 1984).

Tannen (1984) examined New York Jewish people's conversational style, and
reports that they tend to ask short, quick, and latched questions. She calls this style
"machine gun talk," and claims that it exempliﬁes how a culture - speciﬁc style may
cause a negative impression and misunderstanding. She also found that Greek listeners
overlap speakers' utterances more frequently than do American English speakers. She
states that even though both styles result from showing the listener's interest or
enthusiastic attitude, their behavior tends to be interpreted as obstructive or aggressive,
and lacking in patience or interest. As White (1987 :17) states, "when conventions for use

and interpretation of linguistic phenomenon are not shared, stereotyping may well occur."

22

Hinds (1978, 1982) made similar comments about Japanese simultaneous talk,
i.e., that overlapping is not rude in Japanese interactional style and that it is used for
maintaining harmony between participants. According to Locastro (1987), English native
speakers living in Japan for several years adopt the Japanese interactional style and use
backchannels and nods frequently when speaking English.

Lebra (1976: 39) states that "since a Japanese listener in an English conversation
is likely to make more interjections, the American speaker would take such interjections
as a sign of the listener's impatience and demand for a quick completion of the
statement." White (1987) investigated the validity of this claim by Lebra (1976), i.e., the
negative reactions to Japanese interactional style used by native speakers of Japanese
when speaking in English. She found that there was no negative attribution, such as
impatience or intrusiveness on the speakers' character caused by the interactional style.

The acquisition of appropriate backchanneling behavior by learners of Japanese
is considered important for smooth communication and to avoid being misunderstood
(Mizutani 1984; Locastro 1987; Matsuda 1988; Szatrowski 1993; Nishihara 1995;
Horiguchi 1997; Uyeno 1999). The negative evaluations resulting from leamers'
insufficient use of backchannels pointed out in the previous studies, however, are based
on the researchers' speculations or impressions. It is necessary to investigate people's
perception of native and non-native speakers' different listening behavior, using actual
data.

Most of the past studies agree that Japanese interactional style is harmonious,

empathetic, and cooperative in nature reﬂecting cultural values (Lebra 1976; Hinds 1982;

Yamada 1992, 1997; Watanabc 1993; and others). Some scholars, however, argue that

23

these are stereotyped ideological language views (Okamoto 2002).

Sato and Okamoto (1998) state that the simple stereotyped cultural view that
Japanese listeners show cooperative attitudes toward the interlocutors, may not reﬂect
actual language practice. Okamoto and Sato (to appear) analyzed 4 mother-daughter and
4 daughter-friend casual conversations and found that participants in family discourse do
not use backchannels very frequently, while the same participants in conversations with
friends use them frequently. They show that the stereotypical cultural view cannot
account for these results since actual language use is more complex and dynamic. Among
the implications of their study are that careful observations of listening behavior in
various contexts taking social variables of the participants into consideration are needed

to account for real language practice.

2.5. Social variables
This section reviews the effects of social variables such as sex, age, and styles in

Japanese listening behavior. There are few empirical studies of this type.

2.5.1. Sex

In general, researchers agree that women's speech is characterized by a tendency
to be more polite, gentle and indirect than men's (Trudgill 1972; Lakoff 1975; Ide 1982,
1991; Okamoto 1995). Women's speech is thus expected to be more cooperative and
empathetic than is men's speech. Tannen (1990) claims that men's language is
information-centered, while women's language is emotion- or rapport-centered. Such

beliefs regarding gender and linguistic norms in society are common, according to

24

Okamoto (2002). If such beliefs are true, men and women are expected to perform
different listener roles using different kinds of backchannels and with a different
frequency.

Although studies of gender differences in listening behavior are limited in
number, there are two studies that report that women use more backchannels than do men
(Ide 1979; Kurosaki 1987; Maynard 1989). Maynard (1989: 168) examined 40 Japanese
native speakers who were born and raised in Tokyo (19-23 years old) and found that
female participants use backchannels more frequently (401 times) than do male
participants (302 times), and male participants have more frequent tum-exchanges (439
times) than female participants (281 times). These ﬁndings show that there is a difference
between men and women in the frequency of backchannels and turn exchanges.

Kogure (2003) investigated gender difference in the use of backchannels and
nods by 5 single-sex male, 5 single-sex female and 5 mixed-sex dyadic groups whose
ages range between 18 and 25. He found that overall frequency of verbal backchannels
was similar among all groups. The most frequently used was backchannels, and un,
'yeah,' was the most commonly used backchannels among all groups. However, there was
a difference in the use of nods that both accompany verbal backchannels and do not.
Female single-sex dyadic groups use non-verbal behavior most frequently compared with
other groups.

The use of backchannels and nods by male participants does not vary in single-
sex and mixed-sex groups, whereas female participants use backchannels and nods
differently in the two contexts. According to Kogure, female listeners accommodate their

listening behavior in the mixed-sex conversation. That is, they use more reactive

25

expressions, such as 500, 'right,’ fewer evaluative expressions such as honto, 'really,' and
fewer nods that both accompany verbal backchannels and don't to convert their listening
behavior into a pattern of male interlocutors in the mixed-sex conversation more than do
male participants.

Kogure's ﬁndings suggest that there is little difference in terms of the overall
frequency and types of verbal backchannels in the conversation between young male and
female acquaintances. If this is the case, his ﬁndings support Okamoto's claim (1995).

Okamoto (1995) claims that the difference between women's speech and men's
speech is narrowing. Okamoto and Sato (1992), and Okamoto (1995) report that young
Japanese female speakers use "less feminine" language than in the past. These studies
show that young female speakers use more masculine forms at the end of sentences as
well as other masculine expressions in their speech. There may actually be, therefore,

little difference between young women's and men's interactional styles.

2.5.2. Age
Age is one of the major social factors that inﬂuence the way people speak in
Japanese, as in other languages. The relationships between linguistic forms and age or
generation have been studied in terms of the use of keigo 'Japanese honoriﬁcs' (Yoshioka
1996), discourse markers (Phillips 1998), and hedges (Lauwereyns 2000).
Japanese has a highly elaborate honoriﬁc system, or keigo. Since the acquisition
of proper keigo use is developmental, young native speakers often have not mastered it.
Yoshioka (1996) found that high school students had difﬁculty using keigo so that they

consciously avoided it when they were interviewed by the researcher. For example,

26

instead of answering the researcher directly, they talked to their peer on purpose. Follow
up interview showed that these students recognized the importance of keigo in Japanese
society and the necessity of mastering it in the future. Yoshioka's ﬁnding that Japanese
native speakers' keigo use varies according to age suggests that younger people's listening
behavior may differ from that of older people's keigo use.

Lauwereyns (2000) found that high school students use more hedges than
people who are over sixty years old, and female high school students use hedges most
frequently among other groups. She claims that the uses of a wide variety of hedges by
young female speakers show not only their vagueness, and non-committal or self-
protective attitude, but also their attempt to establish solidarity among the users, or show
their identity in society.

Young Japanese female speakers use more assertive, direct expressions, speak
fast, and use masculine forms, compared to traditional Japanese women (Asahi Shinbun
1993 cited Okamoto 1995). These characteristics of young female speakers seem to
contradict their use of hedges. The purpose of the using hedges is to talk like others in the
"group," but to sound 'feminine.’ The language spoken by contemporary young women
does not represent what is often called 'Japanese women's language.‘

Few studies have focused on the age factor on the listening behavior aside from
Sato and Okamoto (1999) and Okamoto and Sato (to appear). Gender and age are

important connecting factors to be considered on the present study.

2.5.3. Styles of conversation

Styles in conversation are deﬁned as speech genres representing different levels

27

of formality. According to Joos (1967), there are ﬁve kinds of 'style': intimate, casual,
consultative, formal, and frozen. These ﬁve styles compose a continuum. The 'intimate
style' is the most informal type, such as conversation between equals who know each
other very well, while the 'frozen style' is at the opposite end of formality continuum.

Japanese listening behavior in intimate, consultative, and formal styles have
been studied in conversations between friends (Maynard 1989; Ide 1979), TV shows
(Mizutani 1984; Horiguchi 1988; Furo 2002), and telephone conversations (Szatorowski
1993). However, there are a limited number of studies that focus on the difference in
listening behavior among different styles.

Okamoto and Sato (to appear) examined the verbal reactive tokens appearing in
three 15-minute conversations each between mother and daughter, and between friends.
They found that young participants (daughters) used RTs much less frequently in family
discourse (total of 128 in 60 minutes) than in friends’ discourse (total of 638 in 60
minutes).

Okamoto and Sato’s study shows that such a stereotyped view that Japanese
listening behavior is harmonious, supportive and co-operative cannot account for
variations in real language use. The results suggest 'speech genre' may determine
interactional styles, including listener behavior, which may be different from the
stereotype.

Nakamura (1998) also shows that conversational style inﬂuences the
interactional style. She found that Japanese listeners use more frequent backchannels
When they are interviewed than when they chat with friends using sentence completion or

other types of utterances.

28

2.6. Preliminary study

This section presents my quantitative analysis of verbal backchannels and nods
used by three female native speakers of Japanese (Miyazaki 2001b). It was found that
there is an interrelationship between listeners' verbal backchannels and nods. The results

also suggest that age is a factor in listening behavior.

2.6.1. Hypotheses and method
The preliminary study tested the following hypotheses:
Hypothesis 1: The ratio of verbal to non-verbal backchannels varies among
individuals, but the total number may be similar.

Hypothesis 2: The frequency of verbal and non-verbal backchannels is affected by
age.

The data consist of three dyadic videotaped face-to-face conversations (90-
minute total). There were four participants, including the researcher. I participated in the
conversations as primary speaker S. The participants were all female native speakers of
Japanese who grew up in Japan. Table 1 shows the demographic data for each participant.
The relationship between primary speaker S and A, B, and C is also shown.

Table 2.1. List of participants in the preliminary study

 

 

 

 

 

 

sex age occupation Relationship ; Place where they
between S and others i grew up

S (speaker) F 35 student Yamaguchi

A (listener) F 25 student Go to the same university Hiroshima

B (listener) F 30 housewife Husbands work at the Saitama ..___ _-._.
same company ........

C (listener) F 37 housewife Husbands work at the Kagoshima
L same company

 

 

 

 

 

 

 

29

A ﬁfteen-minute each conversation for analysis where S was the primary speaker
and the other participants, the listener. The topic of the segments was the researcher's own
childbirth experience. Although the story was not pre-scripted, each participant listened
to a similar story. The number of verbal backchannels and nods was counted and
categorized based on the types and the contexts in which they occurred.

In this preliminary study, verbal backchannels are deﬁned as follows: short
utterances and nods produced by the listener with no intention of taking the turn. The
verbal and non-verbal backchannels analyzed are of three types. The ﬁrst type are
backchannels, or non-lexical vocalic forms such as rm 'yeah,’ aa 'mmhm,’ and brief
messages such as 500 ‘I see,’ and honto ‘really’ which occurred during the speaker's turn
and had no referential meaning. The second type are non-lexical elements, such as waa
‘wow’. The third type are head nods. Only vertical head movements occurred during the
speaker’s turn, and they were interpreted as non-verbal ‘continuers.’ Verbal backchannels
and nods were often repeated in succession, which were regarded as one occurrence.

I consider verbal backchannels to be primary backchannels and nods, secondary.
Nods which accompany verbal backchannels, therefore, were not counted as nods. The
number of verbal backchannels accompanied by nods will be explained in the next

section.

2-6.2. Results
Figure 2.1 shows the total number of verbal tokens used by each of the three
participants. The total numbers of verbal backchannels uttered by participants A, B and C

Were 126, 120, and 63, respectively. While participants A and B used almost the same

30

number of verbal backchannels during the 15 minutes (126 and 120 tokens). Participant C

used verbal backchannels signiﬁcantly less (63 tokens) than the other two.

Figure _2_. 1. Frequency of verbal backchannels

 

 

140 ii . --." ---.....-.--..-. .---.-L.MM.-.___ .. _ .-.--- . .... . . . - “-....- ...—

 

 

 

120

100 -~—— r—————-» tw—~—~—-~—w——-~—*—

 

80

 

 

 

60 126 120

 

 

 

 

 

 

4O

 

 

63
20 -____ ~va~v~wﬁrﬁi .7 “L _mw- —J -4

Number of verbal backchannels

 

 

 

 

 

 

 

 

O

 

A B C

Participants

 

 

 

Figure 2.2. Frequency of nods

I 20 .E.. _.

 

 

 

100 _L__ HHHW-” -. - ._ ___.. _ _.W___M .___._ .__ _- _. a

 

80 . a“. ---______ __ - -. _____‘____,______,,,_,__ .—

 

 

98

 

 

 

 

 

Number of nods
8

59
20 . 45 -_._ ..---. _. _. _ ____.

 

 

 

 

 

 

 

 

0 . s
A B C

Participants

 

 

 

 

Figure 2.2 shows the frequency of nods used by each participant. There is a
negative correlation between the frequency of verbal backchannels and nods [n=3, r=-.98,
p=n.s.]. Participant A nodded 45 times, participant B, 59 times, and C, 98 times.

Figure 2.2. Frequency of nods

31

Table 2.2 display the total number of verbal backchannels accompanied by nods,

verbal backchannels without nods, and nods without verbal backchannels.

Table 2.2. Frequency of verbal backchannels and nods

 

 

 

 

 

 

 

 

 

A B C

1. Verbal backchannels that are not accompanied by nods 66 23 10
2. Verbal backchannels accompanied by nods 60 97 53
3. Nods that do not accompany verbal backchannels 45 59 98
Total 171 179 161

 

 

Figure 2.3. Frequency of verbal backchannels and nods

 

180
I60
140
120
100 -----

Number of listener responses

o8‘588

 

Participants

 

 

 

I" DTbaclcc hanne l {,6}; 08—5“

 

To examine how verbal backchannels and nods interrelate in each participant’s
listening behavior, I combined Figures 2.1 and 2.2 into Figure 2.3. Figure 2.3 shows that
although each participant use verbal and non-verbal backchannels differently, the total
numbers of backchannels is not statistically different. In other words, the primary speaker

received almost the same number of reactions from each participant.

32

2.6.3. Conclusions

The preliminary study found that nods may substitute for verbal backchannels
and that verbal backchannels and nods interrelate. In each of the 15-minute segments, the
primary speaker received the total of 171, 179, and 161 listener responses, respectively.
These results indicate that when a listener participates in a conversation with fewer verbal
backchannels, she nods more frequently, and vice versa. This ﬁnding is signiﬁcant in that
it shows how verbal backchannels and nods are interrelated.

The preliminary study demonstrates that listener responses are not random, in
terms of the total frequency. It does not show, however, whether social factors inﬂuence
listening patterns. Factors such as age, the relationship between the speaker and the
listener, and conversation types, (formal or casual), and regional variations should also be

considered.

2.7. Summary

This chapter presented an overview of previous studies on listening behavior in
Japanese. It is reasonable to assume that the on-going changes in Japanese women's
language inﬂuence their interactional styles.

In the following chapter, I will posit four hypotheses based on my preliminary
studies and present the methods of data collection, analyses and the deﬁnitions of listener

responses used in the present study.

33

CHAPTER 3

HYPOTHESES AND METHODOLOGY
3.1. Introduction
Chapter 2 reviewed previous studies of listening behavior in Japanese and
related areas. The ﬁrst section of Chapter 3 presents four hypotheses based on the
ﬁndings of the previous studies and explains each of the hypotheses. The remainder of

the chapter describes the methodology of the present study.

3.2. Hypotheses
The present study investigates the following four hypotheses.
Hypothesis 1: The ratio of verbal to non-verbal RTs may vary among individuals,
but the total number will be similar in similar contexts.

Hypothesis 2: The frequency of verbal and non-verbal RTs is affected by age. Older
female speakers use more verbal RTs than do younger speakers.

Hypothesis 3: The formality level of conversation inﬂuences not only the choice of
RT types and frequency, but also the interactional styles.

Hypothesis 4: The perception of other's listening behavior varies according to
the perceiver's age.
Hypothesis 1: The ratio of verbal to non-verbal RTs varies among individuals, but
the total number will be similar in similar contexts.
Speakers of a language know what to do and how to do it in the face-to-face
interaction in various contexts and share the same expectation among speakers that tells

them how to react by using linguistic strategies as listeners (White 1989).

Maynard (1989) states that the occurrence of the listeners' backchannels and

34

nods in Japanese are not random and that they are predictable and governed by

'conversational rule' (Goffman 1986). She also claims that it is important for harmonious,

cooperative and empathetic communication.

If native speakers of Japanese in the same generation are given similar social
contexts in which they have the same interlocutor, the same topic, the same goal of
communication, and similar degree of acquaintance, they may show similar listening
behavior appropriate in the contexts.

Although studies of conversation often limit their scope to verbal features,
listening behavior should also be investigated, including non-verbal features such as
nods. In Japanese face-to-face conversation, nods are very pervasive and function as
substitutes for verbal backchannels, according to Maynard (1986, 1989) and Sugito
(1989). Head nods are important features in analyzing the listener’s role in Japanese
interaction.

The ﬁndings of my preliminary study (Miyazaki 2001b) that analyzes three
female participants' (25 to 35 years old) listening behavior show that the total of their
backchannels and nods does not signiﬁcantly differ, but that the ratio of verbal
backchannels and nods varies. Frequent users of verbal backchannels use fewer nods and
vise versa.

Hypothesis 2: The frequency of verbal and non-verbal RTs is affected by age.
Older female speakers use more verbal RTs than do younger
speakers.

Hypothesis 2 is based on the following research. 'Japanese women's language' is

characterized by being less assertive, more indirect, and polite, and by such attributes as

35

use of sentence ﬁnal particles, feminine forms, high pitch, slow speed and so on (Ide
1982; Raynold 1985; McGloin 1990; Shibarnoto 1987; Cook 1990, 1992). It is reasonable
to assume that Japanese women's interactional style reﬂects these characteristics. In other
words, Japanese women are expected to listen to others supportively and cooperatively
by frequently sending backchannels and nods and to use backchannels of feminine forms,
if available.

However, it has been observed that young female speakers speak 'less feminine'
language than do older speakers and that the differences between young male and female
speech are narrowing (Sibamoto 1985; Okamoto 1995). If this is the case, there may be a
large gap between the listening behavior of the younger generation and that of the older
generation especially in the choice of the types of expressions and frequency.

Hypothesis 3: The formality level of conversation influences not only the choice of
RT types and frequency, but also the interactional styles.

The styles of conversation, formal or casual, in hypothesis 3 are differentiated by
the types of speech events that listeners are engaged in. The degree of formality is
determined by the relationship between the participants, or other situational factors. In the
present study the style difference will be determined by the purpose of conversation, i.e.,
receiving instructions or chatting, by the psychological distance between the participants,
but not any particular linguistic forms used in the conversations. As pointed out
elsewhere (Cook 1999; Okamoto 1998), linguistic forms such as the formal (desu/masu)
forms and informal (da/ru) forms are not necessarily restricted to formal or informal
contexts. Rather, it is common to mix both forms within the same speech event to reﬂect

the speaker's emotional or psychological state.

36

Therefore, in the present study, listening to instructions from a stranger or
someone whom the participant doesn't know well is considered to be "formal"
conversation, and chatting with close friends of the same generation, "informal"
conversation. Both types of conversation are between two same-sex participants.

Previous studies on only verbal backchannels show that the styles of
conversation inﬂuence listening behavior. Sato and Okamoto (1998), for example, report
that young participants use more frequent verbal backchannels when chatting with their
close friends than with their mothers, and Nakamura (1998) found that backchannels
were more frequent in interviews than in casual conversation.

I hypothesize that style (formal or informal) inﬂuences listening behavior in the
following way. An informal conversation by same generation presents features that
express solidarity rather than politeness. It possibly contains more verbal responses and
less nods than does a formal conversation. In formal situations, participants try to politely
achieve the goal (i.e. obtaining information) by avoiding face threatening acts (FTA). In
this context, participants do not necessarily have to express empathy or establish rapport
with a stranger, unlike in conversations with a close friend. Thcrfore, it may contain less
verbal RTs and less friendly interaction.

Furthermore, the types of backchannels used during the conversations will be
different. That is, less formal forms of RTs in informal conversation, whereas, more
formal forms or more polite forms in formal conversation.

Hypothesis 4: The perception of others' listening behavior varies according to the
perceiver's age.

Finally, hypothesis four is based on my perception study on male and female

37

native and non-native speakers of Japanese (Miyazaki 2001a). Although the data were
limited in terms of participants' age (the oldest was in the early 30's, with most
participants in their 20's), there was a relatively clear difference in rating of a particular
behavior between the sexes. Based on this, I hypothesize that the perceptions of listening
behavior are inﬂuenced by other social factors such as age. The present study investigates
how the perception differs according to age and whether or not perception reﬂects an

individual’s interaction style.

3.3. Method
This section ﬁrst describes how the data were collected and analyzed and then

explains what are regarded as listener responses by illustrating with examples from the
actual data.
3.3.1. Data

The data for the present study consist of the following.

1 Thirty video-taped face-to-face conversations in which participants listen to
instructions given to them by the researcher (2 minutes each)

2. Fifteen video-taped face-to-face conversations in which 30 participants chat
with their close friends (15 minutes each)

3. The ratings of 5 types of listening behavior in the video clips by 30 participants

4. Fifteen audio-taped interviews about Japanese listening behavior by each
participant.

3.3.2. Participants

Participants in the present study are 30 female native speakers of Japanese, age

19 to 61. They were recruited to satisfy the following conditions. First condition is that

38

participatns were born and raised in Tokyo or in the neighboring prefectures (Saitama,
Chiba, or Kanagawa) and have not lived in other areas of Japan. Second, they are
monolingual native speakers of Japanese. If they have lived overseas, their stay must be
less than one year in total. Finally, each of the 15 women brings a close friend who has
the same characteristics listed above to carry out friendly conversation.

The ﬁrst condition eliminates the inﬂuence of dialect variations in their verbal
and non-verbal behavior as a listener. The second condition minimizes other cultural or
language inﬂuences on their listening behavior.

The participants are divided into three age groups as shown in table3.1. Group I
participants are 8 undergraduate and 2 graduate students, while Group II and 111
participants consist of housewives, teachers and ofﬁce workers. The age of participants in
each group ranges from 19 to 23 in Group I, 28 to 40 in Group II, and 45 to 61 in Group

111.

Table 3.1. List of participants in formal conversation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Group ID Age Hometown Occupation Conversation
partner
1 I YG 19 Kanagawa student OS
2 I OS 20 Kanagawa student YG
3 I TM 21 Chiba student SK
4 1 SK 20 Kanagawa student TM
5 I HW l 9 Saitama student SH
i6 1 SH 19 Tokyo student HW
_ 7 1 SB 20 Kanagawa student $2
"a 1 sz 20 Saitama student SB
M9 1 TY 23 Tokyo student TJ
b l 0 I TJ 22 Tokyo student TY
1 1 II YM 28 Kanagawa teacher EK
E 11 BK 28 Kanagawa clerk YM

 

39

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TablﬁJ (cont’d)
1 3 11 MT 3 1 Saitama teacher IN
1 4 11 IN 3 1 Tokyo teacher MT
1 5 11 MD 39 Tokyo teacher 12
16 11 I2 34 Kanagawa university staff MD
1 7 II HY 3 7 Kanagawa housewife SZ
1 8 II SZ 39 Saitama housewife HY
19 II WB 30 Tokyo housewife PT"
20 II AZ 40 Tokyo university SN in Group 111
staff

21 III MG 45 Tokyo university staff SF
22 III BS 47 Tokyo housewife T2
23 111 T2 53 Tokyo teacher BS
24 III MM 54 Tokyo university staff KM
25 111 KM 54 Tokyo housewife MM
26 111 SN 56 Tokyo university staff AZ in Group II
27 III WK 54 Tokyo teacher ST
28 111 ST 58 Tokyo teacher WK
29 111 MR“ 58 Kawasaki housewife AR
30 III AR 60 Tokyo housewife MR
31 III SF 61 Tokyo university staff RG

 

* WB's conversation partner, FT, participated only as a speaker in chatting since FT had
lived in a foreign country for more than one year.

** MR 's data in formal conversation is not used because there are ten formal data besides
MR's.

3.3.3. Procedure

Each pair of participants met with me in the conference room of university library
or at someone's house by appointment. The participants had been told what they would be
asked to do when they were recruited. When they met the researcher on the scheduled
day, the instructions on the procedures were given to one person at a time in a separate

room. The whole exchange was video— and audio-taped and used as formal conversation

data. The script of the instructions is shown in appendix 2. The researcher gave
instructions in the same manner to each participant.

After each pair of participants was given instructions in a separate room, they sat
at a table showing their proﬁles to the video camera. They were asked to talk for 15
minutes about a school trip or other trips that they had taken. The researcher left the room
immediately after she turned on the video camera and audio -tape recorder.

After the chat, participants were given a sheet to rate different listening
behaviors of the woman in the video. After seeing video clips and rating them,
participants are asked to talk freely about what they thought about the listening behavior

in video and/or in general.

3.3.4. Formal conversation

A 2-minute segment of a participant listening to each set of the instructions was
taken as formal conversation data. When participants (primary listeners) claimed the
speakership and took the ﬂoor, the researcher let them speak until they gave up the ﬂoor
and excluded the segments from the 2-minute data. In each of the 2-minute segments, the

primary speaker (the researcher gives instructions and the listener simply responds.

3.3.5. Informal conversation

Conversations between close friends were recorded right after the instruction
session. Each 15-minute conversation was transcribed and divided into segments, or
wadan (Minami 1987; Szatrowski 1991), which is similar to 'ﬂoor' or 'speakership.‘
Thirteen segments (20 minutes 26 seconds in total) in which one participant talked about

past trip and another participant plays a listening role were selected for analysis. The

41

criteria to select the 13 segments out of 225 minutes informal conversation will be

explained in Chapter 5 as well as quantitative and qualitative analyses.

3.4. Listening behavior to be analyzed in the present study

This section presents the deﬁnitions and examples of listener responses, including
non-verbal types. The term 'Reactive Token (RT)' was introduced by Clancy et a1. (1996)
to explain 'communicative strategies' that listeners engage in. RTs are not restricted to
backchannels; they represent 'a more inclusive range of 'non-primary turns' (Clancy et al.:
357). The reason I have chosen RTs as a cover term is that Japanese listener responses
should be interpreted in the broader sense including not only backchannels, or aizuchi,
but other expressions such as sentence completions, laughter, etc., and non-verbal
expressions, such as nodding. RTs can be used to express different kinds of tokens that
listeners use. The deﬁnition of RTs in the present study is given below, taken from Clancy
et a1. (1996: 356);

A 'Reactive Token' (='RT') is 'a short utterance produced by an interlocutor who is

playing a listener's role during the other interlocutor's speakership'. That is,

Reactive Tokens will normally not disrupt the primary speaker's speakership, and

do not in themselves claim the ﬂoor. If a short utterance served as the second pair

part of an adjacency pair (Sacks et a1., 1874), for example, as an answer to a
question or a response to an offer, it was not considered a Reactive Token.

In the subsequent sections, I will describe the types of RTs analyzed in the
present study. The major categories are verbal and non-verbal RTs, each of which are

further subcategorized.

3.4.1. Verbal reactive tokens

42

I subcategorize Verbal RTs into seven types, that is backchannels, reactive
expressions, repetitions, collaborative ﬁnishes, laughter, short comments, paraphrases,
and resumptive openers. Examples of each category are given in the following sections.
English examples are taken from Clancy et al (1996) and Japanese examples are from my

own data and from Horiguchi (1987).

3.4.1.1. Backchannels, aizuchi

I deﬁne backchannels as non-lexical forms that primarily function as 'continuers'
(Schegloff 1982). Although Clancy et al. (1996) did not include hai, 'yes,' in this
category, the present study includes it, unless it appears as the second part of an
adjacency pair (Sacks et al.1974). According to Kitagawa (1980: 110), hai indicates "that
the speaker has heard (and understood) what the addressee said to him." It is
interchangeable with un, a casual form of hai, depending on the context. Backchannels I
analyzed in my data are listed in (1) below. The reduplicated forms of these (e.g. hai hai)
are also included but these are counted as one occurrence which was intensiﬁed or as an
individual listener's preferences.
(1)
a. hai 'yes' (formal)
b. ee 'uhuh' (less formal than hai)
c. un 'yeah' (informal)
(1. aa 'oh' (showing surprise, etc.)
e. haa 'mhn’ (formal form of hai)
f. hoo 'oh' (the user is impressed)
g. huun ‘uh-hum’ (showing interest)
h. hee ‘really’ (showing surprise, etc.)

3.4.1.2. Reactive expressions

The present study adopts the description of reactive expressions by Clancy et al.;

43

'short non-ﬂoor-taking lexical phrases or words' (1996: 359). These include sugoi, ' great,’

honto, 'really,' soo, 'it is so,‘ a soo/ soo ka, 'is that so,’ and ii na, 'nice.’

Excerpt 3.1: SK listens to the story of TM's trip to Canada and uses a reactive expression,
sugoi ne, 'fascinating, isn't it?' SK's utterance overlaps TM's.

[Japanese]
1 SK: iyokoo tie kanjijanai n da.

Travel QT like NEG NM BE
2 TM: soo da ne. demo nanka [dakara nanka

right BE IP but like so like
3 -) SK: [ee sugoi ne

wow fascinating IP
4 TM: ne
IP

[English]
1 SK: It isn't like a trip, right?
2 TM: Yeah, right. But like [because like
3 -) SK: [Wow, (it is) fascinating, isn't it?
4 TM: right?
3.4.1.3. Repetitions

Clancy deﬁnes a repetition as a situation in which the 'non-primary speaker
repeats a portion of the speech of the primary speaker'. Repetitions have the function of
reinforcing mutual understanding between speakers and listeners and disbelief on the part
of the listener.

Excerpt 3.2: $2 (39 years old) says that she brought a loaf of bread to eat in case she
starved during her school trip; HY (37 years old) listens to SZ.
[Japanese]
1 S2: Demo chichame no, ano re de chigireru kurai no
But small NM well hand with pinch part NM
2 -) HY: chicchame no

small NM
[English]
82: But it was kind of small, ah, it is about the size that it is easy to pinch with your

HY: small one
SZ: =ﬁngers
3.4.1.4. Collaborative finishes

In Japanese conversations, the speaker's utterance may be continued and
completed by an interlocutor providing syntactic units such as a verb and an object
(Lerner and Takagi 1991; One and Yoshida 1996; Hayashi and Mori 1998; Hayashi 1999).
The second part of the co-construction when uttered by a primary listener, is considered to
be a kind of listener backchannel (Mizutani 1984; Horiguchi 1997; Clancy 1996). The

present study will also include collaborative ﬁnishes as RTs.

Excerpt 3.3 taken from Lerner (1991) (Clancy et a1. 1996) is an example in which Josh
ﬁnishes Mary's utterance.

1 Mary : Now most machines don't record that slow. So I'd wanna - when I
make a tape,
2 Josh: be able to speed it up. (Lerner 1991: 441)

The Japanese example is in excerpt 5, taken from the present study.

Excerpt 3.4: MR tells that she leaned the Marimo song when she went on her school trip.

[Japanese]
1 MR: Watashi wa shuugakuryokoo no toki wa ano marimo no uta toka,
I TP school trip when TP ah MARIMO of song etc
2 9 AR: A narai mashita ne. utatte ne.
Oh, learned IP sang IP

[English]

1 MR: As for me, when I went to school trip, marimo song and so on
2 AR: learned, didn't we? Sang, right?

In excerpt 3.4, MR starts her utterance, ' As for me, when I went on my school

trip, marimo (a type of ﬂoating, spherical moss that grows at Akan National Park in

45

Hokkaido) song and so on.’ As soon as MR states the subject of the sentence, AR provides

a verb with a tag question that has the effect of establishing common ground.

3.4.1.5. Laughter

Laughter is considered to be an RT in the present study because laughter often
appears in the same contexts as verbal RTs; in Japanese culture laughter is one of the
vocal signs expressing the listener's attitude toward what the speaker says (Maynarad
1989; Pure 2001). In formal conversation, it plays an important role especially when
people converse with unfamiliar people by mitigating FTAs. On the other hand, in
informal situation, laughter can contribute to a display of solidarity. Laughter that occurs
with verbal RTs is counted as a form of RT but not as laughter. However, laughter that is
associated with nods is counted as laughter because nods are considered as intensiﬁer of
any verbal responses.

Excerpt 3.5: TJ (22 years old) talks about her high school trip to TY (23 years old) but
she found that she has somehow lost some memories.

[Japanese]
1 TJ : Kioku ga {nai n dakedo nande daro=.}
memory SB NEG NM but why wonder

2 -) TY: { }

3 TY: Kookoo no toki no kioku wa aruhazu nanom'.
High school time of memory SB exist though

4 Y]: Are zannen.
Oh, disappointing

[English]

1 TJ: {I don't remember anything. But why}

2 -) TY: { }

3 TY: (You) should have memories of high school.

4 -) TJ: {Oh, I'm disappointed}.

TJ laughs that she does not recall (her school trip) and she wonders why. TY laughs
showing empathy with TJ who feels it is strange and funny. If TY had not responded with
laughter to TJ, this conversation would have turned out very differently and TY's RT may

have had the effect of criticism or surprise.

3.4.1.6. Short comments

This category consists of lexical elements longer than a word and expresses
comments by the listener without claiming speakership. The short utterance can be seen as
talk-in-progress at non-grammatical completion points. Speakers do not relinquish their
turn upon the listeners' utterance but continue their turn.

Excerpt 3.6: TJ (22 years old) listens to TY’s (23 years old) memory of her best friend
who cried during her school trip.

[Japanese]

1 TJ: Nadameta no.
soothed IP

2 TY: Un. Sonna zenzen kiraija nai yo rte. { }
Yeah that not at all dislike said

3 TJ: { }

4 9 TJ: Waa, Sugoi seishun mono da.
Wow very teenage NM BE
5 TY: Sonna sonna koto atta naa. Soo ieba.
That that kind of thing NM happened IP so say if
[English]
TJ: Did you comfort her?
TY: Yeah, I said it was not true at all (that I didn't like you). { }
TJ: { }
9 TJ: Wow, a great teenage story.
TY: That , things like that happened, come to think of it.

MADJNH

TJ asks TY if she comforted a crying friend and TY answers laughingly that she

did. TJ responds with laughter to TY and comments, 'wow, a great teenage story.’ TY

47

continues her recalling in 5, 'that, that kind of thing happened, come to think of it.’ TJ's
utterance can be interpreted as a type of RT that expresses the listener’s involvement

through surprise.

3.4.1.7. Resumptive openers

"Non-lexical elements used at the turn initial points" are deﬁned as resumptive
openers (Clancy et al. 1996). These occur at the beginning of new turn claiming
'speakership.‘ Since they can be followed by RTs such as backchannels or other types of
RTs, they are "hybrid in nature: they themselves do not constitute a new turn, but they are
Reactive Tokens that occur at the beginning of a turn.” (Clancy et a1. 1996: 363) Excerpt
3.7 is an example of an English resumptive opener while excerpt 3.6 illustrates how TJ’s

utterance in 4 also contains the resumptive opener 'Wow' at the beginning of her

utterance.
Excerpt 3.7
l A: ...Ho=w are you doing with the house.
2 9 B: ...Oh, got it all uh...primed,... just about, ...except two sides [of it].
3 A: [Oh you 4A: shoot
a] primer stuff
(Clancy et a1 1996: 363)
3.4.1.8. Paraphrases

Listeners respond by paraphrasing what speakers just said to show their
understanding, interest, or clariﬁcation (Mizutani 1984; Horiguchi 1988). The following

examples are about a TV show, cited in Horiguchi (1988: 20).

Excerpt 3.8:
[Japanese]
1 A: Jukeisha nan desu.
Prisoner BE
2 9 B: Keimusho ni haitte iru hito.
Jail in incarcerated being person

[English]

1 A: He is a prisoner.

2 B: A person who is in jail

Excerpt 3.9:

[Japanese]

1 A: Kohi iro to shiro o nutta no.

coffee color and white OP colored IP
2 9 B: Sutoraipu m’ nutta no ne.
stripes in colored IP

[English]
1 A: I painted coffee color and white.
2 B: You drew stripes, didn't you?

In both excerpts 3.8 and 3.9, listeners paraphrase A's utterance in different wordings but
they function similarly to repetitions in that they conﬁrm what speakers have just said,

but do not intend to claim speakership.

3.4.2. N on-verbal reactive tokens

In the present study, vertical head movement, nod(s) are considered to be RTs
when they are not accompanied by other verbal RTs (see line 6 in exerpt 3.10 below).
When they are associated with other RTs, they are categorized as part of the verbal RTs
(in 3 and 9 below) and they have the function of emphasizing them. Therefore, in the
present study nods that accompany verbal RTs will not be transcribed to simplify
transcription.

It is very common for some listeners to nod more than once during a

49

conversation (in 6 and 9 below ); however, such nods are counted as a single occurrence

when they appear continuously in the context. A single occurrence of nodding ends when

a recognizable gap appears.

Excerpt 3.10: RM (41 years old) gives instructions to YG (19 years old) about what to do

about OS (20 year old).
[Japanese]
1 RM: OS son to,

OS Ms. with
2 YG: hai
3 H

yes

3. RM: nani o shite itadaku ka ni tsuite,

What OP do QP about
4. YG:
5. 9 HH
6. RM: kantan m' setsumei sasete itadakimasu.

Brieﬂy in explain let
7. YG:
8. HH
[English]l
1. RM: With OS,
2. YG: yes
3. H
4. RM: About what to do,
5. YG:
6. HH
7. RM: Let me explain brieﬂy.
8. YG: yes
9. HH

 

hai

yes

“ When Japanese transcripts are translated into English, the word order in the English
translation is different from the original. The location of HTS and/or nods shown in
English translations demonstrate the relative locations in a segment. For example, in
excerpt 11 'yes' and continuous nods follow line 7, but not 'brieﬂy.‘

50

3.5. Social variables
This section explains the social variables that will be investigated in relation to

listening behavior in the present study.

3.5.1. Age

Participants in the present study ranged in age from 19 to 61. The data will be
analyzed the use of RTs in relation to age, that is whether there is any age grade patterns
of the use of RTs in frequency and types of tokens. Furthermore, listening behavior of
female speakers will be investigated as three groups that represent three generations, such
as, college students, young adults (late 20's to 30's), and adults (over 45). The deﬁnitions

of three groups are shown in table 3.2.

Table 3.2. Age groups of participants

 

 

 

 

 

number age sex
Group I 10 19-23 female 8 undergraduate and 2 graduate students.
year old
Group II 10 28-40 female 3 home makers and 7 working women.
year old Half are married and half are single
Group III 10 45-61 female 3 home makers and 6 working women.
year old All are married.

 

 

 

 

 

 

It might be questioned how we can divide people into age groups without any
biologically or socially determined factors. The present study focuses on the people's use
of language, particularly listening behavior. The ﬁrst border between age 23 and 28 is
determined by the following analogy. In Japan it is common for people to change their
behavior, including language use, after they ﬁnish school and go into the real world.

Therefore, college students and 28 year old working people could have different

51

behavior, in other words, 28 year olds may have acquired different variations of listening
behavior from college students.

The border between age 40 and 45 may be controversial since it is very difficult
to draw a line between the mid-thirties and ﬁfties. There is a wider range of individual
difference as participants’ age becomes greater, since their social roles are much more
complicated compared with the youngest group. The reason why group 2 and 3 are
divided between age 40 and 45 is that group II is that it is the closest group to the age (41
year olds) of the researcher who participates as a speaker who gives instruction to all of
the participants. The generational issue of whether a participant recognizes the researcher

to be the same generation or not will be analyzed group by group.

3.5.2. Styles of conversation

Two styles of conversation that will be examined in the present study are formal
and informal conversation. The level of formality is estimated by not only linguistic
forms but the type of speech events and the relationship between participants in each
conversation. The relationship between participants considered for analysis is the degree
of acquaintance, age difference, and social status.

The Japanese language has two distinct systems of linguistic forms beside the
honoriﬁc system, called keigo, i.e., desu/masu form, which is polite and used mainly in
formal situation and the plain form used in mainly informal situations. However, one of
the styles is not exclusively used in one conversation type in actual conversation. For
example, in actual conversation use, some only use plain forms when they speak with

friends but some people do not use plain forms even when they chat with very close

52

friends in real conversation use. In addition, those two systems are often mixed in the
speech to a different degree even in formal conversation (Okamoto 1998; 1999; Cook
1999).

The type of speech events selected as formal in the present study is when the
researcher gives instructions to participants. In contrast, informal conversation is chat
between two close female friends. Each one of participants is asked to participate to the
study by her partner.

The reason formal and informal conversation is deﬁned by linguistic, pragmatic,
and social features is that the way people achieve social goals can not be explained by

just one feature.

3.6. Summary
This chapter explained how the data were collected and deﬁned each RT to be
investigated. The following chapter will present quantitative analyses of the RTs used in

the formal conversations, and discusses how the results can be interpreted.

53

CHAPTER 4

REACTIVE TOKENS IN FORMAL CONVERSATION

4.1. Introduction

This chapter presents a quantitative analysis of the RTs used in a total of 60
minutes of formal conversations taken from 30 respondents who participated as primary
listeners in different age groups and discusses the results with respect to the interrelation
of verbal RTs and nods to the social variable of age. First, the quantitative results for the
total number of occurrences of RTs and the types of RTs used by each respondent will be
presented. Secondly, the relationship between age and use of RTs will be discussed. Then
overall data relative to each hypothesis will be presented with the results of statistical
analysis. Finally, the interactional characteristics of Japanese formal conversation will be

proposed.

4.2. Quantitative results of total reactive tokens

This section presents the quantitative result of RTs used in formal conversation.
First, the overall results of verbal and non-verbal RTs used by all participants, then each
type of verbal RT and types of backchannels used by each age group will be examined in

order.

4.2.1. Total reactive tokens used in 2 minutes

Figure 4.1 shows the participants' ages and total number of RTs used by each

participant during 2 minutes of instruction. The mean score of the total number of RTs is

54

53.4.

Table 4.1 displays the total number of verbal and non-verbal RTs used by all
participants. In the present study nods are considered to be non-verbal RTs. The total
number of verbal RTs used by all participants is 853 (53%) and that of non-verbal ones,
nods. comprise 751 (47%). The results indicate that approximately half of the RTs are
non-verbal RTs. In other words, the high percentage of nods indicates the relative
importance of non-verbal RTs in a formal context such as listening to instructions from an

unfamiliar person.

Figure 4.1. Age and total number of reactive tokens

 

 

 

 

 

 

 

 

 

 

 

 

 

80
o
7 Q
a 0 . r °
.54) 50 ¢ 8 o .
O O O o
‘H ... Q .
it; 50 v . o o o
’3' 40 3’ t
P. ° 0
“5 30
g 20
10
0 l i i i i i
0 10 20 30 40 SO 60 70
I Participants' age

 

 

Pearson Correlation Analysis shows that verbal RTs and non-verbal RTs have a
strong negative correlation [r=-.828, p=.000]. That is, listeners who use more nods use

fewer verbal RTs and vice versa. The next section examines how RTs varies among

55

different age groups.

Table 4.1. Total of reactive tokens by all participants (11 = 30)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Tokens of Tokens of Total to RTs
Group ID Age Verbal RTS Non-verbal RTS,
nods
1 1 YO 19 21 23 44
2 1 OS 20 10 45 55
3 1 TM 21 20 3o 50
4 1 SK 20 19 34 53
5 1 HW 19 4 36 40
6 I SH 19 22 14 36
' 7 1 SB 20 16 40 56
8 1 S2 20 14 28 42
""9 I TY 23 7 46 53
10 1 TI 22 18 42 60
11 11 YM 28 15 37 52
12 11 BK 28 38 7 45
13 11 MT 31 10 51 61
14 11 IN 31 50 9 59
15 11 MD 38 52 7 59
16 11 12 33 18 36 54
17 11 HY 37 60 5 65
18 11 S2 39 3 45 48
'19 11 we 30 39 17 56
20 11 AZ 40 18 51 69
21 111 R0 45 29 12 41
22 111 BS 47 49 17 66
23 111 TZ 53 55 7 62
24 111 MM 54 36 12 48
25 111 KM 54 21 14 35
26 111 SN 54 65 7 72
27 111 WK 54 15 33 48
28 111 ST 58 39 9 48
29 111 AR 60 59 17 76
30 111 SF 61 31 20 51
total 853 751 1604
AVE. 26.7/min

 

56

 

4.2.2. Reactive tokens by age groups

In order to determine whether the respondents use RTS similarly in terms of the
overall ratio of verbal and non-verbal RTS, I divided the 30 participants into three age
groups as shown in Table 4.2. Group I consists of 10 participants whose age ranges
between 19 and 23. They are college undergraduate or graduate students in Tokyo. Group
II and group III consist of 10 participants between 28 and 40 and 10 participants between

45 and 61 , respectively.

Table 4.2. Verbal and non-verbal reactive tokens used in 2-minute instruction by age
groups (n=30)

 

 

 

 

Group I Group II Group 111
Age 19 -23 Age 28-40 Age 45-61 TOTAL ANOVA
(n=10) (n= 10) (n=10)
Verbal reactive tokens 151 303 399 853 p = .004
Non—verbal reactive 338 265 148 751 p = .013
tokens (nods)
TOTAL 489 568 547 1604 ms.

 

 

 

 

 

 

 

 

Table 4.2 shows that group I uses verbal RTS the least frequently among the
three groups and uses nods most frequently. Group III, the oldest group, uses RTS in an
opposite way to the youngest group, that is, they use verbal RTS most and nods least.

Group II is in between group I and III in both verbal RTS and nods.

57

Figure 4.2. Verbal and non—verbal reactive tokens used in 2-minute instruction

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

by age groups
600
8
E 500 148
3 265
g 400 __#_. -—-— .1 — ——-— -
g 300 338
C
8 200 399
E 303
2 10° 151
o . ‘ .
Group I age 19-23 Group II age 28-0 Group III age 45-
(n=10) (n=10) 61 (n=lO)
Age groups
5.179333311339229. £91599: Ellen-verbal resistive telgsus ("9931.3

 

 

The analysis of variation (ANOVA) is used to analyze the difference in the
number of total RTS, verbal RTS and nods among three age groups. ANOVA does not
Show a signiﬁcant difference in total RTS among age groups [F(2, 27) = 1.66, p = .207].

Table 4.3 presents the results of Tukey post hoc test.

Table 4.3. Results of Tukey post hoe test of total reactive tokens among age groups

 

 

 

 

 

 

 

 

Group I Group II Group 111
Age 19-23 Age 28-40 Age 45-61
Group I 1.000
Group 11 0.2014 1.000
Group 111 0.4108 0.8866 1.000

 

 

The results of ANOVA Show that the total number of verbal RTS used by each
group is signiﬁcantly different [F(2, 27)=6.58, p=.004]. The total number of nods of each

group is also signiﬁcantly different [F(2, 27)=5.12, p=.013]. The Tukey post hoc test in

58

Table 4.4 and 4.5 explain in more detail about the difference among the three age groups.
Table 4.4 indicates that in terms of frequency of verbal RTS, group I and III are
Signiﬁcantly different (p=.003), while neither group I and II (p=.088) nor group II and III
(p=.358) are signiﬁcantly different. The difference between group I and II is almost
signiﬁcant. The difference in the results for nods by age group is parallel to the results for
verbal RTS shown in Table 4.5. But it should be noticed that the difference between group

I and II is larger (p =.088) in nods than in verbal RTS (p =.452).

Table 4.4. Results of Tukey post hoc test of verbal reactive tokens among age groups

 

 

 

 

 

 

 

 

Group I Group II Group 111
Age 19-23 Age 28 -40 Age 45 -61
Group I 1.000
Group 11 0.0885 1.000
Group 111 0.0036“ 0.3588 1.000

 

Table 4.5. Results of Tukey post hoc test of nods among age groups

 

 

 

 

 

 

 

 

Group I Group II Group 111
Age 19-23 Age 28-40 Age 45 -61
Group I 1.000
Group 11 0.4528 1.000
Group 111 0.0103* 0.1436 1.000

 

 

 

The following excerpts demonstrate that how verbal RTS and nods are used
differently by participants. These segments are the same part of planned utterances in
which a speaker introduces herself to each listener before giving instruction. The symbol

’H' indicates a single nod and 'HH' means the listener nods twice.

Excerpt 4.1: OS (20 years old) nods at almost every recognizable pause and in the middle

59

of utterance also.

[Japanese]

1 Researcher: Eeto o hanashi shimashita keredomo, watashi wa,
ah, tell did though I-TP

2 OS: HH H

3 Researcher: ano amerika no mishigan shuu ni aru daigaku de,
Ah, America-of Michigan state-in exist university-at

4 OS: H

5 Researcher: genngogaku o senkoo shiteite,
Linguistics - OP major doing

6 OS: H

7 Resercher: anoo, nihongo no kaiwa no kouzou?
Ah, Japanese-of conversation-of structure

8 OS: H HH

9 Researcher: Disukoosu to iu n desu keredomo, kenkyuu shite imasu.
Discourse QT say NM BE but studying

10 OS: H H

[English]

1 Researcher: Ah, I have told you but I

2 OS: H H

3 Researcher: Ah, at a university in Michigan state in America

4 OS: H

5 Researcher: I'm (studying) linguistics as major

6 OS: H

7 Researcher: Ah, the structure of Japanese conversation

8 OS: H H

9 Researcher: Called discourse, though, I'm studying (that)

10 OS: H H

Excerpt 4.2: HY (37 years old) uses hai, 'yes', formal form 6 times and 2 head nods
during a similar utterance to excerpt 1. HY also utters hai. 'yes', at almost every
recognizable pause, while she nods in the middle of sentence.

[Japanese]

1 Researcher:
2 HY:

3 Researcher:
4 HY:

5 Researcher:
6 HY:

7 Resercher:
8 HY:

9 Researcher:
10 HY:

l 1 Researcher:
l 2 HY:
[English]

ano sakihodo chotto ohanashi shimashita ga,
Ah before little tell did but
hai
yes
watashi wa amerika no mishr'gan shuu ni aru daigaku de,
I-TP America-of Michigan state-in exist university-at

H hai
yes
genngogaku o senkoo shiteite,
Linguistics-OP major doing

H hai
yes

nihongo no kaiwa no kouzou?
J apanese-of conversation-of structure

hai
yes

Disukoosu tte iundesu keredomo,
Discourse- QT say but

hai
yes

sore o kenkyuu shiteimasu.
That-OBJ studying
hai

yes

Researcher: Ah, 'I told you before I am in major in Linguistics at a university in

yes yes H yes

Michigan in US and well, I'm studying the structure of Japanese? called discouse.’

H

yes yes yes

Excerpt 4.3: SN (54 years old) responds to the researcher with 3 hai, 'yes', 3 ee, 'yeah,
and 2 un, 'yeah'.

[Japanese]

1

Researcher:

ano konoaida mo ohanashi shimashita kedomo,
Ah last time too tell did but

61

2 SN: ee x2
yeah

3 Researcher: ano watashi wa,

Ah I-TP
4 Researcher: amerika no mishigan shuu ni aru daigaku de,

America-of Michigan state-in exist university-at
5 SN: hai ee

yes yeah
6 Researcher: genngogaku o senkoo shiteimashite,
Linguistics-OP major doing
7 SN: hai
yes

8 Resercher: de, nihongo no kaiwa no kouzou,

Then, Japanese-of conversation-of structure
9 SN: 1m

yeah

10 Researcher: danwa tte iundesu keredomo,

Discourse-QT say but
11 SN: un

yeah

12 Researcher: sore o ano shirabete imasu.

That-OP ah examining
13 SN: hai ee

yes yeah

[English]

Researcher: 'Ah, I told you before I am in major in Linguistics at a university in

yeah yeah yeah

Michigan in US and I'm studying the structure of Japanese? called discouse.’

yes yeah

yeah yeah yeah

SN mixes three 'yes's that vary in degree of formality, but the contexts of RTS are

similar to others. The use of different types of RTS should be attributed to social factors

such as age and relationship. Since these three participants don't know the researcher

62

before the session, hai, 'yes', is considered to be socially appropriate for all participants in
this kind of formal setting. However, the data Show that the youngest participant uses
non-verbal RTS, exclusively, the second youngest uses formal 'yes' and the oldest one
uses three kinds of RTS. We will discuss the possible reasons for this variation later in this

chapter.

4.2.3. Interrelation between age and verbal and non-verbal reactive tokens

The results of a Pearson Correlation analysis show that age and total number of
verbal RTS have a positive correlation [r =.543, p=.002], which means older participants
tend to use more verbal RTS, while the total occurrence of non-verbal RTS and age have a
negative correlation [F-.465, p=.010], which means that older participants use fewer
nods and younger participants use more nods. The total number of RT and age don't have
any signiﬁcant difference [F(2, 27)=1.66, p=n.s.]. That is, the total number and age does
not have any correlation or signiﬁcant difference among the three age groups.

What we can conclude from this result is that the total number of RTS used in 2
minutes conversations by each participant does not vary statistically. In other words,
Japanese speakers of different age group reacted to the same speakers' utterances as with
a Similar frequency of RTS to the extent of statistically not signiﬁcant level. However, the
ratio of verbal RT and nods vary among individuals and it shows age-graded variations.
That is, the results indicate that younger people tend to use fewer verbal RTS and more

nods and older people show the opposite tendency in the given context.

4.2.4. Hypotheses 1 and 2

63

This section explains how hypothesis one and two are tested by the quantitative
results of formal conversation in the present study.

Hypothesis 1: The ratio of verbal to non-verbal RTS may vary among individuals,
but the total number will be similar in similar contexts.

Overall, this hypothesis was supported by a signiﬁcant effect. The number of
verbal RTS and the number of nods have a negative correlation but the total number of
RTS and nods is similar among participants.

Hypothesis 2: The frequency of verbal and non-verbal RTS is affected by age.
Older female speakers use more verbal RTS than do younger
speakers.

The hypothesis two is also supported by the results. The number of verbal RTS
and age have a positive correlation and the number of nods and age have a negative

correlation. The next section will explain how age relates to the choice of RTS or

interactional characteristics by describing the use of RTS by age groups.

4.2.5. Types of reactive tokens by age

Table 4.6 displays the types of RTS used by each age group. Baekchannels are
the most commonly used verbal RTS. 42.9 % of RTS consist of backchannels and 11.4%
of total RTS consist of other verbal RTS, such as reactive expressions or repetitions.
Baekchannels have a positive correlation to age [r=.54l, p=.002] with a Pearson
Correlation analysis. Older participants use more frequent backchannels. However, the
occurrence of other types of RTS seems less important. Only resumptive openers, which
are non-lexical forms followed by speaker change have a positive correlation to age

[r=.575, p=.001]. In fact, it is observed that older participants take turns by uttering

resumptive openers at the initial point of their speaking turns. 1 will discuss whether or

not those features are characteristics of interaction in formal conversation later in this

chapter.

Table 4.6. Types of reactive tokens used by age group

 

 

 

 

 

 

 

 

 

 

 

 

Types of reactive Group I Group II Group 111 total (%) Pearson
tokens Age 19-23 Age 28-40 Age 45-61 Correlation
(n=10) (n=10) (n=10) Analysis
between
age and
types
1 Baekchannels 122 246 321 689 42.9 r =.541,
...- p =.002
2 Reactive 1 1 26 31 68
expressions
3 Repetition 0 2 4 6
4 Collaborative 0 1 0 1
__ ﬁnishes
5 Resumtive 2 1 l 31 44 F575
l opener p=.001
6 Laughter 14 14 6 34
_ 7 Short comment 1 3 4 8
j 8 Paraphrasing 1 0 2 3
9 Nods 338 265 148 751 45.7
Total 489 568 547 1604

 

 

 

 

 

 

 

 

 

Table 4.7. Result of Tukey post hoc test of backchannels among age groups

 

 

 

 

 

Group I Group II Group III
Age 19-23 Age 28-40 Age 45 -61
Group I 1.000
Group 11 0.0762 1.000
Group 111 0.003 1 * 0.3660 1.000

 

 

 

 

 

4.2.6. Types of backchannels used by each age group
Table 4.8 shows the types of backchannels used by each age group. The most

frequently used token is hai 'yes', (397 times) for all age groups. 80.3% of backchannels

65

 

are hai for group I, 54.4% for group II, and 51.4% for group III. The second most
frequent backchannel is un 'yeah', (137 times), which comprises 10.6% of backchannels
for group I, 24.7% for group II, and 22.5% for group 111.

As explained in chapter 3, hai is a formal form of an and normally used in
formal contexts, such as toward the speaker with whom the listener is not acquainted or
who has a higher status, or in classrooms, rather than chatting in everyday life. It is
obvious that group I participants who are younger than the primary speaker, (researcher,
age 41) choose the appropriate linguistic forms in the given context. Participants in group
II and III appear to make their own judgement in the choice of the formal form, hai 'yes’,
or informal form, un, 'yeah'. It is reasonable to assume that many of the group 11
participants thought that the speaker and themselves were in the same generation so that
they can use the less formal form, un, instead of hai. Group III participants make similar

judgements about the speaker's age as shown in the results.

Table 4.8. Types of backchannels used by age groups

 

 

 

 

Types of Group I Group II Group 111 total Tukey
backchannels Age 19-23 Age 28-40 Age 45-61 post hoc test
(n=10) (n=10) (n=10)
1 hai 'yes' 98 134 165 397 n.s.
2 un 'uh 13 61 63 137 n.s
3 cc 'hum' 2 31 69 102 in table4.9
4 aa 'oh' 3 8 13 24 n.s
5 haa 'yeah' 2 3 10 15 n.s
6 hee 4 5 1 10 n.s
7 huun 0 4 O 4 n.s
8 hoo 0 0 0 0 n.s
122 246 321 689

 

 

 

 

 

 

 

The following excerpts show how hai 'yes', and un, 'yeah', are selected by group

 

I, II, and III. The ﬁrst excerpt is taken from SH of group 1. SH chooses hai 'yes', which is
the formal form and uses it only twice in line 4 and 6 of excerpt 4.4, accompanied with
nods. The researcher is talking about where she went for her own school trip, saying "For
example, since I was born and raised in Yamaguchi, I went to Kyushu when I was in
elementary school and went to Kyoto when I was in middle school". SH does not nod

except when accompanied with hai.

Excerpt 4.4: SH (group I: 19 years old) hai 2 times
[Japanese]
1 Researcher: tatoeba,

for example,

2 Researcher: watashi wa yamaguchi ken de umare sodatta node,
I-TP Yamaguchi prefecture-in born raised because

3 Researcher: Shoogakkoo no toki wa kyuushuu e,
Elementary school GEN time-TOPC Kyushu to
4 SH: hai(H)
yes
5 Researcher: chuugakko de wa kyooto e ikimashita.

middle school-for-TP Kyoto-to went

6 SH: hai(H)
yes
7 Researcher29 de, eeto Tokyo de wa chigaun desu yo ne.

and ah Tokyo-in-TP different BE IP IP

8 SH: 500 desu ne. Shoogakko wa Nikko e itte,
So BE- IP elementary school-TP Nikko-to went
[English]
1. RC: 'For example, I ah, was born in Yamaguchi and raised (there)so
2. SH:
3. RC: When (I was) in elementary school, (I went) to Kyushu
4. SH: yes
5. RM: When (I was) in middle school, (I )went to Kyoto but

67

6. SH: yes
7. RM: And, ah, (It is ) different in Tokyo, right?
8. SH: Right. (I) went to Nikko,.
The second excerpt is from MD in group 11. MD used only un, 'yeah', the

informal form of yes 5 times in lines 2, 4, 6, and 9 of exerpt 4.5. MD did not use nods as

RTS, either, except for those that accompany verbal RTS.

Excerpt 4.5: MD (group II: 39 years old) un 5 times
[Japanese]
1 Researcher: tatoeba,
for example,
2 MD: un
yeah
3 Researcher: watashi wa yamaguchi ken de umare sodatta node,

I-TP Yamaguchi prefecture-in born raised because

4 MD: un(H)
yeah
5 Researcher: Shogakko no toki wa shugaku ryokyo de kyuushuu e itte,
Elementary school-of time-TP school trip-for Kyushu-to go
6 MD: un un(H)
yeah yeah
7 Researcher: chuugaku no toki wa, shinkansen de,

middle school-of time-TP Bullet train-by

8 kyooto e itta n desu kedo,
Kyoto-to went but

9 MD: an (H)
yeah
1 0 Researcher: kanto de wa chigau rashii desu ne.

Kanto-in-TP different seem BE IP

1 1 MD: ﬁill turn

68

[English]

‘o?°>’.o‘i":‘i‘i”!\’:"

RC: For example

MD: yeah

RC: I ah, was born in Yamaguchi and raised (there)so

MD: yeah

RC: When (I was) in elementary school, (I ) went to Kyushu for school trip
MD: yeah yeah

'RC: When (I was) in middle school by bullet train went to Kyoto but

MD: yeah

10. RC: (It ) seems different in Kanto area, right?
11. MD: FULL TURN

reacts with a resumptive opener and reactive expression in line 4. Un 'yeah', in line 6
occurs repeatedly although they are considered as one occurrence. Repeated tokens,
reactive expression rather than backchannels, frequent RTS, and co-occurrence with nods

may be features from which the speaker can make an assumption regarding how

In excerpt 4.6, BS from group 111 used different kinds of RTS ﬁve times total and BS

interested the listener is.

Excerpt 4.6: BS (group III: 47 years old) 5 RTS
1 Researcher: tatoeba,
For example,
2 Researcher: watashi wa ano yamaguchi ken de=

I-TOPIC Ah Yamaguchi prefecture in

3 sodatta node,
raised because
4 9 BS: haa so desu ka (H)
'yes. Is that so? (H)'
5 Researcher: Shoogakkoo no tokr' wa shugaku ryoko kyuushuu e itte,
Elementary school GEN time-TOPC school trip Kyushu to go-
6 BS: un x2(H2) u=n (H)
7 Researcher: chuugaku de wa,

Then middle school for-TOPIC

69

7 Researcher: shinkansen ni notte de kyooto e ittan desu kedo,
Bullet train ride Kyoto to go-PAST explain BE but

8 BS: u=n (H)
9 Researcher: kanto de wa chigau rashii desu ne.
Kanto in-TOPIC different seem BE IP FP

10 BS: Full turn

[English]
RC: For example

BS:

RC: I was born ah,in Yamaguchi and grew up (there)so

BS: yes, is that so

RC: When (I was) in elementary school, (I) went to Kyushu (for) school trip
BS: yeahx2 yeah

RC: When (I was) in middle school,

BS:

RC: (We) took a bullet train (and ) went to Kyoto but

BS: yeah

RC: (It) seems different in Kanto area, right?’

BS: FULL TURN

If we contrast these three participants’ behavior, SH in excerpt 4.5 uses the least
RTS and the others use RTS with the same frequency but in different ways. The data
provide evidence that the three listeners' behaviors inﬂuence the speaker’s utterance in
progress. The script that the researcher spoke was prepared in advance; however, the
actual transcripts of the three excerpts differ in wording, pausing, intonation and so on,
Since it is not controllable--as a result, the transcription of SH is shorter than others.

For instance, line 10 of excerpt 4.5 and line 9 of excerpt 4.6 are connected to the
previous utterances that end with kedo, 'but', or 'though.' However, in excerpt 4.4
utterance in line 5, which is the same phrase as in lines 10 and 9, is completed. In excerpt
4.4 the speaker uses the connectives, (sore)de, 'and', and a ﬁller, which may show the
speaker's difﬁculty in continuing to talk. Although it is preplanned instruction, it is also

true that on-going instruction is affected by the listener's RT use. Lack of expected

70

continuers made the speaker complete the sentence at unplanned location. Whereas, the
listening behavior of MD and BS appear to be more supportive and cooperative, so that
the speaker can continue utterances without relinquishing the turn.

Other non-verbal features, such as gaze or facial expressions should be further
examined to investigate the inﬂuence of listening behavior; however, the data in the
present study present evidence that types and frequency of RTS affect the Speaker's
utterance and listeners also participate in the conversation indirectly by using the
linguistic strategy of RTS.

The total number of ee 'yeah' used by all participants is less than hai and un. Ee
is only used 2 times by group I speakers, whereas, it is used 31 times and 49 times by
groups 11 and 111, respectively. Pearson Correlation analysis Shows that there is a positive
correlation [r=.555, p=.001] between the number of ee and age. Be is less formal than hai
but more formal than un and the use of ee varies from person to person. ANOVA shows
that use of ee is signiﬁcantly different among age groups [F(2, 27)=4.42, p=.02]. Tukey
post hoc test explains that the use of ee is signiﬁcantly different between groups I and III

(p=.016) but not between groups I and II (p=n.s.) or groups 11 and III (p=n.s.).

Table 4.9. Results of Tukey post hoc test of ee ‘yeah’ among age groups
Group I Group II Group 111
Age 19-23 Age 28-40 Age 45-61

 

 

 

 

Group I 1.000
Group 11 0.4166 1.000
Group III 0.0168"I 0.2306 1.000

 

 

 

 

 

 

Also, although the number is very small, the token haa, which is a more formal

71

form of hai ‘yes’ (McGloin 1997) is used more frequently by older participants. In
contrast, hee is not used by older participants. The difference in use of particular tokens
might be caused by style preference according to age. The data in the present study does
not allow one to generalize the age-related pattern since the data do not contain sufficient
examples that account for this phenomenon.

The following excerpts illustrate how ee 'yeah', and haa 'yes', are used in the

present study.

Example of ee

Excerpt 4.7: T2 (Group III: 53 years old) uses ee and hai 4 times each.

[Japanese]
I Researcher: nani o shirabeteiruka wa,
What OP examining-TP
2 Researcher: ima wa moshi agerare nai n desu kedo,
Now-TP tell-NEG BE but
3 T2: 9 ee ee
yeah yeah
4 Researcher: naiyo toka,
Content etc.
5 T2: 9 ee
yeah
6 Researcher: hyogen toka,
Expression
7 T2: hai
yes
8 Researcher: kotoba zukai to iu youna koto o shirabete iru wake =
Language use-QT say like NM-OP examining
9 T2: 9 ee

yeah

72

10 Researcher: =de wa nai node,
TP NEG so

11 Researcher: ano ki ni nasara [nai de itsumo doori m',
Ah mind NEG and ordinary way-in

12 TZ: [hai wakarimashita
Yes I understood
l3 Researcher: ohanashi shite itadakere ba kekko desu.
Speak if ﬁne BE
14 T2: hai hai
yes yes

[English]

1. RC: What (I am) studying,

2. RC: (1) can't tell you now but,

3. T2: yeah yeah

4. RC: content or something,

5. TZ: yeah

6. RC: expressions or something,

7. T2: yes

8. RC: I'm not examining things like the use of language =
9. TZ: yeah

10. RC: =(don't )So,
11. RC: please don't [mind too much and then as usual,
12. T2: [yes I understand
13 . RC: It is ﬁne if you can talk (as usual).
14. TZ: yes yes

TZ mixes ee, 'yeah' and hai, 'yes' and all instances of ee and hai are
interchangeable, including ones in line 14 that function to Show agreement. They are
eqlﬁvalent linguistically but If hai is used exclusively in the segment, it may give an
irupression of being more formal and rigid than those in which only ee is used. This
Softened, somehow feminine image that is attributed to ee might be the reason why ee is

preferred by older female Speakers of Tokyo standard Japanese over younger Speakers.

73

Example of haa

Excerpt 4.8: SF (group III: 61 years old)

[Japanese]
1 Researcher:
2 Researcher:
I-TP
3 SF: 9
4 Researcher:
5 SF:
6 Researcher:
7 SF:
8 Researcher:
9 SF:
10 Researcher:
1 1 SF:
1 2 Researcher:
l 3 SF:
[English]

MawN

RC: For example,

tatoeba,
For example,

walashi wa,ano yamaguchi ken de umare sodatta node,
Ah Yamaguchi -in born raised because

haa
yes

Shoogakkoo no toki wa,
Elementary school-of time-TP

H

ee kyuushuu e itte,
Ah Kyushu-to went-

H

chuugaku no toki ni shinkansen de
middle school time-in Bullet train-by

ee
yeah

kyooto e ittan desu kedo,
Kyoto -to go—went BE but

H

kanto de wa chigau n desu ne.
Kanto in-TP different BE IP

aa so desu ka
oh so BE Q

RC: I ah, was born in Yamaguchi and raised (there)so,

SF:

yes

RC: When (I was) in elementary school,
SF:

H

74

6 RC: (1 ) went to Kyushu,

7 SF: H

8 RC: When (I was) in middle school by bullet train,
9 SF: yeah
10 RC :(1 )went to Kyoto but',

11 SF: H

12 RC: (It is ) different in Kanto area, isn't it?

13 SF: Oh, is that so
SF is among the oldest participants in the present study. Excerpt 4.8 shows that SF uses
various forms of 'yes,' including haa. Since SF is older than the researcher and the
conditions are identical to the other participants, there is no reason that She has to
behave more politely than other participants by choosing a more formal 'yes'. The use

of haa by this participant can be explained by individual preference.

4.3. Effects of interactional styles

In this section, I discuss how we can interpret the quantitative results which
demonstrate variation in listening behavior among 30 participants. As the results from the
Previous sections suggest, listeners' behavior represented by the use of RTS varies in
terrns of the ratio of verbal and non-verbal RTS and types of tokens, but not in the total
Occurance of RTS that are perceived as listener feedback for the Speakers, nor the contexts
l“eacted by RTS. In other words, native female speakers of Japanese react to the speaker
Similarly if they are given the same context with one of the variations described in this
c3hapter. The behavior must be governed by conversational rules that are Shared in a

SpeeCh community, i.e., the Japanese community in Japan.
In the following section I propose common interactional patterns and argue that

VariEltions in Japanese listening behavior can be explained through the "multidimensional

75

model" (Tannen 1993a) by showing how the results for formal conversation ﬁt into the

model.

4.3.1. Interactional styles

The results Show that interactional styles of all participants can be divided into
two types: (1) higher ratio of nods and lower ratio of verbal RTs; (2) lower ratio of nods
and higher ratio of verbal RTS. The results also indicate that the two patterns are age

graded. Figure 4.2 shows the frequency of nods, backchannels and verbal RTS that

exclude backchannels.

Figure 4.3. Interactional styles

 

900

800

700

600

Number of reactive tokens
6
o

 

 

high nods-How verbal low nods+high verbal

Interactional styles

 

 

l I backchannels [:1 other verbal RTS El nods l

 

In the following section the interactional styles obtained in the present study will

76

be discussed in terms of social factors, such as age and closeness.

4.3.2. Variation of interactional styles

Brown and Levinson (1987) state that power and solidarity are fundmental to
explain linguistic variations. Power governs asymmetrical relationship where one is
higher than the other, while solidarity governs symmetrical relationship represented by

social equality (Tannen 1993a) as we can see the use or non-use of Keigo, honoriﬁc
forms to the addressee in Japanese.

According to Tannen (1993a), Americans conceptualize power and solidarity as
opposite ends of a Single continuum. In general, to Americans solidarity implies
closeness, whereas power implies distance as manifested in the use of words such as
"brother" or "Sister" to express closeness (solidarity) and equality (power).

Tannen claims that solidarity is a form of control and could be performed by the
interlocutor who has higher status. For instance, “Where’s your coat?” can be interpreted
as hierarchical relationship when it is addressed by female boss to male subordinate even
though She intends to show her friendliness. Linguistic strategies involve ambiguity and
POIYSCmy because they can Signal both power and solidarity.

Tannen proposes that to understand cross-cultural perspectives, a

milltidimentional grid of at least two intersecting continua is desirable with closeness-

d‘StanCe dimension on one axis and hierarchy-equality on another.

77

Figure 4.3 Intersecting dimensions of closeness/distance and hierarchy/equality.
(modiﬁed Tannen 1993a)

 

hierarchy

different generation +Stranger
Japanese: employer - emplo ee American: employer - employee
Parent - child g I:
closeness distance

IE1

same generation + stranger
Close friends
American: Siblings

 

equality

In ﬁgure 4.3, power and solidarity are expressed in the intersecting dimensions. I
name each dimension A, B, C, and D, clock wise for convenience. Tannen explains that
American sibling relationship (close and equal) is placed in dimension D, while the
JaPanese relationship of amae, 'dependency' (close and hierarchical), typiﬁed by the
Parent-child or employer-employee (Yamada 1992) is placed in dimension A.
Dimensions B and C on the right and dimensions A and D on the left are of an opposite
nature with regard to solidarity.

Since the relationship between the researcher (speaker) and each participant
(liSteners) was not familiar at the time of data collection, each case should be placed on
the I‘ight half of the solidarity continuum (distant). In addition, there is a power difference

between the researcher and each participant attributed by age difference, situational role

(g1 Vlng instructions and receiving them), or institutional affiliation, so that they should be

78

placed on the power continuum.

More speciﬁcally, the conversations between the researcher (41 years old) and
participants in Group I (age 19-23) are plotted in dimension B, which is the most distant
and involves a power relationship. Among group I participants, the psychological
distance may vary by how they were recruited, for example, participants YG, OS, and SK
were recruited through a professor of their university and others were students where the
researcher was teaching. The ﬁrst group of participants may be plotted in the right upper
area, whereas the other might be located a little closer to the center on the solidarity axis.

The conversations with group II and group III are assumed to be distant but
equal. Group 11 participants were younger than the researcher but since participants were
not informed of the researcher's age, they must have assumed that the researcher was
approximately the same generation. Therefore, group 11 participants should be placed
toward the right of dimension C. Again, their psychological distance may vary according
to their occupation, education or other factors.

Group 111 participants must be placed in the same dimension as group 11 because
although in Japanese culture age is a signiﬁcant factor inﬂuencing the choice of honoriﬁc
expressions, it is reasonable to assume that the researcher may have had relative power
over the participants who were mostly homemakers, considering her role in the context
and social status.

The results Show that there is age-graded variation in the balance of verbal RTS

and nOdS. Namely, younger listeners tend to use fewer verbal RTS and more nods than
01 der listeners although the total number of RTS is similar. In other words, the dimension

B ' . . . . .
Intel‘aetion lS characterrzed by a greater ratro of nods and fewer verbal RTS inﬂuenced

79

by a hierarchical relationship. The difference in the ratio of verbal to nods among group I
participants can be explained by the different degree of psychological distance from the
researcher or other psychological factors. Nodding is likely to be used as a negative
politeness strategy in a power relationship because nods are less direct and not intrusive.

On the other hand, participants who are of equal status use positive politeness
strategy, verbal RTS, in order to be cooperative and supportive listeners and establish
rapport. The dimension C interaction is more dynamic since listeners can adjust the
distance by choosing more informal tokens, or by using substantial utterances such as
paraphrasing, short comments, repetitions, etc. The fact that older participants used those
tokens more often compared with younger is an evidence of "polysemy of power and
solidarity" argued by Tannen (1993a).

In both dimensions, participants are actively engaged in the listener roles by

sending message continuously to the Speaker through different linguistic strategies.

4.4. Summary
This chapter presented the quantitative results for formal conversation which
su1313011 hypotheses 1 and 2. Next, variations in the use of RTS in relation to the social
variable of age were presented. In face-to-faee interaction, age difference between
PartiCipants is an important factor when other social contexts are similar. That is, younger
listeners tend to use a more indirect strategy—nods—to be polite, whereas older listeners
appear to use more direct strategies, such as verbal backchannels or other types of RTS.
The following chapter presents the analyses of informal conversation in which

th . . . . . . .
e Sallie partlcrpants 1n formal conversation chat wrth then close friends.

80

CHAPTER 5

REACTIVE TOKENS OF INFORMAL CONVERSATION

5.1. Introduction

This chapter examines the use of RTS in informal conversation between close
friends. The focus will be not only on the frequency of RTS but also the interactional style
of the informal conversations in which participants engage. The purpose of this chapter is
to ﬁnd characteristics in casual interaction such as chatting through quantitative and
qualitative analyses. I explore how female speakers of Japanese interact with their close
friends and comparision of the ﬁndings in informal with formal conversation will be

presented in Chapter 6.

5.2. Quantitative results of informal conversation
This section ﬁrst explains the data used as informal conversation data, and then

Presents quantitative analyses.

5.2.1. Data
The data consist of 13 dyadic conversations (20 minutes 26 seconds total). The
1length of one conversation segment varies from 52 seconds to 130 seconds. The

Participants, length of each segment and topics of segments are shown in table 5.1.

81

Table 5.1. Lists of participants in informal conversation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Listener Length of Topic Relationship
Group (speaker) Age segment between
(sec) participants
1 1 OS 20 1'23" YG's school trip Same university
__ (YG) (83) 'Vietnam'
2 I SK 20 1'44" TM's school trip Same university
. (TM) (104) 'Canada'
3 I SH 19 1'08" HW'S school trip Same university
(HW) (68) 'Food in Okinawa'
4 l TY 23 1'18" TJ's school trip Same university
(TJ) (78) 'Hokkaido'
5 1 TI 22 1'30" TY's school trip Same university
(TY) (90) 'argument during trip'
Total 7'03"
(423)
6 11 MT 31 1'45" IN'S school trip Graduate of the
(IN) (105) 'At the hotel in same HS
Kurashiki'
7 11 IN 31 2'10" MT'S school trip Graduate of the
__ (MT) (130) 'Kurashiki' same HS
8 II HY 37 2'05" SZ's school trip Neighbor
(82) (125) Food during the trip
9 II 82 39 52" HY's school trip Neighbor
(HY) (52)
Total 6'52"
(412)
10 111 KM 54 1'37" MT'S school trip Same elementary
(MM) (97) 'Climbing' school
1 1 III ST 58 1'58" WB's private trip Same school
(WB) (118) Former Yugoslavia
12 111 MR 58 1'33" AR's school trip Same church
(AR) (93) 'Hokkaido'
13 111 AR 60 1'23" MR'S school trip Same church
~_ (MR) (83) 'Hokkaido'
Total 6'31"

 

 

 

 

 

451

20'26"
(1226)

 

 

 

 

82

 

A total of 20 minutes 26 seconds of Speech was selected from 3 hours 45
minutes of data for analysis of informal conversations. In order to extract the segments
for analysis, two units of conversation, danwa, 'discourse,' (Minami 1987) and wadan
(Sakuma 1987; Szatrouski 1991), a smaller unit than danwa and close to 'Paratones'
(Brown and Yule 1983) according to Sakuma (1987), were considered. Minarni (1981)
claims that danwa is identiﬁed by the following criteria: (a) pause; (b) continuity of
content; (c) the same participants; ((1) the same functions in communication; (e) the same
level of formality; (f) similar nature of topics, such as everyday conversation, chat and
gossip. Informal conversations in the present study were divided into parts by (b) and (c)
according to Minami's deﬁnition.

The main topic of informal conversations in the present study was controlled in
advance. That is, participants were told to talk about their memory of any of their school
trips; however, if they didn't have any memory at all, they could talk about any trips they
had taken. Therefore, (b) continuity of the topic signiﬁes a speciﬁc episode or memory of
a school trip that is told by speakers. Typically, it is distinguished by an opening by a
speaker and ends with closing remarks. The middle part of danwa consists of several
wadan.

After dividing conversations into smaller segments, 13 segments were selected
under the following conditions: one topic covers the whole segment and the speaker
engages in the conversation as the primary speaker, or as the primary information giver;
the length of total segments is similar; the relationship between participants is similar
among other participants in terms of age, closeness, and so on.

Table 5.1 provides information about the segments selected for analysis. Most of

83

the segments are about episodes from the speakers' school trip when they were in high
school, except for segment 11 in which a speaker talked about a trip to Yugoslavia.
Although the relationship between speakers and listeners was planned to be similar such
that all participants were asked to bring their close friends with them, the conversations,
especially in older groups, exhibited differences in the nature of closeness. Therefore,
among 15 pairs, those whose age was not close-—that is, the same age or an age difference
of less than ﬁve years--and whose workplace is the same, were excluded because we
could expect that the hierarchical relationship at work may inﬂuence their interaction
invisible ways.

Five segments of conversation between 19 and 23 year olds were chosen and
four segments each were selected from group II and group III for the following reason:
the researcher’s intent was to make the total length of segments Similar among the three
groups for comparison. However, the younger participants tended to have shorter speech

segments than older participants. Thus, ﬁve segments were chosen for analysis.

5.2.2. Quantitative results

Table 5.2 presents the total number of verbal RTS and nods used by all
participants. Participants used RTS 414 times in total, which consists of 368 verbal RTS
and 46 nods. The percentages of verbal RTS and nods are 88.9% and 11.1%, respectively.
There is a striking difference between formal conversation and informal conversation in
the ratio of verbal RTS and nods. In formal conversation (see chapter 4), 52% of the total
RTS are verbal and 48% are nods. These results indicate that a great number of RTS are

verbal in informal conversations.

84

Table 5.2. Results of reactive tokens by all participants (n=1 3)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Group listener Age Tokens of Tokens of Total of Occurrence
ID Verbal RTS Non-verbal RTS Per minute
RTS, nods
l I YG 19 31 6 37 26.7
3 I SK 21 24 4 28 ""2417”
“4" _ I TJ 22 25 5 30 230 ,
5 I TY 23 15 2 17 i 1.3" " "
M6" 11 IN 31 34 2 36 IIIII h 20.6 I
8 II HY 37 35 2 37 17.8
9 II SZ 39 14 ll 25 29
10 111 KM 54 34 2 36 22.3
11 111 ST 58 55 m0 .. 55 I 279
12 111” "MR“ 58” ‘ "'27” I ' ’ '5' 32 “ M206"
13 111 AR 60 39 iiiiiiii 2 7 4m] H 29.6
total 368 (88.9%) 46 (11.1%) 414 Ave. 21.1/min

 

 

 

 

 

 

 

 

 

 

The occurrence of total RTS per minute is calculated in table 5.2 to make a
comparison with the formal conversation. The average of total occurrence of RTS is 21.1
times per minute in informal conversation, while it is 26.4 times per minute in formal
conversation. The lower occurrence of nods may partly explain why the average of RTS

per minute is lower in casual conversation than fomial one.

5.2.3. Reactive tokens by age group
The following table shows verbal RTS and nods used by each age group. A
Pearson correlation indicates that age and verbal RTS have a positive correlation [r=.629],

but unlike formal conversation, age and nods do not have a correlation. The occurrence of

85

total RTS per minute Shows that participants in group III use RTS most frequently (25.1
per minutes), whereas participants of group I and 11 use RTS in similar frequency (19.6

per minutes and 19.0 per minutes).

Table 5.3. Verbal and non-verbal reactive tokens used by age groups (n=13)

 

 

 

 

 

 

 

 

 

 

 

Group I Group II Group III
Age 19-23 Age 31-39 Age 54-60 total
(n=5) (n=4) (n=4)
_ Length oftotal . . 424,112 391 1227 .
segments (sec)
Verbal reactive 1 14 99 155 368 r=.629
tokens 16.1 /min 14.4 / min 23.7 / min 17.9 / min
Non-verbal l9 1 8 9 46
reactive tokens 2.6 / min 2.6 / min 1.3 / min 2.2 / min
(nods)
Total 133 117 164 414
18.8 / min 17.0/min 25.1 /min 20.2 / min

 

 

5.2.4. Types of reactive tokens by age

The results Show an overall effect of age on the choice of RTS by respondents.
Age and total number of backchannels have a strong strong positive correlation and
reactive expressions and resumptive openers have a relatively positive correlation, also.
However, laughter and age have a negative correlation, which may be important in

understanding the nature of interaction among younger people.

86

 

 

 

 

Table 5.4. Agggroups and ty S of reactive tokens in informal conversation
Types of Group I Group II Group 111 Total Correlation
reactive tokens Age 19-23 Age 31-39 Age 54-60
(n=5) (n=4) (n=4)

1 Baekchannels 50 44 92 1 86 r=+.6766
2 Reactive expressions 18 15 23 56 r=+.497
3 Repetition 9 5 5 19
4 Collaborative ﬁnishes 6 6 5 17
5 Resumptive openers 14 10 25 49 r=+.499
6 Laughter 1 0 1 5 2 27 r=-.424
7 Short comments 4 4 2 10
8 Paraphrasing 3 0 1 4
9 Nods 19 18 9 46 n.s.

Total 133 117 164 414

 

 

 

 

 

 

 

 

5.2.5. Types of backchannels used by age group

 

The types and frequency of the backchannels used by each group are displayed
in table 5.5. Participants in groups I and 11 used an, 'yeah' which is the most informal
form of 'yes,' exclusively, while group 111 participants used hai, 'yes' 4 times, an, 'yeah' 2
times, and ee, 'yeah'--which is less formal than hai but more formal than un--40 times in
informal conversation. Since we have already observed that ee is used more frequently by
group 111 participants in the formal conversation, we can say that ee is preferred as a less

formal form of yes by the older female Speakers who speak Tokyo standard.

87

 

Table 5.5. TyEs of backchannels used by age groups

 

 

 

 

 

 

 

 

 

 

Types of Group I Group II Group III Total
backchannels Age 19-23 Age 31-39 Age 54-60
(n=5) (n=4) (n=4)
1 hai 'yes' 0 0 4 4
2 un 'uh hum' 43 32 2 97
3 ee 'oh' 0 0 40 40
4 aa 'yeah' 1 5 18 24
5 haa 0 0 2 2
6 hee 2 6 6 l4
7 huun 3 0 0 3
8 hoo l 1 0 2
50 44 92 186

 

Figure 5.1 shows how the ratio of backchannels and the total of other verbal RTS

and nods used by all participants differ in formal and informal conversations.

Figure 5.1. The ratio of RTS used in formal and informal conversation

 

‘ 100%
i 90%
l 80%
l 70%
’ 60%
’ 50%
40%
l 30%
1 20%
l
l
l
1

Percentage

10%
0%

The most prominent diﬁerence between the two styles is the ratio of other verbal

RTS, which is approximately half of the total verbal RTS in informal conversation, and

formal

 

informal

Style of conversation

88

1:] backchannels Clotheriverbal RTS 13 nod;

 

 

nods, which are fairly lower than the formal setting. Since we have already discussed
non-verbal features, the following section will focus on verbal RTS in informal
conversation, and discuss what other verbal RTS consist of and how they are used in the

actual data, in order to uncover the characteristics of chatting between friends.

5.3. Characteristics of female friend talk

The earlier section of this chapter discusses listeners' participation through RTS,
including verbal and non-verbal types, but does not analyze the segments in which
listeners take full turns, nor consider speakers' utterances that facilitate listeners'
cooperative behavior. In order to understand the structure of casual informal
conversation, it is necessary to take a closer look at the actual interaction at the discourse
level, not the sentence level.

This section discusses how verbal RTS are used by listeners and are related to
one another in the contexts by introducing the notion of 'assessment strategy' (Strauss and
Kawanishi 1996). The remainder of this section gives an overview of other interactional
characteristics observed in informal conversations, such as interruption, overlapping,
listeners' questions, and use of masculine/feminine forms, which function to show interest
and establish solidarity between participants. Lastly, three commonly observed listening

behaviors in informal conversation will be proposed.

5.3.1. Assessment strategies

Straus and Kawanishi (1996) argue that Japanese listeners frequently use

assessment expressions in various forms, which allows listeners to Show affect toward

89

Speakers and participate in the Speaker's on-going talk to co-construct the conversation
with the primary speaker. They deﬁne assessments (1996: 150) as 'an interactive activity
which involves the expressed evaluation of some entity, event, situation, or state,‘ and
classify them into three types of expressions: (1) expressions of agreement such as soo
soo soo 'right,’ hai 'yes' ; (2) expressions of emotion/affect such as demo sore wa hidoi
jan? 'but, that's awful, isn't it,’ sugoi desu ne 'how terrible,‘ (3) extreme case expressions
such as expressions of non-literal totality such as minna 'everybody,’ zenzen 'not at al,' or
onomatopoeic expressions.

Straus and Kawanishi's ﬁndings show instances that empathy between
participants is achieved directly by using different assessment strategies, and their notion
accounts for the characteristics of informal conversation of the present study. The rest of

this section will investigate assessment strategies used in the present study.

5.3.1.1. Types of assessment strategies in the present study

In the present study assessment expressions are categorized in one of the
following subcategories of RTS; that is reactive expressions such as soo, 'it is so,‘ sugoi,
'great, or awful,‘ backchannels, hai, 'yes,' repetitions, short comments if it iS a longer
phrase such as sore wa sugoi ne, 'that is great, isn't it,’ and paraphrases. The following is a
list of assessment expressions used during a total of 20 minutes 26 seconds of
conversation by 13 participants.

There are a total of 182 verbal RT tokens besides backchannels (186) in the
informal conversation data: 56 tokens of reactive expressions, 19 repetitions, 17

collaborative ﬁnishes, 49 resumptive openers, 10 Short comments, and 4 paraphrases. Not

.‘e
Mix-

 

 

all are assessment expressions, however; as pointed out in Straus and Kawanishi, Japanese
is rich in assessments compared with English and Korean, thus, we can expect relatively
great amounts of these in the present data.

Resumptive openers are turn initial backchannels that are followed by full turns
or other RTS such as reactive expressions. Resumptive openers are used 49 times out of
414 RTS (12%) in 20 minutes 26 seconds of informal conversation, compared to 61 out of
1528 RTS (3.9 %) in 60 minutes of formal conversation data. The frequent occurrence of
resumptive openers (approximately 2.3 times more) in informal conversation can be
explained by the fundamental difference in interactional style between the two types. The
following is a list of assessment expressions used in the present study.

There are ten tokens of reactive expressions, which signal listeners’
acknowledgement of the speaker’s utterance through variations of so types and other
substantial utterances such as sugoi 'great,’ yokatta '(it was) good,’ and honto 'really.'
Reactive tokens often follow resumptive openers, which are vocalic sounds as in (2)
below. The utterances with resumptive openers can express stronger emotion than those
that are not associated with resumptive openers. Also, utterances associated with
resumptive openers tend to appear as interruptions or overlapping the primary speakers'

utterances.

(1) reactive expressions

30 'right' so desu ne '(it is) right'

50 ‘10 yo ne '(it is) right, isn't it' sugo i ii ne 'very good, isn't it'
sugoi yo ne '(it's) good, isn't it' so do ne 'right'

3148001' 'fantastic' yokatta '(that) was good'

“‘80i n desho 'great, isn't it' homo 'really'

91

(2) resumptive opener + reactive expression

iyaa kowai 'no (it's) scary' a sugoine 'oh great,isn't it?'

a soo 'oh it is so' a soo desu ka 'oh it is so'

a soo no no 'oh is that so?’ aa soo de shoo ne 'oh it would be so, right?’

a soo nan do 'it is so' ara sore wa ii wa ne ' that is nice,isn't it?'
aa ii desu yo ne 'oh it is good, isn't it?' maa shikata not 'well, it can't be helped'
iyaa natsukashi 'oh I miss it' haa naruhodo ne ' well, I see'

(3) collaborative ﬁnishes

ikenain dayo ne, " (it) isn't right, is it"
(4) repetition

naiteru no "Is (she) crying"
nai chau n do "(she) cries"

(5) backchannels

waa 'wow' hoo 'oh well'
(6) short comment

omoshr'roi 'intersting'

de cha ikenai n da yo 'You are not allowed to go out'
sore mo bikkuri dana 'that's also surprising, isn't it'

5.3.1.2. Complex assessment sequences

Straus and Kawanishi also introduce the concept of 'complex assessment

sequences,' which are intricate turn constructions resulting from participants' repetitions

or echoing and collaborative ﬁnishes. They claim that intricate sequences that do not

have a distinction between the primary speaker and listener are not seen in English or

Korean discourse. The following is an example from Straus and Kawanishi (1996).

Excerpt 5.1: Repetition/echoing taken from Straus and Kawanishi (1996: 160)

[English version of (14)]

l

2
3

(It

Hide:

Ai:

Hide:

Ai:

Hide:

Ai:

Hide:

That {ﬁre storm} wasn't small, right?

Since there was a lot of damage, wasn't there?

Yes

And then, what's worse is if it doesn't rain, then
We'll have a water shortage, right?

We'll have a water shortage, and hh? since here, it is
originally, right?

Originally, it is a desert, right?

It is a desert and it is like people force the desert into a livable
place.

Right, right, right, yeah, yeah

Excerpt 5.2: collaborative ﬁnishes taken from Straus and Kawanishi (1996: 162)

[English version of (l 5)]

16
17
18
19

20

Hide:
Ai:

Hide:

Ai:

Hide:

Ai:

Hide:

Ai:
Hide
Ai:

Hide:

Ai:

Hide:

Ai:

Hide:

Ai:

Hide:

Ai:

Hide:

Ai:

(..) But, in Japan, the panic is temble, right?

palnic]?
[the kind] after the earthquakes.
N, well, trains and things stop, right?
(H) Right [right].
[and then] a<[[commuting, or
[[(compared to) here, a little

=well, (compared to) here also, right.
we[:ll, the freeway is..a little
[well, now, commuting is, the freeway {is}
fell [down, so]

[cut down] so,
Yes. It is terrible, right? (H) but, well<
but, see, (that) was good, right?, because (it happened) in the
morning, at four [thirty, right?

[00 1 think [[(SO)
[[IF

[that (happened) around now {in the afternoon},
[=that (happened) in the evening<
now [[or something<

[at rush hour [like ﬁve o'clock ha ha

[right, right, right

Rush hour - ﬁve o'clock in the evening
ha haha haha ha

In the present study 'complex assessment sequences' are also a common

93

phenomenon among all age group participants. Excerpt 5.3 is an example from HY (37

year old)--listener, and 82 (39 year old)--the primary speaker in this segment.

Exerpt 5.3: SZ talks about an episode during her school trip for which she brought a loaf
of bread and margarine in her suitcase for a snack. After SZ '3 story, HY asks questions
how SZ packed it in her bag.

[Japanese]

1

HY: Soo iu no ni shoku pan iretetta no.
That NM-P bread put Q

2 SZ: Soo
right
3 HY: ikkinjanaku te.
One loaf NEG
4 SZ: Demo chichame no.
But small NM
59 HY: Chicchame no repetition
small NM
69 SZ: Ano te de chigireru [kuraino overlapping
ah ﬁnger with fear NM
7 HY: [Wakaru kedo domo hutsuu =
understand but normally
8 HY: nimotsu de gyuu gyuu zume ni suru kara nanka =
stuff with packed because
9 HY: maagarin toka mo tokedashi soo nandakedo,
Margarine something also melt seem though
10 SZ: Heeki dado. collaborative
all right BE
11 HY: Heeki dattanda. repetition
all right BB
129 SZ: Jikr' ga [yokatta overlapping
season SB good
13 HY: [demo komo konara oyamo oya done.
But child also child if parent also parent BE IP
14 SZ: Ee demo zettai ne,
No but absolutely IP
159 HY: Seikaidatta collaborative
right BE
16 SZ: un
yeah
[English]
1 HY: Did you put bread in that kind (of bag)?
2 SZ: (It is) so.

HY: Not a loaf (of bread)?
SZ: But it was kind of small one.

3

4

5 9 HY: kind of small one repetition

6 9 SZ: Well, kind of small that it can be pulled apart with ﬁngers
7 HY: [I understand but normally overlapping
8 HY: A bag is packed with stuff so

9 HY: Margarine and something would start melting but

10 SZ: It was all right collaborative
ll HY: It was all right repetition

12 9 SZ: The season was [good overlapping
13 HY: [But

14 SZ: but (it was) deﬁnitely, you know

15 9 HY: was right collaborative
16 SZ: yeah

HY asks the question, 'Did you put bread in that kind of one (suitcase)?, in line 1
and ' Also, it is not a loaf (of bread)?’ in line 3 to conﬁrm what she has heard. SZ in line 4
explains that the bread was kind of small so she could pull it apart with her ﬁngers. In 5
HY responds by repeating the information, 'kind of small,’ and in 7 HY interrupts SZ's
utterance by saying 'I know but normally (it is) packed with stuﬁ‘....,' in 7-9. In 10 82
concludes HY 's utterance by saying heikr'datta 'it was all right.’ HY repeats the phrase in
11. When SZ continues her turn, adding the reason why it was OK, jikiga yokatta ' the
season was good.’ HY again talks along with S2 by commenting about what SZ did in 13.
S2 takes her turn back from HY in 14 and says, ' Well, but deﬁnitely' and in 15 HY

ﬁnishes SZ'S utterance with '(it was) right' and 82 agrees.

5.3.2. Other characteristics
5.3.2.1. Non-primary speakers' questions
According to Tannen (1984), New York Jewish residents tend to ask short

questions to Show their enthusiasm and interest but such an interactional style is not

95

expected by participants who don't share the cultural background and may be
misunderstood. Questions uttered by non-primary speakers do not always ask for
speakership, but are stylistic devices to encourage the speakers to keep talking. The
linguistic forms of such listener questions are characterized as (1) high pitch, (2) rapid
rate, (3) fast pacing, (4) reduced syntactic forms, and (5) directness of content. They latch
onto the preceding utterances so there may be silence between the questions and
prolonged answers, which there may not exist between adjacency pairs (Sacks, Jefferson,
and Schegloff 1974).

Similar kind of interactions are observed in the present data in which non-
primary speakers ask questions to the primary speakers and sometimes their questions
interrupt the speaker’s utterance or sometimes overlap the speakers. These phenomena
are seen in all age groups of participants, suggesting that they are not age related but
related to individual preference and formality of conversation. The following is an

example taken from a conversation by Group I participants

Exerpt 5.4: When HW (19 years old, primary speaker) talks about her school trip to
Okinawa, SH (19 years old, listener) asks questions abruptly and the topic is shifted to
the food in Okinawa.

[Japanese]
19 SH: E nani tabetano. question 1
what ate-1P
29 SH: Nani tabetano. Okinawa. question 2
what ate-1P Okinawa
3 HW: Ano nee,
Well-1P
49 SH: Goya tabe ta question 3 (interruption)
goya ate
5 HW: Tabeta.
I ate (it).
6 HW: Demo goya rte tiara kono hen nimo hutsuu ni aru desho.

But goya QT said if this area-in normally BE TAG

7 HW: ano nigauri desho,
Ah nigauri TAG
8 SH: un
yeah
99 SH: oishii? question4
135W
10 HW: a mazui. { }
Um awful { }
11 SH: a soo { }
It is so { }
12 SH: Anmari kyoomi nakatta n da.
Much interested NEG
13 HW: Arewa ne tabekata m' yoru ne.
That-1P how to eat depend on-IP
l4 HW: Dakara tamani shokutaku ni noruno. Uchi.
So occasionally dinner table-on appear my home
159 SH: Goya ga questionS
goya-SB
16 HW: Goya. [ano kisetsu dato.
Goya [um, season BE when
179 SH: [( ? ) overlapping
18 HW: Un chikakkuni ne, budoya, budo en ga aruno.
Yeah near-1P grape farm grape farm-SB BE-IP
19 HW: Chotto kuruma de ikuto.
Short car-by go if
20 HW: Soko de goya tsukutte te
There goya grow
219 SH: Nande budoo en de goya tsukutte n no. { } question 6
why grape farm-in goya grow-1P
22 HW: { }
23 HW: Yasai tsukurujanai.
Vegetable grow TAG
24 SH: un
yeah
25 HW: Budo en dakedo yasai mo tsukutte mitaina tokoro de,
grape farm BE but vegetable also grow kind of place BE
26 SH: un
yeah
27 HW: de goya, goya katte kun no.
Then goya goya buy-1P
28 SH: un
yeah
[English]
1 9 SH: What did you eat? question 1
2 9 What did you eat in Okinawa? question 2

\OOOQOUI-bw

ll
12
13
14

15
l6
17
18
19
20
21
22
23
24
25
26
27
28

without giving a space to HW to answer between them. Her questions are reduced forms
that do not contain the particle 0 that marks direct objects. HW starts speaking by using a
ﬁller, ano nee 'well, you know' in 3 but it is interrupted by SH'S question, 'did you eat
goya?‘ HW answers SH'S question and continues her utterance about the new topic posted
by SH, goya, a vegetable that is famous as regional food of Okinawa, so we don't know
what HW tried to say in 3. HW responds that goya is not very special because we can buy
it in Tokyo, using tag questions. SH utters a backchannel, un, 'yeah' whose function is not

clear since it can be interpreted as an agreement or a continuer and asks another latched

HW:

SH:

HW:

SH:
SH:

HW:

SH:
SH:

HW:
HW:

SH:

HW:

SH:

HW:
HW:

HW
SH:

HW:
HW:

SH:

HW:

SH:

HW:

SH:

Well, you know
Did you eat goya? question 3 (interruption)
I ate (it).
But speaking of goya, we can get it around here easily.
Ah it is nigauri, isn't it?
yeah

Does it taste good? question 4
Um , (it tasts ) awful { }

It is so { }
Because you weren't interested in that much.
It depends on how to eat
Because occasionally we have it on the dinner table,
my home (I mean).

(You mean) goya? question 5
Goya, when it is in season.
[( 7 ) overlapping

Yeah near my house you know, there is a grape farm,
If we drive a few minutes.
: (They are) growing goyas there.

Why is a grape farm growing goya? question 6
{ }
They grow vegetables, don't they?
yeah
They have grape farm but also grow vegetables, too
yeah
So goya, (we) go and buy goya.
yeah

In excerpt 5.4, SH repeats the same question in lines 1 and 2 at a fast rate

98

short question, oishii '(does it taste) good?‘ about the vegetable in 9. HW answers her
quickly saying, mazur' 'awful' and both chuckle. SH comments in a full turn in 12,
showing satisfaction that her questions about goya were answered.

Then HW starts explaining why it tastes bad by saying that the dish occasionally
appears for dinner at her home in 13 and 14. SH asks another latched question to conﬁrm
what HW has just said in 15. HW answers in 16 and continues her utterance that is
supposed to follow 14 but SH talks along with HW this time in 17 in which what she says
is not clear. HW continues her utterance in 18-20 to explain how her family get goya
from a farm which also owns a grape farm and SH again interrupt HW'S utterance by
asking a question, 'why does a grape farm grow goya? in 21. HW explains why and
concludes her story of how her family purchases goya in Tokyo.

Although in excerpt 5.4 SH, the listener, keeps asking questions 6 times, SH
does not intend to hold the ﬂoor by asking questions. SH asks questions related to the
previous utterance and lets the speaker answer but she does not talk about the topic that
she introduced by herself except for the comment in 12. SH actually has only non-ﬂoor-
taking turns 4 times that consist of reactive expressions, a soo 'oh is that so,' and
backchannels, un 'yeah.’ This interaction does not present 'complex assessment sequences'
(Straus and Kawanishi 1996) since it does not contain any repetition nor collaborative
ﬁnishes; thus, it is clear that HW is the primary speaker talking about the vegetable goya
and SH plays an enthusiastic listener role.

What makes the interaction between SH and HW remarkable is the use of
questions by SH, the listener. It should be noted that in the present study that such

questions by the listener are not considered RTS so that the analyses of RTS do not explain

the interactional styles associated with listeners' questions.

5.3.2.2. Interruptions and overlapping

According to Goodwin (1986), assessments in English often appear while the
primary speakers are talking, while backchannels tend to appear at grammatical and
intonational completion. It suggests that English allows simultenious talk for
assessments. Unlike English, in Japanese, interruption and overlapping are considered to
have positive meanings (Hinds 1982). In the data of the present study, overlappings are
seen more often in informal conversations than formal conversations, as we have seen in
the two college students conversation in excerpt 5.4 where the listener interrupts the
speaker by asking questions. Excerpt 5.5 is an example in which the listener interrupts the

speaker's utterance but the speaker continues her utterance after interruption.

Exerpt 5.5: TY (Speaker, 23 years old) tells a story about a friend who became emotional
and cried at night. TJ (listener, 23 years old) interrupts while listening to the story by
telling TY about her trip and commenting.

[Japanese]
1 TY: Nanka tomodachi toka sugoi kamatte hoshii tomodachi toka=
You know friends or something very much attention want friend or something
2 TJ: un
Yeah
3 TY: = zenzen kamatte kurenakute nan nano yo mitai na kanji de=
At all don't care about me what BE-IP like
4 TY: = sugoi naiteru ko toka he
a lot cried girl or something BE
5 TJ: naite n no { } repetition
Crying-1P

6 TY: doshiyo doshiyo toka irte, minna de [

what should I do or something said everyone [
79 TJ: [gomen sonna atsuku nakatta Uchr' wa nanka itatte interruption

[Sorry that emotional BE NEG We TP kind of just

8 TY: { } Hutsu datta

100

Normal BE
Collaborative ﬁnishes

9 TJ: un
yeah
10 TY: Sono toki wa chotto tomadotta [kedo
at the time-TP a little confused but
11 TJ: aa
Uh-hum
12 9 TJ: [A, hobo zentai kodo datta karakana. Kobetsude guruupu de=
[Well, almost always a big group BE because smaller group-in=
interruption
13 TJ : =toka anma nakatta.
= or something had NEG
l4 TY: Yoru toka
night like
15 TJ: A, yoru toka repetition
Oh, night like
16 TY: Yado ni modotte kite, 5nin guruupu toka son nan de
Hotel-to return 5 person group like that BE
1 7 TJ : Naichaunda
Cries
Collaborative ﬁnishes
18 TY: Naichatte tayo. Sono ko ga atsui kara kamoshirenai kedo.
Cried .that girl-SB emotional because maybe but
19 TJ: un
yeah
[English]
1 TY: You know, I have kind of friend who wants (my) attention very much,
2 TJ: yeah
3 TY: saying you know like ' what on earth, why you don't care about me at
all'
4 TY: She kind of cried a lot
5 TJ: crying repetition
6 TY: I say like, 'what should I do, what should I do,’ [
7 9 TJ: [Sorry but we were not that emotional. We were like just interruption
8 TY: { } Normal
Collaborative ﬁnishes
9. TJ: yeah
10. TY: I was confused a little at the time, [though
11. TJ: uh-hum
12 9 TJ: [Wei], maybe it is because we went almost always as a big group. We
13 didn't have you know a chance to be in smaller group. interruption
14 TY: Like at night
15 TJ: Oh, like at night repetition
l6 TY: (When we ) return to the hotel, we have like 5 members

101

17 TJ: (she) cries
Collaborative ﬁnishes

18 TY: She did cry. Maybe it is because that girl is emotional but

19 TJ: yeah

TY tells a story about an episode to TJ and TJ interrupts TY's utterance in 7 by
saying 'I'm sorry but we were not that emotional (but we were just),' and in 8 TY
collaboratively ﬁnishes TJ’s utterance, saying futsu data 'was normal.‘ TJ'S interrupts
because She is so enthusiastic that She can not wait until TY ﬁnishes. TY doesn't mind
actually but TY co-participates in TJ's utterance by taking her turn back in 8 and she
completes TJ's utterance and keeps telling her story in 10.

TJ interrupts in 12 again when TJ starts giving reasons for why she did not have
an incident like TY. TY does not talk along with TI but continues her previous story
about her friend who cried, after TJ's interruption. TY does not answer TJ but just
continues her previous utterance in 10, 'I was confused a little at the time...
(interruption)...like at night (when we) return to the hotel, we have like ﬁve members in
our group ...... ' Then TJ collaboratively ﬁnishes in 17.

What is going on in this conversation is that TJ's interruptions do not disturb
TY's ﬂoor much since TY sometimes ignores, or acknowledges them, but takes her turn
back from TI. TJ's interruptions are not intended to claim the ﬂoor or control the topic
because when her comments are mostly ignored, she does not insist on keeping her turn,
rather she gives up easily. Therefore, her interruptions are a strategy demonstrating her

involvement in TY's story as a listener.

5.3.2.3. Use of particle ne

102

Speakers' use of ne functions as a facilitator of listener response (Uyeno 1971;
Maynard 1989; Tanaka 2000). The particle ne is more often observed in informal
conversations than in formal in the present study. This can be explained by the nature of
conversation in informal settings-informal conversations are chats between close friends

and their main purpose is thought to be to establish rapport.

Excerpt 5.6: TM (21 years old, primary speaker) talks about her school trip to Canada to
SK (20 years old, listener).

[Japanese]
1 TM: Sorega nanka nagai=

That-SB like long
2 SK: an an un un

Yeah yeah yeah yeah

3 TK: nanka kenshu mitaina

kind of training (trip) like
4 SK: an un

Yeah yeah

5 TK: kanjide= kanada itta.

Like Canada went
6 SK: Ryoko tte kanjijanain da.

Trip QT like.NEG BE
79 TK: Soda ne. Demo nanka [dakara nanka
right-1P But like so like

89 SK: [ee sugoi ii ne. assessment + ne
Wow, very nice-1P
99 TK: ne ne
IP
109 SK: sugoi ne. assessment + ne
great- IP
11 TK: Shiritsu dakara ne. ne
private school so - IP
12 SK: an an un un
Yeah yeah yeah yeah
13 TK: Joshi ko dashi ne. ne
girl's school BE- IP
149 SK: Sugoi yo ne. assessment + ne
good - IP IP
[English]

103

1 TM: It was like long
2 SK: Yeah yeah yeah yeah
3 TK: Like kind of training (trip)
4 SK: Yeah yeah
5 TK: Like (I) went to Canada.
6 SK: (It) isn't like a trip.
7 9 TK: (It isn't) so, right. But like [that is why like
8 9 SK: [Wow, very nice, isn't it?
assessment + ne
9 9 TK: It is. ne
10 9 SK: (It is) great, isn't it? assessment+ ne
11 9 TK: Because (it is) a private school, you know. ne
12 SK: Yeah yeah yeah yeahTJ
l3 9 TK: (And it is ) a girl's school, you know. ne
14 9 SK: (It is ) good, isn't it? assessment+ne

In excerpt 5.6, TK says that she went to Canada when she was in high school
and that impresses SK very much. SK gives assessment expressions such as, 'it is nice,’ 'it
is great,’ and agreement, 'yeah yeah,‘ repeatedly. In particular, the interaction in lines 7-14
between TK and SK is marked by the interactional particle ne that in this context has the
function of showing that the listener is strongly involved in the conversation and is
impressed by the story. TK's use of interactional particle ne has the function of
facilitating listener's response and rapport. This echo-like effect is a strategy to Show
'camaraderie' in Lakoffs sense (1975).

The interaction in the next excerpt is different from excerpt 5.6 because the

particle ne is used only by WB, the Speaker, but not by ST, the listener.

Excerpt 5.7: WB (speaker: 54 years old) explains why she traveled to Yugoslavia to ST
(listener: 58 years old).

[Japanese]
1 ST: Nande Yugo ni irashita no.
why Yugo to went-1P
2 9 WB: A sorewa ne. ano tamatama warashi no ie no imano ne

104

3 ST:
4 WB:
5 ST:
6 9 WB:
7 ST:
8 WB:
9 ST:
10 WB:
11 ST:
12 9 WB:
13 ST:
14 WB:
15 ST:
[English]

1 ST:
2 9 WB:
3 ST:
4 WB:
5 ST:
6 9 WB:
7 ST:
8 WB:
9 ST:
10 WB:
11 ST:
12 9 WB:
13 ST:
14 WB:
15 ST:

um that-TP-IP um coincidently my house of now

an
yeah
sunderu uchino
living house
an
yeah
chikaku ga Yugosurabia taishikan ga ne. atte ne. ne
near-SB Yugoslav embassy-SB - IP BE- IP
aa
ah
Sokono taishikanin to warito nakayoku natte,
that embassy person with relatively close become
aa aa
Uh-hum uh-hum
de sono kata ga iru uchi ni ,
Then the person-SB BE while
aa aa
Uh-hum
maa Yugo ni ne [kaette ne, ne
I mean Yugoslavia-in-IP return-1P
[aa aa

Uh-hum uh-hum
ichido irashitara tte [iwarete itta n desu.
Once visit if QT being said went
[aa
[Uh-hum

Why did you go to Yugoslavia?
Well it is because, you know, ah, coincidentally, my house, where
now

yeah
I live, near my house
yeah
Yugoslav embassy, you know, is there, you know.
ah
I got to know someone who works in the embassy
Uh-hum uh-hum
Then while the person is there
Uh-hum

I mean in Yugoslavia [After he went back, you knnow
Uh-hum uh-hum
I was invited by him to visit him in Yugoslavia so I went.

[Uh-hum

105

In excerpt 5.7, WB starts her utterance by answering ST's question of why WB
traveled to Yugoslavia. While WB explains the reason in lines 2 to 14, ST only uses
backchannels and no other kinds of RTS. In other words, although ST participates, WB
intends to make sure if ST follows or understands and claim the continuation of WB's

speakership by using sentence ﬁnal ne.

5.3.2.4. Use of masculine and feminine forms

It is reported that young female Japanese use not only less feminine language
but also use masculine forms (Okamoto 1995). In particular, young female Speakers' use
of masculine forms demonstrates their identities and is exclusively used in in-group
conversations, but not in the formal settings such as in classroom situations.

This section will examine how gender speciﬁc expressions or language are used
by participants in the present study and will explain the functions that gender speciﬁc
forms may have in conversations between female friends.

In the present data most of the informal conversations are carried out using plain
forms which are informal most of the time; there are a limited number of polite forms
which are formal, and some honoriﬁc forms which can be used with plain and polite
forms. Gender can be marked lexically or by using particles such as, no, yo, wa, and ne.
These particles can follow RTS in listener responses (e.g. ii wa ne 'it is nice,' so ne 'right,'
and so yo 'right,' etc.) The most commonly observed forms are neutral forms which can

be used by both genders.

Masculine forms

106

The results do not contain any masculine forms of RTS by listeners; however,

some of the speakers use masculine forms as in excerpt 5.8.

Excerpt 5.8: MT told IN that she heard students were not allowed to shampoo their hair
since they caused some problem at the hotel.

[Japanese]
1 IN: demo natsu ni itte arae nai no iya da na.

But summer-in went wash-NEG-NM horrible-BE-IP
29 MT: tte iu ka yama nobori sasennayo.

QT say mountain climb make NEG-1P

3 MT: dattara tte kanji da yo ne.
then QT like BE- IP- IP

4 IN: { }

5 MT: yama itte ase kaite doro darake ni natte =
Mountain go swet get muddy covered become

6 MT: = ohuro wa haire naino mitaina [kanjijanai

Bath-TP take NEG like sound TAG
7 IN: [so so so]
right right right
[English]

1 IN: But it is annoying if I can't wash my hair in the summer
29 MT: Don't let climb mountain then
3 MT: It is like that, isn't it?
4 IN: { }
5 MT: Going to the mountain, and getting sweaty and muddy
6 MT: But you can't take a bath kind of thing, isn't it?
7 IN: right right right

MT and IN went to the same middle and high school in Tokyo known to be a
very conservative school and now IN is teaching at the school. When they talk about their
memories of school days, they speak very casually. The masculine forms used by MT

might demonstrate their solidarity from Spending their teenage time together by Showing

strong objection to the fact that students are not allowed to wash their hair after outside

107

activity. This is a good example of Okamoto's notion that young female speakers,
especially high school age girls, use masculine forms to express their identity in the
group they wish to belong to. Also, this does not Show socio economical strata since
participatns who use masculine forms in the present study obviously are the members of

upper-middle or higher class based on the private school they attended.

Feminine forms

The data in the present study do not contain many feminine forms and all of the
feminine forms are used by participants of the oldest group whose age is over 50. In
excerpt 5.9 ST asks the question, "What were you doing in England?" using an honoriﬁc
form and the particle no, which sounds feminine and polite. In 3 maa, 'oh my,‘ is
considered to be a resumptive opener which appears turn initially. This is followed by
gakko itta no ne "you went to school, didn't you," containing particle no, which can be
used by both genders; when it is used with sentence ﬁnal particle ne, however, it is

considered a feminine form.

Excerpt 5.9: WB (54 years old) talks about how she stayed in England longer than other
countries when she went to Europe and then ST (58 years old) asked what WB did there.

[Japanese]
19 ST: E igirisu de nani shiterashita no.
Well England-in what do IP
2 WB: Igirisu wa chanto ne, eego no benkyo.
England-TP IP English of study
39 ST: maa gakko itta no ne.
Oh my school went - IP-IP

4 WB: so so { }
right right { }
5 WB: samaa koosu no sa,

summer course - IP-IP

108

6 ST: ha= ha=

yes yes

7 WB: 2 kagetsu no. dakara 6 gatsu kara=
2 months NM So June-from
8 WB: =8 gastu gurai made.
August about- to

9 ST: haa haa so

yes yes it is so.
[English]
19 ST: Well, in England what were you doing?
2 WB: In England I properly, you know, studying English
39 ST: Oh my, you went to school, didn't you?
4 WB: right right { }
5 WB: I took a summer course
6 ST: yes yes
7 WB: 2 months (course). So from June
8 WB: until about August
9 ST: yes yes it is so.

Both masculine and feminine forms are rarely used in formal conversations in
the present study, in which neutral polite expressions are used by most of the participants,
except for some of the participants of the older groups. The use of masculine forms by
young female Japanese is considered to be in-group language that can express solidarity
among participants (Okamoto 1995). Thus, participants in excerpt 5.8 intend to create
solidarity by using masculine forms. For similar reasons, older participants may use
feminine forms to express their friendliness or solidarity and the informality of

conversation allows them to do so.

5.3.3. Interactional styles in informal conversation
We have discussed the interactional characteristics exhibited in the informal data
of the present study. Interactional styles are diverse, in contrast to formal conversations,

resulting in a high proportion of various verbal RTS that make structures complicated. In

109

this sense, all individual styles reﬂect the cultural norms of Japanese interactional style;
that is, listeners have supportive, empathetic, and cooperative attitudes toward the
speakers and participate in the conversation actively as listeners.

The ﬁndings suggest that although the norm is right basically, the actual
interactions demonstrate variations according to the degree of empathy and in terms of
the strategies participants use to achieve their communicational goal of solidarity. The
listening styles can be classiﬁed into three types based on the linguistic strategies used.
The following summarize the three most commonly observed interactional styles and

their characteristics.

1 Rapport style: listeners show supportive attitude and acknowledgement
constantly. They tend to use backchannels and repetitions frequently but
remain as passive listeners

2 High involvement style (complex assessment sequences): interruption,
repetition, reactive expressions, collaborative ﬁnishes and laughter are
preferred rather than backchannels.

3. Machine-gun questions: frequent turn Shifts resulting from Short questions
but listeners do not intend to take the ﬂoor.

The ﬁrst type, rapport style in informal conversation, is characterized by a
relatively higher proportion of backchannels and lower proportion of other RTS and nods.
This kind of interaction is seen in formal conversations between older participants who
used more verbal backchannels and fewer nods. In informal conversations, 3 out of 13
participants are deﬁned as this type, namely, YG in group I, TY in group II, and AR in
group III.

The second style is characterized by frequent use of verbal RTS other than

110

backchannels which differentiate it from the rapport style. Since the interactions of high
involvement type contains frequent use of verbal RTS, such as repetitions, reactive
expressions, and collaborative ﬁnishes, there may not be a clear distinction between the
speaker and the listener. There are 4 or 5 participants who are grouped in this style in the
present data.

The last type shows frequent questions uttered by listeners that are not counted
as RTS in the present study since they are considered as full turn shifts. This type of
interaction may have a lower rate of verbal RTS compared with the ﬁrst two types and
gives an impression of being obtrusive 0r interruptive. There are 5 participants who use
these kinds of questions as listener responses.

It should be noted that the distribution of these styles is not related to age--in
fact, the data reveal that participants in all age groups present all three styles. Therefore,
there must be individual differences in preference for certain style resulting from sub-
cultural factors (Tannen 1984) or psychological factors, such as the degree of

involvement in the conversation but not power relations nor solidarity.

5.4. Summary

This chapter examines the use of RTS in informal conversation, i.e., chat
between close female friends in different age groups. The results reveal the common
interactional styles among participants in all age groups. First, nods are not used
frequently; second, verbal RTS other than backchannels are used much more frequently
than in formal conversations. Finally, the frequent use of verbal RTS creates a variety of

interactional styles such as the rapport style, high-involvement style, and the machine-

111

gun question style.
The next chapter will demonstrate the structural differences between formal and
informal conversations based on the results reported in Chapters 4 and 5 and discuss

social factors that inﬂuence such differences.

112

CHAPTER 6

DIFFERENCES IN STRUCTURE
BETWEEN FORMAL AND INFORMAL CONVERSATION
6.1. Introduction
This chapter summarizes the ﬁndings on the use of RTS in formal and informal
interaction in order to compare the interactional type of the two styles. Variations in
interactions in formal and informal conversation are discussed and explained by
considering social factors such as age, relationship between interlocutors, formality level,

and psychological distance. Hypothesis 3 will be discussed at the end of this chapter.

6.2. Quantitative results of formal and informal conversation
This section summarizes the results of quantitative analyses of the use of RTS in
formal and informal conversation presented in Chapters 4 and 5 in order to compare the

characteristics of each style of conversation.

6.2.1. Frequency of reactive tokens

Table 6.1 shows the total number of nods and verbal RTS used during 60 minutes
of formal and informal conversation. The data on informal conversation was modiﬁed to
60 minutes for comparison.

Table 6.1. Number of nods and verbal reactive tokens used in 60 minutes of conversation

 

 

 

 

Formal conversation Informal conversation
Verbal RTS 853 1099
Nods 751 167
Total 1604 1267

 

 

 

 

 

113

In a total of 60 minutes of formal conversation, overall listener responses are
more abundant in formal conversation (1604 tokens) than informal (1267 tokens).
However, in formal conversation there are 853 tokens of verbal RTS compared to are
1099 tokens in informal conversation. Speakers in informal conversation receive
approximately 1.3 times the number of verbal listener responses. On the other hand, non-
verbal RTS, such as nods, are much more frequently used--4.5 times more--in formal
conversations (751 tokens) as opposed to informal conversations (167 tokens). These
results indicate that the way in which listeners participate in conversations iS different in
formal and informal conversations. This chapter will discuss how the listeners' choice of
verbal or non-verbal strategies could be determined and the social factors that may

inﬂuence their decision.

6.2.2. Verbal reactive tokens and nods

The ﬁndings indicate that the total number of RTS used by listeners during 2
minutes of instructions does not vary among participants but that the balance between
verbal RTS and nods varies among participants.

Table 6.2 displays the results of regression analysis of verbal and non-verbal RTS
in formal conversation of the present study. There is a negative correlation between
verbal RTS and nods. This relationship is statistically signiﬁcant. The result conﬁrms the
Pearson Correlation result and provides more reliable statistical evidence regarding the
relationship between verbal RTS and nods. Participants who use more verbal RTS use
fewer non—verbal RTS (nods) and vice versa. In other words, verbal RTS and nods can be

used interchangeably as listener responses in Japanese formal conversation. The

114

frequency of verbal RTS used by someone listening to instructions from the researcher for

two minutes can be predicted from the number of nods he or she uses.

Table 6.2. Results of regression analysis between verbal reactive tokens and nods

 

 

 

(r2 = 70.7)
Predictor Coef. St Dev Probability
Constant 51.615 3.494 0.000
nods -0.95 75 0.1 186 0.000

 

 

 

 

 

 

In informal conversations, verbal RTS and nods do not correlate as seen in
formal conversation. One of the reasons that the data from informal conversations are not
comparable with the data from formal conversation is that the speakers in informal
conversations and the content of each informal conversation vary from pair to pair, while
they do not in formal conversations. Therefore, it may not be relevant to generalize the
results obtained by formal conversation as rules for verbal RTS and nods in face-to-face
conversation. However, the ﬁnding of the present study suggests the possibility that
verbal RTS and nods may have a negative correlation if listeners hear the same story from
the same Speakers in informal settings.

Furthermore, there is an implication that there exist fundamental differences in
interaction between formal and informal conversations with respect to the balance of
verbal and non-verbal RTs; Listeners in the informal conversations use nods less
frequently than in formal conversations (751 times vs. 167 times) and use more verbal

strategies in informal conversations.

6.2.3. The nature of verbal reactive tokens in formal and informal conversation

115

Figure 6.1 (modiﬁed ﬁgure 4.1) presents the average individual participants'

ratio of backchannels, other verbal RTS and nods in formal and informal conversation.

Figure 6.1. Interactional styles in formal and informal conversations

 

100%
90%
80%
70%

 

rn
::
ID
x
.9
o
.2
‘g 60%
3 50%
Co-
0 40%
a) 30°/
8 0°
0 202.
g 10%
‘3‘ 0%
formal (high nods+low formal (10w informal (low
verbal) nods+high verbal) nods+high verbal)

Conversation styles and frequency of reactive tokens

statuses concealment?

 

 

 

In formal conversation participants can be grouped into one of these types: high
frequency of nods and low verbal RTS, or low frequency of nods and high verbal RTS,
whereas in informal conversation the proportion of 3 types of RTS is similar among
participants. Although the proportion of overall verbal and non-verbal varies in the two
types of formal conversation, the percentage of verbal RTS excluding backchannels is
relatively small. This indicates that in formal conversations listener responses have less
variation compared with informal conversation so that listeners are expected to respond
by using mainly either backchannels or nods.

In informal conversations, approximately half of the verbal RTS are

116

backchannels and half are verbal RTS other than backchannels. This implies that listener
responses in informal conversations are not only verbal, but are also more dynamic,
lively and interactive through the frequent use of reactive expressions, collaborative

ﬁnishes, repetitions, laughter and so on.

6.3. Age factor and reactive tokens

This section summarizes the ﬁndings about variations in use of RTS with respect
to age and conversation styles, both formal and informal. We have seen that there is a
variation in listener responses in terms of how many verbal or non-verbal devices
listeners use and the extent of verbal RTS (beside backchannels) used to communicate.
The present study examines age and style as two social factors that inﬂuence variation in
listening behavior. The following is a summary of what was found in the present study.

First, age and total RTS do not have any correlation in either formal or informal
conversations. Therefore, age only inﬂuences the use of verbal RTS, nods, and some types
of verbal RTS but not the total amount of RTS.

In formal conversations there is a negative correlation between age and nods.
Table 6.3 shows the regression results. Age and nods have a statistically Signiﬁcant
relationship. Younger participants tend to use a greater proportion of nods than older

participants.

Table 6.3. Results of regression analysis between age and nods in formal conversation

 

 

(r2=2l .1)
Predictor Coef. St Dev Probability
Constant 46.014 4.674 0.000

 

nods total -O.4256 0.1586 0.012

 

 

 

 

 

 

117

On the other hand, age and verbal RTS have a positive correlation [r=.545,
p=.002] as shown in table 6.4. The number of verbal RTS used by participants is

indicative of the participants' age. This relationship is statistically Signiﬁcant.

Table 6.4. Results of regression analysis between age and verbal reactive tokens

 

 

 

02:24.9)
Predictor Coef. St Dev Probability
Constant 24.122 4.393 0.000
Verbal RT total 0.4061 0.1359 0.006

 

 

 

 

 

 

A Similar relationship between age and verbal RTS is also seen in the relationship
between age and backchannels, which seems consistent considering that backchannels are
the most frequently used verbal RTS. There is a positive correlation between] age and

backchannels. This relationship is statistically signiﬁcant.

Table 6.5. Results of regression analysis between age and backchannels

 

 

 

 

(r2=25.4)
Predictor Coef. St Dev Probability
Constant 23.949 4.392 0.000
[ackchannels total 0.5062 0.1669 0.006

 

 

 

 

 

Among various types of backchannels, some backchannels seem to be
inﬂuenced by age, too. For instance, ee, 'yeah,' which is a less formal 'yes' but not as
informal as un, has a tendency to appear more frequently in conversations between older
participants but rarely between younger participants. This could be the result of the

formality of ee or stylistic differences between generations.

118

01t was also found that age can explain some of the variations characterized by
the use of some types of verbal RTS. There is one type of RTS that is more frequently used
by older participants, that is, resumptive openers, which are non-lexical items uttered in
turn-initial positions or which precede reactive expressions, as in 'Oh, it is nice.' The
tendency observed in older participants suggests that they are more verbally active than
younger participants, or that some older participants attempt to take turns even in formal

situations such as listening to instructions.

Table 6.6. Percentage of ee 'yeah' and resumptive openers by backchannels in formal
and informal conversation

 

 

 

Formal conversation Informal conversation (%)
ee 'yeah' 14.8 % of backchannels 21.5 % of backchannels
resumptive 26.8 % of RTS excludes 25.2 % of RTS excludes
o ener backchannels backchannels

 

 

 

 

 

Unlike in formal conversations, the use of RTS does not have a correlation to age
in informal conversations. Figure 6.1 demonstrates that listener responses in informal
conversations consist of a low occurrence of nods (11%), backchannels (45%), and other
types 0f RTS (44%). Each type of RT and its subcategories have been tested to determine
Whether any of them have a correlation with age, but none Showed a statistically
Signiﬁcant effect. This suggests that when people chat with close friends of the same age
group, there is little variation in listening behavior with respect to the frequency and
types 0f RTS among participants. The remainder of this chapter examines listening
behavior in terms of interaction in order to explain variations in listening behavior in

different contexts.

119

6.4. Interactional styles

This section summarizes the diﬂ‘erences in interaction between formal and
informal conversations in the present study. The previous section revealed the
quantitative results of overall frequency and how subcategories of RTS are interrelated.
However, quantitative analyses alone may fail to explain how listeners participate in
conversation when listeners use strategies other than RTS. Chapter 5 presented qualitative
analyses of interactional styles observed in informal conversation. This section reviews

them by comparing interactional Styles between formal and informal conversation.

6.4.1. Listeners' strategies

Listeners' behavior both in formal and informal conversation in the present study
can be accounted for by the rules of politeness by Lakoff ( 1975). Lakoffs notion of
'Rules of Politeness (Rules of Rapport)‘ consists of three rules.

1. Don't impose (Distance)

2. Give options (Deference)

3. Be friendly (Carnaraderie)
According to Lakoff, a Speaker's choice of linguistic forms observes one of these rules to
fulﬁll the social function of language. For example, the ﬁrst rule applies in formal
Situations better than in informal situations. The third rule would apply in informal
situations more than in formal. I propose that these rules can apply to the listeners' choice
of linguistics form or interactional styles in the present study data.

The ﬁrst rule of 'Distance' is employed in formal conversations in which a

researcher that listeners don't know well gives instructions on what to do. Listeners

Should listen carefully; at the same time they should be polite since they don't know each

120

other. In that situation participants in the present study employ a strategy of using non-
verbal RTS which are less imposing to a different degree and using formal types of
backchannels such as hai 'yes,' rather than more casual tokens such as un, 'yeah.' Another
strategy observed is that listeners use backchannels mostly in their choice of verbal RTS,
which means that they try to refrain from being imposing when they listen. As a result,
RTS such as repetition, collaborative ﬁnishes, resumptive openers, and short comments
are rarely used Since those RTS make interaction dynamic and it is difficult for listeners to
maintain 'Distance’ from the speaker.

However, listeners can also use the rule of 'Camaraderie' even in formal
situations. For example, laughter is one of the strategies that all the participants can
employ to achieve 'Camaraderie' since it is less imposing and yet can contribute to
creating a friendly atmosphere by not saying anything. Laughter RTS are used by younger
participants more than older participants, as discussed in Chapter 4.

If listeners use informal RTS such as an 'yeah, in formal situations where socially
appropriate forms are polite forms, hai 'yes,' in backchannels, or desu 'be' in copula, the
responses have the effect of displaying 'Camaraderie' with speakers. However, this
strategy is normally available only for listeners whose status is equal to speakers or who
have a relatively higher status (power), but not for lower status. The choice of verbal RTS
other than backchannels is also a 'Camaraderie' strategy and works in the same way.

In informal situations such as chatting with close friends, the rule of
'Camaraderie' is the most important rule in the system. Strategies for how to achieve
'Camaraderie' may differ among languages and cultures. The present study focused on

Japanese women's interactions and the next section discusses different strategies

\

121

employed to follow the 'Rules of P01iteness.’

6.4.2. Three commonly observed styles in Japanese listening behavior
The following are three types of listening behavior observed in the present study

that is discussed in chapter 5.

1. Rapport Style: listeners show supportive attitude and acknowledgement constantly.
They tend to use backchannels most frequently among other verbal RTS
or nods.

2. High involvement style: interruption, repetition, reactive expression, collaborative
ﬁnishes and laughter are preferred rather than backchannels. Complex
assessment sequences may be observed.

3. Machine-gun questions style: frequent turn shifts resulting from short questions but
listeners do not intend to take the ﬂoor.

The two types of interaction of formal conversation in the present study, that is,
high nods+low verbal and low nods+high verbal, are considered to be rapport style.
There is no participant whose listening style was identiﬁed as high-involvement style or
machine-gun question style in formal data.

In informal conversation, all the interactions are identiﬁed as one of the styles
according to the rate and types of RTS used by participants and number of questions asked
by listeners. Table 6.7 shows the interactional styles, conversation types, and participants'

age.

122

Table 6.7. Interactional styles and participants in foramal and informal conversation

 

 

 

 

Formal conversation Informal conversation
n =30 Participants' ID (age) 11 =13
Rapport style All the participants OS (19), SK (21), SZ (39),
._ 22...... ... ..(Age 19:91.2 MR(58)AR(60) - _ .‘
High-involvement style N/A TI (22), MT (31), IN (31), HY (37)
Machine - gun style N/A SH (19), TY (23), KM (54),
ST (5 8)

 

 

 

 

6.5. The 'multidimensional model' in relation to social factors
This section reviews the notion of the 'multidimensional model' (Tannen 1993)
and explains how it can account for the different interactional styles in formal and

informal conversation in the present study.

6.5.1. Formal conversation model

Chapter 4 demonstrated how variation in listener responses in formal
conversation is explained in the multidimensional model. Dimension A represents a
hierarchical yet close relationship observed in parent-child, or amae, (Lebra 1976),
typical of Japanese employer-employee relationships. Dimension B shows that there is
hierarchy and distance in the relationship between participants: American employer-
employee relationships. The formal conversation situation in the present study Should be
placed in B in most cases. However, if participants are the same generation as the

researcher or older, the hierarchy level may drop into dimension C as we discussed in

Chapter 4.

123

Figure 6.2. Contexts of formal and informal conversation

 

 

 

 

 

 

 

 

hierarchy
Eistening to instruction]
Different in power + close E] different generation + distant
e.g.) Japanese: employer - employee e.g.)American:employer - employee
Parent - child Teacher - student
closeness distance
[Chatting with friends] [Listening to instructions]
Equal + close [Q same generation + distant
e.g.) American: Siblings
equality

The ﬁndings of the present study demonstrate that there is an age-graded
variation in the balance of verbal RTS and nods. Namely, younger listeners tend to use
fewer verbal RTS than older listeners, and more nods than older listeners, This
phenomenon is parallel to the hierarchy axis. That is, the number of nods and verbal RTS
change on a hierarchy continuum. In dimension B and C contexts, participants can also
employ strategies to change the distance on a solidarity continuum according to their
emotional state or affect toward interlocutors in talk-in-interaction. For example,
laughter, used relatively frequently by young listeners in formal conversation, contributes
to Showing their involvement and creates a less tense atmosphere even while they listen

to instructions.

6.5.2. Informal conversation model

The informal conversation of the present study consists of chats between close

124

friends in an informal setting, so that the relationship between participants is equal but
not hierarchical, which satisﬁes the requirement of dimension D.

The results Show that the 13 participants who play primary listener roles
behaved similarly in terms of overall interactional types in ﬁgure 6.2. That is, listeners
use verbal RTS to a greater extent (88%) and use fewer numbers of nods (11%). The
greater proportion of verbal RTS to backchannels than in formal conversations provides
evidence that listeners co-participate in informal conversations more actively than in
formal conversations.

The analysis of formal and informal conversation suggests that the same
listeners participate in conversation using different strategies as a listener depending on
the context to convey social meanings. Yet these individual behaviors are governed by

rules, such as 'Rules of Politeness.’

6.5.3. Japanese listening behavior

This section provides an explanation of how power and solidarity inﬂuence the
manipulation of different contexts through the use of listener strategies discussed in the
previous sections.

The following ﬁgure 6.3 is a model of Japanese female listening behavior based
on the ﬁndings of the present study. Since the present study lacks data which account for
the contexts in dimension A, the analyses of Okamoto and Sato (forthcoming) will be
adopted as data in dimension A.

Okamoto and Sato (forthcoming) challenge the stereotyped cultural view about

Japanese language, especially Japanese women's speech, that says that Japanese listeners

125

are supportive, cooperative and empathetic. Their ﬁndings show that Japanese mother-
daughter interaction (dimension A) does not exhibit the cultural norm at all in terms of
use of listener responses. On the other hand, daughter-friend (same sex) conversations
(dimension D) present similar types of interaction as the present study.

Their analyses are limited to the verbal RTS so that it does not provide evidence
about how nods are used in both situations. However, it is reasonable to assume that there
may not be many nods in dimensions A and D from the results of the present study. The
reason why I speculate that there are few nods in dimension A is that there is no need to

Show 'Distance' nor Show 'Deference' in such a close relationship.

Figure 6.3. A model of Japanese listening behavior

 

 

hierarchy
Low verbal + low nods Low verbal + high nods
(High backchannels &
low assessment)
E
closeness distance
[E
High verbal + low nods High verbal + low nods
(Low backchannels & (High backchannels &
high assessments) high assessment)
equality

Figure 6.3 is a model of listening behavior in Japanese face-to-face female

conversation. Each dimension is characterized by speciﬁc listening behavior: (A) low

126

verbal RTS and possibly low nods, (B) low verbal RTS and high nods, (C) high verbal RTS
and low nods, and (D) high verbal RTS and low nods. Listening behavior of each
dimension is determined by the relationship between two continua, which moves to either
side depending on the contexts. Power and solidarity relations in a context is not static
but rather dynamic in talk-in—interaction. For example, listeners may Show 'Camaraderie'
by using the features of dimension D or listeners also can be 'Distant' by just nodding,

which is typical in dimension B.

6.6. Hypothesis 3
This section explains how ﬁndings of the present study can test hypothesis 3.
Hypothesis 3: The formality level of conversation inﬂuences not only the choice of
RT types and frequency, but also the interactional styles.
This chapter demonstrates differences in listening behavior in formal and
informal conversations to test hypothesis 3. The ﬁndings show that there are signiﬁcant
differences between formal and informal conversations in terms of types of RTS, the ratio

of RTS, and interactional styles. Therefore, hypothesis 3 is supported.

6.7. Summary

This chapter summarized ﬁndings in terms of the inﬂuence of style of
conversation on the use of RTS and interaction styles. The present study provides only a
tentative proposal; thus there is a need for further investigation in the ﬁeld such as the
analysis of interactions characterized by the features of dimension A. However, the

results have a number of implications for the study of listening behavior.

127

CHAPTER 7

RESULTS OF THE INTERVIEW AND PERCEPTION STUDY
7.1. Introduction
This chapter presents the results of the perception study and follow up interview,
and investigates how they can account for the ﬁndings of actual listening behavior in this
study. I ﬁrst demonstrate what participants know about Japanese listening behavior from
what they said in the follow up interview. Next, I will analyze how perception of video
clips of different listening behaviors reﬂects the actual listening behavior of participants

in informal conversation.

7.2. Follow up interview about listening behavior
This section provides evidence collected by interview of what participants know
or how they perceive Japanese listening behavior in general. Only English translations of

what participants said in Japanese will be presented.

7.2.1. Difference between formal and informal conversation

Chapter 6 summarizes the ﬁndings of listening behavior in formal and informal
conversations by comparing the frequency of RTS, types of RTS, and the interactional
styles in two types of settings. In formal conversation in which the researcher gave
instructions to each participant face-to-face, participants' listening behavior varies in the

ratio of verbal RTS and nods, and displays age-related variations. 1 have argued that this

128

phenomenon Should be explained by the difference in power between the researcher and
participants, since the relationship between the researcher and each participant is the
same among all participants. The following comments may be helpful in understanding
the participants' inner feelings when they were given instructions. SZ (19 years old) said:
(1) SZ: I was kind of anxious about what we are going to do and it is
because I didn't know anything, so that probably appeared in my
responding with RTS and behavior and so on. It was by listening that I
obtained information, wasn't it?

In (1) SZ retrospects that she was nervous so her anxiety could have been
reﬂected in the way she listened. SZ states that she had to listen carefully since that was
the only way to obtain information. TY (23 years old) analyzes her listening behavior in
formal and informal conversations in (2).

(2) TY: When listening to instructions, I don't use many backchannels, I
feel like I don't nod that frequently either.

She thinks that She does not use many backchannels, either verbal or non-verbal.
In formal conversation she used verbal RTS (4 backchannels and 3 laughter) and nods 46
times in 2 minutes. Whereas, she used no backchannels but other type of verbal RTS 20
times, and nods 3 times in 2 minutes informal conversation. Her comments about her
behavior reﬂect what she actually did in the real converation.

In (3) participant ST in group III (58 years old) mentioned what could have
inﬂuenced her listening behavior in the instruction session with the researcher. ST said:
(3) ST: If a listener is interested in what a Speaker says, she asks

questions while she listens. For example, when I talk with you like this
or listen to the story of your university as I did awhile ago, I wonder if I

129

wouldn't respond to you with just yeah, yeah (if I am interested).

According to ST, whether a listener is interested in the story or not determines the
interactional style. Even in the formal conversation of the present study, ST was not
inﬂuenced by the power of the researcher who was younger (ST: 58 years old, researcher:
41 years old) but possibly can exercise relative power as an interviewer in the formal
setting. ST's comments are consistent with her actual performance (i.e. low nods and high

verbal RTS with frequent speaker changes that are excluded from the data).

7.2.2. Types of verbal reactive tokens

In both formal and informal conversations, ee, 'yeah', which is less formal than
hai, but more formal than an, 'yeah,' is used almost exclusively by participants of groups
II and III as discussed in chapter 6. The following dialogue explains how the age-related

preference occurs.

SB and S2 (both 20 years old) were asked if they use ee ‘yeah’ by the researcher.

(4) M: Do you say ee 'yeah?’

SB: [1 don't.

SZ: [1 don't.

M: How does it sound?

SB: It sounds like good breeding. No but we don't see someone who says ee

so on much on campus, do we?
M: What about in job interviews?

SB: 1 don't say that, no.

130

M: What about your mother?
SB: When she calls, she says that, you know?

SZ: Yes, she does.

The conversation in (4) shows their perception toward people of same generation
who use ee ‘yeah’ in their speech. According to them, this linguistic form expresses not
only a different degree of formality for 'yes' but also expresses an image associated with
higher social status. Although these well-educated young participants don’t prefer the
image attributed by the use of the form, they both agree that their mothers use the token,
which is consistent with the results of the present study.

Their comment tells us that participants of group I may not wish to be seen as
well bred by using ee in both formal and informal contexts. However, some participants
of group 11 use ee in formal contexts but not in informal contexts. Participants of group
III who belong to the youngest group’s mothers' generation use ee in both formal and
informal contexts. This phenomenon may be explained as an example of 'hypercorrect
use' (Ide 1982; Reynolds 1985) and the younger generation in the present study is not
inﬂuenced by the covert prestige of ee. The results also suggest that the use of ee in
formal contexts may be developmental in Japanese women's language so that there is a
great chance that participants such as SB and 82 adopt the convention when they become

a certain age.

7.2.3. Perception of nods
One of the remarkable differences in listening behavior between formal and

informal conversations is in the use of nods as RTS.

131

The researcher asked a general question, 'When do you use nods?’ SB (20 years
old) and S2 (20 years old) say in (5) and (6) respectively:
(5) SB: Like in interviews, an interview for a part time job or something I
nod a lot, right? Then all of a sudden, I realize I should speak up and
then I say something. So what is it? It is an occasion in which I stand on

ceremony.

(6) SZ: Toward superiors. When we have to Show respect to a superior or
someone.

SB and SZ analyze when they use nods. SB points out that formality level may
inﬂuence her use of nods, that is, when she needs to "stand on ceremony," SB uses more
nods. SZ analyzes the fact that She uses nods when there is status difference between her
and her interlocutor and she is the one who has the lower status. Their comments account
for the younger group of participants' frequent use of nods in the formal situation.

In addition, the following comment by SB implies that listeners are consciously
changing their listening behavior according to the purpose of communication.

(7) SB: In order to make a good impression, I speak up properly and nod
like when I have an eye contact; otherwise, when the speaker looks
down and says something he/she doesn't know I nod, so I may
consciously try to say 'yes' or something.
What SB (20 years old) relates is the strategy of making a good impression as an
interviewee. Such a strategy of being a successful interviewee represents the cultural
norm expected of her as a young female Speaker of Japanese.
In informal conversation, OS mentions that she nods both accompanying verbal

RTS and not accompanying them when she gets excited while conversing with someone.

132

(8) OS: I am told that I nod a lot, though.
M: Who says so?
OS: Friends or someone. When I listen, she/he says that I nod so
often { }. Well, maybe when I am excited, I say yeah, yeah (nods) like
this. It seems that becomes continuous.
OS'S partner in informal conversation also agrees with OS although it is not shown in (8).
In the present study nods that accompany verbal RTS are not counted Since such nods
function to emphasize verbal RTS and have no other function. OS'S comment provides
evidence that nods accompanying verbal RTS occur to express listeners' stronger
involvement in the conversation.
Another piece of evidence that nods are an indirect way of expressing listeners'
understanding or interest is seen in MD's (37 years old) and YG's (19 years old)

comments. In (9) MD, who teaches at a junior college, comments on nodding based on

her informal observation in the classroom.

(9) MD: I look in the direction of students who nod often. There are Miss
Noddings, you know. I guess I look at those who are nodding naturally
when I notice.

(10) YG: In class there are some people who nod to Show that they
understand when a professor says something, aren't there? I
occasionally do that but not always. I think they understand what the
professor says.

(11) OS: I don't use (nods) in the classroom. I think maybe nobody notices
that I nod. (Because) I don't nod often in class.

MD'S and YG's comments show that both instructors and students understand the

messages conveyed by nods in the classroom situation and use them as a communication

133

strategy, although the actual use of nods varies from person to person. For instance, in

(l 1) OS claims that She doesn't use nods in the classroom. If the listener does not react

properly as expected by the speaker, how does the speaker perceive or understand the

message of the listener’s lack of response? Following comments may answer the

question.

MD'S further observation in (12) suggests that there is a generational difference

in the use of this strategy in the classroom situation.

(12) MD:

Nodding? Yeah, there are students who nod. They nod frequently. So if
I present a lecture to adults, they nod more, you know. It is very clear.
Ah, I can tell that they are concentrating on listening very much...
Well, the rate of nodding of adults, compared with students, well, they
are adults you know, not only is it more frequent... But anyway, adults
nod a lot, that's what I thought at the time.

TM (21 years old) talks about how she feels when she doesn’t get the expected

reaction from the professor when She speaks up in the class:

(13) TM:

When I speak up, the professor does not nod, but stares at me in my
eyes so that I feel like he does not understand or maybe I am saying
something wrong, so, well, I want her/him to express their
acknowledgement, like 'right.’ ...When I ﬁnish, she/he says I am wrong,
then I think, 'Why didn't you Show it by your facial expression at the
point?’ ...Like the frown that teachers or someone gives, frowning
when students' answer is wrong.

TM's comments in (13) shows that she believes that the lack of listener's

response functions as disagreement and speakers are always observing listeners'

responses, including non-verbal features such as smiling, to ﬁgure out listeners'

understanding or attitudes toward what they say.

This section presents an overview of how listeners use nods as a communication

134

strategy in different contexts, and examines what listeners actually do by the use or non-
use of nodding and other non-verbal RTS and how Speakers interpret the listeners’

response.

7.2.4. Interactional style of informal conversation

Listening behavior in informal conversation is characterized by greater
proportion of verbal RTS than nods, frequent use of RTS other than backchannels, and co-
participation of listeners. The interaction of informal conversation tends to consist of
Shorter turns and be abundant in lively expressions. Some of the participants used the
word 'overreaction' to express their interaction between close friends.

(14) SZ: She (in the video) should react to the speaker in a more
exaggerated manner, overreaction, like 'are you serious?’

82 explains the way she talks with friends using the listening behavior in the
video as an example. According to her, overreaction means that someone should be
surprised more and express surprise, for example, by banging (on the table) saying
majide, 'are you serious?’ and move her upper body more like saying ee, 'gee' with rising
intonation. She says that she also uses phrases such as majide bakaja nai no, 'are you
crazy, you are a moron, 'nakune, 'I would cry,’ and sorede sorede, 'and then, and then.‘ TI
(22 years old) also makes similar comments about how she expects the listener to behave

when she talks.

(15) TJ : (I like) one who responds more when I speak. Even too much

reaction
M: As for reaction, what do you mean?
TJ: Ah, well with the one who becomes ( ? ) or something, and one

135

who has lively expressions or something, I feel like I’m comfortable, or
I think that She is listening and it is fun because I also enjoy speaking.

' According to TJ in (15), a good listener for her is one who Shows her/his enthusiasm and

encourages her to keep talking even with an overreacted manner. SZ and TJ have a

common expectation toward listeners' behavior in terms of the types of responses. The

researcher asks TJ's opinion about overlapping.

(16) M:

TJ:

TY:

M:

TY:

TY:

TJ:

Is overlapping someone's utterance rude?
Not at all, right? { }
yeaht }
You don't do that when listening to instructions, do you?

Also when I talk to professors, I listen until the proper boundary
and then say what I want to say.

What about someone of the same generation that you don't
know?

If I don't know her, I listen. I listen to all or let's say I listen a little
more till the boundary of the talk and then say 'Oh, it is so.’

If we agree very much, I may overlap her, though. If she sounds

like a person who can share the same topic, I may overlap her but if she
talks about something that I don't know at all, I listen to her as usual
until the boundary and then say 'ah, it is so.’

In (16) TJ and TY agree that they do not mind at all if someone talks simultaneously or

overlaps their utterance although basically they don't do that to someone who has higher

status or whom they don't know well even in the same generation. The rule for

overlapping might also be explained by the relationship of power and solidarity, similarly

to the rule for nods.

136

The participants' comments are consistent with the ﬁndings that listeners
participate by responding with reactive expressions, collaborative ﬁnishes, repetitions,
and paraphrasing when chatting with close friends; however, fewer active backchannels,
overlapping and interruption are common in formal conversation.

Differences in interaction between formal and informal contexts can be partially
explained by the goal of conversations. That is, formal conversation, such as receiving
instructions, is aimed at being an information-centered interaction. The primary goal for
the speaker for this type of conversation is to convey the meaning accurately for some
purpose. As a result, concern about creating solidarity between a Speaker and a listener is
a minimum in this case. Participants also acknowledge their different behaviors regarding
the purposes of conversation. The following cement in (17) is a part of a speech in which

SZ explains why one nods when listening to instructions.

(17) SZ: Because people listen to instructions to obtain information, if they
can't get information, it will be awful. That is why.

7.2.5. Goals of communication

People communicate in order to pursue their goals of conversation and the goals
might differ according to the contexts that are determined by types of speech events, age,
gender, the relationship between participants, and so on.

The major goal of informal conversation in the present study, chats between
close female friends, is considered the establishment of 'Camaraderie,' whereas that of
formal conversation is information and 'Deference' The different goals could cause

fundamental differences in the structure of formal and informal conversations as we

137

discussed in the previous chapter. The purpose of chat appears in the dialogue (18)
spoken by a participant, HY (37 years old), who also explains how she behaves
differently in different contexts:
(18) HY: ....Conversation between homemakers you know it doesn't
matter if it is yes or no ( ? ). Well ah when we are the same sex and
close. ....If I meet someone oﬁicially as part of work, I listen ﬁrst, then
say 'it is right,‘ but in case of this kind of friend I feel like I want to
understand what She is saying for her sake so say 'yeah yeah.’ Also just
now I realized that when we talked about our school trips, I sometimes

nod like this to help my partner to speak. In a sense it was love { } and
that is what I understood after I did.

HY's own behavior reﬂects her performance overall. She plays a supportive
listener in both settings in different ways (i.e. in formal conversation: low nods and high
verbal response (rapport style), informal conversation: high involvement style. Her

comments explain the listener's internal state while she plays a listener role.

7.2.6. Social contexts and interactional style
The present study investigates the cause of variation in interactional style in
terms of power and solidarity. There is some evidence for how power could inﬂuence
interaction and how a speaker who has more power deals with the situation. SN (56 years
old) participated in the study with AZ (40 years old) who is a subordinate in the same
work place. They both acknowledged that they were close friends but SN said that she
was worried that AZ might not feel free because of the relationship at work.
(1 9) SN: A while ago I guess that she was very nervous because I am her

boss and I am also concerned about her. ..... So I spoke more than her
without listening to her story as much as I should have--it is my

138

personality and I have an impression that my listening behavior is
inﬂuenced by the speaker's personality more than other factors.

SN’s analysis in (19) indicates that she had held the ﬂoor longer than AZ
because of her consideration for AZ'S feelings, i.e., that she might hesitate to speak
because SN is AZ'S boss’ thus, SN tried to help AZ relax by talking a lot. This is an
example of how someone who has more power can exercise solidarity by Showing
friendliness as Tannen pointed out (1993). AZ did not say anything about how she felt
when she talked to SN. However, the present study excludes their conversation from our
analysis of chatting to avoid the possible inﬂuence of a hierarchical relationship at work

on the way they talk.

7.2.7. Perception of family discourse

Okamoto and Sato (to appear) point out that chatting between friends and
chatting between mothers and daughters performed by the same participants (daughters)
demonstrate quite different interactional styles and claim that the traditional view of
Japanese interaction may account for chatting between friends, but not for family
conversation.

Participants in the present study who had children related their thoughts about
young people's listening behavior based on their observation. BS (47 years old) and T2
(53 years old) said:

(20) BS: My child does not remain silent but he is like he starts talking
about something else. Yeah, he doesn't listen to me on the whole. My

two children don't listen but say something next without responding to
me about (what they are asked ﬁrst) so (I think) they don't listen to me.

139

(21) T2: You know, kids want to ﬁnish the conversation (with us). So if they
once respond to us saying, 'well mother,’ the conversation stops at the
point and doesn't move to next topic, right. So they want to avoid that
so they want to ﬁnish conversation { } they want to hear what comes
next. . .. So kids never have manners towards parents, I guess.

MT describes her son's change of listening behavior in terms of non—verbal

features (i.e. gaze directed to her while they talk).

(22) MT: I have a son but when he was younger, I wondered why he didn't
talk looking into my eyes. He at one time while ah, when he ﬁrst lived
in Kyoto, though, had a hard time and came home and then he listened
to me and looked at my eyes or toward a speaker straight. That is what I
feel.

BS in (20) and TZ in (21) agree that their children do not listen to their mothers
as they expect but rather they act as if they don't listen to their mothers. BS and T2 seem
to accept the way that their children act to them and TZ says, 'Kids never have manners
towards parents, I guess.’

It is clear that their children's listening behavior is far from being co-operative or
supportive. In (22) MT mentions that eye contact is a measure to know how much her son
is involved in listening according to her. MT'S comments suggest that listening behavior
is somewhat developmental so that it may change as children grow. Although it is not
clear whether or not MT'S son uses RTS properly, no doubt her son's listening behavior is
also similar to others introduced by BS or T2.

The previous chapters present the ﬁnding that young people participate in both

formal and informal conversations as supportive listeners using diﬁerent interactional

styles. These ﬁndings have implications in that the children of BS, TZ, and MT also

140

switch among different interactional styles according to the context.

The following dialogue presents a perspective from the daughter's point of view,

that is, HW'S (19 years old) observations about the conversations with her mother.

HW (19 year old) describes a conversation with her mother as follows:

(23) HW:

ST:

HW:

HW:

ST:

HW:

...(We) talk at cross purposes. First I say something like this but
she responds to something different although she listens to me you
know.

She does not listen carefully.

So she is listening to the content but the content of her response

is totally different so I think, 'Are you really listening to me?‘ ...My dad
listens to me. Then he responds to me about what I say. But my mom
says like the supermarket was something...like 'What? It is not what
I'm talking about.'...She is annoying, really annoying. I say, 'Are you
listening?‘ So my brother says that he is tired when he listens to the
conversation between me and my mom because we talk about different
topic { }. We talk at cross purposes.

Do you think men talk like that?

I think they don't, right?

Women say only what they want to say
If they are close right?

If they are close right?

It is interesting to know that HW’S mother's problem in (23) is similar to what

BS mentions about her children. That is, their conversation does not consist of 'adjacency

pairs' (Sacks et a1. 1974) but rather two speakers talking simultaneously about different

topics. KM (54 years old) thinks about her own behavior and how She talks to her

daughter and says:

141

(24) KM: I felt like I only responded to (my daughter) like ordering
something. We sometimes use backchannels without listening, don't
we? Just reasonably.

KM'S way of communicating with her daughter may cause the frustration that
HW mentions. Her backchannels used during the conversation with her daughter function
to make an impression that She is listening although it is not clear whether she really
listens or not. At the same time, actual feedback from KM could not be the answer that
her daughter expects but rather something else such as questions or orders from her
mother.

The participants' informal observations provide evidence that some people do
not listen to others in an empathetic manner, particularly in mother-child conversations
and the style is seen in all generations. It is beyond the scope of the present study to
investigate whether it is individual preference or stylistic characteristics that are only
relevant to family discourse. However, comments by the participants about the
conversation between parents and children provide an outline of how mother-child
discourse is constructed, consistent with the notion of Okamoto and Sato (to appear).

Comments from the point of view of a parent or a child not only involve insight
into each other's expectations but also reveal how they perceive the other's listening
behavior. SN (56 years old) in (25) reﬂects upon how her daughter reacted to SN'S

listening behavior to someone else when her daughter was a teenager.

(25) SN: I guess she (my daughter) wanted me to be calm.

M: 'Calm' means fewer noddings and backchannels or something?

142

SN: Yeah yeah I felt that maybe that was the thing you know. From
my child's point of view she felt I was overreacting (and listened to the
doctor).

After discussing people's different listening behaviors in the interview, SN says
that now She understands how her daughter expected her to behave when she listened to a
surgeon who was supposed to operate on her a long time ago although she didn't
understand it at the time.

SZ (20 years old) presents her opinion about why older people use more
backchannels than younger:

(26) SZ: I guess maybe it is because they are shameless { }. That is why
older people can ask questions or respond more verbally than us if their
interlocutors do not respond or something.

In (26) SZ implies that older people should keep their distance from the speaker
by using fewer backchannels. Also, it implies that younger people's behavior is more
polite and appropriate so that they don't dare to conﬁrm when they don't understand what
the speaker says. The ﬁndings of the present study show that participants of group I use
nods as negative a politeness strategy. SZ's comments can explain why they choose an

indirect, less intrusive and, most importantly, less face threatening type of RT, nods, over

verbal RTS.

7.3. Perception study

This section presents the results of rating different listening behavior in video
clips. The following sections explain the methods, results of the rating and follow up
interviews about video scenes that participants see. Discussion will follow at the end of

the chapter including testing hypothesis 4.

143

7.3.1. Method and data

After video taping participants’ conversations with their friends, they were asked
to view 5 50 seconds video scenes. The researcher explained as follows, "Now you are
going to see 5 video scenes. Each scene is 50 seconds. You see a proﬁle of the same
woman who listens to the same story from a woman who doesn't appear on the screen.
Please pay attention to how the woman in the scene listens. I will stop the video after
each scene and give you some time to rate it in the form." And the researcher read the
questions aloud in the form (in Appendix 4): Did the person in the scene behave
naturally? Is the person's behavior attentive? Does the person's behavior annoy you? And
if you were the person in the scene, would you behave like her? The scale of the rating is
one to ﬁve.

The video clips were artiﬁcially made based on the natural conversation between
friends in the following way. The researcher audio-taped the chat with a close female
friend and transcribed it. Then a 50-second segment consisting of a narrative in which the
researcher told an episode that happened when going into the parking lot the other day
was selected and transcribed. The researcher told the same story to 2 people. The
conversations were video taped and transcribed. The transcribed interaction was analyzed
in terms of the frequency of RTS, types of RTS, and contexts that RTS occur in to make 5
variations of listening to the story.

First, the frequency of verbal RTS and nods were divided into 3 levels: high, mid,
and low frequency of each. They were combined into ﬁve styles which vary in only
frequency and ratio of verbal RTS and nods. Table 7.1 shows the combination of verbal

RTS and nods in each video clip.

144

Table 7.1. Number of verbal reactive tokens and nods in 50-second video clips

 

 

 

 

Video 1 Video 2 Video 3 Video 4 Video 5
Number of Verbal RTS 7 l3 13 23 13
(high, mid, low) (low) (mid) (mid) (high) (mid)
Number of Nods 5 5 10 5 0
(high, mid, low) (mid) (mid) (high) (mid) (low)
Total 12 18 23 28 13

 

 

 

 

 

 

 

 

The frequency of the total verbal RTS is either high (23 times per 50 sec), mid
(13 times per 50 sec), or low (7 times per 50 sec). The frequency of total nods is either
high (10 times), mid (5 times) or low (0 times). The 5 video clips consist of one of the
combinations: 1: low verbal+mid nods, 2: mid verbal+mid nods, 3: mid verbal+high
nods, 4: high verbal+mid nods, and 5: mid verbal+low nods. Transcripts of each video
clip are shown in Appendix 3.

The types of verbal RTS used in the video clips are one of four types, that is un
'yeah,' soo 'It is so,' hee 'uh-hum,’ and usoo 'unbelievable.' The following 3 RTS , soo 'It is
so,‘ hee 'really,' and uso 'unbelievable,’ were used once. The rest of RTS used were un,
'yeah' because it is considered most appropriate in the conversation style of the present
study for all age groups (Miyazaki 2001b). The types of RTS are limited to backchannels
(un,'It is so,' hee, 'uh-hum,'), nods, and reactive expressions (soo 'It is so,’ u s o
'unbelievable') although more variety of RTS and more frequent turn shifts were observed
in the actual conversation.

The reason why the video scenes Show a proﬁle of a person is to avoid
unnecessary inﬂuence by the person's other expressions. And artiﬁcially made scenes are
used instead of using the actual listening behaviors which respond to the researcher's

telling a story because participants may be inﬂuenced by other individual factors such as

145

voice quality, face, and types of RTS used by each listener. Lastly, in the video clips, the
locations of all RTS are at clause boundaries which can be reccognized by pauses and
intonation changes. There is no overlap or interruption in the conversation of video clips.
The interaction style is similar to rapport style in the present study. RTS consists of
informal types of backchannels and nods.

The data consist of 30 sets of ratings done by the same participants in formal and

informal conversation data.

7.3.2. Results
7.3.2.1. Rating
This section will present the results of ratings for four questions about
naturalness, attentiveness, annoyance and similarity to their own behavior. The ﬁrst
question that asks how natural the person's behavior is was rated in the following way.
Figure 7.1 shows that video 5 is rated highest (3.8) in naturalness and the
difference in ratings among videos 1, 2, 3, and 4 is small (2.8 or 2.9). Participants
perceive that video 5, which contains mid-verbal RTS (l3) and low-nods (0), is the most
natural behavior for listening to the conversation in the context. The results of ANOVA
show Signiﬁcant effect [F(4,l45)=4.6l, p=.001]. The results of Tukey post hoc test in

table 7.2 Show that only video 5 is signiﬁcantly different from videos 1 to 4.

146

Figure 7.1. The average of rating in terms of naturalness of listening behavior
in each video scene (n=3 0)

 

1.8
1.6

Naturalness rating
H H
.1 iv '1:

.°.°.°
eaten

.0
N

 

0

video 5(mid
VRTs+low
nods)

video 2(mid
VRTs+mid
nods)

video 3(mid
VRTs+hi gh
nods)

Videos

video 4(high
VRTs+mid
nods)

video 1(low
VRTs-I-mid
nods)

 

 

 

Table 7.2. Results of Tukey post hoc test of naturalness among videos

 

 

 

 

 

 

 

 

 

 

 

 

Video 1 Vide02 Video 3 Video 4 Video 5
Video 1 1.0000
Video 2 0.9777 1.0000
Video 3 0.9777 1.0000 1.0000
Video 4 0.9777 1.0000 1.0000 1.0000
Video 5 0.0296* 0.0044" 0.0044" 0.0044M 1.0000

 

 

(*=p<.005, **=p<.0005)

Next, ﬁgure 7.2 Shows the ratings for how each behavior appears to represent

paying attention. Video 2 (mid verbal+mid-nods) is rated the highest followed by video 5

147

(mid verbal+low nods), videos 1 and 3 (low verbal+mid nods and mid verbal+high nods
respectively), and video 4 (high verbal+mid nods). Participants rate the combination of
mid+mid the highest. This result implies that too many RTS do not give an impression of
a good listener who listens carefully and a low rate of RTS means that the speaker may

not understand that their interlocutor is listening attentively.

Figure 7.2. The average of rating in terms of attentiveness of listening behavior in each
video scene (n=30)

 

4.5

4

Attentivenoss rating
Q ~ N u
in -- Ln N in 1» Ln

 

O

video 1(low video 2(mid video 3(mid video 4(high video 5(mid
VRTs+mid VRTs+mid VRTs+high VRTs+mid VRTs+low
nods) nods) nods) nods) nods)

Videos

 

 

 

The results of ANOVA Show signiﬁcant effect [F(4,145)=2.52, p=.043]. The
results of Tukey post hoc test in table 7.3 show that diﬁ'erence between video 2 and 4, is

Signiﬁcantly different. Also video 4 and 5 are almost signiﬁcantly different (p=.059).

148

Table 7.3. Results of Tukey post hoc tests of attentiveness among videos

 

 

 

 

 

 

 

 

 

 

 

 

Video 1 Vide02 Video 3 Video 4 Video 5
Video 1 1.0000
Video 2 0.9637 1.0000
Video 3 0.999 0.9309 1.0000
Video 4 0.2078 0.0412* 0.2670 1.0000
Video 5 0.9840 0.9999 0.9637 0.0595 1.0000

 

(*=p<.005, **=p<.0005)

The third question examines which listening behavior bothers speakers if
participants were to have a conversation with the person in the video scene. Participants
were asked to rate videos on the scale of 1 to 5 (most annoying to least annoying) so that
negative effect is consistent with ratings of other questions. The most dispreferred
behavior is judged to be in video 3 (mid verbal+high nods, total 23) and video 4 (high
verbal+mid nods, total 28). These two video scenes are ones containing the most frequent
responses in total. Too many listener responses may be interpreted as annoying behavior
and there may exist a total number beyond which speakers judge the listeners' behavior to

be supportive or annoying. The result implies that 23 verbal RTS in 50 seconds seems to

be too abundant and 10 nods in 50 seconds also bother speakers.

149

 

Figure 7.3. The average of rating in terms of preference of listening behavior in each
video scene (n=30)

 

4.5

preference rating
r- .~ 1»
H U! N 01 W Ul

.0
U'I

 

0

video 1(low video 2(mid video 3(mid video 4(high video 5(mid
VRTs+mid VRTs+mid VRTs+high VRTs+mid VRTs+Iow
nods) nods) nods) nods) nods)
Videos

 

 

 

The results of ANOVA show signiﬁcant effect [F(4,145)=4.28, p=.0026]. The
results of Tukey post hoc test in table 7.4 Show that video 5 is signiﬁcantly different from
video 2, 3, and 4. Although the rating of video 5 is higher than video 1, they are not
signiﬁcantly different. All participants perceive the ones which contain of low nods as
most preferable and the difference in the number of nods (5 vs. 0) is not perceived with

only signiﬁcant effect.

150

Table 7.4. Results of Tukey post hoc tests of preference among videos

 

Video 1 Vide02 Video 3 Video 4 Video 5

 

Video 1 1.0000

 

 

 

 

Video 2 0.6632 1.0000
I Video 3 0.5909 1.000 1.0000
/ Video 4 0.5909 1.000 1.000 1.0000

 

Video 5 0.3147 0.0102* 0.0070** 0.0070M 1.0000

 

 

 

 

 

 

 

(*=p<.005, “=p<.0005)

Figure 7.4. The average of rating in terms of similarity to their own listening behavior
in each video scene (n=30)

 

Similarity

 

video 1(low video 2(mid video 3(mid video 4(high video 5(mid
VRTs+mid VRTs+mid VRTs+high VRTs+mid VRTs+low
nods) nods) nods) nods) nods)

Videos

 

 

 

The last question is, ‘How are these listening behaviors similar to your own

behavior? Video 5 is rated highest, followed by videos 1, 2, and 3 and 4. All the ﬁgures

151

present a Similar order in the actual rating. That is, the rating is in the order of video
49 39 29 19 5 ((low9 high) which corresponds with the frequency of
28923918912913 times per 50 seconds.

The results of AN OVA do not show signiﬁcant effect among videos in terms of

Similarity [F(4,145)=1.76, p=n.s.].

7.3.2.2. Age and perception

Each rating is tested to determine whether any are related to age. Results of
Pearson Correlation analysis Show that video 3 (mid verbal+highnods) and age have a
positive correlation [r=.40, p=.018]. That is, older participants rate video 3 higher, which
was rated as second lowest by all participants. However, none of the other ratings shows
a correlation between the ratings and age.

Overall, ratings of video sceans vary little among age groups. Table 7.5 shows
the results of ANOVA in terms of age groups and ratings of video scenes.
The rating of naturalness of video 2 and 3 are signiﬁcantly different among age groups.
However other rating and age group do not have signiﬁcant effect. Table 7.6 and 7.7

present the results of Tukey post hoc test of video 2 and 3 respectively.

Table 7.5. Results of ANOVA in terms of naturalness, attentiveness, preference,
and similarity in each video scene

 

The results of ANOVA

 

Naturalness

 

Video 1 F (2, 27)=2.71, p=n.s.

 

 

 

 

152

Table 7.5 (cont’d)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Video 2 F (2, 27)=6.26, p=0.005**
Video 3 F (2, 27)=3.84, p=0.034*
Video 4 F (2, 27)=0.27, p=n.s.
Video 5 F (2, 27)=0.45, p=n.s.
Attentiveness

Video 1 F (2, 27)=0.81, p=n.s.
Video 2 F (2, 27)=0.46, p=n.s.
Video 3 F (2, 27)=0.09, p=n.s.
Video 4 F (2, 27)=2.26, p=n.s.
Video 5 F (2, 27)=2.74, p=n.s.
Preference

Video 1 F (2, 27)=l.l3, p=n.s.
Video 2 F (2, 27)=1.63, p=n.s.
Video 3 F (2, 27)=2.26, p=n.s.
Video 4 F (2, 27)=0.58, p=n.s.
Video 5 F (2, 27)=1.56, p=n.s.
Similarig

Video 1 F (2, 27)=l.35, p=n.s.
Video 2 F (2, 27)=2.l9, p=n.s.
Video 3 F (2, 27)=3.06, p=n.s.
Video 4 F (2, 27)=1.56, p=n.s.
'Video 5 F (2, 27)=0.20, p=n.s.

 

153

 

Table 7.6. Results of Tukey post hoc test of video 2 in terms of naturalness rating
among groups

 

 

 

 

Group I Group II Group III
Group I 1.0000
Group 11 0.0054" 1.0000
5 Group 111 0.0528 0.5944 1.0000

 

 

 

 

 

 

(*=p<.005, **=p<.0005)

Group I participants rated video 2 (mid verbal RTs+ mid nods, total 18) lower
than Group 11 participants. The average of rating of Group I for video 2 is 2, while that of
Group II is 3.4 They are signiﬁcantly different.

Table 7.7 shows the results of Tukey post hoc of video 3. Again, the rating of
Group I and II is signiﬁcantly diﬁerent but none of the relations are. Group 11 participants
also rated video 3 (high verbal RT+low nods, total 23) higher than Group I participants.
The average of rating of Group I is 2.1, while 3.4 for Group 11. Although Group 11
participants rated video 5 as most natural among 5 videos, they rated video 2 and 3 higher

than the other 2 groups.

Table 7.7. Results of Tukey post hoc test of video 3 in terms of naturalness rating

 

 

 

 

 

amonggroups
Group I Group II Group III
Group I 1.0000
Group 11 0.0277“ 1.0000
Group 111 0.2269 0.5483 1.0000

 

 

 

 

 

(*=p<.005, "=p<.0005)

154

 

The results of ANOVA show that there is little difference among native Japanese
female speakers whose age ranges from 19 to 61 in terms of perceiving others' listening
behavior to be natural, attentive, preferable or Similar to their own behavior in this kind
of informal conversation.

The following section will demonstrate the participants' comments in the follow-

up interview and discuss the results of ratings of perception.

7.3.2.3. Follow up interview about video clips

This section introduces what participants thought about the listening behavior of
the woman in the video clips. The comments of participants Show which video scene is
the most unnatural and the reason why they judged it to be so. Then comments about the
gap between listening behavior that participants expect and the behavior in the video will
be presented. Finally, ﬁndings regarding the perception of someone's listening behavior
will follow at the end of this section.

Among 5 video scenes, video 4 is rated as the most dispreferable listening
behavior. For example, SZ (20 years old) states that too many verbal responses (video 4)
may give a negative impression in (27).

(27) SZ: The one in which she uses verbal responses too much is
offensive, isn't it?

M: Too many?

SZ: Right.

M: It is kind of annoying?

SZ: Well, you know it may sound like she doesn't listen to carefully. Like if
she says 'yeah yeah' properly, it sounds like she is listening or

155

something like that [even though She is not listening]

SF (61 years old) and MD (39 years old) also point out that the listening
behavior in video 4 bothers her because of the frequent use of backchannels in (28) and
(29) respectively.

(28) SF: There was one video that bugged me because she punctuated
the interlocutor's words with backchannels one by one. The rest didn't

differ much.

(29) MD: Only video 4 is a little noisy because it has too many 'yeahs.‘

Some participants comment on the types of tokens and interactional style. SB
(20 years old) and SZ (20 years old) agree that they thought the reaction of the woman in
the video clip differs from their expected reaction because She does not laugh when the

speaker told of an incident at the parking lot entrance.

(30) SB: Although that is the high point, she doesn't laugh.

(31) SZ: Well, when I say what I want to say, I expect my friend to laugh,
don't you?

SZ's comment in (31) implies that laughter is a way to express a listener's
involvement and/or strong support toward a speaker. For these young people, a good
listener is one who can create a fun atmosphere through laughter and participate in the on
going conversation in a more direct way through 'exaggerated' expressions. WB (54 years
old) states what she thought about video scenes in general.

(32) WB: ....Though I also hardly recognize the difference (among video
clips), I wonder if it’s nodding or something-there are some clips in

156

which the person nods vigorously. I didn't notice the differences clearly.
But the person only says yeah yeah so I thought maybe I wouldn't listen
to someone like she does after watching the video clips, you know. AS
ST points out, I feel like if I were her, I would use a little more variety
of expressions, such as 'oh my,’ or 'and then,‘ or something....There
could be a variety of 'yeahs' used if I were her.

ST: 'It is annoying, isn't it?' or 'oh, that was tough, wasn't it?' and so on.
What do you think?

WB: But I feel very much that she listens very carefully. I felt this way about

all of them.

In (32) WB says that the difference among video clips is subtle and hardly
noticeable except for one (video 4). WB and ST comment that they wouldn't listen like
the woman in the video does because her reaction lacks variety and emotion, although it
is true that all the scenes could happen in a real conversation and the woman seems to
listen carefully in all the scenes.

In the analysis of interactional style in chapter 6, WB is judged as rapport style
and ST is machine-gun style in their performance in informal conversation. In formal
conversation, both participants respond with high verbal and low nods. Obviously their
performance in informal conversation is distinct from the ones in the video scene.

The comments of the participants reveal the features of listening behavior which
are most important in order for speakers to know the listeners' attitude toward what the
speakers say. Results Show that nods are not important compared with verbal RTS. In fact,
there was little difference between video 2 and 3 which have the same number of verbal
RTS, but the latter has twice the number of nods and the average of naturalness rating is
2.8 for both video 2 and 3. On the other hand, the high rate of verbal RT in video 4 is

judged to be lowest in naturalness among participants.

157

7.3.3. Hypothesis 4 and discussion

This section explains how the ﬁndings account for hypothesis 4 and discusses
the results of the perception data.
Hypothesis 4: The perception of other's listening behavior varies according

to perceiver’s age.

The results of the perception data in the present study did not support hypothesis
4. That is, the perception of other's listening behavior does not Show age-related variation
consistent with all ages. Their perception reﬂects what they do in their own performance.
As I discussed earlier, the results Show that they rated the one which was the closest to
their style of informal conversation in the present study most natural, preferable (least
annoying), and similar to their own performance. The result suggests that female native
speakers of Japanese have similar perceptions of other's listening behavior in chats no
matter what age they are.

However, it should be pointed out that the video clips have problems that
possibly inﬂuenced the results.
1. The video clips are not authentic, but artiﬁcially made, so that participants'

judgements might be inﬂuenced especially in rating naturalness.

2. The ratio and number of verbal RTS and nods determined based on the preliminary
study may not be applicable to the context of the conversation in the video scene.

3. The types of verbal RTS and interactional styles do not reﬂect the expectation of
participants, since in the video clips backchannels are used primarily, while actual
conversation between close friends involves various RTS, such as collaborative
ﬁnishes, repetitions and so on.

4. The interactions of video clips are informal conversation. There might be
age-related difference in perception of formal conversation.

158

These methodological problems should be considered further in future studies to

improve the authenticity of video scenes.

7.4. Summary

This chapter presented an overview of how the ﬁndings of the present study are
supported by participants' comments collected through follow-up interviews. The
participants' dialogues or comments reﬂect how they actually performed in formal and
informal conversation. This chapter presents the data which supports Hymes's notion that
speakers of a Speech communnity share norms although they don't necessarily speak in
accordance with such norms (1972, 1974). It also supports the assumption that native
speakers have rules of listening behavior proposed in the previous chapters. Although
hypothesis 4 is not supported by the results of the perception study, there are some

implications regarding the perceptions of people of different age groups.

159

CHAPTER 8

CONCLUSION

The present study explored the listening behavior of Japanese female listeners,
who were generally thought to be cooperative, supportive, and enthusiastic, in same sex
face-to-face conversations. There were three goals for this study. The ﬁrst goal was to
describe how listeners participate in conversation while they listen by using RTS. The
second goal was to examine the types of linguistic strategies used. The last goal was to
investigate how social factors of age and conversation styles affect listening behavior and
the perception of the listening behavior of others.

The listening behavior investigated in the present study was the use of RTS
comprising 2 types of RTS; verbal RTS--backchannels, reactive expressions, repetitions,
collaborative ﬁnishes, resumptive openers, laughter, short comments, and paraphrases;
and non-verbal RTS, i.e., nods. The following is a review of the main ﬁndings of the

present study.

8.1. Findings of the present study

Data were collected from 30 Japanese female native speakers between the ages
of 19 and 61. Data consist of 2 kinds of video taped conversations, the ratings of others'
behavior in the video, and recorded interviews. Conversational data were transcribed for
analyses and a total of 60 minutes of formal conversations and 20 minutes and 26
seconds of informal conversations were analde quantitatively and qualitatively.

First, the frequency of the use of RTS and the ratio of verbal RTS to nods were

160

quantitatively examined. Second, variation in listener responses was examined among
participants. Next, two social factors of style of conversation (formal and informal) and
age were investigated as social variables that inﬂuence listeners. Finally, the speakers'
metalinguistic knowledge about listening behavior was examined to explain the ﬁndings
in their own performance and perception of other people's listening behavior.

I posited four hypotheses regarding the interrelation between verbal RTS and
nods, the inﬂuence of age and style on the use of RTS and the perception of other
people's listening behavior. The following are the hypotheses and the results.

Hypothesis 1: The ratio of verbal to non-verbal RTS may vary among
individuals, but the total number will be similar in similar contexts.
9 Supported
Hypothesis 2: The frequency of verbal and non-verbal RTS is affected by
age. Older female speakers use more verbal RTS than do younger
speakers. 9 Supported
Hypothesis 3: The formality level of conversation inﬂuences not only the choice of RT
types and frequency, but also the interactional styles.

9 Supported

Hypothesis 4: The perception of other's listening behavior varies according
to perceiver’s age. 9Not supported

In order to test hypothesis 1, the present study examined formal conversation
data in which every participant listens to the same instructions that were planned and
given by the same speaker, the researcher. The ﬁndings reveal that the total number of
RTS used by each participant does not vary statistically, but the ratio of verbal to non-
verbal RTs varies among participants. The total number of verbal and non-verbal RTS had

a negative correlation.

Hypothesis 2 was also supported. The results of formal conversation

161

demonstrated that age and nods had a negative correlation, while age and verbal RTS had
a positive correlation. The hypothesis that older participants respond more verbally than
younger participants was statistically supported.

Analysis of informal conversations showed diverse and complex interaction
compared with that in formal conversations. Overall, listeners used verbal strategies more
frequently in informal settings than they did in formal settings. The ratio of backchannels
to other verbal RTS was smaller in informal than in formal contexts. That is, listeners used
verbal RTS other than backchannels more frequently in informal conversations. Various
listener responses, such as collaborative ﬁnishes or repetitions facilitated listeners' active
involvement in conversations and eventually resulted in very distinct interaction from
formal conversations.

The present study proposed that there are three interaction styles relevant to the
context in the present study--rapport style, high-involvement style, and machine-gun
question style--depending on the choice of RT types and the ratio of each RT. The
findings Showed that interaction during formal conversations tended to be rapport type in
which listeners used a greater proportion of backchannels than other verbal RTS. In
informal conversations, participants' listening types were identiﬁed as each of the 3 types
and age did not relate to the speciﬁc type of interaction.

Hypothesis 4 was not supported because all the participants' ratings were similar
despite their ages. The results showed that female native speakers of Japanese perceived
Others' listening behaviors similarly and their intuition did not vary according to age.
What was found in the present study is consistent with the notion of 'Speech community'

(Gumperz and Hymes 1972; Hymes 1974).

162

8.2. Suggestions for future study
The present study points to a need for further consideration of nods that do not
accompany with verbal RTS aS listener responses. The ﬁndings provide evidence that
nods play a signiﬁcant role, especially in formal conversation, since nods can serve as a
negative politeness strategy. It suggests that future research on Japanese listening
behavior needs to study not only verbal features but non-verbal features.

The results of the present study reveal that there is a Signiﬁcant difference in
listening behavior between formal and informal conversations. It might be pointed out
that the amount of informal conversational data that were actually analyzed was limited
(a total of 20 minutes 26 seconds of data consisting of 13 dyadic conversations)
compared with formal conversational data (a total of 60 minutes data consisting of 30
dyadic conversations). A comparable size of data from different conversation styles could
provide statistically supported results.

The variations caused by the conversation styles were explained by the
'multidimensional model,’ (Tannen 1993a) using four dimensions made by the power and
solidarity continuum. The present study provided empirical data for only 3 dimensions,
and not for the context which involves power difference and close relationship. Informal
natural conversation data that are relevant to such contexts in which participants are close
yet involve power relation between participants are needed to account for the proposed
Stylistic variations.

The present study focused on listening behavior of female Japanese speakers.

Further research on Japanese male Speakers in similar contexts should be conducted in

163

order to investigate common beliefs or norms about Japanese listeners.

The present study is intended to be the ﬁrst step in gaining researchers' attention
toward the occurrence of nods and their function as RTS in individual listening behavior
in face-to-face conversation. It is hoped that this study will make a contribution to the
ﬁeld of discourse analysis as a study of variation in different age groups. It is also hoped
that the ﬁndings will make a contribution to cross-cultural communication by providing

information about Japanese listening behavior.

164

APPENDICES

165

APPENDIX A: Transcription methods and conventions

Hepburn style romanization is used for Japanese translation.

Data, tables, and ﬁgures are numbered for each chapter.

Participants are referred to by the initial of their ﬁrst name in Chapter 3.
Paralinguistic and other interactional symbols are as follows:

falling intonation with noticeable pause

continuing ('little') intonation, with a Slight rise or fall,
followed by a Short pause

omission
linked or continuing utterances without overlap

U

[ ] speech overlap

{ } laughter
H head nod(s)

? Rising intonation
( ? ) unclear utterance

When transcribing, not all morphemes are separated. Only those which are given
glosses are separated.

Abbreviations used

BE
NEG
ID

NM
TAG

IP

Q
QT
SUB
TOP
01>

copulative verb, be

negative

ideophones (including onomatopoeic
[giseigo] and mimetic [gitaigo] words)
nominalizer

tag-question-like auxiliary verb forms
(including jan, ja-nai, desho, daroo, etc
interaction particle

question marker (assigned to ﬁnal particle ka only)
quotative marker

subject marker

topic marker

object marker

166

APPENDIX B: Script of instruction

1: Kyoo wa o isogashii tokoro arigato gozaimasu.
Today TOP busy NM thank you

2: Ana, korekara o hutari de nani o shite itadaku ka kantan ni setsumee =
Ah, now two what OP do QT brieﬂy explain

3: =saseteitadakimasu.
Let me do

(1-3 English translation) Thank you (for coming) while (you are) busy today. Well,
now I am going to briefly explain what you two are going to do (for me).

4: Mae m' chotto o hanashi shimashita ga,
before a little tell did but
5: watashi wa amerika no mishigan shu ni aru daigaku de,
I TOP USA of Michigan state exist university at
6: gengogaku o senko shite ite nihongo no kaiwa no kozo?
Linguistics OP major Japanese of conversation of structure
7: Danwa to iimasu ga sore o kenkyu shite imasu.

Discourse QT say but it OP study

(4-7 English translation)I mentioned a bit before but I am studying linguistics as
major at university in Michigan in America and then studying the structure
of Japanese language? We call (it) 'discourse,' I am studying that.

8: Sorede jissaini,
so actually
9: nihonjin no kataga shitashii shiriai no kata to douiu huu ni o hanashi 0:

Japanese of pe0ple close acquaintance of people with how talk OP

10: = sareru ka no deeta o atshumete imasu.
Do QT of data OP collect

(8-10 English translation) So (I) am collecting data of how Japanese people
actually speak with close acquaintance.

ll: Atode _san to
later Ms._ with

167

12: ano 15 hun kurai shugaku ryoko no omoide ni tsuite hanashite =
ah, 15 minuites about school trip of memory about talk

13: = itadakitai to omotte imasu.
Want QT think

(ll-13 English translation) Later, I think I would like you talk with Ms. about
memory of school trip for about 15 minutes.

14: Nani o shirabete iruka ni tsuite wa ienain desu go,
what OP study about TOP say can't but

15: hanashi no naiyo toka, hyogen toka, kotoba zukai,
talk of content etc. expressions etc. language use

16: to in yona koto o shirabete iru wake deha naino de,
QT like NM OP investigate not so
17: ki ni nasarazu ni,

mind don't and

18: itsumo dori hi 0 hanashi shite itadakereba kekko desu.
Usual as talk if ﬁne BE

(14-18 English translation) I can't tell you what I am studying but I am not studying
content of your talk, expressions, your language use, or something like that,
so it will be ok if you feel free and talk as you do always.

19: Shugaku ryoko toka,
school trip and so on

20: ensoku e itta toki no episodo nanka de hanashi hajimereba=
ﬁeld trip took time of anecdote with talk start if

21: = ii ka to omoimasu.
Good QT think

(19-21 English translation) It will be good if you start talking about like an episode
when you went to the school trip or ﬁeld trip from your school.

22: Tatoeba,
For example
23: watashi wa yamaguchi ken de sodatta node,
1 TOP yamaguchi prefecture in raised so

168

24:

25:

26:

shogakko no shugaku ryoko wa kyushu de,
elementary school of school trip TOP Kyushu and

chugaku wa shinkansen ni notte kyoto e ittan desu kedo,
middle school TOP bullet train OP took Kyoto to went though

Tokyo de wa chigau rashii desu ne.
Tokyo in TOP diﬂ‘erent seem BE IP

(22-26 English translation) For example, I was raised in Yamaguchi prefecture so I

27:

28:

29:

30:

31:

went to Kyushu for elementary school trip, and I went to Kyoto by bullet
train for middle school, but it seems different in Tokyo, doesn't it?

Mao,
Well

sonna koto demo ii shi,
that thing ﬁne and

hutari tomo shuugaku ryoko no omoi de ga tokuni naiyoo nara
two of you school trip of memory SUB especially not exist if

saikin itta ryoko no hanashi demo korekara ikitai tokoro no=
recently went trip of story future want to go place of

= kotodemo kekko desu.
NM ﬁne BE

(27-31 English translation) Well, that kind of thing is ok and if two of you don't have

32:

33:

34:

35:

36:

memory of your school trip particularly, story about a trip you took
recently or a place you would like to go in the future would be ﬁne.

Kyoto no hanashi ni modori masu go,
Kyoto of story to return but

shuugaku ryoko no ato, Kyoto e kojin tekina ryoko de itta no wa,
school trip of after Kyoto to private trip as went NM TOP

10 nen kurai mae ga hajimetede,
10 years about ago SUB the ﬁrst time

sonokoro, ‘
at that time

amerikajin no yuujin ga Kyoto no urasenke de ocha no benkyo o=
American of ﬁ'iend SUB Kyoto of Urasenke at tea ceremony OP study

169

37: =shiteita node at ni ikimashita.
So meet to went

(32-37 English translation) Going back to the story of Kyoto, 10 years ago was the
ﬁrst time when I went to Kyoto privately after school trip and because my
American friend was studying tea ceremony at Urasenke, I went to see her.

38: Tokyo kara,
however

39: kuruma de dekaketara,

car by went
40: yukini natte Shimatte,
snow become
41: koosoku doro ga tochu de heisa ni notte shimatta n desu.

High way SUB on the way had closed

(38-41 English translation) When I drove my car from Tokyo, it became
snowing and then high way was closed between Tokyo and Kyoto.

42: Kaeru hoo mo heisa sareta node,
Going back direction also cholsed been so

43: Shizuoka no saki kurai kara kokudo de Kyoto made ikimashita.
Shizuoka of a little further about from main road by Kyoto till went

(42-43 English translation) The high way to going back was also closed, so I
drove on the main road to all the way to Kyoto from somewhere around
Shizuoka.

44: Taihen na omoi o shimashita ga,

tough feeling OP did but

45: ano yuki no Kyoto wa subarashi katta desu.
Ah snow of Kyoto TOP wonderful BE

(44-45 English translation) It was tough but well, Kyoto in snow was
Wonderful.

46: Kinkakuji nanka mo yukigesho no naka de mita keshiki wa =
Kinkakuji temple and so on also snow of in saw view TOP

4 7: = sorewa sorewa utsukushikute,

170

very very beautiful and

48: ii omoide ni narimashita.
Good memory became

49: To iu yoo ni ohanashi shite itadakemasu ka.
QT say way in talk Q
(4649 English translation) Like Kinkakuji temple, the view seen in snow was

very very beautiful and it has become a good memory. Would you speak like
this?

171

APPENDIX C: Transcrips of conversation of video clips

Video 3 (mid verbal backchannels, 13+mid nods, 5)

1A: Senshuu so,
last week-1P
28: an
yeah
3A: Suiyoobi no asa,

Wednesday-of morning
48: H

5A: Ohiru madeni burokkubasutaa no bideo o kaesa nakuchanannakatta no yo.
Noon by Blockbuster's of video P return must IP + IP

6B: un
yeah

(1-6 English translation) Last week, Wednesday morning I had to return a video to
Blockbuster Video by noon.

7A: Sorede,
so

8A: 1] ji goro uchi dete,
llo'clock home leave

9B: un
yeah
1 0A: Bideo kaeshite kara,
video return after
1 18: H

1 2A: Wells hooru no mae made kite,
Wells Hall of front to came

138: H

1 4A: Sutajiamu no mae no paakingu ni hairoo t shita no ne.
Stadium of front of parking in enter P intended IP+IP

172

1 SB: un
yeah

(7-15 English translation) So I left around 11 o’clock and after returning the video, I
drove to Wells Hall and was about going into the parking lot in front of the
Stadium.

16A: Soshitara ,
then

17A: watashi no mae no mae no hito ga kaado,
I of before before of person SUB card

18A: iretan dakedo geeto ga agannai no yo.
Put but gate SUB didn't liﬁ IP+IP

19B: un
yeah

(16-19 English translation) I found someone in the 2 cars ahead inserted the gate
card but the gate did not open

20A: Watashi mo IIji han no kurasu ni okuresoodattashi,
I too 11 o'clock half of class for being late

21A: hayaku shitette kanji datta no.
hurry up felt like IP

22B: un
yeah

(20-22 English translation) I was almost late for the 11:30 class so I was kind of
irritated by the trouble.

23A: De,
Then
24B: un
yeah

25A: sono hito wa nanto geeto o oshiagete haitte icchatta no.
that person TOP gate P push up enter IP

268: hee:

173

(23-26 English translation) Then I saw that person got into the parking lot by
pushing up the gate by his car.

27A:

28B:

Watashi no moe no kuruma mo sore m' tsuzuito no ne.
I of before of car also it followed IP+ IP

uso:
really

(27-28 English translation) And then another car followed him.

29A:

308:

31A:

32B:

33A:

34B:

De ne,
So IP

un
yeah

watashi mo moo jikonn go nokotto shi,
I also no more time left

un

yeah

moe no kuruma ni tsuzuite issh ni haicchatta no yo.
Before of car to follow together entered Ip+IP

un
yeah

(29-34 English translation) So, I followed them because I didn’t have time.

35A:

36A:

37B:

38A:

39B:

Dokoro,
So

wotoshitochi no see de geeto go koworete taro dooshiyootte ne,
Us gate SUB broken if what should do IP

H

chotto doki doki shiton do kedo,
a little worried though

an
yeah

(35-39 English translation) I was a little worried if we actually broke the gate by

174

pushing up.

40A: yokujitsu koworete nokotto kara onnshinn shitondo.
Next day wasn't broken so relieved

41B: soo:

(40-41 English translation) But when I found it was ok on the next day, I was
relieved.

175

APPENDIX D: Questionnaires

ﬁmnarne;

EB age: ﬁ_ﬁ

mill sex: %(M)_tr(l=)

than! place of birth : #13/ LE [Z/Fﬁ
IE/ unnaonen length of residence in Tokyo or kanto area:
”approximately _ﬂiyears (_m» B_ﬁ§i't’ age : from - to)
names: (current occupation): Circle one.

$5Estudenl, Itﬁ housewife If} clerk

Elsales clerk 133* service industry ﬂﬂteacher {with others

ﬁgfﬂﬁi &: 055% relationship to the conversation partner:
1. 3153' 125 b l. \ very close
2. 793% U a b l. \ relatively close
3. «5‘5 Eb‘tUiliﬁl/U close
4. kiﬁwiﬂ U 3‘“ just an acquaintance

aﬁﬁﬁt ii a ”Silﬁﬂ’i [27:11) Did you talk to your partner naturally?
31531: E ﬁllﬁﬂ‘k Very naturally

7C“?- 5 2’5 lCﬁﬂ‘L‘: Overall naturally

5 K U a ﬂitﬁﬂ’nb‘ ‘3 7:: not very much

as: a museum '3 7: not at all

PPS”?

I76

APPENDIX E: Rating sheet for video clips

ﬁﬁilname

(Questionnaire in Japanesewriting)

1:711:53 ‘CUéAtDEE’ESEE 121:7}1'1‘3 U. {'hi’il’W) €771“?- 71:?th

179‘ B 6 @HMICEZ‘CTE (A (Please pay attention to the person's listening behavior in
the scene and rate each scene according to the questions listed below)

Ei‘ivideo 1.
1 t: @Awﬁﬁli E #575 tans-rm. Did the person in the scene behave naturally?

¢reanésl4lﬂ3lzl FEE?“

:—

very naturally unnaturally

lekliiﬁ¥®ﬁéiiﬁgg< MUTU6 tgui'Tmo Is the person's behavior

attentive?

é‘t‘biiligﬁﬂgu'ti‘éﬂ 4 IE3 | 2 IQGIE éﬂa‘thnh
1

very attentive not at all

3
intﬂ‘ﬁbfhélﬁlccwko); amuse: Bht‘iﬁb bUtEUSE'J'fﬁ

0 Does the person's behavior annoy you?

_t‘C‘bi—atjbh 51 4 | ‘ 3 2 C. ‘ U1
very much not at all

4‘6 Linth‘t‘zm EfiwAiﬁotB. COAODJ: 5H$3Tﬁéﬂl < 2:53.“
i T71). If you were the person in the scene, would you behave like her?

W 5| 4 IEEBIEMUI Z I 2(251

very similar very diﬁ‘erent

The same questions follows for video 2-6.

177

BIBLIOGRAPHY

178

[The translations of the Japanese titles are mine]

Brown, Penelope and Stephen C. Levinson. 1987. Politeness Some universols in
language usage. Cambridge University Press.

Brown, Bilian and George Yule. 1983. Discourse analysis. Dambridge: Cambridge
University Press.

Clancy, Patricia M. 1982. Written and spoken style in Japanese narratives. In Deborah
Tannen, Spoken and written language exploring on language and literacy. 55-
76. Norwood, Ablex.

Clancy, Patricia M, Sadra A. Thompson, Ryoko Suzuki, and Hongyin Tao. 1996. The
conversational use of reactive tokens in English, Japanese, and Mandarin.
Journal of Pragmatics 26. 355-3 87.

Cook, Haruko M. 1990. An indexical account of the Japanese sentence-fmal particle no.
Discourse processes 13: 401-439.

Cook, Haruko M. 1992. Meanings of non-referential indexes: A case study of the
Japanese sentence ﬁnal particle ne. Text 12 (40): 507-539.

Cook, Haruko 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.

Dittman, Allen and Lynn Llewellyn. 1968. Relationship between vocalizations and head
nods as listener responses. Journal of Personality and social Psychology 9: 79-
84.

Duncan, Starkey. Jr., and Donald W. Fiske. 1977. Face-to-foce interaction: Research,
method, and theory. Hillsdale, NJ: Lawrence Erlbaum.

Duncan, Starkey Jr.. 1974. On the structure of speaker-auditor interaction during
speaking turns. Language in Society. Vol. 2, 161-180. Cambridge University
Press: New York.

Duranti, Alessandro and Charles Goodwin. 1992. Rethinking context language as an
interactive phenomenon. Cambridge University Press

Eckert, Penelope. 1993. Cooperative competition in adolescent "Girl Talk." In D Tannen,
Gender and conversational interaction. Oxford University Press.

Ford, Cecilia E., Barbara A. Fox and Sandra A. Thompson, 1996. Practices in the
construction of turns: The "TCU" revisited. Pragmatics 6 (3): 427-454.

179

Ford, Cecilia E and Sandra A. Thompson. 1996. Interactional units in conversation:
syntactic, intonational, and pragmatic resources of the management of turns. In
E. Ochs, E Schegloff, and S. Thompson. Interaction and grammar. Cambridge:
Cambridge University Press.

Fries, Charles C. 1952. The structure of English. New York: Harcourt, Brace.

Furo, Hiroko. 2001. Turn taking in English and Japanese: projectability in grammar:
intonation, and semantics. New York: Routledge.

Goﬁ‘man, Erving. 1986. Frame analysis: An essay on the organization of experience.
Peprint. Boson: Northeastern University Press.

Goodwin, Charles. 1979. The interactive construction of a sentence in natural
conversation. In Geroge Psathas. Everyday language: Studies in
ethnomethodology. 97-121. New York: Irvingtin.

Goodwin, Charles 1986. Between and within: alternative sequential treatments of
continuers and assessments. Human studies 9; 205-217.

Gumperz, John and Dell Hymes. 1972. Directions in sociolinguistics; The ethnohgraphy
of communication. Holt, Rinehart and Winston, Inc.

Hayashi, Makoto, 1999. Where grammar and interaction meet: A study of co-participant
completion in Japanese conversation. Human Studies 22: 475- 499.

Hayashi, M. and J. Mori. 1998. Co-construction in Japanese revised: We do "ﬁnish each
other's sentences". In N. Akatsuka, H. Hoji, S. Iwasaki, Sung-Ock Sohn, and S.

Strauss, eds. Japanese/Korean Linguistics, 7, 77-93. Pp141-157. Stanford, CA:
Center for the Study of Language an Information.

Hill, Beverly, Sachiko Ide, Shoko Ikura, Akiko Kawasaki and Tsunao Ogino. 1986.
Universals of linguistic politeness: Quantitative evidence from Japanese and
American English. Journal of Pragmatics 10, 347-371.

Hinds, John. 1978. Conversational structure: an investigation based on Japanese
interview discourse. In J. Hinds and 1. Howard. Problems in Japanese syntax

and semantics. 79-12 1 . Tokyo: Kaitakush.

Hinds, John. 1982. Japanese conversational structures. Lingua 57: 301-326.
North-Holland Publishing Company.

Horiguchi, Junko. 1988. Komyunikeshon ni okeru kikite no gengo kodo (Listeners'
behavior in communication). Nihongokyoiku 64: 13-26. Nihongokyoikugakkai.

Horiguchi, J. 1997. Nihongo kyooiku to kaiwa bunseki. [Japanese language teaching and

180

Conversation Analysis]. Tokyo: Kuroshio Publishers.

Hymes, Dell. 1974. F undations in sociolinguistics: An ethnographic approach.
University of Pennsilvenia.

Ide, Sachiko. 1979. Daigakusei no hanasi kotoba ni mirareru danzyo sai (Sex differences
in college student conversation). A paper presented at the 6th ICU
Sociolinguistic Symposium.

Ide, Sachiko. 1982. Japanese sociolinguistics: Politeness and women's language. Lingua
57, 357-385.

Ide, Sachiko. 1991. How and why do women speak more politely in Japanese? In
Sachiko Ide and N. H. Mcgloin (eds.), Aspects of women’s language. Tokyo:
Kuroshio. 63-79.

Ikegami, Teizo. 1952. 'Hai' to 'iie' ('yes' and 'no'). Kokugokubun 217. Kyoto: Kyoto
daigaku.

Joos, Martin. 1967. T he ﬁve clocks. NY: Harcourt Brace and World.

Kendon, Adam. 1967. Some functions of gaze-direction in social interaction. Acto
Psychologico 26, 22-63

Kita, Sotaro. 1996. Aizuchi to unazuki kara mita nihonj in no taimen komyunikeshon.
(Japanese face-to-face communication with respect to backchannels and nods).
Nihongogoku vol. 1 5: 58-66. Meijishoin.

Kitagawa, Chisato. 1980. Saying 'yes' in Japanese. Journal of Pragmatics 4: 105-120.

Kogure, Masato. 2003. Gender differences in the use of backchannels: do Japanese men
and women accommodate to each other? Ph. D. dissertation, the University of
Arizona.

Kurosaki, Yoshiaki. 1987. Danwa shiko jo no Aizuchi no unyo to kino-Hyogoken Takino
hogen ni tsuite (Performance and Functions of backchannels in the discourse -
Takino dialect in Hyogo prefecture). Kokugogaku 150: 122-109. Kokugo gakkai.

Lakoff, Robin. 1975. Language and women 's place. New York: Harper and Row.

Lauwereyns, Shizuka. 2000. Hedges in Japanese spoken discourse: a comparison
between younger and older speakers. Ph.D. dissertation, Michigan State

University

Lebra, TS, 1976. Japanese patterns of behavior. Honolulu : University of Hawaii press.

181

Lerner, Gene H. 1991. On the syntax of sentences-in-progress. Language in society
20; 441-458.

Lerner, Gene and Tomoyo Takagi, 1999. On the place of linguistic resources in the
organization of talk-in-interaction: A co-investigation of English and Japanese
grammatical practices. Journal of Pragmatics 31: 49-75.

LoCastro, Virginia. 1987. Aizuchi: A Japanese conversational routine. In Discourse
across culture. Edited by Larry E. Smith 101-13. NewYork: Prentice-Hall
International.

Matsuda, Yoko. 1988. Taiwa no Nihon-go kyooiku-gaku: Aizuchi ni kanren shite
(Teaching of Japanese dialogues: With reference to backchannels). Nihongo-
goku 7.59-66.

Maynard, S. 1986. On back-channel behavior Japanese and English casual conversation.
Linguistics 24: 1079-108.

Maynard, Senko K. 1987. Interactional functions of a nonverbal sign: head movement in
Japanese dyadic casual conversation. Journal of pragmatics 11: 589-606.

Maynard, S. 1989. Japanese conversation: Self-contextuolization through structure and
interactional management. Norwood, NJ: Ablex.

McGloin, H.Naomi. 1990. Sex difference and sentence-ﬁnal particles. Japanese /
Korean Linguistics: 23-41.

McGloin, H. Naomi. 1997. Hai and Be: An Interactional Analysis. Japanese/Korean
Linguistics, vol. 7.

Minami, Fujio. 1981. Nichijo kaiwa no wadai no suii-Matsue text 0 shiryo to shite.
In F ujiwaro yoichi sensei koki kinen ronshu- hogen ronshu. Tokyo: Sanseido
Publisher.

Minami, Fujio. 1987. Danwa kodo ron (Speech acts theory). In Danwakodo no shoos
(Aspects of speech acts). Kokuritsu kokugo kenkyujo hokoku 92. Tokyo:
Sanseido.

Miyaji, Atsuko. 1059. Ukekotae (Response). Kokugogaku 39: 85-98.

Miyazaki, Sachie. 2001a. Native and non-native speakers' perception of other's listening
behavior in Japanese. Term paper for LIN810. Micigan State University

Miyazaki, Sachie. 2001b. Miburi no aizuchi toshite no kikite no unazuki (Listeners' nods
asnon-verbal backchannels). Paper presented at Nihongo kyoiku gakkai shunki
taikai. Tokyo.

182

Mizutani, Nobuko. 1984. Nihongo kyooiku to hanashikotoba no jittai [Japanese language
teaching and the reality of spoken language]. In Kindaichi Haruhiko hokase koki
kinen ronbunshuu [The collection of papers in honor of Dr. Haruhiko

Kindaichi’ s seventieth birthday]. Tokyo: Sanseido.

Mizutani, Nobuko. 1988. Aizuchi-ron (The theory of backchannels). Nihongo-goku
7(12). 4-11.

Mizutani, Nobuko. 1993. Kyowa kara taiwa c (From Taiwa (conversation) to Kyowa
(collaborative speech). Nihongogoku. 4-10. Meijishoin.

Nakamura, Kanae. 1998. Danwa ni okeru nichiei taisho kenkyu _ kikite no gengo kodo
ni chakumokushite (Contrastive study of discourse between Japanese and
English - looking at listeners' speech acts). Studies in Language and culture 7:
38-50. Tokyo Woman's Christian University.

Nishihara, Suzuko. 1995. Ibunka sesshoku ni okeru higengo kodo (Non-verbal behavior
in the contact of different cultures). Nihongogoku 31-41. Meijishoin.

Ochs, Elinor. 1992. Indexing gender. In A. Duranti & C. Goodwin, Rethinking context.
Cambridge. 335-358

Okamoto, Shigeko. 1995. "Tasteless" Japanese: less "feminine" speech among young
Japanese in Women. In K. Hall and M. Bucholts (eds.), Gender articulated:
Language and socially constructed self New York: Routledge. 297-325.

Okamoto, Shigeko. 1998. The use and non-use of honoriﬁcs in sales talk in Kyoto and
Osaka: Area they rude or friendly? In N. Akatsuka, H. Hoji, S. Iwasaki, Sung-
Ock Sohn, and S. Strauss, eds. Japanese/Korean Linguistics, 7: 141-157.
Stanford, CA: CSLI

Okamoto, Shigeko. 1999. Situated politeness: Manipulating honoriﬁc and non-honoriﬁc
expressions in Japanese conversations. Pragmatics 8(2): 51-74.

Okamoto, Shigeko. 2002. Ideology and social meanings: rethinking the relationship
between language, politeness, and gender. In S. Benor and M. Rose (eds.),
Gendered practices in language. Stanford, CA: Center for the Study of
Language an Information .

Okamoto, Shigekko and Shie Sato. 1992. Less feminine speech among young Japanese
females. In Kira Hall, Mary Bucholtz, and Birch Moonwomon (eds.), Locating
power: Proceedings of second Berkeley women and language conference.
Berkeley: Berkeley Women and Language Group. 478-488.

Okamoto, Shigeko, and Shie Sato. (to appear) Culture and interactional sytles: The

183

interpretation of reactive tokens in Japanese conversations. In N. Akatsuka, S.
Iwasaki and S. Strauss (Eds.) Japanese/Korean Linguistics 13. StanfordzCSLI

Ono, Tsuyoshi, and Eri Yoshida. 1996. A study of Co-construction in Japanese: We don't
ﬁnish each other's sentences. In N. Akatsuka, S. Iwasaki and S. Strauss (Eds.).
Japanese/ Korean linguistics 5: 115-119. StanfordzCSLI

Oso, Mieko. 1988. Eigo no aizuchi (English backchannels) Nihongogoku vol.17-12:
46-51 . Tokyo:Meijishoin

Philips, Mieko K. 1 998. Discourse markers in Japanese: connectives, ﬁllers and
interactional Particles. Unpublished Ph.D. dissertation,.Michigan State
University

Reynolds, Katsue A. 1985. Female speakers of Japanese. Feminist Issues, Fall. 13-46.

Sacks, Harvey, Emanuel A. Schegloff, and Gail Jefferson 1974. A simplest systematics
for the organization of tum-taking for conversation. Language 50.4: 696-735.

Sakuma, Mayumi. 1987. Bundan nintei no ichi kijun (I) Teidai hyogen no tokatsu.
Bungeigengokenkyu Gengohen 11. Japan: Tsukuba University.

Sato, Shie and Sigeko Okamoto. 1998. Reexamination of Japanese "cooperative
communication style. In Pragmatics in 1998 selected Papers from the 6th
International pragmatics conference, vol.2.

Schegloff, Emanuel A..l982. Discourse as an interactional achievement: Some use of
"UH-HUH" and other things that come between sentences. In Deborah Tannen
(Ed.), Georgetown University Round Table on Languages and Linguistics,
Analyzing discourse: Text and talk, 71-93. Wachington DC: Georgetown
University Press

Shibamoto, Janet. 8.. 1985. Japanese women's speech. New York: Academic Press.

Shibatani, Masayoshi.1990. The language in Japan. Great Britain: Cambridge University
Press.

Smith, Janet S. 1992. Women in charge: Politeness and directives in the speech of
Japanese women. Language in society 21: 59-82.

Strauss, Susan and Yumiko Kawanishi. 1996. Assessment strategies in Japanese,
Korean, and American English. Japanese/Korean Linguistics 5: 149-166.

Sugito, Kiyoki. 1989. Kotoba no aizuchi to miburi no aizuchi (Verbal backchannels and
non-verbal bakchannels). Nihongokyoiku 67. Nihongo kyoiku gakkai. 48-59.

184

 

Sugito, Kiyoki. 1989. Gengo kodo ni tsuite no kimarikotoba (Fixed expressions about
Speech acts). Nihongogoku vol.8.2. Meijishoin.4-14.

Szatrowski, Polly. 1989. Aizuti to rizumu (Aizuti 'back channel utterances' and rhythm).
Gekaan Nihongo 3. 32-35.Sanada, Shinji. 1990. Sedai to kotoba. Nihongogoku
vol. 9.4. Meijishoin. 4-11.

Szatrowski, Polly. 1991. Kaiwa bunseki ni okeru 'tani' ni tsuite - 'wadan' no teian (on a
'unit' for discourse analysis - proposal for the concept of 'wadan'). Nihongogoku
vol. 10.10. Meijishoin. 79-90.

Szatrowski, Polly, 1993. Nihongo no denwo no koozoo bunseki: Kanyuu no sutoratejii no
koosatsu. (An analysis of Japanese Discourse Organization: An Investigation of
Invitation Strategies). Tokyo: Kuroshio Publishers.

Szatrowski, Polly, 2000. Relation between gaze, head nodding and aizuti 'back channel'
at a Japanaese company meeting. Berkeley Linguistics Society 26. 283-294.

Szatrowski, Polly. 2003. Gaze, head-nodding and aizuti 'back channel utterances' in
information presenting activities. Japanese/Korean Linguistics 11: 119-132. ed.
By Patricia Clancy. Stanford: CSLI.

Tanaka, Hiroko. 2000. The particle ne as a tum-management device in Japanese
conversation. Journal of Pragmatics 32: 1135-1176

Tannen, Deborah. 1982. Ethnic style in male-female conversation. In J. Gumperz,
Language and society. 217-231. Cambridge.

Tannen, Deborah. 1984. Conversational style: Analyzing talk among ﬁ'iends. Norwood,
NJ: Ablex.

Tannen, Deborah. 1990. Gender differences in conversational coherence: Physical
alignment and topical cohesion. In Bruce Dorval ed. Conversational coherence

and its development. Pp. 167-206. Norwood, NJ: Ablex. Rpt. In Gender and
discourse. New York and Oxford: Oxford University Press

Tannen, Deborah. 1993a. The relativity of linguistic strategies: Rethinking power and
solidarity in gender and dominance. In D. Tannen ed., Gender and
conversational interaction, 165-188. Oxford: Oxford University Press.

Tannen, Deborah. 1993b. Framing in discourse. New York and Oxford: Oxford
UniversityPress

Trudgill, Peter. 1972. Sociolinguistics: An introduction. Harmondsworth, England:
Penguin Books.

185

Uyeno, T. 1971. A study of Japanese modality-A performative analy sis of sentence
particles. Unpublished dissertation. University of Michigan.

Uyeno, Tazuko. 1999. Komyunikeshon to nihongo (Communication and Japanese).
Nihongogoku 18: 28-35. Meijishoin

Watanabc, Suwako. 1993. Cultural differences inframing: American and Japanese group
discussions. In Deborah Tannen, Framing in discourse: 176-209. New York and
Oxford: Oxford University Press.

White, Sheida. 1987. Functions of backchannels in English: A crosscultural analysis of
Americans and Japanese. Unpublished doctoral dissertation, Georgetown
University.

White, Sheida. 1989. Baekchannels across cultures: A study of Americans and Japanese.
Center for Development of Early Education Kamehameha Schools, Honolulu.
Language Society. 18: 59-76.

Yamada, Haru, 1992. American and Japanese Business Discourse: A C omporison of
Interactional Styles. Norwood, NJ: Albex.

Yamada, Haru, 1997. Diﬂerent Games, Diﬂerent Rules: Why Americans and Japanese
Misunderstand Each Other. Oxford: Oxford University Press.

Yim and Lee. 1995. Aizuchi kodo ni okeru kachikan no kannichi hikaku (Comparison of
value of backchannel behavior between Korea and Japan). Sekai no nihongo
kyoiku 5: 239-251. Kokusai koryu kikin.

Yngve, V. H. 1970. On getting a word in edgewise. Chicago Linguistics Society 6,
567-578.

Yoshioka, Yasuo. 1992. Gengo seikatsu (Language life). Kokugogaku 169: 102-110.

186

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

m[ll][[Illlllllllllnlllllll