W”. . if.

1:

\A .
. .urém;

s
L
V

X...
.o 1.)

331......
I'llv‘r vll|ibuQ
t. V. 2

\ t A l .1
7.3:: , :1. ‘
if. by...“ :9
. Va . .53..

‘ . {3:
:1
.5
u.

. .9er
I). 3
1’- .
)0. on
. .i Q «I... .Iv

. ‘ 1.5..
amiwmm...»
E.

‘ nr: ,.
r'i‘

1 ) ,
. ’ ‘
.1“?

drain.”

a“; if:
r 4..”
n W,
‘A {‘5
’r

"

 

 

r
I.
,.
r
:1"

 

 

 

A... 1....
5:1:

 

 

.‘u -
V .33.

r .W 53:5!

RSITY LIB

IIWTW [iii 1W; lug! W"

This is to certify that the

thesis entitled

THE FUNCTIONS OF THE JAPANESE
CONJUNCTION (SORE)DE AS A
DISCOURSE MARKER

presented by

Rika Ito

has been accepted towards fulfillment
of the requirements for

M.A. degree in Linguistics

 

 

 

 

Major professor

Date November 13, 1995

0-7 639 MS U i: an Affirmative Action/Equal Opportunity Institution

 

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE ll RETURN BOX to mow. thb checkout from your record.
TO AVOID FINES mum on or baton duo duo.

DATE DUE DATE DUE DATE DUE

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

fi—L—‘fi' ’ 7 \‘m_ ,7

THE FUNCTIONS OF THE JAPANESE CONJUNCTION (SORE)DE
AS A DISCOURSE MARKER

BY

Rika Ito

A THESIS

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

MASTER OF ARTS

Department of Linguistics and

Germanic, Slavic, Asian and African Languages

1995

ABSTRACT

THE FUNCTIONS OF JAPANESE CONJUNCTION
(SORE)DE AS A DISCOURSE MARKER

By

Rika Ito

This paper examines the functions of the Japanese conjunction (sore)de
(the most direct counterpart of English 'and' ). Schiffrin describes the functions
of and as 1) a structural coordinator of ideas and 2) a marker of speaker
continuation (1987: 152). The data show that (sore)de plays the same roles
and additional ones: it functions to elicit an interlocutor's continuation and to
signal the speaker‘s request for further information when prefaced to questions.

Some differences are observed between the two forms. Sorede seems
to prefer coordinating global level of units, while de works more on the local
level. De is more frequently used when signaling continuation and eliciting an
other's continuation. Some gender related differences in a male preference for
sorede are also observed. It seems to be related to what sorede connotes:

formality, distance and difference; hence. it is a ’men's trait' (T annen 1990).

Copyright by
RIKA ITO
1 995

To my parents

ACKNOWLEDGMENTS

My deepest gratitude goes to Dr. Dennis Preston, Chair of my Thesis
Committee. Without his insightful comments, suggestions, encouragement, and
endless patience, I would not have been able to finish this thesis.

I would also like to thank Dr. Mutsuko Endo Hudson and Dr. Mary
Bresnahan for serving as my Committee members and for their variable
suggestions and encouragement.

My thanks go to my friends who let me tape-record their conversations,
which provided the data for this study. i also wish to thank to my friend, Gabriela

Alfaraz who spent so many hours proof-reading.

List of Tables
List of Figures
CHAPTER I
1.1

1.1.1

1.1.2

1.2

1.3

1.4
CHAPTER II
2.1

2.1.1

2.1.2

2.1.3
2.2

2.3

2.3.1

2.3.2

2.3.3
CHAPTER III

TABLE OF CONTENTS

introduction

Background: Definitions and assumptions
Phenomena to be investigated

Goals

Definition of discourse markers

Research Questions

Outline of this study

Review of the related literature

Properties of the conjunction and in English
Logical Connective

Conversational lmplicature and referential
meaning

Discourse marker

Property of the conjunction (sore)de in
Japanese

Properties of other conjunctions in Japanese
Demo, dakedo ’but'

Dafte 'but, because'

Dakara 'so, therefore'

Method and procedures

vi

\immmmhmmd

co

16
16
17
17
19

3.1

3.2

3.3
CHAPTER IV
4.1

4.2

4.2.1

4.2.2

4.2.2a
4.2.2b

4.2.3

4.2.3a
4.2.3b
4.2.30
4.2.3a
CHAPTER V

5.1
5.1.1
5.1.2
5.2
5.2.1
5.2.1.1
5.2.1.1a
5.2.1.1b
5.2.1.2

Data collection

Participants

units of talk

Results

Quantitative results

Qualitative results

(Sore)de as dictionary definition:
expressing reason & cause

In environments shared with other markers
De as demo, dakedo 'but'

(Sore)de to connect two events 'and then'
(Sore)de as a discourse marker
Coordinating units

Speaker's continuation

Eliciting other's speech

Coordination, continuation, and involvement
Analysis of the functions of (sore)de as a
discourse marker

High frequency of (sore)de

Coordination and continuation

Sorede vs. de

Sociolinguistic variables

lnteractional factors

Genre

Written vs. spoken discourse

Narrative vs. non-narrative

Situations and relations

vii

19
20
2O
24
24
27
27

28
28
29
31
31
37
46
49
51

51
51
52
53
54
54
54
55
56

5.2.2
CHAPTER VI
Appendix 1
Appendix 2

List of references

Individual characteristics

Conclusion

Transcription methods and conventions

Selected data

viii

57
61
63
65
71

Table 4.1
Table 4.2
Table 4.3
Table 5.1
Table 5.2

LIST OF TABLES

Frequency of conjunctions in causal dyads in Japanese
Frequency of conjunctions (in same-sex & mixed sex pair)
Sorede vs. de

Frequency of sorede & de in 100 words

Frequency of sorede & de

25
26
26
56
57

Figure 2.1
Figure 2.2
Figure 2.3
Figure 2.4

Figure 3.1
Figure 4.1
Figure 4.2

LIST OF FIGURES

Summary of the conjunction and in English
Summary of and in Japanese

Summary of the definition of sorede, sokode, and de
Summary of the previous studies on functions of
Japanese conjunctions

Participants and topics in their conversations

Planes of talk on which (sore)de functions

Discourse marker as contextual coordinates

1O
11
16
18

2O
50
50

CHAPTER I
Introduction

1.1 Background: Definitions and assumptions
1.1.1 Phenomena to be investigated

This thesis examines the functions of the Japanese conjunction (sore)de
'and' as a discourse marker: not only its unique functions but also functions
shared with other discourse markers such as dakara 'so', datte 'but, because'
and demo, dakedo 'but'. The analysis includes whether there are social
differences, specifically, gender, in its use in casual conversation.

This study was motivated by the following four factors. First, although
research on discourse markers has been done both in English (Blakemore
1987, Fraser 1990, Heritage 1994, Holmes 1986, Romaine and Lange 1990,
Schiffrin 1987), and in Japanese (Cook 1993, Karatsu 1994, Maynard 1992a
1992b 1993, Onodera 1993, Saito 1992, Uemura and Tabuki 1994), the
functions of the Japanese conjunction (sore)de, ranked the most frequently
used conjunction in Japanese conversation (Onodera 1993:9-11), have not yet
been explored. Second, the high frequency of (sore)de usage may correlate
with a wide range of functions. Schiffrin (1987), for example, notes that and
occurs in various environments shared by other conjunctions such as but and
so. Third, even though gender is one of the most important factors in analyzing
Japanese discourse, little research exists on this topic. Only a few studies have
focused on how gender differences influence the use of discourse markers
(Holmes 1986, Saito 1992); several others have looked at variations in men's
and women's use of honorific expressions and final particles to mark politeness
(lde 1982 1991, Loveday 1986, Shibamoto 1985). Fourth, more careful
characterization of the functions of discourse markers can be applicable to

1

2
language teaching and learning. For instance, such infonnaticn may enhance

understanding of coherence and its role in conversation; it is a potentially

important factor in improving proficiency level (Uemura and Tabuki 1993).

1.1.2 Goals

The goals of this study are: (1) to identify and describe the functions of
the conjunction (sore)de as a discourse marker in casual conversation:
specifically, (1a) to identify unique functions and determine; (1 b) to examine
whether there exist any functions shared with other markers, such as demo,
dakedo 'but', dakara 'so, therefore' and datte 'but, because,‘ and (2) to
examine whether differences exist in the use and/or frequency of (sore)de

between males and females.

1.2 Definition of Discourse Markers

Schiffrin's (1987) primary interest lies in the device of coherence in
discourse. She sees discourse markers as one of the important devices which
contribute to discourse coherence. Schiffrin (p. 31) operationally defines
discourse markers as 'sequentially dependent elements which bracket units of
talk.‘

Schiffrin (pp. 24-28) proposes a discourse model containing five
separate planes, the discourse markers play important roles in these planes.
The Exchange Structure is related to the mechanics of turn-taking. For
example, so and y'know work on this plane since both are used at potential
“transition-relevance place (T RP).' The Action Structure reflects the sequence
of speech acts; for example, and marks speaker's continuation and but marks
contrast. The ldeaticnal Structure deals with cohesive, topic, and functional
relationships between the ideas (propositions) found within the discourse. The

3
primary functions of all the conjunctions (and, but, or, so, and because) as well

as temporal adverbs (now, and then) are in this structure since their
grammatical properties contribute to their discourse functions. The Participation
Framework captures both speaker/hearer relations, and speaker/utterance
relations. For instance, the primary function of I mean belongs here because
of its focus on speaker orientation. The Information State focuses on cognitive
capacities, including the organization and management of knowledge and
meta-knowledge. Oh and y'know primarily function on this plane.

Schiffrin proposes that the primary function of and is in the ldeaticnal
structure because it marks textual coordination, but it works in the action
structure and the exchange structure as well since and marks speaker-
continuation (pp. 315-316).

The fact that markers function on different discourse planes and provide
clues to discourse context leads Schiffrin to propose a theoretical definition of
markers. According to her, discourse markers function as 'contextual
coordinates,’ whose fundamental role is to 'index an utterance to the local
context in which utterances are produced and in which they are to be
interpreted (p. 326). More precisely, discourse markers are (1) 'participant
coordinates,‘ which index utterances to the speaker (proximal) and/or the
bearer (distal), and (2) ‘textual coordinates,’ which index utterances to prior
(proximal) and/or upcoming (distal) text (p. 323). Because of these functions,
Schiffrin concludes that discourse markers provide coherence in discourse (p.
330).

Schiffrin (p. 328) summarizes the suggested conditions for a discourse
marker as follows:

1. Syntactically detachable from a sentence

2. Commonly used in initial position of an utterance

4

3. Has a range of prosodic contours
e.g. tonic stress and followed by a pause, phonological reduction

4. Operates at both local and global levels of discourse and on different
planes of discourse
is. it either has to have no meaning, a vague meaning, or to be

reflexive (of the language, of the speaker)

These conditions have also been adopted among other researchers, including
Japanese scholars (Maynard 1992a 1992b 1993, Onodera 1993, Saito 1992,
Uemura and Tabuki 1994).

1.3 Research Questions

Based on the points above, the following research questions are

advanced for this study:

1

Is there any unique function which only (sore)de
performs?

'And' in English functions as a structural coordinator of ideas as
well as a marker of speaker's coordination. Does the Japanese
counterpart also work as a discourse marker? if so, what does it
mark uniquely?

Is there any functional overlap with (sore)de and other
markers?

Although 'and' and 'but' in English signal different effects
(continuation and contrasting, respectively), they both function to
continue rather than relinquish turns in the exchange structures.
Does (sore)de share some functions with other markers? If so,
what functions does it share with others?

5

3 Is there any difference In the way men and women use
(sore)de ?
Japanese has some distinct features in 'women's language,‘
particularly as regards politeness. If (sore)de Operates on the
action structure which reflects the sequence of speech acts,
and/or the exchange structure which is related with the
mechanics of turn-taking, how do these functions affect in the
way men and women use the marker?

These research questions are examined through both quantitative and

qualitative analysis of the data.

1.4 Outline of this study

The outline of this study is as follows: Chapter 1 introduces the topic,
purpose of this study, and background material; Chapter 2 summarizes
previous studies on discourse markers and properties of the Japanese
conjunction (sore)de ; Chapter 3 outlines the method and procedures; Chapter
4 presents both the quantitative and qualitative results of this study; Chapter 5
analyses the functions of (sore)de and discusses the possible reasons for
using it; Chapter 6 summarizes the findings, and discusses the limitations of the
present study and suggestions for future research.

CHAPTER II
Review of the Related Literature

In this chapter, first, the properties of the conjunction and in English are
reviewed from the most restricted interpretation of and as a logical connective
'& ' (2.1.1) to the broadest one as a discourse marker (2.1.3) in order to see .
Second, the origins and properties of the conjunction (sore)de (2.2) are
reviewed. Third, the functions of other Japanese conjunctions which have
already studied such as demo, dakedo, 'but,’ datte 'but, because' and dakara
'so, therefore' are reviewed (2.3) in order to identify the similarity and

differences between (sore)de and these markers later in this study.

2.1 Properties of the Conjunction and in English
2.1.1 Logical Connective

Allwood et al. (1977:26-27) define a sententiallpropositionai connective
as a word or phrase usually belonging to the traditional grammatical category of
the conjunctions; only four of them (and, or, if...then and if and only if) have
been investigated by logicians. The propositional connectives are called
‘logical constants' since their meaning and function are permanent, and they
determine the truth—value of the sentences in whichthey occur. The conjunction
& closely resembles and in everyday language. However, in terms of the

function of the conjunction &, Allwood et al. (p. 32) explain:

The conjunction [& ] is used in logic to construct a
compound sentence which is true only if all the simple sentences
(these are called conjuncts) from which it is constructed are true.
If any simple sentence is false, the compound sentence or the
conjunction (it is a general custom to use the name of the logical
constant for the compound expression it creates) is also false.
Thus, (1) is true, while (2) is false.

(1) George was mad and Pitt was Prime Minister.

6

7

(2) George was mad and Pitt was King. (p. 32, emphasis as in
original).

The order of the two phrases has nothing to do with its truth-value since in logic
p & q is always equivalent to q & p.
Allwood et al. (p. 33) illustrate the distinction between the conjunction &

and its natural language counterpart and as follows:

(3) Gunnar lay down on the bed and died.
(4) Gunnar died and lay down on the bed.

Thus, while the propositional connective & combines sentences atemporally,
phrases combined with and in natural language often imply sequences of
events; the order of the conjuncts thus changes the order of the events. As we
see in the next section, this is accounted by Conversational impiicature

proposed by Grice.

2.1.2 Conversational lmplicature and Referential Meaning

There exists a difference between the logical connective & and its
natural language counterpart and. The phenomenon wherein and imparts
sequential as well as conjunctive meaning is characterized by Grice and
Levinson as Conversational lmplicature.

Grice (1975) discusses the difference between logical devices and their
counterparts in natural language by utilizing the concept of 'implicature,’ based
on general features of discourse. Cooperative conversation involves the least
amount of effort necessary for participants to achieve a common purpose. in
other words, language use needs to be efficient and effective. Grice terms this
tendency the Cooperative Principle, introducing four maxims of conversation:
Quantity ('Don't say more/less than is required'). Quality ('Don't tell a lie'),
Relation ('Be clear, be relevant), and Manner. The fourth maxim, Manner,
requires people to 'be perspicuous,’ and more specifically:

8

1. Avoid obscurity of expression.

2. Avoid ambiguity.

3. Be brief (avoid unnecessary prolixity).

4. Be orderly (p. 46).
Levinson (1983:108), building on Grice's work, cites 'be orderly' as the most
important of the sub-maxim of Manner since it is this submaxim that accounts for
the oddity of (5); in contrast, (6) follows our expectation.

(5) ??T he lone ranger rode into the sunset and jumped on his
horse.
(6) Alfred went to the store and bought some whisky.

Thus, Levinson explains that wherever two events are described, and in natural
language connotes sequentiality, such as 'and then' in addition to expressing
the truth value of both conjuncts.

And in natural language, however, implies not only time sequence but
also reason. Webster's Third New International Dictionary (1981) lists four
entries defining the meanings of the conjunction and which work on phrasal or
sentence levels. One of the definitions relates to time sequence such as 'then'
and 'at the same time ' as in (7) and (8), respectively. The other one relates to a
consequence or sequel as in (9):

(7) They drove five miles and stopped to eat.
(8) They walked and talked.
(9) i said go and he went. (lbid.:80)

Since (9) is given under the definition of 'consequence or sequel,‘ it can be
interpreted as consequence and/or the sequence of two events. However, no
further explanation is provided on the difference between consequence and
sequel in the dictionary. The interpretation seems to be determined by a

particular context in which the utterance occurs.

2.1.3 Discourse Marker

’ Schiffrin (19:37:37) examines the functions of and as a discourse marker
which work on a 'discourse level,‘ i.e. markers are not dependent on the smaller
units of talk of which discourse is composed. According to Schiffrin, items such
as and, but and or are categorized as 'discourse connectives' since these
items have grammatical properties in English, and thus function on the
ideational structure. She also observes that and is the most frequently used
item among the three; in addition, and is used in environments shared by so,
but and even 'zero' conjunction (pp. 128-129).

She proposes that and plays both an ideational role (operating on the
ideational structure) and interactional roles (operating on the exchange and
action structures) simultaneously. More specifically, and is a marker of (1) a
discourse coordinator of idea units, signaling the speaker's identification of an
upcoming unit which is coordinate in structure to some prior unit (p. 141 ), and
(2) a speaker-continuation, marking the speaker's definition of what is being
said as a continuation of his/her own prior talk (p. 147). Schiffrin (p. 152)
illustrates these functions as follows:

(1 O) (Zelda and Debby are wondering why families don't get together
as often as they used to.)

Debby: I don‘t know what it is. Maybe- maybe you're
forced- y'know there's so many
[pressures like to make] a living
Zelda: [And you have your- ]
And you have your- and you have your wizves,
your husband's family, it's too many.
Schiffrin analyzes that Zelda's tum-initial and - repeated with each successive
attempt to state her point~ displays Zelda's point as a continuation of Debby's
prior remarks, i.e. it is marking the speaker's continuation. In addition, it marks
Zelda's point as an additional reason in what becomes a jointly constructed

explanation as to why families are not as close as they used to be. Therefore,

1 0
and coordinates the upcoming unit and the previous utterance (in this data, the

one uttered by Debby). Schiffrin concludes that and serves both ideational and
interactional functions simultaneously.

Although and is a structural coordinator of ideas which has an
interactional effect as a marker of speaker continuation, it does not mean that
and conveys social and/or expressive meaning by itself. Rather, it is deeply
related to planes of discourse in which it plays.

As we have seen, the properties of the English conjunction and have
been viewed in various ways, from the most restricted one in the field of logic to
the most capacious one in the field of discourse analysis. Figure 2. 1
summarizes the prOperties of the English conjunction and.

 

Logical connective & Union (Truth-functional)

 

Conversational Sequentiality of the two events
lmplicature

 

Dictionary definition Time sequence
Webster's Dictionary a) 'and then‘ (same as lmplicature)
b) ‘at the same time‘
Reason/cause

 

 

Discourse Marker 1) Structural coordination of ideas
ideational structure
2) Speaker's continuation
Action & Exchange structure

 

 

Figure 2.1 Summary of the properties of the conjunction and in English

2.2 The property of the conjunction (sore)de in Japanese

In this section, the properties of the conjunction (sore)de are reviewed in
order to identify the functions of (sore)de in spoken discourse.

Some background: while the conjunction and in English is used to '
connect almost any constructions as long as they are of the same type (Kuno
1973:113), its counterparts in Japanese vary depending upon the grammatical

11
type which is connected, i.e. grammatically coordinate nouns, clauses, or

sentences. Figure 2.2 is a summary of Kuno (1973) and Makino and Tsutsui
(1986).

 

 

 

 

Crammaticai Grammatical Type which is Eramples
Form connect
coordinating noun + noun to, ya, ni
rticles
te gerundiveT clause + clause ik-I, omosiro-ku,
i continuative sizuka-de
form of verbs, it-tte, omosiroku—te,
adj, & copula sizuka-de
conjunctions sentence. + sentence (sore)de, sokode

 

 

 

Figure 2.2 Summary of and in Japanese (Source: Kuno 1973, Makino and
Tsutsui 1986)

The first one is 'coordinating particles'. According to Kuno (1973),to, hi, and ya
are used almost exclusively to connect nouns as follows‘:

(11) a. John to Mary to Tom(to)gakita.

D. John nl Mary nl Tom ga kita.

c. John ya Mary ya Tom ga kita.
and and S came (Kuno1973:112)

Although to, hi, and ya are all coordinating particles, there are slight
differences in nuance. Ni is used for listing exclusively and is more restricted in
usage than to, while ya is used to give non-exhaustive examples; thus, (110)
means that John, Mary, Tom and others came. To, on the other hand, does not
convey any of these additional meanings. Kuno claims that (12) sounds slightly
awkward since ni usually requires more than two items to be enumerated - the
condition is required for neither to or ya:

(12) ?John nl Mary ga kita.

 

1The following notation conventions are used (based on Maynard 1989). For example:
S subject marker T topic marker BE copulative verb, be

Q question marker 0 direct object marker LK linker

The complete list is given In appendix 1.

12
Japanese, verbs (e.g. ik-u 'to go‘), i-type adjectives (e.g. omosiroi

'interesting‘), and na-type adjectives (e.g. sizukada 'quite‘) have the 'gerundive
forms‘ (it-to, omosiro-ku-te, sizuka-d e, respectively) and 'continuative form‘
(ik-I, omosiro-ku, sizuka-d e, respectively). These forms are used when
combining predicates such as VPs and APs; the predicate of a first clause
needs be changed into either the gerundive or continuative form:

(13) . John ga Tokyo ni Ik-l, Mary ga Osaka ni iku.
S to go-ing S to go
‘John goes to Tokyo, and Mary goes to Osaka.‘

(14) . Kono hon wa omosiro—ku, wakari-yasui.
this book T interesting-BE-ing understand-easy
'This book is interesting and easy to understand.‘

(15) . Koko wa sizuka d e, suzusii.
BE-ing
'This place is quiet and cool.‘ (Kuno: 112-113)

(16) a. John ga Tokyo ni lt-te, Mary ga Osaka ni iku.
S

to go—and S to go
'John goes to Tokyo, and Mary goes to Osaka.‘

b. Kono hon wa omosiroku-te, wakari-yasui.
this book T interesting-and understand-easy-BE
'This book is interesting and easy to understand.‘

c. Koko wa sizuka de, suzusii.
this T quite and cool
‘This place is quiet and cool.‘ (Makino and Tsutsuiz464—467)

What is notable about the coordinating particles to, hi, and ya as well as the
gerundive and the continuative forms of verbs and adjectives is that they are all
translated as and in English although they do not satisfy the conditions for
discourse markers suggested by Schiffrin (p. 328). As described in (1.2), her
criteria for the prospective expressions for markers need to be, for example, (1)
syntactically detachable from a sentence, and (2) commme used in initial
position of an utterance. Although the coordinating particles, and the 'gerundive

and continuative forms might have functions in conversation, the present study

13
focuses on the examination of the conjunction (sore)de since it seems to be the

most promising candidate for a discourse marker.

Having eliminated the above-discussed elements from inclusion in this
study, we now focus on the conjunction (sore)de as an appropriate candidate
functioning as a discourse marker. in a study of the use of conjunctions in
conversation, the National Language Research Institute (NLRI) grouped
together the variants of conjunctions. For example, sorede, soide, sonde, do,
and nde are grouped together and translated as 'and' (1955, cited in Onodera
199329). Other researchers, for example, Maynard (1989), Onodera (1993), and
Saito (1992) listed sorede as a counterpart of the conjunction 'and,‘ however,
as mentioned earlier, the functions of (sore)de has not yet been studied. For
these reasons, the present study focuses on sorede and do, and not treat other
conjunctions, such as sosite and sorekara even though they are also
translated as 'and. '

De is a conjunction which is derived from other conjunctions such as
sorede and sokode by omitting sore and soko, according to Shogakkan's
Nihon Kokugo Daijiten (1979).

Sorede indicates that what is stated in the preceding sentence is the
reason for what follows in the next sentence, as in (17):

(17) a. Ziko ga atta.

accident S existed
'There was an accident.‘

b. Sorede tikokusita.
and was-late
'And I was late.’ (Shinsen Kokugo Jiten 1981 :620)

Another function of sorede is to elicit another participant's continuation in
conversation, usually accompanying questions, as in (18):

(18) A: Kippu ga nakunatta n desu.
ticket T disappeared NOM BE
‘My ticket disappeared. (= i couldn't find my ticket)‘

14

B: Sorede doo simasita ka?
and how did Q
'So, what did you do about it?‘ (Endo 1987:60-61)

Speech Act Theorists, such as Searle (1975) argues that questions function as
requests. Sorede in (188) marks B's request for further information which
works in the action structure in Schiffrin's discourse model. in addition, the
question followed by sorede seems to be sending a 'meta-message' (T annen
1989) of speaker's interest and/or involvement toward the another interlocutor‘s
talk; thus, sorede also works at the participation framework.

The definition of sokode is almost identical to the one of sorede; sokode
also indicates that what is stated in the preceding sentence is the cause for the
what follows in the next, as in (19):

(19) a. Hidoku tukareta.

terribly tired
'I was terribly tired.‘

b. Sokode hayaku neta.
and early slept
'And i went to bed early.‘ (Shinsen Kokugo Jiten : 612)

in addition, sokode is used when changing the topic which works in the
ideational as well as the exchange structure, as in (20):

(20) Sokode kore kara doo suru ka?
so this form howdo Q
'So what should we do from now?‘ (ibid.)

Since do is an abbreviated version of both sorede and sokode, it can
also indicate the reason/cause. The following excerpt is from the collected data.
Here, R is explaining why she could not go to a job interview:

(21)R: a. Watasi ga nihon e kaetta ato datta kara sore ga
l S Japan to returned after BE so that G
maniawanakatta no yo ne?
make-NEG NOM PP PP
'(T he interview was held) after I went back to Japan so I could
not make it.

15
D. De peke.

no good
'So it was not good.’
Another use of de is identical to the second function of sorede in conversation.
It is used to elicit another participant‘s continuation, usually accompanying
questions:

(22) A: Ee, moo itta toori tootoo kimasendesita.
yes already said like after all come-did-NEG
'Yes, he finally stopped coming after all as I said already.‘

B: De?
and
'And (what happened)?‘
(Sabishikereba, cited in Nihon Kokugo Daijiten)
Although no question follows after De? in (2B), the prosodic contour -- raising
intonation - signal that it is a question, or more specifically, a request for further
information.

The definitions given in Nihon Kokugo Daijiten suggest that the
conjunctions sorede, sokode, and de seem to play roles as discourse markers
- each of the first meanings seem to operate on the ideational structure, while
the second ones function on the action and exchange structure as well as
participation framework. These will be examined in the collected data. Figure
2.3 summarizes the definitions of each conjunction illustrated in Nihon Kokugo

Daijlten.

16

 

 

 

 

”(injunction Definition T Schiffrin's Discourse Planes
81. Sorede S? 1 St: reason for $2 ldeaticnal structure
2 To elicit another Action structure
participant's continuation Participation framework
31.30k0de $2. 1 31: cause for $2 ldeaticnal structure
2 To change the topic ldeaticnal structure
4 _ Exchagqe structure
31. De $2. 1 S12 reason/cause for $2 ldeaticnal structure
2 To elicit another Action structure
participant's continuation Participation Framework

 

 

 

Figure 2.3 Summary of the definitions of sorede, sokode, and de (Source:
Shogakkan's Nihon Kokugo Daijiten)

2.3 Properties of Other Conjunctions in Japanese

In this last section, previous studies on the functions of other Japanese
conjunctions (demo, dakedo 'but,‘ datte ‘but, because,‘ and dakara 'so,
therefore') are reviewed. This is necessary to identify the unique and

overlapping functions of (sore )de more accurately.

2.3.1 Demo, dakedo 'but'

Onodera (1993:122-123) analyzes the so-called adversative
conjunctions demo and dakedo ‘but' as markers of contrastive action, which
she defines as 'contrast between actions, which highlights some contrasting
aspect of the speaker-hearer interactional dynamics in the on-going discourse.‘
Onodera describes four types of contrastive actions which demo, dakedo mark:
(1) point-making such as self-repairs, digressions and interruptions which works
in the ; (2) claiming the floor between interlocutors which works in the exchange
and the action structure; (3) opening conversation which works in the exchange
structure, and (4) changing topic which works in the ideational structure. Thus,
demo, dakedo seem to play roles on the exchange and on the action structure

as well as the ideational structure.

1 7
2.3.2 Datte ‘but, because'

The Japanese conjunction datte 'but, because' has been described as a
logical connector, whose functions are (1) to provide a reason/cause for the
preceding statement, i.e. backward reasoning, and (2) to express feelings of
opposition toward the other interlocutor's statement, or to offer excuses in
conversation (Yokobayashi and Shimomura 1988, cited in Maynard, 1992az63).
Maynard (1992a, 1992b, 1993) claims that datte is a marker of (1) self-
justification: to self-justify one's position by declaring that the upcoming
discourse unit supports one's position (1992a: 84), and sometimes with
additional message -- to warn the listener that the upcoming turn contains self-
justifying information in the context of opposition -- what Pomerantz's (1984;
cited in Maynard, 1992a: 80) calls 'dispreference marker‘ or what Bilmes (1988;
cited in Maynard, Ibid.) calls ‘reluctance markers'; and (2) the claim of turn and
yielding of turn since datte often appears at interactionaliy significant points
such as the utterance and/or tum-initial position, at the turn-transitional period,
and at hesitation points (p. 76). Therefore, datte functions on the participant
framework, the action and exchange structures, in addition to the ideational

structure.

2.3.3 Dakara ‘so, therefore'

According to Maynard (1992a, 1992b), although dakara 'so, therefore'
has been traditionally categorized as a 'cause-and-result' connective (i.e.
forward reasoning), it carries additional discourse functions. She points out that
the functions of dakara are: (1) to signal supplementary explanation starts, i.e.
the following utterance provides explanation related to the preceding utterance,
and sometimes it expresses additional meaning, such as reluctance or irritation,

especially when the speaker is requested to repeat again even though the

18

speaker has already provided information some time in the previous discourse,

and (2) to signal the claim of turn and/or yielding of turn, which is related to

participatory control (p. 97). Thus, dakara operates on various planes of

discourse; the participation framework, the exchange and action structures, and
the ideational structure. Figure 2.4 summarizes the functions of demo, dakedo,

datte, and dakara as discourse markers.

 

 

 

 

2. To claim turn &Ior
to yield turn

 

Conjunctions FLunctions Schiffrin's Discourse Planes
demo, Contrastive action
dakedo (a) point-making Participation Framework
(b) to claim the floor Exchange/Action structure
adversative (c) to open ldeaticnal/ActionIExchange
conjunction conversation structure
(d) to change topic ldeaticnal/ActionlExchange
structure
datte 173elf justification Participation Framework
reason/cause Action structure
connective
(backward) 2. To claim turn and/or Action/Exchange structure
_t:o yield turn _
akara 1. To signal Participation Framework
cause-and-result supplementary
connective explanation starts
(forward)

Action/Exchange structure

 

Figure 2.4 Summary of the previous studies on functions of Japanese

conjunctions

This chapter summarized the properties of and in English, its Japanese

counterpart (sore)de, and other Japanese conjunctions in their roles as

discourse markers. The findings of the previous studies suggest that (sore)de

can play discourse functions in addition to its semantic properties although

such functions have not been yet examined.

 

CHAPTER ill
Method and procedures

3.1 Data Collection
Du Bois et al. (199229) emphasize the importance of naturally occurring

conversation as data for the analysis of discourse as follows:

the first and most important consideration is to record a speech
event which constitutes a naturally occurring interaction. One
should start out with a natural context for interaction, in which the
talk takes place for its own sake rather than for the benefit of the
investigator.

This concern with analyzing naturally-obtained conversation is considered very
seriously in the present study since, as Du Bois found, ‘it represents the most
challenging, and perhaps the most rewarding, kind of data' (pp. 9-10).

Six dyadic conversation between relatively close friends (four same-sex
pairs and two cross-sex pairs) of native speakers of Japanese were audio-
taped and transcribed for examination. Most of the participants (eight out often)
are graduate students at a large university in the Mid-West of the United States,
but two participants are school teachers in Japan. Prospective participants
were asked permission to audio tape an hour-long informal conversation, which
would have occurred anyway, with their relatively close friends (including the
researcher). No specific topics nor tasks were assigned. The conversations
were took place in various places such as in a dorm room, a cafeteria, a coffee
shop, and at home over the telephone. Twenty minute sections of each
conversation were transcribed, using the transcription conventions of Schiffrin
(1987) and Maynard (1987) as models. (See Appendix 1)

19

20
3.2 Participants

Six men and four women participated in this study. They are all native
speakers of Japanese and eight of them are currently students at Michigan
State University. Their ages range from mid-twenties to late-thirties. One of the
conversations was collected in December, 1994 while the researcher was in
Japan, and two were tape-recorded over the telephone. Figure 3.1 summarizes
the participants and the topics covered in their conversations.

 

 

 

 

Participants Setting (a’fime) Topic
M-M 1 M (G) in A's room in the dorm Plans for summer
A (G) (May, 1995) (travels, study etc.)
MM 2 T (G) At a cafeteria in the dorm T's part-time job
K (G) (June, 1995) Plans for summer
(travels)
F-F1 R (G) At a coffee shop in Work situation job searching
O (T) Sapporo, Japan Travels
(December, 1994)
F-F2 R (G) Telephone conversation R's request for borrowing S'
S (G) (February, 1995) book
TAship
M-F1 Y (T) Telephone conversation Weekend
R (G) (May, 1995) Previous coworkers
MP 2 B (G) At a coffee shop Good food and restaurant
N (G) (June, 1995) N's hometown

 

 

Figure 3.1 Participants and topics in their conversations
(M = male, F = female, Bold = male, G = graduate student. T = school teacher)
3.3 Units of talk

The notion of 'units of talk' is crucial to the analysis of discourse markers.
Schiffrin's (1987) operational definition of discourse markers is ‘sequentially
dependent elements which bracket units of talk' (p. 31). However, Schiffrin
does not offer a concrete definition of what 'units of talk,‘ are. She deliberately

21
left it a vague expression which would allow researchers' flexibility: 'sometimes

those units are sentences, but sometimes they are propositions, speech acts,
tone units' (p. 35).

Some researchers define the units of talk in terms of pauses in the
stream of speech instead of propositions or speech acts. For example, Chafe
proposes (1980213) 'idea unit.‘ Maynard (1989220) summarizes three features
which Chafe's idea unit displays (Chafe 1980,1985)2

1. It is spoken with a single, coherent intonation contour,
ending in what is perceived as a clause-final intonation.

2. it is preceded and followed by some kind of hesitation,
ranging from a momentary break in timing to a filled or
unfilled pause lasting several seconds.

3. It is a clause, that is, it contains one verb phrase along with
whatever noun phrases, propositional phrases, adverbs,
and so on, are appropriated.

However, Maynard (lbid.: 23-24) points out the disadvantage of Chafe's 'idea

unit' when analyzing Japanese as follows:

In Japanese a syntactic clause is frequently broken into a number
of smaller units bounded by pauses or skipped beats, and these
units are mostly marked by intonation contour. These smaller
units, which display phonological features similar to idea units in
that they are pronounced in one continuous flow, often do not
contain verbs and may best be described as phrases rather than
clauses. Although these are defined by the phonological feature of
_ pause, they mostly coincide with morphological units of
independent lexical items plus junction words such as particles .

Hence, to analyze Japanese discourse, Maynard proposes 'Pause-bounded
Phrasal Units' (PPUs) which are bounded by pauses and, as a result, are
smaller than Chafe's idea units. PPUs seem to be identical to what Du Bois et
al. (1992: 17) call an 'intonation unit' which is defined as 'a stretch of speech
uttered under a single coherent intonation contour.' Using this smallest unit as

 

2 In more recent work, however, Chafe uses an 'intonation unit' which is identical to Maynard's
Pause-bounded Phrasal Unit (PPU). For more detail. please consult chapter 5 in Chafe (1994).

22
a units of talk has advantages. For instance, Maynard points out that listener

back-channel responses and final particles, such as no, which functions as a
tag question, or yo, which conveys emphatic sense, are often observed at the
end of many PPUs. This means that PPUs are not arbitrary units divided by
pauses, but important units in terms of the progress of conversation. The
occurrence of listener back-channel reveals that interlocutors recognize the
existence of the boundary between two units and that it is an acceptable place
to signal listener's response and more generally, TR P. The role of the sentence
final particles is to 'monitor the progress of the conversation by confirming a
statement, by calling for hearer‘s attention, etc' (Shibatani 1990: 386). in other
words, the use of sentence final particles shows us that the speaker is breaking
down his or her talk into smaller units and sending an interactional message at
the same time. This observation is well illustrated in the example (1 ). Here, M
is talking about an air conditioner which he finally bought.

(1) (MAI8/15—17)3
M: a. Kuuraa neel
air conditioner FP
'Air conditioner,‘
(A12 Ee.)
yes
'Uh huh.'
b. iroiro ano te kono te del
various that way this way by
'(I used) various ways,‘

(A22 Ee.)
yes
'Uh huh.'
c. isya no sindansyo toka tukete-morattel
doctor LK note like have-attach-and
'l asked a doctor to attach a note for me.‘
(A3: Ee.)
yes
'Uh huh.'

 

3 (MA/8I1 5-17) means that the conversation is between participant M and A, and the excerpt is
from page 8, lines 15-17 in the original transcription. More contextualized data are given in
Appendix 2.

23

d. Sorede kyoka o totta n desu yo./
and permission 0 got NOM BE FP
'And i got a permission (from the office), you know.‘

e. Totta wa ii n da kedo/
took T good NOM BE but
'It went well up to that point (= getting a permission) , but...‘

f. Sorede!
and
'And then.‘
((M‘s story continues»

Back-channeling by A is observed after (a) and (b), even though both phrases
are smaller units than Chafe's 'idea units,‘ as Maynard pointed out. And more
importantly, the fact that Japanese speakers use back-channeling without any
overlap suggests that participants recognize this pause as significant. In
addition, (d) and (f) start with the conjunction sorede, which is a candidate for
discourse marker status in Japanese. If PPUs are not adopted, there is no way
to identify these two items.

As described above, there are three reasons for adopting PPUs as units
of talk. First, PPUs are the smallest units in which interactional functions are
observed (e.g. the occurrence of back-channeling, and the use of sentence final
particles). Second, PPUs properly allow us to identify prospective conjunctions.
Third, they are relatively easy to identify. Therefore, PPUs are used as a unit of
talk in this study.

CHAPTER IV

Results

In this chapter, the results of the data examination are illustrated in both
quantitative (4.1) and qualitative (4.2) terms.
4.1 Quantitative Results

The coordinate conjunction and (Schiffrin 1987) and (sore)de, its
counterpart in Japanese (the National Language Research institute 1955, cited
in Onodera 199328-9) were the most frequently used conjunctions in both
studies. Schiffrin reports that 1002 clause-sized idea units in her data were
prefaced by and, mmpared to 440 prefaced by but, and 206 by so; NLRI
reports that in ten hour-long tape-recorded conversations, (sore)de 'and' and
' its variants such as soide, sonde, and nde were used 417 times out of 1558 (or
27%), followed by demo, dakedo 'but' (196, or 12.6%), and dakara 'so,
therefore' (188, or 12.1%). The same tendency is observed in the present
study. A total of 413 uses of all conjunctions were identified within 120 minutes
(20 minutes for each dyad) and (sore)de was the most frequently used
conjunction among the six. Of 413 tokens, 179 (43.4%) were (sore)de 'and,‘
followed by dakara 'so, therefore' 118 (28.6%), demo, dakedo 'but' 96 (23.2%),
and the least frequeme used conjunction was datte 'but, because' 20 (4.8%).
The use of sokode 'and' was not observed at all. The frequent occurrence of
(sore)de 'and‘ suggests that it plays important roles in discourse. The close
examination of the use of (sore)de will be illustrated in (4.2). Table 4.1
illustrates the frequency of the conjunctions with respect to each participant.

24

25
Table 4.1 Frequency of conjunctions in causal dyads in Japanese

 

 

 

 

 

de (sorede) dakara demo datte Tatal
QM
MM 1
M 28(14) 1 6 1 4 1 59
A 4(0) 9 4 o 17
MM 2
T 18 (2) 1 9 8 4 49
K 5(0) 0 3 o a
F—F 1
O 1 (O) 4 5 O 10
n 14(0) 8 1o 3 35
F-F 2
S 8 (1) 1O 1 1 O 29
R 17(2) 15 e 2 42
M-F 1
B 24(1 1) 6 8 1 39
N 25(0) 14 12 1 52
M-F 2
Y 24 (6) 1 1 9 1 45
n 11 (1) 6 4 7 28
TOTAL 179(37) 1 1 8 ' 96 20 413
(43.4%) (28.6%) (23.2%) (4.8%) 500%)

 

Based on six 20-minute long conversations
M = male, F = female, Bold = male
28 (14) represents that 14 are sorede out of 28 tokens of de and sorede

in terms of the frequency of conjunctions, there are three facts which are
related to the gender of the participants and that of their partners . First, both
men and women used conjunctions more frequently with an opposite-sex
partner than with a same-sex partner. The mean score of conjunctions for men
was 33.25 in twenty-minute long conversation when talking to another man, but
42 with a female partner; women used 29 conjunctions with another woman, but
40 when talking to a man. Second, both men and women used (sore)de more
frequently with an opposite-sex partner than with a same-sex partner.
Specifically, the percentage of use of (sore)de among all the uses of
conjunctions increased when talking to an opposite-sex partner. When talking
to another male-partner, 55 out of 133, or 41.4% of the use of the conjunctions

26

are of (sore)de among men. The percentage increased to 57.1% when talking

to a woman. The same tendency is observed among female participants: 40

out of 116, or 34.5% of women's use of conjunctions are (sore)de with a

same—sex partner, while the number increased to 36 out of 80, or 45% when
talking to a man. Third, unexpected but very interesting result was observed:
the choice between do and sorede were asymmetrical: between these two
alternatives, men used sorede more often with a female partner than with
another man (29.1% and 35.4%, respectively), while the use of sorede among

women became less frequent when talking with a man than with another female

(7.5% and 2.8%, respectively). This leads to an assumption that there are some
difference between these two forms although Nihon Kokugo Daijiten (1979)

defines do as derived form of sorede by omitting sore.

Table 4.2 Frequency of Conjunctions (in same-sex and mixed pair)

 

 

 

 

 

 

 

 

 

 

 

de (sorede) dakara demo datte Total
dakedo

MM 55( 16) 44 29 5 133
41.4% 33.1% 21.8% 3.7%

F—F 40(3) 37 34 5 116
34.5% 31 .9% 29.3% 4.3%

MP 48(17) 17 17 2 84
57.1% 20.2% 20.2% 2.5%

PM 36(1) 20 16 8 80
45.0% 25.0% 20.0% 10.0%

TOTAL 179 1 1 8 96 20 413

Table 4.3 Sorede vs. de
de (sorede) Percentage of Percentage of de
sprede

M-M 55(16) 29.1 % 70.9%

F-F 40 (3) 7.5% 92.5%

MP 48(17) 35.4% 64.6%

FM 36 (1) 2.8% 97.2%

TOTAL 179(37) 20.7% 20.7%

 

 

27
4.2 Qualitative results

In this section, a close examination of the use of (sore)de is presented,
based on its dictionary definition and semantic properties (4.2.1), its use shared
with other markers (4.2.2), and its use as a discourse marker (4.2.3). Between
the two items de and sorede, do was used more frequently than sorede (142
and 37 tokens, respectively). It is observed that the difference reflects the
functional differences between these two items, i.e. de occurred in more
various environments than sorede ; moreover, do displayed more functional

flexibility than sorede.

4.2.1 (Sore)de as dictionary definition: Expressing Reason and cause
As Nihon Kokugo Daijiten ( 1979) defines it, the fundamental function of
(sore)de is to conjoin reason and result just as dakara 'so, therefore' does. In

the following example, T is explaining why he decided to come to the States.

(1) (T KI1 135-37)
T: a. Syakaigaku mo... I
sociology also
'(Conversation Analysis is) also (developed in) sociology,‘

b. D el watasi wal sono—I sooitta atarasiil
so i T that such new

0. taipu no syakaigaku 0/
type LK sociology 0

d. maneejimento no sosiki riron no hoo ni
management LK organization theory LK to

e. ooyoo sitakuttel
applied want-to-and

f. ndel Amerika ni kita n desu yo.
and America to came NOM BE PF
'80 I wanted to study such a kind of new sociology and apply it
to organizational management theory. That's why I came to
the States.‘

28
The conjunction do and its variant nde are used in initial position in (b) and (3),

respectively, and both of them conjoin the reasons and the results: (a) is the
motivation (reason) for T to study a new type of sociology and apply it to
organizational management theory, as stated in (b-e); moreover, due to all the
reasons described in (a-e), he came to the States, as stated in (f), which is
preceded by nde.

Sorede is also used to indicate reason. In the following excerpt, B is
talking about why Disneyland became an undesirable place for Disney:

(2) (NB/5I1-3)
B: a. Sorede saal mawari ni saal
and PP neighborhood at FP

b. Dizunii ga kangaetenakatta yoo na/
Disyney S was-thinking-NEG like BE

c. kankyoo ga dandan dekitekite saal
environment S gradually have-built—and FP
'And the neighborhood has changed in a way Disney has
never thought about,‘

d. sorede! zibun no risoo to hodotooi tokoro ni natta to.
so self LK ideal far from place became QT
'80 (the Disneyland) became far from his ideal place.‘

Since the environment around Disneyland has changed a lot after Disneyland
was built, the place became far from Disney's ideal place. The explanation
given in (a-c) provides reason for the following sentence; thus, sorede is used

to signal that the following is the reason.

4.2.2 in environments shared with other markers
4.2.2a Do as demo, dakedo 'but'

De also occurs in environments in which the adversative conjunctions-
demo, dakedo might be used although such a usage was not often. Sorede
was not observed in such environments in the collected data. In the following

29
excerpt, M and A are trying to find an inexpensive place to stay in Detroit. A

showsaieafletofayouth hostelto M:
(3) (MAB/2730)
M: a. E?! Aal Kore neel kore-I kore nee!

what Oh this FP this this FP
What?! Oh, this one, this one,‘

D. Nara-sen kara kiita desyo?l
Mr. Nara from heard TAG
'Did you hear it from Mr. Nara, didn‘t you?‘
A: c. Ee ee as ea!
Yes
'Yeah yeah yeah yeah.‘

M: d. D e horal hurui desyoo./
look old TAG
'(But) see, it's old, isn't it?‘

A: e. Aa hail
oh yes
'Oh, yes.‘

Even though A thought information which he got from one of their mutual friends
was very useful, M already knew about the place as stated in (a-b). Moreover,
M knew the place is not good since it is old, as stated in (d). Therefore, de
seems to be working as the adversative conjunctions such as demo, dakedo
'but' in this context since it's old, isn‘t it? is a negative comment and thus
contradicts A's suggestion.

4.2.2 b (Sore)de to connect two events: 'and then'

According to Labov (1972:359-360), a narrative is defined as 'one
method of recapitulating past experience by matching a verbal sequence of
clauses to the sequence of events which (it is inferred) actually occurred.‘
In English, narrative clauses are usually connected by and. Schiffrin

(1987:131) illustrates that and functions at a local level to link clauses into

30
sections of a story. In Japanese discourse, de seems to assume that role. In

(4), Y is telling a story about his weekend.

(4) (YR/1I4-7)
R: a. Huun./ De higaeri?!
uh huh day trip
'Uh huh. Was it a day trip?‘

Y: b. Uun.l lppaku site./
No one night did-and
‘No, we stayed a night.‘

c. Nde Furiipocto e ittel
Freeport to went-and
'And (on the following day). we went to Freeport!

d. de Aakeezia kokuritukooen tte aru n da yo nej
Arcadia national park QT exist NOM BE FP FP

e. Soko ittel
there went-and
'And we went to a place called Arcadia National Park,‘

f. un/ d e kaette kitel
yes return-and-came
'and we came back,‘

9. kinoo no yonaka no itii zi goro kaette kita.!
yesterday LK midnight LK 1 o'clock around return- and came
'we came back around one o'clock last night.‘

This example illustrates a very simple story about where Y and his family went
to on a weekend. First they stayed a night somewhere, as in (b), sometime on
the following day, they went to Freeport, as in (c), they went to a place called
Arcadia National Park, as in (d-e), and they came back home, as in (f). Thus, do
and its variant nde at the initial position of (c), (d), and (f) function to link
narrative clauses, which are the most local level of idea units, in the sequence
of order within a story as English and does.

31
4.2.3 (Sore)de as a discourse marker

As illustrated in the previous sections, (sore) de connects two adjacent
units of talk (the most local level of talk), identifying a reason/cause and result
relation, and time sequence. However, the examination of the collected data
reveals that (sore)de is sometimes used to express more than the semantic
properties outlined above; it plays important roles in the action and the
exchange structures and sometimes in the participation framework as well as in
the ideational structure. Through the examination of the data, three roles which
(sore)de plays in discourse are identified: it coordinates units (4.2.3a), it
signals the speaker's continuation (4.2.3b) as its English counterpart and does,
and it elicits the other interlocutor's continuation and signals the speaker's
request for further information when prefaced to questions (4.2.30). In many

cases, (sore)de displays these roles simultaneously (4.2.3d).

4.2.3a Coordinating units

Schiffrin (19872141) observes that and marks structural coordination of
idea units: and links structurally equivalent units at the local level as well as
the global level. The Japanese counterpart, (sore)de also plays the same role.
However, the data suggests that there are some differences between do and
sorede. De connects rather local level of ideas such as narrative clauses as
examined in (4.2.2b), while sorede seems to bracket more global level of ideas
when these two are used together. In (5), M is telling a story about having some
trouble when he was installing the air conditioner that he had just bought:

(5) (continuation of (1) in (3.3))
(MAI8/17-21):
a. Sorede kyoka o totta n desu yo.l
permission Otook NOM BE FP
'So I finally got a permission from the housing office.‘

32

b. Totta wa ii n da kedol
took T good NOM BE but
'It went well up that point (=getting a permission), but...‘

c. So rede/ konoaida anool tuke ni kite neel
the other day uh came-to—install-and FP
The other day, a man came to install it‘

(A42 Ee.)
Yes
'Uh huh.'
d. Del sositaral anool
and what happened was uh
'And, what happened was, uh-,'
e. sore yuzuttemoratta kuuraa nan da kedol
that bought air conditioner NOM BE but
'Weli, that's the one I bought from somebody else,‘
(A5: Ee.)
Yes
‘Uh huh.'

f. d el kaette] kore zya tukanai tte.l
left this install—NEG QT
'and he said "I can‘t install itl", and he left.‘

M tells that a man came to install the air conditioner the other day, but he could
not do it. Sorede in (c) marks where the story starts, while de in (f) signals that
the following narrative clause is connected to some prior local level unit,
namely, (c), The other day, a man came to install it. This type of use of de
seems to be different from the one of (4), in which de links two adjacent
narrative clauses. However, there are not so many differences between these
two uses of do for the following reason: although the utterances in (cl-e) look
odd, they need to be said before the narrative furthers because it is crucial
background information for the upcoming talk of why the man could not install it.
Thus, de connects the most local level of units such as narrative clauses which.

are adjacent to one another as in (4), as well as ones that are not adjacent as in
(5).

33
The local level of units which de coordinates are not restricted to

narrative clauses. In the following excerpt, Y starts explaining what other
co—workers have been doing since R left:

(6) (YRl2l44-45)
a. Sonde neel o- omosiroi yol mada./
FP interesting FP besides
'There are some more interesting things,‘

b. D e kono haru neel
this spring FP
'In this spring,’

c. Yosida-sensee ga Meizi ni modottekita no.l
Yoshida-teacher S Meiji to returned FP
'Mr. Yoshida came back to Meiji Junior High.‘

((Y talks about Mr. Yoshida.»

Prior to (6), R and Y were talking about the teachers they know, and the
conversation leads Y to tell more about other teachers. Sorede in (a) signals
that the upcoming is somehow a higher level unit than the adjacent unit; thus, it
marks a topic shift. De in (b), on the other hand, marks the following as the
lowest level of the units within the domain which sorede brackets, i.e. gossip on
Mr. Yoshida. After talking about Mr. Yoshida, Y now starts talking about another
teacher as in (7):

(7) (vars/942)

Y: a. Nde neel mada aru yo./

and PP besides exist FP
'And I know some more.‘

b. Hada-sensee ga neel Keesee no kyootoo na no.l
Hada-teacher S FP LK assistant principal BE NOM
'Mr. Hada is an assistant principal of Keesee Junior High.‘
R: c. Nani?!
what
'What?l'

Nde precedes gossip about another teacher, Mr. Hada; hence (7a-b) which are
preceded by nde parallel to (6b-c) which are preceded by de. Thus, do is

34
coordinating the local level within the domain which sorede bracketed as

higher level.

In the previous two examples, sorede is used to mark the global level of
units: it occurs when the story starts as in (5), as well as when the topic shifts
as in (6). Other evidence that sorede coordinates a somewhat global level of
units, rather than linking purely local units is observed when a speaker wants to
return to the original topic from a digression:

:33) (NB/312427)

a. Demol konkai neel Amerika ni kite neel
but this time FP America to come-and FP

b. aisukuriimu wa neel anmari kangekisinakattaJ
ice cream T FP much impressed-NEG
'But this time, I wasn't impressed much with ice cream.‘

N: c. Mae wa kangekisita n desu ka?
last time T impressed NOM BE Q
'So, were you impressed last time?‘

B: d. 800.! Mukasi neel
Yes old days FP

f. hazimete Amerika ni kita toki saal

for the first time America to came when FP

'Yesl When I came to the States for the first time,‘
((B's story continues.»

After (I), B explains why he thinks that ice cream in the old days tasted much
better than the one in present day because he thinks daily products used to be
much richer than the one of present. However, this leads N to talk about milk in
Japan. Then 8 finally says as in (9):

(9) (NB/3&4I43-2)
B: a. Sorede saal ano- saisyo ne?l
so FP uh- at the beginning FP

b. Saisyo Honoruru ni tuita wake.l
first Honolulu at arrived case
'So anyway, 'we arrived at Honolulu at the beginning.‘

35

c. D el minna ni sal
and every one to FF
((B's story continues.»

Sorede in (a) marks the following as not related to the local level of unit, i.e. N's
talk about milk in Japan, but related to B's original story about the first ice cream
he ate in the States. De in (c), in contrast, is functioning within the frame of the
story, i.e. at a local level.

Although the data illustrated above contains both de and sorede,
co-cccurrence of these two are not obligatory. In (10), N's story begins without

any observable markers of transition:

(10) (NB/3136)
N: a. Atasil ano-l kyanpasu nol konaidal
l uh- campus LK on the other day

b. anoo aisukuriimu no kappu no katta n desu yo.
uh- ice cream LK cup NOM bought NOM BE FP
'Uh-, on the other day, I bought ice cream the one in a cup on
campus.‘

0. lkura ka zenzen siranakutel
how much Q not at all knew-NEG-and
'I did not know how much it is,‘

d. dee- deel ano— hazimete kattemitaral 1.5?
and uh- for the first time bought-and-try-when
'And-, I bought it for the first time but it was a dollar fifty.‘

e. Kekkoo yasui na toka omotte.
quite cheap BE like thought-and

'I thought it's quite cheap.‘

f. Tiisai yatu ne?
small thing FP
'I mean the small one.‘

9. D el dekkail 2 bun no 1 garon?
and big part LK gallon
'And the big one, the half gallon one?‘

h. 2 bun no 1 garon de 2 doru 75?
part LK gallon per dollar
‘the half gallon one is two dollars and 75 cents.‘

36

i. Dakaral maamaa da to omotte.
so decent BE QT think-and
'So, I thought its reasonable.‘

N starts her story on buying ice cream in (a) without any transition markers.
Dee- dee in (d) signals that the following is linked to some prior units; in this
context, the following is linked to (a-b). The background information (o)
intervenes between these two units. Other units (M) and (g-i) are coordinated
by do, the price of the small cup and the one of half gallon. N continues her
story as follows:

(11) (NB/3l9-10)

N: a. D a! demo neel kono aida ittara neel

and but FP on the other day went-when FP
'But when I went there a few days ago,‘

b. bataapiikan ga nakuttel
butter pecan S exited-NEG-and
'they didn't have butter pecanl'

c. kanasikutel okasii naal nande nai n daroo tte/
sad-and strange FP why exist-NEG NOM BE QT
'lt made me sad and I wondered why there's no butter pecan.‘

De in (a) coordinates the following not with a part of (10), but the whole story of
(10), the first visit to the dairy store.

As examined in this section, (sore)de plays an important role as a
discourse coordinator in Japanese, in a similar way that its English counterpart
and does. By using (sore)de, the speaker points out that an upcoming unit
needs to be coordinated with some prior unit in the previous discourse. As and
works at both global and local levels in English discourse, (sore)de also
functions at both levels. However, there are some differences between do and
sorede In terms of the domain in which they operate: sorede prefers to work at
a global level, i.e. when starting a story, shifting the topic, while de tends to
function at a local level, i.e. when connecting narrative clauses and other units
which are usually within the domain bracketed by sorede. Thus, the distinction

37
between sorede and de provides us with key information to understand the

organization of discourse. However, the use of contextual information is still

needed to identify what is coordinated

4.2.3b Speaker‘s continuation

(Sore)de plays an important interactional role as well: (sore)de signals
the speaker's continuation, which operates on the action and the exchange
structure, as its English counterpart and does. This type of function is most
clearly illustrated in environments such as (A) co-occurrence with other

markers, and (B) in combination with fillers.

(A) Multiple usage of conjunctions
De, but not sorede, sometimes occurs in environments in which it
immediately precedes other markers. In (12), B is explaining why Disney World

was built in Orlando after Disneyland in Los Angeles.

(12) (NBI5I3-13)
B: a. Sorede! zibun no risoo to hodotooi tokoro ni natta to!
so self LK ideal far from place became QT

'80 Disneyland became far from his ideal place.‘

b. Nanka koo horal huriiwei o orite saal
like this look freeway 0 get out FP

c. nanka koo—l isi nol ne?]
like this stone LK FP

d. Nanka hokorippoi miti o zuuto hasitteiku tol
like dusty road 0 throughout drive-and-go if

e. Dizuniirando ga aru tte yuu no wal
Disneyland S exist QT say NOM T

f. doomo risoo ni husawasikunai kara saa.
somehow ideal appropriate-NEG so FP
'Like, you know, it's not appropriate for an ideal place if

38

Disneyland is in a place where peOple need to drive through
dusty roads with stones after the freeway exit.‘

9. Sonzyaa neel
if so FP

h. kondo atarasiku Dizuniiwaarundo o tukutte saal
this time new Disney World 0 make-and FP

i. sono mawari ni-l sono mawari o zenbul ano-I
that around that around 0 all well

j. zibun no omoidoori ni sityau.l
self LK to one's satisfaction do
'If that is the case, construct Disney World this time and as
for the neighborhood, do just the way he wants to do.‘

(N1: Huun.)
uh huh
'Uh huh.'

B: k. D e soredel kekkyokul Dizuniiwaarudo o tukuru toki nil
so , after all Disney World O make time at
‘80, after all, when they built Disney World,‘

I. Dizuniiwaarudo no toti no mawari nil
LK property LK around at

m. sono Dizuniiwaarudo no sikiti no 10 bai no toti o kattel
that Disney World LK site LK times LK land 0 buy-and
'they bought the property around Disney World 10 times as big
as Disney World itself and,‘

n. de sore o zenbu sono-l nante yuu no?/
and thathli well what say PP
'and all of them uh... what should I say?‘

p. Sono sikiti nai o katteni sasenai.l
that site inside 0 freely do-CAU-NEG

'won‘t let anybody do anything inside the property.‘
In the previous talk, Bexplained that Disney and his company could not do
anything about the neighborhood although they could control the environment
inside Disneyland - many unexpected and undesirable buildings were I
constructed after Disneyland was made. Therefore, Disneyland became far
from his ideal place , as in (a). B gives background explanation as in (b—f). ln

39
order for Disney and his company to overcome the previous problem, a new

plan, Disney World, was developed, as in (g-j). Sorede in (k) indicates that all
these previous statements in (a-j) are the reason for the following (k-o): why
they bought the property around Disney World 10 times as big as Disney World
itself. On the contrary, de in (k) is void of semantic content: it just signals B's
intention to continue his turn.

In the following example, N is explaining that people in her hometown do

farming as a side business:

(13) (NB/6I21-24)

N: a. A! uti wa anoo-l kengyoo nooka ooi?l
oh home T uh side business farmer many
'Ohl There are many people who do farming as side business
in my hometown.‘

B: b. Maa koozyoo ooi n dattara neel
well factory many NOM BE-if FP
'Well, it's understandable if there are many factories (around
there).'

N: c. 800. Dee-l dakaral
yes so

d. hatake toka tanbo toka wal kisetu dake./
field or rice field or T season only
'Yes. Uh...so, vegetables and rice farming are (available)
only during the crop season.‘

In (a), N tells that neighbors in her hometown do farming as a side business,
then B comments that the fact is understandable since there are many factories
around the area. Dakara in (c) signals that a supplementary explanation starts:
farm work is done only during crop season, which connotes that people put
minimum effort into their farm since they have full-time jobs in the factories. Dee
in (c) is simply used to signal the speaker has something more to say; the way.
de is pronounced, i.e. lengthened way shows her intention to keep the floor.

40
(8) Combination with fillers

Other evidence that (sore)de shows the speaker's continuation is the
use of (sore)de in combination with fillers. Brown (1983:12-13) defines fillers
as expressions used to fill in pauses, such as well, arm, and so on. Maynard
(1989230) categorizes fillers into two groups according to the motivation for their
use: (1) language-production-based fillers which are used when smooth
speech is either cognitively or productively hindered, and (2) socially motivated
fillers which are used to fill a potential silence and avoid potential
embarrassment. She lists uuunto, are, and hora 'Uhh...that,' as part of the first
group, and nanka 'well,‘ in the second group. Thus, the use of fillers is one of
the devices to show the speaker's intention to talk rather than to stop talking. As
examined previously, (sore)de signals speaker continuation as well.

Therefore, the combination of (sore)de and fillers is an understandable strategy
for the speaker to signal his or her intention to continue.

In (14), M talks about a story when he called one of their acquaintance:

(14) (MAI2/39-40)

M: a. Sorede anool neel ree nol anoo- nankal
uh that LK uh well

b. imooto to otooto ga anool
younger sister and younger brother S uh

c. kookoo no sotugyoosiki de neel hora
high school LK graduation day BE FP uhh
'Uh-, I called that person... it was a graduation day for her
brother and sister, uhh...‘

((M continues»

Sorede at initial position in (3) marks M's intention to talk. If he did not have the
desire to talk, he would not need to fill empty slots with fillers such as anoo, and
nanka while he was trying to find words

41
De also occurs in similar environments. (5) is repeated:

(5) (MA/8l17-21):
c. Sorede! konoaida anool tuke ni kite neel

on the other day uh came to install-and FP
'A man came to install it on the other day,‘

(A4: Ee.)
Yes
'Uh huh.'
d. D el sositaral anool
and then uh
'And, then, uh—,'
((M continues his story.»

M's story starts at (c); de in (d) marks M's intention to keep the turn in order to
tell the story.

The excerpt in (15) is a continuation of (13). After (13), B asked N how

much land the average farmers owns. The question led N to digress from B's

original question to talk about some related historical fact - how the tenant

farmers got land for free from the landlords by a policy called the Agricultural
Land Reform after World War II .

(15) (NBl7I4-8)
N: a. D el dakaral
so
b. mukasi no zinusi mitaina no ga otibureta tte.l
old days LK landlord like NOM S fell low QT
‘80 the landlord in old days went down in the world, they say.‘

0. Demol ano-l nan desyoo ne?l
but well what BE FP
'But, what it would be?'

d. Dore kurai desyoo ne?!
how much about BE FP
'How much would that be?'

e. Metyakutya arimasen yo?l
a lot exist-NEG FP
'Not much,‘

42

f. Ano-l sukosi zutu.l
well little by little
'Weli, everybody owns little,‘

9. Meal dakaral
well so
'well, so,‘

h. Demol minna! sono tanbo to hatake dake de wa
but everyone that ricefield and field only by T

i. zettai seekatu sitekemasen yo?!
definitely living do-can-NEG FP
'but nobody can support their living only with their farm.‘

j. Dakaral kekkyoku kengyoo nooka bakkari ..... I
so after all side busincss farmer only
'so that's why everybody does farming as side business,‘
k. Unl d e.. un—l
uh- uh-
‘yeah, yeah.‘

Although N is knowledgeable about the digressed topic in (a-b), N does not
provide a clear answer for the original question raised by B; all N could do is
give information which might help B to get some rough idea, as in from (d—j).
Finally, N can not offer any further information in (n): de, connecting two filler
expressions nnn signals N's intention to continue to talk and hold her turn even
though she ends up failing to do so.

(16) is another example of de usage which is adjacent to two fillers. M

asked A whether it is hard to get a ticket for a baseball game in Chicago:

(16) (MAI2/27-30)

M: a. Toremasu? Ken.l
get-can ticket
'Can you get a ticket?‘

A: b. Ken-I dakaral ima are ga-l konmae itta toki niil
ticket so now that S last time went time at

c. nakatta n desu yol siai ga.l
existed-NEG NOM BE FP game S
‘Ticket. well there was no game when I went there last time.‘

43

d. D el anool henna kaado dake morattel
and well strange card only got-and

e. sukezyuuru dake morattel
schedule only got-and
'I only got a strange card, only got a schedule.‘

f. Dakaral ato denwa de kiitel
so later telephone by ask-and
'So I'll call later and ask,‘

9. n d e nnn.l
hm hmm
hmmmm)

A explains what happened last time when he tried to watch a game in (be) so
that he knows what he needs to do, in (f), 80 I'll call later and ask. Since A has
not yet answered M's question Can you get a ticket?, it seems that A tried to
answer in (9); thus, do in (9) marks the speaker's continuation. However, the
answer is understandable, i.e. A call the place, and if they have a ticket, he'll
buy it, if they don‘t have any, he can‘t get it. Since it is too obvious, it seems that
A simply did not continue to give redundant information although he was
originally planning to talk.

In contrast to the expression nnn de nnn, other markers are never
followed by the filler nnn in the collected data. In addition, there is further
supporting evidence which Illustrates de as a marker of continuation, while
other markers do not play such a function. Other markers may appear at
transition-relevant place (T RP)‘ (Sacks et al. 1974). Thus, it seems to be
acceptable to yield a turn as in (17), (18), and (19):

(17) (T Kl6l1 -3)

K: a. Aa. Demo-l
oh but
'Oh, but,‘

T: b. Del sore ni hikooki o tuke temo 100 doru de sunzyau tte
and that airplane O add though doller for enough QT

44

c. yuu koto wal kanari yasui desyo?!
thing T quite cheap BE
'Isn't it quite cheap that $100 is enough (to go to Mexico)
including the air fare?‘

(18) (T Kl6/13-1 4)
T: a. 100 doru-lpuremiamu ni nattyau n desu yo.l
dollar premium become NOM BE FP
'Air fare goes up $100 more, you know.‘

b. Dakara-l
so
'80,.

K: c. Nanka hantai ka to omotta.l
somehow opposite Q QT thought
'I thought it's the other way around.‘

(19) (YR/3l17-19)

Y: a. 800.!
right
'Right.‘

R: b. Datte-l
because
'Because-'

Y: c. Hayasi-sensee mo da yo?l
Mr. Hayashi also BE FP
'Mr. Hayashi, too.‘

Other markers are sometimes situated in simultaneous speech as in (20):

(20) (BN/6l24-26) (continuation of (15))
N: a. Da[kara-

so

'80"

B: b. [Dakara da yoo.l Kengyoo da kara.l
so BE FP side business farmer BE so
'So they are side business farmers, that‘s why.‘

Only one incident of de is found in simultaneous speech as in (21). However, it

is a case of a 'false start,‘ in which the participants start talking at the same time:

(21) (T Kl6/41 -45)

T: a. Zyaal soich mukoo ryokoo sitel
well then and over there travel do-and
'Well then, you (are going to ) travel over there,‘

45

b. soredel
and
I and,'

c. Nihon ni mo kaerazu kotti nil
Japan to NEG go back-NEG here at
'(you stay) here without going back to Japan,‘

(K1: Un.)
yes
‘Yeah'
T: d. sono mama-I
leave-as-it-is
K: 3. Un. [De— so-
yes and
'Yes, and-'

T: f. [Santafe ittel Gurando kyanion-
Santa Fe go-and Grand Canyon
'(You are going to) Santa Fe and go to Grand Canyon-'

K: 9. Un. Gurando kyanion ittel
yes Grand Canyon go—and
'Yes, (I'm going to) Grand Canyon and,‘

11. del Ransingu modottekitel

and Lansing return-and

'and then come back to Lansing,‘
((K continues»

K uses a back-channel expression between (c) and (d), so T seems to have
misunderstood K's attempt at his turn in (e) as another instance of back-
channeling even though K is trying to continue. Thus, (3) and (f) are a case of a
'false start,‘ not a tum-yielding attempt.

Other markers can be used at the end of units, whereas in the collected

data, (sore)de is never used in such a position.

(22) (MAl9l14-15)

A: a. Kotti no hoo wa suusuu desyo.l
this LK T cool TAG
'This side is cool, isn‘t it?‘

46

M: b. Dakara...ne.l
so TAG
'So.‘

(23) (MA/10l1- -2)
M: a. Eel soredeel neel koko ni iretel
yes and here at put
'Yes, and I put (a computer desk) here,‘

b. soideel ato nimotu o irete-l panpan desita yo]
and other stuff O put stuffed BE PP
'and I put other stuff, it was full.‘

c. Dakara- (5)!
so
.SO-'

Thus, (sore)de seldom appears at a TRP while other markers may. This
distributional difference between (sore)de and other markers tells their
functional difference; while (sore)de signals speaker's continuation others do

not.

4.2.3c Eliciting other's speech

In addition to functioning as a structural coordinator of ideas and a
marker of speaker continuation, (sore)de has another role. As described in
Nihon Kokugo Daijiten, (sore)de is used to signal one‘s request for further
information, especially when prefaced to questions. Although Heritage (1994)
observes and as a feature of questions, he claims that it is typically used in
'institutional' settings such as law courts and certain types of medical encounter,
but is rare in ordinary conversation between peers or acquaintances (p. 1).
However, these data reveal that the Japanese counterpart (sore)de functions
even in ordinary conversation. This use of (sore)de operates on the action and

exchange structure as well as on participant framework. Here, (4) is repeated.

47

(4) (YR/1 I4-5)

R: a. Huun. De higaeri?!
uh huh day trip .
'Uh huh. Was it a day's trip?‘

Y: b. Uun.l lppaku site.l
No one night did-and
‘No, we stayed a night.‘

Before this excerpt, Y only mentioned that he went to an outlet mall on the
weekend. De at the initial position of R's question in (a) signals R's request for
more information, at the same time, showing R's personal interest and/or
'involvement' (T annen 1989) in Y's story. in fact, Y provides not only the answer
for R‘s question (No, we stayed a night ), but he further describes where he and
his family visited in detail in (4c-g). Thus, de in (a) signals not only R‘s request
which is related to the action structure, but also R's involvement which operates
on the participation framework.

(24) is a continuation of (7). Y is explaining what all the other co-workers
have been doing after R left.

(24) (YR/3&4l42-1)
Y: a. Uunl Minato senseel sitteru tokoro de ieba neel
no teacher know-ing part at say-if FP

b. Mori sensee desyoo! Nisi sensee desyoo..l
teacher BE teacher BE

c. Naka sensee mo detyatta si/
teacher also left also
'No, the teachers you know (who are still there) are Mr.
Minato, Mrs. Mori, Mr. Nishi. Mr. Naka has transferred to another
school, too.‘

R: d. A hontoo.l
oh true
'Oh really.‘

Y: e. Kooryoo itta n daa./
went NOM BE
'He transferred to Kooryoo Junior High school.‘

48

R: f. D e kyootoo?!
assistant principal
'So is he an assistant principal?‘

Y: 9. Uunl iyaa tigau tigau.l ((laughter»
no no wrong wrong
'No, no, he is not.‘

Prior to this exchange, Y was saying that many senior co-workers now work as
assistant principals in various schools. As Mr. Naka, who is mentioned in (c) is
about the same age as the other teachers who became assistant principals, R
really wants to know whether he also got the position. Thus, the initial de in (e)
signals R's intention to elicit more information from Y.

Sorede also functions as an elicitor. In (25), Y asks R about her plan

after the graduation:
(25) (YRl5l1 1)
Y: a. Aalsorede kekkyoku nani? Natu ni sotugyoo?
oh after all what summer in graduate

'Oh, so... Are you going to graduate in summer?‘
Sorede functions as an elicitor. but at the same time, it sends a 'meta-message‘
(T annen 1989) -- Y's personal interest in and care for R.

As illustrated in (4), (24), and (25), (sore)de is a marker of a speaker's
request for further information. In addition, (sore)de seems to have another
function besides playing as an elicitor. i.e. it sometimes signals the speaker's
personal involvement and interest in the upcoming information, the other
participant's talk. The speech act of request for more information conveys
positive interpersonal effect between the speaker and the hearer. Thus,
(sore)de works on the participation framework which deals with speaker/nearer

relations, as well as speaker/utterance relations.

49
4.2.3d Coordination, continuation, and involvement

Sometimes (sore)de plays roles of coordination, speaker's continuation,
and eliciting an other’s talk, simultaneously. Here, (4) is repeated below.

(4) (YRl1l4-7)
R: a. Huun.! De higaeri?!
uh huh day trip
'Uh huh. Was it a day trip?‘

Y: b. Uun.l lppaku site!
No one night did-and
'No, we stayed a night.‘

c. Nde Huriipooto e ittel
Freeport to went-and
'And (on the following day). we went to Freeport'l

d. de Aakeezia kokuritukooen tte aru n da yo nej
Arcadia national park QT exist NOM BE FP FP

e. Soko itte.. I
there went-and
'And we went to a place called Arcadia National Park,‘

f. m! d e kaette kitel
yes return-and-came
'and we came back,‘

9. kinoo no yonaka no iti zi goro kaette kita.
yesterday LK midnight LK 1 o'clock around return- and came
'we cameback around one o'clock last night.‘

This is a simple story about what Y and his family did on a weekend. Before this
excerpt, Y only mentioned that he went to a outlet mall on the weekend. De in
(a) plays multiple functions: it marks the speaker's intention to continue the turn
as well as the speaker's request for further information. in addition, de brackets
the prior talk (Y's story) and the upcoming talk (R's question). Thus, de
functions on the action structure (speaker continuation and request), the
exchange structure (continuation), as well as participation framework (personal
interest and involvement). De and its variant nde in (c), (d), and (f), on the
other hand, coordinate more local level of units, which Schiffrin (1987: 135)

50
describes as the 'most local level of idea structure.‘ Again, these de, signal the

speaker's intention to continue his or her turn simultaneously.

As illustrated above, the Japanese conjunction (sore)de works as a
discourse marker. It marks structural coordination of ideas and speaker's
continuation, which function on the action and the exchange structures. The
data reveal that there are some difference between sorede and de in terms of
the level of units they coordinate. Sorede tends to coordinate global level of
units, while de tends to connect local level of units such as narrative clauses.
In addition, (sore)de functions to elicit the other interlocutor's continuation and
to signal the speaker's request for further information when prefaced to
questions. In the collected data, this type of use of (sore )de operates on
participation framework as well since it displays the speakers positive attitude
toward the another participant. Figure 4. 1 and 4.2 summarize the functions of

(sore)de as a discourse marker

 

 

 

Information Participation ldeaticnal Action EYChange
state framework structure structure structure
(sore)de (sore)de (sore)de (screws

 

 

Figure 4.1 Planes of talk on which (sore)de functions

 

 

 

Participation Coordinates Textual Coordinates
speaker/hearer prior/upcoming
(sore)de speaker/hearer prior/upcoming

 

 

Figure 4.2 Discourse marker as contextual coordinates

CHAPTER V

Analysis of the functions of (sore)de as a discourse marker

This section analyzes the reason why (sore)de is the most frequently
used discourse marker (5.1); and how sociolinguistic variables affect the use of
(sore)de in (5.2). The difference between sorede and de is also discussed.

5.1 High frequency of (sore)de

As described in the previous chapter, (sore)de was the most frequently
used mode of connection in the collected data. This is the same as the results
reported in the literature on English and (Schiffrin 1987) as well as on
Japanese (NLRI 1955, cited in Onodera 1993). Frequent use of (sore)de
seems to be related to the nature of the functions it plays since the functions of
(sore)de are less marked than other discourse markers, such as demo, dakedo

'but,‘ datte 'because,‘ and dakara 'so, therefore.‘

5.1.1 Coordination and continuation

Compared to other markers, (sore)de plays more fundamental roles as a
discourse marker: it coordinates ideas and marks speaker's continuation. The
data suggest that (sore)de seems to be a less marked mode of connection.
Schiffrin (1987: 128), for example, explains that the frequent use of and
correlates to its relatively less restricted distribution when contrasted to other
connectors, such as but and so. The same tendency is observed in the
Japanese counterpart: (sore)de occurred in the environments in which demo,
dakedo 'but,‘ and dakara 'so, therefore' might be used. However, the roles -
which (sore)de plays are distinct from the ones of other markers. Since the

51

52
functions of other markers are more specific (demo, dakedo 'but' mark

contrastive action, date 'but, because' expresses reason with the tone of
self-justification, and dakara 'so, therefore' expresses reason while signaling
supplementary explanation follows), it seems to me that speakers select these
discourse markers in order to maximize what he or she wants to say. For
example, a unit in which a speaker expresses a reason of self-justification is
most likely preceded by datte 'because.‘ On the other hand, (sore)de seems
not to have such restrictions since it marks more basic functions in interaction,
such as structural coordination and speaker‘s continuation. Thus, (sore)de
does not have to be as sensitive to the content and context in which it may occur
as other markers need to be. That (sore)de brackets the discourse units in
more general ways than the other markers, and that it signals speaker's
intention to continue, which is a crucial task in simultaneous speech, seem to

indicate why (sore)de is used more frequently than any other markers.

5.1.2 Sorede vs. de

Although sorede and de have been treated as variants in this study, this
section focuses on the difference between the two. When I started this study, 1
assumed that do was simply a variant of sorede. However, the results of this
study suggest that there are differences between them.

First, there exists asymmetrical distribution between sorede and do as
mentioned in (4.1, see Table 4.1 )2 do is more frequently used than sorede (142
and 37 tokens, respectively). For example, nine out of twelve tokens of elicitor
usage were de, while only two were sorede; de seems be a more preferred
marker to function when eliciting an interlocutor‘s continuation.

Second, the distributional difference reflects the functional difference

between these two markers when they signal coordination and continuation. As

53
described in (4.2.2a), the domains which sorede and de coordinate seem to be

different; for example, sorede seems to prefer coordinating global level of units,
while de seems to prefer coordinating local level of units such as narrative
clauses and other smaller units. When sorede and de mark continuation, de
seems to be the first choice. This is well illustrated when these two are
adjacent, in other words when they co-occur. (1) is repeated here from (4.2.3c):

(Bl) (NB/5l10) (repetition of (12) from (4. 2. 30»
B. l. D e soredel kekkyokul Dizuniiwaarudo o tukuru toki nil
so after all Disney World 0 constmct time at
'80, after all, when they construct Disney World,‘
((8 continues his explanation»

Here, sorede is indicating 'reasonlcause' as its main semantic property, while
do is functioning as a marker of speaker continuation. There are no sequences
such as sorede de observed in the collected data. Moreover, there are no
multiple usage starting with sorede, whereas several other examples starting
with de were found in data. This seems to indicate that do is a more
established marker to signal speaker's continuation than sorede.

These observations reveal that sorede and de serve distinct functions.
Thus, in spite of the fact that Nihon Kokugo Daijiten (1979) defines de as an
abbreviated form of sorede, de needs to be recognized independently from

sorede.

5.2 Sociolinguistic variables

In this section, the use of sorede and de are examined in terms of the
sociolinguistic variables: interactional factors (5.2.1), such as genre (5.2.1.1),
situation and relations (5.2.1.2); and individual characteristics (5.2.2), focused
on the sex of the participants. 2

54
5.2.1 lnteractional factors

5.2.1.1 Genre
5.2.1.1a Written vs. spoken discourse

The characteristics of spoken and written discourse provide an
explanation for the frequent use of de. Brown and Yule (1983: 14-19)
summarize the differences in structures between spoken and written language.
According to them, some of the characteristics of spoken language are: (1) it
tends to be much less structured, i.e. it typically contains rather little
subordination, and (2) the largely paratactically organized chunks are related
by and, but, then, and, more rarely, if. In addition, Clancy (1982, cited in
Maynard 1989) points out that in Japanese the tendency for fragmentation
seems to be expressed to an even greater degree than in a comparable
American English situation. Since spoken discourse prefers simplified
structures, de is chosen to coordinate these relatively less structured and
simplified units; at the same time, the speaker can signal his or her intention of
confinuahon.

Because of the difference in structural organization between spoken and
written discourse, some lexical items are preferred in one but not the other. De
seems to be preferred in spoken discourse, but not sorede. NLRl's research
(1955, cited in Onodera 1993: 911) reports that de ranks twentieth in written
language (in a journal) even though it ranks first in spoken language; items
used frequeme in spoken Japanese are used infrequently in written Japanese.
For example, demo and dakedo 'but' are commonly used in spoken discourse,
while sikasi 'but‘ is used in written discourse. It seem that the distributional
difference between de and sorede might be parallel to the differences between

sikasi and demo, dakedo.

55
5.2.1.1b Narrative vs. non-narrative

The characteristics of spoken language are further realized in condensed
fashion in narratives as opposed to non-nanatives, especially, the use of simple
structures. As mentioned in (4.2.2b), a narrative is defined as ‘one method of
recapitulating past experience by matching a verbal sequence of clauses to the
sequence of events which (it is inferred) actually occurred‘ (Labov 19722359-
360). According to Labov (p. 363), a fully formed narrative may display the

following:
1 . Abstract.
2. Orientation.
3. Complicating action.
4. Evaluation.
5. Result or resolution.
6. Coda.

In the collected data, however, such a fully formed nanative is rare. Instead,
most of the narratives contain only some of the elements which Lobov
proposed. Some of the narratives consist of one 'nanative clause.‘ As Lobov
suggests, the structure of narrative clauses is simpler than non-narrative spoken
segments. As discussed in (4.2.2b), narrative clauses are usually connected
with do, while global level of units, such as topic shift and return from a
digression, are marked by sorede. Thus, sorede and de are used more often
in narrative segments than in non-narrative segments. In addition to the
structural need in nanatives for sorede and de, there are other motivations
arising from an interactional need, namely, speakers must demonstrate their
intention to talk. Since speakers need to sustain longer turns in narrative
segments than in non-narratives, sorede and de seem to have a dual roles:
signaling the speaker‘s continuation as well as connecting clauses. For these

two reasons, sorede and do seem to be used more frequeme in narratives

56
than non-narrative text. The data support this tendency. Table 5.1 summarizes

the frequency of sorede and de in a 100 word segment of the data.

Table 5.1 Frequency of sorede and de in 100 words

 

 

 

 

participant narrative non-narrative
M (male) 5(2) 1
N (female) 4(0) 2

 

5.2.1.2 Situations & Relations

According to lde (1991 ), there is no neutral expression to say 'go' in
Japanese; it depends on situations and relations between the speaker and the
addressee. She states that 'we can easily list a few dozens of variants of a
word, say "go," which show some difference of polite level' (p. 78). Some of the
factors that determine the levels of speech in Japanese are (1) the nature of the
addressee (e.g. social status, age, power, and the formality of participants), (2)
the formality of the occasion, and (3) the nature of the topic (lde 1982: 366-77).
The use of sorede and de might to be influenced by the situation: either formal
or informal.

Shibatani (1991: 359) describes that a speech-writing continuum along
the formality parameter. As discussed in (5.2.1.1a), de is frequently used in
spoken discourse, but it is rarely used in written text. As opposed to de, sorede
is not frequently used in spoken data This result seems to follow Shibatani's
formality parameter.

The data were collected among relatively close friends who are mostly
graduate students, and, as a result, their conversations were casual. Although
the present study does not include other situations such as student-professor ‘
interaction, it is assumed that students talk differently when talking to a

professor since it is recognized as a more formal situation because of the nature

57
of the addressee and the situation, as lde points out. Students may try not to

use de when addressing a professor since it sounds too casual. De seems to
be inappropriate in business situations as well since it sounds very
unprofessional. Since de belongs to the lexical items for informal situations, it
was used much more frequently than sorede in the present study. In contrast,

sorede would be used more often in more formal settings.

5.2.2 Individual characteristics

Although individual characteristics such as age and region may be other
variables which might effect the use of sorede and de, these factors were not
controlled in the present study. in this section, the examination focuses on the
gender of the participants.

First, there exists a distributional difference between males and females
in the use of sorede and do. As mentioned in (4.1), there was a difference in
the occurrence of sorede between males and females (33 and 4, respectively).
However, the occurrence of de was symmetrical between males and females
(70 and 72, respectively). Table 5.1 summarizes the frequency of sorede and
de.

Table 5.2 Frequency of sorede and de

 

 

 

 

de sorede ' TOTAL
MALE 70 33 1 03
FEMALE 72 4 76

 

Second, the difference is observed not only in the frequency, but also in
the manner that sorede was used by males and females. Among females, ‘
three out of four uses of sorede were to show a reason/cause relation, which is

the traditional usage. On the other hand, males used sorede in various ways

58
(e.g. elicitor, continuation, and coordination as well as reason/cause). For

example, as examined in (4.2.3a), sorede coordinates units at the global level;
men used sorede when starting a story, returning from a digression.
Nevertheless, no such usage was observed among women. Among women,
both global and local units are coordinated by de. Thus, the status of sorede as
a discourse marker seems to be more established among men, but not among
women.

These differences seem to be very interesting since in sociolinguistic
literature, it is women who are more likely to use the prestige 'standard' forms
than men (T rudgill 1972). However, it seems that it is men who use sorede as
a discourse marker in the collected data although it is supposed to be a more
standard form than de which women preferred. This parallels to Reynolds
(1991: 136) analysis of sikasi and demo:

...the use of sikasi, the conjunction 'howeverlbut' commonly
used in male speech/public discourse, instead of
desukere—do, kere-do, kedo, demo, conjunctions with the
same semantic function as sikasi but more commonly used in
female! conversational discourse.

One of the possible reasons for this may be that sorede may connote some
formality and, as a result, it also displays distance because do is exclusively
used in spoken discourse whereas sorede can be used both spoken and
written discourse. Tannen (1990: 77) points out that man‘s style of talk is
'report-talk,‘ talk is a means to preserve independence and negotiate and
maintain status in a hierarchical social order. While women's style of talk is
'rapport-talk,‘ language is for rapport. If the use of sorede connotes formality
and therefore, distance and difference, it does not show closeness or
involvement. As a result, it does not sound 'soft.‘ Women's tendency not to use

sorede might be related to its use by men. This might be one of the reasons

59
why sorede is used as a discourse marker among male speakers but not

among female speaker.

There are two more questions which have not been answered yet: (1)
why both men and women used more discourse markers in cross-sex pairs than
same—sex pairs, and (2) why men used more sorede in cross-sex dyads, while
women used less in the same situation.

It has been argued that males and females belong to different sub-
cultures, hence they use different strategies in discourse (Maltz and Borker
1982, Tannen 1982). As mentioned earlier, men's talk is 'report-talk,‘ while
women's talk is rapport-talk' Tannen (1990) suggests men and women
misunderstand each other due to their differences in 'conversational style.‘ In
order to avoid misunderstanding, both men and women need to put extra
energy and devices to communicate effectively when talking to opposite-sex
partners. i think the use of discourse markers may be one of the devices used
to maximize communicative effect. As Schiffrin suggests (19872323), discourse
markers provide information on the relation between the prior text and
upcoming (textual coordinates) as well as speaker and hearer (participation
coordinates). By using sorede and de frequently, interlocutors are signaling
cues for their intention of continuation as well as marking textual coordination.

The choice of discourse markers, especially sorede, might be signaling
an additional message. Saito (1992) argues that an overuse of dakara 'so,
therefore' gives the impression that the speaker is very aggressive. The use of
sorede might be perceived as more formal (since it is more likely to be used in
written form), and as a result, connotes some difference and distance, which is
related to 'men‘s talk.‘ It is interesting that men displayed their 'masculinity' by

using sorede more in the cross-sex pairs, while women displayed less

60
masculinity (or more 'femininity') by avoiding the use of sorede in the same

situation.

As discussed above, the use of sorede and do as discourse markers
may vary depending on sociolinguistic variables such as genre, situations,
relations, and individual characteristics (e.g. gender of the interlocutors). This
seems to be deeply related to the Japanese culture.

CHAPTER VI

Conclusion

This section summarizes the findings, and discusses the limitations of the
present study as well as suggestions for future study.

Schiffrin (19872152) describes the functions of and as (1) a structural
coordinator of units of talk and (2) a marker of speaker continuation. The data
reveal that the Japanese counterparts sorede and de play the same roles and
additional ones: sorede and de also function to elicit the other interlocutor's
continuation and to mark the speaker's request for further information when
prefaced to questions. Although both sorede and de play these three roles,
there exist some difference between these two items. First, in terms of
coordination, sorede and de seem to have preferences for the domains in
which they coordinate. Sorede seems to prefer coordinating the global level of
units, whereas de seems to prefer the local level of units. Second, with respect
to the continuation and elicitor usage, de seems to be a more preferred item as
a discourse marker than sorede. These results suggest that de is not just an
abbreviated item of sorede.

Sorede and de share properties as markers of continuation with other
markers, such as demo, dakedo 'but' and datte 'but, because' rather than
relinquishment of turns. However, other markers signal more specifically what
follows next: demo, dakedo 'but‘ signal contrasting action, and datte 'but,
because' marks self-justification. Since sorede and de mark continuation, they
did not occur at the transition-relevance place (T RP)’ or at the end of the units.

There exist differences in terms of the gender of the participants in the
use of sorede and de. Men‘s use of sorede reveals that sorede has status as

a discourse marker, but it does not have that status among women. Men

61

62
distinguished the function of sorede and do when coordinating units: sorede

for global level of units and de works more on the local level. Women's use of
sorede, on the other hand, was rare. Women used do to coordinate both global
and local levels of units. In addition, women seems to prefer using sorede to
express the 'reasonlcause' relation.

In terms of the frequency, both men and women used sorede and do
more frequently with an opposite-sex partner. In addition, men's use of sorede
increased more when talking to a female partner, than with another male, while
women's use of sorede decreased in the same setting. Since sorede is
supposed to be used more often in written discourse than in spoken discourse,
the use of sorede connotes formality, distance, and difference. Tannen (1990:
77) argues that men‘s style of talk is 'report-talk,‘ talk is a means to preserve
independence and negotiate and maintain status in a hierarchical social order.
While women's style of talk is 'rapport-talk,‘ language is for rapport. If this is the
case, then sorede is a lexical item which reflects men's talk. This might be one
of the reasons why sorede was used more often by men, but not by women.
Furthermore, the use of sorede increased when men talked to women, .

As discussed above, the Japanese conjunctions sorede and de play
important roles as discourse markers which contribute to better understanding
of textual cohesion while serving interactional tasks. In addition, in Japanese,
sex-related differences are observed between sorede and de.

The present study examined only one genre of spoken discourse —-
causal conversation between graduate students. Future study needs to
consider other factors such as gender, age, situation, and relation and how they
interact each other, in order to understand more clearly the roles of the

discourse markers among Japanese.

APPENDICES

63

APPENDIX 1

Transcription methods and conventions

NP’S’IPP’

Japanese transliteration is given in Kunree style.

Pair identification and the identification of the data source are given in
parenthesis at the beginning of each data presentation. For example,
(MA/811 517) means that the conversation is between participant M and
A, and the excerpt is from in page 8, lines 15—17 In the original
transcription.

Data are numbered for each chapter.

When proper nouns appeared in the data, pseudonyms are used.
Bold is used to highlight the items which is discussed in the text
CAPS indicates very emphatic stress

Paralinguistic and other interactional symbols are:

I recognizable pause

. utterance-final counter

? rising interrogative contour

l animated tone

noticeable pause or break in rhythm without falling intonation

((3)) indicates long pause lasts 3 seconds.

- self interruption

[ starting point of simultaneous speech

((Iaughter» indicates various types of laughs and laugh-like utterance
When speech from B occurs during what can be heard as a brief

silence from A, then B's speech is under A's silence.
A: I can't wait to go to the party! It'll be fun.
B: Oh! yehl

64

Back—channel expressions in narratives are indicated in the parenthesis
When transcribing, not all morphemes are separated. Only those
morphemes relatively important to the present study are separated with
glosses given. when glossing, hyphens are used only when overt
indication of the connection between grammatical and lexical elements
is useful.
Following abbreviations are used:
BE copulative verb, be
CAU causative morpheme
FP final particle, such as ne, yo, and saa.
lO indirect object
LK linker (linking nominals and nominal adjectives)
NEG negative

nominalizer

direct object marker

NOM
O
Q question marker (assigned to final particle ka only)
QT quotative marker

8 subject marker

T topic marker

TAG tag-question-like auxiliary verb forms (including zyan, zya-nai,

desyo, daroo, etc.)

65

APPENDIX 2
Selected data
(MA/811229)
M: Un un un un.l Soc ne./ (2) lyaa boku mol atukute.. neel iyaa-l moo taimai
hataityaimasita yol
A Nani ga?l
r Kuuraa neelE [iroiro anote konote del isya no sindansyo toka tukete=
2 e. Eel

x2: =morattelE Sorede kyoka o totta n desu yo.l Totta wa ii n da kedol:
2 9.!

M3: =Soredel konoaida anool tukenikite neel Del sositara anool sore
2 Ee.l

M4: =yuzuttemoratta kuuraa nandakedol del kaette! kore za tukanai Its.

A: Ee.l A
soo.l

M Iyaal moo sekkaku koko made kogituketa noni tte omottel MOO koko
made yattara tukeru sika nail tte/ kaini itta n desu yo.l

A Ee ee.l Al ((Iaughter»

M Eel mool taimai! 500 doru mo kakatta.l

A2 ((laughter»

M: Tukattyaimasita yo.l

A: ((laughter»

(NB/48.513543)

B: Dizuniiwaarudo ga dekuru to saal kankyoo ga hozen sarerun da yol are.l

N Aa soo desu ka.l

8 Are saal Dizuniiwaarudo tte doo site tatta ka sitteru?/ Are sal kekkyokul
Dizuniirando o tukutte saal anoo-l Dizunii ga Dizuniirando o tukutte!
kekkyokul Dizuniirando no naka tte yuu nowal jibun de sukina yoo ni
kool sekkee ga dekiru wake da yo ne?!

N: Hai.l

B: Dakedol kekkyoku asoko no-l Anahaimu no atari tte saal anoo-l dotti ka
tte yuu tol sonnani anmari fansiina tokoro za nai za nail Ano kinpen tte
saa.]

N: Un un./

82 Sorede saal mawari ni saal sonoo-l nante yuu no?/ Dizunii ga
kangaete nakatta yoo na/ kankyoo ga dandan dekite kitate saal
((unclear» Sorede! jibun no risoo to hodotooi tokoro ni natta to.l
Nanka koo horal furiiwei o orite saal nanka koo-I isi nol ne?! nanka
hokorippoi michi o zuuto hasitteikuto/ Dizuniirando ga aru tte yuu nowal
doomo risoo ni fusawasikunai kara saa. l Sonzyaa neel kondo
atarasiku Dizuniirando o tukutte saal sono mawari ni-l sono mawari o
zenbul ano-l jibun no omoidoori ni sityau/

N: Fuun. l

B: De soredel kekkyokul Dizuniiwaarudo o tukuru toki nil

Dizuniiwaarudo no toti no mawari nil sono Dizuniiwaarudo no sikiti

66

no 10 bai no toti o katte/ de sore o zenbu sono-l nante yuu no?! Sono
sikici nai o katteni sasenai./

(YRl1-12)

Y' De autoretto no omise ippai aruno.l Honde neel soko itte kita no.I

R. Soc kaa.l '

Y: Un. soo.l

R Fuun.l De higaeri?!

Y Uun.l lppaku site.l Nde Furiipocto e ittel de Aakeezia kokuritukooen tte

5.3.3 $.39

.<

53.35.395.375? $39K"

¥$353<¥<§9<¥<¥r€§9$¥

Nfii‘?

arun da yo ne.l Soko ittel. unl de kaette kitel kinoo no yonaka no itii zi
goro kaette kita./
Samuku nakatta?!
N to neel kyoo-I itiniti me sugoku samukutte/ demo kinoo sugoku hare
tel un sugoku yokattaJ

A honto. Soo ka.l Kotti wa sugoku samukattaJ
((laughter»

R12,3&4l32-2)

Kyonen neel Sakagami sensee to neel issyo ni neel ututta no.I
issyo ni ututta no?l Aryaa.l li neel futari pea de.l
Sosite saal uti no gakkoo no ongaku no sensee mo neel kono haru
sotti ni ututta no.I
Uti no gakkoo tte?!
Meiji no.I
Anoo otoko no hito?!
A sitteru?! Tuzuki sensee tte.l
Nanka hen na hito desyo?! ((laughter»
((Iaughter»
Nanka myoo ni saal seeto no kotol bakobako tataku toka saa.!
A soc sool 500 $00.! Sono him! (1) Ga ututte neel Meizi no nltubo to
kasite iru no.I Sonde neel o- omosirai yol mada.l De kono haru neel
Yosida sensee ga Meien ni modotte kita no.I
Usoo?ll Nanil kyootoo de?!
Soo.l
Doko itteta no.I
To— to
Kurisawa ka dokka?!
Iya.l Takada syoogakkoo no kyootoo da tta no.I Un, dakara slnai idoo=
Aall
=bakkari.l
Demo-I heel
Ndeneelmadaaruyoj Hadasenseeganeel Keeseenokyootoo:
Nanil
=nano.l
USOOI/
Honto.l
U-I Tyotto taimu.l Kyootco na no.I
Soo da yo.l Ima made Tukizi no kyootoo da tta no.I
A hontoo ni.I Zaa ano nendai wa minnal kyootoo ni natte simatta no ne?!

33 2.3

535353535

35353 5353535353 5353

2

2.3 33 .Z"

67

$00.!

Datte—l

Hasimoto sensee mo da yo?!

Hasimoto sensee-l

Hasimoto senseel Hokusee-l kana?! Un! Hokusee no kyootoo!

Sugoi no.I

Un.! Sorekara-l

lsogasii ne/ Meizi no sense datta hito.l

Tagawa senseel ninna taisyoku sita si.l Site Furuno sensee ga neel nnn

Akagawa no kyootoo?! Un! soo.! Ano zidai wa dakara sugoi n da!

dakara/nnn! Meizi o tukkuta hitotati ga neel

Soc kaa.!

Soc 300.! Dareka kikitai hito iru?!

Dareka?!

Ato-l Ato neel Kita san lsiyama ututta n da.l Un daa Meizi dare mo inai

moo.l Ore no sitteru hito inai.l

Are itu ututta no?!

Ore?!

Tigau tigau tigauJ

Kita san?!

Unl

Kita san kono haru.l Un sool Minato sensee ga ichiban furui n datte ima.l

A hontoo ni.l

Un.l

Sika inai no?!

Uunl Minato senseel sitteru tokoro de ieba neel Mori sensee de syoo?!

Nisi sensee de shoo?l.Naka sensee mo detyatta sii/

A hontoo!

Kooryoo itta n daa!

De kyootoo?!

Uun! iyaa tigau tigau.l ((laughter»

((laughter» Tyo- tyotto yabai ne./

B!2,3&4!31-2) .

Demol kono aida syabette ita no wal nanka! Kyooto de kurumi

gisukuriimu o tabetan desu yo.l Del konna-l konna tte ittatte/ tyokkee-l
ore tte—l

3 senti! 5 senti mo nai! 3 senti gurail no aisukuriimu no yama gal 500

en!

Takai yo.!

Un.! De! kotti dato 5 doru?! Honto wa motto! 6 doru gurai desu yo neel

moo ima! 6 doru aisukuriimu o saatliwan de-I za nakutemo ii kedol

kyanpasu de-I

Sore yori saal hora! baketu no yatu kau to sugoi! ((Iaughter»

6 doru bun kattaral dore dake tabereru kanaa toka omotte.! Dee! soc ,

sum to nanka! nihon ni kaettara nanimo taberarenai naa tte omotte.!

Are wa neel baketu de kattaral hito tuki taberareru.!

Baketu tte anool kore gurai no yatu desho?! Anoo-l syaabetto toka

haitteru yoo na.l

3333

33.23

3333 333 33.2.3

3 2.3.2.3? 33 3?

68

$00 soo soo.!

3-/ 3 pondo no yatu?!

Are-l dakedo! syaabetto dake za naku tel aisukuriimu nomo aru yo?!
Atasil ano-I kyanpasu no! konaidal anoo-I aisukuriimu no kappu no
kattan desu yc.! lkura ka zenzen siranakutel dee-l deel ano—l hazimete
katte mitaral 1.5?! kekkoo yasui na toka omotte.! Tiisai yatu ne?! De!
dekkail 2 bun no 1 garon?! 2 bun no 1 garon de 2 doru 75?! Dakaral maa
maa da to omotte.! Motto takai imeezi ga atte.l Dee-l kedo are! konnan
dattara nee! anoo-lkondo kara sore ni siyoo kana tte omottel maiyaa toka
de ironna no kau yori wal De! demo nee! konoaida ittara nee!
bataapiikan ga nakuttel kanasikutel okasiinaa/ nande nain daroo tte!
Noriko san bataa suki nano?!

lyal

Kekkoo! kekkoo kau za nail

Bataapiikan wa./ lyaa! sore made wa tabeta koto nakute! aru toki made!
bataapiikan! De—l mae nee! ano-l Rituko san gal anoo-l waffurukoon nil
bataapiikan to banira to ryoohoo iretel tabesasete kureta toki atta
desyo?!

Aal atta neel

Minna de.l Antoki hazimete tabeta n desu yo.!

Aa hontoo!

Unl are kara KORE WA OISII toka omotte.! Sorekara bataapiikan sika '
kawanai tte.!. A banira wa tokidoki kau kedo!

Ore! banira no hoo ga suki!

Aa hontoo! Mae wa neel tyokominto?! bakari datta no.I

Demol konkai nee! Amerika ni kite neel aisukuriimu wa nee! anmari
kangeki sinakatta.!

Mae wa kangeki sitan desu ka?!

Soo. Mukasi neel hazimete Amerika ni kita toki saal

Aa...soo.l

Ano toozi nee! ano toozi saa.! kekkoo mukasi no hanasi da kedo saal
anoo-l nante yuu no?! Nihon ho gyuunyuu ga mada sugoku usukutte
saal

Aa.!

Amerika no-I Amerika no gyuunyuu ga madal unto kokatta zidai de saal
de! ma dattara sore kosol

Gyaku desu yo nee?!

Del mae no saal mukasi no Amerika no gyuunyuu tte saal eraku kokatta
no. ! Nanka nikibi ga dekiru gurai kokatta no I

4.-I 2 gurai?!

All sonna tokoro za nai yo?!

Motto?!

Motto hidokatta yo?!

Nihon de nookoo gyunyuu del 4. 5 gurai desu yo ne?! Seezee.l Nihon no
nookoogyuunyuu. l

Sorede saal ano-l saisyo ne?! Saisyo Honoruru ni tuita wake. I Del minna
ni sal mazu Honoruru ni tuitara nani suru?! mazu aisukuriimu o kue tte./
Sorede! nani ga umai tte kiitara/ painappuru ga umai tte.l Sorede saal

69

tukunari itiban saisyo ni saal aisukuriimu o tabeta/(2) Kangeki simasita!
boku wa.!

(NB/6&7l13-8)

N: 800.! Dakaral (1).nanka uti no choc wal kekkoo anmari-l yuuzyuufudan
tte yuu ka! anmari toti ni sonna ni syuutyaku aru hito! sonnani inai za nai

- kana! tte ittara are da kedo!

B: Demol are desyo?! Ano-l uun to.. nani?! daitai are?! Anoo-I

N2 Sorekoso-l

B: Hatake yatteru toka! soo yuu hito?!

N: Tanbo bakkari!

B: Tanbo?!

N: All uti wa anoo-l kengyoo nooka col?!

B: Maa koozyoo coin dattara nee!

N: 800.! De...dakaral hatake toka tanbo toka wal kisetu dake! Demo! uti de
taberu okome tol aru teedo uru okome wa aru n desu yo ne?! Da[kara-

B: [Dakara
da yoo.l Kengyoo da kara!

N: Un...Demo! uti-l uti mo soo desu kedol anmari-l uruttatte seezee nee!
Hoondo-I

B2 Kengyoo tte mo saal hora-l nante yuu no?! Zissai nlwa noosagyoo no
hoo ga syu de saal sorekara! betu no hoo ga zuu tte yuu no to saal futuu
no! maal nante yuu no?! Koozyoo zutome nanka no ga syu de saal
noosagyoo no hoo ga zyuu tte yuu no mo aru desyo?! Dotti no hoo ga
ooi n desu?!

N: Koosya./

B: Dakara da yo./

N: Dakaral sigoto ga aru kara betu ni toti ga...nee.l Sore to! uun...l Doc nan
desyoo ne?! lya! mukasi! sono-l watasi no-l

B: Kibo we 00in no?! Kibo tte yuu ka/ toti no kibo.!

N: Hitcri?! 1 kken zutu no?! Kekkoo aru to omoimasu yo?! All tte yuu ka
Hokkaidoo hodo zya nai desu kedol Inaka-l kinkoo noogyoo desu kara
nee.l Demo! watasi! tubo ga wakaranai 11 del sirimasen kedo/....doredake
nan de syoo ne?! Nootikaikaku ga-l

B: Nootikaikaku?!

N2 Arimasita yo ne?! Sengo.l Are del daibu! ironna hito gal toti motta kara!

8: Naomi sanl oboeteru?!

N: Oboetemasen yo.l Kiita hanasi ni yoru to! anoo-l zenbu-l ne?! Kosaku
noo hito ga zenbul toti o te ni ireta desyo?! Tada de.!

8: Un..Tada?l

N: Unl tada.!

B2 Hmmm. Soo yuu zidai ni umareteoku beki datta!

N: Del dakaral mukasi no zimusi mitaina no ga otibureta tte.l Demol ano-l

nan desyoo ne?! Dore kurai desyoo ne?! Metyakutya arimasen yo?!
Ano-I sukosi zutu./ Maal dakaral Demol minna! sono tanbo to hatake '
dake dewa zettai seekatu sitekemasen yo?! Dakaral kekkyoku kengyoo
nooka bakkari....lUn! de.. un-I

7O

(YR/4&5l40-18)

R:
Y
R
Y:_
R:
Y.
R
Y
R

1:

Y:
R2
Y.
R1
Y2
R2:
Y.
R
Y
R
Y
R
Y

lru./ Watasi neel natu no aida ni me! sotugyoo suru yotee datta n
desu kedo!
Un.l
Nobimasita.l ((laughter»
Nasita?! Are wa?! Siisasu wa?!
Siisasu wa neel natu no aida ni kaku.l
A hontoo ni.l
De neel natu no aida ni sotugyoo-l sotugyoo suru tumori da to-l
Un.l
Anoo! fanaru dorafuto o neel 6 gatu 30 nichi ni dasanakya nan na=

Un.l
=katta n da.
Un.
De difensu o 7 gatu no gejun made yattel 8 gatu 5 ka ni anool fainaru:
Un!

=baazyon o ofiisu ni teesyutu na no ne.! Watasi nee! nan tta kke naal
puropoozaru o dasita no gal 4 gatu no sue datta kara-l
Ara! osokatta n daa! ((laughter» Aal sorede kekkyoku nani?! Natu ni
sotugyoo tyuu ka-I
lyaa natu ni-l natu no aida ni nee! 8 gatu no atama gurai made ni
kaityatte!
Un
De ooraru difensu wa-I Anoo! akl gakki ni nobasu kedo-l

um um
Un...dakaral osoku natyatta no.I
Un.l Un un un.l Aal honto ni./

LIST OF REFERENCES

71

REFERENCES

Allwood, Jens, Lars-Gunnar Andersson, and Osten Dahl. 1977. Logic in
Linguistics. (Cambridge Textbooks in Linguistics) Cambridge:
Cambridge University Press.

Blimes, Jack. 1988. The concept of preference in conversation analysis.
Language in Society 17. 161-181.

Blakemore, Diane. 1987. Semantic constraints on relevance. Oxford:
Blackwell.

Brown, Gillian and George Yule. 1983. Discourse analysis. (Cambridge
textbooks in Linguistics) Cambridge: Cambridge University Press.

Chafe, Wallace. 1980. The deployment of consciousness in the production of a
nanative. The pear stories. by Wallace Chafe, 9-50. Norwood, NJ:
Ablex.

. 1985. Linguistic differences produced by differences between speaking
and writing. Literacy, language, and learning: The nature and
consequences of reading and writing, ed. by David R. Olson, Nancy
Torrance, and Angela Hildyard, 105-123. Cambridge: Cambridge
University Press.

. 1994. Discourse, consciousness, and time. Chicago: The University of
Chicago Press.

Clancy, Patrica. 1982. Written and spoken style in Japanese narratives.
Spoken and written language, ed. by Deborah Tannen, 55-76.
Norwood, NJ: Ablex.

Cook, Haruko M. 1993. Functions of the filler am in Japanese. Japanese
Korean Linguistics (Vol. 3), ed. by Soonja Choi, 19-38. Stanford: SLA.

Du Bois, John W., Susanna Cumming, Stephan Schuetze-Coburn, and Danae
Paolino. 1992. Discourse transcription. (Santa Barbara Papers in
Linguistics Vol. 4). Santa Barbara, CA: Department of Linguistics, UC
Santa Barbara.

Fraser, Bruce. 1990. An approach to discourse markers. Journal of Pragmatics
14.383-395. ./ . l

72

Grice, P H. 1975. Logic and conversation. Syntax and semantics 3: Speech
acts, ed. by P. Cole and J. Morgan, 41-58. New York: Academic Press.

Gumperz, John. (ed.) 1982. Language and social identity. Cambridge:
Cambridge University Press.

Heritage, John. 1994. Constituting and maintaining activities across sequences:
Arg- prefacing as a feature of question design. Language in Society 23.
1- .

Hinds, John. 1978. Conversational structure: An investigation based on
Japanese interview discourse. Problems in Japanese syntax and
semantics, ed. by John Hinds and l. Howard, 70-121. Tokyo: Kaitakusha.

Holmes, Janet. 1986. Function of 'you know' in women's and men's speech.
Language in Society 15.1-22.

lde, Sachiko. 1982. Japanese sociolinguistic: politeness and women's
language. Lingua 57. 357-385.

. 1991. How and why do women speak more politely in Japanese?
Aspects of Japanese women's language, eds. by Sachiko lde and
Naomi Hanaoka McGloin, 63-79. Tokyo: Kuroshio Publishers.

lde, Sachiko and Naomi Hanaoka McGloin.(ed.) 1991. Aspects of Japanese
women's language. Tokyo: Kuroshio Publishers.

Karatsu, Mariko. 1994. A functional analysis of DEWA, DAKARA, and SIKASI in
conversation. Paper presented at ATJlMlT conference. University of
Minnesota.

Kindaichi, Kyoosuke, Umetomo Saeki, and Hatsutaro Ooishi. (ed.) 1974.
Shinsen Kokugo Jiten. Tokyo: Shoogakkan.

Kuno, Susumu 1973. The structure of the Japanese language. Cambridge: MIT
Press.

Labov, William. 1972. Language in the inner city: studies in the black English
vernacular. Philadelphia: University of Pennsylvania Press.

Lowday, Leo. 1986. Japanese sociolinguistic.(Pragmatics and beyond VII: 1)
Amsterdam: John Benjamins Publishing Company.

Makino, Selichi and Michio Tsustui. 1986. A dictionary of basic Japanese
grammar. Tokyo: Japan Times.

Maltz, Daniel M, and Ruth A. Barker. 1982. A cultural approach to male-female
miscommunication. Language and social identity, ed. by John Gumperz,
196-216.

73

Maynard, Senko Kumiya. (1989). Japanese conversation: self-
contextualization through structure and interactional management.
Norwood, NJ: Abex.

. 1992a. Speech act declaration in conversation: functions of the
Japanese connective datte. Studies in Language 16(1). 63-89.

. 1992b. Discourse modality: subjectivity, emotion and voice in the
Japanese language. Amsterdam: John Benjamins Publishing Company.

. 1993. Kaiwa bunseki [Discourse analysis]. Tokyo: Kuroshio
Publishers.

National language Research Institute (Kokuritu Kokugo Kenkyuujo). 1955.
Danwago no Jittai (The current situation of the language of discourse).
(NLRI report 8). Tokyo: Shuuei Shuppan.

Nihon Kokugo Daijiten. 1979. Tokyo: Shoogakkan.

Onodera, Noriko Okada. 1993. Pragmatic change in Japanese: conjunctions
and interjections as discourse markers. Unpublished dissertation.
Georgetown University.

Pomerantz, Anita. 1984. Agreeing and disagreeing with assessments: some
features of prefened/dispreferred turn shapes. Structures of social action:
Studies in conversation analysis, ed. by Maxwell J. Atokins and John
Heritage, 57-101 . '

Reynolds, Katsue. Akiba. 1991. Female speakers of Japanese in transition.
Aspects of Japanese women‘s language, ed. by Sachiko lde and Naomi
Hanaoka McGloin, 129-145. Tokyo: Kuroshio Publishers.

Romaine, Suzanne and Deborah Lange. 1991. The use of like as a marker of
reported speech and thought: a case of grammaticalization in
progress. American Speech 66(3). 227-279.

Sacks, H., Schegloff, E. A., and Jefferson, G. 1974. A simplest systematics for
the organization of tum-taking for conversation. Language 50. 696-735.

Saito, Makiko 1992. A study of the discourse marker nanka in Japanese.
Unpublished MA Thesis. Michigan State University.

Schiffrin, Deborah. 1987. Discourse markers. (Studies in interactional
sociolinguistics 5) Cambridge: Cambridge University Press.

. 1994. Approaches to discourse. Cambridge: Blackwell Publishers.

74

Searle, John R. 1975. Indirect speech acts. Syntax and semantics (Vol. 3):
speech acts, ed. by P. Cole and J. L Morgan, 59—82. New York:
Academic Press.

Shibamoto, Janet S. 1985. Japanese women's language. Orlando: Academic
Press.

Shibatani, Masayoshi. 1990. The languages of Japan. Cambridge: Cambridge
University Press.

Simon, Mutsuko E. 1986. Supplementary grammar notes to An introduction to
modern Japanese Part 1. Ann Arbor: University of Michigan.

Tannen, Deborah. 1982. Ethnic style in male—female conversation. Language
and social identity, ed. by John Gumperz, 217-231.

. 1989. Talking voices: repetition, dialogue, and imagery in
conversational discourse. (Studies in interactional sociolinguistics 6)
Cambridge: Cambridge University Press.

. 1990. You just don't understand: Women and men in conversation.
New York: Ballantine Books.

Trudgill, Peter. 1972. Sex, covert prestige and linguistic change in the urban
British English of Nonlvich. Language in Society 1.179-195.

Uemura, Kooichi and Masatoshi Tabuki. 1994. Hypermedia corpus and
Japanese Conversational Analysis. Proceedings of the Second
Princeton Japanese Pedagogy Workshop, ed. by Selichi Makino, 184-
200. Princeton: Princeton University.

Webster's third new international dictionary 1981. Springfield, MA: G. & C.
Merriam Company.

Yokobayashi, Hisayo and Akiko Shimomura. 1988. Setsuzoku no hyoogen.
[The expressions of connectives] Tokyo: Aratake Shuppan.

nICHIGnN STATE UNIV. LIBRARIES
iIIIIIIIIiI iIIIIII iIIIIIIIiIIlII IIIIIIIIII iIIIIIIIIIHI
31293014202786