a storyline analysis in iu-mienh narrative discourse with some insights from cognitive linguistics -...

402
A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS by Tatsuro Daniel Arisawa Presented to the Graduate School of Payap University in Partial Fulfillment of the Requirements for the Degree of MASTER OF ARTS IN LINGUISTICS PAYAP UNIVERSITY, CHIANG MAI, THAILAND June 2006

Upload: frugle133

Post on 30-Jul-2015

231 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

A STORYLINE ANALYSIS IN

IU-MIENH NARRATIVE DISCOURSE

WITH SOME INSIGHTS FROM

COGNITIVE LINGUISTICS

by

Tatsuro Daniel Arisawa

Presented to the Graduate School of Payap University

in Partial Fulfillment of the Requirements

for the Degree of

MASTER OF ARTS IN LINGUISTICS

PAYAP UNIVERSITY, CHIANG MAI,

THAILAND

June 2006

Page 2: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

A STORYLINE ANALYSIS IN

IU-MIENH NARRATIVE DISCOURSE

WITH SOME INSIGHTS FROM

COGNITIVE LINGUISTICS

The members of the Committee approve the master’s thesis of Tatsuro Daniel

Arisawa

Assistant Professor Dr. Thomas M. Tehan ________________________________

Supervising Professor

Professor Dr. Somsonge Burusphat ________________________________

Associate Professor Dr. Kingkarn Thepkanjana ___________________________

Assistant Professor Dr. Kirk P. Person _______________________________

Ajarn Audra Phillips ____________________________________

Approval Date: _____________________________________________________

Page 3: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

Copyright © Tatsuro Daniel Arisawa 2006

All Rights Reserved

Page 4: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

A STORYLINE ANALYSIS IN

IU-MIENH NARRATIVE DISCOURSE

WITH SOME INSIGHTS FROM

COGNITIVE LINGUISTICS

by

Tatsuro Daniel Arisawa

Presented to the Graduate School of Payap

University

in Partial Fulfillment of the Requirements

for the Degree of

MASTER OF ARTS IN LINGUISTICS

PAYAP UNIVERSITY, CHIANG MAI,

THAILAND

June 2006

Page 5: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

RESUME

Name: Tatsuro Daniel Arisawa

Date of Birth: 20 April, 1959

Place of Birth: Asahikawa, Hokkaido, Japan

Institutions Attended:

1983 Hokusei Gakuen University, Sapporo, Japan. Bachelor of Arts (in

English Linguistics and Literature)

2006 Payap University, Master of Arts in Linguistics

Academic Paper Presented:

Forthcoming. A place where time, aspect, and modality meet: A case of

/tu/5/ (duqv) ‘to get’ in Iu-Mienh, The 14th Annual International

Conference of the Southeast Asian Linguistics Society (SEALS 14),

Bangkok, Thailand, 19-21 May 2004

Page 6: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

ACKNOWLEDGMENTS

This thesis could not have been completed without the following people, whom I

sincerely would like to acknowledge for their teaching and help. The Supervising

Professor Dr. Thomas M. Tehan has been repeatedly teaching me to grind the

hands-on work with data and has been patient with and encouraging me when my

progress was slow. Dr. Kirk R. Person found many articles by Longacre, some of

which are hard to find in Thailand, making them available for me from the library

of the University of Texas, Arlington. Ajarn Audra Phillips refined my grammar

chapter by providing some of the latest books and articles and discussion

opportunities. I am grateful to the external examiners, Dr. Somsonge Burusphat

and Dr. Kingkarn Thepakanjana.

Narrowing down to a study of storyline was a wise counsel given by Ajarn Karen

L. Block, our Department Head, to whom my sincere thanks are extended. Ajarn

Terrance D. Gibbs helped me solving technical problems of the computer with

patience and kindness. Ajarn Prang Thiengburanathum, who used to be my fellow

student, now a faculty member, collected many valuable papers and articles for

me in the University of Oregon amid one of the severest challenges of her health,

for whom I cannot be thankful enough. Many fellow students have been an

encouragement to me. Emily Lewis proofread the free translations of the

narrative corpus. I shall cherish many memorable occasions with them.

I appreciate very much Dr. Herbert C. Purnell and Ajarn Wanda Jennings,

speakers of Iu-Mienh, for their comments. I thank Ms. Ann Burgess for sharing

the narrative data with rich grammatical and cultural information, and Greg and

Misako Aumann for sharing their extensive electronic library on Hmong-Mien

studies. I will not forget Dr. Christophor A. F. Court’s encouragement through

correspondence in my early stage of Iu-Mienh language learning. To Dr. Robert

E. Longacre, I wish to express my deep gratitude. Though I have never met him

in person, certainly I have felt his heart in reading his writings, which guided me

to read many Hebrew narrative texts, showing me that he is a lover of the Word.

v

Page 7: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

I would like to acknowledge my debt to my Iu-Mienh colleagues of Associated

Churches in Thailand, Mien Section: especially to Pastor Zeuz Jiem-En, and other

friends, Fuqc Luangh Fin-saeng, Laai Fei Dorc, Naix Yauz Dorc, who generously

tried to free me from pastoral duties. I owe the members of the Japanese Christian

Church in Chiang Mai and many churches in Japan for their continuous, generous,

practical, spiritual and financial support.

I thank my mother Atsuko Arisawa for consenting to my return to Thailand that I

could finish my study despite her health crisis. Due to her sickness a timely

discussion opportunity with Rev. Dr. Yoshinobu Endo was realized in my

hometown and he gave me highly constructive instruction and encouragement. I

am also thankful to Mr. Zanh Gueix-Fongc, Mrs. Dangc Meix Daqv and Mr.

Bienh Wuonh Mengh for teaching me their language. My warm thanks go to

Professor Smokiast Jamlorm and Ajarn Thaworn Foofuang of the Tribal Museum,

Chiang Mai, for giving a special lecture on the Iu-Mienh history and culture to the

Inter-University Iu-Mienh Students Club, allowing me to sit in. Sylvia Lombard,

whose dictionary has been an important source of data, passed away in the course

of my writing last year. My condolences and gratefulness go to Ms. Rosaline

Lombard for sending me her invaluable language materials from the 1960s.

Last but not least, to my wife Tamami, I simply say “thank you.” I myself cannot

understand how I could finish writing this thesis in our busy life fairly deeply

involved in life-and-death issues of many precious people in their marriages,

funerals, and cases of crisis counseling. It was only possible by her love, joy,

peace, cheerfulness, encouragement, humor, patience, sacrifice, practical wisdom

in the face of emergencies, and in-depth prayers in close touch with the Almighty

Triune. Her characteristics are described in the words: qjrw axmy ym lyj-tva

Hrkm <ynynpm (Proverbs 31:10), .hnlk-lu tylu taw lyj wcu twnb twbr (29),

.llhtt ayh hwhy-tary hva (30b). To her I humbly dedicate this thesis with

affection. Sōla Deō glōria.

June 16, 2006

vi

Page 8: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

vii

Page 9: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

TITLE : A storyline analysis in Iu-Mienh narrative discourse with some insights from Cognitive Linguistics

AUTHOR : Tatsuro Daniel Arisawa FACULTY : LINGUISTICS, GRADUATE SCHOOL, PAYAP UNIVERSITY, CHIANG MAI, THAILAND YEAR : 2006 THE MEMBER OF THE COMMITTEES:

1. Dr. Thomas M. Tehan 2. Dr. Kirk Roger Person 3. Ajarn Audra Phillips 4. Dr. Somsonge Burusphat 5. Dr. Kingkarn Thepkanjana

SUPERVISING PROFESSOR: Dr. Thomas M. Tehan

ABSTRACT

Storyline is a foregrounded main line of development in narrative discourse.

Longacre’s textlinguistics storyline analysis distinguishes a storyline verb that is

marked by the preterite tense from supporting materials coded with other tense-

aspect-modality such as imperfect, progressive, modals, irrealis, etc. Among

these he recognizes a cline of ranking from most dynamic to most static (1981,

1996, 2003a). In applying his theory to Iu-Mienh narrative discourse, two issues

arise: 1) the advantage of a verb ranking/salience scheme approach which is

compatible with prototype analysis from a Cognitive Linguistics perspective

(Taylor 2003), as opposed to the binary approach of foreground-background

(Hopper 1979), and 2) the problem of strong dependence on verb morphology

which Iu-Mienh lacks totally as does Thai in identifying storyline (Somsonge

1990b. 1991a). What possible alternatives can Cognitive Linguistics suggest to

such a situation?

Starting with a storyline analysis and salience scheme analysis based on seven Iu-

Mienh narrative discourses within the framework of Longacre’s textlinguistics,

this thesis shows first a prototype effect, or gradient relation, not only in the

supportive materials but also inside each band. The storyline verb in Iu-Mienh is

viii

Page 10: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

unmarked, used 1) by itself, 2) with post verb aspectual verbs, or 3) in serial verb

constructions (SVCs) and topic chains. The pivotal storyline is recognized as a

band of grounding elements mi’aqv ‘resultative aspectual marker,’ aqv ‘perfective

aspectual marker’ and an adverb za’gengh ‘really, indeed, actually.’ Second,

following Somsonge’s (2002) non-verb-morphology dependent storyline analysis,

transitivity and sequentiality as the major factors of storyline are analyzed, using

Langacker’s billiard-ball model (1991b) and updating (i.e. increasing) Current

Discourse Space (CDS) model (2001). From a Cognitive Linguistics perspective,

storyline is foregrounded to the perception of the conceptualizer by two factors: 1)

as a result of the transitivity of energy and event in a clause or sentence composed

of SVCs and topic chains, and 2) as a result of sequentiality expressed by the

sequential marker ziouc ‘then, so, and,’ the conjunction cingx_daaih ‘therefore,’

the topic marker aeqv ‘as for,’ and adverbial clause containing liuz ‘after doing…’

or gau ‘after which.’

To identify storyline, Cognitive Linguistics analyzes conceptual structures of

transitivity and sequentiality, which result in foregroundedness to human

perception. Storyline is an epistemological outcome of transitivity and

sequentiality.

ix

Page 11: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

ช่ือเร่ือง : การวเิคราะหการดําเนนิเรื่องของวจนะเรื่องเลาในภาษาอิ้วเมีย่นในมมุมองของภาษา

ศาสตรเชิงปริชาน

ช่ือผูแตง : Tatsuro Daniel Arisawa

สาขาวิชา : ภาษาศาสตร บัณฑติวทิยาลัย มหาวทิยาลัยพายพั จังหวดัเชยีงใหม

ประเทศไทย

ป : 2549

รายช่ือคณะกรรมการ :

1. Dr. Thomas M. Tehan

2. Dr. Kirk Roger Person 3. Ajarn Audra Phillips 4. Dr. Somsonge Burusphat 5. Dr. Kingkarn Thepkanjana

ประธานควบคุมวิทยานิพนธ : Dr. Thomas M. Tehan

บทคัดยอ

เหตุการณดําเนินเรื่องเปนแกนหลักตอนเดนในการดําเนินเรื่องเลา ลองเอคเคอรไดแบง

ลักษณะไวยากรณของเหตุการณดําเนินเรื่องเลา ดังนี้ คือ คําแสดงรูปอดีตกาล เปนตัวบงชี้กริยา

เหตุการณสําคัญ สวนคําแสดงรูปกาล-การณลักษณะ-modality เชน imperfect, progressive,

modals, irrealis เปนตัวแสดงรายละเอียดเสริมเหตุการณ นอกจากนี้ ลองเอคเคอรไดเรียบเรียง

ลําดับความเดนของขอความจากสวนที่เปนการดําเนินเรื่องที่มีการเคลื่อนไหวที่สุดจนถึงสวนที่

x

Page 12: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

เปนสภาวะหยุดนิ่งที่สุดของเรื่อง(2524, 2539, 2546) จากการประยุกตทฤษฎีของลองเอคเคอร

กับเรื่องเลาในภาษาอิ้วเมี่ยน พบวามีประเด็นที่นาสนใจ 2 ประเด็น คือ 1) แนวคิดการเรียงลําดับ

ความเดนของขอความนั้นตรงกับแนวคิดของ prototype analysis ตามมุมมองของภาษาศาสตร

ปริชาน (เทเลอร 2546) ซ่ึงตางจากวิธีของฮอบเปอร ที่แบง ขอความเปน 2 ขั้ว คือ ขอความเดน

และ ขอความรอง (ฮอบเปอร 2522) 2) วิธีการวิเคราะหของลองเอคเคอร ใชไดดีในภาษาที่มี

ระบบกาล-การณลักษณะ-modality แสดงในรูปคํากริยา จึงทําใหเกิดปญหาการวิเคราะหใน

ภาษาอิ้วเมี่ยน ซ่ึงไมมีการผันรูปคํากริยา เชนเดียวกับภาษาไทย (สมทรง 2533b 2534a) มีทาง

เลือกใดบางในการแกปญหานี้ในมุมมองของภาษาศาสตรเชิงปริชาน

วิทยานิพนธฉบับนี้เร่ิมจากการศึกษาเหตุการณดําเนินเรื่องและการเรียงลําดับความเดน

ของขอความเรื่องเลาในภาษาอิ้วเมี่ยนจํานวน 7 เร่ือง ตามแนวทางของลองเอคเคอร จากการ

ศึกษาพบวา prototype effect ไมเพียงพบแตในรายละเอียดเสริมเหตุการณเทานั้น แตยังพบใน

แตละอันดับขอความดวย คํากริยาแสดงการดําเนินเรื่องในภาษาอิ้วเมี่ยนจะไมมีตัวบงชี้ทาง

ไวยากรณ คํากริยาดังกลาวปรากฎอยู 3 ลักษณะ คือ 1) อยูโดยลําพัง 2) ปรากฏกอนคําแสดง

การณลักษณะ และ 3) ปรากฏในโครงสรางของกลุมอนุพากย และ กลุมภาคแสดงตอเนื่อง

การดําเนินเรื่องหลักมีตัวบงชี้ทางไวยากรณคือ grounding elements ซ่ึงมีอยู 3 ชนิด คือ คําแสดง

ผลการณลักษณะ mi’aqv การณลักษณะแสดงการเสร็จสมบูรณ aqv และ คํากริยาวิเศษณ

za’gengh‘จริงๆ’

วิทยานิพนธฉบับนี้วิเคราะหลักษณะไวยากรณที่แสดงลําดับเวลา และ การเคลื่อนไหว

xi

Page 13: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

ของเหตุการณ ตามแนวทางของสมทรงในการวิเคราะหภาษาที่ไมมีการผันรูปกริยา (สมทรง

2545)โดยใช Billard-Ball Model (2534b) และ Current Discourse Space Model (2544) ของ

แลงเอคเคอรในการวิเคราะห จากมุมมองของนักภาษาศาสตรเชิงปริชาน เหตุการณดําเนิน

เร่ืองจะเปนขอความเดนในมโนทัศนของผูอานหรอผูฟงไดดวยสาเหตุสองประการคือ 1) การ

เคลื่อนไหวของอาการและเหตุการณ ซ่ึงพบในโครงสรางของกลุมอนุพากย และ กลุมภาค

แสดงตอเนื่อง 2) ลําดับเหตุการณในเวลาตอเนื่อง ซ่ึงมีตัวบงชี้คือ ziouc ‘แลว แลวก็ และ’ คํา

เชื่อม cingx_daaih ‘ดังนั้น’ ตัวบอกหัวขอ aeqv ‘สวน’ และ อนุประโยควิเศษณที่มีคําวา liuz

‘หลังจากทํา’gau‘หลังจากที่’

ในการระบุเหตุการณดําเนินเรื่อง นักภาษาศาสตรเชิงปริชานใชวิธีวิเคราะหโครงสราง

เชิงมโนทัศนของการเคลื่อนไหวและลําดับเวลามีผลทําใหเกิดมโนภาพประจักษชัด ดังนั้น

เหตุการณดําเนินเรื่อง จึงเปนผลของการมองการเคลื่อนไหวและลําดับเวลาในเชิงมโนทัศน

xii

Page 14: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

TABLE OF CONTENTS

Acknowledgments..........................................................................................................v

ABSTRACT............................................................................................................... viii

List of figures........................................................................................................... xviii

List of Tables ............................................................................................................. xxi

List of Abbreviations and Symbols........................................................................... xxii

Chapter 1 Introduction ...................................................................................................1

1.0 Introduction..............................................................................................................1

1.1 Overview of the Thesis ............................................................................................1

1.2 Research Background ..............................................................................................2

1.3 Research Questions..................................................................................................3

1.4 Why Cognitive Linguistics? ....................................................................................4

1.5 Hypotheses...............................................................................................................6

1.6 Ethnic, Historical and Cultural Background............................................................7

1.7 Linguistic Background.............................................................................................8

1.8 Language Varieties ................................................................................................15

1.9 Orthography ...........................................................................................................16

1.10 Methodology........................................................................................................17

1.10.1 Steps in Method ..........................................................................................17

1.10.2 Data Collection and Verification ................................................................18

1.10.3 Summaries of the Narrative Texts ..............................................................19

1.10.4 “Cognitive Linguistics” as A Cover Term..................................................22

1.10.5 Analysis Procedure .....................................................................................23

1.11 Scope and Limitations of the Study .....................................................................24

Chapter 2 Literature Review........................................................................................27

2.0 Introduction............................................................................................................27

2.1 Theories of Discourse and Text Analysis in General ............................................27

2.1.1 Discourse Analysis in Tagmemic Theory.....................................................27

2.1.2 Meaning in Textlinguistics ...........................................................................28

2.1.3 Discourse Analysis in Generative Semantics ...............................................29

xiii

Page 15: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

2.1.4 Discourse Analysis in Functional-Typological Linguistics..........................30

2.1.5 Discourse Analysis in the Prague School .....................................................31

2.1.6 Discourse Analysis in Systemic-Functional Linguistics...............................32

2.1.7 Discourse Analysis in Functional Grammar (FG) ........................................33

2.1.8 Discourse Analysis in Relevance Theory (RT) ............................................34

2.1.9 Discourse Markers and Sociolinguistic Discourse Analysis ........................36

2.1.10 Narratology .................................................................................................37

2.1.11 Discourse Study by Thai authors ................................................................39

2.1.12 Discourse Semantics ...................................................................................41

2.1.13 Summary .....................................................................................................41

2.2 Cognitive Approaches to Discourse Analysis........................................................42

2.2.1 An Overall Picture of Cognitive Discourse Semantics.................................42

2.2.2 Knowledge Integration..................................................................................43

2.2.3 Information Management..............................................................................44

2.2.3.1 The Rhetorical Management of Discourse ..........................................44

2.2.3.2 The Thematic Management of Discourse ............................................45

2.2.3.3 The Referential Management of Discourse .........................................46

2.2.3.4 The Focus Management of Discourse..................................................47

2.2.4 Cognitive Functional Linguistics..................................................................48

2.3 Views on Meaning .................................................................................................48

2.3.1 Longacre’s view on meaning ........................................................................49

2.3.2 Langacker’s view on meaning ......................................................................50

2.4 Discourse Analysis in Cognitive Grammar ...........................................................52

2.5 Summary of the Chapter ........................................................................................55

Chapter 3 Some Grammatical Features of Iu-Mienh...................................................57

3.0 Introduction............................................................................................................57

3.1 Basic Clause Types ................................................................................................57

3.2 Serial Verb Constructions (SVC)...........................................................................59

3.3 Tense, Aspect and Modality ..................................................................................62

3.3.1 Tense .............................................................................................................62

3.3.2 Aspect ...........................................................................................................65

xiv

Page 16: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

3.3.3 Modality........................................................................................................71

3.4 Topic Chains ..........................................................................................................75

3.5 Summary of the Chapter ........................................................................................77

Chapter 4 Textlinguistics Storyline Analysis ..............................................................78

4.0 Introduction............................................................................................................78

4.1 Textlinguistics Storyline Analysis in Biblical Hebrew..........................................78

4.2 Definitions of Storyline: Two Approaches ............................................................80

4.2.1 Textlinguistic Definition of the Storyline .....................................................80

4.2.2 Cognitive Linguistic Definition of the Storyline ..........................................83

4.2.3 Summary .......................................................................................................85

4.3 Textlinguistics Storyline Analysis in Iu-Mienh Narrative.....................................86

4.3.1. Profile Analysis............................................................................................86

4.3.2 An Etic Salience Scheme Template..............................................................92

4.3.3 Iu-Mienh Salience Scheme Analysis ............................................................94

4.3.3.1 Band 8: Cohesive .................................................................................95

4.3.3.2 Band 7: Evaluations .............................................................................98

4.3.3.3 Band 6: Irrealis and Modals.................................................................99

4.3.3.4 Band 5: Settings ...................................................................................99

4.3.3.5 Band 4: Flashback..............................................................................102

4.3.3.6 Band 3: Backgrounded Activities ......................................................103

4.3.3.7 Band 2: Backgrounded Action/Events...............................................104

4.3.3.8 Band 1: Primary Storyline .................................................................105

4.3.3.9 Band 1′: Pivotal Storyline ..................................................................108

4.4 Finding Narrative Timeline through Temporal Movement .................................111

4.4.1 Methodological Consideration....................................................................111

4.4.2 Non-Verb-Morphology Dependent Method ...............................................112

4.5 Summary of the Chapter ......................................................................................113

Chapter 5 Cognitive Linguistics Storyline Analysis..................................................115

5.0 Introduction..........................................................................................................115

5.1 Placing A Narrative in the Cognitive Linguistics Setting....................................116

5.2 Salience Scheme as a Prototype Effect ................................................................118

xv

Page 17: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

5.2.1 Primary Storyline Band...............................................................................118

5.2.1.1 Unmarked Verbs ................................................................................118

5.2.1.2 A Gradient within Band 1 ..................................................................119

5.2.2 Pivotal Storyline..........................................................................................128

5.2.2.1 Grounding of a Negated Construction ...............................................128

5.2.2.2 Grounding and Interaction between Band 1 and others.....................131

5.2.3 Setting Band................................................................................................134

5.2.4 Summary .....................................................................................................135

5.3 Foregroundedness through Transitivity ...............................................................135

5.3.1 Energy Transitivity in the Billiard-Ball Model...........................................135

5.3.2 Energy Transition Rate in Action Chain.....................................................138

5.3.3 Summary .....................................................................................................141

5.4 Foregroundedness through Sequentiality.............................................................141

5.4.1 Sequential Markers .....................................................................................142

5.4.1.1 Sequential Marker: Adverb or Conjunction?.....................................143

5.4.1.2 Conjunction Cingx_daaih ..................................................................146

5.4.1.3 Summary of the Sequential Markers..................................................148

5.4.2 Role of Some Cohesive Band Materials for Sequentiality .........................148

5.4.2.1 Conceptual Structure of Adverbial Clauses.......................................149

5.4.2.2 Topic Marker and Incrementation of CDS ........................................157

5.4.2.3 Adverbial Clauses and Incrementation of CDS.................................159

5.4.2.4 Summary ............................................................................................161

5.5 Summary of the Chapter ......................................................................................161

Chapter 6 Conclusion.................................................................................................164

6.0 Introduction..........................................................................................................164

6.1 Summary of Findings...........................................................................................164

6.2 Implications of the Findings to the Hypotheses...................................................167

6.3 Suggested Areas of Further Study .......................................................................168

Appendix A................................................................................................................171

1. Yao or Mien? .........................................................................................................171

2. Linguistic Classification ........................................................................................172

xvi

Page 18: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

Appendix B ................................................................................................................175

A Historical and Cultural Background ......................................................................175

1. A Historical Background .......................................................................................175

2. The Emperor’s Charter ..........................................................................................177

3. The Enigma of the Migration Routes.....................................................................178

4. A Historical Background to the Religion...............................................................180

5. Migration to Thailand ............................................................................................181

6. A Cultural Background ..........................................................................................182

Appendix C ................................................................................................................184

Appendix D................................................................................................................186

Bibliography ..............................................................................................................322

xvii

Page 19: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

LIST OF FIGURES

Figure 1. The place of Hmong-Mien family posited by Benedict (1976:29) ................9

Figure 2. A place of Hmong-Mien family posited by Chinese scholars........................9

Figure 3. The Hmong-Mien family..............................................................................10

Figure 4. The subgroups of Mienic languages.............................................................11

Figure 5. Four Mienic “dialects” in China...................................................................12

Figure 6. Provinces and major districts of Iu-Mienh settlement in the northern

Vietnam.......................................................................................................12

Figure 7. Areas of Iu-Mienh distribution in Vietnam and Lao PDR ...........................13

Figure 8. Iu-Mienh settlements in Thailand and Laos .................................................14

Figure 9. Illustrations to “Cicada Story” and “Faam-Bae’ and Aeng-Doi”.................21

Figure 10. A relationship of Cognitive Linguistics with other disciplines (Tsuji

2002:183) ....................................................................................................23

Figure 11. A blueprint for the speaker .........................................................................43

Figure 12. Signifié-signifiant composit of a sign .........................................................50

Figure 13. The internal and syntagmatic relation of linguistic signs...........................51

Figure 14. Aspects of the usage event (Langacker 2001b:145)...................................53

Figure 15. Polysemous relation within the change-of- state aspectual particle aqv ....69

Figure 16. Composition of the resultative aspectual marker -mi’aqv ..........................71

Figure 17. Basic Epistemic Model (Langacker 1991c:242) ........................................73

Figure 18. The parameters for salience scheme for English narrative (Longacre

1996:26) ......................................................................................................82

Figure 19. A profile of Story 1: A story of an old husband and bamboo shoots .........87

Figure 20. A profile of Story 2: A story of firefly and grasshopper ............................87

xviii

Page 20: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

Figure 21. A profile of Story 3: A story of three piglets..............................................88

Figure 22. A profile of Story 4: Why cicadas don’t have intestine .............................89

Figure 23. A profile of Story 5: A story of Faam-Bae’ and Aeng-Doi........................90

Figure 24. A profile of Story 6: A story of Aahan.......................................................91

Figure 25. A profile of a story of big snake.................................................................92

Figure 26. An etic salience scheme for narrative (Longacre 1996:28)........................93

Figure 27. An etic salience scheme for Iu-Mienh narrative.........................................93

Figure 28. Salience scheme for Iu-Mienh narrative.....................................................95

Figure 29. A stage model ...........................................................................................117

Figure 30. A narrative in the stage model..................................................................117

Figure 31. A gradient of transitivity of gorngv ‘speak’ in contrast to borqv ‘beat’ ..127

Figure 32. The figure-ground relation in the case of affirmative stentence

(Takahashi and Kingkarn 1997:278).........................................................130

Figure 33. The figure-ground relation in the case of negative sentence (Takahashi

and Kingkarn 1997:278) ...........................................................................130

Figure 34. The static and moving objects ..................................................................136

Figure 35. Action chain..............................................................................................136

Figure 36. The conceptual structure of action increase in SVCs and topic chains....138

Figure 37. Energy transitivity ratio............................................................................139

Figure 38. Scattered and condensed objects ..............................................................142

Figure 39. Connectives ..............................................................................................144

Figure 40. Prospective conjunction ziouc ..................................................................145

Figure 41. Retrospective conjunction cingx_daaih ...................................................147

Figure 42. Functions of ziouc and cingx_daaih .........................................................147

xix

Page 21: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

Figure 43. Tail-head linkage as an adverbial clause ..................................................150

Figure 44. Ground and Figure in paragraph linkage..................................................152

Figure 45. Multi-layered topic-comment structure (Court 1986:36) .........................153

Figure 46. Multi-layered figure-ground relation........................................................153

Figure 47. Rightward focus structure.........................................................................156

Figure 48. Topic Marker aeqv ...................................................................................158

Figure 49. CDS increment through adverbial clause .................................................160

Figure A. Chinese tonal influence on Hmong-Mien..................................................173

Figure B. Origine and Migration Route of the Proto-Mienic People ........................178

Figure C. The Crossing of the Sea .............................................................................179

xx

Page 22: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

LIST OF TABLES

Table 1. Distribution and Population of the Iu-Mienh in Ten Provinces of

Thailand ......................................................................................................15

Table 2. Sample lexical and phonological differences between regional varieties .....16

Table 3. Titles and types of seven stories ....................................................................19

Table 4. Temporal adverbs for narrative discourse .....................................................64

Table 5. Aspectual auxiliary verbs...............................................................................66

Table 6. Transitivity parameters (Hopper and Thompson 1980:252)..........................81

Table 7. A gradient relation within the equative constructions .................................134

Table 8. The definitions of ziouc ...............................................................................144

Table A. Four Iu-Mienh groups migrated into Thailand ...........................................181

Table C1. Consonants phonemes and their corresponding orthography ...................184

Table C2. Vowel phonemes and their corresponding orthography ...........................185

Table C3. Vowel glides and their corresponding orthography..................................185

Table C4. Tones and their corresponding orthography .............................................185

xxi

Page 23: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

LIST OF ABBREVIATIONS AND SYMBOLS

1 first person

2 second person

3 third person

1sg 1st person singular

2sg 2nd person singular

3sg 3rd person singular

1pl 1st person plural

2pl 2nd person plural

3pl 3rd person plural

adv adverb

ADVBLZR adverbializer

adv.loc locative adverb

adv.tmp temporal adverb

AFM affirmative

ASST assertive

CAS.COM casual command

CDS Current Discourse Space

CG Cognitive Grammar

CL Cognitive Linguistics

Cl Clause

CLF classifier

cmpr comparative

xxii

Page 24: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

conj conjunction

contr contraction form

CS Cognitive Semantics

DPCL dependent clause marker

DS Direct Speech

GOAL goal or goal marker

idntf identifier

idm idiom

INCHO inchoative aspect

intj interjection

MIM mimetic word

MST Mental Space Theory

NEG negation

NEG.CMD negative command

NOM nominalizer

NP noun phrase

n.prp proper noun

ONOMA onomatopoeia

part particle

part.asp aspectual particle

part.f sentence final particle

part.mod modality particle

part.q question particle

part.voc vocative particle

xxiii

Page 25: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

PL plural

POSS possessor/possessive

PST past

PRS present

PRS.ST present state aspect

PRG progressive

pn pronoun

pn.emph emphatic pronoun

pn.p personal pronoun

q question

q.cont content question

REC.ACT reciprocal action

REPORT reported speech marker

RELT relative marker

RSLT resultative aspect

SFT.CMD soft command

SML.ACT simultaneous action

SOV Subject Object Verb

St sentence

SUSPD suspend (i.e. holding a floor of speech or taking pause, serving as a

dependent clause marker)

SVC serial verb construction

TAM tense, aspect, modality

TOP topic or topic marker

xxiv

Page 26: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

v verb

v.asp aspectual verb

v.aux auxiliary verb

v.cmp compound verb

v.dr directional verb

vi intransitive verb

v.mod modality verb

VP verb phrase

v.st stative verb

vt transitive verb

xxv

Page 27: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

CHAPTER 1

INTRODUCTION

1.0 Introduction

The present study focuses on the issue of foreground and background with reference

to storyline, within the conceptual structures of storyline in Iu-Mienh 1 narrative

discourse. A semantic approach to narrative discourse, especially in storyline and

salience scheme analysis, is pursued in this research.

In explaining the salience scheme of verbs and clauses in relation to the mainline, or

storyline, of Thai narrative, Somsonge (1991a:113) affirms the validity of a

psychological perspective in ranking them with the following example:

Reality is perceived through our cognitive experience or our approach to the world. For example, when we look into the field with nothing in it, the field itself is salient. However, if there is a herd of cows eating grass in the field, the cows are in focus and the field becomes a background. Suppose that a parachute is falling from a plane into the field. Now the parachute is salient and the cows are backgrounded. The field is further outranked.

This is probably an original insight applying Ronald W. Langacker’s concept of

Cognitive Grammar (e.g. 1987a, 1991b, 1991c, 2000) to Robert E. Longacre’s theory

of storyline (e.g. 1981, 1987, 1989a, 2003a) in discourse analysis.

1.1 Overview of the Thesis

Chapter 1 will set out research background, linguistic background, hypotheses, goal,

methodology and scope of this study. Chapter 2 will survey relevant literature in the

domain of discourse analysis and text linguistics in search of the approaches of

various linguists to the semantics of discourse. Chapter 3 will describe some selected

features of Iu-Mienh grammar limiting its scope to the minimum necessity for

storyline analysis. Chapter 4 will present the textlinguistic storyline analysis of Iu-

1 The name of the people “Iu-Mienh” [iu31 mjen31] is sometimes shortened to “Mienh” (in their orthography) or

“Mien” (in Romanized transliteration) [mjen31]. In their own use, [mjen31] refers to both the people and their language, but if the latter needs to be distinguished from the former, “Mienh waac” [mjen31 wa˘11] is used to refer to ‘Mien language.’ Henceforth the Iu-Mienh orthography is used to cite words. See section for the orthography (1.3) and IPA equivelants in Appendix C.

1

Page 28: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

2

Mienh and the salience scheme analysis. The chapter will also compare the definition

of the storyline with its associated concept of foreground from two perspectives; that

is, textlinguistics and Cognitive Linguistics. Setting the subject matter of the present

study, narrative discourse, on the cognitive linguistics basis, chapter 5 will investigate

some selected linguistic constructions that are used in relation to the storyline in Iu-

Mienh to identify their conceptual structures. This final chapter has three main

sections. The first section will discuss (i) that the storyline verbs are unmarked, (ii)

that some grounding elements (in CG term) are used in the pivotal storyline and (iii)

also discuss a gradient of materials in the salience scheme. The second will pick up

the serial verb constructions (SVCs) and topic chains (topic-subject) to investigate

their conceptual structure from a CG perspective. The last section will explain some

selected constructions both from the storyline and cohesive materials to investigate

their conceptual structure in terms of CG. Chapter 5 as a whole argues that

transitivity and sequentialty of events in the narrative are the major factors that make

the storyline perceptively foregrounded in Iu-Mienh. In contrast to chapter 4, which

is descriptive, chapter 5 is explanatory.

1.2 Research Background

Sitting around a fire at night or in the shade under trees during the day, the Iu-Mienh

people enjoy listening to story after story. The storytellers have learned the stories

from their grandparents, parents or friends and are likely to pass them on to the next

generations. Details of the stories may vary from region to region or from narrator to

narrator but their main lines, or storylines, tend to remain the same. The storyline

seems to be remembered well in the minds of storytellers and listeners. What is

actually going on in the minds of a storyteller and their audience?

While tools for analyzing narrative discourse, including identifying storylines as a

field linguistic method with the emphasis on surface structure, are now abundantly

available, an investigation into the meaning of storyline reflected in the conceptual

structure has not been pursued. What kind of discourse analysis can be useful to

reveal such an aspect? Does an investigation of the meaning in a speaker’s mind pose

the danger that an investigator may become subjective, separated from the objective

Page 29: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

3

surface structure of a discourse? After an empirical description of discourse features,

could we not go on to the study of what they mean? This study is an investigation of

such a bundle of questions against the background of the Iu-Mienh lovers of

storytelling.

1.3 Research Questions

With this general background in mind, these inquiries are crystallized into three

research questions.

Storyline is the mainline of narrative discourse (Longacre 1996:21). There is a

difference between storyline and supportive materials, or non-storyline. In narrative

discourse, the former is characterized by past tense (or preterite), the latter by aspect

(e.g. imperfect, durative, descriptive) and modality (modal and irrealis). Hopper

(1979) has proposed that the distinction is binary, referring to the former as

“foreground” and the latter as “background.” This view is endorsed by Niccacci

(1986, English translation 1990) in accordance with the theory of Vordergrund-

Hintergrund (foreground-background) by W. Schneider (1974) and H. Weinrich

(1964),2 who had developed it before American scholars began to use it. On the other

hand, Longacre proposes a gradient of proximity to the storyline as exemplified in his

metaphor of a “spectrum” (1981), which was later described as a “cluster concept” by

Dry (1992:441). So the first research question is: (1) Is the relationship between

storyline and supportive materials (i.e. non-storyline) binary (i.e. foreground vs.

background) or a gradient?

Second, admitting that there is such a distinction between storyline and supportive

materials, whether it is binary or a gradient, how do languages which do not have

morphological inflection on verbs (e.g. past, perfect, imperfect) like Iu-Mienh

identify the storyline? This question might be paraphrased to a wider one on the

2 Schneider’s Grammatik des biblischen Hebräisch (Munich: Claudius, 1974) is an application of Weinrich’s

Tempus. Besprochene und erzählte Welt (Stuttgart, 1964), (which is translated, Tense: The world of discourse and the world of narrative) to Biblical Hebrew. Unfortunately, I do not have access to these works yet, hence they are not found in the bibliography of this thesis. A summary of the two authors is found in chapter 1 of Niccacci 1986. Talstra 1992 gives a detailed assessment of Schneider and Weinrich applying it to texts of 1 Kings 2 and Deutronomy 7.

Page 30: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

4

assumption that the hypothesized answer to the first is that the relation between

storyline and non-storyline is a gradient: (2) What linguistic expressions profile

storyline (in Cognitive Linguistic terms)? On one hand, a simpler question in

descriptive linguistic terms would be “What linguistic expressions carry a storyline in

Iu-Mienh narrative discourse?” On the other hand, the question in Cognitive

Linguistic terms, presupposes that a cline from the most dynamic material (viz.

prototype) on the storyline to the most static supportive material (viz. extensions from

prototype) can be better treated through a Cognitive Linguistic approach (e.g. Taylor

2003).

Thirdly, if the concept of the cline or gradient relationship between storyline and the

supportive materials is valid, then the next question should be asked: (3) What are the

conceptual structures of these linguistic expressions in the storyline? We assume that

a sequentiality of events expressed mainly by cohesive clauses, retrospective and

prospective conjunctions, and marginal clause particles on one hand and transitivity

of events expressed mainly by unmarked verbs, aspectual auxiliaries, aspectual

particles, adverbs, and the composite of serial verb constructions and topic chains on

the other hand collaborate to indicate the storyline. In investigating the conceptual

structures of storyline, these two areas have to be analyzed: sequentiality and

transitivity of events. Following Langacker, who argues that meaning is equated with

the conceptual structure (1991a:278), it is hoped that an attempt to answer these

questions will identify the conceptual structure of storyline in Iu-Mienh narrative

discourse.

1.4 Why Cognitive Linguistics?

With regard to the case for a use of Cognitive Linguistics (hereinafter CL) in this

study, two basic reasons are employed: (1) because of the inseparability of form-and-

meaning, and (2) because of continua across grammatical categories. Firstly, whilst

the cautious attitude about subjectivity in the study of meaning has driven a

generation of linguists to a formal study of language focusing on surface structure,

semantics at the expense of form can safely be avoided as Langacker (1987a:46)

asserts that “cognitive grammar claims that grammatical structure is almost entirely

Page 31: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

5

overt.” He goes on to say, “Surface grammatical form does not conceal a ‘truer,’

deeper level of grammatical organization; rather, it itself embodies the conventional

means a language employs for the structuring and symbolization of semantic content”

(Langacker 1987a:46-7). Our position, therefore, is to believe that CL is a useful tool

for the investigation of meaning through surface grammatical form without divorcing

one from the other.

Secondly, even from among non-strict followers of leading cognitive linguists (e.g.

Langacker or Lakoff 1972, 1987), more and more linguists whose major concern is

description are beginning to recognize the fuzziness and continua that run across

linguistic categories, such as the continuum between verb and adjective (cf. Dixon

1977) or between verb and preposition (e.g. Yip and Rimmington 1997:114-9). With

respect to the latter case, Clark (1989:190) refers a verb that resembles a preposition

as a “locus verb,” citing an example of ‘to go’ which is grammaticalized to a locus

preposition ‘to’ in Khmer.

Further, one can observe a similar attitude among non-cognitive linguists. For

example, Somsonge’s metaphor of the descending parachute into a herd of cows

grazing in a static field cited at the outset of this chapter can well be analogous to a

continuum of a prototypical transitive verb (i.e. figure or trajector) on storyline, a

midway participle, and a setting described in a prototypical stative verb of existence

(i.e. ground or landmark) for her gestalt-like account of verb ranking (Somsonge

1991a:113). Dry (1992:441) also suggests a treatment of foreground as “the cluster

of features which mark transitivity.” This cluster concept enables characterizing the

foreground in terms of prototypical transitivity and its gradual deviations (i.e.

continuum) within the category. Furthermore, we should be reminded that the

terminology Longacre used as early as 1981 was, to begin with, “spectrum”, which is

an optic term implying a gradual change through various linguistic expressions in

relation to storyline. One of his Two hypotheses (1989b:414) contains a phrase

“progressive degrees of departure from the main line” (i.e. a continuum, again). The

last example is that the term “salience scheme” (Longacre 1996:27) suggests the

treatment of constructions placed between the storyline and the farthest band can be

Page 32: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

6

“schematic”, which is a favorite term among cognitive linguists. All these notions in

relation to the storyline as developed by Longacre have a good possibility of

redefining them in terms of CL.

For these two reasons, we will attempt to analyze storyline from a CL perspective.

1.5 Hypotheses

Bearing these research questions in mind, this thesis will argue for the following two

hypotheses (designated H1 and H2):

H1: The relationship between storyline and non-storyline in Iu-Mienh is a gradient.

H2: The storyline in Iu-Mienh is characterized both semantically by transitivity and

sequentiality of events and syntactically also by various linguistic constructions.

Annotations on these two hypotheses may be needed. H1 is stimulated by an

application of Longacre’s storyline theory to Thai by Somsonge (1990b, 1991a,

1992b:419-33) as cited in 1.0. In contrast to Longacre and Somsonge, one

representative from those who have a binary understanding would be Givón (1984):

Foreground vs. background: “In connected discourse, some aspects of the description—coded in some sentences/clauses—are considered the gist, backbone, main line of the episode/description/communication. They are the foreground of the discourse. Others are considered satellites, side-trips, supportive portions of the description/episode/communication. Those are the background portions of the discourse”. (Givón 1984:287-8) [Quotation marks, italics, and bold faces are as in the original.]

Concerning H2 related to the issues of languages without verbal inflection, Somsonge

(1990b) writes:

The study of bipartite [viz. binary] structure of discourse information in Thai reveals that in language without verbal inflection as Thai, the bipartite structure is not expressed solely by the verb system as in English but by a conspiracy of non-systemic ways which include types of verbs, adverbs, time phrases, sequential signals, temporal clauses/phrases/words, auxiliaries, pre-serial verbs, and post-serial verbs. (Somsonge 1990b:76) [Underline added]

In a very similar way, Iu-Mienh exhibits both the difficulty of dependency on the

verbs in finding the storyline and the benefit of utilizing other contextual linguistic

Page 33: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

7

expressions. These constructions are categorized in two groups: sequentiality and

transitivity of events.

For Thai narrative, proposing four categories of storyline verbs (i.e. cognitive

experience, events proper, motion verbs, and action verbs), Somsonge cautions about

an exclusive dependency on them in identification of the storyline and suggests that

“[t]he most important criterion” is sequentiality (1991a:106). Following her, the

present H2 presupposes that the storyline in Iu-Mienh is composed of two major

factors: sequentiality of events and transitivity of events or, in simpler terms,

continuation and movement of events. Sequentiality is profiled by materials such as

cohesive adverbial phrases/clauses, lexical cohesion through tail-head linkage,

retrospective/prospective connectives, and marginal clause particles. Transitivity is

encoded by unmarked verbs, global grounding elements (aspectual sentence level

particles), local grounding elements, other verbs marked by auxiliaries/adverbs,

aspectual verbs, and serial verb constructions and topic chains. Each group has

clusters of materials/constructions. It is as if an intertwined cable of the two groups

composed of clusters of the materials creates a storyline, and not that a particular verb

form is responsible for that storyline.

In summary, it is hypothesized that the storyline is characterized as a gradient with

reference to verb and clause ranking as opposed to the binary concept of foreground

vs. background. As such, a prototype analysis of gradual order from types of Iu-

Mienh verbs, aspectual constructions, aspectual particles, five kinds of copulatives to

marginal cohesive materials is hypothesized to identify a conceptual structure of

storyline.

1.6 Ethnic, Historical and Cultural Background

This study has adopted the name Hmong-Mien,3 rather than formerly used “Miao-

Yao”4 language family. Hence, the name of the people for this thesis is Iu-Mienh, not

3 Ratliff (1992a: 17) says ‘“Hmong” is better than “Miao,” as it does not have any of the derogatory associations

connected with the Southeast Asian form of “Miao”, “Meo.”’ In a similar way, Iu-Mienh want themselves to be referred to as “Mienh” / mien / ‘human’, not “Yao” which, in Thai, sounds like “to tease/jest” as in [jaojk] (เยัาหยอก) or [krasaojaoj] (กระเซาเยัาแหย). In fact “Iu-Mienh”, which is a proper name of the people in

Page 34: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

8

Yao. See Appendix A on confusion concerning the names of the Iu-Mienh people

and a recent consensus about linguistic classification among linguists outside China.

Iu-Mienh historical and cultural aspects are found in Appendix B.

1.7 Linguistic Background

Iu-Mienh belongs to the Hmong-Mien language family. As to the wider affiliation of

the Hmong-Mien language family, the debate concerning whether it belongs to Sino-

Tibetan (mostly by Chinese scholars) or Benedict’s Austro-Thai stock (1975; 1976:1-

36), or the Austro-Asian family assumed by Schliesinger (1998:130), or

Haudricourt’s (1966:56) denial of the Miao-Yao’s contact with that family,5 has not

been settled yet. The place of “Miao-Yao” (viz. Hmong-Mien) under Austro-Thai

posited by Benedict is presented in Figure 1.

their language, is even better in order to avoid mispronunciation (by non-Iu-Mienh speakers) [ mien ] ‘spirits/ghost.’

4 The term “Hmong-Mien” is now more popular than “Miao-Yao,” which seems to be obsolete at least among non-Chinese linguists. For example, Matisoff, at the special lecture held in Chulalongkorn University 10-26 January, 2005 in Bangkok, together with other Thai linguists who were present, never used “Miao-Yao” when talking about typology of South East Asian languages.

5 Haudricourt does so on the basis of Hmong-Mien’s late contact with Tai-Kadai languages, saying, “The Miao-Yao [Hmong-Mien] languages are not in contact with Austroasiatic languages, they are separated from them by the Tibeto-Burman languages, Thai and Kadai. The arrival of the Yao (Man) peoples in Vietnam dates only from the seventeenth century, and that of the Miao (Meo) from the nineteenth” (1966:56).

Page 35: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

9 Austro-Thai

Miao-Yao Kadai [= Hmong-Mien] Austronesian AT-sub. Austroasiatic

Figure 1. The place of the Hmong-Mien family posited by

Benedict (1976:29)

Chinese linguists and anthropologists in mainland China refer to Sino-Tibetan as the

Hàn-Zàng6 phylum in which Hmong-Mien has been considered to be included, as

shown in Matisoff’s summary of the Chinese version (1983:68) in Figure 2.

Hàn-Zàng yŭxì (The Sino-Tibetan)

Chinese-Tai Tibeto-Burman Miao-Yao (Hmong-Mien)

Chinese Tai Miao Yao

(Hmong) (Mien)

(Adapted from Matisoff 1983:68. Dotted lines indicate that the details of the group members are omitted for the sake of simplicity.)

Figure 2. A place of Hmong-Mien family posited by

Chinese scholars

Here in Thailand, while the Tribal Research Institute (Mongkol and Thaworn (eds.)

1995:3, 25) in Chiang Mai and The Inter Mountain Peoples Education and Culture in

Thailand Association (IMPECT Association 2002) adopt Benedict’s hypothesis (also

Schliesinger 2003:272), others like Suriya (1988:4) of Mahidol University are more

careful in their classification. Matisoff (1983:70-1) cautions against a rush

6 Hàn-Zàng (汉藏) is a compound of hàn (汉) ‘the dominant ethnic group in China’ and zàng in Xīzàng (西藏)

‘Tibet,’ hence, the Sino-Tibetan phylum is called Hàn-Zàng yŭxì (汉藏语系).

Page 36: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

10

conclusion, and Clark (1989:175) summarizes, “Wider genetic relationships are

undetermined.”

Figure 3 shows Matisoff’s presentation of the Hmong-Mien family (2001:299). What

he calls “Chiengrai Yao” under “Iu Mien” in Figure 3 is the subject of the present

study.

Paipai [= Dzau Min, or Yau Min]

Lingzhun Proto-Mien Kim-Mun Haininh Daiban

Iu Mien Xing-an Chiengrai Yao

Yongcong

Patengic Pateng

Proto-Hmong-Mien Rongjiang Phö Eastern Guizhou Zhengfeng Taijiang Lushan Kaili Daigong Proto-Hmong Guizhou

Central Hmong Longli Huajie

Chuan Shchuan Suyong -Guizhou [=Magpie -Yunnan Miao]

Green Hmong [= Blue

Western Hmong Hmong] Petchabun

[= White Hmong]

Guangshu

[=YiMiao] Weining Libo Northern Hmong [= West Hunan]

(Slightly altered from Matisoff 2001:299 without a content change)

Figure 3. The Hmong-Mien family

Page 37: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

11

Subgroups of Mienic languages as suggested by Ratliff (1992a:19) are shown in

Figure 4. The “Ho Nte” is an alternative name for Shē.

HMONG-MIEN

MIENIC HO NTE HMONGIC (Shē)

Mien-Mun Min-Meng

Iu Mien Kim Mun Biao Mon Dzao Min Biao Min Chao Kong Meng (Iu-Mienh) (Mun) (= Paipai

or Ba Pai, = Yau Min)

(Adapted from Ratliff 1992a:19)

Figure 4. The subgroups of Mienic languages

Matisoff’s classification is based on place names and pays more attention to the

Hmongic side of the tree. Ratliff presents more detailed members under the Mienic

side (besides her main discussion on the Hmong tones). Matisoff’s “Paipai” Yao or

“Bá Pái Yáo” (八排瑶) (Zee 1991:71-86) corresponds with Ratliff’s Dzao Min, also

called Yau Min (邀敏) (Pan 1991:47). Other than four “dialects” studied by Mao

(2004) (viz. Iu Mien (优勉), Kim Mun (金门), Biao Mon (标曼), and Dzau Min (藻敏

)), Figure 4 also includes Biao Min (标敏) and Chao Kong Meng under the Min-Meng

group.

The name of the people is ‘Iu-Mienh’ [iu mien] in their language, with a hyphen

to indicate tone sandhi (a change from the original mid-level tone [iu] to falling

tone), and with the word-final h to indicate the falling tone. Though Iu-Mienh

themselves distinguish the name of their ethnic group from the name of their

language (Mienh waac [mien wa] ‘Mienh language’), the term Iu-Mienh is used to

refer to both in this study.

Page 38: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

12

Iu-Mienh is spoken in Hunan, Guangtong, Guangxi, Guizhou, and Yunnan provinces

of China (Figure 5), the northern provinces of Vietnam (Ha Giang, Cao Bang, Tuyen

Quang, Lao Cai, Xiang Khoang, Luang Nam Tha, Bokeo, Yen Bai, Lai Chau, Lang

(Lemoine and Chiao 1991:46. Note that the Kim Mun are also on the island of Hainan though the map does not include it.)

Figure 5. Four Mienic “dialects” in China

Nan Ning Thõ

Bình Lư

Vị Xuyên

uyênBìng

Chiê Hóa

ên

A

Phong

Sìn Hô

Lộc Bình

QuỳnhNhai

S

Figure 6. P

se

Bắ Quang Hàm Y

rovinces and majo

ttlement in the nor

Ng Chơ Rãm

Đình Lâp

BìnhLiêu

VIỆ

r districts of Iu-Mienh

thern Vietnam

TiênYên

CHIN

T NAM

LAO
Page 39: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

13

Son, Thai Nguyen, Son La, Hoa Binh, Phu Tho, Bac Giang, Thanh Hoa) (Figure 6)

and Lao PDR (in the provinces of Hua Phan, Luang Prabang, Phongsali, Sayaburi,

and Vientiane) (Figure 7), and the North Thailand provinces (Figure 8). The

distribution of the speakers of Iu-Mienh in Vietnam and Lao PDR is shown in Figure

7. It should be remembered that in many of these areas cohabitation with the Kim

Mun is common (cf. Figure 5).

Figure 7. Areas of Iu-Mienh distribution in Vietnam and

Lao PDR

It is estimated that there are over 700,000 Iu-Mienh in China (Gordon 2005), 474,000

in Vietnam7 (Dang, Son, and Hung 2000:183, following Schliesinger 1998:130), and

20,250 (Gordon 2005) in Lao PDR (Figure 11). In Thailand, the Tribal Research

Institute8 reported that 40,371 Iu-Mienh lived in eight provinces of Chiang Rai,

7 Khong (2002:172) lists 620,538 including Kim Mun but the sum of the detailed population by province is

448,047 (2002:52). 8 This Institute has recently been dissolved and research work is now conducted under the direction of the

Research Institute of Sociology, Chiang Mai University and the Tribal Museum in Chiang Mai. The Institute’s first work on Iu-Mienh reports their migration to the city of Chiang Mai due to economic, social structure, and educational pressures, and cultural changes of their life, as argued by a Iu-Mienh researcher Mr. Yangyon and others (Prasit, Yangyon, Wisut 2004).

Page 40: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

14

Phayao, Nan, Chiang Mai, Lampang, Sukhothai, Kamphaeng Phet, and Tak provinces

in 1995 (Mongkol 1995:25).

Figure 8. Iu-Mienh settlements in Thailand and Laos

More recently (2002), the Department of Social and Welfare Development, Ministry

of Social Development and Human Security of the Kingdom of Thailand has reported

45,571 in ten provinces having added two more provinces, Kanchanaburi and

Phechabun (Table 1).

Page 41: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

15

Population Province Vil-

lage House-hold Family

Male Female Boy Girl Total

Chiang Rai 61 2,193 2,450 4,990 5,258 2,538 2,439 15,225

Nan 34 1,429 1,907 3,538 3,467 1,904 1,775 10,684

Phayao 29 1,326 1,530 3,218 3,120 1,354 1,273 8,965 Lampang 25 838 962 1,643 1,678 843 781 4,945

Kamphaengphet 15 529 594 1,059 1,079 423 429 2,990

Chiang Mai 5 247 250 395 392 271 295 1,353

Sukhothai 6 96 111 221 236 119 107 673

Tak 1 63 69 98 1l2 108 112 420

Kanchanaburi 1 20 20 78 78 31 25 212

Phechabun 1 17 19 30 32 18 24 104

Total 178 6,758 8,022 15,260 15,442 7,609 7,260 45,571

(Department of Social and Welfare Development, Ministry of Social Development and Human Security 2002)

Table 1. Distribution and Population of the Iu-Mienh in Ten

Provinces of Thailand

The trend of migration from hills to plains and cities are reported in Jonsson (1996),

Kimura (1997), and Prasit et al. (2004). About 23,000 Iu-Mienh speakers currently

reside in immigrant communities on the west coast of the U.S.A. and also some 2,000

in France. The total population of Iu-Mienh speakers in all mentioned countries,

estimated by Pourret (2002:12), is 1,600,000.

1.8 Language Varieties

Court (1991:149) observes, “the [Iu-]Mien[h] of Thailand speak a dialect virtually

identical with the [Iu-]Mien[h] of Laos, and very close to the dialect of most of the

same ethnic group in China and Vietnam.” As to the relationship between Thailand

Iu-Mienh and Laotian Iu-Mienh, his observation may well be right since the

migration route was from south China, through Vietnam and Laos, to Thailand (Chob

1997). Within Thailand the alleged difference between the Chiang Rai variety

(spoken in Maechan district, Chiang Mai, Kamphaeng Phet, and Tak) and the Chiang

Kham variety (Phayao, Nan, Lampang, Sukhothai) was discredited by Theeraphan

(1988b, 1989) on the basis of the tone system analysis. They are essentially the same

Page 42: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

16

as far as the tones are concerned. The native speakers, however, are conscious of

some lexical differences as shown in Table 2.

Chiang Rai variety Chiang Kham variety

pumpkin fuqc nyomv / fuü ¯om / famh mbouh / fm bou /

cabbage lai-dorngc / li tü / lai-jaaix dorngc

/ li c˘iÄ tü /

jackfruit da’ norc nih / t nü ni / ma’nun / m nun /

to play nyienx / ¯enÄ / jiuv / ciu /

to water (plant)

fuqv wuom / fu um / pietv wuom / pet um /

gaeng-waen / k wn / cicada gaeng-waen / k wn / (variant)

gaeng-nzen / k zen /

biouv-gomh / pou kom / biouv-gomh / pou kom /

tomato (variant)

loz-laangz biouv / lo l˘/ pou / (variant) muangz hor biouv

/ mw/ h pou /

Table 2. Sample lexical differences between regional

varieties

1.9 Orthography

The orthography used in this study is the Unified Script (a pinyin-like Roman-based

script, which is sometimes called the New Roman Script)9 ratified at the orthography

conference held in Ruyuan County, Guangdong, with the Iu-Mienh delegates from

China and the U. S. A., in 1984 (Purnell 1985).10 The tones are indicated by the

letters suffixed to words except for the mid-level tone, which is unmarked.

Some diacritics are also used in the Iu-Mienh orthography. The hyphen is used for an

indication of tone sandhi (e.g. Iu-Mienh), and syllable breaks of some proper names

9 As opposed to the New Roman Script, there is also the Old Roman Script devised by Cox and Smalley in the

1950s, which is more phonetic than phonemic. See Purnell 1985. On the Thai-based Mienh orthography, see Callaway and Callaway 1976.

10 This orthography is used in Mao’s dictionary (1992), Panh’s dictionary (1995), Aumann and Bienh’s dictionary (2002), Jennings’ literacy lessons (1998), and other literature. Court 1986 and Pán and Shū 1988 also use this orthography but differ slightly in that the former uses a few different vowel letters, the latter some different diacritics.

Page 43: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

17

(e.g. longer than two syllables) are shown by “^” as in A^me^li^ka “America.” Every

syllable is written separately reflecting the monosyllabic tendency of this language.

Within this study, however, the following interlinearization will make use of an

underscore “_” to connect syllables which form one lexical item; e.g. gamh_nziex

“fear” or ninh_mbuo = 3rd person-pronoun_plural-marker, “they.”

A summary of Iu-Mienh phonology 11 (Purnell 1965) is given in Appendix C

concerning 33 consonants, 9 vowels, 12 glides, and 6 tones, with the corresponding

Roman-based orthography (viz. the Unified Script).

1.10 Methodology

1.10.1 Steps in Method

A breakdown of the method into steps is as follows:

(1) Identify verbal and clause structures of Iu-Mienh, based on seven recorded,

transcribed, and interlinearized texts verified through monolingual field

interviews with language helpers/consultants.

(2) Review Longacre’s ten works on storyline analysis in Biblical Hebrew (1961,

1979a, 1981, 1982b, 1985a, 1987, 1989a, 1994a, 1995a, 2003a.), largely

through which, among others, he constructed the theory of verb ranking or

salience scheme.

(3) Define the storyline in relation to “verb ranking,” “salience scheme,”

“foregrounding vs. backgrounding,” and “tense-aspect” as a perceptual

metaphor of foregroundedness that is made salient through the transitivity of

events (including time movement) and sequentiality of events exhibiting a

prototype effect from the prototypical dynamic verbs on-the-line to the

prototypical static materials far-off-the-line, with support from DeLancey

11 Phonology, and its connection to historical-comparative research, is the richest area which has received

attention in the linguistic study of Iu-Mienh for the past five decades (cf. Aumann and Sidwell 2001, Ballard 1985, Callaway and Callaway 1976, Chang 1953; 1966, Chen 1991, Downer 1963, Theeraphan 1988; 1989; 1997, Peiros 1998, Purnell 1965; 1970; 1985.

Page 44: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

18

(1987), Dry (1992), Endo (1996), Hopper and Thompson (1980), Tomlin

(1987), Wallace (1982), and others.

(4) Remove off-the-line materials first, then identify the storyline; that is,

following Somsonge’s insights in Thai (1991:76, 1992b:410-33) and

Longacre’s “peeling off” method (1996:27), an analysis will be done from

margin to nucleus: namely, from cohesion, evaluation, irrealis, setting,

flashback, cognitive state, background activities, then to storyline.

(5) Apply the “billiard-ball model” of Langacker (1991b:282-329) in order to

analyze the transitivity of serial verb constructions and topic chains on-the-

line to reveal their conceptual structures.

(6) Apply Langacker’s “augmentation and grounding” model to an analysis of

retrospective conjunctions and prospective conjunctions (2001b:151) in the

framework of Current Discourse Space (CDS) to analyze the meaning of

sequentiality in storyline.

1.10.2 Data Collection and Verification

Seven narrative texts for this study were tape-recorded and transcribed by Ms. V. Ann

Burgess, who has served as a missionary nurse-midwife and as one of the senior

research associates of the Mien dictionary (Purnell, Zanh G-F, Burgess, forthcoming)

in Doi Luang, Maechan district, Chiang Rai province since the 1970s.

Data verification was done through monolingual interviews and discussions (cf.

Everett 2001:167) with Mr. Bienh Wuonh Mengh (นายพชิญพิเชฐ พนัธุพิสุทธิชน) (mid

20’s), a Chiang Rai variety speaker, and Mrs. Dangc Meix Daqv (นางเหมยตะ

ตั้งกติติกุล) (mid 50’s) from Chiang Khong, Chiang Rai. 12 Though coming from

12 Though coming from Chiang Rai province, her variety is closer to that of Chiang Kham, Phayao province,

probably due to a geographical proximity. Jennings (personal communication 2006) has pointed out that Meix Daqv often uses different words than Chiang Kham, while she has identified herself as a Chiang Kham variety speaker to me.

Page 45: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

19

different areas, both now live in Chiang Mai, hence, providing the present researcher

with the opportunity to learn from them. Mr. Zanh Gueix-Fongc also helped the

author in double-checking of the data.

1.10.3 Summaries of the Narrative Texts

The names, types and number of sentences of the texts are listed in Table 3. These

sentences include quotations of direct speech (DS). Though sometimes DS includes

multiple sentences, these are not counted as a separate sentence but are included

under the heading number. When DS is uttered without a quote marker, it is

numbered as a separate sentence but not analyzed as storyline material.

English Title Mienh Title Type Story-teller

Number of Sentences

Story 1 A Story of an Old Husband and Bamboo shoots

Nqox-Gox Caux Mbiaic Nyei Gouv

Wisdom story Yauz-

Guangv 17

Story 2 A Story of Firefly and Grasshopper

Gaeng-Kiev caux Caah_Laauh_Porngc Nyei Gouv

Animal folk tale (A why story)

Yauz-Guangv 35

Story 3 A Story of Three Piglets

Buo Dauh Dungz-Dorn Western nursery or children’s story

Yauz-Guangv 71

Story 4 Why Cicadas don’t Have Intestines

Gaeng-Waen Maiv Maaih Jaangh Nyei Gouv

Animal folk tale (A why story)

Yauz-Guangv 78

Story 5 A Story of Faam-Baeqv and Aengh Doih

Faam-Baeqv Aengh Doih Nyei Gouv

Love story Yauz-Guangv 120

Story 6 A Story of Aahan Aa^han Gouv Hero story Yunh Zoih 135

Story 7 A Story of Big Snake

Domh Naang Gouv Mysterious tragedy

?Yauz-Guangv 259

Total Sentences 715

Table 3. Titles and types of seven stories

A summary of Story 1 (OH): “A Story of an Old Husband and Bamboo Shoots” is as

follows:

There was a man who had a daughter. He wanted her to marry a very old man because of his riches. “Once you marry him, you will like him” was his persuasion.

Page 46: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

20

But she didn’t want to, so she got an idea to persuade her father to stop thinking about it. She collected old and hard bamboo shoots for a meal. Trying to eat it, her father complained to her that they were too hard to eat. Her reply was, “Once you bite it, you will gradually like it.”

A summary of Story 2 (FG): “A Story of Firefly and Grasshopper” is as follows:

A firefly and a grasshopper took turns working for each other until it got dark. On the way back home they raced to each one’s home. The grasshopper was confident of winning because he had long legs. But the firefly reached home first because he had a light and wings. On the other hand, the grasshopper asked a sparrow to let him sleep in his house. While sleeping, a dead tree snapped and fell down to frighten the grasshopper. He forgot to keep his legs bent and kicked the sparrow hard. He excused himself blaming the dead tree. The dead tree blamed termites. The termites blamed a worm. In the court case the worm was sentenced as guilty and got a punishment of strangling. This is why worms have a ring around their necks today.

Story 3 (3PG) “A Story of Three Piglets” is a Western nursery tale retold by a Iu-

Mienh storyteller Yauz-Guangv, who also narrated four or five other stories.

Therefore, it is completely Mienized having typical Iu-Mienh narrative features. It

goes like this:

Three brother piglets built each one’s house. The first one built his house with stalks of rice. The second one built his house with rods and sticks, and the third one with bricks. A wolf came, blew the first house down, and ate the first piglet. Then he blew down the second house and made it collapse, and then ate the second piglet. The third house was strong. The wolf and the third piglet competed in their cleverness several times. Finally, the piglet won and the wolf died…in what way? The story will tell you.

Story 4 (CI): “Why Cicadas don’t Have Intestines?”13 is another “Just-so story” (cf.

Story 2), which can be summarized as follows:

It explains why cicadas appear hollow, owls have such big eyes, and yellow squirrels’ backs are yellow. All these have reasons, starting from a barking deer that misunderstood the owl’s words. This was the first link of the chain, which caused the barking deer to step on an ash pumpkin’s vine. Cut off, it rolled down to bump a sesame seedpod, which went into a wild chicken’s eyes. Out of agony, the wild chicken scratched black ants, which stung a snake, which in turn went into a nest of a yellow squirrel. Driven out of his nest, the yellow squirrel bit an oil fruit, and then it dropped off a tree to hit a grandfather’s opium lamp, which spilled everything out. A punishment had to be given to each one. Who was most wrong?

13 A Lao Iu-Mienh version of this story can be found in Beard et al. (eds.) (1995:11-19), narrated by Ta Fou

Saechao, under the title “The Owl’s Words” Norqc Guv Long nyei Gouv, in the Unified Script and English translation.

Page 47: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

21

Story 5 (FA): “A Story of Faam-Bae’ and Aeng-Doi”14 is, in fact, of Chinese origin,

well integrated into Iu-Mienh culture. It can be summarized like this:

Aeng-Doi was a girl who desired to study in school in the days when girls never did. She disguised herself to be a boy, and went to a school far from home. She had a classmate boy, Faam-Bae’, who believed Aeng-Doi was a boy. They got along well with each other, did all things together, sharing school books, pencils, and a bedroom too. But a teacher was suspicious about Aeng-Doi, so he tested her twice to uncover her secret. Escaping from these tests, she went home alone. By the time Faam-Bae’ searched and found his old friend, she had become a beautiful young woman. After considerable confusion in his heart, he decided to marry her. But he was too late because Aeng-Doi’s marriage with another man had already been arranged by her parents. Yet, she knew that she was to be united with Faam-Bae’ even after death. So she told him to die first and wait for her. He did. On the day of her wedding to the unwanted man, she called Faam-Bae’s grave to open; she ran into the open grave, then the two ascended to the sky.

Illustrations of Story 4 (CI) and Story 5 (FA), drawn by Kao Fong Saelee in Beard (1995:13, 89)

Figure 9. Illustrations to “Cicada Story” and “Faam-Bae’

and Aeng-Doi”

Story 6 (AS): “A Story of Aahan” is a hero story full of action, adventure and humor.

Unlike Story 4 (CI), Story 5 (FA), and the next one, Story 7 (BS), which are found

among the Iu-Mienh in Laos as well, it is unique to the Thailand Iu-Mienh in that it

shows a sociolinguistic contact with the Northern Thai people, such as the use of

some Northern Thai words with “Iu-Mienh accent,” and interactions with Thai

officials. “Aahan” means “brave,” and this is how he goes about:

14 An English translation of a Lao Iu-Mienh version of this story can be found in Beard et al. (eds.) (1995:75-96);

unfortunately the original Iu-Mienh version was not included.

Page 48: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

22

A poor Iu-Mienh couple, whose occupation was a vender going through Northern Thai villages, suddenly became famous because of their brave actions. The husband was called “Aahan the brave.” First, the wife killed a bear accidentally, and then the husband killed a crocodile that attacked a passenger boat. He further saved a whole community by getting rid of a gigantic snake that had fallen into a village water source, and he killed a tiger too. Each time he solved a problem, he was rewarded with a special title by a Nothern Thai governor. Finally, when he drove back a group of enemy soldiers, he and his wife were given a fortified city.

A summary of Story 7 (BS): “A Story of Big Snake”15 is as follows:

A man had three daughters. When he worked in a forest, he got a help from a big snake. So he gave it the third daughter as a wife. At the snake’s house, she gave birth to a baby boy. When she went her father’s home to pay respect after the wedding, her elder sisters deceived her by taking her out to a forest. They let her fall down from a high tree into a pond and killed her. The first daughter went back to the snake’s house to be his wife. They found a strange bird in the forest and kept it at home, but he noticed that it was his first wife from the way it related with their baby. The elder sister killed it. The husband picked it up and dressed its meat to eat, but his second wife spat it out on the ground, which became mustard green. They ate it, she spat it out again, and it became a big bamboo. A grandmother found that this bamboo had magical power to keep a fire alive in a cooking place, which came from the first wife living in it. The grandmother took the wife home and kept her. The husband noticed again that this woman from the bamboo was his first wife and took her back home. The elder sister realized that her younger sister reincarnated to be a beautiful woman. She wanted her younger sister’s beauty and followed her instructions. Following it, she was boiled to death.

1.10.4 “Cognitive Linguistics” as A Cover Term

In this study the term “Cognitive Linguistics” is used in a hypernymous sense

embracing Langacker’s CG, Fauconnier’s Mental Space Theory (MST) (1994), and

Talmy’s Cognitive Semantics (CS) (2000a, 2000b) as its hyponyms. Primarily, CG is

used for the analysis of the conceptual structure of narrative discourse in regard to the

storyline. Only secondarily, some notions from MST and CS are used. CL’s

overlapping relationships with other disciplines in a broad perspective of cognitive

15 A Lao Iu-Mienh version of this story can be found in Beard et al. (eds.) (1993:62-70), narrated by Muang Yoon

Saechao with the title “Old Black Snake” Naang-Jiev Gouv in the Unified Script and English translation.

Page 49: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

23

science presented by Tsuji (2002:183)16 is reproduced with an emphasis on CG, CS,

and MST in Figure 10.

Figure 10. A relationship of Cognitive Linguistics with

other disciplines (Tsuji 2002:183)

1.10.5 Analysis Procedure

The data was interlenearized and glossed through monolingual interviews with the Iu-

Mienh language consultants with supplements from two dictionaries, Lombard and

Purnell (1968) and Panh (1995). Occasionally, the forthcoming revision of the

former dictionary (Purnell, et al.) was consulted. The text corpus was divided into

two kinds: one is the direct speech (DS) in the narratives, the other the narrative

materials. The sentences in the latter, numbered up to 715, will be analyzed with

respect to the storyline parameters such as punctiliar, sequential, dynamic, realis,

16 Though this diagram appears to be comprehensive, one should be also reminded of CG’s significant

contribution to historical-comparative linguistics, such as an account for diachronic semantic change in a process of grammaticalization studied by Geeraerts (1997) and Sweetser (1990).

Page 50: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

24

narrative, substantive, following Longacre (1996). The corpus in the former kind, the

materials in DS, was separated from the latter because they do not carry the storyline.

Three kinds of analysis were conducted. First, a storyline was analyzed using

Longacre’s “peeling off” method, Langacker’s grounding theory and Tayler’s twelve

transitivity parameters (a cognitive linguistically modified version of Hopper and

Thompson 1980 by Tayler 2003). This resulted in the salience scheme of the

storyline verbs and the supportive materials. The second was an analysis of the

transitivity of the storyline verbs/clause using Langacker’s billiard-ball model with

respect to the action chain in the constructions. The third was an analysis of the

sequentiality of some key constructions in relation to the storyline applying

Langacker’s current discourse space (CDS) model, particularly, the theory of

incrementing CDS.

1.11 Scope and Limitations of the Study

The major limitations of this study are threefold. First, this study is within the

category of narrative discourse. Longacre’s typological classification of discourse

(1968, 1972a:167-85, 1996:10) includes narrative, procedural, behavioral, and

expository discourses. Of these four, narrative discourse is characteri zed by such

features as [+Contingent Succession] and [+Agent-Orientation]. Besides these two

features, if the narrative category has the feature [–Projection], viz. “minus

projection,” this is a story; that is, the events “are represented as having already taken

place” (Longacre 1996:9). It is through this narrative genre among the oceans of

discourse study that we will attempt to wade, narrated by Iu-Mienh storytellers in

seven pieces.

Concerning the term “discourse” analysis, it was used in the early works of Longacre

(i.e. till mid-80s) in the sense of a text unit larger than the sentence, and has been

replaced by “textlinguistics” later (e.g. 1989a, 2003a, 2003b, Longacre and Shin Ja

Hwang 1994). This change is helpful to distinguish the text analysis (cf. chapter 4)

Page 51: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

25

from the kind of “discourse analysis” which has a functional or social aspect.17 For

example, Schffrin (1994) gives three different definitions of discourse, i.e. “discourse

as language above the sentence,” “discourse as language use” and “discourse as

utterances” (viz. from a perspective of the interaction between structure and function)

(1994:23-41). Typically, Longacre’s textlinguistics falls into Schffrin’s first

definition: discourse as language above the sentence. Thus, following Longacre, the

term “narrative discourse analysis” under the present study is used in a textlinguistic

sense rather than the “discourse analysis” in the sociolinguistic, functional or

pragmatic sense.

Second, the scope of this study is storyline, which is foregrounded in a narrative

discourse. In his succinct account for what discourse analysis is to aim at, Payne

(1997:343, 351-6), summarizing Givón (1983:1-41), points out the study of

continuity (cohesion) includes three areas: topic continuity, action continuity and

thematic continuity. In this sense, the present study concentrates on the action

continuity as it seeks to investigate foregroundedness through sequentiality and

transitivity in storyline.

Therefore, admitting that there is rich information in the “off-the-line” materials18 as

well as topic continuity and thematic continuity, this study is limited to the primary

task of investigating the sequentiality and transitivity in the sense of action continuity

in a narrative discourse analysis. By way of explanation, the term sequentiality is a

narrowly defined word; it is a hyponym of continuity. Thus, the examination of

17 The distinction between “text” and “discourse” is necessary from a social/functional perspective. From this

perspective, describing a text of a narrative belongs to textlinguistics; while describing how the genre narrative is used in a social context with all additional components (e.g. symbolic or cultural meaning) is discourse analysis, a text in use. On the other hand, a borderline between the two approaches may be blurred: e.g. Longacre’s approach (2003a) incorporates sociolinguistic analysis, speech acts and dialogue analysis into textlinguistics. In addition, other scholars approache both aspects simultaneously, for example, as in works by Halliday and Hasan (1989) form Systemic Functional Linguistics, viewing language as a social-semiotic phenomena; Brown (1995) from cognitive/pragmatic approach; or Smith (2003) from the discourse modes perspective.

18 Payne (1997:353) writes, “‘Foreground’ information is not the same as the ‘most important’ information in a text. It may be more accurate to say that the foregrounded material is the framework on which the important information is hung. […] The real significance of a narrative, as well as any other kind of text, often is carried in the ‘background’ clauses.”

Page 52: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

26

discourse cohesion is demarcated to the minimum extent as needed for the

sequentiality in storyline only.

Third, this research will be confined to a semantic investigation of storyline in terms

of the conceptual structure in Langacker’s sense. This means that we do not enter an

area of hermeneutics of texts, such as a sense of Johnstone’s heuristic approach to

discourse analysis (2002:230-1). She claims to “consider meaning from all of these

perspectives,” (i.e. what the speaker means, what the text itself means, and its

meaning to its audiences); rather, we limit ourselves to a linguistic study of the

meaning of narrative texts. Halliday and Hasan (1976:327) contend “[t]he linguistic

analysis of a text is not an interpretation of that text; it is an explanation;” thus, we

will try to explore how the linguistic constructions are interconnected to encode

meaning in the storyline reflected in the conceptual structures.

Page 53: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

CHAPTER 2

LITERATURE REVIEW

2.0 Introduction

In contrast to the academic scene five decades ago when Harris blazed a trail in

the analysis of discourse with the distributionalist’s emphasis (Harris 1952a,

1952b), discourse study is now so popular that there is even a high school level

introductory textbook (Chimombo and Roseberry 1998), as well as standard and

general university-level introductions (cf. Brown and Yule 1983, Georgakopoulou

and Goutsos 1997, Jaworski and Coupland eds. 1999, Johnstone 2002, Schiffrin

1994, Schiffrin, Tannen, and Hamilton eds. 2001). Furthermore, the discipline is

now highly interdisciplinary as exemplified in the works of Hatim 1997, Schäffner

and Kelly-Holmes eds. 1996, Scollon and Scollon 1995, van Dijk ed. 1985, etc.

In this chapter the literature on discourse and text analysis in the period of the

1960s to the 1990s will be reviewed with reference to their attention to the study

of meaning in discourse and texts.

2.1 Theories of Discourse and Text Analysis in General

Subsections 2.1.1 to 2.1.12 survey the treatment of meaning in discourse in

various linguistic theories.

2.1.1 Discourse Analysis in Tagmemic Theory

After Harris, Pike was one of the earliest linguists in the American structuralist

tradition who advocated the study of discourse. In Pike’s tagmemics (Pike 1967,

1982; Pike and Pike 1982), the study of meaning is treated in the referential

hierarchy in his triple-hierarchy: phonological, grammatical, and referential.

While his early discourse study (Pike 1964) was concerned with the identification

of tagmemes as composites of situational roles and grammatical roles in matrix

(viz. barely above the sentence level), a later work with his wife Evelyn Pike was

done in a top-down way (i.e. discourse to morphemes) in the referential hierarchy.

27

Page 54: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

28

E. Pike 1983 is subtitled as “systematic exegesis;”19 that is, leading or drawing

meaning out of a text systematically. The meaning of a text is analyzed through

presentations of grammatical constituents and referential constituents, echoing

Pike’s fundamental view of “form-and-meaning composite” (1982:111-7), in the

tree/branch diagrams and the formulas with the four-cell tagmemes (Pike 1981:47-

64, Evelyn Pike 1983:12-6; 1988). With a similar theoretical background of

tagmemics, the issue of form and meaning in relation to participant identification

in discourse was studied by Wise 1971 [1968]. Among a host of writers on

discourse analysis in the tagmemic and structuralist milieux, E. Pike 1983 seems

to be the only work which addresses the issue of referential meaning in an explicit

term “exegesis.”20 As will be seen in 2.2.1, this type of meaning in discourse can

be viewed in terms of “a conduit metaphor of discourse,” that is, meaning is

contained in a text.21

2.1.2 Meaning in Textlinguistics

While the term “discourse analysis” is generally used in the American structuralist

tradition, “textlinguistics” is “a prominent area of linguistics in Europe” (Fasold

1990:65). A difference between the two continents may be approximated by

saying that while discourse analysts are more engaged in the description and

documentation of texts of less-known or never-written languages, the textlinguists

are interested in the processing and interpreting of written texts in well-known

languages. Beaugrande and Dressler (1981), in discussing seven standards of

textuality (i.e. cohesion, coherence, intentionality, acceptability, informativity,

situationality, and intertextuality), freely use the notions of cognitive psychology

and procedural semantics, such as activation, strength of linkage, decomposition,

19 ‘Exegesis’ comes from Greek ejxhgevomai exêgénomai (a compound verb of a prefix ex- ‘out’ and ágô ‘to

lead’), which means ‘to draw (meaning) out from’, i.e. ‘tell, relate, explain, report, make known, reveal’ and a related noun is ejxhvghsi" exêgêsis ‘narrative, description’ and ‘explanation, interpretation’.

20 There is a brief mention in Longacre (2003b:179) that “the role of discourse analysis in general […] is exegetical rather than predictive.”

21 It should be remembered that the text-internal meaning is only one kind of several types of meaning in discourse. Other types involve social and functional meanings.

Page 55: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

29

spreading activation, episodic vs. semantic memory, economy, global patterns,

inheritance, plan, script, schema, mental imagery, and so forth (1981:92, 93).

Their view of meaning is characterized as the distinction between virtual meaning

(i.e. the potential of expression or sign for representing and conveying knowledge)

and sense (i.e. the knowledge actually conveyed by expressions occurring in a

text) (1981:84). A continuity of senses gives text coherence, which is mutually

(i.e. between a sender and a receiver of the text) accessible and relevant within a

configuration of concepts and relations. The receiver’s selecting and verifying of

conceptual schemata “contributes to comprehension” (1981:198) of the situation

or text. Of the seven standards of textuality mentioned above, the study of

intentionality may also reveal meaning by the author of a text.

2.1.3 Discourse Analysis in Generative Semantics

In the U.S.A. in the late 1960s and early 1970s, Generative Semantics was widely

disseminated over the country, while the opposing interpretive semanticists,

namely, those following Chomsky within Extended Standard Theory were

concentrated at MIT. Though the name Generative Semantics may have a

nostalgic connotation to some people22 by now, Grimes’ discourse analysis in that

framework is of importance in that it provided later discourse analysts, as well as

his contemporaries, with tools for capturing different kinds of information in

discourse (Grimes 1971).23 Such notions as a distinction of information between

events and participants (his chapter 3) and non-events (his chapter 4) (e.g. setting,

background, evaluations, and collateral) are developed in Longacre’s profile and

storyline analysis to supplement the semantic side of his discourse study.

22 To some others, such sarcastic comments of Newmeyer’s as “The public lectures given by Lakoff, Ross,

McCawley, Postal, and others resembled political rallies as much as academic seminars” (1980:152) or “The Fall of Generative Semantics” (1980:133) may be recalled.

23 Grimes’ theoretical assumption is that: “[W]e can say the most about language by factoring out two different things: the decisions a speaker can make regarding what and what not to say, and the mechanisms and patterns that are available to him for implementing the results of those decisions in a way that communicates with another person. The decisions that the speaker makes, and the relations among them, are referred to as the underlying formational structure, [...] or the semantic structure” (1975:30, 31).

Page 56: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

30

Working in close association with tagmemists, yet identifying himself as a

generative semanticist, Grimes sought “a more revealing theory” (1975:30) than

the kind of discourse analysis that Pike and Longacre were developing, albeit his

recognition that the latter gradually showed “the direction of including semantics”

(Grimes 1975:23). In other words, semantics in discourse analysis in his day was

not yet a majority concern, at least in American structuralism.

2.1.4 Discourse Analysis in Functional-Typological Linguistics

In a broad range of theoretical similarity shared by structuralism and

functionalism, the following sections will touch on four theories that have been

labeled “functional” in one way or another.

Functional-Typological Linguistics “seeks to account for linguistic phenomena,

based on cross-linguistic data, in relation to human communicative/feeling states

through an on-going interactive discourse between the speaker and the hearer”

(Horie and Sato 2001:1). For example, Givón and others base their investigations

in topic continuity in discourse on the data from Amharic, Biblical Hebrew,

Chamorro, English, Hausa, Japanese, and Spoken Latin-American Spanish (Givón

1983). His theoretical assumption is “that language and its notional/functional

and structural organization is intimately bound up with and motivated by the

structure of human cognition, perception and neuro-psychology” (Givón 1984:11),

as applied to interclausal coherence. For him, “the grammar of referential

coherence is not primarily about reference. Rather it is about indentifying and

activating the locations (‘files’, ‘nodes’) where verbally-coded text is stored in

episodic memory. The nominal referents-topics serve as ‘file labels’; they are

used to access (‘activate’) the storage locations where incoming information is to

be ‘filed’” (Givón 1990:894).

Bybee and others discuss modality in discourse (Bybee 1995, Myhill and Smith

1995, Silva-Corvalán 1995). Myhill’s approach seeks a language-universal

framework to explain why the speakers of a language choose a particular

construction from alternatives (e.g. NP types, tense and aspect, voice alternation,

Page 57: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

31

and word order) with objective contextual evidences in discourse by the use of

quantitative analysis of such parameters as humanness, recency of mention of

referent, specificity, controlled vs. uncontrolled future actions, affectedness of

patient, and temporal sequencing (Myhill 1992, 2001). Hopper (1979) and

Thompson (1997) should also be considered to be important linguists, active in

this theoretical tradition, whose works will be reviewed in relation to foreground-

background in narrative discourse in chapter 4 (cf. Hopper 1979, 1982, 1995,

1998; Hopper and Thompson 1980).

As will be mentioned in subsection 2.2.4 below, those linguists who are working

with the Functional-Typological methodology have much in common with

Cognitive Linguistics concerning the approach to meaning. (cf. Chafe 1994, 1996,

Croft 1991, 1999, Givón 1990, 1994, 1995)

2.1.5 Discourse Analysis in the Prague School

Sharing the common view that communicative function is important in the system

of language with Functional-Typological Linguistics, the approach of the Prague

School may be referred to as structural-functional linguistics. As opposed to the

descriptive emphasis of the former, however, in the latter, “semantics has always

been understood as belonging to the core of the system of language” (Sgall

1994:277). Prague School linguists take discourse as language use in

communication; that is, discourse as a sequence of utterances, rather than a

sequence of sentences. In this sense, the semantics24 of the Prague Linguistic

Circle can be considered as semantico-pragmatics at discourse level, whereas at

the lexical level it endorses “the Saussurean notion of meaning (as linguistic form

of cognitive content)” (Sgall 1994:294). Meaning in discourse is analyzed in

terms of the topic/focus articulation (TFA) of the sentence (Hajičová 1994,

Peregrin 1996:236-7, Sgall 1987) and of the hierarchy of communicative

dynamism; meaning of the sentence, in turn, is distinguished into contextually

24 A historical account of how the Prague School has been studying meaning (e.g. Jakobsen and Mathesius)

can be found in Leška 1996.

Page 58: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

32

bound and non-bound. Despite the high interest in meaning from its early period

and concern with pragmatic discourse (utterance), the School’s analysis is more

focused on sentences than lengthy pieces of text.25

2.1.6 Discourse Analysis in Systemic-Functional Linguistics

Those who discuss more directly the matter of “discourse-semantics” (Eggings

1994:82-4) than those who are mentioned in the previous subsection are scholars

of Systemic-Functional Linguistics, exemplified in the works by Halliday and his

followers (cf. Berry 1981; Coulthard 1987; Coulthard ed. 1994; Eggings 1994;

Enkvist 1987; Fries 1992, 1995, 2002; Halliday 1994; Halliday and Hasan 1976,

1989; Hasan, Cloran, and Butt eds. 1996; Lemke 1988; and G. Thompson 1996).

Eggings, for example, explains that three kinds of meaning (i.e. experiential,

interpersonal and textual meanings) run through a text (1994:83); thus, they are

analyzed, following Halliday and Hasan 1989, in the following ways. The

experiential meaning (i.e. what the discourse is talking about) is investigated by

analyzing the transitivity structure of the clauses. The interpersonal meaning (i.e.

who is taking part) is accounted for by the analyses of mood, modality, and

persons. In order to identify textual meaning (i.e. role assigned to language),

lexico-grammatical organization needs to be described by analyzing the content

words (e.g. verbs, nouns, and circumstances such as prepositional phrases and

adverbs) of transitivity structure of the clauses (Eggings 1994:83) with reference

to theme, information (Fries 1995, Cloran 1995) and cohesion (Aziz 1988).

In sum, Halliday and Hasan have written that “the concept of cohesion accounts

for the essential semantic relations whereby any passage of speech or writing is

enabled to function as text” (Halladay and Hasan 1976:13). It may be safe to say

that the study of textual meaning for Systemic-Functional linguists is closely

25 In this connection, a more extended understanding about topic (or “theme”) as characterized at the

paragraph and discourse level has been presented by Jones (1977) from the Pikean tagmemics perspective.

Page 59: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

33

concerned with the analysis of cohesion26 while the experiential and interpersonal

meanings are also in the scope of the theory.

2.1.7 Discourse Analysis in Functional Grammar (FG)

In the Netherlands, Simon C. Dik and his followers around the world have been

developing Functional Grammar (FG), also with an emphasis on semantics and

pragmatics. The theory has the term “functional” in that the primary function of

language is “a symbolic instrument used for communicative purposes” (Dik

1980:46). As to the relationship among areas of linguistic studies, Dik goes on to

say, “Syntax is subservient to semantics, and semantics is subservient to

pragmatics” (1980:46). FG aims to have four-fold adequacy: typological

adequacy, descriptive adequacy, psychological adequacy and pragmatic adequacy.

Three distinct functions that FG highlights are semantic functions (e.g. Agent,

Goal, Recipient, Beneficiary), syntactic functions (Subject and Object), and

pragmatic functions (Theme and Tail, Topic and Focus).

Though discourse study in FG, or functional grammar of discourse, is not fully

developed yet, Dik has set out an outline (1997:409-41). According to Dik,

discourse is approached from three perspectives: global decision (similar to

Longacre’s discourse typology or genre), global structure (e.g. hierarchy, units,

relations), and coherence.

Integrating FG with textlinguistics, however, Buth (1995:77-100) has analyzed

Biblical Hebrew and Aramaic discourse with an emphasis on the distinction

between Topic and Focus. Topic, in FG, “is a constituent of a clause that has

received special marking […] in order to signal the intended perspective for

relating the clause to the larger context” (Buth 1995:84). It is also termed as

“contextualizing constituent” (or C.C.). “Focus, on the other hand, is a way of

specially marking the salient, important information of a sentence” (ibid.). Topic

26 A cohesion analysis, from the standpoint of stratificational grammar, is found in Gutwinski (1976), where

he applies it to Henry James and Ernest Hemingway’s literature pieces.

Page 60: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

34

provides a discourse with continuity, focus enhances foreground. Relating the

concept of topic (C.C.) and focus with foreground and background, Buth redefines

the foreground as “sequential-foregrounding” structure that has a more author-

oriented pragmatic function, than a traditional understanding that “[t]he

foreground of a narrative is usually defined to be the sequential chain of

completed events” (1995:86). On this ground, Hebrew’s peculiar word order

XSV (where X is any NP or other non-Subject and non-V element), as opposed to

the default VSO, is analyzed as a mark of discontinuity. A significant implication

of his continuity-discontinuity analysis in place of the traditional type of

foreground-background to storyline analysis will be discussed in chapter 6.

2.1.8 Discourse Analysis in Relevance Theory (RT)

Relevance Theory (hereafter RT) is inherently semantic in that it digs out “the

hidden implications” (Morgan 1994:148), being “a theory of cognition which

offers a model of how the mind works” (id. 1994:127). Originally developed by

Sperber and Wilson (1986), RT operates on the principle of relevance:

Principle of relevance

Every act of ostensive communication communicates the presumption of its own optimal relevance. (Sperber and Wilson 1986:158)

In other words, because “[h]umans tend to pay attention to what is relevant to

them,” […] “relevance, and the maximization of relevance, is the key to human

cognition” (Blass 1990:43) and communication. A good communication is

achieved when it is contextually relevant; that is, in RT’s term, communication

has a contextual effect, or “the link-up between an utterance and its context” (Gutt

1992:21). The term “context” in RT is “the cognitive environment of the hearer”

(Gutt 1992:22), which includes perception memory, inference, general knowledge

about science, people’s beliefs, and culture and so forth. The contextual effects

that make communication relevant have three kinds: (i) derivation of a contextual

implication, (ii) strengthening of a contextual assumption, and (iii) elimination of

a contextual assumption through contradiction (Gutt 1992:22-3).

Page 61: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

35

Applying RT to discourse analysis, Blass claims that it “provides a better

theoretical foundation for comprehension and textuality than coherence models”

(1990:43). 27 She argues that discourse analysis in RT has more strength in

understanding a text’s meaning than a mere classification of discourse types.28

She summarizes: “coherence- and topic-based approaches look at purely textual

connectivity: relevance theory is based on the relation between information

content and context. Moreover, discourse analysis within relevance theory is not a

purely formal matter, but involves a consideration of how texts are understood,

how they are processed and what effects are achieved” (Blass 1990:80).

Following Blass (1990) within RT, Follingstad (1994:151-88) analyzes storyline

in Tyap narrative discourse.29 Identifying two sets of particles (one has two

preverbal particles, and the other includes three conjunctions), he argues that the

first set has the contextual effects and that the latter has a. thematic prominence, b.

focus prominence in relation to the foreground of the narrative, and c. emphatic

prominence “with respect to the speaker-hearer axis of communication”

(1994:188). Follingstad’s insight into the nature of the latter is of great

importance for the present study of storyline in Iu-Mienh narrative; one of the

second set, i.e. s, “a coordinating conjunction which tends to imply temporal

succession” (1994:169), “tends to correlate with […] prominent foreground

events” (id. 1994:169, 171, 188). Attention should be drawn to his statement that

it is “conjunction,” as opposed to verbs or narrative tense that marks, in the

particular case of Tyap, the temporal succession and foreground events (e.g. H2 in

27 A discourse marker, or an illocutionary adverb bekicur ‘in short’ in Modern Hebrew within the framework

of RT was studied by Shloush (1998), who agrees with Blass 1990 in saying, “I tend to accept the idea that texts do not occur with fixed coherence relations and that coherence is derivative from the extablishment of relevance. Nevertheless, this does not mean that a text does not exhibit distinct hierarchical structures, from which coherence may follow” (Shloush 1998:78).

28 Blass’s illustration goes like this: “just setting up these types [narrative, conversation, procedural, hortatory] does not say anything about the function of these discourses nor why they have the particular structures they do. This approach is like comparing different games such as football, cricket and polo just by their outward appearances and forgetting that players are actively involved in achieving particular goals, the goals being different in each case” (1990:80).

29 Tyap (also “Kataf”) is a member of the Niger-Congo family, spoken in southern Kaduna State in Nigeria. An application of RT to discourse analysis of another African language, Lobala, a Bantu language of Norhwest Zaïre, has been carried out by Morgan (1994:125-49).

Page 62: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

36

1.1.2). This importance will be discussed in chapters 4 and 5 as Iu-Mienh exhibits

a similar phenomenon.

By way of reference, it is noteworthy that Follingstad has moved from RT to

Mental Space Theory (MST), perhaps through a common emphasis on a

cognitively-oriented approach to language, applying to Biblical Hebrew discourse

analysis with a particular focus on kî ‘because’ (2001).

Due to RT’s deep concern with relevant interpersonal communication, together

with the wider sense designated by the term “context” of interpersonal

communication than a context of texts, it is natural that it extends its scope to the

study of discourse markers. Such examples include Shloush (1998) and Ziv

(1998) in the semantic and pragmatic study of Modern Hebrew: the former on

bekicur ‘in short,’ the latter kaze ‘like this’: both identifying the conceptual and

procedural meanings of these discourse markers within the framework of RT.

2.1.9 Discourse Markers and Sociolinguistic Discourse Analysis

Congenial to the study of discourse markers within RT, others have done similar

research with no particular theoretical brand name. They include studies on oh,

well, and, but, or, so, because, now, then, y’know, I mean (Schiffrin 1987), and

yeah and like (Jucker and Smith 1998) in English, on rotsè lishmoa kéta? ‘wanna

hear something weird/funny’ (Maschler 1998) in Modern Hebrew, on X nante, X

nanka, X nado, X dano, X toka, and X tari, pejorative expressions of ‘the likes of

X’ (Suzuki 1998) in Japanese, on k3 ‘then, also, consequently’ (Yajai 1985), na5,

la5, si5, and th2 (Paensiri 1998, Cooke 1979) in Thai, and on amen and hallelujah

in African American sermons by Wharry 2003 to name a few.

As is obvious from these, the study of discourse markers resides at the intersection

of the sociolinguistic study of discourse (Fasold 1990:65-75, Gee 1999, Stubbs

1983), conversation analysis (Coulthard ed. 1992, Fox 1987:6-7630), pragmatics

30 After the treatment of anaphora in conversational analysis in English, the latter half (pp. 77-136) of Fox’s

study focuses on the anaphora in monologue expository written texts.

Page 63: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

37

(e.g. Clark and Haviland 1977), and the study of discourse markers and reported

speech in written texts (e.g. Larson 1978). Furthermore, a sociolinguistic

discourse analysis which incorporates Halliday’s social-semiotics (Halliday and

Hasan 1989) provides the discourse marker study with a wider and more adequate

perspective. Some of this research may be interesting; however, they are of

limited relevance to the present study (except for Thai k3, cf. 2.1.12, infra).

2.1.10 Narratology

Quid est enim tempus? (“What, then, is time?”) was the question Augustine asked

in his Confession, which lead Paul Ricoeur (1984 [in French 1983], 1985 [1984],

1988 [1985]) to argue for temporality of narrative. On the other hand, however,

he also approves an “emplotment” (1984:31-51) of narrative, following Aristotle,

the culturally polar end of Augustine. According to Ricoeur, Augustine analyzed

time in terms of a threefold present, 31 whereas Aristotle argued that time is

something passing from the past through the present to the future in a plot of

mimesis/muthos (viz. narrative). Ricoeur’s analysis is illuminating in suggesting

that the aspectual nature of the Iu-Mienh language, as opposed to tense-oriented

Indo-European languages, is akin to Augustine’s theory of time in narrative, while

the underlying philosophy of the plot/profile analysis in narrative discourse

theorized by Longacre (1996:33-50) is from an Aristotelian understanding of

plot.32 Rather than separating them, though, Ricoeur puts forth the correlation

between “the activity of narrating a story [with plot therein] and the temporal

character of human experience” in the following way: “time becomes human to

the extent that it is articulated through a narrative mode, and narrative attains its

31 Meaning, “By saying that there is not a future time, a past time, and a present time, but a threefold present,

a present of future things, a present of past things, and a present of present things, Augustine set us on the path of an investigation into the most primitive temporal structure of action in terms of this threefold present. The present of the future? Henceforth, that is, from now on, I commit myself to doing that tomorrow. The present of the past? Now I intend to do that because I just realized that… The present of the present? Now I am doing it, because now I can do it. The actual present of doing something bears witness to the potential present of the capacity to do something and is constituted as the present of the present” (Ricoeur 1984 :60).

32 It is noteworthy that Ricoeur (1985:7) points out that, “[t]he Aristotelian theory of plot was conceived during an age when only tragedy, comedy, and epic were recognized as ‘genres’ worthy of philosophical reflection.”

Page 64: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

38

full meaning when it becomes a condition of temporal existence” (1984:52)

(italicized by Ricoeur).

In addition to Ricoeur’s philosophical narratology, there is also: a.

structural/functional narratology (Prince 1982), 33 b. sociolinguistic narratology

(Toolan 2001), c. literary narratology (Abbott 2002), and d. exegetical/theological

narratology (Coats 1985).34 An interesting common feature among them is their

acute consciousness of the distinction between the narrative time and the real-

world time. 35 For example, Toolan (2001:42) states that, “there is something

unreal and heavily convention-laden about so-called story time and text time. In

neither case are we referring to actual temporal progression, but the linear verbal

representation of temporality” (emphasis by Toolan). Furthermore, “[g]iven a

narrative recounting a chrono-logical sequence, where sequence is taken to be a

group of non-simultaneous topic-comment structures the last one of which

constitutes a modification of the first, events can be distinguished in terms of their

relevance to that sequence” (Prince 1982:68).

In other words, change/transition/movement of events is inherently bound with the

time passage in narrative. It is this inseparability of event sequentiality and event

transitivity along the time-line that makes the narrative time more complex than

the real-world time. This is also a fundamental nature of a prototypical narrative,

since “[t]his dependence of time with regard to change (movement) is a sort of

primitive fact […]” (Ricoeur 1988:15, [1985]). This significant nature of

33 Prince (1982:163) summarizes the goals of narratology as “to discover, describe and explain the mechanics

of narrative, the elements responsible for its form and function.” He goes on to say, “narratology gives us an insight into the principles governing systems of signs and signifying practices as well as our interpretation of them” (idem 164).

34 Coats (1985:15) categorizes Hebrew Old Testament narratives into Saga, Legend, Tale, Novella, and Fable, asserting that recognition of different narrative forms affects exegesis: “interpretation advances under the careful control of genre definition.”

35 Nevertheless, in contrast to the distinction, it is also necessary to note that Abbott (2002) tends to integrate narrative time and real-world time. He says, “In narrative […], though it is the incidents that give shape and that dominate our sense of time, the regularity of abstract time, which is also an integral part of all our lives, unavoidably adds its own counterpoint to the time structured by incidents. […] We have always been aware of the recurring cycles of the sun, moon, and seasons, and at the same time we have always been shaping and reshaping time as a succession of events, that is, as narrative” (2002:5).

Page 65: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

39

inseparability between the sequentiality of events and the transitivity of events in

relation to the storyline (e.g. H2 in 1.1.2) will be discussed in chapter 5.

2.1.11 Discourse Study by Thai authors

It has been mentioned three times thus far in this section (by Buth (1995) in 2.1.8;

Follingstad (1994) in 2.1.9; and Riccoeur (1984, 1985, 1988) in 2.1.11) that the

central issue of inseparability between change/movement/transitivity of events and

sequentiality of events along the time-line is crucial to the storyline. In a similar

vein, a fourth author, who is from Thailand, Somsonge has also stated that the

sequentiality marked by a group of certain conjunctions and serial clauses is “the

most important criterion” (1991:105-6) in determining whether clauses and verbs

in question are really on the storyline. She has also stated that these markers often

co-occur with temporal adverbs. For instance, she says, “k3 signals that the

following event is sequential to the previous one” (Somsonge 1990:71). Further,

Yajai (1985:2) claims that, among other functions of k3, it has the function of

“addition, temporal sequence, sequence of events,” “discourse cohesion” (id. 6),

and “prominence marker” (id. 9). In addition, Somsonge (1992a: 48-9) has found

a similar function with the Sgaw Karen particle sw as highlighting “an

important chronological juncture”, “a backreference” (id. 50-1) which indicates

“the crucial backbone event” and “progression from the preceding backbone

event” (id. 50), “a connectivity of successive events” (id. 51).

Finally, her recent study on Hlai (Li) discourse (Somsonge 2002) has integrated

the above-mentioned findings in her own version of “storyline” theory, “narrative

timeline.” “The narrative timeline is defined as the main line of development, i.e.

the line of sequential, punctiliar happenings (Longacre 1996)” (Somsonge

2002:142). Incorporating Dry’s point that “for a structure to move time, it must

present new information” (1983:33), Somsonge summarizes that the narrative

timeline must have two major components: sequentiality and presentation of new

information. Her statement (ibid.) is presented visually as follows:

Page 66: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

40

The clauses ↑which refer to (i) sequenced points on timeline and constitute the foreground (ii) present new information of a narrative.

Hence, H2 of the present study has been hypothesized for Iu-Mienh, which also

has a conjunction ziouc [zjou] meaning approximately ‘and then, so, after which’

similar to Thai k3 ‘then’ and cak2nan4 ‘after that’. A distant brother of the Iu-

Mienh in the Hmong-Mien family, the White Hmong, has been reported by Clark

(1988) to have a similar conjunction los [l11] “with the approximate meaning

‘(and) then/so, thus, therefore, yet, and it happens, and it turns out, with the

result’” (1988:93). Her naming of it as “an inchoative conjunction” suggests that

it has a sort of force of aspectual/verbal element to contribute to movement in a

discourse. Clark’s later investigation (1992) has added data of the Vietnamese

conjunction thì and the Black Tai conjunction kò “all seem to have meanings

something like ‘well, (and) then, (and) so, and it happens that, and it turns out

that’”(1991:87), adding one more function, that of topicalizer.

Therefore, insights from these Thai authors and other Southeast Asian languages

can be summarized as follows: (1) it is not only verbs but particles (i.e.

conjunction-like particles) that mark the storyline in these languages that do not

have tense-aspect marking on verbs, (2) these conjunctions mark sequentiality,

which is the most important criterion for the identification of storyline, and (3) the

sequentiality is indicated in corroboration not only with conjunctions but also

serial clauses and temporal adverbs.

For other Thai discourse studies, see Person’s (1993:6-7) lucid summary. There

are two narrative-related MA theses of Chulalongkorn University, which studied

the Iu-Mienh. Uthai’s (1982) sociological and pedagogical study of Iu-Mienh

folktales includes 35 narratives from Ban Huai Mae Sai village, Chiang Rai

province, which is the same variety as the subject of the present study.

Kasamaporn (1990) investigated the participant reference in five narratives

loosely following Longacre’s discourse analysis.

Page 67: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

41

2.1.12 Discourse Semantics

Sharing the same term “discourse semantics” with Eggings (1994:82) mentioned

in 2.1.6, Seuren (1985) has more polemic characteristics. Seuren’s book is

primarily a departure from “surface semantics,” “truth conditional semantics,”

“calculus semantics,” and an “autonomous” view of components of language. He

criticizes them with numerous citations of insufficiency of propositional analysis

and entailment analysis, terms that come from a logical point of view. He claims

that his semantics is a “discourse-dependent linguistic interpretation” (1985:1),

holding the view that discourse is a “psychological reality.” He even asserts, “that

logic is psychologically real in so far as humans carry out logical operations as

part of their cognitive behaviour” (1985:211).

Yet, despite the same dissatisfaction with autonomously disconnecting grammar

from semantics as cognitive semanticists and linguists have been lamenting,

Seuren’s theory is not CL. Discourse semantics from a more cognitive oriented

perspective has been outlined by Tomlin, Forrest, Pu, and Kim 1997. Tomlin et

al. (1997:104-5) discuss “the central issues and concepts of discourse semantics”

in two main problems as follows:

The first is the problem of knowledge integration: how the individual propositions in a text and discourse are integrated to reflect well the speaker’s conceptual representation and to optimize the creation of an appropriate conceptual representation in the listener. The second is the problem of information management: how information is organized and distributed as the speaker and listener interact during the blueprint creation process.

In the next section (2.2), cognitive approaches to discourse analysis will be

reviewed following the basic outline by Tomlin et al. (1997): the knowledge

integration and information management.

2.1.13 Summary

To summarize this section on the treatment of semantics in discourse analysis in

various theories, one can observe that the study of meaning in discourse was not a

major concern in structuralism. Functionalists are confined to sentence level

Page 68: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

42

study or studying meaning in the pragmatic domain. Some theories have the

psychological and cognitive range to capture the meaning beyond the sentence

level. One such example is the discourse semantics from the cognitive linguistic

perspective. Concerning the study of storyline in narrative discourse, some

linguists who pay attention to semantics have found that certain conjunctions

exhibit both the sequencing function and the developing function along the

timeline of narrative.

2.2 Cognitive Approaches to Discourse Analysis

In this section, following Tomlin et al. (1997) mentioned in 2.1.12, an overall

picture of cognitive discourse semantics will be reviewed. Therein, works that

have been done within various cognitive theories and cognitive oriented

approaches will be integrated. Such theories include CL (CG, CS, MST),

cognitive-minded functionalists (e.g. Chafe and Givón), an application of

cognitive psychology to linguistics (e.g. van Dijk and Kintsch), and linguists of

other theoretical schools who approach their subject matter in a highly cognitive

way without reference to CL. Subsequently, views of meaning by Longacre and

Langacker will be compared concerning the matter of the inseparability of the

form-and-meaning composite or surface-and-deep structure. Finally, the section

will propose a model of a CG approach to discourse.

2.2.1 An Overall Picture of Cognitive Discourse Semantics

There are two basic ways of looking at discourse metaphorically speaking: a

conduit metaphor of discourse and a blueprint metaphor of discourse. Cognitive

discourse semantics views discourse and texts utilizing a blueprint metaphor. The

conduit metaphor represents discourse and text as things that contain meaning,

where an act of text comprehension can be reduced to “exegesis” or drawing

meaning out of the text. In contrast, in the blueprint view of discourse, “the

speaker holds a conceptual representation of events or ideas which he intends

should be replicated in the mind of the listener” (Tomlin et al. 1997:64). In this

view, creation and comprehension of discourse have two major areas: knowledge

Page 69: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

43

integration and information management as shown in Tomlin et al. (1997:71),

who use Levelt’s (1989) concept of blueprint for the pre-verbal message (Figure

11).

└───────────────┘ └──────────────┘ Knowledge Integration Information Management

(adapted from Tomlin et al. 1997:71, based loosely on Levelt 1989)

Figure 11. A blueprint for the speaker

Starting from the left end of the Figure 11, the knowledge integration stage

involves the formation of the conceptual representation as “the fundamental

‘meaning’ the speaker works with in constructing a discourse” (Tomlin et al.

1997:68). That is, the conceptual representation is formed by receiving raw

materials from remembering, perceiving and creating activities. The conceptual

representation in turn is processed through four kinds of information management,

namely, rhetorical management, thematic management, referential management

and focus management. Thus in this model, cognitive discourse semantics deals

with knowledge integration and information management leaving the utterance

formulator to cognitive morphology and cognitive phonology (cf. Figure 10).

2.2.2 Knowledge Integration

In order to build and interpret discourse or text successfully, the semantic

information provided in each utterance or sentence has to be integrated into a

coherent whole (Tomlin et al. 1997:65). This integration involves three areas: (i)

morpho-syntactic coding, (ii) implicatures, and (iii) planning and inference. MST

Page 70: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

44

can be placed in the morpho-syntactic coding of the model inasmuch as

Fauconnier organizes his model by saying “[t]he spaces set up by a discourse in

this way [a new mental space is set up relative to an existing space that is in focus]

are organized into a partially ordered lattice. At any given stage of the discourse,

one of the spaces is a base for the system, and one of the spaces […] is in focus.

[…] As “the discourse participants move through the space lattice; their view

point and their focus shift as they go from one space to the next” (1997:38-9).

Implicatures and inference are not purely morpho-syntactic but important in

producing and understanding discourse. In this sense, Langacker’s domain-based

meaning and encyclopedic semantics in CG (1987a:62-3; 154-66) has more power

to explain implicatures and inference than formal semantics.

2.2.3 Information Management

This section includes four areas of discourse management: rhetorical management,

thematic management, referential management and focus management.

2.2.3.1 The Rhetorical Management of Discourse

The process of producing discourse (spoken and written) involves the information

component and the action component,36 both guided by the speaker’s/writer’s

goal. The information component includes study areas like propositional content

and pragmatic matters (e.g. emphasis, importance, presupposition). From a

decoder’s point of view, all of them guide how the semantic content should be

interpreted. From an encoder’s viewpoint, a selection process is at work. In

regard to the information management component, Langacker has investigated the

selective nature of linguistic coding, exploiting CG notions of focal adjustment of

selection (1987a:116-20), construal (1991b:294-8), and action chain (1991b:213-

7). With respect to the pragmatic side of discourse, the study of discourse markers

36 The sentence final particles that have illocutionary force in Asian languages (e.g. -oc [o] ‘POLITE

PARTICLE,’ –aex [E] ‘I-REALLY-MEAN-IT,’ –maah [ma] ‘COMMAND,’ nii [ni] (with an emphatic rising intonation) ‘I-AM-TELLING-YOU’ in Iu-Mienh) should also be studied in the cognitive discourse semantics under the area of the rhetorical management.

Page 71: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

45

in RT (though dismissed as irrelevant if they are only from a classifying approach,

cf. 2.1.10) Andersen (1998), Ariel (1998), Jucker and Smith (1998), Shloush

(1998), Ziv (1998), among others, should belong here due to their attention to the

cognitive status of the discourse markers.

2.2.3.2 The Thematic Management of Discourse

In the thematic management of discourse, at least three major problems are

recognized: (1) theme or topic at clause level, (2) discourse theme, and (3)

foreground and background proposition, i.e. the problem of propositional

centrality. Commonly understood definitions of ‘theme’ include: (i) theme as

aboutness (i.e. it is that which the predicate talks about), (ii) theme as a starting

point or what Lambrecht (1994) terms “a point of departure for the clause,” and

(iii) theme as a center of attention. As will be seen in chapter 6, the thematic

marker or topic marker –nor ‘talking about,’ -aeqv ‘as for’ in Iu-Mienh come

under the term “theme as aboutness” and they will be analyzed in relation to the

conceptual structure of sequentiality (cf. 2.1.12).

From the perspective of cognitive psychology, Van Dijk and Kintsch (1978, 1983)

studied “cognitive information processing,” which they defined as “understanding

discourse” using the notion “macro-structure”. They state, “The global meaning

of a discourse is represented by semantic macro-structures” (1978:68). In other

words, it is “an abstract semantic description of the global content, and hence of

the global coherence of discourse” (Van Dijk and Kintsch 1983:189). Originally

taken from Generative Semantics, in the conviction that there must be “THE DEEP

STRUCTURE OF A TEXT” (van Dijk 1972:130), van Dijk’s notion of macro-structure

of texts (1977:130-63; 1981:195-214) (except for its cognitive psychological

emphasis) was incorporated into Longacre’s discourse analysis (1979a, 1985a) to

reinforce the semantic part of his theory (cf. 2.1.3).37

37 Note the cross-academic cooperation among the American Generative Semantics, European Textlinguistics,

and American discourse analysis.

Page 72: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

46

Rong (1992) identifies English logical connectors express foreground-

backgrounding in discourse. The background connectors include though, despite,

in spite of, regardless, notwithstanding, even though, since, as due to, for

example, that is, in other words, by the way and in passing. The foregrounding

connectors are but, however, nevertheless, yet, so, therefore, also, moreover,

besides, to sum up, overall, anyhow, to resume, etc. It should be noted, as has

been pointed out in 2.1.9 and 2.1.12, that conjunctions can express foreground-

background when they are analyzed from a CL perspective.

The present study of storyline, which can be broken down to be an analysis of

foreground and background, receives its due position under the thematic

management of discourse in a broader perspective of cognitive discourse

semantics: specifically, in the issue of propositional centrality.

2.2.3.3 The Referential Management of Discourse

Under the term referential management, information structure such as given, or

old, and new information in the traditional investigations receive a more

cognitively oriented treatment. Though belonging to no particular theoretical

group, Chafe (1973, 1974) researched them in a new light of “psychosemantics”

as “active memory” (given information), “semi-active memory,” and “inactive

memory” (new information) in the flow of discourse (1979, 1987, 1994:53-191).

Givón’s (1983) referential distance model is devised to measure the gap between a

previous occurrence of the referent/topic and its current occurrence by counting

numbers of clauses between them. He also claims that coherence is a mental

entity rather than being found in texts (1995). For Givón “text comprehension is

synonymous with the construction of a structured mental representation of the

text” (1995:64) on the assumption that coherence is grounding (i.e. the more

grounded the clause in a discourse is, the more mentally-accessible it is). In a

similar vein with Chafe (1974) and Givón (1983), Epstein (1999, 2002)

investigated the accessibility of the English definite article in discourse within the

MST framework.

Page 73: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

47

Research into anaphora can be located under the term referential management of

discourse. Van Hoek (1997) has researched the conceptual-semantic relationship

of anaphora within a CG perspective resulting in the establishment of the

“reference point model.” Conceptual connectivity is argued to be a continuum

ranging from the most strongly interconnected structures to more weakly

connected structure, and to the weakest, that is, discourse unit boundaries. Her

area of research is reinforced by Langacker (1996) in that “restrictions on the

location of a pronoun vis-à-vis its antecedent are best described in terms of

conceptual configurations” (Langacker 1996:333). Liang’s (1996) work has

revealed that zero anaphora in Chinese discourse is an “emergent reference” using

a cognitive strategy term.

2.2.3.4 The Focus Management of Discourse

The focus management segment is a counterpart of the theme or topic at clause

level under the thematic management of discourse when it is interpreted in Prague

School terms: namely, the theme-rheme pair. In cognitive discourse semantics,

however, the term focus management has broader perspective encompassing the

focus as prominence, focus as pragmatic function (Dik 1980, Buth 1995:84, cf.

2.1.8), and Lambrecht’s (1994) predicate focus, argument focus and sentence

focus. Following Lambrecht, Heimerdinger (1999) analyzed topic, focus and

foreground in Biblical Hebrew with some cognitive psychological concerns.38

Holmqvist and Holšánová (1997) have investigated empirically the focus

movement of a text during a time of telling and listening to a spoken discourse.

They found that a visualization of language is taking place in the mind of the

listener.

38 Heimerdinger’s work is different from Johnstone’s heuristic approach to discourse analysis (2002:230-1, cf.

1.11 supra.). With regard to the focus structure in a third person narrative clause, Heimerdinger analyzes predicate-focus structures, argument-focus structure and sentence-focus structure in both wayyiqtōl (the preterite verb or Longacre’s foregrounded material) clause and NP + qātal (perfective verb) clause. He concludes from them that the latter construction, which Longacre claims to be the background material, indicates “a change of information structure in the sentence” and sometimes indicates “foregrounded rather than backgrounded material” (1999:219). See 4.1 concerning its implication to Longacre’s theory of storyline.

Page 74: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

48

Subsections 2.2.1 to 2.2.3 have reviewed literature related to cognitive linguistics

approaches in a scaffold-like framework of broad cognitive discourse semantics

outlined by Tomlin et al. (1997).

2.2.4 Cognitive Functional Linguistics

As has been noted above, there are number of linguists who are cognitive-minded

or cognitive-psychology-concerned with regard to the semantic issue of discourse.

These approaches were recently summarized by the Japanese cognitive linguists

(Horie and Sato 2001) under a new term “Cognitive-Functional Linguistics.”

The last item in this chapter of literature review, Dooley and Levinsohn (2001),

(written independently from the aforementioned Horie and Sato (2001) in the

same year with the same interest from different backgrounds), possibly fall into

this category of cognitive-functional discourse analysis. Though intending to

provide linguistic field workers with an introductory manual for discourse

analysis, it contains “a functional and cognitive approach” (2001:vii) such as the

notion of “mental representation” (2001:21-5, 49-60) as a conceptual framework

within the traditional method of Longacre’s genre analysis and text charting

techniques. In fact, Levinsohn co-authored with Longacre the “field analysis of

discourse” twenty-three years ago (Longacre and Levinsohn 1978:103-22). It

seems to be natural for Levinsohn to have increased his concern about semantics

of discourse as an exegete of Biblical languages; Hebrew (Levinsohn 2000) and

Greek (1992a, 1992b, 1992c).

2.3 Views on Meaning

In the last subsection 2.2.4, it was suggested that field linguistics could

legitimately proceed to the interpretation of meaning of texts after an initial stage

of description. A comparison of two field linguistics manuals, i.e. Dooley and

Levinsohn (2001) and Longacre (1964), shows that both have a serious concern

with the issue of meaning. However, at the same time the comparison may

suggest that their approach to the study of meaning has changed in the period of

nearly forty years. In this section, two views on meaning will be compared.

Page 75: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

49

2.3.1 Longacre’s view on meaning

In Longacre’s field manual (1964) for linguists who work in a structural and

tagmemics framework, his view on meaning can be seen:

Obviously, meaning lurks in the background everywhere. The background of meaning would seem to be essential to the analysis. […] We assume with Pike that the units we handle are form-meaning composites. Nevertheless, it seems necessary to insist that it is the formal side of the composite which is amenable to initial systemic analysis. (1964:23) [underline added]

While both formal and semantic sides are recognized as important, it should be

noted that a major interest in an “initial systemic analysis” necessitates a

concentration on the former (or the surface structure) rather than the latter, which

is reasonable within the context of the discovery procedure and descriptive field

linguistics.

Furthermore, in the process of development from descriptive linguistics at

morpheme-to-sentence levels (1964) to a full-fledged discourse analysis,

Longacre (1972b) accentuates the inseparability of form and meaning as follows:

Here [i.e. in questions of universality], I accept without further cavil the point eloquently made by post-structuralist linguistics that deep or semantic structures are relatively universal, while surface structures are relatively restricted and language specific. I reject, however, any implication that ‘surface’ is to be regarded as superficial, or ‘deep’ as more relevant. While this implication is a hazard built into the very terms themselves, the contention here is that both deep and surface structure are prime concerns of the student of language and that meaning is found in both. (1972b:xi)39 [underline added]

This contention that meaning is found in both surface and deep structures based

on the conviction of inseparability between them urges us to revisit de Saussure’s

signifié-signifiant composite (de Saussure 1910-11:93 in Komatsu and Harris

1993; English translation by Baskin 1959:67) as a transitional link to Langacker’s

39 Longacre goes on to say “[s]urface structures are of obvious importance, in that all that is said, or ever will

be said, in a given language must be expressed in its surface structures. Tagmemics (of which I am a practitioner) need not be apologetic for its many years of preoccupation with the discovery procedures for, and the taxonomy of, surface structures” (1972:xi).

Page 76: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

50

concept of grammar as symbolic units comprising semantic and phonological

poles (Langacker 1987a:76-86).

2.3.2 Langacker’s view on meaning

De Saussure’s axiomatic diagrams are in Figure 12 as follows:

(The left diagram is from Emile Constantin’s notebook VII of the 2nd of May, 1911, in Komatsu and Harris 1993:75. The two diagrams on the right are from Baskin’s translation of Saussure [1911] 1959:67)

Figure 12. Signifié-signifiant composit of a sign

The diagrams in Figure 12 show that the linguistic sign is composed of concept

and acoustic image, namely, signifié and signifiant (signified and signifier).

Further, Saussure says that “the linguistic sign is based on an association made by

the mind between two very different things, but which are both mental and in the

subject: an acoustic image is associated with a concept” (1910-11; Komatsu and

Harris 1993:74a).40 From the above, it is clear that both the symbolic (sign) and

mental nature of linguistic expressions have been carried into Langacker’s CG.

It is this symbolic nature or the oneness of the signified-and-signifier composite of

the linguistic sign in the mental domain that keeps investigators from becoming

subjective in the analysis of meaning. This symbolic understanding demands

consistency in grammatical analysis as well, which will be shown in section 2.4.

40 “[L]e signe linguistique repose sure une association faite par l’esprit entre deux choses très différentes,

mais qui sont toutes deux psychiques et dans le sujet: une image acoustique est associée à un concept” (Constantin’s notebook VII, 2nd of May 1911 in Komatsu and Harris 1993:74)

concept arbre

image acoustique arbos

“tree”

arbos

arbos

Page 77: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

51

Before that, however, attention should be drawn to one more statement about the

“surface structure” that is closely associated with an objectivity of linguistic

analysis. As has been observed above (2.3.1), Longacre (1964) said, “that it is

[in] the formal side of the composite…” and “in both” (1972b:xi) of the composite

that meaning should be found. In 1976, he refined his position by saying, “that

the surface structure patterns in themselves have meanings which are imposed

over the deep structure patterns” (Longacre 1976:10). Even though he inclines

toward the formal side of the composite as a place of meaning’s existence,41 a

balance between the form and meaning can be secured as expressed in Saussure’s

following metaphor:

[T]aking up [an] example of the sheet of paper that is cut in two, it is clear that the observable relation between the different pieces A, B, C, D, etc. is distinct from the relation between the front and back of the same piece as in A/A′, B/B′, etc. (Saussure 1959:115 [1910-11]).

Here Saussure talks about the following diagrams in Figure 13

(Saussure from Constantin’s notebook VIII, 12 May 1911 in Komatsu and Harris 1993:89a)

Figure 13. The internal and syntagmatic relation of

linguistic signs

41 Indeed, Longacre elsewhere writes, “I think we must insist that surface constructions have meaning which

is imposed over and sometimes in tension with deep structure meaning” (1976:256).

On the one hand we have this relation, already mentioned:

and on the other hand this relation

Term A Term B

Concept

acoustic image

Page 78: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

52

The internal relationship between the “concept” and “acoustic image” is likened to

the front and back of a piece of paper. Therefore, what is needed is a model that

deals with both sides simultaneously. For this purpose, Saussure’s diagram is

utilized to provide CG with the symbolic assumption of language (Langacker

1987a:11-13).

The notion of the “sign” composed of the “concept” and the “acoustic image” are

incarnated in Langacker’s modern charicterization that linguistic “expressions”

associate “a semantic representation of some kind” with “a phonological

representation” (1987a:11). Not only that, the two ovals connected by an arrow in

Figure 15 suggest that a syntagm (i.e. a sequenced linguistic form of more than

two items) also has the same semantic value in itself. This is the basis for

Langacker’s claim that grammar is symbolic: “Grammar is simply the structuring

and symbolization of semantic content” (1987a:12).

Having looked at the comparison in subsections 2.3.1 and 2.3.2, one should be

reminded of the fact that Longacre’s apparent rejection of the “deep structure” or

reluctance to delve into the semantic side of the form-meaning composite and

Langacker’s deep concern about the meaning in grammar are not in conflict.

Rather, their standpoints actually stem from a common dissatisfaction with an

abstract treatment of syntax detached from the reality of language.

2.4 Discourse Analysis in Cognitive Grammar

Developing the Saussurean diagram (Figure 13) into a discourse compatible

device, Langacker (2001b:145) presents the following syntagmatic windows

(Figure 14):

Page 79: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

53

Figure 14. Aspects of the usage event (Langacker

2001b:145)

Parallel to Saussure’s diagram, the upper frames represent the conceptualization

pole, and the lower the vocalization pole. The usage event consists of the speaker

(S) and the hearer (H) (or the “interlocutors” collectively) interacting with each

other or paying attention to each other as indicated by the dotted two-hand arrow

between them. The upward arrows indicate that both the speaker and hearer are

actively focusing on the conceptualization. That the upward arrows cross the

upper and lower frames means that “each has to deal with both a

conceptualization and a vocalization” (Langacker 2001b:144). A discourse occurs

in a usage event, which is “carried out by the speaker and hearer.” A meaningful

communication can only be actualized when both the speaker and the hearer are

grounded (G), that is, “the interlocutors’ apprehension of their interactive

circumstances,” as the inner frame has the label “Ground,” which is in turn

situated in the context. This context is referred to as the context of speech,

“interpreted broadly as including the physical, mental, social, and cultural

circumstances” (2001b:145) (cf. Halliday’s ‘context of situation’ in Halliday and

Hasan 1989:3-14), thus it is surrounded by the “Shared Knowledge” (cf.

Halliday’s ‘context of culture’ in Halliday and Hasan 1989:46-7). As pointed out

Page 80: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

54

in subsection 2.2.2, shared knowledge is a part of the encyclopedic knowledge of

the language users, which has to be taken into consideration in the cognitive

analysis of discourse.

The “Viewing Frame” represents a limit or amount that the mind can

conceptualize at one moment. In the viewing frame, the interlocutors’ attention is

drawn to a selected entity that is focused. The bold-lined circle represents the

focus of attention or profiling.

The whole window containing the above-mentioned relations and configuration

within is termed the current discourse space (CDS). “The CDS is defined as the

mental space comprising those elements and relations construed as being shared

by the speaker and hearer as a basis for communication at a given moment of the

flow of discourse” (2001b:144). This flow is indicated by “> >” in the passage of

“Time” along the penetrating arrow. An occurrence of a discourse is thus

conceptualized in the usage event, “i.e. actual instance of language use”

(2001b:144). All the necessary elements involving the usage event of a discourse

are schematized as a representation of conceptual structures in the CDS in Figure

14.

Arrows, circles, and boxes in CG diagrams, various profiling, relations and

conceptual configurations can be expressed in this schematic representation of

CDS by the technique of highlighting lines. This is a tool to capture the meanings

of linguistic expressions without disconnecting them from grammar. In other

words, this device is an entry into the conceptual structure through the surface

structure, with the conviction that “Cognitive grammar equates meaning with

conceptualization…[…] Linguistic semantics must therefore attempt the analysis

and explicit description of conceptual structures” (Langacker 1991a:278).

Langacker also states, “[l]ingusitc meaning is largely a matter of construal, and

dimensions of construal reasonably described as matters of prominence are critical

to both semantics and grammar” (2001b:158). In chapter five, this will be utilized

to analyze an increment of discourse spaces and a conceptualization of some

conjunctions.

Page 81: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

55

2.5 Summary of the Chapter

Chapter 2 has surveyed how various linguists from different theoretical

backgrounds approach the semantic issues manifest in discourses and texts. Four

approaches to this issue and one specific area of Iu-Mienh will be summarized

below.

Broadly speaking, there are four basic approaches to the meaning of language at

the discourse level. The first is the formal approach. In this approach, the field-

based discovery and descriptive procedure, the texts documentation, and the

structural-functional methodologies are employed from a holistic perspective (i.e.

discourse to clause and lower) or a build-up perspective (i.e. sentence syntax

applied to a larger unit) to explain how a text is organized as a coherent whole.

The study of meaning is conducted in relation to the wholeness of text, thus

encompassing the study of cohesion and information structure at the sentence

level. A rigorous semantic study at the discourse level from this approach is

difficult to find. However, their concentration on empirical field-based

description of surface structure and the analysis of a text corpus have provided

later generations with a solid foundation of data and methodology toward a deeper

investigation into the semantic domain of discourse.

The second approach is mental. The meaning of texts are approached from

cognitive psychology, psycholinguistics, and neuro-psychology, by both linguists

who are psychology-minded and psychologists who are linguistic-minded,

providing an interdisciplinary environment internationally. These researchers

approach texts in such a way as to “process” and “interpret” them. In this

approach, the process of the interpretation of texts is the central area of study,

rather than the texts themselves.

Third approach is a sociolinguistic study of discourse. Here the meaning of

discourse can be found in the middle ground between the sender and receiver of a

message in a relevant context. Therefore, the study of meaning in this type of

discourse analysis is perceived as a subject of communicative, pragmatic, and

Page 82: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

56

social-semiotic research. In Halliday’s term, the interpersonal meaning of

discourse is treated in this area (sociolinguistics), while the ideational meaning

belongs to the second group (mental) mentioned above.

The fourth approach is symbolic, reconciling the formal and mental sides.

Language and grammar are construed as a symbol comprising the semantic

content and its phonological representation. In this view, grammar and semantics

are inherently one, and this oneness strives to answer the concern about ignoring

the surface structure in the study of discourse as held by the majority in the first

group. Discourse, whether written or interpersonal, is also taken as a symbol, that

is, a conceptual entity. As such, the study of discourse cannot be exempt from the

analysis of its meaning in this approach of CG.

One specific issue with regard to Iu-Mienh, suggested through several cases of

other languages (i.e. from Africa and Asia), is the multifunction of some

conjunctions in relation to the storyline. This group of conjunctions seems to have

both a connecting function and a foregrounding function (as opposed to a

traditional understanding that it is a verb with special tense and aspect that

foregrounds the storyline). This issue will be discussed in chapter 6 from the CL

perspective.

Page 83: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

CHAPTER 3

SOME GRAMMATICAL FEATURES OF IU-MIENH

3.0 Introduction

Chapter 3 lays a grammatical foundation for chapters 4 and 5. Court (1986)

describes Iu-Mienh syntax from noun phrases to simple clauses, as well as

including a basic phonology, based on Purnell (1965). Building upon the

foundation laid by Court, this chapter will recapitulate basic clause types, serial

verb constructions, tense-aspect-modality and topic chains.

As opposed to languages that have morphosyntactic coding with respect to tense

and aspect, Iu-Mienh does not depend on such features to identify the storyline

since it does not have any morphological marking on verbs. The main purpose of

this chapter is not an exhaustive description, but a presentation of verb and clause

types sufficient for identifying a distinction between storyline and non-storyline

materials.

3.1 Basic Clause Types

The basic word order in Iu-Mienh is SVO. The verb in (1) is intransitive and the

subject is an Agent.

(1) Mbopv tiux squirrel run

‘A squirrel ran.’

The prototypical transitive clause has the subject as an Agent and the object as a

Patient as in (2).

(2) (CI .034) Mbopv ngaatc yie Squirrel bite 1sg

‘A squirrel bit me.’

An adverbial clause often precedes the main clause as in (3).

57

Page 84: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

58

(3) (3PG.003) Cl.Adv Cl.Main

Dorh ninh nyei buo dauh dorn hlo deix ninh mbuox neiz42 dorn raise 3sg POSS three CLF cub big some 3sg tell her cub

‘(When she) brought up three cubs (to) big enough, she told them’ (followed by a direct speech).

Adverbial clauses are often marked with the topic marker aeqv ‘as for’

(sometimes nor aeqv), an independent clause marker gau ‘after which,’ ‘as,’ or

‘while’ or an aspectual verb liuz ‘after finishing.’ The aeqv stands at the end of

the adverbial clause (4); gau and liuz come immediately after the verb in the

clause as in (5).

(4) (CI.023) Mbopv maiv_lamh yiem aeqv squirrel lack_a_place_for live TOP

cingx_daaih la'guaih cuotv mingh therefore aimlessly go_out go

‘Because the squirrel did not have place to live, therefore (he) just came out aimlessly.’

(5) (BS.122) Yie zouv gau hnaangx, douz jaamh nzengc mi'aqv 1sg cook DPCL rice fire singe consumed RSLT

‘As I cooked rice, the fire singed all my hair.’

An adverbial clause marked by hnangv ‘like, as’ often comes after the main clause

as in (6).

(6) (AS.090) Nqa'haav aengx deic-bung aengx faanv Later again country again unsettled

hnangv ih_zanc naaiv mborqv jaax like now here hit REC.ACT

‘Later, the country became unsettled again; just as people here fight each other nowadays.’

42 neiz is a contraction of ninh nyei ‘his/her’ (3sg + POSS).

Page 85: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

59

3.2 Serial Verb Constructions (SVC)

Iu-Mienh uses serial verb constructions and multiclausal constructions

extensively, and sentences are often juxtaposed without conjunctions, resulting in

a long string of verbs. In such a situation, it is important to identify which

component in an SVC encodes an event and an action that are on the storyline and

which do not.

A serial verb construction (SVC) is defined by Aikhenvald as follows:

A serial verb construction is a sequence of verbs which act together as a single predicate, without any overt marker of coordination, subordination or syntactic dependency of any other sort. Serial verbs describe what can be conceptualized as a single event. They are monoclausal; their intonational properties are those of a monoverbal clause, and they have just one tense, aspect and polarity value. Serial verbs may also share arguments and obliques. (Aikhenvald 2005)43

Thus, six defining factors of SVC are 1) serial verb as a single predicate, 2)

monoclausality, 3) no intonation breaking between components, 4) shared tense,

aspect, mood, modality and polarity value, 5) serial verb as one event, 6)

prototypical serial verb sharing at least one argument.

As such, each SVC is expected to be on the storyline. However, some SVCs

exhibit a complex internal structure; hence, the following examples should be of

help in recognizing the head of the construction.

Starting from a simple SVC, gorngv mbuox ‘speak tell,’ gorngv heuc ‘speak

call/order,’ gorngv naaic ‘speak ask’ are frequently used both in a conversation

and narrative discourse. The relationship between two verbs in all of these SVCs

seem to be that V1 provides a general meaning followed by V2 with specific

information about how the general meaning of speaking is performed. (7) is from

a narrative.

43 The citation of Aikhenvald 2005 is from the pre-publication draft.

Page 86: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

60

(7) (CI.008) Loz-hnoi norqc_guv_long gorngv mbuox jung old-day owl speak tell barking_deer

"Maqv_gorkv_biouv zuoqc aqv olive_fruit ripe PFT

‘Once upon a time, an owl told a barking-deer, “The olives are ripe.”’

Switch-subject SVCs (Aikhenvald 2005) are exemplified in (8), where the object

of haeqv ‘frighten’ hieh_juv ‘black wolf’ is the subject of the verb gamh_nziex

‘fear.’

(8) (3PG.056) Haeqv nc dauh hieh_juv_jieqv gamh_nziex haic frighten that CLF wolf_black fear very

‘(A piglet) frightened the black wolf into being very afraid.’

An example of a switch-subject cause-effect SVCs, where V1 causes an effect on

the object, which in turn becomes a subject of V2 expressing the effect, is shown

below in (9):

(9) (3PG.069) douz-sioux congx zuqc ninh m'zing sui smoke zap GOAL 3sg eye sour

‘Smoke zapped his eyes making them sore.’

In (10) Cl2 which is a monoclausal SVC is embedded in Cl1. This is explained by

Aikhenvald (2005) as the switch-subject SVC used for complement clause

serialization. The complement clause is in brackets.

(10) (3PG.067) Cl1 Cl2

Dungz-dorn buatc [ hieh_juv_jieqv nyorng jienv faaux] Piglet see wolf_black creep SML.ACT ascend

‘The piglet saw the black wolf creeping up.’

The next four SVCs include adverbial information. The location adverbial SVC is

shown in (11), where the adverbial component precedes the head of the

construction.

Page 87: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

61

(11) (FA.083) ninh yiem horqc_dorngh doqc_sou 3sg be_there school study_book44

‘He studied at school.’

An associative adverbial SVC uses caux ‘accompany’ or ‘to be with’ as in (12):

(12) (BS.167) Njang_hnoi meih caux yie mingh tomorrow 2sg be_with 1sg go

‘Tomorrow, you will go with me.’

Manner adverbial SVCs have two different constructions. One type has the same

constituent order as (11) and (12); the other uses duqv ‘get, aquire, attain.’ The

former type is illustrated in (13), the latter in (14).

(13) (3PG.033) Dauh dungz-dorn njiec qaqv jatv ninh CLF piglet let_go_down strength laugh 3sg

‘That piglet laughed at him very hard.’

(14) (Lombard and Purnell 1968:72) Meih dorngx jauv duqv go haic ni aa! 2sg do_wrong road attain far very part.f

‘Goodness! You missed the trail a long way back!’ (Translation by Lombard and Purnell) (or ‘You’re WAY off the path!’)

This usage of duqv in (14) is referred to as “attainment aspect marker” by Li

(1991:32). Though he does not consider this kind of sentence as an SVC, his

analysis of Hmong examples are labeled “simplex sentences,” thus, fitting the

definition of SVCs by Aikhenvald.

Temporal adverbial SVCs also use duqv ‘get, aquire, attain’ as in (15).

(15) (FA.044) Yie njaaux mh45_mbuo duqv i buo hnyangx aqv 1sg teach 2_PL attain two three year PFT

‘I have been teaching you for two or three years already.’

44 Doqc_sou literally means ‘read book’ (at school). That gives the gloss ‘study.’ See (25) below for ‘read.’ 45 The second person mh is a contraction form of meih.

Page 88: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

62

In directional SVCs, the construction ends with mingh ‘go,’ daaih ‘come,’ faaux

‘ascend,’ and njiec ‘descend.’ They indicate movement of a participant. In (16)

mingh is a directional verb describing the action taking palce.

(16) (BS.190) Ninh auv zorqv pietv guangc mingh 3sg wife take sprinkle throw go

‘His wife took (the meat) and threw (it) away.’

Lastly in this section, a brief mention on a quasi-serial verb construction should be

made. In this, a verb follows a head noun or NP to modify it, as in English ‘food

to eat’. A similar translation for the noun phrase which contains a verb in (17) is

possible, i.e. ‘wings to fly with.’

(17) (FG.011) gaeng-kiev maaih douz yaac maaih ndaatv ndaix fire_fly have light and have wing fly

‘The fire fly had a light and wings to fly with.’

In the identification of a storyline clause, any verb that modifies a head noun

should be demoted to a lower band.

3.3 Tense, Aspect and Modality

A strong connection between the verbal morphology and the narrative storyline

can be seen in Longacre’s statement: “each surface structure type has

characteristic tense/aspect/voice features in the verbs that occur on its mainline in

discourse in a particular language” (1996:12). The crux of the matter seems to be

that it is the verbal tense/aspect/voice that indicates whether a particular verb is or

is not a storyline verb. However, this does not work in Iu-Mienh since its system

for tense-aspect-modality (TAM) has more complex features.

3.3.1 Tense

Tense is defined as “the grammaticalisation of location in time” (Comrie 1985:1).

A comment such as “Chinese [is] an example of a language which lacks tense

entirely” (Trask 1993:276) is possible when one adopts a definition as follows:

Page 89: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

63

Tense Fundamental grammatical (morphological) category of the verb which expresses the temporal relation between a speech act (S) and the state of affairs or event (E) described in the utterance, i.e. which places the event spoken of in relation to the temporal perspective of the speaker. (Bussmann 1996:478)

Though Iu-Mienh does not have a “morphological category of the verbs to express

the temporal relation,” it has tense and a way to express it. Whaley discusses

different strategies in different languages which have “no special tense

morphology” (1997:206-9). He cites an example of Burmese, in which “time

adverbials are used in conjunction with modal affixes” (Whaley 1997:206) to

express tense.

Iu-Mienh also uses temporal adverbs to indicate time reference. For example,

am’hnoi ‘yesterday,’ wuov norm leiz-baaix ‘last week,’ wuov zanc ‘then, that

time’ etc. for past; ih zanc ‘now,’ ih hnoi ‘today,’ jangc jangc ‘just now’ etc. for

present; and aav hitv deix ‘in a minute,’ njang hnoi ‘tomorrow,’ dieh norm leiz-

baaix ‘next week’ etc. for future. The word order is : Adv.tmp + SV as in (18).

(18) (FA.005) Wuov_ zanc hnoi m'sieqv_dorn mv haih doqc_ sou that_time day girl NEG be_possible study_book

‘In those days, women could not study.’

In narrative46 discourse the following temporal adverbs used to refer to past tense

are listed in Table 4.

46 These are found in the narrative corpus for this study. Other temporal adverbs in conversations or other

situations include: hitv deix dangh ‘in a minuite,’ a’hitv deix ‘later,’ aav deix dangh ‘shortly,’ dieh hnoi (dih hnoi) ‘some other day,’ zinh ndaangc ‘in the past, previously,’ nqa’haav ‘later, in the future,’ nqa’haav hingv ‘in the future,’ V + jiex ‘have experienced V,’ jiex daaih wuov norm leiz-baaix ‘last week,’ jiex daaih buo hnyangx ‘in the past three years,’ zanc-zanc ‘always,’ manc-manc ‘gradually,’ etc.

Page 90: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

64 a loz-hnoi in the old days, once upon a time b wuov hnoi that day c wuov zanc that time, then d wuov zanc hnoi in those days e da’nyeic hnoi the following day (lit. the second day) f da’faam hnoi another day (lit. the third day) g da’nyeic ndorm the following morning (lit. the second morning) h njang ndorm the following morning (lit. tomorrow morning) i njang hnoi the following day (lit. tomorrow) j wuov nzunc hnoi this47 time, another occasion k wuov nzunc hnoi aeqv as for this time, another occasion l mv nzunc hnoi aeqv as for this time, another occasion

Table 4. Temporal adverbs for narrative discourse

The items (k) and (l) employ a topic marker aeqv ‘as for’ forming a full length

construction; the whole phrase meaning ‘talking about this time,’ or ‘as for this

time of event’.

The expressions (h) njang ndorm ‘tomorrow morning’ and (i) njang hnoi

‘tomorrow’ refer to the future in a direct speech or an actual conversation.

However, in a narrative discourse, they refer to the future from the perspective of

discourse participant, which is the past from the perspective of audience.

Therefore, in (19) the verb mingh ‘go’ refers to past time reference even though

njang hnoi ‘tomorrow’ is used. By the inherent narrative perspective of telling a

story about something that happened in the past, or an initial temporal setting

indicator that provides the past setting, njang hnoi ‘tomorrow’ means ‘the next

day.’

47 Direct translation of wuov is ‘that’ and ‘that time’ in English refers to what had already happened prior to

the time of narration. However, in narrative discourse, the construction wuov nzunc hnoi means ‘another occasion’ referring to an event that takes place after a previous one. Therefore, the translation for j and k are ‘this time’ and ‘as for this time.’

Page 91: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

65

(19) (FA.090) Njang_hnoi neiz maa mbuo za'gengh mingh lorz tomorrow his mother PL really go look_for

‘The next day (or ‘the following day’), his mother and others [i.e. parents] indeed went to look for (her).’

Some Iu-Mienh native speakers refer to njang_hnoi ‘tomorrow’ of this situation

“tomorrow of the past day,” as opposed to “tomorrow of now.” This is a kind of

relative future tense in Comrie’s sense (1985:56, 124).

3.3.2 Aspect

Comrie offers following definition: “aspects are different ways of viewing the

internal temporal constituency of a situation” (1976:3). In Iu-Mienh aspects are

expressed by serial verb constructions and sentence final particles.

Similar to the directional serial verb constructions discussed in 3.2, perfective

aspects can be expressed by mingh ‘go’ and daaih ‘come.’ In (20) mingh is not

involved in a locomotion in space but indicates that the action has just taken place.

This is different from mingh in (19) which indicates movement in space.

(20) (FG.033) ninh_mbuo zorqv caah_laauh_porngc nyei zaux wuotv ngau mingh 3_PL take grasshopper POSS leg bend bent go

‘They took the srasshopper’s legs (and) bent them.’

(21) (OH.010) Mbenc daaih bun ninh nyei die nyanc prepare PFT let 3sg POSS father eat

‘(She) has prepared (food) (and) let her father eat.

In both (20) and (21) mingh and daaih can be replaced with ziangx ‘finish.’ In

addition, one cannot answer such questions as Ninh mbuo mingh haaix? ‘Where

did they go?’ concerning (20); Ninh yiem haaix daaih? ‘Where did she come

from?’ with respect to (21).

The following five examples contain posthead aspectual verbs: (22) V + ziangx

‘finish V-ing,’ (23) V + liuz ‘finish/done V-ing,’ (24) V + baac ‘finish V-ing,’

Page 92: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

66

(25) V + dorng ‘completed a process of V-ing,’ (26) V + nzengc ‘consume V-ing,

completely V-ed.’

(22) (BS.226) Mbenc ziangx kuv lai kuv hnaangx prepare finish delicious vegetable delicious rice

‘(the old woman) finished preparing delicious vegetables and delicious rice.’ (i.e. finished preparing a delicious meal or a feast).

(23) (CI.061) Laapc liuz cingx_daaih ih_zanc nguaiv nguaiv wuov nor aex put finish therefore now big-eyed big-eyed that as I_mean_it

‘(They) finished pouring (ginger juice into the owl’s eyes); therefore, now the eyes of owls are as big as saucers; I mean it.’

(24) (Lombard and Purnell 1968:22) doqc baac sou read finish book

‘to have finished reading the book; to have finished studying’

(25) Ninh doqc dorng sou 3sg read complete_a_process book

‘He has graduated (from school).’

(26) (FA.054) Yiem wuov haaix_nyungc waac yaac gorngv nzengc be_in there what word TOP speak consumed

‘There (they) talked about all kinds of thing.’ (in a sense that “what ever words were consumed”)

The constructions are summarized in Table 5.

Construction Semantic value Description 1 V-liuz Past TR marker: ‘finished V-ing’: Simple past 2 V-baac Past TR marker: ‘finished V-ing’: Simple past, similar to liuz 3 V-ziangx Perfective

Aspectual Verb ‘completed V-ing’: the perfective sense is stronger than the simple past liuz and baac.

4 V-nzengc Completive Aspectual Verb

‘have V-ed all’: Expresses consumption, an exhaustive or universal plural, some emphasis or surprise.

5 V-dorng Completed Aspectual Verb

‘have completed V-ing throughout’

Table 5. Aspectual auxiliary verbs

Page 93: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

67

V-liuz and V-baac are almost interchangeable in most contexts. As will be seen

in chapter 4 (and example (23)), they are often used in an adverbial clause

preceding a main or independent clause: ‘after finishing V-ing, S + main Cl.’

When used in an independent clause, they require the aspectual sentence final

particles aqv or mi’aqv to ground the whole clause in a usage event as in (27):

(27) (BS.104) Ninh nziez ziouc maeqv zoi njiec bun liuz aqv 3sg younger_sister then peel throw descend give finish PFT

‘Her younger sister then finished taking off (the bracelet and silver neckpiece and) throwing down (and) finished giving (them).

V-ziangx indicates completion of a process or action. Its difference from liuz and

baac is in line with Bybee’s definition of perfective aspect:

Perfective also designates a temporal notion, but not one that is reckoned deictically in relation to the moment of speech, but rather one that is determined by the viewpoint taken on the situation. That is, perfective presents the situation described by the clause as having temporal boundaries, as being a single, unified, discrete situation. (Bybee et al. 1994:83)

The completive nzengc originally comes from a sense that a quantity of something

is consumed. Bybee’s elaboration of completive into three semantic nuances fits

nzengc well:

1. The object of the action is totally affected, consumed, or destroyed by the action. To repeat, ‘to end up’ is a good example.

2. The action involves a plural subject of intransitive verbs or object of transitive verbs, especially an exhaustive or universal plural, such as ‘everyone died’ or ‘he took all the stones’.

3. The action is reported with some emphasis or surprise value. (1994:57) Durative aspect and simultaneous action are expressed by an extended use of the

stative verb jienv ‘to be stable, firm.’48 Example (28) is in a construction V +

jienv ‘keep in a state of V,’ and (29) in V1 + jienv + V2 ‘V2 while/at the same time

V1-ing.’

48 Another meaning is ‘to be important, urgent, pressing’ (Lombard and Purnell 1968:95), but it is not known

yet if the two different meanings are polysemous or they are homophonous.

Page 94: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

68

(28) (BS.209) Ninh ziouc dorh mingh 3sg so_then take go

dapv jienv kapv gu'nyuoz aqv put_into DUR four-sided_woven_basket_with_a_lid inside PFT

‘Then she (the old woman) took (the wife and) kept (the wife) in a woven basket.’

(29) (BS.077) Aengx jiez_sin daaih gan jienv naang mingh further get_up come follow SML.ACT snake go

‘Then again (she) got up (and) went following the snake.’

The following seven examples discuss aspect as expressed by the sentence final

particle aqv ‘change-of-state.’ Though aqv is often used in a perfective sense,

there are other situations where the perfective meaning does not fit, such as a case

of referring to a future time reference. The interpretation of aqv depends on

context largely.

Court analyzes the sentence final particle aqv as a “change-of-state particle” (CH-

O-ST) (1986:248). A contrast between (30a) and (30b) shows that the “change-of-

state” has a perfective sense.

(30a) (Court 1986:24) Yie mv mingh 1sg not go

‘I will not go.’49

(30b) (Court 1986:25) Yie mv mingh aqv 1sg not go CH-O-ST

‘I am not going any more.’ < ‘I have changed my mind from going farther.’

“Change-of-state” as being a generic meaning of the particle aqv can be

interpreted in two ways: 1) perfective aspect and 2) inchoative aspect. (30b)

shows that the change of the speaker’s mind is from the past (i.e. ‘I was thinking

of going before.’) to the present (i.e. ‘Now I have changed my mind not to go’ or

‘I’ve made up my mind, and I’m not going to go,’ or ‘I’ve decided not to go.’)

49 If the speaker means ‘I did not go,’ yie maiv duqv mingh is usually used.

Page 95: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

69

Another example of this perfective sense is (31), in which the change can be

interpreted as from the fruit’s growth in the past to the present state of ripenness.

(31) (CI.008) Maqv_gorkv_biouv zuoqc aqv olive_fruit ripe PFT

‘The olives are ripe.’

On the other hand, the extention of the generic meaning “change-of-state” to the

inchoative aspect is indicated in the right-hand tree of Figure 15.

aqv ║ Change-of-State

From Past to Present From Present to Future

Perfective Inchoative

1st person 2nd person 3rd person

Resolution Casual Command Inception (I’ve decided to…) (Will you do?/You just do) (He is about to…)

Figure 15. Polysemous relation within the change-of-

state aspectual particle aqv

In the inchoative sense, there are three possibilities as in (32)-(35). They are all

used in the present and future time references. Since the subjects are all animate

and can have volition, they correlate with modality. When a speaker is in the first

person, it means, “I have changed my mind from what I have been doing” or a

resolution (32) in a sense that “I have decided to do a new thing.”

(32) (AS.022) Aa, yie m'sieqv_dorn, yie maiv nongc mengh aqv Ah 1sg woman 1sg NEG want fame I_am_sure

‘Ah, I am a woman; I’m sure I don’t want an honor.’

When used to address the second person, it means, “You should change your

present situation,” that is COMMAND but casual (33).

Page 96: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

70

(33) (FG.005) Gaeng-kiev gorngv heuc, "Nzuonx aqv” n vi vt return CAS.CMD

‘The firefly called saying, “Go home!”’

In the second person but in a inclusive sense, it means INVITATION, “Let’s

do…”(34).

(34) (OH.011) Mbuox ninh nyei die, "Nyanc hnaangx aqv." tell 3sg POSS father eat rice CAS.CMD

‘(She) said to her father, “Come eat!”’ [If she were speaking to her peers/friends, it would be ‘Let’s eat!’]

When the subject is in the third person, it means “he is about to…,” or “he is on

the verge of doing…” (35).

(35) (BS.173) Ninh daix yie aqv 3sg kill 1sg INCHOAT

‘Otherwise, he is about to kill me.’

Thus, glossing of aqv varies depending on its context, whether the sentence occurs

in the past (i.e. realis) modality or in the future (i.e. irrealis) modality.

Resultative aspect is “a state that was brought about by some action in the past”

(Bybee et al. 1994:63), and this is expressed by mi’aqv. In the previous

paragraph, it was mentioned that the second person with inclusive sense under the

inchoative usage of the CH-O-ST aspectual marker aqv has the meaning of casual

command or invitation. This is exemplified in Panh (1995:250) where mi’aqv is

given a translation “let’s go” with his note that the expression is a “contraction of

mingh aqv.” Compare this with (36).

(36) (Panh 1995:255) mingh mi’aqv go go-CH-O-ST

‘gone already’

Page 97: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

71

The seeming duplication of “go + go” is not a problem in a native speaker’s mind

due to a well-entrenched contraction of mi’aqv as an inseparable unit.50 It means

that mi’aqv is an independent construction by itself.

Formed in this way mi’aqv is a resultative aspectual marker. This concurs with

Bybee’s comment that “resultative constructions may be lexically restricted, or a

more generally used construction may have a resultative use only with change-of-

state verbs” (Bybee et al. 1994:65-6). The combination of the aspectual verb and

the aspectual particle is shown in Figure 16.

Aspectual Verb Perfective Aspectual Particle mingh aqv

Composition Entrenched

Resultative Aspectual Marker mi’aqv

Figure 16. Composition of the resultative aspectual

marker mi’aqv

In contrast to the subtle modal senses of aqv, the listener/hearer of a narrative can

be sure that an event is over and there is a definite result of it whenever mi’aqv is

uttered (except when addressed to the second person).

3.3.3 Modality

Modals include oix ‘want, like, will’ (37) and oix zuqc ‘must’ (lit. want + passive

or want + forced) (38). The latter construction sometimes varies to aqv zuqc in a

rapid speech. Both oix zuqc and aqv zuqc are commands when addressed to the

second person.

50 Some incidents of elicitation turned out that most native speakers of Iu-Mienh cannot identify the

etymological break down of mi’aqv; some can guess this might have come from the two words mingh + aqv. Panh (1995) is a native Iu-Mienh.

Page 98: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

72

(37) (3PG.009) Yie oix bieqc meih nyei biauv 1sg want enter 2sg POSS house

‘I want to come into your house.’

(38) (FA.011) Mingh aeqv, aqv_zuqc faix_fim go TOP must careful

‘When you go, you must take care .’

Not all modal uses of oix are desiderative, while a non-desiderative use has an

irrealis sense. An inanimate subject can take this modal verb as in (39).51

(39) Biauv oix mbaang house will collapse

‘The house is about to collapse.’

Example (40) is a consecutive speech consisting of two sentences, where the first

oix is non-desiderative (i.e. near future without volition) while the second is

desiderative.

(40) Ninh se gox oix daic haic niaa 3sg TOP old is_about_to die very I_wonder

‘He is so old that he is about to die.

Aengx oix mauv haaix_nhungc camv? further want desire what much

What else (what other things) does he want/desire?’

The non-desiderative use of oix correlates with futurity as a derivation from a

modal use. That is, it only refers to the future in a secondary sense after the

primary sense that an Agent wants to do something, which means that the thing to

be done lies ahead time-wise. This correlation between the future time reference

and the modality of the verb oix is represented in Langacker’s “basic epistemic

model” (1991c:242-44) (Figure 17).

51 Purnell (2006, personal communication) mentions that “oix can be treated as a future time marker”. Also

Burgess (personal communication).

Page 99: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

73

(Known) Reality Irreality

Immediate

Reality

C

(C = Conceptualizer or the speaker)

Figure 17. Basic Epistemic Model (Langacker

1991c:242)

What is wanted/willed has not come true yet; it is irrealis, and thus, naturally

belongs to a domain yet to come to the conceptualizer, i.e. future time reference.

Another case of cross-category interaction is the relationship between the

aspectual use and modal use of mingh ‘go’ and daaih ‘come.’ Court contends:

The verbs mingh and daaih [are] used as aspectual/evaluative/diectics components. These indicate a change of state, and by the choice of mingh one can indicate that the change is not desirable, or neutral, while daaih marks it as desirable. (Court 1986:248).

Since these senses of “desirable” and “not desirable” are determined in relation to

a conceptualizer or the speaker, it is suggested to refer to them as modal verbs. A

context of (41) is at the occasion of the arrival of a newborn child, hence daaih is

used with a desirable sense.

(41) (BS.091) Longc jienv buo hnyangx, yungz duqv dauh dorn daaih. Marry DUR three years, bear get CLF son come

‘Having been married for three years, (she) has given birth and got (happily) a baby son.’

In contrast to the “desirable” sense of ‘come,’ (42a) shows an undesirable

situation with the aspectual verb mingh ‘go.’

Page 100: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

74

(42a) (Court 1986:248)52 Duqv mi(ngh) aqv Get go PFT

‘He’s (gone and) taken it off (when he shouldn’t have)’ (translation by Court)

The contrastive sense will become clearer when compared with (42b):

(42b) (Court 1986:248) Duqv daac aqv (daac = a variant of daaih ) get come PFT

‘He’s taken it off (as he hoped he would)’ (translation by Court)

Another form with a range of meaning is duqv ‘to get, acquire, attain, be able to,

can.’ That is, its lexical meaning has extended to an aspectual use. Clark reported

on tau ‘get/acquire’ in Hmong, in which she recognizes a perfective aspect

(1989:217). Furthermore, Li (1991) advanced the argument that tau is an

“attainment aspect marker.”53 Duqv in Iu-Mienh has a similar aspect shown in

(43a-b).

(43a) (AS.138) Za'gengh! ninh yaac duqv zoux hlo Really 3sg also get be big

‘Really he also has become big.’ (i.e. ‘He reached/gained an important position.’)

That is, he attained the event of becoming an important person. Besides the

original sense of ‘to get,’ the perfective meaning in the sense of accomplishment

and achievement after some period is evident.

The verb duqv in (43b) does not take a verb as its complement but a noun phrase,

and yet still has a perfective sense.

(43b) (FA.043) Yie njaaux mh mbuo duqv i buo hnyangx aqv 1sg teach 2 PL attain two three years PFT

‘I have been teaching you both for two or three years.’

52 Though Court’s example lacks the context, I assume that he has drawn his conclusion from sufficient

examples. 53 Recently, Enfield (2001) has referred to this kind of polyfunctional construction as AQUIRE, “a marker of

complex descriptive complement constructions such as resultative, adverbial, and potential expressions” (Enfield 2001:268).

Page 101: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

75

The object of attainment by the verb duqv is the event of “two or three years.”

The result of attaining the event is naturally interpreted with the perfective

aspectual sense. It seems to be more suitable to adopt Li’s “attainment aspect” to

designate the Iu-Mienh equivalent duqv.

3.4 Topic Chains

Topic chains are “chains of clauses which share a common topical participant”

(Van Valin 2005:103). The primary topical participant in a clause is also termed

“topic subject” (Somsonge 2002:142). Once a topic subject is established in the

first clause, it will be realized by a pronoun or zero anaphora (Van Valin

2005:104). Since zero anaphora can be used frequently in topic chains in

discourse, the chains appear to be juxtaposed clauses, and it is sometimes difficult

to distinguish them from sequenced serial verb constructions. Aikhenvald

(2005:42), referring to such constructions as “sequence of actions,” distinguishes

them as follows: 1) In a sequence of actions, “A pause or an intonation break

indicating a clause boundary can appear between the verbs” (vs. SVCs cannot), 2)

a topic chain or a sequence of actions is compositional (vs. SVC as a single

predicate), 3) a topic chain encodes “several events in sequence which may be

conceptualized as connected to each other” (vs. SVC as one event), 4) it is “chains

of clauses” (Van Valin 2005) (vs. SVC as monoclausal).

In (44) the subject topic is in bold face; established by a pronoun in clause 1, then

carried over by zero anaphora in clauses 2 and 3 (the subject being a piglet

running away from a wolf in the story).

Page 102: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

76

(44) (3PG.046) Cl1

Ninh youc bangc wuov norm zanc 3sg also depend_on that CLF period

Cl2

Ø nqongh deix njiec daaih  it hurry some descend come

Cl3

Ø biaux nzuonx mingh  it flee return go

‘As for (the piglet), it used that occasion, came down quickly (from a tree), fled (and) went home.’ (i.e. ‘...it took the opportunity to quickly come down and run off/away home.’)

Each component may have its composite internal construction. Cl2 in (44) is an

adverbial SVC with a directional verb (i.e. [[nqongh deix]ADV.PH [njiec]V.MAIN]

daaihV.DR), and Cl3 is a directional SVC.

An even more complex structure can be found in (45). The serial verb

components within the clauses are in brackets. The topic subject of this sentence

is caah laauh porngc ‘grasshopper’ established in the preceding sentence

(FG.018), hence, (45) begins with the zero anaphora. The whole thing is a

multiclausal construction, consisting of topic chains. However the internal

structure is as follows: Cl1 is an adverbial clause marked by aeqv the topic marker

and an intonation break. Cl1 is dependent on Cl2. Cl2 is a switch-subject cause-

effect SVC followed by a short pause. Cl3 is another adverbial clause marked by a

topic marker and a pause, and is dependent on Cl4. Cl4 cannot be broken by a

pause; thus it is a switch-subject SVC.

Page 103: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

77

(45) (FG.019) Cl1

Ø bueix gau aeqv   [Pause] it sleep DPCL TOP

Cl2

[ndiangx-kou nauv] [haeqv caah_laauh_porngc] [Ø nyie ] [Short Pause] dead_tree snap frighten grasshopper it wake_up

Cl3

Ø mv jangx zuqc aeqv [Pause it NEG remember GOAL TOP

Cl4

[Ø sung ninh nyei zaux] [Ø ndamc zuqc norqc_meix nyei beuh ] it stretch 3sg POSS leg it stamp GOAL sparrow POSS goiter

[mbienv faaux jaang-ngunz mingh] overturn go_up nape go 1After (they) had been asleep for awhile, 2a dead tree snapped down (and) frightened (and) woke the grasshopper up, 3because (it) did not remember (keeping its legs bent), 4(it) stretched out its legs (and) kicked the sparrow's goiter hard (and the goiter) turned up to his nape.

An overall structure of this topic chain is that a sentence comprising Cl1 and Cl2

provides a setting for the succeeding sentence comprising Cl3 and Cl4. Thus Cl2

and Cl4 carry the story forward.

3.5 Summary of the Chapter

In chapter 3 we have surveyed the basic clause types, serial verb constructions,

tense-aspect-modality (TAM) and topic chains. A semantic extension is obvious

in a certain class of verbs (e.g. mingh ‘go,’ daaih ‘come,’ faaux ‘ascend,’ njiec

‘descend,’ duqv ‘get, attain, aquire, can,’ oix ‘like, want, will’ or ‘be about to,’ liuz

‘finish,’ baac ‘finish,’ ziangx ‘finish,’ nzengc ‘be consumed,’ jienv ‘be stable’).

They extend their original meanings to adverbs of direction and TAM. All these

verbs can be strung together into SVCs, multiclausal constructions and topic

chains. The sentence final particles, aqv (change of state) and mi’aqv (resultative)

provide additional aspectual and modal meanings. Thus the interaction between

TAM through verbs is frequent with no morphological inflection, implying a need

for a non-TAM based storyline theory. The SVCs and main clauses in topic

chains advance the narrative forward.

Page 104: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

CHAPTER 4

TEXTLINGUISTICS STORYLINE ANALYSIS

4.0 Introduction

Starting with a brief review of Longacre’s principle behind the storyline and verb

ranking analysis, this chapter will first present definitions of storyline from two

different perspectives, i.e. textlinguistics and cognitive linguistics. Secondly, the

chapter will discuss textlinguistic storyline and salience scheme analysis of Iu-

Mienh narratives, preceded by their profile analysis. In the course of doing so,

some rudimentary insights into cognitive-linguistic aspects of storyline analysis

will be pointed out. Thus, chapter 4 is a transitional chapter moving from a

textlinguistics storyline analysis to a cognitive linguistics storyline analysis.

4.1 Textlinguistics Storyline Analysis in Biblical Hebrew

The second edition of Joseph (2003a), a culmination of rigorous empirical

research, analytical techniques, comprehensive methodology and the holistic

theory of Longacre’s version of discourse analysis, stands on the basis of nearly

fifty years54 of his study of Biblical Hebrew among some eighty languages.55

Obviously Biblical Hebrew56 is one of the important languages for him, largely

54 The earliest mention, perhaps, of the Biblical Hebrew among his publications may be Longacre 1960,

where he illustrates hypertagmemic analysis of the clause types, predicate tagmemes and object tagmemes. He testifies that it was in 1944 at seminary that he first encountered so-called waw-consecutive as a “special narrative tense”, which sounded strange to him, through the Gesenius-Kautzsch-Cowley Hebrew grammar (1910) (Longacre 2003a:64).

55 Examples include New Guinea (Longacre 1972a), Colombia, Panama, and Ecuador (Longacre and Woods (eds.) 1976, 1977), India, Nepal, Vietnam, and the Philippines (Longacre ed. 1984), and East and West Africa (Longacre 1987, 1990).

56 But again why Hebrew? It was perfectly legitimate for Longacre to choose Biblical Hebrew to be one of many soils on which the study of narrative discourse should be cultivated. It is appropriate because of two-fold richness. First, it is because it provides abundant corpus: 484 chapters out of 912 chapters of the whole Hebrew Bible are entirely narratives (other 417 chapters are entirely poems or narratives that contain poetic forms and 11 chapters are written in Aramaic). Second, it is because no other narrative corpus in the world has undergone such an acute scrutiny of literary, linguistic, semantic studies, and textual criticism over two millennia as these texts. Thus, Hebrew of the Bible offers a laboratory where various methodologies can be tested.

78

Page 105: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

79

through which, if not exclusively, he constructed the theory of storyline and

salience scheme.

The first half of this section summarizes Longacre’s ten works on storyline

analysis in Biblical Hebrew (1961, 1979a, 1981, 1982b, 1985a, 1987, 1989a,

1994a, 1995, 2003a). 57 These works have been compared with five recent

authors, and a summary of the comparison will be presented in the second half.58

First, having reviewed these ten works, three premises concerning Hebrew

narrative discourse can be summarized as follows:

(1) The preterite (i.e. “wāw plus prefixal verb”) is the special narrative tense59 for the back-bone of a story (viz. a storyline tense/verb). It is this tense of a verb that pushes the story forward.

(2) A distinction between the foregrounded event-line (i.e. storyline) and supportive material is binary. An implication of this premise is that once the storyline tense is identified, all the other supportive materials will be (almost automatically) eliminated as “off-the-line” materials into a background from the “on-the-line” material.

(3) On the other hand, there is a “spectrum” (i.e. “ranking” or “cline”) across the supportive materials.

Concerning the third point, Longacre states:

Possibly Biblical Hebrew narrative is one of the clearest places to posit with

confidence a spectrum which involves considerable diversity of verb and clause

structure (1981:341) [underline added]

Second, comparing the points above with Buth 1995, Endo 2001, 2003,

Fokkelman 1996, Heimerdinger 1999, Levinsohn 2002, four things can be

summarized as follow:

(1) The prototypical function of the preterite (wayyiqtōl form) verb is to keep

sequentiality in recounting past events in a narrative.

57 One should be reminded, however, of the fact that Hebrew is not the only language Longacre’s theory is

built on. 58 For details see Arisawa 2006. 59 By assigning the term “preterite,” which is non-aspect, Longacre set a right course for the future analysis of

Hebrew verb in discourse. See Blau who argues “that in Biblical prose the Hebrew verbal system denotes tenses, rather than aspect” (1998b:208-10).

Page 106: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

80

(2) The sequential function of the preterite can be extended to some non-

sequential functions off the storyline pragmatically by the author’s

intention.

(3) The storyline emerges as a foreground phenomenon because of the

sequentiality expressed by the sequential verb wayyiqtōl, not that this

particular form is the storyline verb/tense or foregrounding verb/tense.

(4) A mechanical application of storyline verb form as a measuring stick to all

cases of occurrence in order to single out the storyline may either lump

subtle semantic differences within a category into a rough grouping or put

an exegetical process of texts into a peril of eisegesis.60

A difference between Longacre and the others is as follows: Longacre emphasizes

the importance of dependence on the preterite verb form in identifying the

storyline, while others see the preterite’s primary function as the sequential verb;

thus the foregrounded storyline is an outcome of the verb’s sequentiality.

4.2 Definitions of Storyline: Two Approaches

In this section the definitions of storyline from different authors are summarized

into two groups: one is from a textlinguistics perspective, another is from a

cognitively oriented approach.

4.2.1 Textlinguistic Definition of the Storyline

Summarizing Hopper’s analysis of foreground and background (1979), his

version of story line can be defined as follows in the present author’s words:

A definition of storyline (1)

The story line is foregrounded events that are succeeding “one another in the narrative in the same order as their succession in the real world,”61 and that are

60 Waltke and O’Connor say in a footnote, “In reality, a discovery procedure is usually an inductive extension

of a definition. Any linguistic analysis results from much more than mechanical application of fully specifiable technique; discovery-procedure linguistics tends to be a form of empiricist extremism” (Waltke and O’Connor 1990:56).

61 Hopper (1979:214).

Page 107: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

81

denoted by verbs with preterite tense or perfective aspect. The foregrounded story line is in a binary opposition to the background events, which are concurrent with the foregrounded events, and are marked by imperfective aspect of the verbs.

Hopper and Thompson (1980) hypothesized that there is a correlation between the

high transitivity of the foreground, and the low transitivity of the background.62

Their ten transitivity parameters are displayed in Table 16.

HIGH LOW A. PARTICIPANTS 2 or more participants, A and O 1 participant B. KINESIS Action non-action C. ASPECT Telic atelic D. PUNCTUALITY Punctual non-punctual E. VOLITIONALITY Volitional non-volitional F. AFFIRMATION Affirmative negative G. MODE Realis irrealis H. AGENCY A high in potency A low in potency I. AFFECTEDNESS OF O O totally affected O not affected J. INDIVIDUATION OF O O highly individuated O non-individuated

(A = Agent, O = Object)

Table 6. Transitivity parameters (Hopper and

Thompson 1980:252)

Longacre, however, questions if transitivity is the only parameter to identify the

foreground, that is, the storyline. For him sequentiality is also important. He

states, “I have often thought it somewhat curious that, although their [Hopper and

Thompson 1980] parameters apply primarily to narrative, a further parameter,

sequentiality (sequenced vs. nonsequenced) was not included” (Longacre

1996:24). Furthermore, he asserts, “Sequentiality must in fact be taken as [a]

prime characteristic of the storyline, i.e., of foregrounding in narrative”

(1996:25).63

62 They characterize the foreground (i.e. storyline) with two points. “First, the foregrounded portions together

comprise the backbone of the text, forming its basic structure; the backgrounded clauses put flesh on the skeleton, but are extraneous to its structural coherence. […] Second, the foregrounded clauses […] are ordered in a temporal sequence; a change in the order of any two of them signals a change in the order of real-world events. Backgrounded clauses, however, are not ordered with respect to each other, and may even be movable with respect to the foregrounded portions” (Hopper and Thompson 1980:281).

63 These two passages seem to be the only places where Longacre emphatically states the importance of sequentiality. Interestingly, he does not relate this importance of sequentiality with wayyiqtōl in Hebrew, though it has been pointed out elsewhere by other researchers (e.g. Buth 1995, Endo 1996, 2003, Myhill

Page 108: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

82

Thus he has included sequentiality as an important semantic value of the storyline

in the salience scheme for English narrative (1995c:7, 1996:26) shown in Figure

18 as follows:

A. B. C. D. E. F.

Actions Motions Cognitive events Contingencies

Band 1

Punc

tilia

r

“Hap

peni

ngs”

Sequ

entia

l

Non-punctiliar (activities) Band 2

Dyn

amic

Non-sequential (flashback, projection) Band 3

Rea

lis

Non-dynamic (= static) Band 4

Nar

rativ

e

Irrealis (modals, some negatives) Band 5

Subs

tant

ive

Non-narrative (author evaluation and comment) Band 6 Non-substantive (= cohesive) Adv. Clauses in backreference Band 7

Figure 18. The parameters for salience scheme for

English narrative (Longacre 1996:26)

Collected from various publications by Longacre (e.g. 1995c:6-8;1996:18, 27,

passim), his version of storyline, particulariy in the framework of textlinguistics,

is defined as follows:

A definition of storyline (2)

The storyline is the most foregrounded main line of narrative discourse whose semantic values are substantive, narrative, realis, dynamic, sequential and punctiliar, and whose morpho-syntactic marking is the preterite or completive verb form.64

As observable from the above summaries, while both Hopper and Longacre

developed the storyline/foreground theory with its correlaton to transitivity, one

difference is that Longacre acknowledges the importance of sequentiality as a

property of storyline.

This point has been recognized by some Asian scholars and some who have

studied languages of Africa (Follingstad 1994, cf. 2.1.8) or Semitic languages

2001:168). In his second edition of Joseph (2003a) he has changed the term BU (build up) of the 1989a edition assigned to wayyiqtōls to ST (Sequential Thesis) without reference to these scholars.

64 This statement is constructed by the present author from several definitions of storyline by Longacre.

Page 109: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

83

(Endo 1996). For example, Somsonge (2002) defines the foreground of a

narrative (viz. storyline) as follows:

A definition of storyline (3)

The clauses which refer to sequenced points on a timeline and present new information, rather than referring anaphorically to happenings that have already occurred, constitute the foreground of a narrative. The foregrounded clauses are thus defined as clauses in which the action of the narrative takes up and time begins to move forward. (Somsonge 2002:142)

In this statement the important thing is that she points out that sequentiality is an

essential part of the storyline. Moreover, the presentation of new information is

also captured as a component of it. To repeat this, two important components,

namely, the sequentiality on a timeline and the increment of new information,

constitute the foreground of a narrative, that is, a storyline.

While the textlinguistic storyline analysis in Iu-Mienh in 4.3 is based on the

definition (2), Somsonge’s definition will be incorporated into a re-definition of

storyline in terms of CL in 4.2.2.

4.2.2 Cognitive Linguistic Definition of the Storyline

Dry (1992) has reviewed various authors’ different understandings about storyline

and the foreground-background issue, and categorized them into two large groups

of elements that characterize the foreground. One group views the foreground

from importance (i.e. thematic, human, causal, formal (or timeline eventline)); and

the other group views foreground from salience (i.e. from unexpectedness, figural

properties and cognitive accessibility). It is the latter type of foreground that

provides the present study with the definition of storyline from a CL perspective.

Together with Dry’s reminder that the very notion of foreground is originally from

“gestalt theories of perception” (1992:436), two more concepts should be

recognized as important: the cluster concept (1992:441) and the continuum

concept (1992:445, Hopper and Thompson 1980).

Dry has developed the concept that foreground markings are best described as a

cluster concept, proposing it based on Longacre’s idea of the “bag of tricks”, i.e.

Page 110: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

84

many different items that are available to mark discourse peak (Dry 1992:441). In

a similar vein, foreground (i.e. storyline) can be marked by a variety of markers,

forming a cluster of markers. Furthermore, different marking features in the

cluster are in the continuum relation. This understanding is feasible and plausible

when the foreground (i.e. storyline) is perceived as a cognitive phenomenon. The

foreground (storyline) in this view is consistent with Buth’s and Endo’s

understanding of the sequentiality factor. For instance, Endo states, “The

distinction between ‘foreground’ vs. ‘background’ as a result of observation rather

seems to be a phenomenon triggered by the ‘sequentiality’ of verbs” (Endo

2003:88). That is, the foreground is the perceptive outcome of a phenomenon that

events expressed by verbs are strung in a sequence: not that it exists ontologically

and should be detected by identification of its marks.65 When recognized in this

way, the problem of missing the reality of language by limiting analysis to

morpho-syntactic markings, cautioned by Waltke and O’Connor (1990), will be

safely avoided. This point is summarized by Dry also as follows:

Many of the markers identified have aspectual meaning as well as pragmatic function, and this may bolster the identification of foreground with temporally successive clauses, while the identification of specific markers may lead away from conceptualizing foreground as relative” (Dry 1992:448).

In other words, the identification of foreground or storyline is not absolute nor

mechanical but relative. A definition of storyline from the CL perspective,

therefore, is proposed as follows:

A storyline is a perceptually foregrounded line of a narrative discourse. The line is foregrounded by two major causes: the sequence of events and the movement of events along a timeline toward its goal or conclusion, which is pragmatically planned by the narrator. The former is referred to as sequentiality of events, the latter transitivity of events. Thus, storyline consists of sequentiality and transitivity of events.

65 As it may be noticeable, sequentiality comes first in CL perspective, and then the foreground is identified

subsequently. Once this order is recognized, such an opinion seen as follows may be eased to some extent: “Endo’s attempt to take sequentiality as basic rather than foregrounding initially impressed me as a mere terminological squabble. After all, foregrounded clauses in narrative are sequential. Foregrounding is a general term which is extendible to other discourse types (e.g. expository or descriptive) where sequentiality is not a concern” (Longacre 1998:217).

Page 111: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

85

Note the expansion of the meaning of the term “transitivity” used in the above

definition. Traditionally, as in works in Halliday (1967-8, 1970:146-55,

1994:106-75) or Hopper and Thompson (1980), Hopper and Thompson eds.

(1982), transitivity means a transition of energy from an Agent to a Patient/Object

within a clause. However, in the above proposal in the present study, the term is

used to capture the energy transition both in and across clauses, hence an event

transition at the inter-sentence level as well. Thus, the definition is expanded

from the standard use so it will be fitted to an analysis of strings of the serial verb

constructions (SVCs) and topic chains in 5.3.

4.2.3 Summary

Various understandings from different disciplines on the nature of storyline can be

summarized into two fundamental attitudes. One is an ontological approach to the

storyline. In this approach, the the storyline is assumed to exsist first, and the

linguist’s responsibility is to discover it through an identification of the most

foregrounded line of development, particulary, an identificaton of a marking on a

verb for preterite tense or perfective aspect, and an application of such a verb to

text corpus to confirm or check the initial identification. This spiral procedure of

finding the storyline will continue until the non-storyline linguistic constructions

are ranked in the order of the most dynamic to the most static in a salience

scheme.

Another approach to the storyline is epistemological. In this view, the storyline,

which is foregrounded, is a result of perceiving the sequenced events and

dynamically moving events along the timeline. Just as grouped dots arranged in

sequence emerge to form a line, so the sequenced events in a narrative also

cognitively stand out to the observer’s perception as a foreground, rather than

disconected items. Moving events or a transitional relation from clause to clause,

sentence to sentence, has more vividness to the perception of an observer of a

narrative. Thus, once the narrative is set on a theater stage as a cognitive entity,

Page 112: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

86

the storyline is explained as a foregrounded line of narrative facilitated by the

transitivity of events and sequentiality of events.

This subsection is summarized by comparing two different opinions of one

linguist in his change with regard to this issue. Compare the Givón’s opinion in

1984 quoted in 1.1.3 with the revised one quoted below:

Like all discrete, binary distinctions we have constructed in the past, the foreground/background distinction is both useful and dangerous. It is useful in carrying us the first step toward a function-based definition of an important strand in the thematic coherence of discourse. It is dangerous if we wed ourselves to it rigidly and do not eventually trade it in for more elaborate, more specific, less circular and empirically better grounded notions (Givón 1987:185).

4.3 Textlinguistics Storyline Analysis in Iu-Mienh Narrative

In this section, a profile analysis of the seven stories and a salience scheme of Iu-

Mienh narrative will be discussed.

4.3.1. Profile Analysis

A profile of narrative discourse is analysed in terms of its surface and notional

structures. The notional structure is “etic and heuristic devices to uncover emic

structures in given languages” and “a schema on which climactic narrative

discourses are built” (Longacre 1996:34). The notional structure is manifested in

the surface structure, expressed by various linguistic devices.

A profile of Story 1 (OH) is as follows:

Page 113: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

87

Surface Structure

Sent. No.

Sentence at the boundary Notional Structure

Aperture A story of an old husband and bamboo shoots Surface features only

Stage 1 (1) There was a very very old man who tried to force his daughter to marry that old man.

Exposition

Prepeak Episode 1 2-4 (2) His daughter didn’t like the idea. Inciting moment Prepeak Episode 2 5-8 (5) As for his daughter, thinking about that old

man’s wealth, yes, it’s true, but too old to be suitable as a husband.

Peak episode 9-15 (9) Then after that, his daughter came up with an idea and boiled old bamboo shoots which no one could chew.

Developing conflict

Postpeak episode 16 (16) The daughter said, Climax Closure 17 (17) As his daughter had her strategy, he could

not persuade her any longer. Therefore she did not have to marry that old man.

Conclusion

Finis none

Figure 19. A profile of Story 1: A story of an old

husband and bamboo shoots

Usually the Aperture is a formulaic phrase, which appears only on the surface

structure (Longacre 1996:36). However, in this short story, the Aperture is only a

title of the story.

Story 2 (FG) exhibits a profile as follows:

Surface Structure Sent. No.

Sentence at the boundary Notional Structure

Aperture 1 A story of Firefly and Grasshopper Surface features only

Stage 2-3 (2) In the old days, Firefly and Grasshopper exchanged work

Exposition

Episode1 4-9 (4) When they worked it became dark, then they talked about it.

Inciting moment

Episode 2 10-12 (10) They competed to see who would get home first.

Pre-Peak episode 13-18 (13) Grasshopper didn’t reach his house, so on the way he stayed Sparrow’s house.

Developing conflict

Peak episode 19-27 (19) As they slept, a dead tree snapped and woke Grasshopper up…

Climax

Postpeak episode 28–30 (28) This time they brought law to discuss. Denouement Closure 31-35 (31) Today, if don’t believe, do look at an

earthworm’s neck and check it. Conclusion

Finis None

Figure 20. A profile of Story 2: A story of firefly and

grasshopper

Page 114: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

88

Story 3 (3PG) has the following surface structure:

Surface Structure Sent. No.

Sentence at the boundary Notional Structure

Aperture 1 (1) Now I am going to tell a story of a wolf and piglets.

Surface features only

Stage 2 (2) In the old days, a female pig gave birth to three piglets. Exposition

Ep 1 3-16 (3) After bringing them up to big enough, she told them, "Each of you must go and find your own place to live."

Inciting moment

Ep 2 17-25

(17) The wolf took the first piglet and ate it; and again he went to the second piglet's house and told him again like that, "You, open the door to let me in."

Ep 3 26-33

(26) (The wolf) further went to the third piglet's house, (and) called the third piglet to open the door to let him come in.

Ep 4 34-38

(34) He again planned to deceive the piglet, enticed to say, "Let's go and dig potatoes tomorrow."

Prepeak Episodes

Ep 5 39-47

(39) He again planned to deceive the piglet, (saying) "Let's go and collect some fruits, tomorrow; collect apples."

Ep 6 48-58

(48) He again planned to entice the piglet, (saying) "Let's go and see the fair tomorrow."

Ep 7 59-64

(59) And then the black wolf returned following the piglet.

Developing conflict

Peak episode 65-70

(65) The roof of the piglet's house has a hole; a hole that they made to let smoke go out. Climax

Closure 71 (71) There, that the piglet's plan really tortured that wolf to death. Conclusion

Finis none Surface features only

Figure 21. A profile of Story 3: A story of three piglets

The prepeak episodes consist of two cycles. One is a cycle of the wolf’s eating of

two piglets and a failure to eat the third piglet (Ep 1-3). The second is from Ep 4

to 6 where the wolf tried to deceive the third piglet and failed. The second cycle

is appended with a conversation between the two and the wolf’s failure is

emphasized (Ep 7), which suspends the story and prepares for the peak.66

66 Concerning the Western- (and Thai-) loaned folk tales, Jennings has pointed out that “the pattern of ‘three

little animals’ is a Western discourse feature. The Lao-Thai loans would not be found in Mienh stories in China and Vietnam” (personal communication 2006). A higher level plot scheme which is unique to the Iu-Mienh is an interesting topic of study, whilst a difficulty of excluding elements of language contact remains. With regard to the pattern of three, Purnell points out that “three daughters as primary actors is also found in the Big Snake story” (personal communication).

Page 115: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

89

A profile of Story 4 (CI) is presented as follows:

Surface Structure Sent. No. Sentence at the boundary Notional

Structure Title A story of why cicadas don't have

intestines Aperture 1-7 (1) I am going to tell you a story.

Surface features

only

Stage 8 (8) In the old days, an owl told a barking-deer; "Olive fruits are ripe, Olive fruits are ripe," deceiving the barking-deer to eat some.

Exposition and inciting moment

Ep 1 9-16 (9) The barking-deer came there to eat those olive fruits.

Ep 2 17-26 (17) This time, a wild-chicken scratched the ground carelessly and hit ants accidentally making them to come out.

Prepeak Episode

Ep 3 27-59 (27) Talking about that squirrel, he got very angry criticizing that the snake took his place to live.

Developing

conflict

Peak episode 60-63 (60) This old man, therefore, has come to the conclusion there, found out the root (of the problem)…

Climax

Postpeak episode 64- (64) Talking about the cicada, Denouement Closure 65-72 This story is like (I) have told, and... Conclusion Finis 73-78 (73) Now, one person has one story, is't that

right? Surface features only

Figure 22. A profile of Story 4: Why cicadas don’t have

intestine

This story is an unedited one; it has a long Aperture and Finis. They are outside

the narrative text, though socially or interpersonally they may have some more

meanings.

Page 116: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

90

Story 5 (FA) is profiled as follows:

Surface Structure

Sent. No. Sentence at the boundary Notional Structure

Title A story of Faam-Bae’ and Aeng-Doi Aperture 1 (1) I am going to tell a story of Faam-Bae’ and

Aeng-Doi. Surface features only

Stage 2-5 (2) In the old days, there was a man whose name was Faam-Bae’. Exposition

Ep 1 6-11 (6) She couldn't do anything, but she made a plan. Inciting moment

Ep 2 12-25 (12) She didn't know that Faam-Bae’ was going (to study).

Ep 3 26-34 (26) They indeed arrived at the Meisen school. Ep 4 35-43 (35) After studying for three years, her parents

at home felt that their daughter had gone very long and she had never returned home.

Ep 5 44-60 (44) The next day, her teacher said, Ep 6 61-80 (61) When Aeng-Doi returned home, she took

off men's clothes and put on women's clothes, as a beautiful lady.

Ep 7 81-89 (81) (When he) returned home, his heart was sad and only wept without eating (anything).

Prepeak Episode

Ep 8 90-99 (90) The next day, his mother and others really went to look for; as they arrived they found that Maajaa had arranged for a marriage with her first.

Developing conflict

Peak episode 100-113 (100) And then when it was a day of wedding, Maaja and the relatives were taking Aeng-Doi back home (as a bride).

Climax

Postpeak episode

114-118 (114) That is, in the past, from the beginning Aeng-Doi had known that she was destined to marry Faam-Bae’ but she did not tell him anything.

Denouement

Closure 119-120 (119) Later, (they) really married. Conclusion Finis none

Figure 23. A profile of Story 5: A story of Faam-Bae’

and Aeng-Doi

Page 117: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

91

A profile of Story 6 (AS) is presented below:

Surface Structure Sent. No.

Sentence at the boundary Notional Structure

Aperture/Title Yun Zoi narrated Surface features only

Stage 1-3 (1) There was a couple who were very poor. Exposition Ep 1 4-15 (4) As they went, it became dark. Inciting

moment Ep 2 16-28 (16) Then the couple discussed. Ep 3 29-46 (29) Then another day, like the river down there

where we ride a boat, there was a kind of crocodile, which was big.

Ep 4 47-64 (47) Then another day, there was a gigantic snake, which fell into a city water source for all the residents.

Prepeak Episode

Ep 5 65-89 (65) And then another day, there was a disturbing tiger.

Developing conflict

Peak episode 90-130 (90) Later, the country became unsettled again, like people here have a war nowadays.

Climax

Postpeak episode 131 (131) The officer further let people go to get guns and so forth back.

Denouement

Closure 132-134

(132) Then this time, (the officer) assigned one corner of a region with one capital city for him to make a living.

Conclusion

Finis 135 (135) I have told you. Surface features only

Figure 24. A profile of Story 6: A story of Aahan

It seems that the tape-recorded data used to have a longer aperture than this.

Page 118: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

92

Finally, a profile of Story 7 (BS) is shown in Figure 26:

Surface Structure Sent. No.

Sentence at the boundary Notional Structure

Aperture none Surface features only

Stage 1-2 (1) In the old days, there was an old man who had seven daughters and he raised them up. Exposition

Ep 1 3-5 (3) Because their father saw that they were lazy, he then went to look a piece of land. Inciting moment

Ep 2 6-7 (6) The following day, then, his daughters went. Ep 3 8-58 (8) The following morning, after having a

breakfast, their father shapened his axe; carrying it on his shoulder to chop a tree.

Ep 4 59-90 (59) The following forming, after eating breakfast, then, their father told the First daughter, saying "Wrap your blanket, and follow the big snake to be his wife.

Ep 5 91-120 (91) Having married for three years, (she) bore and got a baby son.

Ep 6 121-139 (121) At night when they went to bed, her husband then asked her,

Ep 7 140-178 (140) This time, he has got a permanent servant for the house.

Ep 8 179-198 (178) The following morning, his wife get up carring the baby; before she washes his face, the baby cried.

Ep 9 199-210 (199) Later when this bamboo grew tall, its middle part became swollen up; so he did not cut it.

Prepeak episode

Ep 10 211-243 (211) Later a son of the shop owner went that old woman's house for a visit.

Developing conflict

Postpeak episode 224-255 (224) He then told that old woman to kill a pig to have a birthday feast.

Climax

Closure 256-258 (256) As for her elder sister, she was boilt to death; she didn't become anything.

Conclusion

Finis 259 (259) The story of the Big Snake has finished. Surface structure only

Figure 25. A profile of a story of big snake

4.3.2 An Etic Salience Scheme Template

As a working tool for doing a storyline analysis, an “etic salience scheme” is

proposed by Longacre. He explains that “[t]his etic salience scheme can serve as

a beginning frame of reference for analyzing verb/clause functions in narrative,

and so served for a number of languages in various parts of the world” (1996:26-

7), shown in Figure 25.

Page 119: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

93 1'. Pivotal storyline (augmentation of 1) 1. Primary storyline (S/Agent > S/Experiencer > S/Patient 2. Secondary storyline 3. Routine (script-predictable action sequences) 4. Backgrounded actions/events 5. Backgrounded activity (durative) 6. Setting (exposition) 7. Irrealis (negatives and modals) 8. Evaluations (author intrusions) 9. Cohesive and thematic

Figure 26. An etic salience scheme for narrative

(Longacre 1996:28)

Two modifications are needed for Iu-Mienh: first, concerning “flashback”, second,

“Routine.” Longacre explains “flashback” in this salience scheme for narrative

“as an ill-defined category, it [flashback] can group with [Band] (2) or (4); as a

well-defined morphosyntactic category it can be added after (5)” (1996:28). Iu-

Mienh flashback is not clearly defined if one follows Longacre’s explanation.

However, a judgment from the context leads to the conclusion that the flashback

band in Iu-Mienh has its own position between (5) and (6).

Another modification of the etic salience scheme is the “Routine.” This was

posited by Longacre to make the etic template operative to capture such materials

of Halbi, and often Indo-European language (1996:27-8). However, it is not

relevant to Iu-Mienh. As a result, a modified etic salience scheme for Iu-Mienh is

presented in Figure 26 as follows:

1'. Pivotal storyline (augmentation of 1) 1. Primary storyline (S/Agent > S/Experiencer > S/Patient) 2. Secondary storyline 3. Backgrounded actions/events 4. Backgrounded activity (durative) 5. Flashback 6. Setting (exposition) 7. Irrealis (negatives and modals) 8. Evaluations (author intrusions) 9. Cohesive and thematic

Figure 27. An etic salience scheme for Iu-Mienh

narrative

Page 120: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

94

4.3.3 Iu-Mienh Salience Scheme Analysis

A Iu-Mienh salience scheme is obtained by applying the etic salience scheme

posited in 4.3.2 (Figure 27) to the seven stories; it is presented in Figure 28.

Following Longacre’s “peeling off” method, the analysis begins with the most

remote band, viz. the cohesive band, going up to the storyline band.

Page 121: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

95 Band 1′. Pivotal storyline

1′.1 Sentence + mi’aqv (Resultative aspect) 1′.2 Sentence + aqv (Perfective aspect) 1′.3 za’gengh+V (Realis adverb ‘really, indeed, actually’)

Band 1. Primary storyline

1.1 Unmarked action (dynamic) verbs 1.2 aengx+V (Developmental adverb) 1.3 S + ziouc + V (Sequential marker ‘and then,’ ‘after which’, ‘then soon’) 1.4 Serial Verb Constructions (SVC) including: Aspectual SVCs (V+daaih ‘come,’ V+mingh ‘go,’ V+ ziangx ‘finished V-ing,’ V + nzengc ‘completely V or

‘be consumed’) Directional SVCs (V+daaih ‘come,’ V+mingh ‘go,’ V+faaux ‘ascend,’ V+njiec ‘descend’)

1.5 Topic Chains (multiclausal constructions sharing the same topic-subject) 1.6 Repetition of unmarked verbs

Band 2. Backgrounded actions/events

V1+jienv (in the construction V1+jienv+V2, where V1+jienv- indicates prolonged, repetitive, ongoing, simultaneous action to modify V2)

Band 3. Backgrounded activity (durative)

3.1 V+V+V (repetitive), V+neyi+V (repetitive) 3.2 V+jienv (Durative aspect) 3.3 Mimesis/Onomatopoeia + nyei+V (descriptive),

Band 4. Flashback

4.1 Unmarked verb 4.2 wuov_zanc ‘that time,’ loz-hnoi ‘old days’ or ‘in the past,’ m’daaih ‘from the beginning’

Band 5. Setting (exposition)

5.1 maaih+Subject (Presentational) 5.2 Stative verb 5.3 VST+nyei (present state aspect) 5.4.1 (Eq1) A benx B ‘be’ or ‘become’ 5.4.2 (Eq2) A zoux B ‘be’ or ‘make’ 5.4.3 (Eq3) A zeiz B ‘be_correct’ 5.4.4 (Eq4) A dongh B ‘the same as’ 5.4.5 (Eq5) A se B ‘that_is’ 5.5 Neg+Vst (Negation of antonym in a synonymous parallel to a positive stative verb)

Band 6. Irrealis (negatives and modals)

6.1 Neg + dynamic/action verb 6.2 Negative idiom 6.3 Modal. oix+V ‘want to do’ 6.4 Unmarked verb in a complemental VP of a causative verb

Band 7. Evaluations (author intrusions)

7.1 1st person pronouns yie ‘I’, mbuo ‘we’ (inclusive) 7.2 Empathy with participants/audience by an interjection 7.3 Marginal clause hnangv…nor ‘like, as’ taking an illustration from audience

Band 8. Cohesive and thematic

8.1 Cl+aeqv (Topic marker) ‘as for’ 8.2 Cl+nor (Topic marker) ‘talking about’ 8.3 V+gau (Indpt. Cl marker ‘and then’ 8.4 V+liuz ‘after finishing V’ 8.5 Tail-head link

Figure 28. Salience scheme for Iu-Mienh narrative

4.3.3.1 Band 8: Cohesive

The most obvious band that does not move the story forward is the cohesive band.

The materials in this band repeat what is stated previously. There are two groups

Page 122: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

96

here: one is marked, the other not. In the first group, the adverbial clauses are

marked by one of four particles. Two particles are clause final, forming the

structure as in (46) and (47). Both can function as a topicalizer. When the clause

is topicalized, its meaning can be interpreted as the temporal setting, reason, cause,

and condition. The adverbial clause in (46) indicates the temporal setting, and Cl2

in (47) the reason.

(46) (OH.010) Cl-adverbial

zouv ziangx aeqv cook finish TOP

‘(After) finished cooking,

Cl-main

mbenc daaih bun ninh nyei die nyanc prepare come give 3sg POSS father eat

(she) prepared (food) for her father to eat.’

One more construction of this kind is Cl + nor ‘talking about’ or ‘as/when/since’

as in (47).

(47) (BS.078) Cl1 Cl2

Mingh gau taux diuh ndoqv nor, go DPCL reach CLF stream as

‘When (they) went, as (they) reached a stream,

Cl3

ninh gorngv mbuox ninh nyei auv 3sg speak tell 3sg POSS wife

it (the snake) spoke (and) told its wife

Cl4

yiem wuov ndoqv-hlen zuov ninh be_there there stream-side wait_for 3sg

(that she should) stay there at the side of the stream (and) wait for him.’

The other kind of clauses have a post verb markers, forming the construction V +

gau ‘having done V and then’ or ‘when/as S V-ed…’ and V + liuz ‘after finishing

V’. All these four constructions precede the main clause providing the main

clause with the old information from the previous sentence. A surface structural

difference between the first two constructions and the latter two is that aeqv and

Page 123: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

97

nor are clause final while gau and liuz are clause internal, placed immediately

after the verb.

Example (48) shows gau functioning as a dependent clause marker (DPCL):

(48) (FG.003) Buic gau gong exchange DPCL work

wuov_hnoi ninh_mbuo gorngv taux ninh_mbuo ganh nyei za'eix that_day 3_PL speak reach 3_PL self POSS plan

‘When/as (they) took turns working for each other, that day they talked about each other's plan.’

Example (49) shows that the aspectual verb liuz occupies the same position as gau

forming an adverbial clause, which precedes the main clause.

(49) (BS.049) Adverbial Clause

Nyanc liuz hnaangx eat finish rice

Main Clause

ninh die aengx heuc Guh_Meix dimc dorngx 3sg father again order the_first_daughter overlay place

bun domh naang bueix give big snake sleep

‘After they finished having a meal, their father told again the first daughter (to) lay out [a bed or mattress on] the place [for the snake to sleep on] (to) let the big snake (to) sleep.’

One more structure, which has no identifying particle, is ‘tail-head’ linkage. It

repeats the last word or phrase of the previous sentence before the main clause as

in (50).

Page 124: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

98

(50) (3PG.007) Aengx maaih dungz-dorn nyei za'eix longc hui-zun gomv biauv and_then have piglet POSS plan use brick build house

‘Another piglet's plan was 7.2(that he) built a house with bricks.’

(3PG.008) Gomv jienv mv nangc lauh build DUR NEG very long

maaih dauh hieh_juv_jieqv daaih nduov ninh_mbuo nyanc have CLF wolf_black come deceive 3_PL eat

‘Building (it) not very long, there was a black wolf (and it) came (to) deceive them (to) eat.’ (i.e. ‘They had not been building it for very long when a black wolf came (and tried to) trick (them) so he could eat them.’)

4.3.3.2 Band 7: Evaluations

The evaluation band includes author intrusions by using the first person pronouns

yie ‘1sg’ and mbuo ‘1pl’ (‘we’ inclusive). Besides the ordinary use of the first

person plural yie_mbuo ‘1_PL’ which has exclusive meaning, mbuo sometimes has

inclusive meaning. When the inclusive ‘we’ uttered by a narrator is directed to

the audience, the narrator’s intention is to involve the audience in a story (50).

Such an involvement or non-forcing invitation may be called a narrator-audience

tie. This tie is sometimes expressed by an interjection of empathy ov ‘oh!’, aah

‘ah!’ This interjection is directed from the mouth of a participant in the narrative

toward the audience; and sometimes it is addressed to the audience directly from

the narrator. The marginal clause preposition hnangv ‘like’ or ‘as’ is used to draw

an illustration from the audience’s world or experience and incorporate it in the

narrative (51).

(51) (AS.029) 1 2 3

Da'nyeic hnoi hnangv wuov ndiev ndaaih naaic, mbuo geh jakv second day like that under river there 1pl ride boat

4 5

maaih douh_taanh yietc nyungc, hlo nyei have big_lizard one kind big PRS.ST

‘1Then another day, 2like (in) the river down there .3(where) we ride a boat, 4there was a kind of crocodile, 5(which) was big.’

Page 125: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

99

These rapport-building techniques are highly pragmatic. Though they are far

down from the primary storyline, they certainly contribute to the vividness of the

narrative and help the audience to be drawn to continue to listen.

4.3.3.3 Band 6: Irrealis and Modals

Negatives and modals belong to this band. Negatives include a simple negation of

dynamic/action verbs and idioms which contain the negator maiv ‘not’ (or its

shortened form mv) as in maiv_haih_fungc_zoux ‘there is nothing one can do

about it.’ The idiom that contains a negative marker is used more frequently when

the story comes near the peak, in the process of accumulating tension.

The modals oix ‘want’ and haih ‘can’ or ‘be possible’ do not push the story

forward. These modals refer to possibility but not actual happenings, i.e. irrealis.

Another use of verbs which do not carry the storyline ahead should be mentioned.

The regular verbs in a complementation of the main verb do not realize the action

described in them. In (52) a complement is in brackets. After the actual

happening of heuc ‘call’ or ‘order’, all the following verbs buov ‘burn’, hnyutv

‘boil’, bun ‘let’ and nzaaux ‘wash’ are not realized yet.

(52) (BS.032) Aengx heuc [Guh_Meix buov douz hnyutv wuom further order the_first_daughter burn fire boil water

bun domh-naang-gorx nzaaux sin.] let big-snake-elder_brother wash body

‘(Father) further ordered the first daughter (to) make a fire (i.e. make a fire) (and) boil water (to) let the big snake elder brother take a bath.’

4.3.3.4 Band 5: Settings

In this band, there are four kinds of expressions. The first is a presentational

construction using an existential verb maaih ‘to have, to be, there was…’ This

verb, in the SVO order, means ‘to have’ or ‘to possess.’ However, at the

beginning of a story, or at the episode boundary inside a story, it “serves to

introduce a new participants into the discourse. The subject NP appears in the

Page 126: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

100

postverbal position normally reserved for objects, the unmarked focus constituent

in a predicate focus construction” (Van Valin 2005:71). See (53).

(53) (FA.002) Loz-hnoi maaih dauh m'jangc_dorn heuc Faam-Baeqv in_the_old_days there_is CLF man call Faam-Bae’

‘In the old days, there was a man (whose) name was Faam-Bae’.’

Similar to the function of laying a story out in the presentational construction,

temporal adverbs also provide the setting.

The second kind of setting expressions is stative verbs describing the nature and

characteristics of the subject. These stative verbs are often accompanied by nyei

‘present state affirmative’ (PRS.ST.AFM.) at the end of the clause. Nyei does not

occur with negative as in a contrast between (54a) and (54b).

(54a) (BS.187) Ninh nqox nyanc aeqv, kuv nyei 3sg husband eat TOP delicious PRS.ST.AFM

‘When her husband ate (it), (it) tasted delicious.’

(54b) (BS.188) Ninh auv nyanc aeqv, maiv kuv 3sg wife eat TOP NEG delicious

‘When his wife ate (it), (it) was not delicious.’

The third kind of setting expression has a wide variety, all of which are

copulatives or constructions in equative sentences. The construction ‘A benx B’

means ‘A is B’ describing A’s internal characteristic (almost permanent), e.g.

Aengh Doi benx m’sieqv dorn ‘Aeng-Doi was a girl.’ The ‘A zoux B’ means ‘A is

B’ in a sense of role, occupation, or social position as in (55abc) (these three

examples are all from Lombard and Purnell 1968:158).

(55a) (Lombard and Purnell 1968:158) zoux auq be wife

‘to be a wife’

Page 127: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

101

(55b) zoux baeng be soldier

‘to be a soldier’

(55c) zoux bou be servant

‘to be a servant’

In contrast to the previous two constructions in this subset, ‘A zeiz B’ is associated

with evaluation; that is, right or wrong as in (56).

(56) (FA.077) Ninh_mbuo gengh i dauh zorqv ninh_mbuo nyei sou doix 3_PL really two CLF take 3_PL POSS book match

za'gengh zeiz nyei really right PRS.ST

‘They two indeed took out their books (and) compared; really (they) were right [i.e. they really did match].’

Zeiz is often used in the negative, e.g. Aengh Doi maiv zeiz m’jangc dorn ‘Aeng-

Doin is not a boy’ (<lit. Aeng-Doin is not correct a boy). Another example is (57).

(57) (BS.042) Mangc buatc maiv zeiz naang look see NEG right snake

‘(She) looked (and) saw (that) it was not a snake.’

Furthermore, zeiz can negate a proposition (58).

(58) (FG.023) Ndiangx-kou youc gorngv mv zeiz ninh ganh oix nauv dead_tree also speak NEG right 3sg self want snap

‘The dead tree also said (that) it was not that he himself wanted to snap.’

Among this group, se is the loosest connector. ‘A se B’ means ‘A, that is B.’

After presenting A as a topic, the description or explanation follows after se.

Example (59) is a repetition of (57) with a full ending.

Page 128: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

102

(59) (BS.042) Mangc buatc maiv zeiz naang look see NEG be_right snake

se yietc dauh kuv mienh that_is one CLF good person

‘(She) looked (and) saw (that) it was not a snake, (but) it was a good man.’

The fifth setting expression is dongh, originally meaning ‘the same.’ Thus, ‘A

dongh B’ means ‘A is identical with B’ or ‘A is identified as B.’ See (60):

(60) (FA.074) Zinh_ndaangc mbuo juangc horqc dorngh doqc_sou wuov previously 1pl share school place study_book that

dongh yie that_is 1sg

‘The one (who) we studied in the same school before; that’s me.’

The last kind of expression in this band is the negation of a stative verb which is

used in a synonymous parallel affirmative expression. An example is maiv kuv,

im nyei ‘not delicious, bitter’, where the first phrase is in the negative, which is

coupled with a synonymous paraphrase in the affirmative stative verb. The whole

phrase as a unit functions as a descriptive expression as in (61).

(61) (BS.192) Ninh auv nyanc yaac maiv kuv, im nyei 3sg wife eat and_yet NEG delicious bitter PRS.ST

‘His wife ate but (it was) not delicious, (it was) bitter.’

Note that a negation of dynamic/acton verbs is not used in this construction. Such

a negation belongs to the irrealis band.

4.3.3.5 Band 4: Flashback

Basically, flashbacks are expressed with unmarked verbs; there is no specific

marking on the verb for this band. The context has clues to indicate that certain

clauses are pointing to a deeper past time reference than the time of narration. In

(62) (sentences 1-3), (St1) is on the storyline with the verb gornv ‘speak’ followed

by an indirect speech maiv zeiz ninh dorngc ‘it was not that he was wrong.’ This

Page 129: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

103

indirect speech is carried on in (St2) and (St3), which report an event that

happened previous to the narrative storyline. Verbs in boldface are all flashbacks.

(62) (FG.025) St1

Nziouv-mbiauh_loc gorngv maiv zeiz ninh dorngc termites speak NEG right 3sg wrong

(FG.026) St2

Naang-nzung ganh zun ninh nyei biauv earthworm self cement 3sg POSS house

(FG.027) St3

Ninh mv maaih dorngx yiem 3sg NEG have place live

cingx_daaih cuotv naaic daaih ngaatc ndiangx-nzungh dangx therefore come_out there come bite tree_root break_off

‘25The termite said (that) he was not wrong. 26The earthworm himself was cementing his (termite’s) house. 27He (termite) had had no place to live, therefore (he) had come out from there, (and) bit the tree roots (and the roots) broke off.’

Other than interpreting from the flow of context, flashbacks are marked by such

time phrases as wuov_zanc ‘that time’, loz-hnoi ‘in the old days’ or ‘in the past’

and m’daaih ‘from the beginning.’

4.3.3.6 Band 3: Backgrounded Activities

The materials in this band do not advance events in the narrative, but they

describe how the actions are carried out. They are repetitive, habitual, progressive

and durative. The examples are (63) V1 + V1 + V1 +…(repetitive), (64) V + nyei

+ V (repetitive), (65) V + jienv ‘keeping in the state of V’ (durative aspect), and

(66) Mimesis/onomatopoeia + nyei + V (manner).

(63) (AS.110) Repetitive Sin zinx za'gengh nyanh nyanh nyanh nyanh nyanh nyanh nyanh body tremble really jerk jerk jerk jerk jerk jerk jerk

nyanh nyanh deix jerk jerk some

‘(His) body really trembled continuously so much.’

Page 130: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

104

Another repetitive construction involves nyei ‘adverbializer’ between V1 and V2,

where V1+nyei modifies V2 (64).

(64) (AS.084) Repetitive Baqv nyei baqv, baqv daic mingh pierce ADVBLZR pierce pierce die go

‘(They) pierced and pierced (the tiger), (they) pierced (it) dead.’ (They stabbed the tiger repeatedly and stabbed until it died.)

(65) (FA.063) Durative Ninh za'gengh hnyouv nzauh yiem jienv 3sg really heart sad exist DUR

‘He was in a really sad situation.’

Similar to (64) the next example has mimetic words (sometimes onomatopoeia)

adverbialized by nyei, and as a whole phrase it modifies the main verb taux

‘reach’ as in (66).

(66) (AS.075) Descriptive Nda'maauh suoh suoh nyei taux wuov tiger MIM MIM ADVBLZR reach there

‘The tiger arrived there slinking along.’

4.3.3.7 Band 2: Backgrounded Action/Events

The background action and events expressed by the use of jienv ‘be stable’ is

syntactically in close connecton with the verbs in Band 1. The verb phrase nyiemv

jienv mingh ‘cry-ing-go’ can be translated as ‘go while crying’, ‘go in a state of

crying,’ ‘go at the same time crying’ or ‘went away crying.’ That is, in ‘V1 +

jienv + V2’, where V2 is the main verb, the phrase ‘V1 + jienv’ modifies/explains

V2. The V1 expresses a simultaneous action with the main verb V2, thus it is V2

that carries the storyline (67). Hence, V1 in this construction belongs to Band 2.

Semantically, besides the simultaneous action, V1 indicates prolonged, repetitive,

on-going action. Example (67) consists of sequenced clauses (Cl1 and Cl2), where

Cl2 contains an expansion of ‘V1 + jienv + V2’, namely ‘V1 + jienv NP + V2.’

Page 131: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

105

(67) (AS.079) Cl1 V1 V2

Wuov_deix huei, huei nyei zunc jienv daaih those_people MIM MIM ADVBLZR chase SML.ACT come

Cl2

bungx jienv congx daaih release SML.ACT gun come

‘Those people came chasing (the tiger) with a loud noise; (they) came shooting guns.’ (or ‘shooting as they came.’)

Another construction expressing a simultaneous action without using jienv is yietc

binc + V1, yietc binc + V2 ‘V1-ing and V2-ing’ (lit. one time V1, one time V2).

However, this will be described in the section of primary storyline (4.3.3.8) for the

reason given therein.

4.3.3.8 Band 1: Primary Storyline

Once a narrative is introduced by the presentational expression maaih ‘there

was..’, and/or the setting expression loz-hnoi ‘in the old days’ or ‘once upon a

time’, the events following are all in the past in the narrative; hence there is no

need to mark the tense as a narrative past. In fact, Iu-Mienh does not have such a

coding for this band. Thus, this band employs the unmarked verbs as a default

storyline verb. However, they are distinguished from the verbs used in Band 6,

which are presentational, stative and equative. As in Somsonge’s analysis for

Thai storyline (1992:113-4), Iu-Mienh also uses verbs of cognitive experience (e.g.

hiuv_duqv ‘know’, buatc ‘see’), event proper, action and motion in the storyline.

These unmarked verbs are also used in other grammatical constructions such as

repetition (for emphatic expression and repetitive action), serial verb constructions,

multiclause constructions and topic chains. They are also used with the

directional verbs daaih ‘come,’ mingh ‘go,’ faaux ‘ascend’ and njiec ‘descend.’

Among them daaih and mingh have aspectual meaning when a transportation in

space is not involved. They all have the properties of Band 1 that Longacre has

listed in Figure 18. The repetition of unmarked verbs signals the dynamics and

punctiliar parameters; SVCs and topic chains provide the narrative with clear and

Page 132: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

106

rapid transitivity of actions, and aspectual verbs express the completive and realis

aspects of actions.

One more important parameter for the storyline is “sequentiality” as has been

discussed previously (in the subsections 2.1.7, 2.1.8, 2.1.10, 2.1.11).67 In Iu-

Mienh sequentiality is overtly marked by the sequential marker ziouc, whose

approximate translation would be ‘and then,’ ‘then soon’ and ‘after which.’ This

conjuntion can be used with all the other verbs and constructions listed above in

this subsection. An example of the way this sequential marker ziouc is used can

be found in (68).

(68) (BS.116) Ninh nziez ziouc nyorng jienv mingh 3sg younger_sister then creep SML.ACT go

taux ndiangx-nquaah dueiv-mueiz mi'aqv reach branch end_of_tail RSLT

(BS.117) Ninh dorc ziouc goix ndiangx-gorn, ndiangx ziouc nauv 3sg elder_sister then cut_down tree-root tree then snap

(BS.118) Ninh nziez ziouc zong njiec wuom-njaangh mi'aqv 3sg younger_sister then strike descend water-pond RSLT

ziouc daic wuov njaangh mi'aqv then die there pond RSLT

(BS.119) Ninh dorc ziouc zorqv ninh nziez nyei lui-houx 3sg elder_sister then take 3sg younger_sister POSS clothes

siou-setv daaih zorng jienv a_set_of_woman's_dress come dress DUR

ziouc nyiex jienv guh_nguaaz nzuonx taux biauv then carry_on_back DUR baby return reach house

‘116Her younger sister then went creeping (and) reached the end of the branch. 117Her elder sister then cut down a tree base; then the tree fell down. 118Her younger sister then struck down into the pond, then died there in the pond. 119Then her elder sister took her younger sister’s clothes and the whole set of dress (and) put them on; (she) then went home carrying the baby.’

67 Cf. Buth (1995:86, 88, 99); Endo (1996:189-90); Follingstad (1994:169, 171, 188); Longacre (1996:25);

Ricoeur (1988:15 [1985]); Somsonge (2002:142).

Page 133: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

107

The development adverb aengx ‘again’ should receive some mention. Lombard

and Purnell (1968:4) give the definition ‘again, still, also.’ Panh (1995:4) defines

it as ‘and, again, another, also, either.’ However, longer narrative texts show that

it often introduces a new development of the story, rather than just a repeat of the

previous happening. Hence another meaning is ‘further, furthermore, moreover’

as an extention of ‘still’ and ‘also’ by the two dictionaries above-mentioned. In

(69) a cicada speaking is not a repetition of its previous action but the first

mention of the new participant. In the preceding sentence CI.001, an owl talked

to a deer (CI.008), a squirrel talked to the deer (CI.010), then comes (69).

(69) (CI.011) Gaeng-waen aengx gorngv cicada further/also speak

‘A cicada also spoke up.’ [in addition to the owl and the squirrel having spoken]

Example (70) shows that aengx functions as a developer of the story. It recounts

the sequence of events which led to the injury or damage. This kind of discourse

level grammatical device is termed a “development marker” by Levinsohn

(1992b:32-37).

(70) (CI.014) Sopc luangh dangx mingh youc aengx sopc njangx ash_pumpkin vine break_off go also further ash_pumpkin roll

(CI.015) Sopc njangx jienv njiec ash_pumpkin roll SML.ACT descend

youc aengx mborqv zuqc saqv-ndongh kung also further hit GOAL sesame-container spill_out

(CI.016) Mborqv zuqc saqv-ndongh kung hit GOAL sesame-container spill_out

youc aengx kung bieqc norqc_jai m'zing mingh also further spill_out enter wild_chicken eye go

‘14The ash pumpkin vine broke and a pumpkin rolled away. 15The ash punbkin rolled down and smashed into a sesame pod spilling out the contents. 16(The pod) being smashed open, (the seeds) flew out and got into a wild fowl’s eye.’

There is a construction which expresses a simultaneous action similar to that of

the Band 2-backgrounded actions/events band. However, the sentence containing

Page 134: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

108

this construction is on the storyline. In contrast to (67) in 4.3.3.7 (i.e. V1 + jienv +

V2), (71) uses yietc binc + V1, yietc binc + V2 ‘V1-ing and V2-ing’ (lit. one time

V1, one time V2):

(71) (BS.010) Ninh ziouc yietc binc nyiemv yietc binc heuc lungh_ndiev 3sg then one time weep one time call world

‘He then one time cried, one time called (to) the world.’ (i.e. ‘So then he was weeping and calling on Heaven and Earth [as witnesses] all at the same time.’)

In this construction, both VP1 and VP2 are in equal status (i.e. crying and calling),

as opposed to the modifier-modified relationship between V1 + jienv and V2 in

(67). Nonetheless, as will be seen in Figures 50-52 in 5.4.2.1 (Court 1986:36),

VP2 seems to receive more attention by a hearer than V1 due to a fundamental

nature of putting more focus on a rightward element in a sentence in Iu-Mienh.

However, the two VPs describe one event, hence the whole sentence (71) is on the

storyline (note that the sentence has the sequential marker ziouc). In regard to the

isolated VP yietc binc + V1 itself, it encodes a background action.

4.3.3.9 Band 1′: Pivotal Storyline

Though unmarked verbs serve as default verbs for the storyline, there are three

more items that are used in a special way to augment or supplement the storyline.

They include sentence final aspectual markers aqv ‘perfective aspect,’ mi’aqv

‘resultative aspect’ and an adverb za’gengh ‘really.’ They all belong to the pivotal

storyline. Here, mi’aqv and za’gengh are discussed, leaving aqv for 5.2.2.

The pivotal storyline is an augmentation of the primary storyline. Longacre (1996)

categorizes it into two groups: “(i) happenings which are marked as pivotal

constitute a rough abstract of the story, and (ii) happenings which are marked as

pivotal, although they are for some reason weighted, when taken together do not

constitute such an abstract” (Longacre 1996:28-9). It is the second group that aq,

mi’aqv and za’gengh ‘really’ belong to.

The construction ‘Sentence + mi’aqv’ (resultative aspect) expresses a strong

punctiliarity and a complete passage of an event. For example, ninh mingh

Page 135: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

109

mi’aqv ‘he is gone’ has a sense that ‘he is not here any more’ besides its explicit

meaning. (72) shows that an event of Faam-Bae’s death is conclusive, definite

and punctiliar. This additional information is indicated by the resultative

aspectual particle mi’aq, though the unmarked verb daic ‘die’ is enough for

indicating the storyline.

(72) (FA.098) Nqaengx jaang sim baqv daic mi'aqv choke_on throat needle pierce die RSLT

‘(A needle) choked on the throat, the needle pierced (his throat), (and he) died.’

The construction often marks a discrete point of ending at the closure of one

paragraph as in (73):

(73) (AS.130) Buonv baaic mi'aqv, ninh_mbuo biaux nzengc mi'aqv shoot defeat RSLT 3_PL flee consumed RSLT

‘(He) shot (and) defeated (them) all; they all completely ran away.’

(73) is the last sentence at the end of the peak episode (cf. Figure 24). Similarly,

the sentence containing mi’aqv in (FA.113) in Story 5 and (BS.255) in Story 7

marks the end of the peak episode in both stories (cf. Figure 23 and Figure 25).

The adverb za’gengh ‘really, indeed, actually, surely,’ has two basic usages: 1) as

a manner adverb as in ‘za’gengh + Stative Verb,’ and 2) as an adverb emphasizing

an actual occurence of the action, ‘za’gengh + Action/dynamic Verb.’ It is the

second kind of usage that is on the pivotal storyline, while the first is used in the

setting band ((74a) is in DS.). The first kind of usage is exemplified in (74a) and

the second in (74b):

(74a) (AS.026) "U..., Meih za'gengh henv haic." Wow! 2sg really strong very

“Wow! You are really very strong!”

(74b) (FA.097) Ninh za'gengh nyanc nc zeiv fienx. 3sg really eat that paper letter

‘He actually ate the letter.’

Page 136: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

110

An implication of (74b) is “Believe it or not, he actually ate the letter!”

emphasizing the actual happening of the action it modifies. Borrowing a term

from Whaley (1997),68 this adverb can be referred to as a “realis adverb.” A realis

adverb means that it “asserts that an event actually occurs” (cf. Whaley 1997:206).

A contrast between (75a) and (75b) may solidify the definition.

(75a) (AS.101) Baeng za'gengh daaih soldier really come

‘The soldiers really came.’

(75b) Baeng mv gaengh daaih soldier NEG-yet come

‘The soldiers haven’t come yet.’

Za’gengh in (75a) indicates realis; hence on the storyline, whereas mv gaengh ‘not

yet’ in (75b) indicates irrealis.69

The realis adverb za’gengh ‘really,’ ‘indeed,’ ‘actually’ is used immediately after

direct speech, flashback, the participant’s internal thinking and author intrusions

in order to bring the audience back onto the storyline. This usage will be

discussed in 5.2.2.2.

In contrast to punctiliar adverbs like than1 daj1 nan4 ‘suddenly’ in Thai, the Iu-

Mienh equivalent liemh zeih ‘suddenly’ does not appear often in Iu-Mienh

narratives. Instead, za’gengh ‘really’ is very frequently used for the furtherance of

a story, not only in the environment just mentioned above but also in other places

which are clearly on the storyline and its augmentation.

68 In describing Burmese past tense which is marked by “a realis affix,” Whaley defines it: “roughly, this affix

is used to assert that an event actually occurs” (1997:206). Thus, for za’gengh in Iu-Mienh I have adopted the term “realis adverb,” as opposed to a manner adverb, which emphasizes the stative verb.

69 The component gengh itself means ‘really’ (as in CI.003, FA.077, FA.094, FA.106, passim). Thus, za’ by itself probably comes from zien gengh (‘true + really’) or zengh gengh, with the first part undergoing the common neutralization to za’ (Purnell, 2006: personal communication). Za’gengh and gengh mean ‘really’; mv gaenh means ‘not yet.’

Page 137: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

111

In summary, two things should be emphasized. First, the pivotal storyline in Iu-

Mienh is the second kind which Longacre (1996:28-9) describes as a non-abstract-

constituting pivotal storyline, as opposed to an abstract-constituting pivotal

storyline. Hence, one should be reminded of the primary function of an unmarked

verb is as a storyline verb. Second, belonging to the non-abstract-constituting

pivotal storyline band, aqv, mi’aqv and za’gengh occur to supplement the primary

storyline. That is, other constructions in the primary storyline band and the three

constructions in the pivotal storyline band are not mutually exclusive; the latter

reinforces the effect of the former.

4.4 Finding Narrative Timeline through Temporal Movement

Somsonge, mainly working within the framework of Longacre, seems to have

been aware of the need for a method to identify the storyline in languages without

verbal morphological inflections such as Thai (1991a:76). She states:

Even though these four categories of on-the-line verbs [i.e. verbs of cognitive experience, event proper, motion and action] are recognized as being on the storyline, one should not expect that whenever they are present, they are always on the storyline. (Somsonge 1991a:99)

The following subsections will recapitulate an observation of the Iu-Mienh

salience scheme in the similar vein with Somsonge’s caution, and will mention

briefly that her proposed alternative is in concordance with the identification of

storyline in Iu-Mienh.

4.4.1 Methodological Consideration

From the observation of verbs and clauses in arranging the salience scheme in Iu-

Mienh, the same construction is used in different bands. First of all, the fact that

the verbs on the storyline are unmarked is a complication for a morpho-syntactic

based methodology of identifying the storyline. For example, the forms in the

flashback band are the same as those in the storyline. There is no morphological

distinction between irrealis verbs in complementation in Band 6 and storyline

verbs in Band 1. Other examples include the verb mingh as a regular verb ‘go’

Page 138: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

112

and the directional verbs in different bands, and jienv in different bands, zoux as a

regular verb ‘do’ or ‘make’ and a copular ‘be’. In the Iu-Mienh case, it is difficult

to identify the storyline by searching for a particular verb form. This situation is

due to the fact that many aspectual particles and auxiliaries were originally regular

verbs and the same forms are still used.70

A similar phenomenon has been recognized in other genetically unrelated

languages such as Greek (Tehan 2000:8-10). He acknowledges that 1) different

verb forms can be on a storyline band without excluding each other, 2) therefore,

an exclusion of either form will lose significant information from the

macrostructure (viz. semantics).71

4.4.2 Non-Verb-Morphology Dependent Method

As a proposal of a non-verb-morphology dependent method of identifying the

storyline, taking the contextual and semantic information into consideration,

Somsonge argues for two parameters: time-movement in sequence and

presentation of new information in sequence (2002:142). They are inseparable.

According to her, these parameters are found in four types of grammatical

constructions in Hlai (Li) narrative discourse: 1) a sequential indicator fan53

‘then,’ 2) topic subject, i.e. “the noun phrase designating the topic of the discourse

used as the subject of a clause” (2002:148-9), 3) temporal connectives and 4)

anaphoric temporal clauses. Likewise, four corresponding groups of constructions

are evident in Iu-Mienh.

First, in Iu-Mienh narratives, the sequential marker ziouc belongs to Band 1. In

discussing the sequential marker in Hlai, Somsonge also mentions an adverb lom11

‘again,’ whose equivalent in Iu-Mienh is the development adverb aengx ‘again,’

70 A similar situation in Thai is discussed in Kingkarn (1986:245): “…lexical verbs are used to perform what

a speaker of other languages might expect to be performed by two types of elements: lexical and grammatical. In fact, lexical verbs in Thai still retain their semantic properties in every case.”

71 The exclusion of present indicative, as opposed to aorist which appears to be a storyline verb, would lead an exegete to a deficiency because such a mechanical application of alleged storyline criterion will eliminate substantive information. See Tehan 2000:9.

Page 139: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

113

‘further,’ ‘moreover.’ Belonging to the primary band, both ziouc and aengx carry

the storyline forward: the former by keeping the sequentiality, the latter by

developing an event and presenting a new happening (discussed in 4.3.3.8 and

5.4.1).

Second, what Somsonge refers to as the “topic subject” is discussed as “topic

chains” within Band 1 in 4.3.3.8. This construction is also a significant device to

keep sequentiality while presenting new events at the same time.

Third, the temporal connectives in Iu-Mienh are assigned to Band 5: settings.

Fourth, the anaphoric temporal clauses in Iu-Mienh belong to Band 8: cohesive

band. Particularly, the two expressions, V + gau (independent clause marker)

‘after V,’ ‘V and then,’ and V + liuz (aspectual verb) ‘after finishing V’ have both

a connecting function with the previous sentence and a presenting function of a

new development.

The third and fourth groups are not main-verb-related constructions; they even

belong to “supportive materials” bands. However, they play a significant role in

advancing the storyline forward.

4.5 Summary of the Chapter

In this chapter the storyline analysis theorized by Longacre has been reviewed

with particular focus on the storyline verb or the preterite. In doing so, the

importance of sequentiality was pointed out as a common characteristic of the

foregrounded storyline studied by many scholars. Subsequently two types of

definition as to storyline were compared: one being ontological and the other

epistemological. Belonging to the latter, the cognitive approach sees the

foregroundedness as a result of the sequentiality phenomenon. At the end of the

chapter, a salience scheme for Iu-Mienh narrative discourse was presented

according to the textlingusitics storyline analysis. Finally, it was observed that the

same verb form is used across different bands and that different constructions are

used in one band. This phenomenon has suggested that a “spectrum” can be

recognized not only across the supportive materials but even within a band, e.g.

Page 140: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

114

inside the storyline band. In dealing with this phenomenon Somsonge’s four

categories of constructions were considered: the sequential and development

markers, topic-subject (i.e. topic chains), temporal connectives and anaphoric

temporal clauses.

Page 141: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

CHAPTER 5

A COGNITIVE LINGUISTICS STORYLINE ANALYSIS

5.0 Introduction

In 4.2.2 a definition of storyline from a CL perspective was proposed. In 2.2 and

2.4, the CL framework for discourse analysis was presented. This chapter affirms

further that the conceptual structure of storyline can be analyzed in such a way

that reveals its meaning if the narrative discourse is taken as a cognitive system,

following Langacker (1991c, 2001b) and Talmy (2000b). In Talmy’s “cognitive

framework for narrative structure” (2000b:423-7), he sets forth two factors of

prototypical narrative: the degree of progression and the degree of coherence and

significance. These two factors are also confirmed by Somsonge (2002:142) as

major parameters of the main line of narrative development, when she says,“The

clauses which refer to sequenced points on a timeline [viz. Talmy’s “coherence”]

and present new information [viz. Talmy’s “progression”] […] constitute the

foreground of a narrative.” For the analytical feasibility of sequenced SVCs and

topic chains in Iu-Mienh, the latter can be summarize as event transitivity (i.e.

“progression” or “presentation of new information”) and the former as event

sequentiality (i.e. “coherence” or “sequenced points on a timeline”), which are the

crucial components in a prototypical narrative storyline. First, in this chapter, a

notion of narrative discourse will be situated on a stage as if we were viewing a

storyline in a theater; then, the conceptual structure of transitivity of events, as a

contributor to foregroundedness in the storyline, will be explained. Finally, the

conceptual structure of sequentiality will be discussed at the end of this chapter.

Describing and analyzing the storyline and salience scheme in chapter 4, this

chapter explains what they mean in terms of conceptual structure of some selected

linguistic constructions in Iu-Mienh.

115

Page 142: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

116

5.1 Placing A Narrative in the Cognitive Linguistics Setting

“Inside or outside?” is the issue in this section. Does a narrative discourse exist

outside of a mind or inside? As has been shown thus far in chapter 4 (except for

4.2.2), textlinguistics places a narrative discourse outside of a mind, often as a

written text. By contrast, CL views it as a thing taking place in a mental domain

(with an attention to interpersonal meaning too, though the present study does not

include such an aspect).72 A fundamental assumption in approaching narrative in

Talmy’s cognitive semantics (2000b:417-8) is “the existence of a mind that has

produced the narrative as well as of a mind that is cognizing the narrative.” This

cognition of the narrative is carried out perceptually. One of the six73 tenets of

cognitive semantics suggested by Gärdenfors (1999:21) is that “[c]ognitive

models are mainly perceptually determined (meaning is not independent of

perception).” The perception of a clause (an archetypal/prototypical transitivity in

this case) by a viewer (V) is schematised as the “stage model” by Langacker

(1991a:211) in Figure 29 and its extention to a discourse level is shown in Figure

30, which is a modificaton of Figure 14 in 2.4.

72 For an argument for the social and cultural component of discourse in cognitive semantics, see Langacker

1997. He writes, “I would claim, then, that despite its mental focus, cognitive linguistics can also be described as social, cultural, and contextual linguistics” (Langacker 1997:240).

73 The 0ther five are quoted as follows: “1) Meaning is conceptualisation in a cognitive model (not truth conditions in possible worlds). 2) [quoted above]. 3) Semantic elements are based on spatial or topological objects (not symbols that can be composed according to some system of rules). 4) Cognitive models are primarily image-schematic (not propositional). Image-schemas are transformed by metaphoric and metonymic operations (which are treated as exceptional features in the traditional view). 5) Semantics is primary to syntax and partly determines it (syntax cannot be described independently of semantics). 6) Concepts show prototype effects (instead of following the Aristotelian paradigm based on necessary and sufficient conditions [end quote]” (Gärdenfors 1999:21-5).

Page 143: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

117

setting

V

Figure 29. A stage model

The clause level viewer (V) in Figure 29 is divided into a discourse level speaker

(S) and hearer (H) in Figure 30 (See 2.4 for explanations on the parts of the

diagram). In Figure 30, the two upward broken arrows indicate that S and H are

mentally viewing a narrative.

Figure 30. A narrative in the stage model

Thus, a narrative discourse is viewed as a cognitive entity in a CL perspective.

Page 144: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

118

5.2 Salience Scheme as a Prototype Effect

This section is concerned with the phenomenon that there is a gradient both in the

storyline material and in the supportive materials. That is, a primary task here is

in reference to hypothesis one (H1) in 1.1.3.

5.2.1 Primary Storyline Band

In the primary storyline Band, the unmarked verbs and the gradient relationship

within the band are considered.

5.2.1.1 Unmarked Verbs

As has been analyzed in 4.3.3.8, Band 1, primary storyline verbs are unmarked.

See example (76), where the boundary between the aperture (CI.001-007) and the

first sentence of the actual narrative (CI.008) is clear (e.g. Figure 22).

(76) (CI.008) Cl1

Loz-hnoi norqc_guv_long gorngv mbuox jung in_the_old_days owl speak tell barking_deer

‘In the old days, an owl told a barking-deer;’

Cl2 Cl3

"Maqv_gorkv_biouv zuoqc aqv, Maqv_gorkv_biouv zuoqc aqv" olive_fruit ripe PFT olive_fruit ripe PFT

“Olive fruits are ripe! Olive fruits are ripe!”

Cl4

nduov jung daaih nyanc deceive barking_deer come eat

‘(and he) deceived the barking-deer to eat some’

In Cl1, the temporal adverb loz-hnoi ‘in the old days’ sets the scene; and

immediately after that, the clause norqc_guv_long gorngv mbuox jung begins the

storyline. The verbs gorngv mbuox ‘spoke & told’ and nduov ‘deceived’ are all

unmarked; and both Cl1 and Cl4 are on the storyline. The whole sentence and

what follows to the end is understood as a narrative and in the past tense once the

presentational phrase loz-hnoi ‘in the old days’ opens the narrative.

Page 145: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

119

In Fauconnier’s term, this presentational phrase is “A space builder […] a

grammatical expression that either opens a new space or shifts focus to an existing

space. Space builders take on a variety of grammatical forms, such as

prepositional phrases, adverbials, subject-verb complexes, conjunctions +

clauses…” (1997:39-40). Besides loz-hnoi ‘in the old days’ (FG.002), other space

builders include maaih ‘there was…’ (< ‘to have’) (OH.001, AS.001), and the two

expressions combined loz-hnoi maaih ‘in the old days there was…’ (BS.001,

3PG.002, FA.002). They all can be referred to as narrative space builders.

As soon as the narrative space is opened by one of these space builders, both the

narrator (S in Figure 30) and the audience (H in Figure 30) are grounded to

understand that what is going to be told belongs to a Past TR; hence, there is no

need to encode the verbs as such. This can be supported by the comment by

Dooley and Levinsohn: “Typically, the body of a text is UNMARKED for

prominence. Thus, the storyline or foreground events of a narrative do not

normally carry a marker” (2001:84). 74 This situation goes with all the

constructions in Band 1: S + ziouc + V ‘and then V,’ repetition of verbs, SVCs,

topic chains, V + directional verbs (daaih ‘come’ and mingh ‘go’), and V +

aspectual verbs.

5.2.1.2 A Gradient within Band 1

A selective approach is taken to suggest that there is a gradient within a band. The

discussion in this subsection is limited to a gradient within Band 1. Another

specimen of a gradient is recognized in Band 5-setting band, which will be briefly

mentioned in 5.2.3.

As Dooley and Levinsohn (2001:84) state that the unmarked materials are (proto)

typical body text of a narrative, it is possible to recognize that the unmarked verbs

in Iu-Mienh are the prototypical storyline verbs. Moreover, it seems to be feasible

to posit a graduated cline of salience among the constructions mentioned above.

74 Also Levinsohn (1992b) and (1994).

Page 146: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

120

For example, the verb gorngv ‘speak, say, talk’ can have various transitivity

depending on how it is used:

(i) gorngv ‘speak, talk, say’ (FG.028, IC.011, both on the storyline;

passim)

(ii) gorngv ‘have a marriage arranged’ (FA.090, FA.120, both in

flashback.)

(iii) gorngv_ziangh ‘engage’ (FA.091 in DS)

(iv) gorngv waac ‘say words’ or ‘speak a language’ (FA.054, FA.068,

both on the storyline)

(v) gorngv + DS ‘say’ (OH.007, FG.006, both on the storyline; passim)

(vi) gorngv_mbuox + person ‘say & tell + person’ (3PG.037, CI.008,

both on the storyline; passim)

(vii) gorngv-baeqc ‘lie’ (BS.103 on the storyline)

(viii) gornv_taux ‘talk about’ (FG.003, FG.004, both on the storyline)

All of these but (ii) and (iii) are unmarked verbs and used on the storyline.

Though (ii) and (iii) are not found in the corpus for this study, they also may well

be used on the storyline.

Furthermore, taux ‘reach’ in (viii) is used in the following expressions: (ix) taux

biauv ‘arrived home,’ (x) mingh taux biauv ‘went as far as a house.’ A comparison

of (ix) and (x) with (xi) nzuonx biauv ‘go home’ seems to show an expansion of

grammatical constructions in the same semantic domain.

Developing upon Hopper and Thompson’s ten transitivity parameters (1980) (e.g.

Table 16 in 4.2.1), Taylor (2003:232-3) has presented twelve parameters or

semantic properties from a more cognitively oriented approach to analyse the

prototype transitive category. They are listed below:

(a) The construction describes events involving two, and only two participants, encoded by the subject and direct object NPs respectively;

Page 147: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

121

(b) The identity of the two participants can be determined, that is to say, the subject and direct object nominals have specific reference;

(c) The two participants are highly individuated, distinct from each other and from the background environment;

(d) The event is initiated by the referent of the subject NP, i.e. by the agent. Responsibility for the event thus lies exclusively with the agent. Furthermore, the subject NP is the sentence topic; the subject is what the sentence is about;

(e) The agent acts consciously and volitionally, and thus controls the event. Since consciousness and volition are typically human attributes, it follows that the agent is typically a human being;

(f) As a consequence of the agent’s action, something happens to the patient, i.e. the referent of the object nominal. The effect on the patient is intended by the agent. Typically, though by no means necessarily, the patient is inanimate;

(g) After the occurrence of the event, the patient is in a different state from before the event. Usually, the difference is one which would be highly perceptible to an onlooking observer;

(h) The event is construed as punctual. Even though the event necessarily has temporal extension, the internal structure of the event, and the intermediate states between its inception and termination, are not in focus;

(i) The agent’s action on the patient usually involves direct physical contact, and the effect on the patient is immediate;

(j) The event has a causative component, i.e. the agent’s action causes the patient to undergo a change;

(k) Typically, agent and patient are not only clearly differentiated entities, often they also stand in an adversative relationship;

(l) The event reported by the construction is real, not imaginary, hypothetical, or counterfactual. Central instantiations of the construction are realis.

These twelve parameters will be applied to the constructions (i) – (viii) mentioned

above. They are applied categorically to all of them, even though some examples

are not found on the storyline due to the limitation of corpus. Nevertheless, it is

possible to know their degree of transitivity as independent constructions, while

there is a high possibility that all of them occur as storyline verbs.

This set of parameters is intended to identify a prototypical transitive verb. A

deviation from a prototypical transitive verb will be identified if the conditions of

the construction under investigation do not meet all the descriptions in the

Page 148: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

122

parameters. For instance, a participant reference indicated by zero as regard to

(a), (b) and (c) suggests a less typical transitive verb. As Taylor (2003:232-3)

emphasizes above, “(two participants) encoded by the subject and direct object

NPs” (of a), “specific reference” (of b), “[t]he two participants are highly

individuated” (of c), they should be on the surface structure in search for a

prototypical transitive verb. This means that one and the same construction can

have a different degree of transitivity depending on context. Comparing the

different constructions which contain the same element will certainly show a

variety of degree.

The expressions (i) – (viii) are exemplified in (77a-h):

(77a) (CI.011) Gaeng-waen aengx gorngv cicada also speak

‘A cicada also spoke up.’ (i.e. entered the conversation)

Here the verb gorngv does not have an object, describing a situation in which a

cicada simply made a sound or said something. Thus it is very low in transitivity:

parameters are: (a)-0, (b)-0, (c)-0, (d)-0 since the narrator explains that “the cicada

just called without purpose” in CI.012. The adverb aengx ‘also,’ ‘furthermore’ or

‘again’ introduces a new participant, cicada, after mentioning an owl (CI.008) and

a squirrel (CI.010) said something. Transitivity of this sentence is 16.7%.

Parameters a b c d e f g h i j k l Total Transitivity Value of gorngv 0 0 0 0 0 0 0 1 0 0 0 1 2 16.7%

(77b) (FA.119) Cl1

Ih_zanc ninh_mbuo longc jienv nyei now 3_PL marry DUR PRS.ST

Now they are married.

Cl2 Cl3

Maaz Jaa gorngv gau hnangv, m'daaih mv duqv longc Maajaa speak as only of-course NEG can marry

‘Maajaa just completed the arrangements (of his marriage with Aeng-Doi); (but) of ourse (he) didn’t get to marry her.’

Page 149: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

123

Though Cl2, where the verb gorngv ‘have a marriage arranged’ occurs, is an

adverbial clause, the twelve parameters are applied to see how transitive this

construction is because it is perfectly possible to have this construction on the

storyline. Even though it may be on the storyline, an event of arranging a

marriage involves many steps in a process; thus, the parameter (h = punctiliarity)

is null. However, zero punctiliarity does not necessarily mean that this verb

cannot occur on the storyline. The transitivity of this sentence is 25%.

Parameters a b c d e f g h i j k l Total Transitivity Value of gorngv 0 0 0 1 1 0 0 0 0 0 0 1 3 25%

Example (77c) is a typical usage of this speech verb gorngv waac ‘speak words.’

(77c) (FA.053) Obj-Topicalized

Yiem wuov haaix nyungc waac yaac gorngv nzengc be_in there what word TOP speak consumed

‘There (they) talked about anything altogether.’ (i.e. ‘There, whatever there was to talk about, they covered everything.’)

The Object is marked by yaac TOPICALIZER. Even though the verb seems to be

more transitive, the whole sentence does not have an explicit Subject-Agent as is

often case in Iu-Mienh, thus it computes to a lower level of transitivity. The

transitivity of this construction is 25%, the same degree as (77b).

Parameters a b c d e f g h i j k l Total Transitivity Value of gorngv waac 0 0 0 1 1 0 0 0 0 0 0 1 3 25%

Example (77d) is an idiomatic compound gorngv-baeqc ‘lie.’ It is a part of a long

sentence, where the verb is followed by nduov ‘deceive;’ thus, the transitivity of

the verb by itself and the clause as a whole differ.

(77d) (BS.103) Ninh dorc ziouc gorngv-baeqc nduov ninh nziez 3sg elder_sister then lie deceive 3sg younger_sister

‘Her elder sister lied and deceived her younger sister.’

Page 150: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

124

The parameters applied only to the verb itself are:

Parameters a b c d e f g h i j k l Total Transitivity Value of gorngv-baeqc

0 0 0 1 1 0 0 0 0 0 0 1 3 25%

The transitivity of the verb itself is 25%. For the sake of a comparison, however,

if the parameters are applied to the serial verb construction gorngv-baeqc nduov

‘lied and deceived,’ its transitivity is 66.7%.

Parameters a b c d e f g h i j k l Total Transitivity Value of gorngv-baeqc nduov

1 1 1 1 1 1 0 0 0 0 1 1 8 66.7%

Example (77e) is from a DS in FA.090.

(77e) (FA.090) Ih_zanc Maaz Jaa gorngv ziangh mi'aqv now Maajaa engage RSLT

‘Now Maajaa has already become engaged to me.’

The transitivity of this sentence is 41.7%.

Parameters a b c d e f g h i j k l Total Transitivity Value of gorngv ziangh 0 0 0 1 1 1 1 0 0 0 0 1 5 41.7

The verb gorngv ziangh ‘become engaged, complete the engagement process’

inherently implies that two participants are involved even though Iu-Mienh

surface structure often omits the Patient. On this basis of the semantic sturucture,

rather than the surface, the parameter (f) is valued positive. The parameter (g) is

positive because of the resultative aspectual particle mi’aqv.

In example (77f) the verb gorngv has DS as its object, which results in a greater

degree of transitivity.

(77f) (OH.007) Ninh die gorngv "Longc maah” 3sg father speak marry SFT.CMD

‘Her father said, “Marry (him)!”’

Page 151: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

125

The transitivity of this verb is 50%.

Parameters a b c d e f g h i j k l Total Transitivity Value of gorngv+DS 1 1 1 1 1 0 0 0 0 0 0 1 6 50%

Example (77g) is a compound verb gorngv mbuox ‘speak tell’ taking an addressee

as a Patient. The complement of the speech can be either DS or an embedded

indirect speech. Here (77g) is the latter case.

(77g) (FA.095) Aengh^Doih gorngv mbuox ninh aqv_zuqc nyanc Aeng-Doi speak tell 3sg must eat

‘Aeng-Doi told him that he must eat (it = her letter).’

The transitivity of this construction is 50%, the same degree of transitivity as

(77g).

Parameters a b c d e f g h i j k l Total Transitivity Value of gorngv mbuox 1 1 1 1 1 0 0 0 0 0 0 1 6 50%

The last example (77h) is also an idiomatic compound comprising V+V, gorngv

taux ‘talk about’ (literally, ‘talk reach’).

(77h) (FG.004) ninh_mbuo gorngv taux hmuangx nyei jauv 3_PL speak reach dark POSS NOM

‘They talked about the darkness’ (i.e. talked about that it became dark)

The parameters applied to this expression are:

Parameters a b c d e f g h i j k l Total Transitivity Value of gorngv taux 1 1 1 1 1 0 0 0 0 0 0 1 6 50%

The transitivity of this expression is 50%.

Since the verb gorngv is a cognitive experience verb, it is not surprising to have a

relatively low transitivity (the highest is 50% among the eight expressions here).

For a heuristic purpose, a typical transitive verb mborqv ‘hit’ or ‘beat’ is drawn

from BS.185 as (78):

Page 152: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

126

(78) (BS.185) Ninh ziouc ziang_naaic mborqv norqc daic mi'aqv 3sg then just beat bird die RSLT

Then she just beat the bird to death.

The parameters of this verb are:

Parameters a b c d e f g h i j k l Total Transitivity Value of mborqv 1 1 1 1 1 1 1 1 1 1 1 1 12 100%

The transitivity of this verb is 100%. This is the only fully transitive verb in these

examples.

Though the numerical indications have some significance in understanding the

transitivity of the verb, they should be at best interpreted as a tendency of these

semantic properties as a cluster. There are intricate combinations of factors that

may influence the transitivity of the whole sentence. For example, when gorngv-

baeqc in (71d = 25% transitivity) is considered as a part of an SVC, its transitivity

increases to 66.7%. Some transitive verbs inherently have an Object as a semantic

component; e.g. ziaaux means to call animals (back to a cage), which can be

perfectly understood with or without saeng-kuv ‘animal’ as an Object (BS.113-

115). Furthermore, the verbs which have a preposition-like function can alter the

transitivity of the main verb. For example, gorngv taux +NP ‘talk reach/about

NP’ in (77h) has an explicit and specific object (hence, a relatively higher

transitivity), while gorngv waac ‘say words’ is too general. In the expression

gorngv + preposition-like verb, the verb taux ‘reach’ has lost its concrete force of

reaching a physical goal. Indeed, the force of transitivity may not be numerically

measured as the sum of the value of gorngv and the value of taux.

At this point, one should be reminded of Taylor’s description of these twelve

semantic properties: “many of the following properties are understood relative to a

prior understanding of the gestalt, 75 the gestalt does not emerge from the

75 Gestalt (literally meaning ‘shape, form, configuration’ in German) is an organized whole which is seen as

more than the mere sum of its elements. Gestalt psychologists argue “that the whole is greater than the sum of its parts, that the configuration of the whole has properties of its own, over and above the properties of the individual objects that constitute it” (Miller and Johnson-Laird 1976:49).

Page 153: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

127

summation of independently conceptualized attributes” (2003:232). From a gestalt

psychological perspective, which argues for “the primacy of perceptual whole

over sensory atoms” (Miller and Johnson-Laird 1976), these various transitivity

degrees are not more than a tool for getting the feel of a tendency of different

verbs. The “summation” of “atoms” of computed transitivity parameters does not

exactly reflect what the conceptualizer or listener of a narrative perceives when

listening to a story.

However, to summarize, one clear point is that there is a cline or gradient among

the verbs and constructions of the primary storyline band from prototypical

transitivity to marginal (low) transitivity. A specimen of this gradient relation

within a category is shown in Figure 31.

Transitivity of gorngv and mborqv

0

20

40

60

80

100

120

gorn

gv's

peak

'

gorn

gv'h

ave

am

arria

gego

rngv

waa

c's

peak

gorn

gv-

baeq

c 'lie

'

gorn

gvzi

angh

'bec

ome

gorn

gv +

DS

'say

'

gorn

gvm

buox

'spe

ak &

gorn

gvta

ux 't

alk

abou

t'

mbo

rqv

'bea

t'

CI.011 FA.119 FA.053 BS.103 FA.090 OH.007 FA.095 FG.004 BS.185

71a 71b 71c 71d 71e 71f 71g 71h 72Constructions

Tran

sitiv

ity

exp

Figure 31. A gradient of transitivity of gorngv ‘speak’

in contrast to borqv ‘beat’

Page 154: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

128

5.2.2 Pivotal Storyline

In 4.3.3.9, it was pointed out that the pivotal storyline in Iu-Mienh is the kind that

is non-abstract-constituting, as defined by Longacre: “happenings which are

marked as pivotal, although they are for some reason weighted, when taken

together do not constitute such an abstract” (1996:28-9). Then, for what reasons

are they “weighted”? This section will propose an interpretation of the pivotal

storyline in Iu-Mienh from a CG perspective. That is, that pivotal storyline

materials function as grounding elements.

5.2.2.1 Grounding of a Negated Construction

The CG term ‘grounding’ here must be clearly distinguished from the

discourse/textlinguistics term ‘foreground vs. background’ used among linguists

like Hopper and Thompson (1980). As a specific technical term in CG, it is

defined by Brisard (2002) as follows:

Grounding is proposed as a technical term in Cognitive Grammar to characterize grammatical predications that indicate the relationship of a designated entity to the ground or situation of speech, including the speech event itself, its participants, and their respective spheres of knowledge. By definition, grounding predications are obligatory grammatical elements needed to turn nouns into full nominals, and verbs into finite clauses. (Brisard 2002:xi).76

In short, it can be described as “the epistemic footing of deixis and reference” as

the subtitle of Brisard’s book indicates. This fits well with the use of aqv

‘perfective aspectual particle’ and mi’aqv ‘resultative aspectual marker’ in Iu-

Mienh, though their behavior is not as strictly regulated as “obligatory

grammatical elements” as defined above. Compare (79a) and (79b).

76 Langacker’s glossary defines them as follows: “ground; The speech event, its participants, and its

immediate circumstances” (Langacker 1991c:548), and “grounding; A semantic function that constitutes the final step in the formation of a nominal or a finite clause. With respect to fundamental ‘epistemic’ notions (e.g. definiteness for nominals, tense/modality for clauses), it establishes the location vis-à-vis the ground of the thing or process serving as the nominal or clausal profile” (1991c:549).

Page 155: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

129

(79a) *Yie nyanc ziangx hnaangx 1sg eat finish rice

*I finish eat rice

(79b) Yie nyanc ziangx hnaangx aqv 1sg eat finish rice PFT

‘I have finished eating rice’ (i.e. I have finished eating.)

Iu-Mienh speakers are uncomfortable with (79a) since the sentence is not finished

yet; they would wait for something more to be said from the speaker.77 By

contrast, (79b) is situated in a usage event such as an actual situation at a meal

table or as a response to a question like “Have you eaten yet?;” thus, it is

grounded.

As a clausal grounding element in Iu-Mienh, aqv ‘perfective aspectual particle’

and mi’aqv ‘resultative aspectual marker,’ among others, are tentatively proposed.

Their grounding function in the pivotal storyline can be seen in (80):

(80) (AS.048) Za'gengh mv duqv wuom nyanc aqv really NEG get water eat PFT

‘Indeed, the water became undrinkable.’

The context is that a gigantic snake fell into a city water source for all the

residents (AS.047). Then comes (80). If the sentence is mv duqv wuom nyanc

‘did not get the drinking water,’ it belongs to Band 7-irrealis. However, the whole

sentence, lead by the realis adverb za’gengh ‘really’ or ‘indeed,’ is grounded by

aqv, whose original meaning is ‘change-of-state aspect’ as discussed in 3.3.2. The

water situation has changed from the previous drinkable state to the undrinkable

state as a reality. From a figure-ground perspective, against the usual (expected)

ground of the drinkable state, the negative situation of undrinkable state

77 Dangc Wuonh Mengh, one of my Iu-Mienh language helpers, has described other (quasi/pseudo) sentences

like (74a) without aqv or mi’aqv as naaiv joux waac maiv maaih gorn-baengx, meaning, “this sentence has no foundation.” His description of the situation and the term gorn-baengx ‘foundation,’ ‘root’ or ‘base’ seem to be a good explanation that aqv and mi’aqv are the clausal grounding elements in Iu-Mienh. The nominal grounding elements in Iu-Mienh include classifiers, the demonstrative pronouns naaiv ‘this,’ naaic ‘that,’ wuov ‘that,’ etc.

Page 156: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

130

(unexpected) has become the figure. This analysis can be supported by Takahashi

and Kingkarn (1997:273-82), though their focus is on Thai serial verb negation.

The event is informative saliently only if the background of inertia or non-eventness of the universe is more frequent, normal, and routine. When negative sentences occur, there must be a reversal of the figure-ground relation, so that the corresponding affirmative somehow becomes the expected, unsurprising background. Against such a background, the negative sentence then becomes salient or informative as illustrated in Figure 1 and 2 [reproduced as Figure 32 and 33] (Takahashi and Kingkarn 1997:277).

fi gure: positive event or eventness

ground: inertia or non-eventness

Figure 32. The figure-ground relation in the case of an

affirmative sentence (Takahashi and Kingkarn

1997:278)

figure: negative event or non-eventness ground: (expected) eventness

Figure 33. The figure-ground relation in the case of a

negative sentence (Takahashi and Kingkarn 1997:278)

Such a reversal of negation into a figure by the grounding element aqv can be

found in (81a-b), where the particle aqv is glossed as the ‘change-of-state’ aspect,

reflecting its prototypical semantic property (cf. Court 1986:248). The examples

(81a-b) are the part of a story where the participant’s attitude changed78 from

(81a) to (81b). That is, the participant (a young man) has been visiting an old

woman’s house but he would go home when the time came (up to the point of

(81a)), but in (81b) his attitude changed and he would not go home. Thus, maiv

nzuonx ‘not return’ is marked by the perfective aspectual particle aqv which

reflects the original meaning “change-of-state.”

78 Elsewhere Court glosses aqv as “New-Situation-Aspect” (1986:24, 38, 219).

Page 157: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

131

(81a) (BS.213) Mv_baac ninh nziaauc nziaauc yaac nzuonx nyei but 3sg visit visit and return PRS.ST

‘But though he visited her repeatedly, he still would go home.’

(81b) (BS.214) Nqa'haav aeqv, maiv nzuonx aqv later TOP NEG return CHG-O-ST

Later, he wouldn't go home.

The CHG-O-ST aspectual particle in (81b) is in contrast to the present state

aspectual particle nyei, which is often used with the affirmative, in (81a). This

contrast in the flow of the discourse shows the contrast between the previous

expected state (i.e. he would go home) and the changed situation (i.e. he would

not go home), and the negative sentence (81b) is grounded as an event that pushes

the story forward by the particle aqv. A similar example can be found in (82).

(82) (BS.169) Mingh taux ninh hnoi-hnoi yiem nyei dorngx go reach 3sg day-day live RELT place

norqc yaac maiv heuc aqv bird and_yet NEG call CHG-O-ST

‘(They) went to the place where he stayed every day, (but) the bird wouldn’t make its call.’

Before (82), or BS.169 in the story, the bird always made its cry when the

participant went to his work place. When he took his employer to the place to let

him listen to the bird, it would not cry as expressed by maiv heuc aqv. That is, the

expression is negative, and yet propels the story forward.

The resultative aspectual marker mi’aqv has a similar function, since it is

composed of mingh ‘go away from the speaker with adversative effect’ and aqv

(e.g. Figure 16).

5.2.2.2 Grounding and Interaction between Band 1 and others

As described in 4.3.3.9, the realis adverb za’gengh ‘really, indeed, actually’ often

occurs immediately after a DS, flashback, the participant’s internal thinking and

the author’s intrusions. Here a focus is given to the realis adverb za’gengh

Page 158: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

132

‘really’ as a grounding element, rather than its ordinary meaning of emphasis (e.g.

za’gengh jomc ‘really poor’ in AS.001, za’gengh henv ‘really strong’ in AS.026,

za’gengh ceng ninh haic ‘really praised him much’ in AS.057, etc.).

Example (83) occurs immediately after a DS. The context is that the piglet jeered

at the wolf, which is in the DS (3PG.029); then comes the narration which

contains za’gengh in (83), i.e. 3PG.030.

(83) (3PG.030) Ninh qiex_jiez haic, ninh za'gengh biomv aqv 3sg get_angry very 3sg really blow PFT

‘He got angry, so he really blew away (at the house).’

After the storyline is put on hold for a while by the DS, za’gengh retrieves it. This

usage is found in BS.176, FA.025, FA.089, FA.105, passim.

Example (84) is a case where za’gengh occurs immediately after a flashback.

Here, AS.008 reports the hero and his wife went into a granary, then AS.009

explains, as an author’s intrusion, that a bear had given birth in that granary, i.e.

flashback.

(84) (AS.008) Nyorng bieqc lamz mingh creep enter granary go

‘(They) crept into the granary.’

(AS.009) Jiepv njiec dorn bear give_birth_to cub

‘A bear had given birth to a cub there.’

(AS.010) Jiepv za'gengh tiux cuotv oix ngaatc bear really run go_out want bite

‘The bear really ran out and wanted to bite the couple.

The realis adverb za’gengh, thus, brings the audience back to the storyline.

(Another example of this kind is found in FA.118 after a flashback of FA.113-117

in Appendix D.)

Page 159: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

133

The last example is the case of za’gengh occurring immediately after a

participant’s internal thinking. In (85a-d), the sentences (85a)-(85c) are the

narrator’s empathetic representation of the participant’s thought. Then (85d)

narrates what really happened in the story.

(85a) (AS.031) Janx aengx mv haih fungc zoux non_Iu-Mienh again NEG be_able_to how-come do

‘There was nothing the northern Thai people could do about it again.’

(85b) (AS.032) Aav mv noic duqv aqv Ah NEG manage can PFT

Ah, (we) can't manage to do anything.

(85c) (AS.033) Aqv_zuqc mingh lorz naaic dauh Aa^Han_ Mix daaih zorqv hnangv must go look_for that CLF Bear_the_Brave come take only

‘(We) must go find that Bear Aahan and bring him here to let him catch (the crocodile).’

(85d) (AS.034) Za'gengh mingh lorz really go look_for

‘(They) really went to look for (him).’

It should be noted that all (85a)-(85c) are irrealis, two negatives and a modal: mv

haih ‘not able to’ in (85a), mv noic ‘not manage’ in (85b), and aqv_zuqc ‘must’ in

(85c). After them, za’gengh in (85d) resumes the realis of the story.

To summarize the observations conderning (83)-(85), the adverb za’gengh brings

the audience’s attention back to the storyline after the narrator establishes rapport,

explains a background through flashback or intrudes to address the audience.

Therefore, za’gengh is another grounding element in the sense that the narrator

pragmatically restores the audience to the common ground of perceiving the

narrative event (cf. S and H sharing a ground in Figure 30 in 5.1). Thus, the

materials in Band 4-flashback, Band 6-irrealis and Band 7-author intrusion are

interacting with Band 1-Primary storyline through the pivotal storyline material.

Page 160: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

134

5.2.3 Setting Band

A similar situation to the gradient in transitivity category analyzed in 5.2.1.2 can

be posited in the Band 6-setting, with regard to equative clause construction.

Space does not allow for a detailed discussion, and yet a general cline of strength

in connecting A and B in an equative sentence can be proposed as follows in

Table 17.79

Cline Constructions Function Examples and description

Reg

ular

ver

b A benx B ‘A is B’ profiles a genetic relation, inner character, instinct.

Aeng-Doi benx m’sieq dorn ‘Aeng-Doi is a girl.’ The equation between A and B is considered permanent. Negation of benx means ‘no good,’ ‘bad in terms of a moral or ethical issue,’ or ‘not rich.’

A d

eriv

ed v

erb

A zoux B ‘A is B’ or ‘A as B’ profiles a temporal equation, a social role, occupation.

Ninh zoux baeng ‘He is a soldier’ or ‘He serves as a soldier.’ Zoux originally means ‘do’ and ‘make.’ The equation between A and B can be temporary as in Ninh zoux jiex baeng ‘he has ever (jiex) served as/been a soldier.’

A re

stric

tedV

erb A zeiz B ‘A be_correct B’ Ninh maiv zeiz baeng ‘He is not a soldier.’

This is also used for the negation of a situation expressed by A benx B. Zeiz can be negated, showing that it is a verb. But the usage is restricted to zeiz nyei ‘it is right’ and maiv zeiz ‘it is not right/correct.’

Parti

cle A dongh B ‘A same as B’ or

‘A identical B’ profiles identification.

Aa^han Nguv dongh yie aqv! ‘The Dragon Aahan, that’s me myself!’ Dongh cannot be negated.

Parti

cle

A se B ‘A, that is B’ or ‘A, in other words B’ profiles presentation of a topic and its explanation.

Naaiv se baatv ‘This is a pen.’ A before se can be a whole sentence or a situation, not only an NP. B after se often explains, elaborates, or paraphrases A. Se cannot be negated.

Table 7. A gradient relation within equative

constructions

On the upper side of the scale, the constructions A benx B, A zoux B, A zeiz B are

all regular verbs, allowing negation and utilizing the present affirmative aspectual

particle nyei. Among them, zoux originally means ‘do’ or ‘make,’ and its derived

79 Examples in the table are from interviews with Iu-Mienh language consultants.

Page 161: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

135

meaning is similar to “is” in English, linking A and B in an equal status. On the

lower side of the scale, the constructions A dongh B and A se B use particles,

neither allowing negation nor the aspectual marker. While dongh indicates an

identifying relationship between A and B, se signifies an explanatory function of

B to A. Thus, the strength of connecting the two elements in the first construction

is stronger than that of the second. From the light of the above observation, it is

possible to recognize a gradient relationship in the setting band from the verbs to

the particles. The gradient can be interpreted as a bundle of different

characteristics which are spread across the cline from the prototypical-to-deviation

relationship.

5.2.4 Summary

The analyses in this section have been selective. Nonetheless, they show that a

gradient relationship can be continued both inside the storyline and in other

supportive materials, exemplified by the transitivity cline and in the category of

equative constructions. The pivotal storyline is approached as the band of

grounding elements in the sense that the narrator pragmatically employs them to

restore the audience back to the storyline.

5.3 Foregroundedness through Transitivity

This section is concerned with one part of H2: The storyline in Iu-Mienh is

characterized semantically by transitivity of events.

In this and the following sections, the storyline is construed as being foregrounded

in that it perceptually emerges into the conceptualizer’s perception through energy

and action transitivity (5.3) and through sequentiality (5.4).

5.3.1 Energy Transitivity in the Billiard-Ball Model

A moving object is more conspicuous than a static one in the perception of the

viewer as schematized in Figure 34.

Page 162: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

136

<

Figure 34. The static and moving objects

In 3.2, it was mentioned that serial verb constructions (SVC) can be strung

together to form a multiclausal construction, and topic chains80 express a sequence

of actions. Aikhenvald (2005) states, “The order of components [in a sequence of

actions] is iconic (that is, it follows the temporal sequence of the subevents).”

Somsonge also states that one of the storyline materials that secure sequentiality

(sequentiality being the most important criterion that marks the storyline in Thai

according to her) is “a series of clauses” (1992:106). Therefore, SVCs, a

multiclause composed of SVCs, and topic chains are analyzed here with regard to

their conceptual structure of transitivity. For this purpose, Langacker’s billiard-

ball model, which is more suitable for grammatical analysis than the schema in

Figure 29 in 5.1, is utilized. Langacker’s model (1991b:283) expresses an action

chain as in Figure 35.

R R R R

Head Tail

Figure 35. Action chain

80 Note that Van Valin’s “topic chain” (2005) (originally labeled by Dixon), Aikhenvald’s “a sequence of

actions” (2005), and Somsonge’s sequence of “topic subject” (2002) are almost equivalent.

Static

Moving

V V

Page 163: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

137

Figure 35 can be applied to example (86a-b). (86a) has a complex composition.

An intonation break shows that it is composed of two sentences, but connected

tightly describing one event comprising subevents. Cl1 and Cl2 are SVCs

presenting a setting for Cl3. There is no intonation break between Cl4-Cl6, which

comprise the three-member switch-subject SVCs. Cl3-Cl6 is a topic chain.

(86a) (AS.030) St1 [Cl1 ] [Cl2 ] [Cl3 ] [Pause]

Mienh geh jakv jiex, nzaeng nzangv jiex, ninh cuotv daaih person ride boat pass paddle boat pass 3sg come_out come

St2 [Cl4 ] [Cl5 ] [Cl6 ]

ngatv nzangv mbienv ndortv duqv mienh nyanc press_down_on_hard boat overturn fall get person eat

Cl1 People rode a motorboat, passed (river), Cl2 paddled, passed, Cl3 it [the crocodile] came out; Cl4 pressed down on the boat, Cl5 (the boat) turned over (and people) fell off, Cl6 (crocodile) got people (to) eat.

Impressionistically, the audience would feel that this sentence is dynamic in its

movement and speed. Analytically, the billiard-ball model makes the dynamism

explicit. In (86b) below, the action chain schema in Figure 40 is superimposed.

(86b) (AS.030) Mienh geh jakv jiex, nzaeng nzangv jiex People ride boat pass paddle boat pass

ninh cuotv daaih ngatv nzangv 3sg (crocodile) come_out come press_down_on boat

mbienv ndortv duqv mienh nyanc overturn fall get people eat

The participants are circled, and the actions are labeled by the arrows. Each time

the verb is uttered by the narrator, an action occurs. This increase of new

information, particularly the information concerning action in the case of SVCs

and topic chain, is conceptualized as follows in Figure 36, where the action

increase is profiled with the black arrows.

Page 164: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

138

Act

ion

incr

ease

in a

sent

ence

Y ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

X Perceptual movement of action by the audience

Figure 36. The conceptual structure of action increase

in SVCs and topic chains

Axis Y indicates the action incrementation in the sentence spoken by the narrator.

Each time the narrator adds another verb, energy is transfered to the following

participant and the action/event goes forward. Axis X indicates the perceptual

movement of the actions from the viewpoint of the audience. As the audience

hears the sentence, they mentally visualize the movement of action. Particularly

in the case of SVC and topic chains, the density of verbs is high; thus, the

perceptual movement becomes evident; that is, it is foregrounded in the mind of

the hearer. From the CG point of view, the foregroundedness of the storyline is

thus conceptualized as a result of this energy/action transitivity capturing the

hearer’s attention. This is probably because the “order of components” or

“temporal sequence of the subevents” is “iconic” (Aikhenvald 2005:42).81

5.3.2 Energy Transition Rate in Action Chain

A faster object is more conspicuous than a slow-moving one in the perception of a

conceptualizer. From a CL perspective, a sentence composed of SVCs and

sequential serial verbs expressing action chains seem to have a sense of speed in

regard to presentation of information.

81 The page number of Aikhenvald 2005 is from the pre-publication draft.

Page 165: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

139

In example (86) used in 5.3.1, the sentences contain sixteen elements, of which six

are participants and ten are actions. The rate of action is 10/16 = 62.5%. A

paraphrase of (86) with explicit participants is given below as (87). The added

participants are indicated in bold.

(87) (modification of AS.030) Mienh geh jakv jiex, mienh nzaeng nzangv jiex people ride boat pass people paddle boat pass

‘People rode a boat to pass (river), people paddled a boat to pass,’

ninh cuotv daaih ninh ngatv nzangv 3sg(crocodile) come_out come it press_down_on boat

‘It (crocodile) came out and it pressed down on the boat,’

nzangv mbienv mienh ndortv ninh duqv mienh nyanc boat overturn people fall crocodile get people eat

‘the boat turned over, the people fell off, the crocodile got people to eat.

The energy transitivity of this artificial sentence is schematized in a modified

billiard-ball model in Figure 37.

Distance 1 2 3 4 5 6 7 8 9

1 0

1 1

1 2

1 3

1 4

1 5

1 6

1 7

18

19

20

21

Action chain ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○Energy transition

1 2 3 4 5 6 7 8 9

(The numbers in the upper row indicate the total number of words in the sentences; the circled numbers

actions expressed by the verbs.)

Figure 37. Energy transitivity ratio

The upper row indicates the total number of elements in these sentences, and this

is interpreted as the distance through which the action is transmited. The middle

row indicates the action chain composed of the participants and actions, starting

from the head terminating at the tail. The bottom row, with the circled numbers,

indicates the total number of energy transitions. Thus the rate of energy transition

of the sentence is gained by the following calculation:

Energy Transition ———————— = Energy Transition Rate Distance

Page 166: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

140

The energy transition rate of the sentence (87) is 47.6% (= 10/21). Compare this

with 62.5% of sentence (86a-b). This difference is reflected in the response from

the Iu-Mienh speakers, who describe (87) as “too long,” “boring,” “unnatural,” or

“We don’t speak like that.”82 The energy transition rate can be interpreted to

contribute to the sense of speed of the sentence. Metaphorically, it takes many

short steps to walk through the sentence like (87), whereas a setence like (86a-b)

can be walked through in fewer steps.

The Iu-Mienh SVC, sequential serial verbs and topic chains create sentences as

exemplified in (86), (88) and (89), which are far above the 47.6% information rate

of the artificial dull sentence (87).83

(88) (AS.015) Nzopv jiepv, nzopv gau nzopv! nzopv nzuiz gau pierce bear pierce DPCL pierce pierce shove DPCL

‘(She) stabbed and stabbed and shoved (the pole) and then

jiepv nyei nzuih baengx, daic mi'aqv bear POSS mouth die RSLT

(into) the bear's mouth (and) the bear died.’ (or ‘She jabbed and jabbed the bear, shoving the pole into its mouth, until the bear died.’)

The energy transition rate of this sentence is 75%.

(89) (AS.131) Guien aengx bun mienh mingh zorqv congx lo haaix Officer further let person go take gun and_so_forth

nzuonx daaih return come

‘The officer then let the people go get (their) guns and other things and come back.’

The energy transition rate of (89) is 62.5%.

82 Although example (87) is not ungrammatical, it is just unnatural. 83 “(87) is dull because all the implicit elements are made explicit, overloading the communicative task. But

the elements are there in the underlying structure because they can be correctly recovered. So the deletions do not undermine the coherence despite the lack of explicit lexical cohesion. In terms of storytelling, the deletions are obligatory if naturalness and hearer interest are to be maintained” (Purnell 2006, personal communication).

Page 167: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

141

This sense of speed in SVCs and topic chains is an important factor for

foregroundedness in the storyline as much as these constructions are frequently

used throughout the narrative. The faster the energy moves forward in a sentence,

the more foregrounded the storyline becomes. This is an interpretation of what it

means to have SVCs, sequenced SVCs in multiclausal constructions and topic

chains on Band 1-Primary storyline from the CL perspective.

5.3.3 Summary

In this section, the conceptual structure of foregroundedness through transitivity

has been investigated using the modified billiard-ball model, originally posited by

Langacker (1991c). In terms of the perception, a moving object, as opposed to a

static one, and faster movement, as opposed to slower one, are more perceptive to

the conceptualizer. The SVC and topic chains being the important and common

storyline materials, it is evident that their transitivity, both its energy movement

and the sense of speed and dynamism indicated by the information stride, has a

significant part to make the sentence stand out to the viewer of the narrative.

5.4 Foregroundedness through Sequentiality

This section is concerned with one more part of H2: The storyline in Iu-Mienh is

characterized semantically by sequentiality of events.

A condensed cluster of objects in a linear order is more conspicuous to the

perception of the viewer than scattered objects with no connection between them

as schematized in Figure 38. This is in accordance with two of the gestalt

principles of organization:84 “the principle of proximity” and “the principle of

good continuation” (Anderson, J. R. 1995:45).

84 They are: (1) the principle of proximity: “elements close together tend to organize into units,” (2) the

principle of similarity: “objects that look alike tend to be grouped together,’ (3) the principle of good continuation: a good continuation or well-shaped line tends to be recognized easily as opposed disrupted or a sharp bent, (4) the principle of closure and good form. (Anderson, J. R. 1995:44-47)

Page 168: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

142

< <

V V V

Figure 38. Scattered and condensed objects

The proposal in this section is that two groups of discourse markers (i.e.

sequential markers and topic markers) connect events in a sequence in the

storyline. The sequenced events, in turn, stand out as foregrounded, as depicted in

the right hand diagram in Figure 38.

The first group of discourse markers includes ziouc ‘then, and then, then soon,

consequently, so’ and cingx_daaih ‘therefore.’ The second group involves

nor_aeqv ‘as for,’ aeqv ‘talking about,’ nor ‘in case of, if,’ V + gau ‘V and then,

while,’ V + liuz ‘after finishing…’ and V + baac ‘after finishing….’ As has been

described in 4.3.3.8, the sequential marker ziouc identifies storyline material (cf.

Figure 28), in a similar vein to cµN1 ‘consequently’ in Thai analyzed by

Somsonge (1992b:197). By contrast, the materials in the second group all belong

to Band 8-cohesive as in 4.3.3.1 (cf. Figure 28). One of the aims in this section is

to show how these “off-the-storyline” materials contribute to sequentiality in

relationship to the foregroundedness of the storyline.

5.4.1 Sequential Markers

Strictly speaking, the conjunction cingx_daaih ‘therefore’ is not a sequential

marker per se. It may be correct to define it as a pure conjunction. It is, however,

grouped together with the sequential marker ziouc due to a common syntactic

Page 169: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

143

behavior; (i.e. both occur in the second position after the Subject and before the

Verb in a sentence) and a few semantic similarities, as will be seen.

5.4.1.1 Sequential Marker: Adverb or Conjunction?

Concerning the nature of ziouc, first, a seeming problem has to be pointed out.

Then, it will be shown that that very problem plays an important role in the

storyline. See example (90):

(90) (BS.117) Cl1

Ninh dorc ziouc goix ndiangx-gorn 3sg elder_sister then cut_down tree-root

‘Her elder sister then cut down the tree at its base;’

Cl2

ndiangx ziouc nauv tree then snap

‘then the tree fell down.’

(BS.118) Cl1

Ninh nziez ziouc zong njiec wuom-njaangh mi'aqv 3sg younger_sister then strike go_down water-pond RSLT

‘Her younger sister then crashed down into the pond,

Cl2

ziouc daic wuov njaangh mi'aqv then die there pond RSLT

then (she) died there in the pond.’

Several points can be observed from these examples. (i) Syntactically, the basic

word order of this construction is S + ziouc + V as in BS.117 and Cl1 of BS.118,

while S can sometimes be omitted as in Cl2 of BS.118. (ii) It is a sequential

marker at the discourse level, occurring in every clause in this example. (iii)

Semantically, it is difficult to separate the sequential function of ziouc from the

immediacy of a consequent event, especially in Cl2 of BS.117, Cl1 and Cl2 of

BS.118.

Page 170: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

144

Compare the dictionary definition of ziouc; one by Panh, a Iu-Mienh

lexicographer, another by Lombard-Purnell,85 both western linguists in Table 8.

Table 8. The definitions of ziouc

The problem is this: Is ziouc an adverb (Panh) or a conjunction (Lombard and

Purnell)? To this question, a broad description of “connectives” by Bussmann

(1996) may be illuminating in that he includes “conjunctions” and “conjunctive

adverbs.” He defines the “conjunctive adverbs” as follows:

Adverb which occurs as an independent constituent before the finite verb and which has a co-ordinating function, e.g. so in It was raining, so we stayed at home. Conjunctive adverbs can have other semantic and syntactic functions besides co-ordination, such as particles or adverbials (Bussmann 1996:95).

A summary of his description is diagrammed in Figure 39.

Connectives

Conjunctions Conjunctive Adverbs

Panh (1995:449) Lombard and Purnell (1968:171)

ziouc forthwith, right away, immediately so, then, after which

joc, -jouc (variants)

immediately (1995:180) Ø86

Co-ordinating Subordinating Co-ordinating Particles/Adverbials

Figure 39. Connectives

Furthermore, Clark (1988) has reported that the White Hmong has what she

defines “an inchoative conjunction,” meaning “an indication of the

85 Purnell et al. (forthcoming) puts the usage for ziouc, “Expresses a sequence of events happening without

delay, the second of which marks a result or consequence of the previous action or state. In some cases, it is used as a marker of future time.”

86 The absence of joc in Lombad and Purnell may be due to a regional difference. They researched mainly in Chiang Rai province in Thailand, whereas Panh is a speaker of the Lao Iu-Mienh variety, and migrated to the U.S.A. Zanh Gueix-Fong, a Chaing Rai variety speaker and Burgess have reported that joc is an old form of ziouc (personal communication 2005).

Page 171: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

145

commencement of actual or perceptual change, including such concepts as

realization and consequence” (1988:93). The Hmong los means “(and) then/so,

thus, therefore, yet, and it happens, and it turns out, with the result” (1988:93).

Whereas these meanings are closer to Lombard and Purnell’s conjunction-like

definition of ziouc in Iu-Mienh, Clark’s insight that the conjunction has the

inchoative (i.e. adverb-like or even aspectual) semantic value seems to be a

parallel situation to that of Iu-Mienh. Moreover, Zanh Gueix-Fong has explained,

from a native speaker’s point of view, that ziouc indicates “future.”87

Therefore, it should be safe to accept both properties in one entity ziouc: it

connects clauses and sentences into a discourse, and at the same time it pushes a

story forward with an adverb-like function.

Secondly, having recognized this “double function” in ziouc, its conceptual

structure can be elucidated. According to Langacker, a “particular linguistic

element can be retrospective, in the sense of making a specification concerning

the prior discourse, and/or prospective, by virtue of evoking the subsequent

discourse” (2001b:151). Figure 40 depicts the conception that ziouc as a

prospective conjunction “evokes” the subsequent event.

Figure 40. Prospective conjunction ziouc

Three viewing frames in the CDS are labeled from the left hand as “minus frame,”

“zero frame” and “plus frame.” The presently focused event (E1) in the zero

87 Personal communication 2005. Also Burgess.

CDS

··· > > ···

– 0

ziouc

+

E1 E2E E2

Page 172: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

146

frame (also called the “focus frame”) exists as a result of the previous event (E) in

the minus frame. While ziouc in the zero frame profiles the E1, it simultaneously

evokes E2 in the immediate prospect of the plus frame. The rightward projected

frame based on the zero frame represents the incrementation of CDS by ziouc.

Thus, the conceptual structure of ziouc is to profile two relations simultaneously:

the expectation of a prospective event and the conjoining of the present event with

the prospective event. Grammatically, the evoking of E2 and the incrementation

of CDS by ziouc is obligatory, in that a narrator is compelled to speak something

once he/she utters the word ziouc.

This analysis concurs with the aforementioned observation of similar kinds of

conjunctions in other languages, e.g. Thai by Somsonge, White Hmong by Clark.

Another example can also be cited from an African language (Tyap) as referenced

in 2.1.9, which is by Follingstad: “In the context of chronologically arranged

material such as narratives and procedures, s is a coordinating conjunction which

tends to imply temporal succession. As such, it tends to correlate with, but not

exclusively mark, prominent foreground events” (1994:169).

In sum, ziouc is a prospective conjunction. This is the reason (not the problem of

conflicting definitions, cf. Table 8) why ziouc indicates clauses on the storyline:

being a conjunction it keeps the sequentiality; being a prospective element, it

increments the discourse space.

5.4.1.2 Conjunction Cingx_daaih

In contrast to the prospective conjunction ziouc, cingx_daaih ‘therefore’ is a

retrospective conjunction. See Figure 41.

Page 173: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

147

– +

Figure 41. Retrospective conjunction cingx_daaih

Langacker’s analysis of therefore in English is quoted here due to its parallelism

to cingx_daaih: “therefore [and cingx_daaih] introduces and profiles a

relationship in which one proposition, labelled P2, follows from another, P1.

Moreover, P1 is identified as the proposition expressed in the previous usage

event, just earlier in the flow of discourse” (2001:149). Two differences between

cingx_daaih and ziouc are that (i) P1 is associated with the proposition in the

minus frame, as opposed to E1 which is projecting E2 expecting the plus frame (cf.

Figure 45); (ii) cingx_daaih profiles a logical relation between two propostions, as

opposed to ziouc which profiles the temporal sequence along the storyline. The

relation of the two conjunctions is diagrammed in Figure 42.

Conjunction Prospective Conjunction Retrospective Conjunction profiles profiles Temporal Sequence Logical Sequence ziouc cingx_daaih

Figure 42. Functions of ziouc and cingx_daaih

Page 174: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

148

5.4.1.3 Summary of the Sequential Markers

Both ziouc and cingx_daaih are conjunctions. They function to indicate the

sequence in a narrative. They differ, however, in that the former is a prospective

conjunction; the latter, a retrospective. Defining ziouc corresponds with Clark’s

(1988) analysis of a similar conjunction in Hmong as being the “inchoative

conjunction,” Bussmann’s (1996) definition of “conjunctive adverb” and the Iu-

Mienh native speakers’ intuition that it has the “future” sense. While cingx_daaih

sustains the sequentiality, ziouc introduces it. In this way, the progressive

sequentiality of ziouc makes the storyline proactive/transitive and tightly cohesive.

This causes the storyline to be foregrounded in the perception of the audience.

Thus, ziouc has more qualification than cingx_daaih to be on the storyline.88

5.4.2 Role of Some Cohesive Band Materials for Sequentiality

It has been explained that the materials in the cohesive band are placed far from

the storyline band in the salience scheme. They are used to repeat a part of the

previous sentence/discourse or thematic materials to secure cohesion. Thus, they

are assumed to be non-story-pushing elements.

The materials under investigation in this subsection are mostly clauses with topic

markers and adverbial clauses, including the “tail-head” linkage. These topic

markers are: aeqv ‘as for,’ nor ‘as, like,’ nor_aeqv ‘talking about’ and naaic

‘concerning.’ They follow immediately after the NP which they topicalize, and

the thus-formed topic clause stands before the main clause. The adverbial clauses

are constructed with gau ‘while doing…’ and ‘after doing…,’ liuz ‘after finishing

…,’ and baac ‘after finishing…’ They come immediately after a V followed by

an Object NP if the verb is transitive. Formed in this way, adverbial clauses are

followed by the main clause.

88 Purnell et al. (forthcoming) defines cingx_daaih as ‘and so, and then, after which, as a result of which,’ and

it ‘Expresses a sequence of events, though there may be some timebetween them. In some cases, it is used as a marker of future time’.

Page 175: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

149

Aeqv for the topicalized constituents and liuz for the adverbial clauses are selected

for the present investigation. It will be shown that the topic marker and adverbial

clauses also have a function of discourse incrementation; that is, pushing a story

forward along the timeline, when analyzed from the CG perspective.

5.4.2.1 Conceptual Structure of Adverbial Clauses

Here it will be shown that Iu-Mienh prefers to use topic-comment structure and

that a conceptual structure of this nature is important in understanding an aspect of

information structure in a sentence.

In Figure 43, the sentences of AS.001-009 are charted to show the tail-head

linkage, where the tail of the previous word/phrase is repeated in the first position

of a succeeding sentence. The repeated constituent functions as an adverbial

phrase/clause for the coming main clause. The left hand column is the adverbial

clause; the second, the topic or subject; the third, the VO or the predicate; and the

right hand column shows the grounding elements. The second, third and fourth

elements are for the nucleus of the matrix clause. The grounding of the last

column is described in the sense of the epistemic footing (i.e. in this case, the

clausal grounding through the aspectual markers), in 5.2.2.1. The storyline verbs

are in boldface, and they are repeated in the adverbial clause.

Page 176: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

150 Ground

(margin) Figure

(nucleus) S# Adverbial Cl. Topic/Subject V + O Grd./Asp (1) Maaih i_gox mienh

have couple person za'gengh jomc really poor

nyei. PRS.ST

(2a) Jomc nyei, poor PRS.ST

ninh 3sg

mingh maaic gong go sell work

(2b) lorz hnaangx nyanc. look_for rice eat

(3a) I_gox married_couple

ndaam jienv ndaamx carry SML.ACT carrying_pole

(3b) hnangv Janx-Kor^lormx nor like Northern Thai as

hnaeng, dangle

(3c) hnaeng hnaeng hnaeng mingh. dangle dangle dangle go

(4) Mingh gau go DPCL

hmuangx dark

aqv CHG-O-ST

(5a) Hmuangx aqv, dark CHG-O-ST

taux wuov ndaamv-jauv, reach there half-way

(5b) hnangv yie nyei lamz nor like 1sg POSS granary as

zoux daaih lauh haic make come long-period very

aqv CHG-O-ST

(6) Taux wuov reach there

hmuangx dark

aqv CHG-O-ST

(7) Aav, Ah

mingh haaix yaac go where TOP

mv mingh taux bieqc lamz bueix no go reach enter granary sleep

(8) [they] Nyorng bieqc lamz creep enter granary

mingh go

(9) Jiepv bear

njiec dorn give_birth_to cub

Figure 43. Tail-head linkage as an adverbial clause

Semi-literal translation: (1) There was a couple, who were poor. (2a) Being poor, they would go for selling their labor, (2b) earning their living. (3a) The couple would carry a pole (on their shoulders for hanging goods on both ends), (3b) just as the Northern Thai would do, (3c) went swinging, swinging and swinging. (4) As they went, it started to get dark. (5a) It was getting dark; (while they were still) on the way, (5b) (and there was an old granary) just like my granary, (which) had been used for a long time. (6) (By the time they) reached there, it was dark. (7) Ah! Wherever they went, they had never gone as far as this─having to sleep in a granary. (8) Anyway (they) crept into the granary. (9) A bear had given birth to a cub (there).

Sentence (1) is a presentational structure marked by maaih ‘there was’ or ‘have,’ a

stative verb and the present state aspectual particle nyei. After this (2a) starts

pushing the story with mingh ‘go’ in the main clause. (2a)-(2b) consist of a two

component-SVC. (3a) has a construction V1 + jienv + V2 ‘simultaneous action’

material from Band 2, followed by (3b) an author’s explanation with hnangv…nor

‘like…as’. Thus (3a) and (3b) provide a setting for (3c). Two verbs of mingh in

(2a) and (3c) are repeated as an adverbial clause in (4). Though the verb in (4) is

Page 177: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

151

a stative hmuangx ‘dark,’ it is grounded as a storyline verb by the change-of-state

aspectual particle aqv; that is ‘it became dark’ or ‘it started to get dark’ indicating

the early stages of getting dark, rather than ‘it was dark.’ This event of hmuangx

aqv is repeated as an adverbial clause providing a setting for the further event taux

‘reached’ in (5a). (5b) is another author intrusion marked by hnangv ‘as’

introducing an illustration from his situation outside the story. After the author

intrusion, the audience has to be brought back to the story; thus the event taux

‘reached’ of (5a) is repeated as an adverbial clause, and the event of hmuangx aqv

‘it became dark’ or ‘it had become dark (by then)’ is also repeated as a main

clause marked by aqv, which has the effect of restoring the storyline in (6). (7) is

irrealis. (8) with zero anaphora topic subject is a directional SVC, advancing the

storyline forward. (9) is a flashback.

In (2a), (4), (5a) and (6) of this example, their matrix clauses, of which the

adverbial clause is a repeat, are on the storyline. Though (6) does not present a

new event, it restores the narrator-audience rapport back to the storyline after the

flashback. An adverbial clause in the tail-head construction is in Band 8, and is

used to indicate “anaphoric temporal clauses” (Somsonge 2002). These three

examples (2a, 4, 5a) accord with Somsonge’s statement: “the main clause

preceded by the anaphoric temporal clause is sequenced with the previous time-

moving clause and thus triggers the perception of temporal movement”

(2002:146).

In terms of Thompson and Longacre (1985:210-1), the adverbial clause column of

Figure 43 is called “ground” or margin, and the columns ranging from the

topic/subject to the grounding aspectual markers “figure” or nucleus. In analyzing

interparagraph linkages in Philippine languages by identifying many devices,89

Longacre (1968:1-222) globally uses a concept of ground and figure. His example

of one paragraph from the “Bakid” story in Itneg is shown in Figure 44 (1968:57):

89 E.g. Longacre discusses tail-head linkage, linkage through summary, consecutive time horizon linkage,

linkage through particles, lexical linkage, head-head linkage, etc. (1968:1-222)

Page 178: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

152 Ground Figure

He went. BU1

When he arrived in the forest he chopped the trees. BU2

When he had chopped them he shaped them. BU3

When he had shaped them he went home again. BUn

Figure 44. Ground and Figure in paragraph linkage

It can be observed immediately that the clauses in the ground column represent

old (given) information and the figure new. This simple chart has considerable

potential for analyzing the conceptual structure of information in sequentiality.

Such is the case in Iu-Mienh, which uses topic-comment structure extensively.

Figure 43 shows that within the figure column there is a distinction between the

topic subject carrying the old (given) information while the VO column carries the

new. An exception is lamz ‘granary’ in (5b), which was drawn from a shared

world between the narrator and the audience outside the story marked by

hnangv…nor ‘like/as.’ This multi-layered topic-comment structure is analyzed by

Court (1986:36) as shown in Figure 45.

Page 179: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

153 S Top Com Cl Cl

Top Com Top Com AdvP Cl AdvP Cl Adv VP Adv Top Com Cl Cl VP VP V Ih_hnoi maaih maengc yiem njang_hnoi daic yaac mv hiuv duqv today have life DUR-ASP tomorrow die indeed not know be-possible

‘Even if today we are alive, we still do not know whether or not we will be dead tomorrow.’

Figure 45. Multi-layered topic-comment structure

(Court 1986:36)

This multi-layered topic-comment structure can be expressed in accordance with

the ground-figure schema as in Figure 46.

Top 1 ‘today’ Ground

Com 1 (= Top2) ‘have life’

Com 2 (=Top3) ‘tomorrow’

Com 3 (=Top4) ‘die’

Com 4 (=Top5) ‘indeed we do not know’ Figure

Figure 46. Multi-layered figure-ground relation

This can be referred to as a multi-layered figure-ground structure of a sentence.

In Iu-Mienh the outermost layer is coded by the adverbial clause with or without

the topic marker. Almost anything can serve in this position. It can be filled with

Page 180: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

154

NPs, VPs, clauses, sentences with or without the perfective marker aqv and

summaries of what has been told already in the narrative. They can be marked by

the topic marker aeqv or not. The innermost layer receives the highest focus by its

end position. In other words, the more rightward the constituent goes, the more it

is focused.

This underlying multi-layered figure-ground structure allows the Iu-Mienh

speaker to keep sequencing adverbial clauses before the main clause. Especially,

it is possible to do so when he/she utilizes the topic marker aeqv. This is the

reason why this particle is defined in two seemingly different ways: as (i) a

“topicalizing particle” by Court (1986:52) and (ii) a “[p]article marking the end of

an introductory conditional clause” by Lombard and Purnell (1968:4).90 Then,

when should the speaker stop adding clauses? The speaker can add conditional

clauses (six_gorngv…nor ‘if’), reason clauses (weic_zuqc ‘because,’ laaix ‘due

to’), descriptive clause, until he/she grounds the sentence with the clausal

grounding elements: the perfective aspectual particle aqv and the resultative

aspectual marker mi’aqv. Example (91) ends with another grounding element, an

illocutionary particle na’maah ‘I am telling you’ or ‘you should understand me.’

90 Purnell et al. (forthcoming) gives a fuller definition, ‘particle used to indicate the end of an introductory

conditional clause or a topic clause’.

Page 181: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

155

(91) (CI.60) Cl1

Dauh ong cingx_daaih zimh jomc nc norm dorngx CLF old_man therefore investigate end that CLF place

‘This old man, therefore, has brought the investigation to a close there’

Cl2

zimh cuotv gorn daaih aeqv investigate go_out root come TOP

‘discovering the root (of the problem); namely…’

Cl3

norqc_guv_long nduov jung mingh nyanc biouv owl deceive barking_deer go eat fruit

‘The owl enticed the barking deer to go and eat the fruit’

Cl4 Cl5

ninh_mbuo zorqv sung mborqv mborqv daaih 3_PL take ginger pound pound come

‘they crushed a ginger’

Cl6

nanv ninh nyei wuom cuotv daaih aeqv squeeze 3sg POSS water go_out come TOP

‘they squeezed the juice out of it, and so,’

Cl7

dorh sung-wuom mingh laapc norqc_guv_long m'zing na’maah take ginger_juice go put owl eye I_am_telling_you

‘they took the ginger juice to pour into the owl's eyes, you see.’

By the use of the topic marker aeqv, the narrator can suspend the progression of

the storyline for the audience and is able to prepare them for the next clause. Each

time a clause with new information is spoken, it immediately becomes a ground

for the next clause. The suspended audience will be settled when the whole

sentence is grounded by the grounding element, usually aqv or mi’aqv. Though

they can be used in an adverbial clause, when the sense of change-of-state aspect

in the particle aqv and the sense of resultative aspect in mi’aqv are pragmatically

used by the speaker, they serve to indicate an ending point of suspension. The

ending of (91) is rhetorical. Unlike the expected aqv or mi’aqv, the narrator used

an illocutionary particle, directly addressing the audience; thus, Cl1 to right before

the particle in Cl7, the audience are listening to the narrative objectively or with

Page 182: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

156

less emotional engagement. Suddenly, at the end, that is, at the focus positon of

Iu-Mienh sentence structure, the audience is caught by the narration with greater

empathy. Figure 47 shows that the more rightward a constituent goes in a string

of multiclauses, the more it is focused.

Ground

aqv/mi’aqv FocuTopic Figure

or s

na’maah

Figure 47. Rightward focus structure

This explains several phenomena. First, it can explain a difference between (92a)

and (92b). While Court (1986:252) analyzes (92a) and (92b) as having the same

meaning, an alternative analysis from the perspective of the rightward focus

structure will show a difference in profile.

(92a) (Court 1986:252) Yie zoux gong yiem biauv 1sg do work be_in house

‘I work at home.’ (translation by Court)

(92b) (Court 1986:252) Yie yiem biauv zoux gong 1sg be_in house do work

‘I work at home.’ (translation by Court)

In (92a) yiem biauv is profiled, whereas in (92b) what is profiled is zoux gong.

This difference should be reflected in the translation, e.g. “In the house I work”or

“I work in the home” (92a) vs. “I work at home” (92b).

From the perspective of the rightward focus structure, which iconically

corresponds with the “from-general-to-specific” order of the topic-focus structure

in Iu-Mienh, (92b) seems to be less marked. That is, it is more natural than (92a).

Page 183: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

157

The pattern of (92a) has been increasing during the past ten to twenty years in

Thailand, as well as in the U.S.A. as in Court’s analysis above (92a-b) (c.f. his

informants were in California). This is probably due to language contact. See the

same word order with Thai (93a) and with English (93b).

(93a) (Thai)

pHom5 tHam1 Na˘n1 tHi˘3 ba˘n3

1sg do work at home ║ ║ ║ ║ ║

Yie zoux gong yiem biauv 1sg do work be_in house

‘I work at home.’

(93b) (English) I work at home

║ ║ ║ ║ Yie zoux gong yiem biauv 1sg do work be_in house

It seems that the younger generation of Iu-Mienh speakers feel as if the word yiem

in this pattern is a preposition rather than a verb.

Second, the topicalized constituents and adverbial clauses share the same

conceptual structure in that they set the ground in which the succeeding clause

gets the attention of the hearer/audience. This function of setting the ground for

the upcoming figure is similar to the discourse expectation of the prospective

conjunction discussed in 5.4.1.1. As such, the adverbial clauses and the topic

marker, though ranked in Band 8-cohesive, have an effect of foreshadowing an

upcoming story in the mind of audience before the story is told. This is a

significant contribution of Band 8 to Band 1.

5.4.2.2 Topic Marker and Incrementation of CDS

How can a narrator continue holding the floor of narration by the use of aeqv?

Levinsohn (1994:7) proposes that the topicalization’s main function is to indicate

discontinuities while maintaining “the overall unity and continuity”. He states:

Page 184: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

158

“This is because topicalized constituents have a BIDIRECTIONAL function: (a) they

serve as a POINT OF DEPARTURE for what follows and (b) they indicate the primary

BASIS for linking what follows to its context.

The conceptual structure of this bi-directional function of the topic marker is

discussed by Langacker (2001b), which is applicable to aeqv in Iu-Mienh. See

Figure 48.

> >

– 0

aeqv

+

PROP

Figure 48. Topic Marker aeqv

Langacker’s explanation is as follows:

A topic marker refers schematically to the thing profiled by the noun phrase with which it combines; the nominal referent is represented as a circle in the focus frame. Such marker is prospective by virtue of signaling that the profiled entity will function as a conceptual reference point for purposes of interpreting a subsequent proposition, as shown in the plus frame. But it is also retrospective, in the sense that the topic needs to be an entity already accessible in the prior discourse. Thus the profile is shown as correspoinding to a thing present in the minus frame. (Langacker 2001b:152)

This bi-directional or retro-prospective function of the topic marker can be

observed in (94), where the sentence starts with this topic marker. Here nor aeqv

is a variant of aeqv.

(94) (FG.020) Nor aeqv, norqc_meix ziouc hnyouv mv nqaai aqv So then sparrow then be-dissatisfied CHG-O-ST

‘So, the sparrow then became unhappy.’

The preceeding sentence (FG.019) recounts as this:

After going to bed, a dead tree snapped down to frighten to wake the grasshopper up; without remembering keeping his legs bent, he stretched out his legs and kicked the sparrow's goiter hard turning up the nape of its neck.

Page 185: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

159

Then comes (94), viz. (FG.021), followed by a DS of the sparrow, saying:

“From the beginning I told you I wouldn't give you my house to sleep in. No matter how you said that you would keep your legs bent, this time (you) kicked my goiter up to the nape!”

5.4.2.3 Adverbial Clauses and Incrementation of CDS

As opposed to the bi-directional topic markers discussed in 5.4.2.4, liuz ‘after

finishing…,’ baac ‘after finishing…’ and gau ‘while doing…’ or ‘after doing…,

are more monodirectional. They are prospective. Compare (95a-d).

(95a) *nyanc liuz hnaangx eat finish rice

(unnatural)

(95b) nyanc liuz hnaangx aqv eat finish rice PFT

‘I have finished having a meal.’

(95c) nyanc liuz nhaangx, (yie) cingx_daaih daaih eat finish rice (1sg) therefore come

I have finished having a meal, so I came.

(95d) nyanc liuz nhaangx, (yie) ziouc daaih eat finish rice (1sg) consequently come

‘Having finished a meal, I came right away.’

All but (95a) are used in actual situations. The Iu-Mienh speakers’ response to

(95a) would be “So then what?” or “and then what happened?” This means the

clause nyanc liuz hnaangx is a fragment until either it is grounded by an aspectual

marker or followed by a main clause. The former likely occurs in a conversation

or DS in a discourse; the latter tends to be found in the storyline of a narrative.

An example of the latter is (96).

Page 186: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

160

(96) (BS.081) Cl1 Cl2

Mbuox liuz ninh ziouc mingh mi'aqv tell finish 3sg then go RSLT

‘Having finished telling her, he then has gone.’

Both the narrator and hearer cannot stop after the liuz clause. This adverbial

clause inevitably evokes in the hearer’s mind an expected comment or focus.

Utilizing Langacker’s CDS incrementing model (2000:265), the conceptual

structure of the liuz clause can be explained as in Figure 49. Diagram (a)

specifically depicts liuz while (b) gives the general idea of the increment.

(a) (b)

CDS

increment

CDS

anchor

V(old)+liuz new

Figure 49. CDS increment through adverbial clause

In the construction V + liuz, the V carries the old information (a). The whole

construction’s primary function is to profile the space for new information. In

general (b), this kind of adverbial clause (using liuz, baac and gau) takes an

anchor in the CDS and increments a new CDS. This conceptual structure differs

from aeqv in that while aeqv is both a retrospective and a prospective element

(e.g. Figure 48), this advervial clause is prospective. The effect on the audience in

hearing the adverbial clause that contains liuz, baac and gau is to cause an

expectation or a whetting of the appetite to listen for more. Naturally, it has an

impact on the sequentiality, which is created in the narrative mental space of the

audience.

Page 187: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

161

5.4.2.4 Summary

In this subsection, a role for both adverbial clauses and topic markers has been

identified as having some contribution to the storyline by their double functions.

That is, they are conceptually retrospective and prospective in relation to the

proceeding and succeeding discourses. Based on the rightward focus characteristic

of Iu-Mienh, where the leftward constituents provide ground and the right-most

constituent profiles focus, both the topic markers and adverbial clauses take the

left-most position in a sentence. While the leftward elements retain a relation with

the preceding discourse, the topic marker aeqv and adverbial clauses V + liuz

(additionally, V + gau and V + baac) increment or update the CDS with new

information; thus, they are prospective. A subtle difference between the topic

marker aeqv and the adverbial clause V + liuz is this: the former is bi-directional

(i.e. both retrospective and prospective), the latter more prospective. The

conceptual function of these linguistic constructions, the members of Band 9-

cohesive, is to keep an overall cohesiveness of the narrative while at the same

time evoking an expectation of new information. The kind of sequentiality caused

by them is implicit but existing in the mind of the narrative audience.

5.5 Summary of the Chapter

At the outset of this chapter, a narrative was established as a cognitive entity on

the stage, mentally viewed by the conceptualizer. The chapter has investigated

three areas: a gradient relation in the salience scheme, the conceptual structure of

event/action transitivity and the conceptual structure of event sequentiality. First,

a prototypical transitive verb mborqv ‘hit/beat’ and a less dynamic verb gorngv

‘say/speak’ were analyzed in terms of the twelve transitivity parameters (Taylor

2003) to show how the other constructions that use gorngv are distributed on a

scale of a dynamic-static nature. This prototype analysis of transitivity category

has suggested that there is a gradient within verbs and constructions of the

storyline not just in the spectrum or in gradual degree of more dynamic and more

static constructions across off-the-line materials. This gradient is caused by two

Page 188: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

162

factors, although at first glance it appears disordered due to the same forms being

used in different bands and different forms in one band. That is, (i) some lexical

verbs in Iu-Mienh have a variety of grammatical functions, such as aspectual

markers and directional verbs functioning similar to English prepositions and

adverbs; and (ii) the aspectual markers and illocutionary sentence final particles

function pragmatically and epistemically as grounding elements, relating some

materials in lower bands with the storyline and the pivotal storyline.

Secondly, the conceptual structure of action/event transitivity manifested in some

specimens of SVCs, a series of action chains and topic chains have been analyzed

using a modified billiard-ball model, originally developed by Langacker (1991c).

It has shown that the Iu-Mienh SVCs together with topic chains are one category

of constructions that give the audience of a narrative a sense of movement and

speed due to the iconic relation between constituents and subevents with regard to

time-movement. Such an analysis assumes the existence of human cognitive

ability to perceive a fast moving object as more salient than a static or slow-

moving one.

Thirdly, the conceptual structure of sequentiality has been analyzed through the

investigation of two groups of constructions: the sequential markers and cohesive

materials. Group one includes ziouc and cingx_daaih. Using Langacker’s notion

of “retrospective” and “prospective” linguistic elements, ziouc has been analyzed

as having a prospective profiling semantic feature, while cingx_daaih profiles a

retrospective referent. This explains why this kind of conjunction in other

languages is sometimes called “inchoative conjunction” or “adverbial

conjunction.” From the CG perspective, ziouc is a prospective conjunction;

cingx_daaih is a retrospective conjunction. The former naturally pushes the

storyline forward even though it is not a verb. Group two involves two kinds of

cohesive materials: the topic marker aeqv and the adverbial clause V + liuz (as a

representative of the group including baac and gau). The former profiles bi-

directional cohesion; and the latter profiles prospective reference point in a new

mental space. Though ranked in the cohesive band, adverbial clauses have a

Page 189: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

163

prospective semantic property in the narrative, contributing to a formation of

sequentiality. The underlying principle of these analyses is this: a cluster of

tightly sequenced objects in line is more salient to the human mind than

disconnected and scattered ones. This sequentiality is marked primarily by the

prospective conjunction ziouc and adverbial clauses, and secondarily by the topic

marker aeqv. It influences the viewers (listeners) of narrative to see the storyline

emerge as perceptually foregrounded.

Recapitulating the main idea of this chapter, three words are the keys: gradient,

movement and sequence.

Page 190: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

CHAPTER 6

CONCLUSION

6.0 Introduction

In the beginning, was the storyline verb, and the verb was with tense/aspect/voice,

and the verb was preterite. The preterite was a narrative tense. It was in the

beginning on the storyline. “Find the narrative tense and the storyline will be

given unto you,” said a textlingist. There came a cognitive linguist sent from a

narrative space, whose name was Foregroundedness. And he said, “Let

transitivity and sequentiality serve the storyline;” and the storyline became

foregrounded perceptually in a canonical narrative space. Tense was a relation

between the textlinguist and the cognitive linguist. Then, a voice of an author

intrusion was heard from another space, “A new aspect I give unto you. That is,

go hand in hand.”

6.1 Summary of Findings

A semantic investigation into the storyline in Iu-Mienh has taken us through the

preceding four chapters. Our original goal was to account for the meaning of

storyline in discourse without separating grammar and semantics.

For that goal, in chapter 2, we surveyed twelve different linguistic theories and

schools or approaches and have categorized them into four basic attitudes with

regard to the semantic study of discourse and text. These are formal (or

descriptive), mental, sociolinguistic and symbolic approaches. While the second

approach provides a framework of discourse semantics from a broad cognitively

oriented perspective, the fourth approach, represented by Langacker’s Cognitive

Grammar, has been chosen as appropriate to achieve our goal of analyzing the

conceptual structure of storyline. This is because of its principal claim that a

linguistic unit, including a discourse, is a symbolic entity comprising a

phonological representation of concept and its semantic content, viz. the form and

meaning are inseparable within a linguistic and social context.

164

Page 191: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

165

In chapter 3, we have described some selected grammatical features of Iu-Mienh.

Some important characteristics of Iu-Mienh grammar that are relevant to the

investigation of storyline involve the following aspects:

(i) the extensive use of lexical verbs together with their derived

grammatical functional verbs, e.g. directional, aspectual and modal

verbs,

(ii) the frequent use of the Serial Verb Constructions, sequenced SVCs in

multiclause constructions and topic chains,

(iii) all these verbs have the same form, and

(iv) the pragmatic use of aspectual and sentence final particles.

Chapter 4 has taken us back to the inception of Longacre’s storyline theory with

his particular adherence to the preterite verb in Biblical Hebrew (wayyiqtōl) as a

guiding principle in the development of his theory. The chapter has also

compared two different definitions of storyline: from a textlinguistics perspective;

the storyline is the most foregrounded main line of narrative discourse whose semantic values are substantive, narrative, realis, dynamic, sequential and punctiliar, and whose morpho-syntactic marking is the preterite tense or completive aspectual verb form.

From a CL perspective:

the storyline is a perceptually foregrounded line of a narrative discourse. The line is foregrounded by two major causes: the sequence of events and the movement of events along a timeline toward its goal or conclusion, which is pragmatically planned by the narrator. The former is referred to as sequentiality of events, the latter transitivity of events. Thus, the storyline consists of sequentiality and transitivity of events.

Subsequently, the fourth chapter has presented the profile/plot analysis of seven

Iu-Mienh narratives, has analyzed the storyline, and has proposed the salience

scheme based on 715 sentences in this corpus. The chief characteristics found

there are:

(i) the storyline verbs are unmarked by default,

Page 192: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

166

(ii) clauses containing SVCs, topic chains, directional SVCs with

directional verbs (mingh and daaih), aspectual SVCs with aspectual

verbs (mingh and daaih) are on-the-line,

(iii) clauses containing not only verbal constructions but also the

conjunction ziouc and the development marker aengx are on the

storyline, and

(iv) the pivotal storyline band has (a) the aspectual markers, aqv

‘perfective’ and mi’aqv ‘resultative,’ and (b) the realis adverb

za’gengh.

In chapter 5, we discussed three areas: a prototype analysis of salience scheme,

the conceptual structure analysis of action/event transitivity and the conceptual

structure analysis of sequentiality. First, the prototype effect of transitivity has

been found in reference to the prototypical transitivity of the verb mborv ‘beat/hit’

on a continuum with varying constructions to the verb gorngv ‘say/speak.’ We

have also found a small example of the gradient relation among five

equation/copulative particles in Band 5-setting: benx, zoux, zeiz, se and dongh.

The materials in the pivotal storyline have been analyzed as grounding elements;

viz. the narrator pragmatically and epistemically establishes rapport with the

audience by bringing their attention back to an anchor into the storyline

(particularly after an author intrusion and empathy, a long direct speech, a

participant’s internal monologue).

Second, we have analyzed the conceptual structure of action/event transitivity, and

how it is related to the foregroundedness of storyline. For that, we have utilized

Langacker’s billiard-ball model with a few modifications, and applied it to some

selected SVCs and topic chains. It was shown that a construction of this kind

profiles “movement” and “speed”; viz. high transitivity and a high rate of action

movement. These two conceptual factors have a foregrounding effect in the

perception of storyline by the conceptualizer (hearer). Underneath this perceptual

foregroundedness lies the principle of human cognitive ability that perceives a fast

moving object as more salient than a static or slow-moving one.

Page 193: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

167

Third, we investigated the conceptual structure of event sequentiality and its

relation to the foregroundedness of storyline. Here the constructions analyzed fall

into two major groups. Group one contains the conjuction/sequential marker

ziouc and the conjunction cingx_daaih. Langacker’s notion of “retrospective” and

“prospective” linguistic elements proved to be useful in analyzing the conceptual

storucture of these entities. While the conjunction cingx_daaih profiles a

retrospective referent, ziouc profiles a prospective referent. Thus, the latter has a

function of incrementing the current discourse space (CDS), which, in turn,

creates the sequentiality resulting in making the storyline foregrounded

perceptually. Group two includes the topic marker aeqv and the adverbial clause

V + liuz. Though both are materials of Band 8-cohesive, they have some qualities

that move the storyline forward. Whereas aeqv profiles the bi-directional

reference in CDS, V + liuz profiles a prospective reference point in the newly built

mental space. This CDS incrementing function of the V + liuz is enhanced by the

characteristic rightward focus structure of Iu-Mienh. That is, V + liuz occupying

the left most position in a sentence with its prospective reference function (i.e.

pointing toward the right), the expectation and focus of attention by the hearer is

strongly directed to the right-most position of the sentence. This mechanism

creates sequentiality in the perception of narrative and evokes expectation for an

upcoming discourse space. The bottom line is that the tightly sequenced objects

in a line are more conspicuous to our cognitive perception than scattered and

unrelated ones. This perception of salience helps us mentally see the storyline

standing out or foregrounded.

6.2 Implications of the Findings to the Hypotheses

We started with the following hypotheses.

H1: The relationship between storyline and non-storyline in Iu-Mienh is a

gradient.

Page 194: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

168

H2: The storyline in Iu-Mienh is characterized both semantically by transitivity

and sequentiality of events and syntactically also by various linguistic

constructions.

Based on the findings in the present study, some conclusions are summarized

below:

1. The relationship between storyline and non-storyline in Iu-Mienh is a gradient.

Not only that, the gradient relation can be found inside the storyline itself and the

off-the-line bands as well. The nature of the gradient relation is that the cluster of

different features are lined up from a prototypical feature of a construction to the

deviations of it. Moreover, it is recognized that there is a pragmatic interaction

between Bands 1 and 7-9, through the grounding function of the pivotal storyline.

Thus, the relationship between storyline and non-storyline in Iu-Mienh is not a

binary distinction of foreground vs. background.

2. The storyline in Iu-Mienh is characterized both semantically and syntactically.

It is not characterized exclusively by syntactic codings.

3. The storyline in Iu-Mienh is characterized by various kinds of linguistic

constructions and configurations. It is difficult to identify the storyline

exclusively depending on a particular coding of verbs.

4. Semantically, the storyline is identified by perceptive foregroundedness

resulting from transitivity and sequentiality of events in narrative.

6.3 Suggested Areas of Further Study

The present study has been of a selective nature, and there are many other items

awaiting further investigation. First, from the category of ziouc, other similar

conjunctions should be studied further: joc (jouc) ‘immediately,’ yaac

‘and/also/even/besides/ still,’ youc (yoc) ‘again/and/also.’ It seems that ziouc and

joc (jouc) can be grouped together whereas yaac and youc form another group.

Clark (1991) seems to lump these two different groups, which correspond with the

Iu-Mienh items of these kinds, in other languages in one discussion of

Page 195: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

169

conjunction-cum-topicalizer. It seems that there is a distinction between ziouc and

yaac, which probably corresponds with c1 and k3 in Thai respectively.

Second, another conjunction weic_naaiv ‘for this reason’ has to be studied with

regard to its similarity and difference to cingx_daaih.

Third, on the pivotal storyline, where za’gengh has received some investigation,

there is one more important entity, viz. aengx ‘again/furthermore/and_then.’ This

can be termed as a development marker, which obviously works for the

furtherance of the storyline. It has some similarity with de (δέ) and kai (καί) in

Koine Greek discussed in Levinsohn (1992b:31-7).

Fourth, the construction V + liuz has two more members, baac and gau in this

group within adverbial clauses. The distinction between liuz and gau is that the

former is clearly prospective wheareas the latter sometimes profiles a temporally

overlapping relation between the event of the preceeding sentence and the event of

the succeeding main clause. As to baac it is difficult to find its difference from

liuz as far as our corpus of seven stories and discussion with the native speakers

are concerned. A speculation is that baac is an Iu-Mienh original word and liuz is

a loan from Chinese (了), functioning in the same way. If this is the case, a topic

of historical background of synonyms and how loan words are integrated into Iu-

Mienh until their origins are completely forgotten is an interesting area of study.

Fifth, the issue of gramaticalization in Iu-Mienh should be studied more. Court

may well be right in saying, “My impression is that the coverbs are not as

‘deverbalized’ as in Mandarin: for instance, they may take aspectual markers

freely” (1986:249). Kingkarn also states “lexical verbs in Thai still retain their

semantic properties in every case” (1986:245). If so, a kind of storyline theory

that depends on a particular verb form as a criterion for its identification has to be

revised, at least for Iu-Mienh (and possibly Thai).

Sixth, since some insights from common areal features in other authors’ studies

have been helpful to describe some aspects of Iu-Mienh grammar, further studies

Page 196: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

170

in other cognate languages will be needed and rewarding; viz. Kim Mun, Biao

Mon, and Dzau Min, as well as other non-related languages of the area.

Seventh, the development of metalanguage in Iu-Mienh will be an enterprise but

may be possible. In the course of elicitation and discussion with the Iu-Mienh,

one comment was interesting: naaiv joux waac maiv maaih gorn-baengx ‘this

sentence has no foundation.’ This explanation exactly fitted Langacker’s concept

of grounding. More tools of this kind, as well as encouragement, will be needed

for the Iu-Mienh to think and talk about their language in their language.

Page 197: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

APPENDIX A

A NAME OF THE PEOPLE AND LINGUISTIC

CLASSIFICATION

1. Yao or Mien?

There is confusion concerning the names of the Iu-Mienh people. A source of the

confusion concerning the various names of the people primarily comes from a

discrepancy between the name in their own language and names given by

outsiders in the national languages of China and Vietnam, which are often based

on characteristics of their traditional attires or names of place they inhabit. In

China the “Yao” is the name given to one of the fifty-five ethnic minority peoples

of the country, called Yáo-zú, the “Yao nationality.” Within the “nationality,”

some member groups are genetically related, and others are unrelated to each

other. This fact increases the confusion of language names. As research in

linguistics and anthropology has revealed this name confusion, Chinese scholars

have become more aware of the medley nature of the “Yao nationality.” For

example, Shū (1992:1), after mentioning that the Yao includes Mien, Bunu, and

Lakkia, clearly states, “the reality, however, is that they do not belong to the same

linguistic family: the Mien belongs to the Yao branch of Miao-Yao language

family; Bunu to Miao branch of the Miao-Yao; and Lakkia to the Dong Shui [viz.

Kam-Sui] branch of Zhuang Dong [viz. Tai] family.” In a recent study of Mao’s

Mien dialectology (2004:7), Bunu and Lakkia are not included; instead, four

separate but closely related “dialects” under the pure Mienic group are discussed

in terms of detailed lexical comparison. Their names in their own languages are

Mien [mjen31] (勉) or Iu Mien [ju31 mjen31] (优勉), Kim Mun [kim33 mun33] (金

门), Biao Mon [bjau31 mçn31] (标曼), and Dzau Min [dzau53 min53] (藻敏).

171

Page 198: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

172

The Iu-Mienh people in the north of Thailand under the present study speak an

almost identical language with Mao’s [mjen31] or [ju31 mjen31] 91 above.

Although the Thailand Iu-Mienh (or Mien) are also called as “Yao” by the Thai,

there is no other group under this term; the Yao in Thailand are equated to the Iu-

Mienh. Thus, when we talk about the Iu-Mienh, we mean the only “Yao” by Thai

definition in Thailand. In some contexts, particularly in reference materials

written in Chinese, “Yao” refers to the Iu-Mienh, the Kim Mun, the Biao Mon,

and the Dzau Min but excludes the Bunu and the Lakkia as is seen in Mao (2004).

Strictly speaking, however, a preferable terminology referring to these four

languages of Mao’s study is the “Mienic languages” following Aumann and

Sidwell (2001).

2. Linguistic Classification

A recent approach to the Hmong-Mienh language family in its own right, not

subsuming it in other groups, is from two directions. One is a practical and

pedagogical approach. For example, Suriya (1988) recommends the Hmong-

Mien language family should be separated from Sino-Tibetan and not be classified

in Austro-Thai so that it will get more attention by more scholars for further

research (1988:165-6).

Another direction is from an investigation on the reasons for the apparent

similarities between the Hmong-Mien languages and Chinese. For instance,

LaPolla (2001) has implied that further scrutiny of Hmong-Mien and languages

which share similar words, tone system, and classifiers may well reveal that

Hmong-Mien should be separated from the Sino-Tibetan phylum. This

disconnection hypothesis, he claims, parallels “the case of Vietnamese, which at

one time was also thought to be related to Chinese, due to its many Chinese-like

91 “Iu-Mienh” in Thailand is pronounced as [/iu31 mjen31], in contrast to Mao’s [ju31 mjen31] in China.

Page 199: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

173

features and words, but is now thought to be a Mon-Khmer language heavily

influenced by Chinese” (LaPolla 2001:227).92

Concerning the tonal and lexical similarities between Chinese and Hmong-Mien

which have been used for the claim that they are genetically related, Matisoff

(1983, 1991, 2001) has suggested an alternative solution; namely, an areal

diffusion explanation based on his theory of tonogenesis (Matisoff 1973:71-95) by

tentatively adopting Benedict’s Austro-Tai hypothesis (e.g. Figure 1, supura).

Matisoff writes, “these branches [Tai and Hmong-Mien] of the originally atonal

and disyllabic Austro-Tai stock became monosyllabic and tonal under Chinese

influence, diverging from Austronesian, which remained atonal and disyllabic”

(2001:317). This hypothesis is presented in Figure A, adapted from Matisoff

(1983:75, 1991:488, 2001:316) leaving the Chinese tone influence to Vietnamese

under Austroasiatic out from the original chart.

Austro-Tai Sino-Tibetan (atonal polysyllabic) (toniferous monosyllabic)

Austronesian Chinese Tibeto-Burman

Tai-Kadai Hmon

(adapted from

Figure A. Chinese ton

Furthermore, in agreement with the

hypothesized by Dixon (1997), Aikhe

92 LaPolla has stated this in the following con

[Chinese, Tai, and Homng-Mien] are related, bvery old loans, and the other features, such as tfor definite marking, etc. spread areally” (2001Hmong, similar points are stated by Li, “The sohave been residing suggests that (i) the superstand (ii) all of the languages in the area have beeis probably the major factor for an array of areanalytic (isolating) type of grammatical structuof the genetic classification of many of the lang

tones

g-Mien

Matisoff 2001:316)

al influence on Hmong-Mien

recent theory of punctuated equilibrium

nvald and Dixon (2001:1-23), Matisoff has

text: “Many scholars in China argue that the languages ut most linguists outside China feel the shared words are he similarities in the tone systems, the use of the classifier :227). Concerning the language contact in relation to the cio-historical background of the area in which the Hmongs ratum languages in that area have been Thai and Chinese, n subject to intense contact with each other. Thus, contact al features, such as mono-syllabicity, lexical tones, and an res. Contact is also the reason for the indeterminate status uages in the area” (Li 1991).

Page 200: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

174

suggested that the Mongol invasions in Eurasia in the twelfth and thirteenth

centuries might have caused that vast range of diffusion. 93 Thus, Matisoff

(2001:295) asserts, “Every one recognizes the validity of these basic macro-

groupings:” i.e. Austroasiatic, Sino-Tibetan, Tai-Kadai, Hmong-Mien and

Austronesian. That is, he does not subsume the Hmong-Mien under other

phylum. It seems that an increasing number of linguists outside China have

adopted the theory that Hmong-Mien should not be placed under other stocks.

93 Matisoff writes: “It is remarkable fact that a tremendous spate of tonogenetic and registrogenic activity

occurred all over the South-East Asian linguistic area in the twelfth and thirteenth centuries, triggered by the devoicing of the previously *voiced series of obstruents in many Middle Chinese and Hmong-Mien dialects, in Siamese and other Tai languages, in Karenic, in Burmese and many Loloish languages, and in Vietnamese, Khmer, and other Mon-Khmer languages” (Matisoff 2001:323).

Page 201: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

APPENDIX B

A HISTORICAL AND CULTURAL BACKGROUND

1. A Historical Background

According to Zhèn (1993:1), at the beginning of the Qín (221-206 B.C.) dynasty,

there was a mention in a Chronology about “three Miaos” or Sānmiáo (三苗),

which includes Miáo (苗), Shē (畲), and Yáo (瑶), a mixed group of the Proto-

Miao-Yao (i.e. Proto-Hmong-Mien) people. The Proto-Yao people lived in

Jiānghàn plain (江汉平原), a vast region extending from its west end Yíchāng (宜

昌) in Húbĕi province (湖北省) to the east end at around Wǔhàn (武汉) along the

Hànshuĭ river (汉水) and the Chángjiāng river (长江) with its south end at the lake

Dongting or Dòngtínghú (洞庭湖) in Húnán province (湖南省). They also lived in

a region east of Wǔhàn, west of the lake Poyang or Póyánghú (鄱阳湖). They

gradually crossed the Chángjiāng river southward and lived in the region south of

the lake Dòngtíng due to pressure from the feudal Hàn people who moved

southward (cf. Figure B, infra). Their names given by the Hàn people were after

the places of their habitation, such as Wǔlíng mán (武陵蛮),94 Chángshā mán (长

沙蛮), Wǔxī mán (五溪蛮) (Zhèn 1993:7). Collectively they were called as Mán

zú ( 蛮 族 ), namely “undeveloped people” or “barbarians” whose internal

ethnic/linguistic composition was unknown to the dominant people who came

from the north.

Anthropologicaly, in the Wèi (魏) (A.D. 220-265), the Jìn (晋) (265-316) dynasties,

and the Nánběicháo (南北朝) (or South-North dynasty) (316-589) period, those

who associated themselves around a legend of Pán Hù (盘瓠),95 a dragon dog96

94 Wǔlíng (武陵) is present day Chángdé (常徳), a city on the west coast of the lake Dòngtíng. 95 They are also called the “Bang Ku community.” 96 Lombard (n.d. 1st ed) mentions an importance of dog as an alleged progenitor of the tribe; dog is one of

several possible meanings of the word “Pan Hu.” Also the Yao’s abstain from eating dogs. However, not

175

Page 202: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

176

that has a human body riding on a deer,97 gradually became a distinct ethnic group,

the Yáo, separating themselves from other ethnic groups (Zhèn 1993:13-4).

Linguistically speaking, Nishida (1989:174) explains that the speakers of Hmong-

Mien (i.e. Miao-Yao) languages diverged to the west and north becoming the

Miao (i.e. Hmong) and the Bunu speakers, and to the east and south becoming the

Mien-Kimmun (i.e. Zhèn’s Yáo) speakers and the Shē speakers.

The aforementioned Pán Hù or Bang Ku community was originally referenced in

some Han Chinese historical books in the periods of the Sòng (宋) (420-479) and

the Liáng (梁) (502-557) in Southern dynasties, as Mò Yáo (莫徭), meaning “not

subject to compulsory service” (Fei 1991:18). Later, in the Suí (隋) (581-618) and

the Táng (唐) (618-907) dynasties their name was shortened to Yáo (徭). It was in

the Sòng dynasty that they migrated to Guăngxī (广西) and Guăngdōng (广东),

after which the two provinces have been the major residence area of the Yáo

where they inhabit half of both provinces (Zhèn 1993:16). Further down in

history in the Qīng (请) dynasty (1632-1911), they moved into Guìzhōu (贵州) and

Yúnnán (云南) provinces.

The Chinese character that designated the name Yáo during the Suí and Táng

dynasties up to the liberation was 徭, which means “used for compulsory labor

works.” After 1949, it has been changed to 瑶 meaning “gem,” the two characters

being homophonous. Thus, the people have been given different names beginning

with “barbarian” (for over 400 years), “those who are exempt from compulsory

labor works” (360 years), next, “those who are used for compulsory labor works”

(370 years), then “gem people” (50 years) and finally to the recent academically

recognized name Mien, that is, “human.”

in all areas of China. Purnell reported that his Mienh informant in China ate several meals of dog meat. (personal communication).

97 The Pán Hù legend is widely spread among the Iu-Mienh, Kim Mun (Pourret 2002:18), and Shē (Tán 1993:736).

Page 203: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

177

2. The Emperor’s Charter

An important Yao document written in Chinese came to the light of

anthropological and historical research in the 1970s, that is, a Charter or passport (

券牒) for the Yao people issued by Emperor (King) Ping or Píng Huáng (评皇).

Huáng’s study (1989) on this Charter includes 101 documents collected from five

provinces of southern China. Huang (1991:108) assumes that “King Ping’s

Charter was produced in the Zhen Guan era of the Tang dynasty.” Zhèn

(1993:72-92) emphasizes the importance of Pan Hu mythology found in this

document in understanding the Yao cosmology and the importance of its social

function.

The existence of this document became known to academics through extensive

research on various kinds of Yao documents in Thailand lead by Shiratori (1974),

followed by the publication of Yao Documents in Japan. In Thailand, Theraphan

(1991) translated a manuscript into Thai and English of the Charter or Jiex sen

borngv (in Iu-Mienh), bought from Mr. Woun Fei Sae Pan (Bienh Wuonh-Fei) of

Khun Haeng village, Ngau district, Lampang province in 1974. In Singapore, Tan

(1986) has translated one of the hand-written copies of the Charter into English,

several of which were found in Khun Haeng, the same village in which Theraphan

obtained her manuscript. Its first part tells of a legendary origin for the Iu-Mienh

people, while the latter part tells of the protection and privileges of the Iu-Mienh

people in crossing the mountain areas under the Emperor’s jurisdiction.

Comparing several scrolls of this Charter, Tan summarizes four common themes

found in them:

(1) The Yao [Iu-Mienh] ancestor was a dog having the surname Pieun [Bienh]. He was popularly called Pieun Hu [Bienh Hungh].

(2) This dog crossed a sea or seas to take the head of an enemy of a monarch, the latter most probably a Chinese emperor.

(3) The dog married a human being. She belonged to the monarch’s court and was given in marriage to the dog as his reward for his killing the enemy.

(4) The offspring of this dog-human union were the ancestors of the twelve Yao [Iu-Mienh] clans (Tan 1986:58-9).

Page 204: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

178

However, there is an intervarsity Iu-Mienh Students Club consisting of students

from Chiang Mai University, Maejo University, Rachaphat University, doing

ethnographic and historical research concerning point #1 above, disagreeing with

the idea that their ancestor was a dog.

3. The Enigma of the Migration Routes

The legends or myths which are included in many different manuscripts found in

various places bewilder researchers with an enigma of two possible migration

routes. From the description at the outset above, the first route can be postulated

as shown in Figure B.

(Base map downloaded from Google Earth)

Figure B. Origine and Migration Route of the Proto-

Mienic People

Page 205: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

179

On the ot

they saile

Guangdon

(1983:194

(2005: pe

preceded

with Figu

The seco

ethnograp

them as fo

(McKinnon & Bhruksasri 1983:194. drawn by Manus M.)

Figure C. The Crossing of the Sea

her hand, the second route starts from Nanjing. Setting out to the sea,

d through the Formosa strait offshore of Fujian and landed in

g. This is presented as a hypothesis in McKinnon and Bhruksasri

) (Figure C) and endorsed by the Tribal Museum, Chiang Mai. Somkiat

rsonal communication) explains that the crossing of the sea by the Yao

their arrival at Hunan province. Compare the second route in Figure B

re A.

nd theory is based on a sea odyssey recorded in the Charter and

hic interviews with the Iu-Mienh storytellers. Chob (1997) summarizes

llows:

According to the songs and stories of the Yao living in Thailand, they lived for some times on mountains, cultivated the ground. They then migrated until they finally established their abode at Nanking, which is near the sea. Later, during the years of the Tiger and Rabbit, the soil was devastated by drought, so that they could no longer cultivate it. Many of the people died. The twelve Yao clans fled, crossing the big sea by boats for seven days and nights until finally, with the help of the Great spirits, [the] Three Pures Ones, they were able to land at Le Chang district in Lao Chiao province of

Page 206: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

180

Guangdong, where they dispersed to establish themselves on the mountains. These Yao groups were called Pan Yao [viz. Iu-Mienh].98 (Chob 1997:5)

One of the reasons for this discrepancy between the two migration routes is a Iu-

Mienh word’s polysemy koriv [kHçi

], which can mean both “sea” and “lake” as

pointed out by Zhèn (1993:15). He implies “the sea” that the Yao are said to have

crossed in the Charter must have been the lake Dongting (e.g. Figure B). In order

to reach a convincing conclusion with regard to the actual course of their

southward journey, it will be necessary to investigate geographical feasibility,

careful differentiation between legends and historical facts in the Charter, textual

criticism of numerous manuscripts found across the international borders, the

successive migration’s psychological effects on the people’s epistemology formed

throughout the history, historical-comparative study between the four “dialects” of

the Mienics, and dating of loaned Cantonese99 lexical items in Iu-Mienh.100

4. A Historical Background to the Religion

Nonetheless, it is clear from both these documents and “the songs and stories of

the Yao living in Thailand” that their religion took its present form only after “the

crossing of the sea/lake.” The religious side of their life is summarized, by

Somkiat (2005: personal communication), as an amalgamation of the following

three: (i) “Three Pures Ones” (三清) (Chob 1997:5, supra) or Faam Cing with a

Daoistic flavored complex sacrifice system (which Tán refers to as guĭshén

chóngbài “worshipping demons and gods” (鬼神崇拝)(1993:629-32 [1982]), (ii)

98 Lombard (n.d., 2nd ed.) transcribed “crossing the sea,” originally tape recorded by Herbert C. Purnell, Jr. in

Maechan district, Chiang Rai province, Thailand in 1964. Beard, Warrick, and Saefong (eds.) (1995:1-9) record “The story of Bienh Hungh” in English. I interviewed and taperecorded a similar story told by Zeuz Gueix-Zoih (趙貴財) (then 77 years old) of Thammajarik village, Maechan district, Chiang Rai province on 16th Augst 1998.

99 Robert S. Bauer, at the Asia Lexicography Conference, 24-26 May 2004, Payap University, Chiang Mai, pointed out to me that comparative studies between Cantonese and Iu-Mienh would be extremely revealing because of a considerable percentage of loan words of the former into the latter. Ramsey (1987:286) briefly explains the phonological similarity between “Yao” (Iu-Mienh) and Cantonese. Matisoff (1991:489) mentions, “More than 50% of the lexicon in some Mien dialects are Chinese loanwords,” without reference to whether the loanwords are Mandarin or Cantonese. Purnell points out, “In the literary language and among many loanwords, the tone correspondences are very regular with a type of Mandarin. The ritual language would have more correspondences with Cantonese” (personal communication).

100 It is also possible that they took both routes.

Page 207: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

181

ancester worship, and (iii) animistic belief. Iu-Mienh “cultures and mode of living

run parallel with deep-rooted beliefs in ancestor worship, animism and spirit

worship, sometimes interplaced [i.e. replaced] as a result of missionary work with

a little Buddhism or Christianity. These practices govern every aspect of their

sexual, marital, family and economic life” (Preecha 1987:9).

5. Migration to Thailand

Migration from southern China to Thailand through Laos took place in the middle

of 19th century, estimated to be about 1880 by such scholars as Tsunemi (1980)

and others. A detailed study by Chob (1997:7-8) shows that one of the first Iu-

Mienh clans who migrated from China was that of Tang Jan Kuan (Dangc Zanh

Kuon (?) in the Unified Script) arriving at Laos in 1877, and eventually came into

Nan province of Thailand in 1896. Chob claims that there are four groups which

migrated into Thailand in large numbers at different stages, which can be

summarized as follows in Table A.

Group Lead by Clan Woman’s turban Areas

1 Chiang Rai – Nan group

Tang Jan Kuan (Phaya Kirisrisombat)

Tang, Pian

m’nqorngv-beu ping (rolled turban)

Laos→ Doi Ji, Doi Nam Mong, Doi Pha Chang Noi, Doi Pha Ji, Doi Pha Lom, Nam Lao (Nan), Chinag Khong, Thoeng (Chiang Rai)

2 Doi Ang Khang group

Tang Yao Foey

Pian, Zhao, Tang

m’nqorngv-beu paanx (criss-cross turban)

Laos→ Thoeng (Chiang Rai) → Doi Ang Khang (north of Chiang Mai)→ Myanmar→ Doi Ang Khang→ Mae Ngon Luang (south of Fang district) → Huai Chmphu, Doi Phalang (Chiang Rai), Fang and Mae Ai districts (Chiang Mai), Nong Waen (Mae Chan district, Chiang Rai)

3 Upper Chiang Rai group

Zhao Wang Seng and later by Bianh Lao Luu

Li, Zhao, Pian, Fong, Wang, Tang

m’nqorngv-beu paanx (criss-cross turban)

Nam Khyeung of Laos→ Nam Kham and Huai Kwang (Chiang Saen district, Chiang Rai), Doi Luang (Chiang Khong district), Lao Sip, Lao Shi Kuai, Phalae (Mae Chan district) → Wang Nua district (Lampang), Khlong Lan district (Kamphaeng Phet)

4 Evacuee group

(criss-cross turban)

Khun Mae Bong (Mae Chan district, Chiang Rai)

Table A. Four Iu-Mienh groups migrated into Thailand

Page 208: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

182

6. A Cultural Background

As is listed in the fourth column in Table A, a group identity is best expressed in

women’s attire, an important semiological code of the people, though different

dress does not mean different dialect. A study of elegant and intricate geometric

patterns of embroidery with color photographs of Laos can be found in Goldman

(1995), of Thailand in Lewis and Lewis (1984:138-49), the garments, head gears,

poems and songs of China Yao in Chen and Li (1995), and silver ornaments used

by the Iu-Mienh and the Kim Mun of China, Vietnam, Laos, and Thailand in

Pourret (2002:142-73).

“[S]widden (slash-and-burn) agriculture” and “opium cultivation” mentioned in

Kunstadter (1967:585) has gradually been discarded. The protection of the forest

through the Royal project began in 1931 and was reinforced by the National

Policy on the hill tribes in 1989,101 and the drug control began in 1952 with the

successful result of leaving about 200 rai of opium growing fields in 2002.

Instead, a tension between maintenance of village life on the mountains and a shift

to urban life, pursuit of university education and international communication

through the internet and the use of the Unified Script (a Roman based

orthography; c.f. Apendix C) are the present state of affairs. The unchanging

aspect, however, is that Iu-Mienh “exhibit a spirited, congenial and independent

nature, […] initial shyness when confronted by strangers, […] hard-working, […]

gracious and honest manner” (Schrock et al. 1970:689). The Iu-Mienh are “the

industrious, thrifty, wise and kind” people, inheriting these characteristics from

their “brave and persevering forerunners” (Huang 1991:122).

To summarize, the fostering of the aforesaid noble characteristics of the Iu-Mienh

in the face of the harsh nature of mountain life and the hardship of generations of

migration due to feudal rulers’ pressure and wars, and their progressive nature of

adjustment to bigger ethnic groups or host countries observable in their adoption

101 The Technical Service Club at The Tribla Musium, Chiang Mai (2004:71-9) lists the hill tribal policy

concerning these facts.

Page 209: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

183

of various religious practices and acceptance of contact languages are still in

process.102

102 Besides these positive characteristics, there are also problems. Cf. The Technical Service Club at The

Tribla Musium, Chiang Mai (2004:64-6).

Page 210: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

APPENDIX C

IU-MIENH PHONOLOGY AND ORTHOGRAPHY

Consonants

Bilabial Labio- Dental Alveolar Palatal Velar Glottal

Aspirated P p / ph / T t

/ th / Q q / ch /

K k / kh /

Voi

cele

ss

Unaspirated B p / p / D d

/ t / J j / c /

G g / k /

q / / /

Plos

ives

Voiced Mb mb / b / Nd nd

/ d / Nj nj

/ ∆ / Nq nq

/ g /

Aspirated C c / t Ésh /

Voi

cele

ss

Unaspirated Z z / t És /

Aff

rica

tes

Voiced Nz nz / d Éz /

Voiceless Fricatives Hu hu / „ /

F f / f /

S s / s /

Hi hi / C / H h

/ h / Voiceless Hm hm

/ m8 / Hn hn / n 8 /

Hny hny / ¯8 /

Hng hng / N /

Nas

als

Voiced M m / m / N n

/ n /* Ny ny

/ ¯ / Ng ng

/ N /

Voiceless Hl hl / l 8 /

Lat

eral

Voiced L l / l /

Approximant W w / w / Y y

/ j /

*This voiced nasal is preglottalized. The voiced nasals are preglottalized on three tones:

unmarked, -v, and –x, and not preglottalized on the other three tones. Purnell 1965

Table C1. Consonants phonemes and their

corresponding orthography

184

Page 211: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

185

Vowels Front Central Back

Close I i / i /

U u / u /

Close-mid E e / e /

O o

Mid er

/ ´ /

Open-mid Ae ae

/ E / Or or

/ ç /

Open A a Aa aa

/a/ / a˘ /

/ o /

Table C2. Vowel phonemes and their corresponding

orthography

Vowel glides Iu iu / iu /

Ui ui / ui /

Ei ei / ei /

Ou ou / ou /

Ie ie / i´ /

Uo uo / u´ /

Eu eu / eu /

Oi oi / çi /

Ai ai / ai /

Au au / au /

Aai aai / a˘i /

Aau aau / a˘u /

Table C3. Vowel glides and their corresponding

orthography

Tones

The tone marks by the Unified Script are in < > shown in the Table C4.

1. Mid-high level:

‹Unmarked› / /

3. High rise falling:

‹-v› / /

5. Mid-low rising:

‹-x› / /

7. checked syllable

‹-Cv› or ‹-qv› [ ]

2. Mid-falling:

‹-h› / /

4. Low rise falling: ‹-z› / /Ÿ /

6. Low level:

‹-c› / /

8. checked syllable

‹-Cc› or ‹-qc› [ ]

C = consonant, ‹-q› = glottal stop. (adapted from Downer 1963:136 and Purnell 1991:374)

Table C4. Tones and their corresponding orthography

Page 212: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

APPENDIX D

THE TEXTS OF SEVEN STORIES Notes on the translation: The “free translation” presented as follows is not actually

“free” in a sense that it is natural and flowing English. It reflects the boundaries

between clauses, and sometimes between components of serial verb constructions

in a clause, as indicated by subnumbers (e.g. 1.1) under the sentence number (e.g.

OH.001). The clause sometimes omits the verb, thus some clauses appear to be a

noun phrase.

Direct speeches are rendered more freely or dynamically since they do not affect

the analysis of the storyline. DSs are indented and are not numbered except for

long multi-sentence monologues. In the latter cases, the sentences are numbered

instead of clauses.

When words which do not exist in the original language are needed to be

expressed in English, they are in parentheses. Cultural information that needs to

be supplemented to understand the sentence is in brackets.

186

Page 213: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

187

Story 1 (OH): “A Story of an Old Husband and Bamboo Shoots” Old Husband Nqox-Gox Caux Mbiaic Nyei Gouv A story of an old husband and bamboo shoot

OH.001 1.1 Maaih dauh lov nor, have CLF man TOP vt clf n top 1.2 1.3 aapv ninh nyei sieqv longc wuov dauh nqox-gox. force 3sg POSS daughter marry that CLF old-husband vt pn.p poss n vt pn.dem clf n 1.1There was a man; 1.2he (tried to) force his daughter 1.3(to) marry that old husband.

OH.002 Ninh sieqv mv oix longc. 3sg daughter NEG want_to marry pn.p n neg v.mod vt

His daughter didn't want to.

OH.003 3.1 Ninh ndongc_haaix yaac 3sg no-matter-how and_yet pn.p adv conj 3.2 3.3 oix aapv ninh sieqv longc dauh wuov butv_zoih. want_to force 3sg daughter marry CLF that rich v.mod vt pn.p n vt clf pn.dem adj 3.1No matter what, 3.2he wanted (to) force his daughter 3.3(to) marry that man, (who was) rich.

OH.004 "Longc jienv aqv, maiv_zuqc kouv, maaih zinh_nyaanh longc." marry DUR CAS.CMD no_need_to tired have wealth use vt asp part. mod adv v.st vt n vt

"Just marry him, it shouldn't be bad. He is wealthy."

OH.005 5.1 5.2 5.3 Ninh sieqv aeqv, hnamv zuqc butv_zoih aeqv, zeiz nyei, 3sg daughter TOP think GOAL rich TOP be_right PRS.ST pn.p n part vt part adj part v.st asp 5.4 5.5 mv_baac gox haic ba'laqc mv puix. but old very too_much NEG appropriate conj v.st adv adv neg v.st 5.1As for her, 5.2(she) thought about the wealth, 5.3it was right, 5.4but (he) was so old 5.5(it) was not appropriate.

Page 214: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

188 OH.006 Ninh sieqv mv oix longc. 3sg daughter NEG want_to marry pn.p n neg v.mod vt

She did not want to marry.

OH.007 Ninh die gorngv, 3sg father say pn.p n vi "Longc maah. Meih gunv longc maah. marry SFT.CMD 2sg go-ahead-and-do marry SFT.CMD vt part.f pn.p adv vt part.f Ih_zanc mv hnamv hnangv lorqc. now NEG love only you-see adv.tmp neg vt adv part.f Gox mv_baac longc jienv meih manc-manc hnamv nyei." old but marry DUR 2sg gradually love AFM v.st conj vt asp pn.p adv vt asp

Her father said, “Marry him! You go ahead and marry him. It’s just now that you don’t like him, you see. Though he is old, once you marry, you will gradually love him.”

OH.008 Hnangv nc nor mbuox ninh sieqv. like that as tell 3sg daughter adv pn.dem top vt pn.p n

Like that, (he) said to his daughter.

OH.009 9.1 Zoux_gau after_that conj 9.2 nqa'haav ninh sieqv daav cuotv za'eix daaih later 3sg daughter plan come_out plan come adv.tmp pn.p n vt v n v.dr

9.3 zouv caeng mbiaic-gox nziuc mv bieqc wuov nyungc. cook pot bamboo_shoot-old chew NEG enter that kind vt n n v.st vt neg vi pn.dem n 9.1Do (like this) then, 9.2later, his daughter planned a plan; 9.3(she) cooked a pot of old bamboo shoots, (which people) cannot chew—such a kind.

OH.010 10.1 Zouv ziangx aeqv, cook finish so_then vt vi top 10.2 10.3 Mbenc daaih bun ninh nyei die nyanc. prepare come let 3sg POSS father eat vt v.dr vt pn.p poss n vt 10.1She finished cooking; 10.2(she) prepared (food) 10.3(she) let her father eat.

Page 215: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

189 OH.011 Mbuox ninh nyei die, tell 3sg POSS father vt pn.p poss n "Nyanc hnaangx aqv." eat rice INCHO vt n part.asp

She said to her father, "We're going to have a meal."

OH.012 12.1 12.2 Ninh nyei die gorngv naaic gaax ninh, 3sg POSS father speak ask try_and_see 3sg pn.p poss n v v part pn.p "Ih_hnoi zouv haaix_nyungc lai?" today cook what vegitabel adv.tmp vt q.cont n 12.1Her father spoke, 12.2(he) asked her, “Today, what did you cook?”

OH.013 13.1 13.2 Ninh nyei sieqv dau ninh nyei die gorngv, 3sg POSS daughter answer 3sg POSS father speak pn.p poss n vt pn.p poss n v "Zouv mbiaic." cook bamboo-shoot vt n 13.1His daughter answered her father, 13.2(she) said, “(I have) cooked bamboo shoots.”

OH.014 14.1 14.2 14.3 Nduov ninh nyei die zorqv daaih nyanc aeqv, deceive 3sg POSS father take come eat TOP vt pn.p poss n vt v.dr vt part 14.4 nziuc daaih aeqv, chew come TOP v v.dr part 14.4 naqv maiv njiec. swallow NEG go_down vt neg v.dr 14.1She deceived her father, 14.2he took (it) toward (him) and then; 14.3(he) chewed (it) and then; 14.4(he) could not swallow (it).

Page 216: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

190 OH.015 15.1 15.2 Ninh nyei die gorngv naaic gaax ninh nyei sieqv, 3sg POSS father speak ask try_and_see 3sg POSS daughter pn.p poss n v v part pn.p poss n "Meih zoux_haaix_nyungc zouv mbiaic ndongc naaiv gox? 2sg why cook bamboo-shoot as_much_as this old pn.p q vt n adv pn.dem v.st Fungc nyanc duqv bieqc?" how eat can enter adv vt v.aux vi 15.1Her father spoke, 15.2(he) asked his daughter, “Why did you cook such old bamboo shoots as this? How can I eat?”

OH.016 Ninh nyei sieqv gorngv, 3sg POSS daughter speak pn.p poss n v "Bieqc nyei lorqc. Mbiaic-gox gauh kuv. enter AFM you-see bamboo-shoot_old more delicious vi part part.f n-v.st cmpr v.st

His daughter said, “Yes you can. Old bamboo shoots are more delicious.”

Ih_zanc die nziuc mv bieqc now father chew NEG enter adv.tmp n v neg vi mv_baac manc-manc nziuc lauh deix kuv nyei. but gradually chew long some delicious AFM conj adv v adj adv v.st part

Father, you can’t chew it now, but if you gradually chew for some period, it will be delicious.

Yie mv oix longc nqox-gox. Die yaac aapv yie longc. 1sg NEG want_to marry old-husband father also force 1SG marry pn.p neg v.mod vt n-v.st n conj vt pn.p vt

I don’t want to marry an old man. But father, you force me to marry (him).

Die gorngv mbuox yie nqa'haav manc-manc hnamv nyei father speak tell 1sg later gradually love AFM n v vt pn.p adv.tmp adv vt part m'daaih dongh naaic mbiaic-gox. rightly that_is that bamboo_shoot-old adv idntf pn.demn n-v.st

You told me that I would gradually love him; that’s exactly the same with the old bamboo shoots.

Page 217: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

191

Se_gorngv die nziuc nqa'haav kuv nyei nor aeqv, if father chew later delicious AFM as TOP conj n v adv.tmp v.st asp top top naqv duqv njiec nor aeqv, swallow can go_down as TOP vt v.aux v.dr top top yie longc nqox-gox mv_baac duqv nyei. 1sg marry old-husband but can AFM pn.p vt n-v.st conj v.aux asp

If you, father, chew (old bamboo shoots) and feel it’s delicious, and if you can swallow it, I can marry the old man too.

Se_gorngv die haiz mv kuv, if father feel NEG delicious conj n vi neg v.st ndongc_haaix nziuc naqv mv njiec nor aeqv, no_matter_how chew swallow NEG go_down as in_that_case adv v vt neg v.dr top top yie yaac mv oix longc." 1sg also NEG want_to marry pn.p top neg v.mod vt

If you don’t feel it’s delicious, and no matter how you chew well if you can’t swallow it, then I also will not marry.

OH.017 17.1 Ninh nyei sieqv daav cuotv za'eix daaih aeqv, 3sg POSS daughter plan come_out plan come and_then pn.p poss n vt v.dr n v.asp top 17.2 17.3 gorngv ninh die nyei jauv jomc mingh speak 3sg father POSS way end go v pn.p n poss n vi v.asp 17.4 cingx_daaih maiv_zuqc longc wuov dauh nqox. therefore no_need_to marry that CLF husband conj aux vt pn.dem clf n 17.1His daughter planned a plan, and then; 17.2(she) talked (about) her father’s way, 17.3(her father’s way) came to an end; 17.4therefore, (she) did not have to marry that old man.

Page 218: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

192

Story 2 (FG): “A Story of Firefly and Grasshopper”

FG.001 gaeng-kiev caux caah_laauh_porngc nyei gouv. firefly and grasshopper POSS story n conj n poss n

A story of Firefly and Grasshopper

FG.002 Loz-hnoi gaeng-kiev caux caah_laauh_porngc buic gong. in_the_old_days firefly and grasshopper exchange work adv.tmp n conj n vt n

In the old days, a firefly and a grasshopper exchanged works [i.e. offer labor reciprocally or ‘you work for me and I work for you in return’].

FG.003 3.1 Buic gau gong, exchange DPD.CL work vt part n 3.2 wuov hnoi ninh_mbuo gorngv taux ninh_mbuo ganh nyei za'eix. that day 3_PL speak reach 3_PL self POSS plan pn.dem n pn.p v vt pn.p pn.reflexive poss n 3.1(They) exchanged works and then, 3.2that day they talked about each other's plan.

FG.004 4.1 4.2 Zoux gau gong hmuangx deix aqv, do as work dark some INCHO vt top n adj adv part.asp 4.3 ninh_mbuo gorngv taux hmuangx nyei jauv. 3_PL speak reach dark RELT NOM pn.p v vt v.st relt nom 4.1(They) did works and then, 4.2(it almost) became dark, 4.3they talked about (that it was becoming) dark.

FG.005 Gaeng-kiev gorngv heuc, n vi v "Nzuonx aqv, meih mv maaih douz, aqv_zuqc nzuonx aqv. return CAS.CMD 2sg NEG have light must return CAS.CMD v part. mod pn.p neg vt n aux v part.mod Aav_deix hmuangx daaih meih mangc mv buatc.” shortly dark come 2sg look NEG see adv.tmp adj v.dr pn.p v neg v

The firefly called saying, “Go home! You don't have a light, you must go now. It's going to be so dark soon so that you can't see.”

Page 219: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

193 FG.006 6.1 Caah_laauh_porngc gorngv, grasshopper speak n v 6.2 “Nzauh_heix haaix_nyungc laeh? be_warried_about what_kind will_you_tell_me? v q n part.f 6.3 Yie nyei zaux ndaauv nyei weqv. 1sg POSS leg long PRS.ST ASSERT pn.p poss n adj asp part 6.4 Yie ziex borngh_lorngh_qorngh haac_mv taux biauv mingh.” 1sg many long_leaps don't_you_think_I_could reach house go pn.p adj n idm v n vi 6.1The grasshopper said, “6.2What are you worring about? 6.3Don't you know that my legs are long. 6.4Don’t you see it won't take that many leaps to get home.”

FG.007 Gaeng-kiev gorngv, firefly speak n v “Yie korh_lienh meih aah. 1sg feel_sorry_for 2sg listen_Please! pn.p vt pn.p part.f Yie maaih douz nyei aah.” 1sg have light PRS.ST listen_Please! pn.p vt n asp part.f

The firefly said, “I just feel sorry for you. I have a light, you see.”

FG.008 8.1 Caah_laauh_porngc gorngv, grasshopper speak n v 8.2 “Meih maiv_zuqc korh_lienh yie lorqc. 2sg no_need_to feel_sorry_for 1sg you-see pn.p aux vt pn.p part.f 8.3 Yie nyei zaux ndaauv, yie tiux duqv henv. 1sg POSS leg long 1sg run can strong pn.p poss n adj pn.p vi v.aux v.st 8.4 8.5 Yie korh_lienh meih. Meih maaih douz mv_baac meih faix. 1sg feel_sorry_for 2sg 2sg have light but 2sg small pn.p vt pn.p pn.p vt n conj pn.p v.st 8.6 8.7 Meih nzuonx aqc_duqv maiv taux biauv. Yie tiux dangh taux biauv mi'aqv.” 2sg return difficult NEG reach house 1sg run awhile reach house RSLT pn.p v aux neg v n pn.p vi adv vt n asp 8.1The grasshopper said, “8.2You don't have to worry about me. 8.3I have long legs, I can run strongly. 8.4It's you I feel sorry for. 8.5You have a light, but you are small. 8.6It's too difficult for you to get home. 8.7If I run for a while I can get home.”

Page 220: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

194 FG.009 Cuoqv gaeng-kiev qiex_jiez haic. harass firefly get_angry very vt n v adv

He teased the firefly (and he) got angry.

FG.010 10.1 10.2 10.3 Zoux gau gong hmuangx aqv, ninh_mbuo ziouc saaix nzuonx. do DPD.CL work dark CHG-O-ST 3_PL so_then compete return vt part n v.st part.asp pn.p seq.mk v v 10.1After (they) did works 10.2it got dark, 10.3then they competed go(ing) home.

FG.011 11.1 11.2 Nzuonx aqv, gaeng-kiev maaih douz yaac maaih ndaatv ndaix. return CHG-O-ST firefly have light also have wing fly v part.asp n vt n conj vt n vi 11.1(They) started to set out, 11.2the firefly had a light and wing to fly too.

FG.012 12.1 12.2 Ninh ganh ziux ninh ganh nyei jauv yangh taux biauv mi'aqv. 3sg self shine_on 3sg self POSS road walk reach house RSLT pn.p pn.reflexive vt pn.p pn.reflexive poss n vi vt n asp 12.1He could light up his own road himself 12.2(and) walked back (and) reached home.

FG.013 13.1 13.2 Caah_laauh_porngc nzuonx maiv taux biauv, taux ndaamv-jauv aeqv, Grasshopper return NEG reach house reach half-way so_then N v neg v n v n top 13.3 mingh gaav norqc_meix nyei biauv bueix. go borrow sparrow POSS house sleep vi vt n poss n vi 13.1Grasshopper returned (but) did not reach house, 13.2when (he) went half way, 13.3(he) went (and) borrow [i.e. asked for a room to stay] the sparrow’s house to sleep in.

FG.014 Norqc_meix gorngv, sparrow speak n v “Meih naaic zaux ndaauv haic aah! Aav_deix ndamc zuqc yie aex.” 2sg TOP leg long very I_am_annoyed shortly stamp GOAL 1sg you_know pn.p pn.dem n v.st adv part.f adv.tmp vt partp n.p part.f

The sparrow said, “You, your legs are too long. Soon (you) will stamp on me, I’m sure.”

FG.015 Caah_laauh_porngc gorngv, grasshopper speak n v “Maiv gamh_nziex lorqc. Yie nyuotv jienv, yie maiv sung zaux.” NEG fear you_see 1sg bend DUR 1sg NEG stretch leg neg v.cmp part.f pn.p vi asp pn.p neg vt n

The grasshopper said, “Don’t worry. I will bend (my legs), I won’t stretch my legs.”

Page 221: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

195 FG.016 Norqc_meix gorngv, sparrow speak n v "Maiv aex, yie buatc meih naaic zaux gengh ndaauv. NEG I_mean_it 1sg see 2sg that leg really long neg part.f pn.p vi pn.p pn.dem n adv v.st Aav_deix ndamc zuqc yie. Yie naaiv biauv hepc nyei." shortly stamp GOAL 1sg 1sg this house narrow PRS.ST adv.tmp vt part pn.p pn.p pn.dem n v.st asp

The sparrow said, “You are sure you won’t, aren’t you? I see your legs are really long. You will soon stamp on me. My house is narrow.”

FG.017 Caah_laauh_porngc gorngv, grasshopper speak n v "Maiv haih fungc zoux. NEG be_able_to how do Neg aux adv vt Yie za'gengh jangx jienv nyuotv jienv, yie maiv sung yie nyei zaux." 1sg really remember DUR bend DUR 1sg NEG stretch 1sg POSS leg pn.p adv vt asp vi asp pn.p neg vt pn.p poss n

The grasshopper said, “You can't help it. I will really remember (that) I should bend (my legs), I will not strech my legs.”

FG.018 18.1 18.2 Norqc_meix maiv_haih_fungc_zoux, bun caah_laauh_porngc bueix. sparrow cant’t_do_anything let grasshopper sleep n idm vt n vi 18.1There was nothing the sparrow could do about it, 18.2(he) let the grasshopper sleep.

FG.019 19.1 19.2 19.3 19.4 Bueix gau aeqv, ndiangx-kou nauv, haeqv caah_laauh_porngc nyie, mv jangx zuqc aeqv, sleep DPCL TOP dead_tree snap frighten grasshopper wake_up NEG remember GOAL TOP vi part top n vi vt n vi neg vt part top 19.5 19.6 sung ninh nyei zaux ndamc zuqc norqc_meix nyei beuh stretch 3sg POSS leg stamp GOAL sparrow POSS goiter vt pn.p poss n vt part n poss n 19.7 mbienv faaux jaang-ngunz mingh. overturn go_up nape go vt v.dr n v.dr 19.1After (they) had been asleep for awhile, 19.2a dead tree snapped down 19.3(and) frightened (and) woke the grasshopper up; 19.4because he did not remember (to keep his legs bent), 19.5(he) stretched out his legs 19.6(and) kicked the sparrow's goiter hard 19.7(it) turned up to his nape.

FG.020 20.1 20.2 Nor aeqv, norqc_meix ziouc hnyouv_mv_nqaai aqv. as so_then sparrow so_then be_dissatisfied CHG-O-ST top top n seq.mk v.cmp part.asp 20.1So, 20.2the sparrow then became unhappy.

Page 222: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

196 FG.021 21.1 "Yie m'daaih gorngv mv bun biauv meih bueix. 1sg from_the_beginning speak NEG offer house 2sg sleep pn.p adv vt neg vt n pn.p vi 21.2 Meih ndongc_haaix gorngv meih nyuotv jienv zaux, 2sg no_matter_how speak 2sg bend DUR leg pn.p adv vt pn.p vi asp n 21.3 21.4 mv nzunc zorqv yie nyei beuh ndamc faaux jaang-ngunz mi'aqv." this time take 1sg POSS goiter tread_on go_up nape RSLT pn.dem n vt pn.p poss n vt vi n asp

“21.1From the beginning I told you I wouldn't give you my house to sleep in. 21.2No matter how you said you would keep your legs bent, 21.3this time (you) took my goiter 21.4(and) kicked it up to the nape.”

FG.022 22.1 22.2 22.3 22.4 Mv_nzunc_hnoi aeqv, caah_laauh_porngc youc gorngv laaix ndiangx-kou nauv, ndiangx-kou dorngc. this_time TOP grasshopper also speak because dead_tree snap dead_tree wrong adv.tmp top n seq.mk vt conj n vi n v.st 22.1This time 22.2the grasshopper also said 22.3(that is) because the dead tree snapped, 22.4the dead tree was wrong.

FG.023 23.1 23.2 Ndiangx-kou youc gorngv mv zeiz ninh ganh oix nauv. Dead_tree aslo speak NEG be_right 3sg self want_to snap n seq.mk vt neg vi pn.p pn.emph v.mod vi 23.1The dead tree also said 23.3(that) it was not that he himself wanted to snap.

FG.024 24.1 24.2 Nziouv-mbiauh_loc ganh ngaatc ninh, ninh nyei ndiangx-nzungh dangx nzengc mingh, termites self bite 3sg 3sg POSS root_of_tree break_off consumed go n pn.emph vt pn.p pn.p poss n vi vi v.dr 24.3 maiv haih yiem mv jienv. NEG be_able_to exist NEG stable neg aux vi adv v.st 24.1Termite himself bit him, 24.2his roots broke off altogether, 24.3(he) was not able to be stable.

FG.025 Nziouv-mbiauh_loc gorngv maiv zeiz ninh dorngc. termites speak NEG be_right 3sg wrong n vt neg vi pn.p v.st

The termite said (that) he was not wrong.

FG.026 Naang-nzung ganh zun ninh nyei biauv. earthworm self cement 3sg POSS house n pn.emph vt pn.p poss n

It was an earthworm that cemented his house.

Page 223: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

197 FG.027 27.1 27.2 27.3 Ninh mv maaih dorngx yiem cingx_daaih cuotv naaic daaih ngaatc ndiangx-nzungh dangx. 3sg NEG have place live therefore come_out there come bite tree_root break_off pn.p neg vt n vi conj vi pn.dem v.dr vt n vi 27.1He had no place to live 27.2therefore (he) came out from there, 27.3(and) bit the tree roots (and the roots) broke off.

FG.028 Wuov_nzunc_hnoi aeqv, ninh_mbuo baeng leiz daaih gorngv. this_time TOP 3_PL pull law come speak adv.tmp top pn.p vt n v.dr v

This time, they drew a law (and) discussed.

FG.029 29.1 29.2

Gorngv gau aeqv, naang-nzung suei mingh. speak DPCL TOP earthworm fail go v part top n v v.dr29.1When they discussed, 29.2the earthworm (was declared) guilty.

FG.030 30.1 30.2 Naang-nzung suei mingh zuqc Nyutc_Daaix_Hungh ndoh ninh nyei jaang ngaeng-ngaeng wuov. earthworm fail go PASS god bind 3sg POSS neck be_ringed that n vi v.dr part n vt pn.p poss n v.st np.dem 30.1The earthworm failed 30.2(he) got a Nyut-Daai god bound around his neck (and it) became round-ringed.

FG.031 31.1 31.2 Ih_zanc meih_mbuo maiv sienx mangc naang-nzung nyei jaang maah! now 2_PL NEG believe look earthworm POSS neck SFT.CMD adv.tmp pn.p neg vt vt n poss n part.f 31.1Now (if) you don’t believe, 31.2look at the neck of an earthworm.

FG.032 Za'gengh ndoh jienv nyei. really bind DUR PRS.ST adv vt asp asp

(It’s) really bound.

FG.033 33.1 33.2 Caah_laauh_porngc yaac dorngc diqv_dien, ninh_mbuo zorqv caah_laauh_porngc nyei zaux wuotv ngau mingh. grasshopper also wrong a_little 3_PL take grasshopper POSS leg bend bent go n top v.st adv pn.p vt n poss n vt v.st v.asp 33.1The grasshopper also was wrong a little, 33.2(so) they took the grasshopper’s legs (and) bent them.

FG.034 Yaac mangc caah_laauh_porngc nyei zaux maah! and look grasshopper POSS leg SFT.CMD conj vt n poss n part.f

And look at a grasshopper’s legs.

FG.035

Page 224: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

198 Za'gengh ngau-ngau wuov. really bent-bent that adv v.st-v.st pn.dem

(They are) really bent and bent.

Story 3 (3PG) “A Story of Three Piglets”

Page 225: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

199 3PG Buo Dauh Dungz-dorn three CLF piglet numb clf n

Three piglets 3PG.001 Ih_zanc yie gorngv taux hieh_juv caux dungz-dorn nyei gouv. now 1sg speak reach wolf be_with piglet POSS story adv.tmp pn.p v v n vt n poss n

Now I am going to tell a story of a wolf and piglets.

3PG.002 2.1 2.2 Loz-hnoi maaih dauh dungz-nyeiz njiec duqv buo dauh dorn. 0ld_days have CLF female_pig give_birth_to get three CLF cub n vt clf n vt vt numb clf n 2.1In the old days, there was a female pig 2.2(and she) gave birth to three piglets.

3PG.003 3.1 3.2 Dorh ninh nyei buo dauh dorn hlo deix ninh mbuox neiz dorn, raise 3sg POSS three CLF cub big some 3sg name her cub vt pn.p poss numb clf n v.st adv pn.p n contr n "Meih_mbuo aqv_zuqc mingh lorz dorngx gorqv-mienh yiem gorqv-mienh nyei." 2_PL must go look_for place each_person live each_person POSS pn.p aux vi vt n adj-n vi adj-n poss 3.1(She) brought up three cubs (to) big enough, 3.2she told them, “Each of you must go and find your own place to live.”

3PG.004 Deix dungz-dorn ziouc cuotv mingh lorz dorngx yiem. some piglet so_then go_out go look_for place live

vt clf n poss n vt vt n vt n

adv n seq.mk v v.dr vt n vi

So these piglets went out (and) looked for a place (to) live.

3PG.005 5.1 5.2 5.3 Maaih dauh dungz-dorn nyei za'eix lorz duqv mbiauh_nqaauv gomv biauv. have CLF piglet POSS plan look_for get stalk_of_rice build house

5.1One piglet’s method was 5.2(that he) looked for (and) got rice stalks 5.3(and) built a house.

3PG.006 6.1 6.2 Aengx maaih dauh dungz-dorn nyei za'eix lorz duqv biaav gomv biauv. further have CLF piglet POSS plan look_for get rod build house adv vt clf n poss n vt vt n vt n 6.1And the another piglet’s method was 6.2(that he) looked for rods (and) built a house.

3PG.007 7.1 7.2

Page 226: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

200 Aengx maaih dungz-dorn nyei za'eix longc hui-zun gomv biauv. and_then have piglet POSS plan use brick build house adv vt n poss n vt n vt n 7.1Another piglet's plan was 7.2(that he) built a house with bricks.

3PG.008 8.1 8.2 Gomv jienv mv nangc lauh, maaih dauh hieh_juv_jieqv daaih nduov ninh_mbuo nyanc. build DUR NEG very long have CLF wolf_black come deceive 3_PL eat vt asp neg adv v.st vt clf n-v.st vi vt pn.p vt 8.1(When they were) building not very long, 8.2there was a black wolf (and it) came (to) deceive them (to) eat (them).

3PG.009 9.1 9.2 Mingh taux da'yietv dauh dungz-dorn nyei biauv mbuox da'yietv dauh dungz-dorn, go reach first CLF piglet POSS house tell first CLF piglet vi v nmb clf n poss n vt nmb clf n "Meih aqv_zuqc koi gaengh bun yie bieqc. 2sg must open door let 1sg enter pn.p aux vt n vt pn.p vi Yie oix bieqc meih nyei biauv." 1sg like enter 2sg POSS house pn.p v.mod vi pn.p poss n 9.1(When he) arrived at the fist piglet's house, 9.2(the wolf) told him, “You must open the door to let me come in. I want to come into your house.”

3PG.010 10.1 10.2 Dungz-dorn yaac gamh_nziex, ninh maiv koi gaengh bun ninh bieqc. piglet TOP fear 3sg NEG open door let 3sg enter n part v.cmp pn.p neg vt n vt pn.p vi 10.1And the piglet was fearful, 10.2he did not open the door (to) let him come in.

3PG.011 Hnangv_haaix yaac maiv koi. no_matter_what TOP NEG open adv part neg vt

No matter what (he) would not open.

3PG.012 "Meih gorngv meih za'gengh maiv koi? 2sg speak 2sg really NEG open pn.p v pn.p adv neg vt Meih za'gengh maiv koi nor aeqv, 2sg really NEG open as TOP pn.p adv neg vt top top yie ziouc biomv meih biauv mbaang aex." 1sg so_then blow_out_from_mouth 2sg house collapse I_mean_it pn.p seq.mk vt pn.p n vi part.f

“Don’t you really open it? If you really won’t open it, I will really blow your house down.”

3PG.013

Page 227: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

201 13.1 13.2 Dungz-dorn yaac maiv_haih_fungc_zoux mbaang yaac maiv_haih_fungc_zoux. piglet TOP can't_do_anything collapse TOP can't_do_anything n part idm vi part idm

He was so afraid (that he) didn’t open.

13.1There was nothing the piglet could do about it, 13.2(if the house) collapsed (he) can’t do anything.

3PG.014 Ninh za'gengh gamh_nziex maiv koi. 3sg really fear NEG open pn.p adv v.cmp neg vt

3PG.015 15.1 Hieh_juv_jieqv za'gengh biomv aqv, wolf_black really blow PFT n-v.st adv vt part.asp 15.2 15.3 fuz^^ dangh, youc hluo ninh nyei biauv biomv mbaang mi'aqv. blow awhile also touch_softly 3sg POSS house blow collapse RSLT vi adv seq.mk vt pn.p poss n vt vi asp 15.1The black wolf indeed blew; 15.2(he) blew for a while and touched softly his house, 15.3(his house) was blown down (and) collapsed.

3PG.016 16.1 16.2 Ninh maiv haih biaux mingh haaix, hieh_juv_jieqv youc nyanc mi'aqv. 3sg NEG be_possible_to flee go where wolf_black also eat RSLT pn.p neg aux v vi q n_v.st seq.mk vt asp 16.1He could not flee to anywhere, 16.2so the black wolf ate it.

3PG.017 17.1 17.2 Zorqv da'yietv dauh nyanc mingh aengx mingh taux da'nyeic dauh dungz-dorn nyei biauv take first CLF eat go further go reach second CLF piglet POSS house vt nmb clf vt v.asp adv vi v n clf n poss n 17.3 aengx hnangv_nc_nor mbuox, further like_that_as tell adv adv vt "Meih koi gaengh bun yie bieqc." 2sg open door offer 1sg enter pn.p vt n vt pn.p vi 17.1(He) has taken the first piglet and ate it; 17.2and again (he) went to the second piglet's house 17.3and told him again like this, “You, open the door (to) let me in.”

3PG.018 18.1 Da'nyeic dauh dungz-dorn yaac gamh_nziex hieh_juv_jieqv bieqc mingh nyanc ninh, second CLF piglet TOP fear wolf_black enter go eat 3sg n clf n part v.cmp n-v.st vi v.dr pn.p

Page 228: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

202 18.2 ninh yaac maiv haih koi. 3sg TOP NEG be_possible_to open pn.p part neg aux vt 18.1The second piglet also was fearful (that) the black wolf (would) enter (to) eat him, 18.2so he would not open.

3PG.019 Ninh aengx mbuox dauh dungz-dorn, 3sg further name CLF piglet pn.p adv n clf n "Meih mv koi yie ziouc biomv meih nyei biauv mbaang aex." 2sg NEG open 1sg so_then blow 2sg POSS house collapse I_mean_it pn.p neg vt pn.p seq.mk vt pn.p poss n vi part.f

He told the piglet again, “If you won't open the door, then I will really blow down your house”

3PG.020 Ninh yaac maiv_haih_fungc_zoux. 3sg TOP can't_do_anything pn.p part idm

There was nothing he could do about it either.

3PG.021 Mbaang yaac maiv_haih_fungc_zoux. collapse TOP can't_do_anything vi part idm

(If the house) collapses, (he) would not be able to do anything.

3PG.022 Gamh_nziex maiv koi gaengh bun hieh_juv_jieqv bieqc. fear NEG open door offer wolf_black enter v.cmp neg vt n vt n-v.st vi

(He) was fearful (so he) would not open the door (to) let the black wolf come in.

3PG.023 23.1 Hieh_juv_jieqv ziouc za'gengh biomv aqv, wolf_black so_then really blow CHG-O-ST n_v.st seq.mk adv vt part.asp 23.2 23.3 fuz^^^ dangh, aengx hluo ninh nyei biauv biomv mbaang mi'aqv. blow awhile further touch_softly 3sg POSS house blow collapse RSLT vi adv adv vt pn.p poss n vt vi asp 23.1Then the black wolf really blew; 23.2(he) blew a short while, 23.3further (he) touched his house (and it) collapsed.

3PG.024 24.1 24.2 Ninh maiv lamh bingx, hieh_juv_jieqv youc aengx nyanc mi'aqv. 3sg lack_a_place_for hide wolf_black also further eat RSLT pn.p idm n n-v.st seq.mk adv vt asp 24.1He (the second piglet) had no place to hide, 24.2then the wolf ate (him) also.

Page 229: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

203 3PG.025 25.1 25.2 Hieh_juv_jieqv nyanc i dauh mi'aqv, haiz za'gengh kuv nyei. wolf_black eat two CLF RSLT feel really be_delicious PRS.ST n-v.st vt numb clf asp vi adv v.st asp

3sg speak

25.1The black wolf ate two piglets, 25.2(he) felt (they) were really delicious.

3PG.026 26.1 Aengx mingh taux da'faam dauh dungz-dorn nyei biauv, further go reach third CLF piglet POSS house adv vi v adj clf n poss n 26.2 heuc da'faam dauh dungz-dorn koi gaengh bun ninh bieqc. order third CLF piglet open door let 3sg enter v adj clf n vt n vt pn.p vi 26.1(The wolf) further went to the third piglet's house, 26.2(and) called the third piglet (to) open the door (to) let him come in.

3PG.027 Da'faam dauh dungz-dorn yaac gamh_nziex maiv koi. third CLF piglet TOP fear NEG open adj clf n part v.cmp neg vt

The third piglet was also fearfull (and he) did not open.

3PG.028 Ninh gorngv,

pn.p v "Meih za'gengh maiv koi? 2sg really NEG open pn.p adv neg vt Meih za'gengh maiv koi yie biomv meih nyei biauv mbaang aex!" 2sg really NEG open 1sg blow 2sg POSS house collapse I_mean_it pn.p adv neg vt pn.p vt pn.p poss n vi part.f

He said, “Do you really not open? If you really don’t open, I will really blow your house down.”

3PG.029 Dungz-dorn yiem wuov biauv jatv ninh, piglet be_there that_place house laugh_at 3sg n vi pn.dem n vt pn.p "Yie nyei biauv ndongc naaiv longx hnangv_haaix_nor biomv duqv mbaang?" 1sg POSS house as_much_as this good how blow can collapse pn.p poss n adv pn.dem v.st adv vt v.aux vi

There in the house the piglet laughed at him, saying “My house is so good that you can't blow it to collapse.”

3PG.030 30.1 30.2 Ninh qiex_jiez haic, ninh za'gengh biomv aqv. 3sg get_angry very 3sg really blow CHG-O-ST pn.p v adv pn.p adv vt part.asp 30.1He got angry, 30.2(so) he really blew away (at the house).

Page 230: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

204 3PG.031 Fuz^^, fuz nyei biomv aqv. blow blow ADVBLZR blow PFT vi vi part vt part.asp

(He) blew and blew.

3PG.032 32.1 32.2 Hnangv_haaix_nor biomv yaac maiv noic duqv mbaang. how blow TOP NEG be_possible can collapse adv vt part neg vi v.aux vi 32.1No matter how he blew, 32.2(the house) could not collapse.

3PG.033 Dauh dungz-dorn njiec qaqv jatv ninh. CLF piglet let_go_down strength laugh_at 3sg clf n vt n vt pn.p

That piglet laughed at him very hard.

3PG.034 34.1 34.2 Ninh aengx daav za'eix nduov dungz-dorn, yuoqc dungz-dorn, 3sg further plan plan deceive piglet entice piglet pn.p adv vt n vt n vt n "Njang_hnoi mbuo mingh wetv ndoih." tomorrow 1pl go dig potato adv.tmp pn.p vi vt n 34.1He again made a plan (and) deceiveed the piglet, 34.2(he) enticed the piglet, “Let's go to dig potatoes tomorrow.”

3PG.035 35.1 Dungz-dorn yaac laengz caux ninh mingh wetv ndoih piglet TOP promise be_with 3sg go dig potato n part v vt pn.p vi vt n 35.2 mv_baac da'nyeic ndorm dungz-dorn jiez sin nziouv nyei daaih but second morning piglet raise body early ADVBLZR come conj n n n vt n v.st part v.asp 35.3 35.4 mingh ndaangc wetv jienv ndoih nzuonx daaih aqv. go first dig DUR potato return come PFT vi n vt asp n v v.dr part.asp 35.1Then the piglet promised (that) he would go (to) dig potatoes with him, 35.2but the following morning the piglet got up early 35.3went first (and) dug potatoes 35.4(then) came back.

3PG.036 Hieh_juv_jieqv taux mbuox ninh, Wolf_black reach tell 3sg n-v.st v vt pn.p "Mingh aav_lov! Mingh wetv ndoih aav_lov." go INVITATION go dig potato INVITATION vi intj vi vt n intj

The black wolf came (and) told him, “Let’s go! Let’s go dig potatoes!”

Page 231: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

205 3PG.037 Ninh gorngv mbuox hieh_juv_jieqv, 3sg speak tell wolf_black pn.p v vt n-v.st "Yie wetv nzuonx daaih aqv. 1sg dig return come PFT pn.p vt v v.dr part.asp Meih oix mingh meih ganh mingh aqv." 2sg want_to go 2sg self go CAS.CMD

40.2

v.mod vi pn.p pn.reflexive vi part.mod

He said to the black wolf, “I’ve been back from digging. If you want to go, go by yourself.”

3PG.038 38.1 38.2 38.3 Nduov hieh_juv_jieqv youc qiex_jiez haic, mv duqv nc dauh dungz-dorn nyanc aqv. deceive wolf_black also get_angry very NEG get that CLF piglet eat CHG-O-ST vt n-v.st seq.mk v.cmp adv neg vt pn.dem clf n vt part.asp 38.1(He) deceived the black wolf, 38.2(the black wolf) got very angry, 38.3(he) could not get the piglet (to) eat (him).

3PG.039 Ninh aengx daav za'eix nduov dungz-dorn, 3sg again plan plan deceive piglet pn.p adv vt n vt n "Njang_hnoi mbuo mingh gaeqv biouv, gaeqv apc^ben biouv." tomorrow 1pl go pick fruit pick apple fruit adv.tmp pn.p vi vt n vt n n

He again made a plan (to) deceive the piglet, “Let's go and collect some fruits, tomorrow; collect apples.”

3PG.040 40.1 Da'nyeic ndorm dungz-dorn aengx jiez sin nziouv nyei, second morning piglet again raise body early AFM n n nn adv vt n v.st asp

ninh youc aengx mingh ndaangc aqv. 3sg then further go first PFT pn.p seq.mk adv vi n part.asp 40.1The following morning the piglet got up early again, 40.2then he went first again.

3PG.041 41.1 41.2 Mingh taux biouv-ndiangx faaux biouv-ndiangx mingh. go reach fruit-tree go_up fruit-tree go vi v n-n vi n-n v.dr 41.1(When he) reached a fruit tree, 41.2(he) climbed up the tree.

3PG.042 42.1 42.3 Jang-jang faaux taux gu'nguaaic hnangv, hieh_juv_jieqv taux aqv. just_now go_up reach above only wolf_black reach PFT adv vi v adv.loc adv n_v.st v part.asp 42.1As soon as (he) went up above, 42.3the black wolf just arrived.

Page 232: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

206 3PG.043 Ninh gamh_nziex hieh_juv_jieqv nyanc ninh aqv. 3sg fear wolf_black eat 3sg INCHOAT pn.p v.cmp n-v.st vt pn.p part.asp

He was afraid (that) the wolf was about to eat him.

3PG.044 44.1 44.2 44.3 Dungz-dorn aengx daav cuotv za'eix gaeqv apc^ben biouv guaengx cuotv wuov_huaav. piglet again plan go_out plan pick apple fruit throw_by_whirling go_out yonder n adv vt v n vt n n vt vi adv.loc 44.1Again the piglet made a plan 44.2(he) picked an apple 44.3(and) threw (it) away over there.

3PG.045 45.1 Ndortv wuov_huaav go deix nyei dorngx fall yonder far some RELT place vi adv.loc v.st adv relt n 45.2 nduov hieh_juv_jieqv mingh zorqv wuov norm biouv. deceive wolf_black go take that CLF fruit vt n-v.st vi vt pn.dem clf n 45.1(It) fell there in quite a far place, 45.3(he) deceived the black wolf (and the wolf) went (to) pick the fruit.

3PG.046 46.1 46.2 46.3 Ninh youc bangc wuov norm zanc nqongh deix njiec daaih biaux nzuonx mingh. 3sg also depend_on that CLF period hurry some descend come flee return go pn.p seq.mk vt pn.dem clf n vi adv vi v.dr vi v v.dr 46.1He then used that time 46.2came down quickly 46.3fled (and) went home.

3PG.047 47.1 47.2 Hieh_juv_jieqv youc maiv noic duqv nyanc aengx gaanv jienv nzuonx. Wolf_black also NEG manage can eat further hurry DUR return n-v.st seq.mk neg vi v.aux vt adv vi asp v 47.1So the black wolf could not manage (to) eat (him) 47.2again (he) hurried to home.

3PG.048 Ninh aengx daav za'eix yuoqc dungz-dorn, 3sg again plan plan entice piglet pn.p adv vt n vt n "Njang_hnoi mbuo mingh mangc ngaanv." tomorrow 1pl go look fair adv.tmp pn vi v n

He again planned to entice the piglet, (saying) “Let's go and see the fair tomorrow.”

Page 233: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

207 3PG.049 49.1 Da'nyeic ndorm dungz-dorn aengx jiez sin nziouv nyei mingh ndaangc, second morning piglet again raise body early ADVBLZR go first n n n adv vt n v.st part vi n 49.2 49.3 mingh taux ngaanv wuov maaiz jienv. go reach fair that buy DUR vi v n pn.dem vt asp 49.1The following morning, the piglet got up early again (and) went first, 49.2(he) went (and) reached there 49.3(and) was buying (things).

3PG.050 50.1 50.2 Maaih norm, maaih nyungc ndongh, have CLF have kind container vt clf vt n n 50.3 50.4 ninh_mbuo zoux nyorx-hmei nyei ndongh, zoux nyorx-hmei nyei tongv hlo nyei. 3_PL do milk_oil POSS container do milk_oil POSS bucket big PRS.ST pn.p vt n-n poss n vt n-n poss n v.st asp 50.1There was…50.2there was a kind of container; 50.3they make a can for powdered milk, 50.4(they) make a big powdered milk’s bucket.

3PG.051 Maaiz jienv nc norm tongv nzuonx aqv. buy DUR that CLF bucket return PFT vt asp pn.dem clf n v part.asp

(He) bought the bucket (and) returned.

3PG.052 52.1 52.2 Nzuonx taux ndaamv-jauv maaih norm dorngx faaux deix jiex nor, return reach half-way have CLF place go_up some pass as v v n vt clf n vi adv vi top 52.3 ninh taux wuov_ndiev jiex-juoqv. 3sg reach down_yonder foot_of_hill pn.p v adv.loc n 52.1(He) returned half way, 52.2as there was a place where the slope went up a bit, 52.3he reached the foot of the hill.

3PG.053 53.1 53.2 Ninh youc maiv_haih_fungc_zoux, mangc jienv hieh_juv nyanc ninh aqv. 3sg also can't_do_anything look DUR wolf eat 3sg INCHOAT pn.p seq.mk idm vt asp n vt pn.p part.asp 53.1Then there was nothing he could do about it but 53.2(he) could see the wolf was ready to eat him.

3PG.054 Ninh ziouc nyorng bieqc nc norm nyorx-hmei tongv gu'nyuoz. 3sg then creep enter that CLF Milk_oil bucket inside pn.p seq.mk vi vi pn.dem clf n-n n n

So then he crept into the powdered milk container.

Page 234: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

208 3PG.055 55.1 55.2 55.3 Nyorx-hmei tongv njangx, njangx jienv hungx jienv hieh_juv_jieqv njiec. milk_oil bucket roll roll DUR go_on_without_seeing SML.ACT wolf_black go_down n-n n vi vi asp vi asp n-v.st vi 55.1The powder milk container rolled, 55.2rolling 55.3(and) went down directed straight toward the black wolf.

3PG.056 56.1 56.2 56.3 Haeqv nc dauh hieh_juv_jieqv gamh_nziex haic youc biaux mingh. frighten that CLF wolf_black fear very then flee go vt pn.dem clf n_v.st v.cmp adv seq.mk vi vi 56.1(It) frightened the black wolf 56.2(the wolf) was very fearful 56.3then ran away.

3PG.057 57.1 Hieh_juv_jieqv biaux mingh wolf_black flee go n_v.st vi vi 57.2 57.3 ninh youc aengx bangc mc norm saaix aengx nzuonx. 3sg also again depend_on that CLF competition again return pn.p seq.mk adv vt pn.dem n adv vi 57.1The black wolf ran away, 57.2then he [piglet] depend on that competition again 57.3(and) went home again.

3PG.058 Nzuonx taux biauv daaih. return reach house come v v n v.dr

(He) came back home.

3PG.059 Hieh_juv_jieqv aengx gan jienv ninh nzuonx aqv. wolf_black further follow SML.ACT 3sg return PFT n_v.st adv vt asp pn.p v part. asp

And then the black wolf returned following the piglet.

3PG.060 60.1 60.2 Nzuonx taux biauv daaih mbuox dungz-dorn. return reach house come tell piglet v v n v.dr vt n 60.1(When) he came home, 60.2(he) told the piglet.

Page 235: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

209 3PG.061 "A'ndorm yie mingh taux ngaanv yaac mv buatc meih. this_morning 1sg go reach fair and_yet NEG see 2sg adv.tmp pn.p vi v n conj neg vt pn.p Mingh taux ndaamv-jauv maaih nyungc ga'naaiv haeqv yie kouv haic aqv. go reach half-way have kind thing frighten 1sg tired very PFT vi v n vt n n vt pn.p v.st adv part.asp Mv_maaih_doic kouv haic aqv. Haeqv yie oix daic aqv." fear tired very PFT frighten 1sg want_to die INCHOAT v.cmp v.st adv part.asp vt pn.p v.mod vi part.asp

“This morning I went to the fair but I didn’t see you. As I went half way there was a thing that scared me very much and made me tired. I was very fearfull. (That thing) scared me so much that I was almost dead.”

3PG.062 Dungz-dorn jatv, piglet laugh n vi "Naaic se dongh yie. Yie zoux haeqv meih." that is same 1sg 1sg make frighten 2sg pn.dem copl idntf pn.p pn.p vt vt pn.p

The piglet laughed, “That was me. I made you frightened.”

3PG.063 Nduov ninh qiex_jiez haic. deceive 3sg get_angry very vt pn.p v.cmp adv

(He) deceived him (to) make (him) angry very much.

3PG.064 64.1 64.2 Mc dauh dungz-dorn za'eix ndongc naaiv longx maiv noic duqv nyanc. that CLF piglet plan as_much_as this good NEG manage can eat pn.dem clf n n adv pn.dem v.st neg vi v.aux vt 64.1The piglet’s method was so well planned that 64.2(the wolf) could not manage to eat (him).

3PG.065 65.1 Dungz-dorn biauv-ngorh tong norm kuotv piglet roof_of_a_house break_open CLF hole n n v clf n 65.2 ninh_mbuo liouh douz-sioux cuotv nyei kuotv. 3_PL keep_aside smoke go_out RELT hole pn.p vt n vi relt n 65.1The roof of the piglet's house has a hole; 65.2a hole that they [i.e. Westerner] make to let smoke go out [i.e. chimney].

Page 236: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

210 3PG.066 66.1 66.2 Nc dauh hieh_juv_jieqv ziouc aengx daav za'eix nyorng faaux mc norm kuotv, that CLF wolf_black so_then further plan plan creep go_up that CLF hole pn.dem clf n_v.st seq.mk adv vt n vi vi pn.dem clf n 66.3 66.4 gan nc norm kuotv bieqc nyanc nc dauh dungz-dorn aqv. follow that CLF hole enter eat that CLF piglet INCHOAT vt pn.dem clf n vi vt pn.dem clf n part.asp 66.1So the black wolf planned again 66.2(and he) crept up the chimney hole, 66.3along the hole went in 66.4(to) eat the piglet.

3PG.067 67.1 Dungz-dorn buatc hieh_juv_jieqv nyorng jienv faaux, piglet see wolf_black creep SML.ACT go_up n v n_v.st vi asp vi 67.2 67.3 dungz-dorn aengx daav cuotv za'eix zorqv domh tongv zaangh wuom daaih piglet again plan go_out plan take big bucket dip water come n adv vt v n vt adj n vt n v.dr 67.4 67.5 hnyutv jienv wuov douz-nzauc nduov hieh_juv_jieqv faaux boil DUR that fire_place deceive wolf_black go_up vt asp pn.dem n vt n_v.st vi 67.5 67.6 faaux taux wuov norm kuotv za'gengh bieqc aqv. go_up reach that CLF hole really enter PFT vi v pn.dem clf n adv vi part.asp 67.1The piglet saw the black wolf was creeping up, 67.2the piglet made a plan again, 67.3(he) took a big bucket (and) filled water (in it), 67.4boiled on the fire place 67.5tricked the black wolf go up, 67.5(he) went up (to the top of) the chimney 67.6(and) indeed went in.

3PG.068 Gan nc norm douz-sioux-kuotv bieqc. follow that CLF smoke_hole enter vt pn.dem clf n-n vi

(He) followed the chimney hole (and) entered.

3PG.069 69.1 69.2 Bieqc daaih yietv bieqc taux gu'nyuoz maengx ziqc biauv aeqv, enter come as_soon_as enter reach inside part the_side house TOP vi vi adv vi v n n adv.loc n part 69.3 69.4 69.5 69.6 douz-sioux congx zuqc ninh m'zing sui, mongh_longh, mangc mv buatc haaix ndau. smoke shoot GOAL 3sg eye sower dizzy look NEG see what place n vt part pn.p n v.st v.st vt neg v q n 69.1(wolf) came in, 69.2as soon as (he) came in to the inner part of the house, 69.3a smoke shot at his eyes 69.4(eyes) got sore, 69.5got dizzy, 69.6could not see anywhere.

Page 237: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

211 3PG.070 70.1 70.2 Ndortv njiec, cingx_daaih ndortv bieqc mc norm wuom-mbueix tongv mingh, fall descend therefore fall enter that CLF water_boil bucket go vi vi conj vi vi pn.dem clf n-vi n v.dr 70.3 70.4 cingx_daaih wuom hluqv daic mingh. therefore water scald die go conj n vt vi v.asp 70.1(He) fell down, 70.2therefore (he) fell in to the boiling water in the bucket, 70.3therefore (he) the water scalded (him) 70.4(he) died.

3PG.071 Yiem naaic be_there there vi pn.dem mc dauh dungz-dorn nyei za'eix za'gengh zeix zuqc nc dauh hieh_juv daic mi'aqv. that CLF piglet POSS plan really torture GOAL that CLF wolf die RSLT pn.dem clf n poss n adv vt part pn.dem clf n vi asp

There, that the piglet's plan really tortured that wolf (and he) died.

Page 238: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

212

Story 4 (CI): “Why Cicadas don’t Have Intestines?” Cicada Gaeng-waen mv maaih jaangh nyei gouv cicada NEG have intestine RELT story n neg vt n relt n

A story of Why cicadas don't have intestines

CI.001 Yie gorngv diuh gouv bun meih muangx oc. 1sg speak CLF story let 2sg listen PLT pn.p v clf n vt pn.p vt part

I am going to tell a story (and) let you listen (to it).

CI.002 Oix muangx nyei? want_to listen q v.mod v part

Do you want to listen?

CI.003 Gengh oix muangx? really want_to listen adv v.mod vt

Do you really want?

CI.004 Oix muangx haaix_nyungc gouv? want_to listen what_kind story v.mod vt q n

Which story do you want to listen to?

CI.005 5.1 5.2 Gaeng-waen ga'sie mv maaih jaangh nyei gouv meih haiz jiex nyei? cicada stomach NEG have intestine RELT story 2sg hear experienced q n n neg vt n relt n pn.p vt aux part 5.1The story that there is no intestine in a cicada's stomach, 5.3have you ever heard?

CI.006 Mv haiz jiex? NEG hear experienced neg vt aux

You haven't, have you?

CI.007 7.1 7.2 Mv haiz jiex yie gorngv bun meih muangx. NEG hear experienced 1sg speak let 2sg listen neg vt aux pn.p v vt pn.p vt 7.1(If) you haven't, 7.2I will tell you.

Page 239: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

213 CI.008 8.1 Loz-hnoi norqc_guv_long gorngv mbuox jung, old-day owl speak tell barking_deer adv.tmp n v vt n "Maqv_gorkv_biouv zuoqc aqv, Maqv_gorkv_biouv zuoqc aqv", olive_fruit ripe PFT olive_fruit ripe PFT n v.st part.asp n v.st part.asp 8.2 nduov jung daaih nyanc. deceive barking_deer come eat vt n vi vt 8.1In the old days, an owl told a barking-deer; “The olives are ripe, The olives are ripe,” 8.2deceived the barking-deer (to) eat (some).

CI.009 9.1 9.2 Jung daaih taux wuov nyanc ndau_wuov ma'gorkv_biouv. barking_deer come reach there eat ...uh... olive_fruit n vi v pn.dem vt int n 9.1The barking-deer came there 9.2(and) ate…uh… those olive fruits.

CI.010 Mbopv-yangh youc gorngv mbuox jung, squirrel-yellow also speak tell barking_deer n-v.st seq.mk v vt n "Aqv_zuqc zorqv meih, aqv_zuqc zorqv meih." must take 2sg must take 2sg aux vt pn.p aux vt pn.p

A yellow-squirrel also told the barking-deer, “You must be caught, must be caught.”

CI.011 Gaeng-waen aengx gorngv. cicada further speak n adv v

A cicada also spoke up.

CI.012 12.1 Gaeng-waen la'guaih heuc hnangv lorqc, cicada just call only you_see n adv v adv part.f 12.2 mv_baac jung ganh hnamv ninh gamh_nziex mv_bei. but barking_deer self think 3sg fear I_guess conj n pn.emph vt pn.p v.cmp part.f 12.1You see, the cidada just cried without any intention, 12.2but the barking deer himself thought (that) he was fearful, I guess.

Page 240: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

214 CI.013 13.1 13.2 Jung gorngv haiz gaeng-waen gorngv, barking_deer speak hear cicada speak n v vt n v "Oix zorqv, oix zorqv, oix zorqv," want_to take want_to take want_to take v.mod vt aux vt aux vt 13.3 haeqv jung tiux mingh frighten barking_deer run go vt n vi vi 13.4 13.5 caaiv zuqc sopc luangh dangx norh aah. step_on GOAL ash_pumpkin vine break_off REPORT listen_Please!! vt part n n vi part part.f 13.1The barking deer said 13.2(that he) heard the cicada saying, “I will catch you, I will catch you, I will catch you,” 13.3(the words) frightened the barking deer (and let him) run away, 13.3(he) stepped on an ash-pumpkin, 13.5(and) its vine broke off.

CI.014 14.1 14.2 Sopc luangh dangx mingh youc aengx sopc njangx. ash_pumpkin vine break_off go also further ash_pumpkin roll n n v v.asp seq.mk adv n vi 14.1A vine of the ash pumpkin broke off 14.2and further the ash pumpkin rolled down.

CI.015

16.2

15.1 Sopc njangx jienv njiec, ash_pumpkin roll SML.ACT go_down n vi asp v.dr 15.2 youc aengx mborqv zuqc saqv-ndongh kung. also further hit GOAL sesame-container spill_out seq.mk adv vt part n vi 15.1The ash pumpkin was rolling down and down, 15.2and further (it) hit a sesame container (and it) spilled out.

CI.016 16.1 Mborqv zuqc saqv-ndongh kung, hit GOAL sesame-container spill_out vt part n v

youc aengx kung bieqc norqc_jai m'zing mingh. also further spill_out enter wild_chicken eye go seq.mk adv vi vi n n v.asp 16.1(It) hit the sesame container (and it) spilled out, 16.2and further (the semame) spilled (and) went into a wild chicken's eyes.

Page 241: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

215 CI.017 17.1 17.2 Wuov_nzunc_hnoi aeqv, norqc_jai la'guaih mbuonc, this_time TOP wild_chicken carelessly claw_ground adv.tmp part n adv vt 17.3 17.4 mbuonc zuqc nziouv cuotv daaih. claw_ground GOAL ant go_out come vt part n vi v.dr 17.1This time, 17.2a wild chicken scratched (the ground) carelessly 17.3(and) scratched (and) hit an ant accidentally 17.4(the ant) came out.

CI.018 Oh, maiv zeiz la'guaih nziouv hnangv, nziouv-muotc. Oh NEG be_right just ant only black_ant intj neg vi adv n adv n

Oh, he was not just an ordinary ant (but) a black ant.

CI.019 19.1 19.2 19.3 Norqc_jai la'guaih mbuonc zuqc nziouv-muotc cuotv daaih danx zuqc naang. wild_chicken carelessly claw_ground GOAL black_ant come_out come sting GOAL snake n adv vt part n vi vi vt part n 19.1The wild chicken carelessly clawed the ground (and) hit the black ant, 19.2(the black ant) came out, 19.3(they) stung a snake.

CI.020 Naang youc la'guaih! la'guaih mingh. snake also aimlessly aimlessly go n seq.mk adv adv vi

Then the snake went away aimlessly.

CI.021 21.1 21.2 Mingh gau aeqv, nzun bieqc mbopv nyei lauz. go DPCL TOP crawl_under enter squirrel POSS nest vi part part vi vi n poss n 21.1When (it) went, 21.2(it) crawled into a squirrel's nest.

CI.022 22.1 22.2 Mingh nzun bieqc mbopv nyei lauz mingh, go crawl_under enter squirrel POSS nest go vi vi vi n poss n vi 22.3 mbopv youc maiv_lamh yiem. squirrel also lack_a_place_for live n seq.mk idm vi 22.1(The snake) went 22.2(and) crawled into the squirrel's nest, 22.3the squirrel also did not have place to stay.

CI.023 23.1 23.2 Mbopv maiv lamh yiem aeqv cingx_daaih la'guaih cuotv mingh. squirrel lack_a_place_for live TOP therefore aimlessly go_out go n v vi part conj adv vi v.dr 23.1Because the squirrel did not have place to live, 23.2therefore (he) just came out aimlessly.

Page 242: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

216 CI.024 Ngaatc biouv-youh. bite oil_fruit vt n

He (squirrel) bit a kind of cucumber.

CI.025 Biouv-youh youc yietv_tin_yietv_deic mv ndortv. oil_fruit also randomly NEG fall n seq.mk adv neg vi

The oil fruit also did not fall without an specific goal.

CI.026 26.1 26.2 Mangc ziangx wuov dauh ong nyei in-dang njiec mingh zong. look finish that CLF old_man POSS opium_smoking_lamp go_down go bump v vi pn.dem clf n poss n vi vi vt 26.1(The oil fruit) aimed straight at that old man’s opium lamp 26.2(and) went down (and) bumped (it).

CI.027 27.1 27.2 27.3 Dauh mbopv aeqv, qiex_jiez haic, nqemh naang nzaeng neiz dorngx yiem. CLF squirrel TOP get_angry very accuse snake quarel his place live clf n part v.cmp adv vt n vt contr n vi 27.1As for that squirrel, 27.2he got very angry 27.3(and) criticized (that) the snake took his place to live.

CI.028 28.1 28.2 Ninh mv maaih dorngx yiem aeqv, za'gengh ngaatc nc norm biouv-youh aeqv, 3sg NEG have place live TOP really bite that CLF oil_fruit TOP pn.p neg vt n vi part adv vt pn.dem clf n part 28.3

vi vi part vi adv vi pn.dem vi intj

hnaeng! Hnaeng nyei mbapc dangh njiec wuov mingh aeqh. dangle dangle ADVBLZR thud quickly descend there go SUSPND

28.1As he lost his place to live, 28.2because (he) really bit the oil fruit, 28.3(it) dangled and dropped down there with a thud…and…

CI.029 29.1 29.2 Za'gengh hluo wuov dauh ong nyei in-dang zong kung nzengc mingh ae. Really touch that CLF old_man POSS opium_lamp hit spill_out consumed go SUSPND adv vt pn.dem clf n poss n vt vi vi vi intj 29.1(It) indeed hit that old man's opium smoking lamp 29.2(it) completely spilled out...and…

Page 243: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

217 CI.030 Nc dauh ong qiex_jiez haic, that CLF old_man angry very pn.dem clf n v.cmp adv "Yie yiem naaiv lomc-ndiev buov in 1sg be_here this forest-area smoke opium pn.p vi pn.dem n vt n meih la'guaih hnangv_nc_nor daaih hoic yie nyei in-dang 2sg carelessly like_that_as come harm 1sg POSS opium_lamp pn.p adv adv vi vt pn.p poss n zoux haaix nyungc?" do what kind vt q n

That old man got very angry, “I was smoking opium in this forest and you just came randomly (and) destroyed my opium lanp, why?”

CI.031

Biouv-youh youc gorngv,

Naaic biouv-youh zan_naaic ninh. that oil_fruit interrogate 3sg pn.dem n vt pn.p

That cucumber asked him back.

CI.032

oil_fruit also speak n seq.mk v "Yie maiv zeiz, yie ganh oix ndortv lorqc. 1sg NEG be 1sg self want_to fall you-see pn.p neg vi pn.p pn.emph v.mod vi part.f Yie mv dongh maaih mienh hoic yie ndortv?" 1sg NEG same have person harm 1sg fall pn.p neg adj vt n vt pn.p vi

The cucumber said, “Not me! Not that I myself wanted to fall. How can I fall without someone damaging me?”

CI.033 Dauh ong gorngv naaic gaax ninh, CLF old_man speak ask try_and_see 3sg clf n v v part pn "Haaix nyungc hoic meih laeh? what kind harm 2sg will_you_tell_me? q n vt pn.p q.part Maaih haaix nyungc hoic meih ndortv laeh?" have what kind harm 2sg fall will_you_tell_me? vt q n vt pn.p vi part.q

That old man asked saying, “What damaged you? Can you tell me what damaged you to fall?”

Page 244: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

218 CI.034 Ninh gorngv, 3sg speak pn v "O mbopv ganh ngaatc yie, yie fungc ziangh duqv jienv? Oh squirrel self bite 1sg 1sg how_come alive can stable intj n pn.emph vt pn.p pn.p q vi v.aux v.st

The cucumber said, “Oh, a squirrel himself bit me, how can I be connected to the stem tightly?

Mbopv ngaatc yie, yie ziangh mv jienv cingx_daaih ndortv niaa. squirrel bite 1sg 1sg alive NEG stable therefore fall I’m_thinking n vt pn.p pn.p vi neg v.st conj vi part.f

pn.p vt pn.p vt vi pn.p poss n

Yie tov meih siex cuotv yie nyei zuiz." 1sg ask_for 2sg rescue go_out 1sg POSS sin

The squirrel bit me, I couldn't get connected tightly, that's why I fell, don't you know? Please forgive my mistake.”

CI.035 35.1 35.2 35.3 Nc dauh ong gorngv ninh longc mbopv, mbopv ganh ngaatc. that CLF old_man speak 3sg use squirrel squirrel self bite pn.dem clf n v pn.p vt n n pn.emph vt 35.1That old man said 35.2(that) he would charge the squirrel; 35.3the squirrel himself must have bitten.

CI.036 "Hnangv meih maih_zingh_mouc_ziev la'guaih ngaatc biouv-youh like 2sg without_thinking carelessly bite oil_fruit adv pn.p idm adv vt n ndortv mborqv yie nyei in-dang zoux_haaix_nyungc?" fall hit 1sg POSS opium_smoking_lamp why vi vt pn.p poss n q

“You just carelessly bite the old fruit without thinking anything, letting it hit my opium lamp, why?”

Page 245: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

219 CI.037 Mbopv aengx gorngv mbuox ninh, squirrel again speak tell 3sg n adv v vt pn.p "Naang ganh daaih bieqc yie nyei biauv, snake self come enter 1sg POSS house n pn.emph vi vi pn.p poss n yie mv_maaih_doic naang yie cuotv daaih qiex_jiez haic. 1sg scared_of snake 1sg come_out come angry very pn.p v.cmp n pn.p vi vi v.cmp adv La'guaih! la'guaih ngaatc cingx_daaih ngaatc zuqc ninh ndortv. randomly randomly bite therefore bite GOAL 3sg fall adv adv vt conj vt part pn vi Naaiv yie funx yie maiv ziqc_zuiz meih. this 1sg count 1sg NEG offend 2sg pn.dem pn.p vt pn.p neg v.cmp pn.p Yie mv dorngc meih. 1sg NEG be_wrong 2sg pn.p neg v.st pn.p Tov meih siex cuotv yie nyei zuiz." ask_for 2sg rescue come_out 1sg POSS sin vt pn.p vt vi pn.p poss n

The squirrel also said to him, “The snake himself came into my house. I was scared of the snake so I came out, and was very angry. I just aimlessly bit, that’s why I accidentally bit it and it fell. This, I don’t consider that I offended you. I didn’t do wrong to you. Please, overlook my wrong.”

CI.038 38.1 38.2 Nduov dauh ong aengx mingh naaic naang. deceive CLF old_man further go ask snake vt clf n adv vi vt n 38.1(The squirrel) tricked the old man 38.2(the old man) further went (to) ask the snake.

CI.039 Naaic gaax naang, ask try_and_see snake v part n "Hnangv meih maaih_zingh_mouc_ziev, meih nyei dorngx yiem mv yiem, like 2sg without_thinking 2sg POSS place exist NEG exist adv pn.p vt pn.p poss n vi neg vi nzuangv mv nzuangv, la'guaih daaih bieqc mbopv nyei biauv restlessly intensely restlessly carelessly come enter squirrel POSS house adv intj adv adv vi vi n poss n zoux_haaix_nyungc?"

why q

(The old man) asked the snake, “Seems like you are not thinking anything. You should have stayed in you place but came out restlessly and just carelessly went into the squirrel’s house. Why?”

Page 246: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

220 CI.040 Gorngv, speak v "O nziouv-muotc ganh danx yie. Oh black_ant self sting 1sg intj n pn.emph vt pn.p Yie la'guaih biaux gau haiz naaiv kuv yiem nyei 1sg just flee DPCL hear here comfortable_to be_in PRS.ST pn.p adv vi part vt pn.dem adv vi asp yie ziang_naaic biaux bieqc naaiv daaih. 1sg so flee enter here come pn.p conj vi vi pn.dem vi Naaiv se maiv zeiz yie nyei zuiz, nziouv-muotc nyei zuiz. this is NEG be 1sg POSS fault black_ant poss fault pn.dem copl neg vi pn.p poss n n poss n Meih mingh baatc nziouv-muotc oc." 2sg go punish black_ant SFT.CMD pn.p vi vt n part

(The snake) said, “Oh, it’s the black ant that stung me. When I just fled, I felt it’s comfortable to be here. So I came in here. It’s not my fault, it’s the black ant’s fault. You go and punish the black ant.”

CI.041 Dauh ong aengx mingh naaic nziouv-muotc. CLF old_man further go ask black_ant clf n adv vi vt n

The old man further went (to) ask the black ant.

CI.042 42.1 42.2 Mingh mingh taux nziouv-muotc wuov naaic nziouv-muotc, go go reach black_ant that ask black_ant vi vi v n pn.dem vt n "Hnangv meih la'guaih maaih_zingh_mouc_ziev danx naang zoux_haaix_nyungc? like 2sg carelessly without_thiking sting snake why adv pn.p adv idm vt n q Danx liuz naang, naang biaux gau sting finish snake snake flee DPCL vt aux n n vt part ninh zoux dorngc sic mi'aqv niaa." 3sg do be_wrong matter RSLT I-am-thinking pn.p vt v.st n asp part.f 42.1(The old man) went, went to the black ant’s place without thinking anything you stung the snake carelessly, why? You stung the snake, because it ran away, it commited a crime, won’t you see?”

42.2(and) asked the black ant, “Like you,

Page 247: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

221 CI.043 43.1

43.4

claw_ground 1_PL come_out come vt pn.p v vi

ot that I myself wanted to sting. ld chicken himself clawed me out. y house to enter I got angry; it was because I faced something (problem). It’s not that I stung without a reason, you see. themselves came to bother me and drove me out? ou want to punish, go and punish the wild chicken. e wild chicken who clawed the black ant out.

CI.044 44.1 44.2 44.3 Nduov nv dauh ong youc mv_haih_fungc_zoux aengx mingh naaic norqc_jai. deceive this CLF old_man also can’t_do_anything further go ask wild_chicken

Nziouv-muotc youc gorngv, black_ant also speak n seq.mk v 43.2 "Mv zeiz yie ganh oix danx lorqc. NEG be 1sg self want_to sting you_see neg vi pn.p pn.emph v.mod vt part.f 43.3 Norqc_jai ganh mbuonc yie cuotv daaih. wild_chicken self claw_ground 1sg come_out come n pn.emph vt pn.p v vi

Lorz mv buatc biauv bieqc qiex_jiez haic, look_for NEG see house enter angry very vt neg vt n vi v.cmp adv buangh haaix_nyungc mv zeiz la'guaih danx aqv la'maaqh. meet what_kind NEG be_right carelessly sting PFT I_am_telling_you vt q n neg vi adv vt part.asp part.f 43.5 Ninh_mbuo ganh daaih hoic yie nyei biauv hoic yie cuotv daaih weqv. 3_PL self come harm 1sg POSS house harm 1sg come_out come ASST pn.p pn.emph vi vt pn.p poss n vt pn.p v vi part 43.6 Meih oix baatc nor aeqv, mingh baatc norqc_jai. 2sg want_to punish as TOP go punish wild_chicken pn.p v.mod vt top top vi vt n 43.7 Norqc_jai ganh mbuonc nziouv-muotc cuotv. wild_chicken self claw_ground black_ant come_out n pn.emph vt n iv 43.8 Mbuonc yie_mbuo cuotv daaih."

43.1The black ant also said, “43.2It’s n 43.3The wi43.4Because I could not find m

43.5Don’t you see that they 43.6If y

43.7It was th 43.8He clawed us out.”

vt pn.dem clf n seq.mk idm adv vi vt n 44.1(The black ant) deceived the old man 44.2there was nothing (the old man) could do about it, 44.3(he) further went (and) asked the wild chicken.

Page 248: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

222 CI.045 45.1 Norqc_jai youc gorngv, wild_chicken also speak n seq.mk v 45.2 "Saqv-nyim ganh bieqc yie nyei m'zing, sesame-seed self enter 1sg POSS eye n-n pn.emph vi pn.p poss n yie mangc mv buatc haaix_nyungc. 1sg look NEG see what pn.p vt neg vt q 45.3 Mv zeiz la'guaih mbuonc aqv la'maaqh. NEG be_right carelessly claw_ground PFT I_am_telling_you neg vi adv vt part.asp part.f 45.4 La'guaih mbuonc carelessly claw_ground adv vt mv zeiz mbuonc cuotv daaih la'maaqh. NEG be_right claw_ground come_out come I_am_telling_you neg vi vt vt vi part.f 45.5 Gorngv yie m'zing longx m'zing njang nor aeqv, if 1sg eye good eye bright as TOP conj pn.p n v.st n v.st top top yie mv_baac simv jienv nziouv, yie mv mbuonc lorqc." 1sg but avoid DUR ant 1sg NEG claw_ground you-see pn.p conj vt asp v.st pn.p neg vt part.f 45.1The wild chicken also said, “45.2The sesame seeds got into my eys, I couldn’t see anything. 45.3It’s not because I clawed carelessly. 45.4I didn’t carelessly claw and make the ant come out. 45.5If I could see well, even I could have avoided ants and clawed different places, you see?”

CI.046 46.1 46.2 46.3 Nduov dauh ong youc maiv_haih_fungc_zoux aengx mingh naaic saqv, deceive CLF old_man also can’t_do_anything further go ask sesame vt clf n seq.mk idm adv vi vt n "Hnangv meih maaih_zingh_mouc_ziev la'guaih mingh like 2sg without_thinking carelessly go adv pn.p idm adv vi bieqc norqc_jai nyei m'zing zoux_haaix_nyungc?" enter wild_chicken POSS eye Why? vi n poss n q 46.1(The wild chicken) deceived the old man, 46.2there was nothing (the old man) could do about it, 46.3(he) further went (and) asked the sesame, “Like you without thinking anything went carelessly and entered the wild chicken’s eyes, why?”

Page 249: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

223

CI.047 Saqv youc gorngv, sesame also speak n seq.mk v "Maiv zeiz yie ganh oix bieqc lorqc, NEG be_right 1sg self want_to enter you-see neg vi pn.p pn.emph v.mod vi part.f sopc ganh mborqv zuqc yie, yie cingx_daaih yiem mv jienv, ash_pumpkin self hit GOAL 1sg 1sg therefore be_in NEG stable n pn.emph vt part pn.p pn.p conj vi neg v.st ndortv mingh bieqc norqc_jai m'zing." fall go enter wild_chicken eye vi vi vi n n

The sesame also said, “Not that I myself wanted to enter, you see. The ash pumpkin itself hit me; that’s why I could not stay there, fell down and went into the wild chicken’s eyes.”

CI.048 48.1 48.2 48.3 Nduov dauh ong youc maiv_haih_fungc_zoux aengx mingh naaic sopc, deceive CLF old_man also can't_do_anything further go ask ash_pumpkin vt clf n seq.mk idm adv vi vt n 48.4 naaic gaax sopc, ask try_and_see ash_pumpkin vt part n "Hnangv meih maaih_zingh_mouc_ziev la'guaih mingh like 2sg without_thinking carelessly go adv pn.p vt adv vi mborqv saqv cuotv bieqc norqc_jai m'zing zoux_haaix_nyungc? hit sesame come_out enter wild_chicken eye Why? vt n vi vi n n q Ninh_mbuo zoux dorngc sic mi'aqv niaa. 3_PL do wrong matter RSLT I-am-thinking pn.p vt v.st n asp part.f Yie oix baatc meih aqv." 1sg want_to punish 2sg INCHOAT pn.p v.mod vt pn.p part.asp 48.1(The sesame) deceived the old man, 48.2there was nothing (the old man) could do about it, 48.3(he) went (to) ask the ash pumpkin, 48.4(he) asked the ash pumpkin, “Like you, without thinking anything you went carelessly and hit the sesame to come out and entered the wild chicken’s eyes, why? They have done wrong, you see. I want to punish you.”

Page 250: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

224 CI.049 49.1 49.2 49.3 Gorngv nduov ninh youc mv_haih_fungc_zoux aengx gorngv, speak deceive 3sg also can't_do_anything further speak v vt pn seq.mk idm adv v 49.4 "Yie mv_baac maiv zeiz yie ganh oix njangx lorqc. 1sg but NEG be_right 1sg self want_to roll you-see pn.p conj neg vi pn.p pn.emph v.mod vi part.f 49.5 Jung ganh caaiv zuqc yie nyei luangh dangx barking_deer self step_on GOAL 1sg POSS vine break_off n pn.emph vt part pn.p poss n v yie mv dongh yiem mv jienv, yie cingx_daaih aqv_zuqc njangx. 1sg NEG * be_there NEG stable 1sg therefore must roll pn.p neg neg v.st pn.p conj aux vi 49.6 Njangx mingh youc simv maiv hingh cingx_daaih mborqv zuqc saqv nyei gorn. roll go also avoid NEG win therefore hit GOAL sesame POSS base vi vi seq.mk vt neg vt conj vt part n poss n 49.7 Meih mingh naaic jung oc. 2sg go ask barking_deer SFT.CMD pn.p vi vt n part

49.8 Naaic gaax ninh zoux_haaix_nyungc daaih caaiv yie." ask try_and_see 3sg Why? come step_on 1sg vt part pn.p q vi vt pn.p 49.1(The ash pumpkin) spoke (and) deceived (the old man), 49.2there was nothing (the old man) could do about it again, 49.3(the ash pumpkin) also said, “49.4I myself didn’t want to roll. 49.5The barking deer himself stepped on me (and) the vine broke off; no way for me to keep staying there, that’s why I had to roll down. 49.6Rolling down then I couldn’t avoid and then hit the sesame’s bottom. 49.7You go and ask the barking deer. 49.8Ask him why he came and stepped on me.?

CI.050 50.1 50.2 Nduov ninh aengx mingh naaic jung. deceive 3sg further go ask barking_deer vt pn.p adv vi vt n 50.1(The ash pumpkin) deceived him, 50.2(the old man) again went (and) asked the barking deer.

Page 251: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

225 CI.051 Jung gorngv, barking_deer speak n v "Laaix wuov_deix i buo nyungc ga'naaiv, gorngv haeqv yie, because those two three kind thing speak frighten 1sg conj pn.dem numb numb n n v vt pn.p yie cingx_daaih tiux. 1sg therefore run pn.p conj vi Ninh_mbuo mv gorngv haeqv yie, yie mv_baac maiv dorngc sic." 3_PL NEG speak frighten 1sg 1sg but NEG wrong matter pn.p neg v vt pn.p pn.p conj neg v.st n

The barking deer said, “Because those two or three things said (something and) frightened me, then I ran away. If they hadn’t frightened me, I myself wouldn’t have done wrong.”

CI.052 Naaic gaax, ask try_and_see v part "Hnangv_haaix_nor gorngv, ninh_mbuo hnangv_haaix_nor gorngv haeqv meih?" how speak 3_PL how speak frighten 2sg q v pn.p q v vt pn.p

(The old man) asked, “What did they say? How did they frighten you?”

CI.053 Dauh jung gorngv, CLF barking_deer speak clf n v "Oh, norqc_guv_long ganh nduov ninh gorngv Oh owl self deceive 3sg speak intj n pn.emph vt pn.p v se ma'gorkv_biouv zuoqc, maqvgorkv_biouv zuoqc." that_is olive_fruit ripe olive_fruit ripe copl n v.st n v.st

The barking deer said, “Oh, the owl himself deceived (and) he said that olive fruits are ripe, olive fruits are ripe.”

CI.054 Nduov ninh mingh nyanc ma'gorkv_biouv. deceive 3sg go eat olive_fruit vt pn.p vi vt n

(The owl) enticed him (to) go (and) eat the olive fruits.

CI.055 55.1 55.2 Nyanc haiz kuv nyei, yiem wuov nyanc. eat feel delicious PRS.ST be_there there eat vt vi v.st asp vi pn.dem vt 55.1(I) ate (it and it) felt delicious, 55.2(I) ate there.

Page 252: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

226 CI.056 56.1 56.2 56.3 Nyanc gau, haiz mbopv_yangh gorngv, nqongh zorqv, nqongh zorqv, nqongh zorqv. eat DPCL hear yellow_squirrel speak hurry take hurry take hurry take vt part vt n v vi vt vi vt vi vt

vi vt vi vt v vt

as TOP 1sg NEG fear q

58.6 Mv tiux ninh_mbuo zorqv yie niaa." NEG run 3_PL catch 1sg I_am_thinking

56.1When (he)ate, 56.2(he) heard a yellow squirrel say, ‘56.3Seize quickly, seize quickly, seize quickly!’

CI.057 57.1 Aav_dangh gaeng-waen youc aengx heuc gorngv se, just_a_minute cicada also further call speak that_is adv.tmp n seq.mk adv vi vt copl 57.2 'nqongh caa, nqongh caa, nqongh caa.’ ' hurry catch hurry catch hurry catch

57.1Soon, a cicada also cried (and) said ’57.2Catch quickly, catch quickly, catch quickly!’

CI.058 58.1 Ov yie m'daaih hnamv naaic aeqv, SURPRISE 1sg from_the_beginning think that TOP intrj pn.p adv vt pn.dem part ninh_mbuo i dauh zungv hnangv nc nor caangh_laangh. 3_PL two CLF already like that as discuss pn.p numb clf adv.tmp adv pn.dem top v 58.2 Aav_dangh ninh_mbuo haac_mv za'gengh zorqv yie? just_a_minute 3_PL don't_you_think_possible really catch 1sg adv.tmp pn.p idm adv vt pn.p 58.3 Yie haac_mv daic? 1sg don't_you_think_possible die pn.p idm vi 58.4 Nor aeqv, yie mv_dongh gamh_nziex nyei?

top part pn.p neg v.cmp part 58.5 Yie mv_dongh aqv_zuqc tiux? 1sg NEG must run pn.p neg aux vi

neg vi pn.p vt pn.p part.f 58.1Oh, of course I thought like that, that two of them discussed like that. 58.2Don’t you think it’s possible that they would really catch me soon? 58.3Don’t you think I could die? 58.4If so, I couldn’t be unafraid. 58.5I couldn’t help running away. 58.6If I would not run away, they would catch me, you see.”

Page 253: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

227 CI.059 59.1 59.2 Nqemh mbopv ngaatc biouv-youh ndortv mborqv ninh nyei in-dang norh.

bite(ing) the oil fruit (and letting it) drop, t) hit his opium smoking lamp, that’s what he [deer] said.

60.4 60.5 60.6

62.1 62.2

accuse squirrel bite oil_fruit fall hit 3sg POSS opium_smoking_lamp REPORT vt n vt n vi vt pn.p poss n part 59.1(The barking deer) accused the squirrel (of) 59.2(and i

CI.060 60.1 Dauh ong cingx_daaih zimh jomc nc norm dorngx CLF old_man therefore investigate end that CLF place clf n conj vt vi pn.dem clf n 60.2 zimh cuotv gorn daaih aeqv, investigate go_out root come TOP vt v n vi part 60.3 norqc_guv_long nduov jung mingh nyanc biouv, owl deceive barking_deer go eat fruit n vt n vi vt n

ninh_mbuo zorqv sung mborqv mborqv daaih nanv ninh nyei wuom cuotv daaih aeqv, 3_PL take ginger pound pound come squeeze 3sg POSS water go_out come TOP pn.p vt n vt vt v.asp vt pn.p poss n vi v.dr part 60.7 60.8 dorh sung-wuom mingh laapc norqc_guv_long m'zing na’maah. take ginger_juice go put owl eye I_am_telling_you vt n v.dr vt n n part.f 60.1This old man, therefore, has come to the conclusion there, 60.2found out the root (of the problem); 60.3(that is,) the owl enticed the barking deer (to) go (and) eat the fruit; 60.4(so) they took ginger 60.5(and) crushed (it) 60.6(and) squeezed out the juice of it, and so, 60.7(they) took the ginger juice (and) 60.8poured (it) into the owl's eyes, you see.

CI.061 61.1 61.2 Laapc liuz cingx_daaih ih_zanc nguaiv nguaiv wuov nor aex. put finish therefore now big-eyed big-eyed that as I_mean_it vt vt conj adv.tmp v.st v.st pn.dem top part.f 61.1(After they) finished pouring (ginger juice in his eyes); 61.2therefore, now the eyes of owls are so big and round like that, I mean it.

CI.062

Haah maaih mbopv-yangh aeqv, ninh_mbuo qiex_jiez haic aeqv *** have squirrel-yellow TOP 3_PL angry very TOP intrj vt n-v.st part pn.p v.st adv part 62.3 dorh mingh jaamh norh aex. take go singe REPORT I_mean_it vt vi vt part part.f 62.1As for the yellow squirrel, 62.2because they were very angry, 62.3(they) took (him) away (and) singed (him), that’s what they say, I mean it.

Page 254: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

228 CI.063 63.1 63.2 Jaamh ninh nyei diqc_daanz ih_zanc cingx_daaih ngoi yangh nyei.

ninh_mbuo cingx_daaih zorqv ninh nyei jaangh wetv cuotv mingh

CI.065

Weic_naaiv gaeng-waen ganh, gaeng-waen mv_baac mv funx jienv ninh nduov lorqc.

singe 3sg POSS upper_back now therefore INTENSE-FOR-YELLOW yellow PRS.ST vt pn.p poss n adv.tmp conj part v.st asp 63.1(They) singed his back, 63.2therefore, nowadays, (its back) is totally yellow.

CI.064 64.1 Aengx maaih gaeng-waen aeqh, further have cicada SUSPND adv vt n intrj 64.2 64.3 64.4 Faix haic hnangv_haaix_nor zoux yaac maiv haih zoux aeqv, small very how do TOP NEG be_possible_to do TOP v.st adv adv vt part neg aux vt part 64.5 64.6

3_PL therefore take 3sg POSS intestine dig come_out go pn.p conj vt pn.p poss n vt vi v.asp 64.7 div wuov dauh ong nyei in-dang norh aah. substitue that CLF grandfather POSS opium_smoking_lamp REPORT listen_Please!! vt pn.dem clf n poss n part part.f 64.1Moreover, talking about the cicada…64.2(he was) too small, 64.3whatever (they tried to) do 64.4(they) could not do (it) and so, 64.5they, therefore, took his intestine 64.6dug (it) out, 64.7replaced that old man’s opium smoking lamp (with it), that’s what they say, listen to me!

Nv diuh gouv ninh_mbuo hnangv nc nor gorngv ae. this CLF story 3_PL like that as speak SUSPND pn.dem clf n pn.p adv pn.dem top v part

This story, they tell like this, and...

CI.066 Naaiv gorn-baengx mv_baac maaih eix_leiz nyei oh. this foundation but have meaning AFM Oh pn.dem n conj vt n asp intj

Underneath it, there is a meaning too.

CI.067

for_this_reason cicada self cicada but NEG count DUR 3sg deceive you-see conj n pn.emph n conj neg vt asp pn.p vt part.f

For this reason, the cicada himself, even the cicada doesn’t consider (that) he deceived, you see.

CI.068 Ninh nyei qiex m'daaih hnangv nc nor heuc. 3sg POSS voice from_the_beginning like that as call pn.p poss n adv adv pn.dem top v

From the beginning his voice sounded like that.

Page 255: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

229 CI.069 69.1 69.2 Jung ganh hngongx, ninh ganh hnamv haeqv zuqc ninh ganh. barking_deer self stupid 3sg self think frighten GOAL 3sg self n pn.emph v.st pn.p pn.emph vt vt part pn.p pn.emph 69.1The barking deer himself was stupid; 69.2he himself thought he was frightened.

CI.070

n-v.st top v.cmp

The yellow squirrel also lost in court.

pn.p conj vt pn.p adv pn.dem top

CI.073

Now, one person has one story, isn’t that right?

CI.074

CI.075

CI.076

70.1 70.2 La'guaih tiux aeqv, nqa'haav ninh_mbuo cingx_daaih baeng daaih gorngv gau carelessly run TOP later 3_PL therefore pull come speak DPCL adv vi part adv.tmp pn.p conj vt v.dr vt part 70.3 70.4 gaeng-waen cingx_daaih ndortv_leiz, norqc_guv_long yaac ndortv_leiz. cicada therefore lose_in_court owl also lose_in_court n conj v.cmp n conj v.cmp 70.1(The barking deer) ran carelessly and, 70.2that’s why, later, they pulled over (and) talked and then, 70.3 the cicada, therefore, lost in court, 70.4the owl also lost in court.

CI.071 Bbopv-yangh yaac ndortv_leiz. squirrel-yellow also lose_in_court

CI.072 Ninh_mbuo cingx_daaih hoic ninh_mbuo hnangv nc nor. 3_PL therefore harm 3_PL like that as

Therefore they punished them like that.

Ih_zanc yietc laanh maaih diuh gouv la'maaqh. now one CL have CLF story I_am_telling_you adv.tmp numb cl vt clf n part.f

Jangx duqv nyei, meih? remember can AFM 2sg vt v.aux asp pn.p

Can you remember?

Jangx mv duqv dih_hnoi yie aengx gorngv. remember NEG can some_day 1sg again speak vt neg v.aux adv.tmp pn.p adv v

If you can’t remember, I will tell it to you again sometimes.

Jangx duqv nor aeqv, dih_hnoi meih gorngv bun yie muangx oc. remember can as TOP some_day 2sg speak let 1sg listen SFT.CMD vt v.aux top top adv.tmp pn.p v vt pn.p v part

If you can remember, someday you tell it (and) let me listen.

Page 256: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

230 CI.077 Mv qaqv-jangx duqv haic mbor_lor. NEG memorize can very I_wonder neg vt v.aux adv part.f

Can’t memorize it, I wonder.

CI.078 Za'gengh jangx duqv aeqv, longx la'maaqh.

If you really remember it, it’s good, you see.

really remember can TOP good you_should_agree_with_me adv v v.aux top v.st part.f

Page 257: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

231

Story 5 (FA): “A Story of Faam-Bae’ and Aeng-Doi” Faam Baeqv Aengh Doih Faam-Baeqv Aengh^Doih Nyei Gouv (Yauz-Guangv gorngv) Faam-Bae’ Aeng-Doi story Yau-Guang speak n.prp n.prp n n.prp v FA.001 Yie gorngv taux Faam-Baeqv caux Aengh^Doih. 1sg speak reach Faam-Bae’ and Aeng-Doi pn.p v vt n.prp conj n.prp

I am going to tell a story of Faam-Bae’ and Aeng-Doi.

FA.002 Loz-hnoi maaih dauh m'jangc dorn heuc Faam-Baeqv. in_the_old_days have CLF man call Faam-Bae’ adv.tmp vt clf n v n.prp

In the old days, there was a man (whose) name was Faam-Bae’.

There was (also) a girl (whose) name was Aeng-Doi

FA.004

She wanted (to) go to Meisen school (to) study.

FA.005 5.1 5.2

5.1In those days, women could not study she wanted to go.

3sg can’t_do_anything 3sg plan go_out one kind plan

FA.007

FA.003 Maaih dauh m'sieqv_dorn heuc Aengh^Doih. have CLF girl call Aeng-Doi vt clf n v n.prp

Ninh oix mingh Meih^Sen horqc doqc_ sou. 3sg want_to go Mei-Sen school study_book pn.p v.mod vi n.prp n v.cmp

Wuov_ zanc hnoi m'sieqv_dorn mv haih doqc_ sou mv_baac ninh oix mingh. that time day girl NEG be_possible_to study_book but 3sg like go pn.dem n n n neg aux v.cmp conj pn.p v.mod vi

5.2but

FA.006 6.1 6.2 Ninh mv_ haih_fungc_zoux, ninh daav cuotv yietc nyungc za'eix.

pn.p idm pn.p vt v numb n n 6.1There was nothing he could do about it, (but) she made a plan. 6.2

7.1 7.2 Ninh zorqv m'sieqv_dorn lui jaiv guangc nzengc mingh, 3sg take girl clothes undress discard consumed go pn.p vt n n vt vt vi v.asp 7.3 7.4 zuqv jienv m'jangc_dorn nyei ga'naaiv zorng benx m'jangc_dorn mingh. put_on DUR man POSS thing dress be man go

7.1She took off girl’s clothes undressed all, (and) put on man’s clothes 7.4dressed up as a man. 7.2 7.3

vt asp n poss n vt vi n v.asp

Page 258: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

232 FA.008 8.1 8.2

8.3 aengx zorqv norm nqopv jienv gu'nguaaic,

10.1 10.2

pn.p top neg vt n.prp vi

Ninh mingh wuov_zanc ninh yaac zorqv norm yienv zaangh wuom daaih, 3sg go that_period 3sg also take CLF bowl dip water come pn.p vi adv.tmp pn.p top vt clf n vt n v.dr

again take CLF turn_over&cover DUR above adv vt clf v asp adv.loc 8.4 heuc neiz maa mbuo yiem biauv, order her mother PL be_in house vt contr n pl vi n "Mv dungx koi ninh mc norm yienv oc". NEG.CMD open 3sg that CLF bowl PLT neg vt pn pn.dem clf n part 8.1When she (was about to) go, 8.2she took a bowl (and) dipped water, 8.3and took another (and) turned (it) over above the other as a cover, 8.4told her mother and others (who) were at home, “Please don’t open this bowl.”

FA.009 9.1 9.2 Zorng jienv m'jangc_dorn aeqv, mingh aqv. dress DUR man TOP go INCHOAT vt asp n part vi part.asp 9.1As (she) dressed up as a boy, 9.2(she) was ready to go.

FA.010

Ninh nyei maa mbuo fungx jienv ninh cuotv jauv. 3sg POSS mother PL send DUR 3sg go_out road pn.p poss n pl vt asp pn.p v n 10.1Her mother and others sent her off 10.2(she) set out.

FA.011 Haaix_nyungc yaac mbuox ninh, what_kind also tell 3sg q.cont top vt pn.p "Mingh aeqv, aqv_zuqc faix_fim". go TOP must take_care vi top aux v.cmp

(Mother) told her everything, “When you go, you must take care.”

FA.012 Ninh yaac mv hiuv_duqv Faam-Baeqv mingh. 3sg also NEG know Faam-Bae’ go

She didn't know (that) Faam-Bae’ was going (to study too).

FA.013 Faam-Baeqv yaac mv hiuv_duqv ninh mingh. Faam-Bae’ also NEG know 3sg go n.prp top neg vt pn.p vi

Faam-Bae’ didn’t know (that) she was going either.

Page 259: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

233 FA.014 14.1 14.2 14.3 Mingh taux ndaamv-jauv buangh zuqc Faam-Baeqv naaic gaax Aengh^Doih, go reach half-way meet GOAL Faam-Bae’ ask try_and_see Aeng-Doi vi v n vt part n.prp v part n.prp "Meih mingh haaix?" 2sg go where pn.p vi q 14.1(Aeng-Doi) went half way, 14.2(she) ran into Faam-Bae’ 14.3(and he) asked Aeng-Doi, “Where are you going?”

FA.017 Aengh^Doih gorngv mbuox Faam-Baeqv, Aeng-Doi speak tell Faam-Bae’ n.prp v vt n.prp "Ov, yie mingh Meih^Sen horqc." SURPRISE 1sg go Mei-Sen school exclam pn.p vi n.prp n

Aeng-Doi said to Faam-Bae’, “ Oh, I am going to Meisen school.

FA.018 Aengh^Doih aengx naaic gaax Faam-Baeqv, Aeng-Doi again ask try_and_see Faam-Bae’ n.prp adv v part n.prp "Meih mingh haaix?" 2sg go where pn.p vi q

Aeng-Doi also asked Faam-Bae’, “Where are you going?”

FA.019 Faam-Baeqv yaac mbuox Aengh^Doih, Faam-Bae’ also tell Aeng-Doi n.prp top vt n.prp "Yie yaac mingh Meih^Sen horqc." 1sg also go Mei-Sen school pn.p top vi n.prp

Faam-Bae’ also told Aeng-Doi, “I am going to Meisen school too.”

FA.020 20.1 20.2 Ninh_mbuo ziouc yiem wuov mbaih jaapv-zaangv 3_PL so be_there there put_side_by_side age pn.p seq.mk vi pn.dem vt n 20.3 mangc gaax haaix_dauh gauh gox. look try_and_see who more old v part q_clf cmpr v.st 20.1They be(ing) there 20.2compared 20.3(to) see who was older.

Page 260: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

234

FA.021 21.1 21.2 Mbaih gau, Faam-Baeqv gox i hnyangx put_side_by_side DPCL Faam-Bae’ old two year vt part n.prp v.st numb n 21.3 mv_baac Faam-Baeqv funx Aengh^Doih zoux m'jangc_dorn hnangv. but Faam-Bae’ reckon Aeng-Doi be man only conj n.prp vt n.prp vi n adv 21.1As (they) compared, 21.2Faam-Bae’ was two years older; 21.3but Faam-Bae’ only thought (that) Aeng-Doi was a boy.

FA.022 Wuov_zanc ninh mv haih funx Aengh^Doih zoux m'sieqv_dorn. that_time 3sg NEG be_able_to count Aeng-Doi be girl adv.tmp pn.p neg aux vt n.prp vi n

At that time, he could not think (that) Aeng-Doi was a girl.

FA.023 23.1 23.2 Lomh_nzoih zoux m'jangc_dorn doic mingh doqc_sou hnangv. together make man friend go study_book only adv vt n n vi v.cmp adv 23.1They made boy friends together 23.2went to study only.

FA.024 24.1 24.2 Ninh_mbuo mingh yaac hnamv, yiem haaix yaac hnamv. 3_PL go and love be_in what also love pn.p vi top vt vi q top vt 24.1Wherever they went (they) liked (each other), 24.2wherever (they) were (they) liked (each other).

FA.025 Aengh^Doih gorngv mbuox Faam-Baeqv, Aeng-Doi speak tell Faam-Bae’ n.prp v vt n.prp "Mbuo mbaih liuz nin-gaeng, mbaih liuz jaapv-zaangv aqv, 1pl compare finish birthday compare finish age PFT pn.p vt vi n vt asp n part.asp mbuo mingh aqv. Meih gox i hnyangx, meih mingh ndaangc. 1PL go INCHOAT 2sg old two year 2sg go front pn.p vi part.asp pn.p v.st numb n pn.p vi n Yie lunx i hnyangx yie nqa'haav hnengx sou-faang." 1sg young two year 1sg later carry school_bag pn.p v.st numb n pn.p adv.tmp vt n

Aeng-Doi said to Faam-Bae’, “We have finished comparing (our) birthdays; finished comparing (out) ages. Let’s go. You are two years older (so) you go in front of me. I am two years younger, I will carry your school bag behind you.”

Page 261: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

235

FA.026 Ninh_mbuo za'gengh mingh taux Meih^Sen horqc. 3_PL really go reach Mei-Sen school pn.p adv vi v n.prp n

They indeed arrived at the Meisen school.

FA.027 27.1 27.2 Yiem haaix ndau yaac nziaauc duqv maaih doic. be_in where place also play_with get have friend vi q n top vi vt vt n 27.1Wherever (they) might be, 27.2(they) could have friends to play with.

FA.028 Aengh^Doih hiuv_duqv Faam-Baeqv m'jangc_dorn nyei. Aeng-Doi know Faam-Bae’ boy PRS.ST n.prp vt n.prp n asp

Aeng-Doi knew (that) Faam-Bae’ was a boy.

FA.029 Faam-Baeqv mv hiuv_duqv Aengh^Doih m'sieqv_dorn. Faam-Bae’ NEG know Aeng-Doi girl n.prp neg vt n.prp n

Faam-Bae’ didn’t know (that) Aeng-Doi was a girl.

FA.030 30.1 30.2 Mingh haaix ndau yaac yietc jauv mingh. go where place also one road go vi q n top numb n vi 30.1Wherever (they) went, 30.2(they would) go together.

FA.031 31.1 31.2 Lungh_muonz bueix yaac juangc coux bueix. night sleep also share bed sleep n vi top vt n vi 31.1Sleeping at night, 31.2(they would) share a bed (to) sleep.

FA.032 32.1 32.2 32.3 Lungh_hnoi doqc_sou yaac juangc norm dieh yaac juangc zung eiv zueiz. day_time study_book also share CLF desk and share CLF chair sit n v.cmp top vt clf n conj vt clf n vi 32.1In regard to studying in the day time, 32.2(they) shared a desk 32.3and shared one chair.

FA.033 33.1 33.2 Fiev nzaangc yaac juangc baengx batv. write letter TOP share CLF pencil vt n part vt clf n 33.1As for write(ing) characters, 33.2(they) shared a pencil.

Page 262: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

236

FA.034 34.1 34.2 Doqc_sou yaac juangc buonv sou. study_book TOP share CLF book v.cmp part vt clf n 34.1As for studying, 34.2(they) shared books.

FA.035 35.1 35.2 Doqc gau duqv buo hnyangx, read DPCL attain three year vt part vt numb n 35.3 35.4 35.5 ninh nyei maa mbuo yiem biauv haiz ninh nyei sieqv mingh duqv lauh haic, 3sg POSS mother PL live house feel 3sg POSS daughter go attain long very pn.p poss n pl vi n vi pn.p poss n vi vt v.st adv 35.6 35.7 ninh sieqv yaac mv nzuonx. 3sg daughter also NEG return pn.p n top neg v 35.1After studying 35.2for three years, 35.3her mother and others 35.4at home 35.5felt (that) their daughter had been gone very long 35.6as for their daughter 35.7(she) had never returned home.

FA.036 Ninh nyei maa mbuo jouc koi ninh mc norm yienv. 3sg POSS mother PL then open 3sg that CLF bowl pn.p poss n pl seq.mk vt pn.p pn.dem clf n

So her mother and others opened that bowl of hers.

FA.037 37.1 37.2 Neiz sieqv yiem Meih^Sen horqc ziouc bingx nyorx mv mbueiz, her daughter be_there Mei-Sen school so hide breast NEG be_concealed contr n vi n.prp n seq.mk vt n neg vi 37.3 37.4 37.5 ninh nyei sai-diex hiuv_duqv Aengh^Doih se mv zeiz m'jangc_dorn, m'sieqv_dorn 3sg POSS teacher knowAeng-Doi that_is NEG be boy girl pn.p poss n vt n.prp copl neg vi n n 37.6 mv_baac Faam-Baeqv mv gaengh hiuv. but Faam-Bae’ NEG really know conj n.prp neg adv v 37.1Her daughter 37.2at Meisen school, then could not hide her breasts (from growing); 37.3her teacher knew (that) 37.4Aeng-Doi was not a boy, 37.5(but actually) a girl.

Page 263: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

237

FA.038 38.1 38.2 Ninh nyei sai-diex se weic ninh guai haic 3sg POSS teacher that_is because 3sg clever very pn.p poss n copl conj pn.p v.st adv 38.3 ninh hiuv_duqv deix m'nziex m'sieqv_dorn, 3sg know some herhaps girl pn.p vt adv adv n 38.4 ninh nyei sai-diex ziouc seix ninh aqv. 3sg POSS teacher so try 3sg INCHOAT pn.p poss n seq.mk vt pn.p part.asp 38.1Her teacher, 38.2because he was quite clever, 38.3he knew somehow (that) she might perhaps be a girl; 38.4so her teacher was going to try her.

FA.039 39.1 39.2 39.3 Ninh nyei sai-diex daaih caux ninh_mbuo bueix. 3sg POSS teacher come be_with 3_PL sleep pn.p poss n vi vt pn.p vi 39.1Her teacher came 39.2(and) being with them 39.3(he) slept.

FA.040 40.1 Aengh^Doih yaac aengx daav cuotv diuc za'eix, Aeng-Doi TOP again plan go_out CLF plan n.prp part adv vt v clf n 40.2 40.3 zorqv yienv zaangh duqv yienv wuom daaih an jienv ninh_mbuo i dauh mbu'ndongx, take bowl dip get bowl water come put DUR 3_PL two CLF middle vt n vt vt n n v.dr vt asp pn.p numb clf n 40.4 an jienv suangx gu'nguaaic. put DUR blanket above vt asp n adv.loc 40.1Aeng-Doi, on the other hand, made a plan again: 40.2(she) took a bowl (and) filled (it) with water, 40.3placed (it) between the two (men), 40.4(that is) put it on the blanket.

FA.041 Yietc dauh yaac dongz mv duqv. one CLF and_yet move NEG can numb clf top vi neg v.aux

No one could move.

Page 264: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

238 FA.042 42.1 42.2 Gorngv hnangv dongz mc yienv speak only move that bowl v adv vi pn.dem n 42.3 wuom ziouc kung pietv suangx ndorn nzengc aqv, water so_then spill_out sprinkle blanket wet consumed INCHOAT n seq.mk v vt n v.st v part.asp 42.4 mv haih bueix. NEG be_possible_to sleep neg aux vi 42.1Just speaking 42.2would move that bowl, 42.3then the water would spill over the blanket; 42.4(they) could not sleep.

FA.043 43.1 Nduov ninh nyei sai-diex deceive 3sg POSS teacher vt pn.p poss n 43.2 43.3 caux ninh bueix yietc muonz yaac mv haih hiuv_duqv, be_with 3sg sleep one night and_yet NEG be_possible_to know vt pn.p vi numb n top neg aux vt 43.4 gamh_nziex maiv dongz. fear NEG move v.cmp neg vi 43.1(She) tricked her teacher: 43.2though (he) slept with her (i.e. them) for a night, 43.3(he) could not find out (anything); 43.4(he) was afraid of moving.

FA.044 Da'nyeic hnoi aengx ninh nyei sai-diex gorngv, second day again 3sg POSS teacher speak n n adv pn.p poss n v "Yie njaaux mh_mbuo duqv i buo hnyangx aqv. 1sg teach 2_PL attain two three year PFT pn.p vt pn.p vt numb numb n part.asp Mbuo ih_hnoi seix gaax haaix_dauh bungx yiez ndiu gauh go?" 1pl today try try_and_see who release urine spurt more far pn.p adv.tmp vt part q_clf vt n v cmpr v.st

The following day their teacher said again, “I have been teaching you for two or three years already. Today let’s try and see who can spurt urine farthest.”

FA.045 Aengh^Doih aengx daav cuotv diuc za'eix. Aeng-Doi again plan go_out CLF plan n.prp adv vt v clf n

Aeng-Doi made a plan again.

Page 265: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

239 FA.046 46.1 Ninh mingh nqen duqv nqanx hlauv-dorn daaih 3sg go chop get CLF small_bamboo come pn.p vi vt vt clf n v.dr 46.2 46.3 topv jienv ninh nyei qam-gorn saaix bungx yiez. insert DUR 3sg POSS genital_area compete release urine vt asp pn.p poss n vt vt n 46.1She went (and) chopped a small bamboo, 46.2interted in her genital part, 46.3(and) competed (to) release urine.

FA.047 47.1 47.2 Neiz sai-diex nyei yaac ndiu mv ndongc ninh go. her teacher POSS TOP spurt NEG as_much_as 3sg far contr n poss top vt neg adv pn.p v.st 47.1Even her teacher 47.2did not spurt as far as she (did).

FA.048 Ninh nyei sai-diex seix ninh i nyungc. 3sg POSS teacher try 3sg two kind pn.p poss n vt pn.p numb n

Her teacher tried two kinds (of tests).

FA.049 Ninh haiz za'gengh mv haih yiem. 3sg feel really NEG be_possible_to be_in pn.p vi adv neg aux vi

She felt she really could not stay there.

FA.050 Ninh za'gengh oix nzuonx aqv. 3sg really want_to return CHG-O-ST pn.p adv v.mod vi part.asp

She really wanted to go home.

FA.051 51.1 Aengh^Doih ziouc mingh maaiz zeiv daaih, Aeng-Doi so go buy paper come n.prp seq.mk vi vt n vi 51.2 51.3 ceu cuotv ninh nyei buonc sou dorh jienv nzuonx. transcribe go_out 3sg POSS part book take SML.ACT return vt v pn.p poss n n vt asp vi 51.1So Aeng-Doi bought (some) paper, 51.2transcribed her part of the textbook 51.3(to) take home.

FA.052 Faam-Baeqv fungx jienv ninh nzuonxv taux ndaamv-jauv. Faam-Bae’ send DUR 3sg return reach half-way n.prp vt asp pn.p v v n

Faam-Bae’ sent her off half way.

Page 266: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

240 FA.053 Yietc dauh yaac siev mv duqv mingh. one CLF also give_up NEG managed_to go numb clf top vt neg vt vi

Neither of them could manage to go.

FA.054 Yiem wuov haaix_nyungc waac yaac gorngv nzengc. be_in there what word TOP speak consumed vi pn.loc q n part v vi

There (they) talked about anything altogether.

FA.055 Faam-Baeqv gorngv mbuox Aengh^Doih, Faam-Bae’ speak tell Aeng-Doi n.prp v vt n.prp "Meih nzuonx mingh, jangx deix yie oc." 2sg return go remember some 1sg PLT pn.p vi vi vt adv pn.p part

Faam-Bae’ said to Aeng-Doi, “When you reach home, remember me.”

FA.056 Aengh^Doih gorngv mbuox Faam-Baeqv, Aeng-Doi speak tell Faam-Bae’ n.prp v vt n.prp "Meih nzuonx meih nyei biauv nyei hnoi 2sg return 2sg POSS house POSS day pn.p v pn.p poss n poss n nzuonx taux yie wuov mingh oc." return reach 1sg that_place go PLT v v pn.p pn.dem vi part

Aeng-Doi said to Faam-Bae’, “When you go home, please visit my place.”

FA.057 Mv_baac Aengh^Doih hiuv_duqv, but Aeng-Doi know conj n.prp vt ninh haih caux Faam-Baeqv zoux I_hmuangv. 3sg be_possible_to be_with Faam-bae be husband_and_wife pn.p aux vt n.prp vi n

But Aeng-Doi knew (that) she and Faam-Bae’ could be husband and wife.

FA.058 Faam-Baeqv mv hiuv_duqv Aengh^Doih m'sieqv_dorn. Faam-Bae’ NEG know Aeng-Doi girl n.prp neg vt n.prp n

Faam-Bae’ didn’t know (that) Aeng-Doi (was) a girl.

FA.059 59.1 59.2 Nqa'haav hnangv_haaix_nor zoux ninh mv hiuv. later how do 3sg NEG know adv.tmp adv vt pn.p neg vt 59.1How he would do later, 59.2he did not know [i.e. He did not know what would happen to them later].

Page 267: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

241 FA.060 Funx Aengh^Doih za'gengh a'nziaauc doic hnangv. count Aeng-Doi really play friend only vt n.prp adv v n adv

(He) reckoned (that) Aeng-Doi was really a good friend only.

FA.061 61.1 61.2 Aengh^Doih nzuonx taux biauv yienc cuotv ninh nyei m'jangc_dorn nyei ga'naaiv, Aeng-Doi return reach house change go_out 3sg POSS man POSS thing n.prp vi v n vt v pn.p poss n poss n 61.3 61.4 zuqv jienv m'sieqv_dorn nyei lui, zorng benx m'sieqv_dorn, put_on DUR girl POSS clothes dress be girl vt asp n poss n vt vi n 61.5 benx yietc dauh sieqv nzueic nyei. be one CLF lady beautiful PRS.ST vi numb clf n v.st asp 61.1(When) Aeng-Doi returned home, 61.2(she) took off men’s clothes 61.3(and) put on women’s clothes, 61.5(she) was a beautiful lady.

FA.062 62.1 62.2 Yietc hnoi yiem jiex yietc hnoi Faam-Baeqv nzuonx taux horqc dorngh. one day live pass one day Faam-Bae’ return reach shcool place numb n vi vi numb n n.prp vi v n n 62.1Days passed by, 62.2Faam-Bae’ had been back to school.

FA.063 Ninh za'gengh hnyouv nzauh yiem jienv. 3sg really heart sad exist DUR pn.p adv n v.st vi asp

He was really in a sad situation.

FA.064 64.1 64.2 64.3 64.4 Doqc_sou yaac mv haih doqc, bueix yaac bueix mv m'njormh. study_book and_yet NEG be_able_to read sleep and_yet lie_douwn NEG sleep v.cmp top neg aux vt vi top vi neg vi 64.1As for study(ing), 64.2(he) could not do it; 64.3as for sleep(ing), 64.4(he) did not sleep.

FA.065 65.1 65.2 Aengh^Doih yiem jienv zanc-zanc maaih i laanh mienh, Aeng-Doi exist DUR time-time have two CLF person n.prp vi asp adv.tmp vt numb clf n 65.3 65.4 65.5 zoux haaix_nyungc i dauh, bueix yaac i dauh. do what two CLF sleep TOP two CLF vt q numb clf vi part numb clf 65.1(When) Aeng-Doi was there, 65.2(they) always had two people; 65.3(when they) did whatever, (there were) two; 65.4as for sleep(ing) 65.5(there were) two.

Page 268: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

242 FA.066 66.1 66.2 66.3 Ih_zanc ninh ganh hnangv, ndoqc haic, nzauh haic, now 3sg self only lonely very feel-sorry very adv.tmp pn.p pn.emph adv v.st adv v.st adv 66.4 ninh aengx mbuox sai-diex ninh nzuonx aqv. 3sg further tell teacher 3sg return CH-O-ST pn.p adv vt n pn.p vi part.mod 66.1Now he was only by himself, 66.2very lonely, 66.3very sad; 66.4further he said to his teacher (that) he decided (to) go home.

FA.067 67.1 67.2 67.3 Nzuonx aeqv, naaic jienv nzuonx, naaic taux Aengh^Doih nyei biauv. return TOP ask SML.ACT return ask reach Aeng-Doi POSS house v part vt asp vi vt vt n.prp poss n 67.1When (he) returned, 67.2(he) returned asking; 67.3(he) asked about Aeng-Doi’s house.

FA.068 68.1 68.2 Bieqc Aengh^Doih nyei biauv mingh, zueiz njiec aqv, enter Aeng-Doi POSS house go sit go_down PFT vi n.prp poss n v.dr vi vi part.asp 68.3 68.4 buatc Aengh^Doih cuotv daaih, caux ninh gorngv waac. see Aeng-Doi come_out come be_with 3sg speak word vt n.prp vi vi vt pn.p v n 68.1(He) entered Aeng-Doi’s house, 68.2sat down; 68.3saw Aeng-Doi came out, 68.4(and he) talked with her.

FA.069 69.1 Ninh corc se zieqv mv duqv Aengh^Doih, 3sg yet that_is recognize NEG can Aeng-Doi pn.p adv copl vt neg v.aux n.prp 69.2 gorh_nziex laaic haaix ndau mienh. wonder suspect where land person v.cmp v q n n 69.1He still could not recognize Aeng-Doi, 69.2(he) wondered (and) suspected where she was from.

FA.070 Ninh nyei maa mbuo gorngv naaic gaax Aengh^Doih, 3sg POSS mother PL speak ask try_and_see Aeng-Doi pn.p poss n pl v vt part n.prp "Naaic haaix ndau nyei mienh?" that where land POSS person pn.dem q n poss n

Her mother and others asked Aeng-Doi, “That (man), (he is) a man of which land?”

Page 269: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

243 FA.071 Aengh^Doih dau ninh nyei maa mbuo, Aeng-Doi answer 3sg POSS mother PL n.prp vt pn.p poss n pl "Naaic mv zeiz haaix ndau mienh. that NEG be where land person pn.de m neg vi q n n Naaic yie_mbuo juangc horqc dorngh doqc_sou, duqv buo hnyangx. that 1_PL share school place study_book attain three year pn.dem pn.p vt n n v.cmp vt numb n Yie nyei a'nziaauc doic." 1sg POSS play friend pn.p poss adv n

Aeng-Doi answered her mother and others, “He is not a stranger. He is…, We studied together in the same school for three years. He is my friend.”

FA.072 Aengh^Doih nyei maa mbuo sienx Aengh^Doih nyei waac. Aeng-Doi POSS mother PL believe Aeng-Doi POSS word n.prp poss n pl vt n.prp poss n

Aeng-Doi’s mother and others believed Aeng-Doi’s words.

FA.073 Faam-Baeqv corc mv sienx. Faam-Bae’ yet NEG believe n.prp adv neg vt

Faam-Bae’ didn’t believe yet.

FA.074 Aengh^Doih gorngv mbuox Faam-Baeqv, Aeng-Doi speak tell Faam-Bae’ n.prp v vt n.prp "Zinh_ndaangc mbuo juangc horqc dorngh doqc_sou wuov dongh yie." previously 1pl share school place study_book that that_is 1sg adv.tmp pn.p vt n n v.cmp pn.dem idntf pn.p

Aeng-Doi said to Faam-Bae’, “We studied in the same school before; that’s me.”

FA.075 75.1 75.2 Faam-Baeqv buatc naaic m'sieqv_dorn ninh mv sienx. Faam-Bae’ see that woman 3sg NEG believe n.prp v pn.dem n pn.p neg vt 75.1Faam-Bae’ saw that woman, 75.2he would not believe.

Page 270: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

244 FA.076 76.1 76.2 Aengh^Doih gorngv, Aengh^Doih aengx mbuox Faam-Baeqv, Aeng-Doi speak Aeng-Doi again tell Faam-Bae’ n.prp v n.prp adv vt n.prp "Mv sienx mv nyungc, NEG believe this kind neg vt pn.dem n mbuo koi mbuo nyei sou daaih doix mangc gaax. 1pl open 1PL POSS book come match look try_and_see pn vt pn poss n v.dr v v part Wuov_zanc mbuo yiem horqc_dorngh juangc buonv sou doqc. that_period 1PL be_in shcool share CLF book read adv.tmp pn vi n vt clf n vt Ih_zanc mbuo doix mangc gaax." now 1PL match look try_and_see adv.tmp pn v v part 76.1Aeng-Doi said, 76.2Aeng-Doi said again to Faam-Bae’, “If you don’t believe this, let’s open our textbooks (and) compare them (and) see. That time we were in the same school (and) shared the book to read. Now we compare them.”

FA.077 77.1 77.2 Ninh_mbuo gengh i dauh zorqv ninh_mbuo nyei sou doix, za'gengh zeiz nyei. 3_PL really two CLF take 3_PL POSS book match really be PRS.ST pn.p adv numb clf vt pn.p poss n vt adv vi asp 77.1They two indeed took out their books (and) compared, 77.2really (they) were right.

FA.078 Faam-Baeqv nzauh haic aqv. Faam-Bae’ sad very CHG-O-ST n.prp v.st adv part.asp

Faam-Bae’ became very sad.

FA.079 "Wuov_zanc yiem horqc_dorngh yie funx ninh m'jangc_dorn zoux a'nziaauc doic. that_period be_in shcool 1sg reckon 3sg boy be play friend adv.tmp vi n pn.p vt pn.p n vi adv n Ih_zanc buatc m'sieqv_dorn, now see woman adv.tmp vt n yie mh_nziex aqv_zuqc caux ninh zoux i_hmuangv mv_bei." 1sg afraid must be_with 3sg be husband_and_wife I_guess pn.p v.cmp aux vt pn.p vi n part.f

“At that time, in school I thought s/he was a boy, being my playmate. Now I see a woman; I am afraid I must make husband and wife with her, I guess.”

FA.080 Faam-Baeqv ziouc hnyouv mv longx nzuonx aqv. Faam-Bae’ so heart NEG good return PFT n.prp seq.mk n neg v.st v part.asp

Then Faam-Bae’ became unhappy (and) went home.

Page 271: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

245 FA.081 81.1 81.2 81.3 81.4 Nzuonx taux biauv hnyouv nzauh kungx nyiemv hnangv yaac mv nyanc. return reach house heart feel-sorry only weep only and NEG eat v v n n v adv vi adv conj neg vt 81.1 (When he) returned home, 81.2his heart was sad 81.3(and he) only wept 81.4and did not eat (anything).

FA.082 Ninh nyei maa mbuo gorngv naaic gaax ninh, 3sg POSS mother PL speak ask try_and_see 3sg pn.p poss n pn v v part pn.p "Meih yiem horqc_dorngh zoux_haaix_nyungc nzuonx? 2sg be_there shcool why return pn.p vi n q v Zoux_haaix_nyungc mv doqc_sou? why NEG study_book q neg v.cmp Zoux_haaix_nyungc ndongc naaic jaic? why as_much_as that skinny q adv pn.dem v.st Zoux_haaix_nyungc mv gorngv waac?" why NEG speak word q neg vt n

His mother and others asked him, “Why did you come back from the school? Why do you not study? Why are you so skinny? Why do you not say anything?”

FA.083 83.1 Ninh gorngv mbuox ninh nyei die maa mbuo 3sg speak tell 3sg POSS father mother PL pn.p v vt pn.p poss n n pn 83.2 83.3 ninh yiem horqc_dorngh doqc_sou maaih dauh a'nziaauc doic. 3sg be_there school study_book have CLF play friend pn.p vi n v.cmp vt clf adv n 83.1He said to his parents 83.2(that when) he was studying at school 83.3(he) had a friend.

FA.084 84.1 84.2 84.3 Wuov_zanc ninh mv hiuv_duqv m'sieqv_dorn , funx lomh_nzoih m'jangc_dorn, zoux a'nziaauc doic. that_period 3sg NEG know girl reckon together boy be play friend adv.tmp pn.p neg vt n vt adv n vi adv n 84.1At that time he did not know (she was) a girl, 84.2(he) regarded both as boys, 84.3made playmates.

FA.085 85.1 85.2 85.3 85.4 Ninh_mbuo doqc_sou juangc buonv sou doqc, fiev nzengc juangc baengx batv. 3_PL study_book share CLF book read write consumed share CLF pencil pn.p v.cmp vt clf n vt vt vt vi vt clf n 85.1They used to study 85.2shared a book (to) read; 85.3(they) used to write every thing 85.4(by) share(ing) a pencil.

Page 272: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

246

FA.086 Lungh_muonz bueix juangc zung coux bueix. night lie_douwn share CLF bed sleep n vi vt clf n vi

At night (when they) splept, (they) shared a bed (to) sleep.

FA.087 87.1 87.2 Bueix gau Aengh^Doih nzuonx daaih. sleep DPCL Aeng-Doi return come vi part n.prp v vi 87.1When (he) went to sleep, 87.2Aeng-Doi would come back (to their bed room).

FA.088 88.1 Ninh ganh yiem wuov ndoqc haic 3sg only live there lonely very pn.p pn.emph vi pn.dem v.st adv 88.2 88.3 aengx nzuonx taux Aengh^Doih nyei biauv buatc m'sieqv_dorn, again return reach Aeng-Doi POSS house see woman adv vi v n.prp poss n v n 88.4 ih_zanc ninh oix longc Aengh^Doih zoux auv, now 3sg want marry Aeng-Doi be wife adv.tmp pn.p v.mod vt n.prp vi n 88.5 naaic gaax ninh maa mbuo hnangv_haaix_nor zoux? ask try_and_see 3sg mother PL how do v part pn.p n pn adv vt 88.1He was at home by himself (and) felt very lonely, 88.2(he) again visited Aeng-Doi’s house 88.3saw a lady; 88.4now he wanted to have Aeng-Doi as his wife, 88.5(so he) asked his parents what to do.

FA.089 Neiz maa mbuo dau Faam-Baeqv, his mother PL answer Faam-Bae’ contr n pl vt n.prp "Meih mv_dungx nzauh, 2sg NEG.CMD sad pn.p neg v.st njang_hnoi yie_mbuo m'daaih mingh tengx meih lorz." tomorrow 1_PL ourse go help 2sg look_for adv.tmp pn.p_pl adv vi vt pn.p vt

His parents answered Faam-Bae’, “You don't have to worry. Tomorrow we of course will go and help you to find (her).”

Page 273: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

247

FA.090 90.1 Njang_hnoi neiz maa mbuo za'gengh mingh lorz, tomorrow his mother PL really go look_for adv.tmp contr n pl adv vi vt 90.2 90.3 mingh taux youc Maaz_Jaa daaih gorngv ndaangc mi'aqv. go reach TOP Maa_house come speak first RSLT vi v top n.prp vi v n asp 90.1The next day, his mother and others [i.e. parents] indeed went to look for (her); 90.2as they arrived 90.3(they found that) Maajaa [i.e. Maa household] had arranged for a marriage with her first.

FA.091 91.1 91.2 Mv haih hnangv_haaix_nor zoux, Aengh^Doih gorngv, NEG be_possible_to how do Aeng-Doi speak neg aux adv vt n.prp v "Mh_mbuo aengx taux nziouv buo hnoi nv nyungc duqv mi'aqv. 2_PL again reach early three day this kind get RSLT pn.p adv v v.st numb n pn.dem n vt asp Ih_zanc Maaz_Jaa gorngv ziangh mi'aqv." now Maa_house engage finish RSLT adv.tmp n.prp v vi asp 91.1Being unable to do anything, 91.2Aeng-Doi said, “If you had revisited me three days earlier, you could have engaged me. Now Maajaa has already engaged me.”

FA.092 92.1 Aengh^Doih mv_haih_fungc_zoux, Aeng-Doi can’t_do_anything n.prp idm 92.2 92.3 betv ninh nyei lui-siaam fiev duqv zeiv fienx tear 3sg POSS jacket's_long-flaps write attain paper letter vt pn.p poss n vt vt n n 92.4 bun ninh nyei maa mbuo dorh nzuonx mingh. let 3sg POSS mother PL take return go vt pn.p poss n pl vt vi vi 92.1There was nothing Aeng-Doi could do about it, 92.2(but she) tore her jacket’s long flap 92.3(and) got (to) write a letter (on it) 92.4(and) let his parents take it home.

FA.093 "Bun Faam-Baeqv doqc oc." give Faam-Bae’ read PLT vt n.prp vt part

“Please let Faam-Bae’ read it.”

FA.094 94.1 94.2 Ninh nyei maa mbuo dorh nzuonx mingh gengh bun Faam-Baeqv. 3sg POSS mother PL take return go really give Faam-Bae’ pn.p poss n pl vt v v.dr adv vt n.prp 94.1His parents took it home 94.2(and) actually gave it to Faam-Bae’.

Page 274: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

248

FA.095 Faam-Baeqv koi daaih mangc. Faam-Bae’ open come look n.prp vt vi v

Faam-Bae’ opened (it and) read (it).

FA.096 Aengh^Doih gorngv mbuox ninh aqv_zuqc nyanc. Aeng-Doi speak tell 3sg must eat n.prp v vt pn.p aux vt

Aeng-Doi told him (in the letter that) he must eat (it).

FA.097 Ninh za'gengh nyanc nc zeiv fienx. 3sg really eat that paper letter pn.p adv vt pn.dem n n

He actually ate the letter.

FA.098 98.1 98.2 98.3 Nqaengx jaang sim baqv daic mi'aqv. choke_on throat needle pierce die RSLT vt n n vt vi asp 98.1(It) choked in his throat 98.2a needle pierced (his throat) 98.3(he) died.

FA.099 99.1 Mc zeiv fienx nqaengx daic mingh, that paper letter choke_on die go pn.dem n n vt vi v.asp 99.2 ninh nyei maa mbuo nzauh haic 3sg POSS mother PL sad very pn.p poss n pl v.st adv 99.3 99.4 mingh cingv fin-saeng daaih daapc_deic zangx Faam-Baeqv, go invite teacher come search_for_a_site_to_bury bury Faam-Bae’ vi vt n vi v.cmp vt n.prp 99.5 99.6 daapc duqv norm deic zangx jienv wuov domh jauv-hlen. tread_on get CLF site bury DUR there big road_side vt vt clf n vt asp pn.dem adj n 99.1(When) that letter choked (him and he) died, 99.2his parents were very sad; 99.3(they) went (and) invited a teacher [i.e. priest] 99.4searched for a site (to) bury Faam-Bae’, 99.5(he) treaded on (and) got a site 99.6(and they) buried (him) at the side of a big road.

FA.100 100.1 100.2 Aengx taux wuov norm hnoi Maaz_Jaa mbuo dorh jienv Aengh^Doih nzuonx aqv. further reach that CLF day Maa_house PL take SML.ACT Aeng-Doi return PFT adv v pn.dem clf n n.prp pl vt asp n.prp vi part.asp 100.1And then when it was the day of the wedding, 100.2Maajaa and the relatives were taking Aeng-Doi back home [as a bride].

Page 275: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

249

FA.101 101.1 101.2 Taux ninh_mbuo zoux cing-jaa nyei hnoi, ninh_mbuo fungx_cien nyei hnoi, reach 3_PL do wedding POSS day 3_PL send_a_daughter_to_marriage RELT day v pn.p vt n poss n pn.p vt+n relt n 101.3 ninh_mbuo fungx jienv Aengh^Doih mingh. 3_PL send SML.ACT Aeng-Doi go pn.p vt asp n.prp vi 101.1(It) reached the day (that) they would have a wedding, 101.2the day (when) they send a daughter to marriage, 101.3they went sending Aeng-Doi.

FA.102 102.1 102.2 Ninh_mbuo nyei siang-mbuangz jaaix haic, mv bun yangh jauv, 3_PL POSS bride precious very NEG let walk road pn.p poss n v.st adv neg vt vi n 102.3 zieqv jiuc daaih gaeng. weave sedan_chair come carry vt n v.asp vt 102.1Their bride was so precious 102.2(that they) would not let (her) walk on a road, 102.3(so they) wove a sedan chair (and) carried it.

FA.103 103.1 103.2 Gaeng jienv mingh taux wuov, Faam-Baeqv nyei zouv ga'hlen, carry SML.ACT go reach there Faam-Bae’ POSS grave beside vt asp vi v pn.loc n.prp poss n n 103.3 Aengh^Doih gorngv tov bun ninh njiec dangh ndau. Aeng-Doi speak ask_for let 3sg let_go_down awhile ground n.prp v vt vt pn.p vt adv n 103.1(As) they carried the sedan chair (and) reached there, 103.2Faam-Bae’s grave side, 103.3Aeng-Doi spoke out (and) asked (them to) let her come down to the ground for a while.

FA.104 104.1 104.2 Ninh_mbuo mv hiuv_duqv, za'gengh bun Aengh^Doih njiec ndau. 3_PL NEG know really let Aeng-Doi let_go_down ground pn.p neg vt adv vt n.prp vt n 104.1They didn’t know [why she asked] they) actually let Aeng-Doi come down. 104.2(but

Page 276: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

250 FA.105 105.1 105.2 Njiec ndau mingh tiux njiec ndiev maengx Faam-Baeqv nyei zouv descend ground go run go_down down_there direction Faam-Bae’ POSS grave vi n vi vi vi adv.loc n n.prp poss n 105.3 105.4 ndamc buo zaux-ndamc, ninh gorngv, stamp three foot-stamp 3sg speak vt numb n-vt pn.p v "Faam-Baeqv maaih lingh maaih singx aeqv, zouv-gaengh nqoi, Faam-Bae’ have power have spirit TOP grave-door open n.prp vt n vt n top n-n vi mv maaih lingh mv maaih singx nor aeqv, Maaz_Jaa dorh mingh aqv." NEG have power NEG have spirit TOP TOP Maa_house take go INCHOAT neg vt n neg vt n top part n.prp vt vi part.asp 105.1(She) went down to the ground, 105.2ran down to the lower place (of the road side) to the grave of Faam-Bae’, 105.3stamped her feet three times, 105.4 she said, “If Faam-Bae’ has power and spirit, let the grave door open. If (you) have no power, no spirit, Maajaa is going to take (me) away.”

FA.106 Yaac za'gengh zouv-gaengh gengh nqoi. and really grave-door really open conj adv n-n adv vi

And indeed, the grave door really opened.

FA.107 Aengh^Doih tiux bieqc Faam-Baeqv nyei zouv. Aeng-Doi run enter Faam-Bae’ POSS grave n.prp vi vi n.prp poss n

Aeng-Doi ran into Faam-Bae’s grave.

FA.108 108.1 Maaz_Jaa mbuo yietv hnyapv Maa_house PL as_soon_as catch_quickly n.prp pl adv vt 108.2 hnyapv duqv deix lui-siaam-jieqv hnangv. catch_quickly manag_to some jacket's_back_flap only vt vt adv n adv 108.1Maajaa’s people (tried to) catch (her) quickly, 108.3only managed to catch a bit of her jackets back flap.

FA.109 Maaih deix gorngv, have some speak vt adv v "Mv buatc mi'aqv, bieqc naaiv zouv mi'aqv." NEG see RSLT enter this grave RSLT neg v asp vi pn.dem n asp

Some said, “Disappeared, (she) has gone into the grave.”

Page 277: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

251 FA.110 Maaih deix gorngv, have some speak vt adv v "Mv sienx." NEG believe neg vt

Others said, “I don’t believe (it).”

FA.111 Maaz_Jaa mbuo gorngv, Maa_house PL speak n.prp pl v "Mv sienx mv nyungc aeqv, ndaam ciu wetv." NEG believe this kind TOP carry spade dig neg vt pn.dem n top vt n vt

People of Maajaa’s household said, “If you don’t believe, why not get a spade and dig the grave?”

FA.112 Ninh_mbuo gengh mingh wetv. 3_PL really go dig pn.p adv vi vt

They really went down (and) dug.

FA.113 113.1 113.2 Wetv dangh, wetv tong mc norm zouv aeqv, dig awhile dig break_open that CLF grave TOP vt adv vt v pn.dem clf n part 113.3

n.prp vt pn.p vi n

buatc i norm sopc_bang ndaix jienv faaux lungh mi'aqv. see two CLF butterfly fly SML.ACT go_up sky RSLT vt numb clf n vi asp vi n asp 113.1(They) dug awhile, 113.2when (they) dug through (and) reached inside the grave, 113.3(they) saw two butterflies fly up to the sky.

FA.114 114.1 Dongh naaic loz-hnoi Aengh^Doih m'daaih hiuv_duqv that_is that old-days Aeng-Doi from_the_beginning know idntf pn.dem n n.prp adv vt 114.2 114.3 ninh aqv_zuqc longc Faam-Baeqv mv_baac ninh mv mbuox nzengc Faam-Baeqv. 3sg must marry Faam-Bae’ but 3sg NEG tell consumed Faam-Bae’ pn.p aux vt n.prp conj pn.p neg vt vi n.prp 114.1That was exactly what Aeng-Doi had known in the past from the beginning, 114.2(that) she was destined to marry Faam-Bae’ 114.3but she did not tell him all.

FA.115 Faam-Baeqv funx ninh zoux a'nziaauc_doic. Faam-Bae’ assume 3sg be friend

Faam-Bae’ considered her as a friend.

Page 278: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

252 FA.116 116.1 Aengh^Doih mv funx Faam-Baeqv zoux a'nziaauc_doic, Aeng-Doi NEG consider Faam-Bae’ be friend n.prp neg vt n.prp vi n 116.2 funx nqa'haav ninh_mbuo m'daaih aqv_zuqc longc zoux i_hmuangv nyei. consider later 3_PL rightly must marry be husband_and_wife PRS.ST vt adv.tmp pn.p adv aux vt vi n asp 116.1Aeng-Doi did not consider Faam-Bae’ as a (mere) friend; 116.2(she) knew (that) later they would have to be husband and wife.

FA.117 117.1 117.2 Weic naaic Maaz_Jaa gorngv mv_baac ninh mv nzauh, because that Maa_house speak but 3sg NEG sad conj pn.dem n.prp v conj pn.p neg v.st 117.3 117.4 hiuv_duqv daic mingh ninh_mbuo duqv longc. know die go 3_PL attain marry vt vi vi pn.p vt vt 117.1Because of that, (though) Maajaa had engaged (with her first) 117.2but she was not worried; 117.3(she) knew (that after they) died 113.4they would be able to marry.

FA.118 118.1 Dongh naaic ninh fiev fienx mbuox Faam-Baeqv that_is that 3sg write letter tell Faam-Bae’ idntf pn.dem pn.p vt n vt n.prp 118.2 118.3 daic ndaangc zuov jienv ninh nqa'haav longc. die first wait_for DUR 3sg later marry vi n vt asp pn.p adv.tmp vt 118.1That was what she had written (to) tell Faam-Bae’ (to do) in the letter, 118.2(that is, he should) die first (and) wait for her 118.3(then) later (they) would marry.

FA.119 Nqa'haav za'gengh longc mi'aqv. later really marry RSLT adv.tmp adv vt asp

Later, (they) really married.

FA.120 120.1 Ih_zanc ninh_mbuo longc jienv nyei, now 3_PL marry DUR PRS.ST adv.tmp pn.p vt asp asp 120.2 120.3 Maaz_Jaa gorngv gau hnangv, m'daaih mv duqv longc. Maa_house speak DPCL only of_course NEG attain marry n.prp vt part adv adv neg vt vt 120.1Now they are married. 120.2Maajaa just completed the arrangements (of his marriage with Aeng-Doi); 120.3(but) of ourse (he) didn’t get to marry her.

Page 279: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

253

Story 6 (AS): “A Story of Aahan” Aa^han Gouv Yunh Zoih gorngv Yunzoi speak n.prp v

Yunzoi narrated.

AS.001 1.1 1.2 Maaih i_gox mienh za'gengh jomc nyei. have married_couple person really be_poor PRS.ST vt n n adv v.st asp 1.1There was a couple 1.2(who were) very poor.

AS.002 2.1 2.2 2.3 Jomc nyei, ninh mingh maaic gong lorz hnaangx nyanc. be_poor PRS.ST 3sg go sell work look_for rice eat v.st asp pn.p vi vt n vt n vt 2.1Poor they were, 2.2they went (and) did labor for other people 2.3looked for rice (to) eat [i.e. made their living.]

AS.003 3.1 I_gox ndaam jienv ndaamx married_couple carry SML.ACT carrying_pole n vt asp n 3.2 3.3 hnangv Janx-Kor^lormx nor hnaeng, hnaeng hnaeng hnaeng mingh. as-if Northern Thai as dangle dangle dangle dangle go adv n top v v v v vi 3.1The couple would carry baskets on their shoulders with poles 3.2just like Northern Thai would do, 3.3(they) went dangling, dangling and dangling.

AS.004 4.1 4.2 Mingh gau hmuangx aqv. go DPCL dark CHG-O-ST vi part adj part.asp 4.1As they went, 4.2(it) became dark.

AS.005 5.1 5.2 Hmuangx aqv, taux wuov ndaamv-jauv, dark CHG-O-ST reach there half-way adj part.asp v pn.dem n 5.3 5.4 hnangv yie nyei lamz nor zoux daaih lauh haic aqv. like 1SG POSS granary as make come long-period very CHG-O-ST adv pn.p poss n top vt v.asp v.st adv part.asp 5.1It became dark 5.2(when they) were on the way, 5.3(then there was) a granary just like mine, 5.4it has been built for long time.

Page 280: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

254 AS.006 6.1 6.2 Taux wuov hmuangx aqv. reach there dark CHG-O-ST v pn.dem v.st part.asp 6.1(When they) arrive there 6.2(it) became dark.

AS.007 7.1 7.2 Aav, mingh haaix yaac mv mingh taux bieqc lamz bueix. Ah go what and_yet NEG go reach enter granary sleep intj vi q conj neg vi v vi n vi 7.1Ah, wherever (they) went 7.2(they) have not gone as far as to enter a granary (to) sleep.

AS.008 Nyorng bieqc lamz mingh. creep enter granary go vi vi n vi

They crept into the granary.

AS.009 Jiepv njiec dorn. bear give_birth_to cub n vt n

A bear had given birth to a cub (there).

AS.010 10.1 10.2 Jiepv za'gengh tiux cuotv oix ngaatc ninh_mbuo. bear really run go_out want-to bite 3_PL n adv vi v v.mod vt pn.p 10.1The bear really ran out 10.2(it) wanted (to) bite the couple.

AS.011 Ninh nyei nqox biaux! biaux. 3sg POSS husband flee flee pn.p poss n vt vt

Her husband ran away.

AS.012 12.1 12.2 Ninh nyei auv maiv_haih_fungc_zoux, maiv henv. 3sg POSS wife not_be_able_to_anything NEG strong pn.p poss n idm neg v.st 12.1There was nothing his wife could do, 12.3(she) was not strong.

AS.013 Ndaam jienv ga'naaiv. carry DUR thing vt asp n

(She) was carrying things.

Page 281: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

255 AS.014 14.1 Hnungh hnaeng! hnungh hnaeng ONOMA dangle ONOMA dangle onoma v onoma v 14.2 ziouc zorqv ninh nyei ndaamx hnangv nv nor, so_then take 3sg POSS carrying_pole like this as seq.mk vt pn.p poss n adv pn.dem top 14.3 se ninh nyei biaav-ndaauv nyei, that_is 3sg POSS stick-long PRS.ST copl pn.p poss n-v.st asp 14.4 14.5 14.6 bungx njiec daaih ndamc mv bung ndutv mingh, release let_go_down come stamp this end come_apart go vt vt v.dr vt pn.dem n vi v.asp 14.7 14.8 14.9 ndamc mv bung ndutv mingh nzopv jiepv. stamp this end come_apart go pierce bear vt pn.den vi vi vt n 14.1(She) swung (it) dangling, dangling, 14.2then (she) took her carrying pole like this, 14.3that is, her pole was long, 14.4(she) released (it and) let (it) go down 14.5(and) stamped on this end 14.6(and it) came off; 14.7(she) stamped on this end 14.8let it drop (on the bear) 14.9(and it) pierced the bear.

AS.015 15.1 15.2 Nzopv jiepv, nzopv gau pierce bear pierce DPCL vt n vt part 15.3 15.4 nzopv! nzopv nzuiz gau jiepv nyei nzuih baengx, daic mi'aqv. pierce pierce shove DPCL bear POSS mouth die RSLT vt vt v part n possen vi asp 15.1(She) stabbed the bear('s mouth) 15.2as (she) stabbed, 15.3(she) shoved the pole really hard 15.4(and the bear) died.

AS.016 Wuov_nzunc_hnoi i_gox caangh_laangh aqv. this-time married_couple discuss PFT adv.tmp n v.cmp asp

Then the couple discussed.

AS.017 17.1 17.2 Ninh oix paaix nc dauh jiepv daaih dorh mingh bun Am^pev. 3sg want_to divide that CLF bear come take go give Ampe pn.p v.mod vt pn.dem clf n v.asp vt vi vt n.prp 17.1He wanted to divide the bear (to) take it 17.2(and) give to the district township official.

Page 282: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

256

AS.018 Ninh_mbuo i_gox gorngv, 3_PL married_couple speak pn.p n v "Aav!, hnangv_haaix_nor zoux. Mbuo nzuqc mv maaih, congx mv maaih. ah how do 1pl knife NEG have gun no have intj adv vt pn.p n neg vt n adv vt

They said, "Ah, what shall we do? We don't have a knife or a gun."

AS.019 19.1 19.2 Gorngv meih, ninh nyei auv mborqv daaih nae... speak 2sg 3sg POSS wife hit come SUSPD v pn.p pn.p poss n vt vi part 19.3 ninh nyei m'jangc_dorn gauh nyaiv. 3sg POSS husband more feel_ashamed pn.p poss n cmpr vi 19.1If you say…19.2(if) his wife beat (the bear to death)…19.3her husband would be ashamed more.

AS.020 20.1 Gorngv ninh nyei nqox mborqv nae... speak 3sg POSS husband beat SUSPD v pn.p pos n vt part 20.2 20.3 yaac mv zeiz neiz nqox mborqv, ninh auv mborqv daaih. TOP HEG be_right her husband beat 3sg wife beat come part neg vi contr n vt pn.p n vt v.asp 20.1(If) say her husband beat…20.2that’s not (that) the husband beat; 20.3(the truth is) his wife beat.

AS.021 Hnangv_haaix_nor zoux? how be adv vi

What shall we do?

AS.022 Mv_baac ninh auv gorngv, but 3sg wife speak conj pn.p n v "Aa, yie m'sieqv_dorn, yie maiv nongc mengh aqv. Ah 1sg woman 1sg NEG want fame I_am_sure intj pn.p n pn.p neg vt n part.mod Bun mengh meih gorngv meih mborqv daaih." give fame 2sg speak 2sg hit come vt n pn.p v pn.p vt v.asp

But his wife said, “Ah, I am a woman; I’m sure I don’t want an honor. I give it to you, you tell them you beat it.”

Page 283: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

257 AS.023 23.1 23.2 Gengh wuov_nzunc_hnoi paaix orv daaih dorh mingh bun Am^pev. really this_time divide_up meat come take go give Ampe adv adv.tmp vt n v.asp vt vi vt n.prp 23.1Really this time (they) divided up the meat 23.2(and) took it (to) give to the governer.

AS.024 24.1 24.2 Taux wuov Am^pev naaic gaax ninh_mbuo reach there Ampe ask try_and_see 3_PL v pn.dem n.prp v part pn.p “Mv maaih congx, mv maaih nzuqc hnangv_haaix_nor duqv jiepv?" NEG have gun NEG have knife how get bear neg vt n neg vt n adv vt n 24.1(They) reached there, 24.2the governer asked them, “You don’t have a gun, don’t have a knife. How could you get the bear?”

AS.025 "Oh yie_mbuo mborqv daaih. Caa jienv mborqv nyei mborqv. Oh 1_PL hit come catch SML.ACT hit ADVBLZR hit intj pn.p vt v.asp vt asp vt part vt Dorh biaav nzuiz nyei nzuiz nzuih_baengx daic mingh." take stick shove ADVBLZR shove mouth die go vt n v part v n vi v.asp

“Oh, we beat (it to death). (We) caught it (and) beat and beat. (We) took a stick and shoved (it) into its mouth repeatedly (and it) died.”

AS.026 "U..., Meih za'gengh henv haic." Wow! 2sg really strong very intj pn.p adv v.st adv

“Wow! You are really very strong!”

AS.027 Heuc naaic dauh mienh zoux Aa^Han_Mix. name that CLF person be Bear_the_Brave vt pn.dem clf n vi n.prp

(The governer) entitled that man (to) be Bear the Brave.

AS.028 Wuov_nzunc_hnoi duqv yietc norm mbuox hlo deix aqv. this-time get one CLF name big some PFT adv.tmp vt numb clf n v.st adv part.asp

At this occasion (he) got one considerably big name.

Page 284: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

258 AS.029 29.1 29.2 29.3 Da'nyeic hnoi hnangv wuov ndiev ndaaih naaic, mbuo geh jakv second day like that under river there 1pl ride boat n n adv pn.dem adv.loc n pn.dem pn.p vt n 29.4 29.5 maaih douh_taanh yietc nyungc, hlo nyei. have big_lizard one kind big PRS.ST vt n numb n v.st asp 29.1Then another day, 29.2like the river down there 29.3(where) we ride a boat, 29.4there was a kind of crocodile, 29.5(which) was big.

AS.030 30.1 30.2 Mienh geh jakv jiex, nzaeng nzangv jiex, person ride boat pass paddle boat pass n vt n vi v n vi 30.3 30.4 30.5 30.6 ninh cuotv daaih ngatv nzangv, mbienv ndortv duqv mienh nyanc. 3sg come_out come press_down_on_hard boat overturn fall get person eat pn.p v vi vt n v vi vt n vt 30.1People rode a motorboat to pass (the river), ed to pass, t [the crocodile] came outpressed down on the boat; he boat) turned over (and people) fell off, le) got people (to) eat.

30.2paddl 30.3i 30.4 30.5(t 30.6(crocodi

AS.031 Janx aengx mv haih fungc zoux. non_Iu-Mienh again NEG be_able_to how-come do n adv neg aux adv vt

There was nothing the Northern Thai people could do about it again.

AS.032 Aav mv noic duqv aqv. Ah NEG manage can PFT intj neg vi v.aux part.asp

Ah, (we) can't manage to do anything.

AS.033 Aqv_zuqc mingh lorz naaic dauh Aa^Han_Mix daaih zorqv hnangv. must go look_for that CLF Bear_the_Brave come take no_other_way aux vi vt pn.dem clf n.prp v.dr vt adv

(We) must go find that Bear the Brave and bring him here to let him catch (the crocodile).

AS.034 Za'gengh mingh lorz. really go look_for adv vi vt

(They) really went to look for (him).

AS.035 Lorz buatc nc dauh Aa^Han_Mix. look_for see that CLF Bear_the_Brave vt v pn.dem clf n.prp

(They) found the Bear the Brave.

Page 285: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

259 AS.036 36.1 36.2 Mingh taux aeqv, aengx nzaeng jienv nzangv jiex. go reach TOP again paddle SML.ACT boat pass vi v top adv v asp n vi 36.1As (they) reach there, 36.2(they) again passed (the river) paddling a boat.

AS.037 37.1 Jiex taux ndaaih mbu'ndongx hnangv, pass reach river middle only vi v n n adv 37.2 37.3 37.4 douh_taanh za'gengh hmatv jieqv nyei, jangv nyei cuotv daaih. big_lizard really absolutely black PRS.ST wide PRS.ST come_out come n adv part v.st asp v.st asp vi vi.dr 37.1As soon as (the boat) passed (to) reach the middle of the river, 37.2the crocodile, (which was) really absolutely black 37.3(and) wide, 37.4came out.

AS.038 38.1 38.2 38.3 Ninh gamh_nziex haic sin zinx nyanh nv nyanh nyei. 3sg fear very body tremble jerk intensly jerk PRS.ST pn.p v.comp adv n vi vi intj vi asp 38.1He was very afraid, 38.2(his) body trembled, 38.3jerked intensely.

AS.039 Nqaiv cuotv. feces come_out n vi

Feces came out.

AS.040 40.1 40.2 40.3 Wuov_deix, yiem nzangv gu'nyuoz wuov_deix zorqv caang nzopv, those_people be_in boat inside those_people take spear pierce pn.dem vi n n pn.dem vt n vt 40.4 40.5 zorqv congx buonv, buonv daic mingh. take gun shoot shoot die go vt n vt vt vi v.asp 40.1Those people, 40.2those who were inside the boat, 40.3took spears (and) pierced (it), 40.4took guns (and) shot, 40.5shot (it and it) died.

Page 286: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

260 AS.041 "Meih_mbuo mv_dungx gaengh baqv saa. 2_PL NEG.CMD really pierce SFT.CMD pn.p neg adv vt part

“Don’t you yet pierce it.”

Yie jaang-jaang bungx nqaiv bun ninh nyanc. 1sg on_the_process release feces let 3sg eat pn.p adv vt n vt pn.p vt

“I’m just releasing feces to let it eat.”

Yie oix zorqv nangh. 1sg want_to take be_alive pn.p v.mod vt v.st

“I want to catch it alive.”

Oix zorqv ndoh jienv dorh mingh fongc guen jien. want_to take bind DUR take go show_gratitude province officer v.mod vt vt asp vt vi vt n n

“I want to take it bound to show gratitude to the provincial officer.”

Meih_mbuo fungc zorqv baqv daic mingh?" 2_PL how_come take pierce die go pn.p adv vt vt vi v.dr

“How come you guys pierced it to death?”

AS.042 Hnangv nc nor gorngv. like that as speak adv pn.dem top v

He said like that.

AS.043 Hmz, wuov_nzunc_hnoi aengx mingh aqv. Yes! This_time again go CHG-O-ST intj adv.tmp adv vi part.asp

Yes, this time (he)’s got to go again.

AS.044 Dorh jienv nv dauh douh taanh mingh Am^pev wuov. take SML.ACT this CLF big_lizard go Ampe place vt asp pn.dem clf n vi n.prp pn.dem

(He) taking this crocodile (and) went to the provincial office.

AS.045 Aengx orn ninh zoux Aa^Han_Suv^Tiqc. further_more appoint 3sg be Suthi the Brave adv vt pn.p vi n.prp

(The governor) appointed him again (to) be Suthi the Brave.

Page 287: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

261 AS.046 46.1 46.2 Duqv i norm mbuox Aa^Han_Mix, Aa^Han_Suv^Tiqc duqv daaih aqv. get two CLF name Bear_the_Brave Suthi_the_Brave get come PFT vt numb clf n n.prp n.prp vt v.asp part.asp 46.1(He) has got two names, Bear the Brave (and) Suthi the Brave; 46.2(he) got (them).

AS.047 47.1 47.2 47.3 Da'nyeic hnoi youc maaih diuh domh naang, hlo nyei, ndortv wuov laangz mienh, second day also have CLF big snake big PRS.ST fall that village person n n seq.mk vt clf adj n v.st asp vi pn.dem n n yietc norm mungv, yietc norm zingh nyei mienh nyei wuom-kuotv. one CLF town one CLF city POSS people POSS water_source numb clf n nmb clf n poss n poss n 47.1Then another day, 47.2there was a gigantic snake, (which) was big 47.3fell into a city water source for the residents of the whole town, the whole city.

AS.048 Za'gengh mv duqv wuom nyanc aqv. really NEG get water eat CHG-O-ST adv neg vt n vt part.asp

Indeed, (people) became unable to get the drinking water. (i.e. the water became undrinkable.)

AS.049 Naang za'gengh hlo haic bieqc gu'nyuoz mingh. snake really big very enter inside go n adv v.st adv vi n v.asp

The really very big snake fell into (the well).

AS.050 Aengx caangh_laangh naaic laangz mienh, yietc norm fouv-zingh nyei mienh. again discuss that village people one CLF capital_city POSS people adv v pn.dem n n numb clf n poss n

And again the people of that village, of the whole capital discussed.

AS.051 Hm, maiv haih fungc zoux. well... NEG be_able_to how-come do intj neg aux adv vt

Well, there was nothing (they) ccould do about it.

AS.052 Aqv_zuqc mingh lorz Aa^Han_Mix, Aa^Han_Suv^Tiqc daaih zorqv hnangv. must go look_for Bear_the_Brave Suthi_the_Brave come take no_other_way aux vi vt n.prp n.prp v.dr vt adv

There is no other way than (we) must go look for Bear the Brave, Suthi the Brave (and ask him to) catch (the snake).

Page 288: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

262

AS.053 53.1 53.2 53.3 Mingh lorz ninh mingh taux wuov norm laangz buangh zuqc ninh go look_for 3sg go reach that CLF village meet GOAL 3sg vi vt pn.p vi v pn.dem clf n vt part pn.p 53.4 53.5 hnangv tengx meih weih laatc. like help 2sg wall_in fence adv vt pn.p vt n 53.1(They) went (to) look for him, 53.2went (and) reached that village, 53.3found him (there), 53.4(he was) like help(ing) you 53.5(to) make a fence.

AS.054 54.1 54.2 54.3 Yiem wuov weih laatc cipv jienv yie wuov nyungc nzuqc_paiv be_there there wall_in fence stick DUR 1sg that kind scabbard vi pn.dem vt n vt asp pn.p pn.dem n n 54.4 heuc ninh, call 3sg v pn.p "Aa yie mingh mangc gaax wuov nyungc." Ah 1sg go look try_and_see that kind intj pn.p vi v part pn.dem n 54.1Being there, 54.2(he was) build(ing) in a fence 54.3sticking a scabbard like mine (in his back part of the belt); 54.4(and they) called him. (He answered) “Ah, I will go and see that kind (of thing).”

AS.055 55.1 Ndau juiz deix land be_steep some n v.st adv 55.2 55.3 eix_leiz wuom-kuotv yiem ndiev maengx, ninh yiem jiex maengx. meaning well exist down_there side 3sg be_in up_there side n n vi adv.loc n pn.p vi n n 55.1The land was steep; 55.2which means the well was on the lower side, 55.3he was on the upper side.

AS.056 56.1 56.2 56.3 Mingh taux yietv ngamv njiec, nzuqc_paiv yietv donx naaiv go reach as_soon_as squat go_down scabbard as_soon_as hit this vi v adv vi vi n adv vt pn.dem 56.4 ninh ziang wuov biu gu'nyuoz mi'aqv. 3sg just that jump inside RSLT pn.p adv pn.dem vi n asp 56.1(They) went (and) reached (the well), 56.2as soon as (he) squatted down 56.3his scabbard hit this [i.e. his back], 56.4he just fell down there inside (the well).

Page 289: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

263 AS.057 U Janx za'gengh ceng ninh haic. Wow! non_Iu-Mienh really praise 3sg very intj n adv vt pn.p adv

Wow! The Northern Thai really praised him very much.

AS.058 58.1 58.2 58.3 Mangc gaax naaic daaih taux ziang_naaic biu gu'nyuoz mi'aqv. look see that come reach just_then jump inside RSLT v part pn.dem vi v adv vi n asp 58.1Look at him, 58.2(he) arrived here 58.3(he) immediately jumped into the well.

AS.059 Zorqv naang-jaang nanv jienv. take snake-neck grasp DUR vt n-n vt asp

(He) took the snake by the neck, grabbing (it)

AS.060 60.1 60.2 60.3 60.4 Wuov_deix zorqv caang nzopv nzopv nzopv nzopv daic mi'aqv. those_people take spear pierce pierce pierce pierce die RSLT pn.dem vt n vt vt vt vt vi asp 60.1Those people 60.2took spears 60.3pierced (and) pierced repeatedly (the snake) 60.4(and it) died.

AS.061 Cingx_daaih aengx orn yietc nzunc ninh zoux Aa^Han_Nguv. therefore again appoint one time 3sg be Snake_the_Brave

adv v.st adv asp pn.dem n

conj adv vt numb n pn.p vi n.prp

Therefore, (the governer) again gave him another title (to) be Snake the Brave.

AS.062 Za'gengh longx haic aqv wuov nzunc. really be_good very CH-O-ST that time

This time it was really good.

AS.063 Za'gengh duqv nyaanh duqv lui-houx. really get money get clothes adv vt n vt n

(He) actually got money and clothes.

AS.064 Guien (Jien) za'gengh a'hneiv haic ninh. officer official really be_happy very 3sg n n adv v.st adv pn.p

The governer was really happy about him.

AS.065 65.1 65.2 Aengx taux da'nyeic hnoi, aengx maaih nda'maauh faanv. and_then reach second day further have tiger be_disturbing adv v n n adv vt n v.st 65.1And then another day, 65.2again there was a disturbing tiger.

Page 290: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

264 AS.066 66.1 66.2 66.3 66.4 Maaih laanh mienh nda'maauh daaih ngaatc gau, jai ngaatc gau, dungz ngaatc gau, have CLF person tiger come bite DPCL chicken bite DPCL pig bite as vt clf n n vi vt part n vt part n vt top 66.5 mienh oix nzengc aqv. people want consumed INCHOAT n vt v part.asp 66.1There was a person 66.2the tiger came (and) bit (him); 66.3then (it) bit chickens, 66.4then (it) bit pigs, 66.5(it) wanted all people!

AS.067

v intj idm

Wuov hnoi lungh_aanx ziouc haiz wuov_jiex laangz mienh.

Caangh_laangh aa, maiv_haih_fungc _zoux. discuss Ah can’t_do_anything

(They) discussed (saying) Ah, there’s nothing (we) can do (about it).

AS.068 68.1 Aqv_zuqc mingh lorz Aa^Han_Mix, Aa^Han_Nguv, must go look_for Bear_the_Brave Snake_the_Brave aux vi vt n.prp n.prp 68.2 Aa^Han_Suv^Tiqc daaih zorqv hnangv. Suthi_the_Brave come take no_other_way n.prp vi vt adv 68.1(We) must go look for Bear the Brave, Snake the Brave, Suthi the Brave 68.2(and have him) catch (the tiger), there is no other way.

AS.069 Aengx mingh lorz buatc ninh. and_then go look_for see 3sg adv vi vt v pn.p

And then (they) went (and) found him.

AS.070 70.1 Ninh mingh wuov ndiev 3sg go there down_there pn.p vi pn.dem adv.loc 70.2 70.3 hnangv laangz yiem naaiv, ninh mingh wuov ndiev ndoqv wuov. like village exist here 3sg go there down_there stream there adv n vi pn.dem pn.p vi pn.dem adv.loc n pn.dem 70.1He went down there, 70.2like in this village, 70.3he went down there at the stream.

AS.071

that day after_noon so_then hear upper_side village people pn.dem n adv.tmp seq.mk vt n n n

In the afternoon of that day, (they) heard (a sound of) the upper village people.

Page 291: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

265 AS.072 72.1 72.2 72.3 Mbiouh mv mbiouh nyei zunc jienv nda'maauh daaih aqv. moisy intensly noisy ADVBLZR chase SML.ACT tiger come PFT v.st intj v.st part vt asp n vi part.asp 72.1(They) made a lot of noise 72.2chasing the tiger 72.3(and they) came.

AS.073 73.1 73.2 73.3 Zunc jienv nda'maauh taux wuov ninh faaux ndiangx mingh yiem gu'nguaaic. chase SML.ACT tiger reach there 3sg go_up tree go be_there above vt asp n v pn.dem pn.p vi n v.dr vi adv.loc 73.1(When they) came chasing the tiger (and) reached there, 73.2he climbed up a tree 73.3(and) stayed above.

AS.074 Ga'ndiev ndiangx njongz nyei. underneath tree be_hollow PRS.ST adv.loc n v.st asp

Underneath the tree was hollow.

AS.075 75.1 75.2 Nda'maauh suoh suoh nyei taux wuov youc maiv bieqc. tiger MIM MIM ADVBLZR reach there but NEG enter n v.st v.st part v pn.dem seq.mk neg vi 75.1The tiger arrived there slinking along 75.2but did not come in (the hole).

AS.076 76.1 76.2 Haaix ndau maiv mingh youc bieqc wuov ndiangx-nqunx mingh. what ground NEG go also enter there tree-hole go q n neg vi seq.mk vi pn.dem n-n v.dr 76.1(It) did not go anywhere 76.2but entered that hole of the tree.

AS.077 Dueiv yiem ga'nyiec. tail be_in outside n vi n

(Its) tail was outside.

AS.078 Dueiv ndaauv nyei yiem ga'nyiec. tail long PRS.ST exist outside n adj asp vi n

The tail was long (and) it stayed outside.

AS.079 79.1 79.2 Wuov_deix huei, huei nyei zunc jienv daaih bungx jienv congx daaih. those_people MIM MIM ADVBLZR chase SML.ACT come release SML.ACT gun come pn.dem v.st v.st part vt asp vi vt asp n vi 79.1Those people came chasing (it) with a loud noise, 79.2came shooting guns.

Page 292: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

266 AS.080 80.1 80.2 Wuov_deix daaih fatv deix ninh njiec aiv deix. those_people come near some 3sg descend low some pn.dem vi v.st adv pn.p vi v.st adv 80.1(As) they come a little closer, 80.2he came down a little.

AS.081 81.1 81.2 Wuov_deix daaih fatv deix ninh njiec aiv deix. those_people come near some 3sg descend be_low some pn.dem vi v.st adv pn.p vi v.st adv 81.1(As) they come a little closer, 81.2he came down a little.

AS.082 82.1 82.2 Hnangv wuov_deix taux wuov ndau ninh tiux njiec daaih as_if those_people reach that place 3sg run go_down come adv pn.dem v pn.dem n pn.p vi vi vi 82.3 baeng jienv nda'maauh nyei dueiv. pull SML.ACT tiger POSS tail

yell_at those_people

85.1 85.2 85.3

vt asp n poss n 82.1It looked like (that) when those people arrived there 82.2he ran down 82.3pulling the tiger’s tail.

AS.083 Nauc wuov_deix,

vt pn.dem "Nqongh daaih nqongh daaih yie zorqv duqv dauh nda'maauh aqv. hurry come hurry come 1sg take can CLF tiger PFT v vi v vi pn.p vt v.aux clf n part.asp

(He) shouted to them, “Come quickly! I’ve just managed to catch the tiger.”

AS.084 Wuov_deix daaih taux buonv nyei buonv. those_people come reach shoot ADVBLZR shoot pn.dem vi v vt part vt

Those people arrived, shooting repeatedly.

AS.085

Baqv nyei baqv, baqv daic mingh. pierce ADVBLZR pierce pierce die go vt relative vt vt vi v.asp 85.1(They) pierced and pierced (the tiger) 85.2pierced 85.3(it) dead.

AS.086 Aengx dorh mingh bun Guien mangc. and_then take go let officer look adv vt vi vt n vt

(He) took it back (to) let the governor look at (it).

Page 293: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

267 AS.087 Guien aengx heuc ninh zoux Aa^Han_Siex, Aa^Han_Siex. officer further_more name 3sg be Tiger_the_Brave Tiger_the_Brave n adv vt pn.p vi n.prp n.prp

The governer furthermore called him Tiger the Brave, Tiger the Brave.

AS.088 Aa^Han_Siex duqv biei nzunc aqv. Tiger_the_Brave get four time PFT

Nqa'haav aengx deic-bung aengx faanv hnangv ih_zanc naaiv mborqv jaax. Later again country again unsettled like now here hit REC.ACT adv.tmp adv n adv v.st adv adv.tmp pn.dem vt adv

Later, the country became unsettled again; just as people here fight each other nowadays.

n.prp vt numb n part.asp

Tiger the Brave got (awarded) four times.

AS.089 Za'gengh henv. really strong adv adj

(He) was really strong.

AS.090

AS.091 91.1 Mborqv jaax mienh camv, baeng camv, hit REC.ACT people many soldier really many vt adv n adj n adv adj 91.2 haaix_deix yaac mborqv waaic nzengc. wherever also hit be_destroyed consumed n conj vt v.st v 91.1(There) were many fighters, many soldiers; 91.2(they) destroyed everywhere.

AS.092 Maaih naaiv dauh Guien yiem naaic yietc norm zingh. have this CLF officer live there one CLF city vt pn.dem clf n vi pn.dem numb clf n

There was this officer living in that city.

AS.093 Wuov_deix baeng aengx oix daaih mborqv. those soldier again want_to come hit pn.dem n adv v.mod vi vt

Further those soldiers wanted to come (to) fight.

AS.094 Maiv_haih_fungc_zoux. can’t_do_anything idm

There was nothing (he) could do about it.

Page 294: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

268 AS.095 95.1 Aa heuc Aa^Han_Mix Aa^Han_Suv^Tiqc Aa^Han_Nguv Aa^Han_Siex mingh Ah call Bear_the_Brave Suthi_the_Brave Snake_the_Brave Tiger_the_Brave go intj v n.prp n.prp n.prp n.prp vi 95.2 zorqv hnangv. take no_other_way vt adv 95.1Ah, (we only) order Bear the Brave, Suthi the Brave, Snake the Brave, Tiger the Brave (to) go 95.2(and) catch (the soldiers), there is no other way.

AS.096 96.1 96.2 96.3 Mv zeiz nor aeqv, mv hingh aeqv, mbuo naaiv mv maaih faatv. NEG be as TOP NEG win TOP 1pl here NEG have magic neg vi top top neg vt top pn.p pn.dem neg vt n 96.1If we don’t, 96.2(we) can’t win and then 96.3we do not have magic here.

AS.097 97.1 97.2 Mingh wuov ndiev! gorngv ziangx go there down_there speak finish vi pn.dem adv.loc v vi 97.3 wuov ngaanc domh zorng yietc norm ndaamh. that opposite_side big mountain one CLF slope pn.dem n adj n numb clf n 97.1(The governor’s people) went far down there 97.2(and) finished talk (with the brave man); 97.3on the slope on the opposite side of the great mountain.

AS.098 98.1 98.2 Yiem wuov ninh faaux taux wuov aeqv, maaih yietc nyungc nyaaix-weih. be_in there 3sg go_up reach there TOP have one kind fern vi pn.dem pn.p vi v pn.dem part vt numb n n 98.1From there as he went up there, 98.2there was one kind of fern.

AS.099 Ndiangx yiem naaiv mbu'ndongx yietc nyungc nyaaix-weih gormx mingh. tree exist here middle one kind fern traverse_completely go n vi pn.dem n numb n n vt v.dr

Among the trees here a kind of fern traversed completely.

AS.100 100.1 100.2 Ninh faaux mingh zueiz jienv wuov. 3sg go_up go sit DUR there pn.p vi vi vi asp pn.dem 100.1He climbed up 100.2sitting there.

AS.101 Baeng za'gengh daaih camv haic. soldier really come many very n adv vi v.st adv

The numerous soldiers really came.

Page 295: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

269 AS.102 Ziex cin mbu'ziex cin mbu'ziex waanc yaac maiv hiuv. many thousand how_many thousand how_many ten_thousand TOP NEG know adj numb q numb q numb part neg v

Many thousand, (or we) don’t know if it was how many thousand, how many ten-thousand.

AS.103 Daaih taux youc haaix ndau maiv hitv. come reach also what ground NEG rest vi v seq.mk q n neg vi

Didn’t know where they stayed either.

AS.104 104.1 104.2 Mingh hitv jienv ndiangx-gorn-hlen, camv! nyei hitv jienv wuov. go rest DUR tree-base-side many ADVBLZR rest DUR there vi vi asp n-n-n v.st part vi asp pn.dem 104.1(They) went (and) were resting at the tree base, 104.2numerously resting there.

AS.105 Yiem wuov caangh_laangh. be_there there discuss vi pn.dem v

They were there (and they) discussed (i.e. they were discussing there).

AS.106 "Uv! Gamh_nziex haic, wuov laangz mienh yie_mbuo se mv gamh_nziex, Oh! fear very that village people 1_PL that_is NEG fear intj v.cmp adv pn.dem n n pn.p copl neg v.cmp wuov norm fouv-zingh nyei mienh yie_mbuo maiv gamh_nziex. that CLF capital_city POSS people 1_PL NEG pn.dem clf n poss n pn.p neg v.cmp Gamh_nziex Aa^Han_Nguv Aa^Han_Mix hnangv." fear Snake_the_Brave Bear_the_Brave only v.cmp n.prp n.prp adv

“Oh, it’s scary! We are not afraid of the village people, we are not afraid of the people of the city. We only fear Snake the Brave, Bear the Brave.”

AS.107 Ninh yiem gu'nguaaic ninh haiz nyei. 3sg be_in above 3sg hear PRS.ST pn.p vi adv.loc pn.p vt asp

He could hear (the soldiers’ conversation) staying above.

AS.108 “Uv! se_gorngv duqv Aa^Han_Mix Aa^Han_Nguv daic mingh nor aeqv, Oh! if get Bear_the_Brave Snake_the_Brave die go as TOP intj conj vt n.prp n.prp vi vi top top yietc laangz mienh maiv gaux yie_mbuo nyanc lungh ndorm.” whole village people NEG enough 1_PL eat sky morning adj n n neg v.st pn.p vt n n

“Oh! If (we) get Bear the Brave, Snake the Brave killed, the people of the whole village are not enough for us to eat for breakfast.”

Page 296: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

270 AS.109 109.1 109.2 109.3 Ninh gamh_nziex haic, gamh_nziex haic sin zinx qam jienv ndiangx. 3sg fear very fear very body tremble hold DUR tree pn.p v.cmp adv v.cmp adv n vi vt asp n 109.1He was terrified very much, errified so much (that his) body trembled, clinging to a tree.

112.1 112.2

109.3t

AS.110 Sin zinx za'gengh nyanh nyanh nyanh nyanh nyanh nyanh nyanh nyanh nyanh deix. body tremble really jerk jerk jerk jerk jerk jerk jerk jerk jerk some n vi adv vi vi vi vi vi vi vi vi vi adv

(His) body trembled much continuously.

AS.111 111.1 111.2 111.3 Caang wuov_deix nyei baeng caang_congx baeqc nzengc yiem wuov zoux nyanc. spear those RELT soldier gun white consumed be_there there make eat n pn.dem relt n n v.st v vi pn.dem vt vt 111.1(The area was) all white (with) the soldiers (holding) spears (and) guns, 111.2(they) were there 111.3cook (ing) food.

AS.112

Ciev jienv geh daaih siqv nyei. direct DUR ride come red PRS.ST vt asp vt vi v.st asp 112.1(Others) came driving horses 112.2(they) were red.

AS.113 113.1 113.2 Ninh aeqv, sin zinx haic aav lov! 3sg TOP body tremble very SURPRISE pn.p top n vi adv part.f 113.1As for him, 113.2his body trembled terribly.

AS.114 114.1 114.2 114.3 Naaiv aeqv, mbuo yiem naaiv, gorv_ laaic ninh_mbuo maiv daaih. here TOP 1pl live here guess 3_PL NEG come pn.dem top pn.p vi pn.dem v.cmp pn.p neg vi 114.1As for this place, 114.2we are here, 114.3(we) suspect (that) they won’t come here.

AS.115 115.1 115.2 115.3 Za'gengh daaih naaiv aeqv, congx mv maaih, nzuqc mv maaih. really come here TOP gun NEG have sword NEG have adv vi pn.dem top n neg vt n neg vt 115.1If (they) really come here, 115.2(we) don’t have guns, 115.3don’t have swords.

AS.116 Fungc zoux. adv v

Can't do anything.

Page 297: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

271 AS.117 117.1 117.2 117.3 Aav_dangh ninh_mbuo ciepv buatc ndortv aqv, za'gengh gamh_nziex haic. in_a_short_while 3_PL look_at see fall PFT really fear very adv.tm pn.p vt v vi part.asp adv vt adv 117.1Suddenly they saw 117.2(the brave man) fall down, 117.3(they) were really terrified very much.

AS.118 Ninh oix daic haic. 3sg want_to die very pn.p v.mod vi adv

He was scared very much.

AS.119 Sin zinx nyanh gau. body tremble jerk very n vi vi adv

His body trembled very much.

AS.120 120.1

cry that ground come

n.prp n.prp n.prp idntf pn.p part.mod

Nyaaix-weih nquatv fern crack_down n vi 120.2 ziouc suz! suz nyei so_then sound_of_spurting_air sound_of_spurting_air ADVBLZR seq.mk onoma onoma part fortc wuov ndau daaih.

vi pn.dem n vi 120.1The fern cracked down, 120.2so (he) zoomed down to the ground with a sound of spurting air.

AS.121 Maiv_haih_fungc_zoux aqv. can’t_do_anything CHG-O-ST idm part.asp

He couldn’t help.

AS.122 122.1 122.2 122.3 Nquaqv sin daaih mbaix buoz mbaix zaux. get_up_quickly body come clap hand slap foot vt n v.dr vt n vt n 122.1(He) got up quickly 122.2clapped (his) hands 122.3beat (his) thighs.

AS.123 "Aa^Han_Mix Aa^Han_Nguv Aa^Han_Suv^Tiqc dongh yie aqv.

Bear_the_Brave Snake_the_Brave Suthi_the_Brave that_is 1sg I_am_sure

“Bear the Brave, Snake the Brave, Suthi the Brave, that’s me myself!”

Page 298: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

272 AS.124 124.1 124.2 124.3 Ih_hnoi ziangh, ih_hnoi daic yaac ih_hnoi haeqv! wuov_deix aeqv, today alive today die TOP today frighten those_people TOP

look try_and_see 1pl be_in here discuss only

hough) today (he) is alive, , today, 124.5there is no way to describe where he flew from.

NEG remember look fern

adv.tmp v.st adv.tmp vi top adv.tmp vt pn.dem top 124.4 mangc gaax mbuo yiem naaiv caangh_laangh hnangv,

v part pn.p vi pn.dem v adv 124.5 ninh mv beiv duqv yiem haaix_ndau ndaix daaih aqv. 3sg NEG compare can be_there what_ground fly come PFT pn.p neg v v.aux vi q vi vi part.asp 124.1(Even t 124.2(or) dies today 124.3(since he) frightened them124.4let’s see, we only discuss here,

AS.125 Maiv jangx mangc nyaaix-weih.

neg v v n

(He) didn’t remember to look at the fern.

AS.126 126.1 126.2 Uv! biaux ninh ziouc zorqv baeng nyei congx daaih Oh! flee 3sg so_then take soldier POSS gun come intj v pn.p seq.mk vt n poss n v.dr 126.3 126.4 buonv nyei buonv nzuangv nyei nzuangv zunc jienv mingh. shoot ADVBLZR shoot aimless ADVBLZR aimless chase SML.ACT go vt part vt v.st part v.st vt asp vi 126.1Oh! (they) ran away, 126.2so he took the soldier’s guns 126.3(and) shoot repeatedly, 126.4(and) chased them away furiously.

AS.127 Wuov_nzunc_hnoi ninh aengx ndaam wuov_deix nyei congx. this-time 3sg again carry those_people POSS gun adv.tmp pn.p adv vt pn.dem poss n

Then he futher carried their guns.

AS.128 Wuov_deix biaux nzengc mi'aqv. those_people flee consumed RSLT pn.dem vi vi asp

They all ran away.

AS.129 129.1 129.2 Ndaam jienv congx nzuonx mingh bun Guien mangc. carry SML.ACT gun return go let officer look vt asp n v vi vt n vt 129.1(He) went back carrying the guns on his shoulders, 129.2(and) let the governer look (at them).

Page 299: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

273 AS.130 130.1 130.2 Buonv baaic mi'aqv, ninh_mbuo biaux nzengc mi'aqv. shoot defeat RSLT 3_PL flee consumed RSLT vt vt asp pn.p v vi asp

I have told you.

130.1(He) shot (and) defeated (them) all; 130.2they all completely ran away.

AS.131 Guien aengx bun mienh mingh zorqv congx lo_haaix nzuonx daaih. officer further let person go take gun and_so_forth return come n adv vt n vi vt n n vi vi

The officer then let the people go get (their) guns and other things and come back.

AS.132 132.1 132.2 132.3 Wuov_nzunc_hnoi paaiv yietc gorqv deic-bung yietc norm fouv-zingh bun ninh nyanc. this_time appoint one corner region one CLF capital_city let 3sg eat adv.tmp vt numb n n numb clf n vt pn.p vt 132.1Then this time, (the governer) assigned one corner of a region 132.2(with) one capital city 132.3(to) let him make a living.

AS.133 Za'gengh! ninh yaac duqv zoux hlo. really 3sg also get be big adv pn.p conj vt vi v.st

Really he also became big.

AS.134 I_gox loz-hnoi za'gengh jomc haic. married_couple old-days really poor very n n adv v.st adv

Before that the coulple was really very poor.

AS.135 Nae! See intj

Page 300: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

274

Story 7 (BS): “A Story of Big Snake” Big snake BS.001 1.1 1.2 Loz-hnoi maaih dauh ong yungz duqv siec dauh sieqv, old-days have CLF old_man give_birth get seven CLF daughter n vt clf n vt vt n clf n

Ninh_mbuo lueic, mv oix zoux gong.

nzuonx daaih, ziouc gorngv mbuox ninh sieqv mbuo,

1.3 dorh hlo daaih. raise big come vt v.st vi 1.1In the old days, there was an old man; 1.2(he) begat seven daughters 1.3(and he) raised them (that they become) big.

BS.002 2.1 2.2

3_PL lazy NEG like do work pn.p v.st neg v.mod vt n 2.1They were lazy, 2.2(they ) didn't want to work.

BS.003 3.1 3.2 Ninh die buatc ninh_mbuo lueic haic ziouc mingh mangc ndau. 3sg father see 3_PL lazy very so go look land pn.p n v pn.p v.st adv seq.mk vi vt n 3.1 Their father saw that they were very lazy, 3.2he then went to look at a piece of land.

BS.004 4.1 Mangc duqv siec norm zorng, siec norm horngz aeqv, look can seven CLF mountain seven CLF gully TOP v v.aux n clf n n clf n top 4.2 4.3

return come so_then speak tell 3sg daughter PL v vi seq.mk v vt pn.p n pn "Njang hnoi mingh zaqv liangx. tomorrow go clear field adv.tmp vi vt n Oix_zuqc yietc hnoi zaqv taux nzengc." must entire day clear reach consumed aux adj n vt v vi 4.1(He) could look at seven mountains and seven gullies, 4.2(he) came back; 4.3(he) then spoke (and) told his daughters, “Tomorrow go and clear a field. You must clear (it) all day until (you get it) done.”

Page 301: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

275 BS.005 5.1 5.2

5.3

3sg want_to kill he daughter PL

The following day, then, his daughters went.

adj n vt vi asp

BS.008

Se_gorngv zaqv maiv nzengc, nzuonx daaih if clear NEG consumed return come conj vt neg vi v vi

ninh oix daix ninh sieqv mbuo.

pn.p v.mod vt pn.p n pn 5.1If (they) did not complete (and) cut, 5.2(and they) would come back, 5.3he would want to kill his daughters.

BS.006 Da'nyeic hnoi ninh sieqv mbuo ziouc mingh. second day 3sg daughter PL so_then go n n pn.p n pn seq.mk vi

BS.007 7.1 Yietc hnoi zaqv ziangx mi'aqv, whole day clear finish RSLT

7.2 7.3 ziouc nzuonx daaih gorngv mbuox ninh die, so_then return come speak tell 3sg father seq.mk v vi v vt pn.p n "Yie_mbuo zaqv nzengc aqv. 1_PL clear consumed PFT pn.p vt vi part.asp Aa die njang hnoi zoux, yietc hnoi goix taux nzengc." dear father tomorrow do whole day cut_down reach consumed part n adv.tmp vt adj n vt v vi 7.1The whole day (they) finished cleared all, 7.2(they) then came back, 7.3(they) spoke (and) told their father, “We have cleared them all. Dad, tomorrow (you) do all day until (it will be) complete.”

8.1 Njang ndorm ninh die nyanc liuz hnaangx next_morning 3sg father eat finish rice n pn.p n vt aux n 8.2 8.3 8.4 ziouc nzioux bouv daaih ndaam jienv mingh goix ndiangx aqv. so_then sharpen axe come carry SML.ACT go chop tree INCHOAT seq.mk vt n vi vt asp vi vt n part.asp 8.1The following morning, their father ate a breakfast, 8.2(he) then sharpened (his) axe; 8.3(he) carried_(it)_on_shoulder (and he) went 8.4(to) start chopping a tree.

Page 302: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

276 BS.009 9.1 9.2 Mingh taux ndeic-zorng mangc buatc ndiangx camv haic, go reach mountain-field look see tree many very vi v n v v n adj adv 9.3 nzauh_heix goix maiv hingh. warried cut_down NEG win v.cmp vt neg v 9.1(When he) arrived at the mountain-field, 9.2(he) saw many trees, 9.3(he) became worried (that he would) not manage to cut (them).

BS.010 10.1 Ninh ziouc yietc binc nyiemv yietc binc heuc lungh_ndiev, 3sg then one time weep one time call world

q_clf v.mod vt n vi vt pn.p vt n

hen one time cried, one time called [i.e. crying and calling] (to) the world, ing) whoever wants a wife, e help him cut down the tree.

BS.011 11.1 11.2 11.3 Heuc liuz, aav dangh ziouc buatc domh naang daaih.

come reach then ask try_and_see 3sg why? weep

pn.p poss n neg vt n pn.p vt

pn.p seq.mk numb n vi numb n v n 10.2 10.3 haaix_dauh oix longc auv daaih gan ninh goix ndiangx. who want marry wife come help 3sg cut_down tree

10.1He t 10.2(say10.3com

call finish in_a_short_while so_then see big snake come v aux adv.tmp seq.mk v adj n vi 11.1(When he) called, 11.2then soon (he) saw 11.3(that) a big snake came.

BS.012 12.1 12.2 12.3 Daaih taux ziouc naaic gaax ninh zoux_haaix_nyungc nyiemv.

vi v seq.mk v part pn.p q vi 12.1(When it) came (and) reached (to the man), 12.2then (it) asked him 12.3why (he was) crying.

BS.013 Ninh gorngv, 3sg speak pn.p v "Nzauh haic, goix ndiangx maiv nzengc. feel-sorry very cut_down tree NEG consumed v adv vt n neg vi Hitv hmuangx nzuonx mingh later dark return go adv.tmp adj v vi yie nyei sieqv maiv bun hnaangx yie nyanc." 1sg POSS daughter NEG give rice 1sg eat

He said, “Very sad (that) I can’t cut the tree down completely. Later when it’s dark, when I return home, my daughters would not give me food to eat.”

Page 303: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

277 BS.014

"Maiv_dungx nzauh. Yie tengx meih goix. NEG.CMD sad 1sg help 2sg cut_down

one CLF tree chop one axe only

Goix nzengc maiv nangc hmuangx.

Domh naang ziouc gorngv, big snake then speak adj n seq.mk v

neg v.st pn.p vt pn.p vt Meih simv nqoi. 2sg avoid be_apart pn.p vt vi Meih faaux wuov jiex ndeic-dauh mingh mangc jienv yie." 2sg go_up there up_there field go look DUR 1sg pn.p vi pn.dem adv.loc n vi vt asp pn.p

The big snake then said, “Don’t be sad. I will help you cut down. You stay back. You go up there, the upper field and watch me.”

BS.015 15.1 Ninh ziouc zorqv wuov dauh ong nyei bouv daaih, 3sg then take that CLF old_man POSS axe come pn seq.mk vt pn.dem clf n poss n vi 15.2 15.3 njiec wuov ndiev ndeic-jieqv mingh, ziouc goix aqv. go_down there down_there lower_field go then cut_down INCHOAT vi pn.dem adv.loc n vi seq.mk vt part.asp 15.1It [snake] then took that old man’s axe, 15.2went down to the lower field, 15.3then began to cut down.

BS.016 Yietc diuh ndiangx goix yietc bouv hnangv.

numb clf n vt numb n adv

(Cutting) one tree (was a matter of) just one stroke (of the axe).

BS.017 17.1 17.2 Faaux taux ndeic-dauh ziouc zoi bouv njiec, go_up reach upper-field so_then throw axe go_down vi v n seq.mk vt n vi 17.3 ndiangx ziouc nauv nzengc mi'aqv. tree so_then snap consumed RSLT n seq.mk vi vi asp 17.1(It) came up to the upper field, 17.2then (he) threw down the axe, 17.3so the tree snapped completely.

BS.018 18.1 18.2

cut_down consumed NEG so dark vt vi neg adv adj 18.1(The snake) cut down (the tree) completely, 18.2(but it) was not so dark (yet).

Page 304: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

278 BS.019 Mv dauh ong ziouc dorh jienv domh naang nzuonx aqv. this CLF old_man so_then take SML.ACT big snake return PFT pn.dem clf n seq.mk vt asp adj n v part.asp

Then this old man returned carrying the big snake.

BS.020 20.1 Nzuonx taux gaengh ndaangc, return reach door front v v n n 20.2 20.3 20.4 heuc ninh nyei sieqv Guh_Meix koi gaengh bun domh naang bieqc. order 3sg POSS daughter the_first_daughter open door let big snake enter v pn.p poss n n.prp vt n vt adj n vi 20.1(When he) returned to the door front, 20.2(he) ordered his first daughter 20.3(to) open the door 20.4(to) let the big snake enter.

BS.021 Ninh nyei sieqv-Guh_Meix ziouc gorngv, 3sg POSS daughter-the_first_daughter so_then speak pn.p poss n-n.prp seq.mk v "Yie mv koi. Domh naang aqc_mangc haic. 1sg NEG open big snake ugly very pn.p neg vt adj n adj adv Domh naang fungc bieqc biauv laeh." big snake how enter house will_you_tell_me? adj n adv vi n q.part

His first daughter then said, “I won’t open. A big snake is very ugly. How can a big snake enter the house?”

BS.022 Aengx heuc ninh nyei M'Naix daaih koi. further call 3sg POSS Second_daughter come open adv v pn.p poss n.prp vi vt

(He) further called his second daughter to come (and) open (the door).

BS.023 M'Naix yaac fih_hnangv nyei gorngv, Second_daughter also same ADVBLZR speak n.prp top adj part v "Yie maiv koi." 1sg NEG open pn.p neg vt

The second daughter also said in the same way, “I won’t open.”

BS.024 Heuc taux M'Faam. call reach Third_daughter v v n.prp

(He) called as far as the Third daughter.

Page 305: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

279 BS.025 25.1 Maiv haih fungc zoux aqv, NEG be_able_to how do PFT neg aux adv vt part.asp 25.2 25.3 nyiemv jienv mingh koi gaengh bun domh naang bieqc daaih. weep SML.ACT go open door let big snake enter come vi asp vi vt n vt adj n vi vi 25.1(She) could not help; 25.2(she) went crying 25.3(and) opened the door (to) let the big snake come in.

BS.026 Ninh die aengx heuc sieqv-Guh_Meix 3sg father again call the_third_daughter pn.p n adv v n zorqv don bun domh-naang-gorx zueiz. take stool let big-snake-elder_brother sit vt n vt adj-n-n vi

Her father again ordered the third daughter (to) take a stool (to) let the big snake elder brother sit.

BS.027 Sieqv-Guh_Meix aengx gorngv, the_third_daughter further speak n adv v "Yie maiv zorqv. 1sg NEG take pn.p neg vt Domh naang fungc hiuv_duqv zueiz don laeh!" big snake how know sit stool will_you_tell_me? adj n adv vt vi n q.part

The third daughter further said, “How can a big snake ever know sitting on a stool?

BS.028 Aengx heuc sieqv-M'naix zorqv. further order the_second_daughter take adv v n vt

(He) further order the second daughter to take (a stool).

BS.029 Sieqv-M'naix yaac fih_hnangv nyei gorngv, mv zorqv. the_second_daughter also same ADVBLZR speak NEG take n top adj part v neg vt

The second daughter also said in the same way (that she would) not take (a stool).

BS.030 Aengx heuc sieqv-M'Faam zorqv. further order the_third_daughter take adv v n vt

(He) further ordered the third daughter (to) take (a stool).

Page 306: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

280

BS.031 31.1 31.2 Heuc taux sieqv-M'Faam maiv_haih_fih_nyungc_zoux, order reach the_third_daughter can't_do_anything v v n idm 31.3 31.4 ziouc nyiemv jienv mingh zorqv don daaih so_then weep SML.ACT go take stool come seq.mk vi asp vi vt n vi 31.5 bun domh-naang-gorx zueiz. let big-snake-elder_brother sit vt adj-n-n vi 31.1(He) called as far as the third daughter; 31.2(she) could not do anything [to reject him], 31.3(she) then went crying 31.4(and) took the stool (and) came 31.5(and) let the big snake elder brother sit (on it).

BS.032 32.1 32.2 32.3 Aengx heuc Guh_Meix buov douz hnyutv wuom further order the_first_daughter burn fire boil water adv v n.prp vt n vt n 32.4 bun domh-naang-gorx nzaaux sin. let big-snake-elder_brother wash body vt adj-n-n vt n 32.1(He) further ordered the first daughter 32.2(to) make a fire (i.e. make a fire) 32.3(and) boil water 32.4(to) let the big snake elder brother take a bath.

BS.033 Guh_Meix gorngv ninh maiv hnyutv. the_first_daughter speak 3sg NEG boil n.prp v pn.p neg vt

The first daughter said (that) she would not boil.

BS.034 Domh naang fih_nyungc haih nzaaux sin mv bei. big snake how be_possible_to wash body I_guess adj n adv aux vt n part.f

How can a big snake possibly take shower, I wonder.

BS.035 35.1 35.2 Aengx heuc M'Naix yaac fih_hnangv nyei gorngv, further order the_second_daughter and_yet same ADVBLZR speak adv v n.prp top adj part v 35.3 ninh maiv hnyutv. 3sg NEG boil pn.p neg vt 35.1(He) further ordered the second daughter 35.2and (she) in the same way said 35.3(that) she would not boil (water).

Page 307: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

281 BS.036 36.1 36.2 Aengx heuc taux sieqv-M'Faam yaac maiv_haih_fih_nyungc_zoux. further order reach the_third_daughter also can't_do_anything adv v v n top idm 36.1(He) further called as far as the third daughter 36.2(she) also could not do anything.

BS.037 37.1 37.2 Ndongc haaix youx yaac zuqc nyiemv jienv mingh no-matter-how hate and_yet be-compelled weep SML.ACT go adv v top part vi asp vi 37.3 37.4 hnyutv wuom bun naang nzaaux sin. heat_up water let snake wash body vt n vt n vt n 37.1No matter how much (she) hated (it), 37.2(she) had to go crying 37.3(to) heat up water 37.4(to) let the snake take a shower.

BS.038 38.1 38.2 38.3 Hnyutv daaih dox ziangx aqv, ziouc heuc naang mingh nzaaux sin. boil come pour finish PFT so_then order snake go wash body vt vi vt vi part.asp seq.mk v n vi vt n 38.1(When she) boiled (it) 38.2(and) finished pouring (into a basin), 38.3(she) then called the snake (to) go and take a shower.

BS.039 Domh naang mingh nzaaux sin aqv. big snake go wash body PFT adj n vi vt n part.asp

The big snake went (and) took a shower.

BS.040 Sieqv-M'Faam ziouc hnamv taux, the_third_daughter so_then think reach n seq.mk vt vt "Aa die ndongc_haaix yaac heuc yie gan naang mingh hnangv." Ah father no-matter-how and_yet order 1sg follow snake go only Intj n adv top v pn.p vt n vi adv

The third daughter then thought about, “Ah, (my) father no matter what orders me to go and follow the snake only.”

BS.041 Ninh ziouc mingh cienh mangc naang nzaaux sin. 3sg so_then go inspect look snake wash body pn.p seq.mk vi vt v n vt n

She then went (to) inspected (and) saw (that) the snake washing himself.

BS.042 42.1 42.2 Mangc buatc maiv zeiz naang, se yietc dauh kuv mienh. look see NEG be_right snake that_is one CLF good person v v neg vi n copl numb clf adj n 42.1(She) looked (and) saw (that) it was not a snake, 42.2(but) that (it) was a good man.

Page 308: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

282

BS.043 43.1 Aav dangh nyanc hnaangx nyei ziangh_hoc, in_a_short_while eat rice RELT time adv.tmp vt n relt n 43.2 43.3 ninh die aengx heuc Guh_Meix mbenc hnaangx 3sg father again order the_first_daughter prepare rice pn.p n adv v n.prp vt n 43.4 bun naang-gorx nyanc. let snake-elder_brother eat vt n-n vt 43.1In a short while, when (it became) time (to) eat rice [i.e. have a meal], 43.2her father called the first daughter again 43.3(to) prepare a meal 43.4(to) let the snake elder brother eat.

BS.044 Guh_Meix aengx gorngv, the_first_daughter again speak n.prp adv v "Yie maiv zeiz mbenc. 1sg NEG be_right prepare pn.p neg vi vt Domh naang fih_nyungc hiuv_duqv nyanc hnaangx." big snake how_on_earth know eat rice adj n adv vt vt n

The first daughter said again, “I won’t prepare (anything). How could a big snake possibly know eating rice?”

BS.045 Ninh die aengx heuc M'Naix mbenc. 3sg father again order the_second_daughter prepare pn.p n adv v n.prp vt

Their father order again the second daughter to prepare (food).

BS.046 46.1 46.2 M'Naix yaac fih_hnangv nyei gorngv ninh maiv lorz. the_second_daughter also same ADVBLZR speak 3sg NEG look_for n.prp top adj part v pn.p neg vt 46.1The second daughter also said in the same way 46.2(that) she would not look for (food).

BS.047 47.1 47.2 Aengx taux M'Faam, maiv_haih_fih_nyungc_zoux. further reach the_third_daughter can't_do_anything adv v n.prp idm 47.1Again (it came) as far as the third daughter, 47.2(she) could not do anything [to avoid the father’s order].

Page 309: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

283 BS.048 48.1 48.2 Ndongc haaix youx, nyiemv jienv no-matter-how hate weep SML.ACT adv v vi asp 48.3 48.4 yaac oix_zuqc mingh lorz hnaangx bun naang nyanc. and_yet must go look_for rice offer snake eat top aux vi vt n vt n vt 48.1Now matter how she dislike (the idea), 48.2(as to the degree of her) crying, 48.3even if so, (she) had to go and find food 48.4(to) let the snake eat.

BS.049 49.1 Nyanc liuz hnaangx, eat finish rice vt aux n 49.2 49.3 49.4 ninh die aengx heuc Guh_Meix dimc dorngx bun domh naang bueix. 3sg father again order the_first_daughter overlay place offer big snake sleep pn.p n adv v n.prp vt n vt adj n vi 49.1(After they) finished having a meal, 49.2their father told again the first daughter 49.3(to) lay out [a bed or mattress on the] place [for the big snake to sleep on] 49.4(to) let the big snake sleep.

BS.050 Guh_Meix gorngv, the_first_daughter speak n.prp v "Yie maiv dimc. 1sg NEG overlay pn.p neg vt Domh naang fih_nyungc oix_zuqc dimc dorngx bueix laeh! big snake how_can must overlay place lie_down I_object_you adj n adv aux vt n vi q.part La'guaih kuing jienv wuov ndau haac_maiv_duqv?" just coil_around DUR that ground don't_you_think_it_can_be_done? adv vi asp pn.dem n idm

The first daughter said, “I won’t overlay (a mattress). How come I should overlay a place for a big snake to sleep? Can’t it just coil around on the ground there?”

BS.051 51.1 51.2 Aengx heuc ninh nziez M'Naix mingh dimc dorngx. further order 3sg younger_sister the_second_daughter go overlay place adv v pn.p n n.prp vi vt n 51.1(He) again ordered her younger sister, the second daughter, 51.2(to) go and overlay a (sleeping) place.

BS.052 52.1 52.2 Ninh nziez yaac fih_hnangv nyei gorngv, ninh mv dimc. 3sg younger_sister also same ADVBLZR speak 3sg NEG overlay pn.p n top adj part v pn.p neg vt 52.1Her younger sister also said in the same way 52.2(that) she would not overlay (a sleeping place).

Page 310: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

284 BS.053 Aengx heuc ninh nyei sieqv-M'Faam mingh dimc. again order 3sg POSS the_third_daughter go overlay adv v pn.p poss n vi vt

(He) again ordered his third daughter to go and overlay.

BS.054 54.1 54.2 54.3 Sieqv-M'Faam aeqv, ndongc_haaix youx yaac maiv_haih_fih_nyungc_zoux. the_third_daughter TOP no-matter-how dislike and_yet can't_do_anything n part adv v top idm 54.1As for the third daughter, 54.2no matter how (she) disliked [the idea], 54.3and yet (she) could not do anything [to reject it].

BS.055 55.1 55.2 Oix_zuqc muangx ninh die nyei waac, mingh dimc dorngx aqv. must listen 3sg father POSS word go overlay place CHG-O-ST aux v pn.p n poss n vi vt n part.asp 55.1(She) had to listen to her father’s word 55.2(and she) went (to) overlay a place.

BS.056 Domh naang gorngv mbuox sieqv-M'Faam, big snake speak tell the_third_daughter adj n v vt n "Mv_zuqc hnangv_haaix_nor dimc oc. no_need_to how overlay PLT adv adv vt part La'guaih zorqv norm siang hietv jienv duqv aqv." just take CLF woven_tray put DUR can CAS.CMD adv vt clf n vt asp v.aux part.mod

The big snake told the third daughter, “You don’t have to specially overlay (a bed). Just pick up a woven tray and put it there; that’s just fine.”

BS.057 57.1 Sieqv-M'Faam ziouc mingh zorqv siang daaih the_third_daughter so_then go take woven_tray come n seq.mk vi vt n v.dr 57.2 57.3 hietv jienv ting, gorngv mbuox domh naang, place DUR central_area_of_a_house speak tell big snake vt asp n v vt adj n "Gorx aac, kouv haic, bueix aqv." elder_brother listen_please tired very lie_down CAS.CMD n part.voc v.st adv vi part.mod 57.1So the third daughter went and took a woven tray hither, 57.2(and she) placed it in the central part of the house, 57.3(and) said to the big snake, “Elder brother! You must be tired. Come and lie down.”

Page 311: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

285 BS.058 58.1 Naang ziouc luih mv luih nyei bieqc siang mingh snake so_then slither ** slither ADVBLZR enter woven_tray go n seq.mk vi part vi part vi n v.dr 58.2 58.3 kuing jienv siang gu'nyuoz, bueix jienv siang gu'nyuoz wuov. coil_around DUR woven_tray inside sleep DUR woven_tray inside there vi asp n n vi asp n n pn.dem 58.1Then the snake slitheringly went into the woven tray thither, 58.2(and it was) coiling around in the woven tray, 58.3(was) sleeping inside the woven tray there.

BS.059 59.1 Da'nyeic ndorm nyanc liuz hnaangx, second morning eat finish rice n n vt aux n 59.2 ninh die ziouc paaiv ninh nyei sieqv-Guh_Meix, 3sg father so_then send_with_an_errand 3sg POSS the_third_daughter pn.p n seq.mk vt pn.p poss n "Meih oix_zuqc paanx suangx 2sg must tie_with_straps_cross_over blanket pn.p aux vt n gan domh-naang-gorx mingh zoux auv." follow big-snake-elder_brother go be wife vt adj-n-n vi vi n 59.1The following morming, (after they) ate breakfast, 59.2their father then told the First daughter, saying, “Wrap your blanket, and follow the big snake to be his wife.”

BS.060 Guh_Meix gorngv, the_first_daughter speak n.prp v "Yie fungc mingh laeh. 1sg how_come go I_object_you pn.p adv vi q.part Domh naang aqc_mangc daic. Yie maiv longc." big snake ugly intensely 1sg NEG marry adj n adj adv pn.p neg vt

The first daughter said, “How come I should go? The big snake is terribly ugly. I won’t marry (it).”

BS.061 Aengx paaiv M'Naix gan domh-naang gorx mingh. further appoint the_second_daughter follow big-snake elder_brother go adv vt n.prp vt adj-n n vi

(He) further appointed the second daughter to go and follow the big snake elder brother.

Page 312: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

286 BS.062 62.1 62.2 M'Naix yaac fih_hnangv nyei gorngv ninh mv mingh. the_second_daughter also same ADVBLZR speak 3sg NEG go n.prp top adj part v pn.p neg vi 62.1The second daughter also said in the same way 62.2(that) she would not go.

BS.063 Aengx paaiv sieqv-M'Faam. further appoint the_third_daughter adv vt n

Again (he) appointed the third daughter.

BS.064 64.1 64.2 Maiv_haih_fih_nyungc_zoux, nyiemv jienv jiez_sin mingh can't_do_anything weep SML.ACT get_up go idm vi asp v.cmp vi 64.3 64.4 nzipv suangx gan jienv domh naang mingh. fold blanket follow SML.ACT big snake go vt n vt asp adj n vi 64.1(She) could not do anything, 64.2(she) got up crying (and) went, 64.3folded up a blanket 64.4(and) following the big snake went.

BS.065 Jauv zangc aah_loc zoux cunv ninh nqox. road above frequently do pester 3sg husband n adj adv vt vt pn.p n

On the way (she) pestered her husband frequently.

BS.066 66.1 66.2 Ninh longc zaux-ndoqv-nyeiz neqv domh naang nyei dueiv. 3sg use big_toe claw_into_by_the_toe big snake POSS tail pn.p vt n vt adj n poss n 66.1She used (her) big toe 66.2clawed the big snake’s tail. (i.e. clawed the big snake’s tail with her toe.

BS.067 67.1 67.2 Naang huin nzuonx daaih mangc mangc deix, ziouc gorngv mbuox ninh, snake turn_around return come look look some so_then speak tell 3sg n vi v v.dr v v adv seq.mk v vt pn.p "Meih maiv_dungx zoux." 2sg NEG.CMD do pn.p neg vt 67.1The snake turned around hither (and) looked (at her) well; 67.2then said to her, “Don’t do that.”

Page 313: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

287 BS.068 Ninh gorngv, 3sg speak pn.p v

"Yie maiv jangx mangc, caqv-bouc caaiv zuqc hnangv." 1sg NEG remember look accidentally step_on GOAL only pn.p neg v v adv vt part adv

She said, “I just forgot to look; it’s only I accidentally stepped on you.”

BS.069 Naang aengx mingh. snake again go n adv vi

The snake went (on) again.

BS.070 70.1 70.2 Mingh dangh ninh aengx yietv neqv. go awhile 3sg again immediately claw_into_by_the_toe vi adv pn.p adv adv vt 70.1(They) went for a while, 70.2she immediately clawed (it) again.

BS.071 Naang huin nzuonx daaih gorngv, snake turn_around return come speak n vi v v.dr v "Mv_dungx zoux maah! Aav_dangh yie zoux bun meih. NEG.CMD do SFT.CMD in_a_short_while 1sg do give 2sg neg vt part.f adv.tmp pn.p vt vt pn.p Meih youc gorngv mun haic lov." 2sg then speak feel_hurt very *** pn.p seq.mk v vi adv part.f

The snake turned around toward (her and) said, “Don’t do that! Next time I will do it to you. Then you will cry, ‘Oh, it hurts!’”

BS.072 Ninh aengx gorngv "Caqv-bouc weqv!" 3sg again speak accidentally ASSERT pn.p adv v adv part

She said again, “Can’t you see it’s an accident!”

BS.073 73.1 73.2 Naang yaac maiv gorngv haaix nyungc, aengx mingh aqv. snake also NEG speak what kind further go CHG-O-ST n top neg v q n adv vi part.asp 73.1The snake also didn't say anything either 73.1(but) again went on.

Page 314: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

288 BS.074 74.1 Da'faam wuov nzunc ninh ziouc huin nzuonx daaih third that time 3sg so_then turn_around return come adj pn.dem n pn.p seq.mk vi v v.dr 74.2 nyorqv zuqc ninh nyei cingh_mborqc. peck GOAL 3sg POSS knee v part pn.p poss n 74.1The third time [when she clawed it], then it (the snake) turned around toward (her) 74.2(and) pecked her (in the) knee.

BS.075 75.1 75.2 75.3 Ninh ziouc gorngv mun haic, njangx bieqc njangx cuotv, nyiemv gorngv, 3sg so_then speak feel_hurt very roll enter roll go_out weep speak pn.p seq.mk v vi adv vi vi vi v vi v "Mun haic. Mingh maiv duqv aqv." feel_hurt very go NEG can CHG-O-ST vi adv vi neg v.aux part.asp 75.1So she said (that it) hurt very much, 75.2(and she) rolled in, rolled out, 75.3cried (and) said, “It hurts very much. I can’t go any longer.”

BS.076 76.1 Naang aengx huin nzuonx daaih snake further turn_around return come n adv vi v vi 76.2 76.3 76.3 tuiv wuom-nzuih gan ninh nzaatv jienv ziouc longx mi'aqv. spit saliva help 3sg smear DUR so_then good RSLT vt n vt pn.p vt asp seq.mk v.st asp 76.1The snake turned around again (and) came back, 76.2spat saliva 76.3helped her smeared 76.3then (the wound) got healed.

BS.077 Aengx jiez_sin daaih gan jienv naang mingh. further get_up come follow SML.ACT snake go adv v.cmp vi vt asp n vi

And then (she) got up (and) went following the snake.

BS.078 78.1 78.2 Mingh gau taux diuh ndoqv nor, go DPCL reach CLF stream as vi part v clf n top 78.3 78.4 ninh gorngv mbuox ninh nyei auv yiem wuov ndoqv-hlen zuov ninh. 3sg speak tell 3sg POSS wife be_there there stream-side wait_for 3sg pn.p v vt pn.p poss n vi pn.dem n-n vt pn.p 78.1When (they) went, 78.2as (they) reached a stream, 78.3it (snake) spoke (and) told its wife 78.4(that she should) stay there at the side of the stream (and) wait for him.

Page 315: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

289 BS.079 Ninh mingh wuov_nyuoz maengx nzaaux sin. 3sg go in_the_valley side wash body pn.p vi adv.loc n vt n

It went to the side of the valley (and) washed (his) body [i.e. took shower].

BS.080 "Meih yiem naaiv buatc wuom cuotv maeng siqv yangh mbuov 2sg be_here here see water come_out green red yellow blue pn.p vi pn.dem v n v v.st v.st v.st v.st mv_dungx nuqv oc." NEG.CMD point_with_finger PLT neg vi part

“You stay here. (If) you see water coming out with its colors in green, red, yellow and blue, don’t put your finger in it.”

BS.081 81.1 81.2 Mbuox liuz ninh ziouc mingh mi'aqv. tell finish 3sg so_then go RSLT vt aux pn.p seq.mk vi asp 81.1(He) finished telling (her), 81.2then he has gone.

BS.082 Ninh auv yiem wuov. 3sg wife be_in there pn.p n vi pn.dem

His wife was there.

BS.083 83.1 83.2 83.3 Aav dangh ziouc buatc cuotv aqv. in_a_short_while so_then see come_out PFT adv.tmp seq.mk v v part.asp 83.1Soon, 83.2(she) saw 83.3(the colored water) came out.

BS.084 84.1 84.2 84.3 Ninh buatc wuom nzueic haic ziouc nuqv, 3sg see water beautiful very so_then point_with_finger pn.p v n v.st adv seq.mk vi 84.4 za'gengh nitv jienv buoz-ndoqv aqv. really close_to DUR finger PFT adv vi asp n part.asp 84.1She saw 84.2(that) the water was very beautiful, 84.3so she pointed her finger (into the water); 84.4(she) really kept her finger touching (the water).

Page 316: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

290 BS.085 85.1 Buatc ninh nqox cuotv, see 3sg husband come_out v pn.p n vi 85.2 85.3 85.4 haeqv ninh bingx jienv buoz-ndoqv, mv bun ninh nqox mangc. frighten 3sg hide DUR finger NEG let 3sg husband look vt pn.p vt asp n neg vt pn.p n v 85.1(When she) saw her husband came out, 85.2(it) frightened her, 85.3(she) hid her finger, 85.4(she) did not let her husband see (it)

BS.086 86.1 86.2 86.3 Ninh nqox taux daaih ziouc heuc ninh auv bun buoz-ndoqv ninh mangc. 3sg husband reach come so_then order 3sg wife give finger 3sg look pn.p n v v.dr seq.mk v pn.p n vt n pn.p vt 86.1(When) her husband reached hither, 86.2(he) then ordered his wife 86.3(to) give him her finger (to) look at.

BS.087 Ninh auv mv bun ninh mangc. 3sg wife NEG let 3sg look pn.p n neg vt pn.p v

His wife would not let him look at (it).

BS.088 88.1 88.2 Tov gau ninh cingx_daaih bun ninh mangc. ask_for DPCL 3sg therefore let 3sg look vt part pn.p conj vt pn.p v 88.1When (he) asked, 88.2(she) therefore let him look at (it).

BS.089 Za'gengh ziangh jienv aqv. really alive DUR PFT adv vi asp part.asp

Actually the finger is attatched (to the hand).

BS.090 90.1 90.2 90.3 Ninh cingx_daaih gan ninh auv nzaaux nzengc mi'aqv, 3sg therefore help 3sg wife wash consumed RSLT pn.p conj vt pn.p n vt vi asp 90.4 cingx_daaih aengx mingh taux domh naang nyei biauv. therefore further go reach big snake POSS house conj adv vi v adj n poss n 90.1Therefore it (snake) 90.2helped his wife 90.3cleansed (wound) completely; 90.4therefore (they) further went (and) reached the big snake’s house.

Page 317: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

291 BS.091 91.1 91.2 Longc jienv buo hnyangx, yungz duqv dauh dorn daaih. marry DUR three year give_birth get CLF son come vt asp numb n vt vt clf n v.asp 91.1(They) married for three years, 91.2(she) gave_birth and got a baby son.

BS.092 92.1 92.2 Nyiex jienv ziouc nzuonx_ngoih_jaa aqv. carry_on_one's_back DUR so_then visit_parents'_home_after_marriage PFT vt asp seq.mk vi part.asp 92.1(she was) carring her baby on the back, 92.2then (she) returned home to pay respect to her parents after the wedding.

BS.093 93.1 93.2 Nzuonx taux ninh die nyei biauv daaih, yiem duqv buo hnoi. return reach 3sg father POSS house come be_in attain three day v v pn.p n poss n v.dr vi vt numb n

here (and) got three days [i.e. she stayed there for three days].

"Wuov ndiev n jaangh maaih diuh biouv kuv haic,

93.1(She) returned (and) reached her father’s house, 93.2(she) was t

BS.094 Ninh dorc ziouc nduov ninh nziez gorngv, 3sg elder_sister so_then deceive 3sg younger_sister speak pn.p n seq.mk vt pn.p n v

there down_there pond have CLF fruit delicious very pn.dem adv.loc n vt clf n v.st adv oix mingh gaeqv biouv nyanc." want_to go pick fruit eat v.mod vi vt n vt

Her elder sister then deceived her younger sister (and) said, “There are very delicious fruits down there near the pond. I want to go get fruits and eat.”

BS.095 Ninh nziez gorngv, 3sg younger_sister speak pn.p n v "Nyiex jienv guh_nguaaz mingh maiv duqv." carry_on_one's_back SML.ACT baby go NEG can vt asp n vi neg v.aux

Her younger sister said, “I can’t go carrying my baby on my back.”

BS.096 Ninh dorc M'Naix gorngv, 3sg elder_sister the_second_daughter speak pn.p n n.prp v "Mv gamh_nziex, mingh duqv nyei." NEG fear go can AFM neg v.cmp vi v.aux asp

Her elder sister, the second one, said, “Don’t be afraid, we can go.”

Page 318: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

292 BS.097 97.1 97.2 Mv_haih_fih_nyungc_zoux, gan jienv mingh taux biouv-gorn. can't_do_anything follow DUR go reach fruit-root idm vt asp vi v n-n 97.1(She) could not do anything; 97.2(she) went following (them and) reached a root of the fruit (tree).

BS.098 98.1 98.2 98.3 Ninh dorc ziouc laengz gan ninh nziez nyiex guh_nguaaz aeqv, 3sg elder_sister so_then promise help 3sg younger_sister carry baby TOP pn.p n seq.mk v vt pn.p n vt n top 98.4 98.5 heuc ninh nziez faaux ndiangx gaeqv biouv. order 3sg younger_sister climb tree pick fruit v pn.p n vt n vt n 98.1Then as her elder sister promised 98.2(to) help her younger sister 98.3carry the baby, 98.4(she, i.e. the elder sister) called her younger sister 98.5(to) climb the tree (and) pick fruits.

BS.099 Ninh nziez gorngv, 3sg younger_sister speak pn.p n v "Faaux maiv duqv." go_up NEG can vi neg v.aux

Her younger sister said, “I can’t go up.”

BS.100 Ninh dorc gorngv, 3sg elder_sister speak pn.p n v "Faaux duqv nyei." go_up can AFM vi v.aux asp

Her elder sister said, “You can go up.”

BS.101 Ninh nziez ziouc faaux aqv. 3sg younger_sister so_then go_up CHG-O-ST pn.p n seq.mk vi part.asp

So her sister climbed up.

BS.102 Faaux taux ndiangx-dueiv mi'aqv. go_up reach tree-tail RSLT vi v n-n asp

She climbed up to the end of a branch.

Page 319: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

293 BS.103 103.1 103.2 103.3 Ninh dorc yiem ndiangx-gorn ziouc nietv guh_nguaaz nyiemv aqv, 3sg elder_sister be_there tree-root so_then pinch baby weep PFT pn.p n vi n-n seq.mk vt n vi part.asp 103.4 ziouc gorngv-baeqc nduov ninh nziez gorngv, so_then tell_a_lie deceive 3sg younger_sister speak seq.mk v.cmp vt pn.p n v "Guh_nguaaz oix_longc buoz-jiemh caux ja'waanh." baby want bracelet and silver_neckpiece n v.cmp n conj n 103.1Her elder sister stayed at the root of the tree, 103.2then (she) pinched the baby 103.3(and made the baby) cry, 103.4then (she) lied (and) deceived her younger sister (and) said, “(Your) baby wants (your) bracelet and silver neckpiece.”

BS.104 Ninh nziez ziouc maeqv zoi njiec bun liuz aqv. 3sg younger_sister so_then peel throw descend give finish PFT pn.p n seq.mk vt vt vt vt vt part.asp

Her younger sister then took (them off and) threw down (and) finished giving (them).

BS.105 105.1 105.2 Ninh aengx nietv da'nyeic nzunc, guh_nguaaz aengx nyiemv. 3sg again pinch second time baby again weep pn.p adv vt n n n adv vi 105.1She pinched again for the second time, 105.2the baby cried again.

BS.106 Ninh dorc aengx gorngv, 3sg elder_sister again speak pn.p n adv v "Guh_nguaaz oix_longc nyaanh_limc, m'normh_hiun norh." baby want silver_chain earring REPORT n v.cmp n n part

Her elder sister said again, “The baby wants a silver chain (and) earring, he’s saying.”

BS.107 Ninh nziez aengx bun njiec daaih. 3sg younger_sister again give go_down come pn.p n adv vt vi v.dr

Her younger sister gave (them) down (toward) again.

BS.108 108.1 108.2 Aengx nyiemv da'faam nzunc ninh dorc gorngv, again weep third time 3sg elder_sister speak adv vi adj n pn.p n v "Guh_nguaaz oix_longc m'nqorngv-beu, la'sin." baby want turban waist_sash n v.cmp n n 108.1Again (the baby) for the third time, 108.2her elder sister said, “The baby wants a turban (and) waist sash.”

Page 320: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

294 BS.109 Da'feix nzunc gorngv, fourth time speak nmb n vt "Oix_longc lui-houx." want clothes v.cmp n

She said for the fourth time, “(The baby) wants (your) clothes.”

BS.110 110.1 110.2 Gorngv taux naaiv_deix ninh nziez gorngv, speak reach these 3sg younger_sister speak v vt pn.dem pn.p n v "Mv fungc bun, nyaiv haic." NEG how_come give feel_ashamed very neg adv vt vi adv 110.1(When) the elder sister talked about these things, 110.2her younger sister said, “How come I can give (them)? (I) feel ashamed very much.”

BS.111 Ninh dorc gorngv, 3sg elder_sister speak pn.p n v "Mv gamh_nziex. Mv maaih mienh, kungx mbuo ganh hnangv. NEG fear NEG have people only 1pl self only neg v.cmp neg vt n adv pn pn.emph adv Maiv nyaiv, bun duqv nyei." NEG feel_ashamed give can AFM neg vi vt v.aux asp

Her elder sister said, “Don’t be afraid. There are no people around us, only us. No shame, you can give them all.”

BS.112 Ninh nziez ziouc bun nzengc ninh yietc sin nyei siou-setv mi'aqv. 3sg younger_sister so_then give consumed 3sg whole body POSS set_of_dress RSLT pn.p n seq.mk vt vi pn.p adj n poss n asp

So her younger sister gave up all of her dress.

BS.113 Taux da'hmz nzunc ninh aengx gorngv, reach fifth time 3sg further speak v nmb n pn.p adv v "Guh_nguaaz oix_longc wuov ndiev baby want there down_there n v.cmp pn.dem adv.loc ndiangx-nquaah dueiv-mueiz wuov joih biouv norh." branch end_of_tail that bunch fruit REPORT n n pn.dem n n part

In the fifth time, the elder sister said again, “The baby wants that bunch of fruits at the end of the branch down there, he said.”

Page 321: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

295 BS.114 Ninh nziez gorngv, 3sg younger_sister speak pn.p n v "Mingh mv duqv." go NEG can vi neg v.aux

The younger sister said, “I can’t go.”

BS.115 Ninh dorc gorngv, 3sg elder_sister speak pn.p n v "Mingh duqv nyei, nyorng jienv mingh hnangv." go can AFM creep SML.ACT go only vi v.aux asp vi asp vi adv

The elder sister said, “Yes, you can; just creep and go.”

BS.116 Ninh nziez ziouc nyorng jienv mingh taux ndiangx-nquaah dueiv-mueiz mi'aqv. 3sg younger_sister so_then creep SML.ACT go reach branch end_of_tail RSLT pn.p n seq.mk vi asp vi vt n n asp

Her younger sister then went creeping (and) reached the end of the branch.

BS.117 117.1 117.2 Ninh dorc ziouc goix ndiangx-gorn, ndiangx ziouc nauv. 3sg elder_sister so_then cut_down tree-root tree so_then snap pn.p n seq.mk vt n-n n seq.mk vi 117.1Her elder sister then cut down the tree at its base; 117.2then the tree fell down.

BS.118 118.1 Ninh nziez ziouc zong njiec wuom-njaangh mi'aqv, 3sg younger_sister so_then strike go_down water-pond RSLT pn.p n seq.mk vt vi n-n asp 118.2 ziouc daic wuov njaangh mi'aqv. so_then die there pond RSLT seq.mk vi pn.dem n asp 118.1Her younger sister then crashed down into the pond, 118.2then died there in the pond.

Page 322: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

296 BS.119 119.1 Ninh dorc ziouc zorqv ninh nziez nyei lui-houx 3sg elder_sister so_then take 3sg younger_sister POSS clothes pn.p n seq.mk vt pn.p n poss n 119.2 siou-setv daaih zorng jienv a_set_of_woman's_dress come dress DUR n vi vt asp 119.3 119.4 ziouc nyiex jienv guh_nguaaz nzuonx taux biauv. so_then carry_on_one's_back DUR baby return reach house seq.mk vt asp n v v n 119.1Then her elder sister took her younger sister’s clothes and the whole set of dress 119.2(and) put them on; 119.3(she) then carrying the baby 119.4went home.

BS.120 120.1 Ninh nziez nyei nqox yaac hiuv nyei 3sg younger_sister POSS husband also know PRS.ST pn.p n poss n top v asp 120.2 mv_baac ninh maiv gorngv haaix nyungc. but 3sg NEG speak what kind conj pn.p neg v q n 120.1The younger sister’s husband also knew [what had happened] 120.2but he did not say anything.

BS.121 121.1 Aa_hitv_hmuangx nyei ziangh_hoc, mingh bueix, later_tonight POSS time go sleep adv.tmp poss n vi vi 121.2 ninh nqox ziouc naaic ninh auv gorngv, 3sg husband so_then ask 3sg wife speak pn.p n seq.mk vt pn.p n v "Meih da'hnaav_muonz bun meih nyei mba'biei yie zoux nzomx-dauh. 2sg couple_evening_before offer 2sg POSS hair 1sg be pillow pn.p adv.tmp vt pn.p poss n pn.p vi n Ih_muonz meih fih_nyungc mv bun?" tonight 2sg why NEG offer adv.tmp pn.p adv neg vt 121.1At night when they went to bed, 121.2her husband then asked her, “A couple evenings before you offered your hair to me as a pillow. Tonight why do you not offer?”

Page 323: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

297 BS.122 Ninh auv dau ninh gorngv, 3sg wife answer 3sg speak pn.p n vt pn.p v "Ov mv maaih aqv. SURPRISE NEG have PFT exclam neg vt part.asp Nzuonx daaih ninh_mbuo lueic haic, mv zouv hnaangx. return come 3_PL lazy very NEG cook rice v vi pn.p v.st adv neg vt n Yie zouv gau hnaangx, douz jaamh nzengc mi'aqv." 1sg cook DPCL rice fire singe consumed RSLT pn.p vt part n n vt vi asp

His wife answered him (and) said, “Oh, I don’t have hair. When I came home, (I found) they were lazy, they didn’t cook rice. As I cooked rice, the fire singed all my hair.”

BS.123 Ninh nqox yaac mv gorngv haaix nyungc.

BS.126

3sg husband and NEG speak what kind pn.p n top neg v q n

Her husband didn’t say anything.

BS.124 124.1 124.2 Da'nyeic ndorm ninh_mbuo ziouc nziouv nziouv nyei nzipv suangx nzuonx aqv. second morning 3_PL so_then early early ADVBLZR fold blanket return PFT n n pn.p seq.mk v.st v.st part vt n v part.asp 124.1Then the following morning they folded blankets very early 124.2went home.

BS.125 125.1 Nzuonx taux biauv nyei ziangh_hoc return reach house RELT time v v n relt n 125.2 ninh nqox ziouc heuc ninh auv mingh koi gen-gaengh. 3sg husband so_then order 3sg wife go open bedroom_door pn.p n seq.mk v pn.p n vi vt n 125.1When (they) returned to home, 125.2her husband then ordered his wife (to) go (and) open the bedroom door.

Ninh auv mingh koi yaac koi maiv nqoi. 3sg wife go open and_yet open NEG apart pn.p n vi vt top vt neg vi

His wife (tried to) open (the door) and yet (it) would not open.

BS.127 Ninh nqox aengx mingh tengx ninh auv koi nqoi mi'aqv. 3sg husband further go help 3sg wife open be_apart RSLT pn.p n adv vi vt pn.p n vt vi asp

Her husband further went (to) help his wife (and it) opened.

Page 324: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

298 BS.128 128.1 128.2 128.3 Ninh nqox aengx heuc ninh auv koi faang koi kapv. 3sg husband further order 3sg wife open box open woven_basket_with_a_lid pn.p n adv v pn.p n vt n vt n 128.1Her husband called his wife again 128.2(to) open the box 128.3(and) open the basket.

BS.129 Koi yaac maiv nqoi. open and_yet NEG be_apart vt top neg vi

(She tried to) open but (they) would not open.

BS.130 Wuov nzunc hnoi ninh nqox ziouc gorngv, this-time 3sg husband so_then speak adv.tmp pn.p n seq.mk v "Yietc_gau meih koi duqv nqoi. usually 2sg open managed_to be_apart adv pn.p vt vt vi Ih_muonz meih fih_nyungc koi maiv nqoi aqv?" tonight 2sg why open NEG be_apart CHG-O-ST adv.tmp pn.p q vt neg vi part.asp

This time her husband then said, “Usually you can manage to open. Tonight how come you became unable to openn?”

BS.131 Ninh auv dau ninh nqox gorngv, 3sg wife answer 3sg husband speak pn.p n vt pn.p n v "Maiv hiuv. NEG know neg v Naaiv hieh_guaiv_ga'naaiv weic_haaix_nyungc koi mv nqoi mv_bei." this monster why open NEG be_apart I_wander pn.dem n q vt neg vi part.f

His wife answered her husband (and) said, “I don’t know. I wonder why these monster things won’t open.”

BS.132 132.1 Ninh nqox aengx mingh tengx ninh koi, 3sg husband again go help 3sg open pn.p n adv vi vt pn.p vt 132.2 aengx zorqv koi nqoi nzengc mi'aqv. again take open be_apart consumed RSLT adv vt vt vi vi asp 132.1Her husband went again (and) helped her open, 132.2(he) again took (it) open completely.

Page 325: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

299 BS.133 Da'nyeic ndorm aengx heuc ninh auv mingh ziaaux saeng-kuv. second morning again order 3sg wife go call animal n n adv v pn.p n vi vt n

The following morning (he) again ordered his wife (to) go (and) call [domestic] animals [to come for feeding].

BS.134 Ninh auv aengx mingh ziaaux saeng-kuv. 3sg wife further go call animal pn.p n adv vi vt n

His wife further (i.e. indeed) went (and) called animals.

BS.135 135.1 135.2 Ziaaux, yietc nyungc yaac maiv nzuonx. call_animals one kind TOP NEG return vt nmb n part neg v 135.1[Though she] called (animals), 135.2even one kind (of anmal) did not return.

BS.136 Ngongh maaz jai dungz yaac maiv nzuonx. cow horse chicken pig also NEG return n n n n top neg v

[None of] a cow, a horse, a chicken or a pig did not return.

BS.137 Ninh nqox aengx gorngv, 3sg husband again speak pn.p n adv v "Meih ndorm-ndorm goux nyei saeng-kuv, 2sg morning-morning take_care RELT animal pn.p n-n vt relt n ih_ndorm fih_nyungc ziaaux mv nzuonx?" this_morning why call NEG return adv.tmp adv vt neg v

Her husband said aganin, “This morning why don’t the animals you take care of every morning return?”

BS.138 Ninh auv dau, 3sg wife answer pn.p n vt "Mv hiuv_duqv. NEG know neg vt Janx-daic saeng-kuv ih_ndorm fih_nyungc ziaaux maiv nzuonx mv_bei." dead_person animal this_morning why call_animals NEG return I_wander n n adv.tmp q vt neg v part.f

His wife answered, “I don’t know. I wonder why a dead person or an animal won’t return at all this morning?”

Page 326: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

300 BS.139 139.1 Ninh nqox aengx mingh gan ninh ziaaux 3sg husband further go help 3sg call_animals pn.p n adv vi vt pn.p vt 139.2 ziouc ziaaux nzuonx nzengc daaih. so_then call_animals return consumed come seq.mk vt v vi v.dr 139.1Her husband again went (and) helped her call (the animals to return) 139.2then all the animals came back.

BS.140 Wuov nzunc hnoi ninh maaih dauh bou yietc_liuz gan ninh. this-time 3sg have CLF servant forever follow 3sg adv.tmp pn.p vt clf n adv.tmp vt pn.p

This time, he has got a permanent servant (to) help him.

BS.141 141.1 Ninh bou ziouc mingh gaatv maaz-miev, 3sg servant so_then go cut horse-grass pn.p n seq.mk vi n n-n 141.2 141.3 haiz maaih norm norqc-nqo daaih heuc gorngv, hear have CLF dove come order speak vt vt clf n vi v v "Maav cauv nuh nuh, wongh_waa zaix jaa kuh buv kuh." (Chinese) horse grass servant servant baby in house cry NEG cry n n n n n prep n vi neg vi 141.1Then his servant went (to) cut horse grass, 141.2(and he) heard (that) there was a dove come(ing) 141.3(and it) called (and) said, “Horse grass servant, servant! A baby is crying in the house or not?”

BS.142 142.1 142.2 Nv dauh bou ziouc hnyouv nzauh haic, mv gaatv maaz-miev, this CLF servant so_then heart sad very NEG cut horse-grass pn.dem clf n seq.mk n v.st adv neg n n-n 142.3 kungx muangx norqc heuc hnangv. only listen bird call only adv v n v adv 142.1So the servant became sad very much, 142.2(he) did not cut horse grass, 142.3(he) only listened to the bird cry(ing).”

BS.143 143.1 143.2 Yiem yietc hnoi hmuangx gaatv duqv yietc zeiv maaz-miev hnangv. be_in one day dark cut can one CLS horse-grass only vi nmb n v.st n v.aux nmb cls n-n adv 143.1In one day (till) dark, 143.2(he) could cut only one unit of horse grass.

Page 327: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

301 BS.144 144.1 144.2 Nzuonx daaih maaz maiv gaux nyanc. return come horse NEG enough eat v vi n neg v.st vt 144.1(When he) came back, 144.2the horse did not have enough food.

BS.145 Ninh loz-benv hemx ninh. 3sg employer scold 3sg pn.p n vt pn.p

His employer scolded him.

BS.146 Ninh gorngv, 3sg speak pn.p v "Nqaan-nziem camv haic, gaatv maiv duqv camv." prickly_thatch_grass many very cut NEG get many n v.st adv n neg vt adj

He said, “There are too many prickly thatch, I can’t cut much [horse grass].”

BS.147 147.1 147.2 147.3 Ninh loz-benv ziouc bun heh ninh daapc jienv mingh gaatv. 3sg employer so_then give shoes 3sg put_on SML.ACT go cut pn.p n seq.mk v n pn.p vt asp vi n 147.1So his employer gave him (a pair of) shoes 147.2(he) went wearing (them) 147.3(to) cut.

BS.148 148.1 Mingh taux gaatv maaz-miev nyei dorngx go reach cut horse-grass RELT place vi v n n-n relt n 148.2 aengx haiz wuov norm norqc aengx heuc. further hear that CLF bird again call adv vt pn.dem clf n adv v 148.1(He) went to the place to cut horse grass, 148.2(he) again heard that bird cry again.

BS.149 149.1 149.2 Ninh aengx muangx norqc heuc hnangv, mv maaih henh_hnyouv gaatv. 3sg again listen bird call only NEG have peaceful_heart cut pn.p adv v n v adv neg vt n vt 149.1He again listened to the bird cry(ing), 149.2(he) did not have a peaceful heart (to) cut (grass).

Page 328: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

302 BS.150 150.1 Yietc hnoi taux hmuangx duqv nduqc zeiv miev nzuonx hnangv, entire day reach dark get single paper grass return only adj n v adj vt adj n n v adv 150.2 maaz yaac maiv gaux nyanc. horse and NEG enough eat n top neg v.st vt 150.1The whole day till dark (he) got only one unit of grass (and) returned, 150.2and the horse did not have enough (to) eat.

BS.151 Ninh loz-benv aengx hemx ninh. 3sg employer again scold 3sg pn.p n adv vt pn.p

His employer scolded him again.

BS.152 Ninh aengx gorngv, 3sg again speak pn.p adv v "Njimv camv haic, gaatv maiv duqv." thorns many very cut NEG can n v.st adv n neg v.aux

He said again, “There are lots of thorns; I can’t cut (grass).”

BS.153 Ninh aengx bun lui-houx ninh. 3sg further give clothes 3sg pn.p adv vt n pn.p

He further gave him clothes.

BS.154 Da'nyeic hnoi ninh aengx zorqv jienv mingh gaatv. second day 3sg again take SML.ACT go cut n n pn.p adv vt asp vi n

The following day, he went taking (the clothes) (to) cut.

BS.155 155.1 155.2 Mingh taux hnangv, norqc aengx heuc aqv. go reach only bird again call PFT vi v adv n adv v part.asp 155.1As soon as they arrived there, 155.2the bird cried again.

Page 329: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

303 BS.156 156.1 Ninh aengx muangx norqc heuc, 3sg again listen bird call pn.p adv v n v 156.2 156.3 yietc hnoi hmuangx gaatv duqv nduqc zeiv maaz-miev dorh nzuonx daaih, entire day dark cut get single CLS horse-grass take return come adj n adj n vt adj cls n-n vt vi v.dr 156.4 maaz yaac maiv gaux nyanc. horse and_yet NEG enough eat n top neg v.st vt 156.1He listened to the bird again, 156.2the whole day till dark (he) could cut only one unit of horse grass 156.3 (and) took it back (home); 156.4and the horse did not have enough (to) eat.

BS.157 Ninh loz-benv aengx hemx ninh, 3sg employer further scold 3sg pn.p n adv vt pn.p "Meih zoux_haaix_nyungc maiv gaatv maaz-miev?" 2sg Why? NEG cut horse-grass pn.p q neg n n-n

His employer scolded him again, “Why do you not cut horse grass?”

BS.158 Ninh aengx gorngv, 3sg further speak pn.p adv v "Nyutc doqc haic, gaatv maiv duqv. sunlight fierce very cut NEG can n v.st adv n neg v.aux Yie kungx bingx laangh hnangv." 1sg only hide cool only pn.p adv vt v.st adv

He said again, “The sunshine is too severe, I can’t cut. I only hide myself in the shade.”

BS.159 Ninh loz-benv aengx bun muoc ninh ndongx. 3sg employer further give hat 3sg put_on pn.p n adv vt n pn.p v

His employer further gave him a hat (to) put on.

BS.160 Da'faam hnoi ninh aengx ndongx jienv muoc mingh gaatv. third day 3sg further put_on SML.ACT hat go cut adj n pn.p adv v asp n vi n

The third day he further went putting on the hat (and) cut.

Page 330: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

304 BS.161 161.1 161.2 Mingh yietc hnoi hmuangx yaac duqv nduqc zeiv maaz-miev hnangv, go entire day dark and_yet get single CLS horse-grass only vi adj n adj top vt adj cls n-n adv 161.3 dorh jienv nzuonx daaih. take SML.ACT return come vt asp v vi 161.1(he) went the whole day till dark 161.2yet (he) got only one unit of horse grass, 161.3returned home carrying (it).

BS.162 Ninh loz-benv aengx hemx ninh aqv. 3sg employer again scold 3sg PFT pn.p n adv vt pn.p part.asp

His employer scolded him again.

BS.163 163.1 163.2 Ninh ziouc maiv_lamh gorngv aqv, ninh cingx_daaih gorngv mbuox ninh loz-benv, 3sg so_then lack_a_place_for speak CHG-O-ST 3sg therefore speak tell 3sg employer pn.p seq.mk idm v part.asp pn.p conj v vt pn.p n "Mv zeiz haaix nyungc. NEG be what kind neg vi q n Weic_zuqc maaih norm norqc-nqo aah_loc daaih because have CLF dove frequently come conj vt clf n adv vi yiem yie gu'nguaaic ndiangx-dueiv heuc. be_there 1sg above tree-tail call vi pn.p adv.loc n-n vi Da'faanh yie taux, norqc ziouc heuc aqv. as_soon_as 1sg reach bird so_then call PFT adv.tmp pn.p v n seq.mk vi part.asp 163.1So he had no excuse to say anything; 163.2he therefore said to his employer, “There is nothing but because the bird came frequently and cries on a tree above me. As soon as I reach there, then the bird cries.”

BS.164 Ninh loz-benv naaic ninh, 3sg employer ask 3sg pn.p n vt pn.p "Norqc hnangv_haaix_nor heuc?" bird how call n adv vi

His employer asked him, “How did the bird cry?”

Page 331: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

305 BS.165 Ninh bou ziouc gorngv, 3sg servant so_then speak pn.p n seq.mk v "Norqc heuc, bird call n v 'maav cauv nuh nuh, wongh_waa zaix jaa kuh buv kuh’” (Chinese) horse grass servant servant baby in house cry NEG cry n n n n n prep n vi neg vi

His servant then said, “The bird cried ‘A horse grass servant, servant! Is the baby crying in the house?’”

BS.166 Ninh loz-benv ziouc hnamv taux gamh_nziex ninh auv. 3sg employer so_then think reach fear 3sg wife pn.p n seq.mk vt v v.cmp pn.p n

Then his employer thought that might be his wife.

BS.167 Ninh bou ziouc gorngv, 3sg servant so_then speak pn.p n seq.mk v "Meih mv sienx, njang_hnoi caux yie mingh. 2sg NEG believe tomorrow be_with 1sg go pn.p neg vt adv.tmp vt pn.p vi Ih_zanc mv_dungx gaengh daix yie. now NEG.CMD really kill 1sg adv.tmp neg adv vt pn.p Njang_hnoi meih caux yie mingh. Norqc maiv heuc m'daaih daix yie." tomorrow 2sg be_with 1sg go bird NEG call rightly kill 1sg adv.tmp pn.p vt pn.p vi n neg v adv vt pn.p

Then his servant said, “If you don’t believe (me), come with me tomorrow. Don’t kill me yet, now. Tomorrow, when you and I go together, if the bird still does not cry, then of course you can kill me.”

BS.168 Da'nyeic ndorm ninh loz-benv ziouc gan jienv ninh mingh aqv. second morning 3sg employer so_then follow DUR 3sg go PFT n n pn.p n seq.mk vt asp pn.p vi part.asp

The following morning his employer then went following him.

BS.169 169.1 169.2 Mingh taux ninh hnoi-hnoi yiem nyei dorngx norqc yaac maiv heuc aqv. go reach 3sg day-day live RELT place bird and_yet NEG call CHG-O-ST vi v pn.p n-n vi relt n n top neg vi part.asp 169.1(They) went to the place where he stayed every day, 169.2(but) the bird wouldn’t make its call.

Page 332: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

306 BS.170 Ninh loz-benv gorngv, 3sg employer speak pn.p n v "Hmx, meih gorngv norqc heuc. Ih_hnoi yaac maiv heuc aqv." well...2sg speak bird cry oday and_yet NEG call CHG-O-ST intj pn.p vt n vi adv.tmp top neg v part.asp

His employer said, “Well…you said that a bird cries. But today it doesn’t.”

BS.171 Loz-benv ziouc baeng cuotv nzuqc daaih oix daix ninh aqv. employer so_then pull go_out knife come want_to kill 3sg INCHOAT n seq.mk vt v n v.asp v.mod vt pn.p part.asp

So the employer pulled out a knife (and) wanted to kill him.

BS.172 Ninh bou gorngv, 3sg servant speak pn.p n v "Mv_dungx gaengh daix. Zuov dangh yie." NEG.CMD really kill wait_for awhile 1sg neg adv vt vt adv pn.p

His servant said, “Don’t kill me yet! Wait for a second.”

BS.173 Ninh ziouc heuc, 3sg so_then call pn.p seq.mk vi "Meih naaic norm norqc, maaih lingh maaih singx gaanv daaih heuc. 2sg that CLF bird have power have spirit hurry come call pn.p pn.dem clf n vt n vt n vi vi v Meih hnoi-hnoi heuc nduov yie hnyouv nzauh, 2sg day-day call deceive 1sg heart sad pn.p n-n vi vt pn.p n v.st gaatv maiv duqv maaz-miev bun maaz nyanc. cut NEG can horse-grass give horse eat n neg v.aux n-n vt n vt Ih_zanc gaanv daaih heuc bun loz-benv muangx. Mv_zei_aeqv, ninh daix yie aqv." now hurry come cry let employer listen otherwise 3sg kill 1sg INCHOAT adv.tmp vi vi vi vt n v adv pn.p vt pn.p part.asp

Then he called out, “You, that bird! If you have life and spirit, come quickly and cry. You cried and entice me to be sad everyday, making me unable to cut grass for horses to eat. Now, come out quickly to let my employer listen to you. Otherwise, he is about to kill me.”

Page 333: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

307 BS.174 174.1 174.2 Wuov_nzunc_hnoi norqc za'gengh ndaix daaih nzopc jienv gu'nguaaic ndiangx-nquaah this-time bird really fly come sit_on_a_tree_branch DUR above branch adv.tmp n adv vi vi vi asp adv.loc n 174.3 ziouc yiem wuov heuc aqv. so_then be_there there cry PFT seq.mk vi pn.dem v part.asp 174.1This time the bird actually flew in 174.2(and it) sat on a branch above, 174.3then cried there.

BS.175 Nv dauh loz-benv ziouc gorngv, this CLF employer so_then speak pn.dem clf n seq.mk v "Meih naaic norm norqc zien-zien yie nyei auv nor, 2sg that CLF bird really 1sg RELT wife as pn.p pn.dem clf n adv pn.p relt n top ndaix daaih nzopc yie nyei buoz-zaangv." fly come sit_on 1sg POSS palm vi vi vi pn.p poss n

This employer then said, “If you are really my wife, fly over here and sit on my palm.”

3sg so_then take SML.ACT return PFT

BS.176 176.1 176.2 Norqc ziouc za'gengh ndaix daaih nzopc jienv ninh nyei buoz-zaangv, bird so_then really fly come sit_on DUR 3sg POSS palm n seq.mk adv vi vi vi asp pn.p poss n 176.3 176.4 ziouc gan lui-mueiz-kuv bieqc daaih, bieqc lui-mbuoqc mi'aqv. so_then follow opening_end_of_sleeve enter come enter coat_pocket RSLT seq.mk vt n vi vi vi n asp 176.1Then the bird really flew in, 176.2(it) sat on his palm; 176.3then following the sleeve came in, 176.4entered a pocket.

BS.177 Ninh ziouc dorh jienv nzuonx aqv.

pn.p seq.mk vt asp v part.asp

So he took it home.

BS.178 178.1 178.2 178.3 178.4 Dorh nzuonx daaih zieqv longh wuonx jienv, aeqv kuangx jienv domh gaengh ndaangc. take return come weave cage confine DUR so_then hang_on_hook DUR big door front vt vi vi vt n vt asp top vt asp adj n n 178.1(He) took (it) (and) returned home 178.2(he)wove a cage 178.3(he) kept (it in the cage); 178.4and then hung it over at the big front door.

Page 334: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

308 BS.179 179.1 Da'nyeic ndorm ninh auv dorh jienv guh_nguaaz jiez sin daaih, second morning 3sg wife take SML.ACT baby raise body come n n pn.p n vt asp n vt n v.asp 179.2 179.3 mv gaengh gan guh_nguaaz nzaaux hmien, guh_nguaaz nyiemv. NEG yet help baby wash face baby weep neg adv vt n vt n n vi 179.1The following morning, his wife got up carrying the baby; 179.2before she helped wash her baby’s face, 179.3the baby cried.

BS.180 180.1 180.2 Ninh auv nduov mv guai aeqv, dorh mingh bun guh_nguaaz mangc norqc, 3sg wife deceive NEG clever TOP take go let baby look bird pn.p n vt neg v.st part vt vi v n v n 180.3 mingh duqv fatv haic, go get near very vi vt v.st adv 180.4 180.5 180.6 norqc mangc buatc korh_lienh haic guh_nguaaz m'zing-nqaiv yaac camv haic. bird look see feel_sorry_for very baby eye_mucus TOP many very n v v vt adv n n part v.st adv 180.1His wife was not good at soothing the baby, 180.2(she) took the baby (and) let him look at the bird; 180.3(the baby) went very close (to the bird); 180.4the bird looked (at the baby) 180.5(and) felt very sorry 180.6the baby had a lot of eye mucus.

BS.181 Ninh maiv gan ninh nyei guh_nguaaz nzaaux hmien. 3sg NEG help 3sg POSS baby wash face pn.p neg vt pn.p poss n vt n

She didn’t help her baby wash his face.

BS.182 182.1 182.2 Ninh ziouc yiem longh gu'nyuoz ndeng jaang cuotv daaih 3sg so_then be_there cage inside thrust_out neck come_out come pn.p seq.mk vi n n vt n v v.dr 182.3 182.4 gan ninh nyei guh_nguaaz nzemx m'zing-nqaiv. help 3sg POSS baby pick eye_mucus vt pn.p poss n vt n 182.1Then she from inside the cage 182.2thrusted out her neck 182.3(and) help her baby 182.4picked eye-mucus.

BS.183 183.1 Yietv ndeng jaang cuotv daaih, as_soon_as thrust_out neck come_out come adv vt n v v.dr 183.2 nyorqv ninh guh_nguaaz m'zing-nqaiv hnangv. peck 3sg baby eye_mucus only v pn.p n n adv 183.1As soon as (she) thrusted out her neck, 183.2(she) only pecked her baby’s eye mucus.

Page 335: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

309 BS.184 Ninh auv gorngv, 3sg wife speak pn.p n v "Janx-daic norqc nyorqv maengh guh_nguaaz m'zing hnangv." dead_person bird peck blind baby eye only n n v v.st n n adv

His wife said, “That ghost bird is only pecking my baby’s eye blind.”

BS.185 Ninh ziouc ziang_naaic mborqv norqc daic mi'aqv. 3sg so_then just beat bird die RSLT pn.p conj adv vt n vi asp

Then she just beat the bird to death.

BS.186 186.1 Wuov_nzunc_hnoi ninh nqox aengx zorqv daaih hliangv nyanc, this-time 3sg husband further take come dress_meat eat adv.tmp pn.p n adv vt vi v vt 186.2 aengx bun ninh auv nyanc. further give 3sg wife eat adv vt pn.p n vt 186.1This time her husband further took (the dead bird), dressed the meat (to) eat, 186.2(he) further gave his wife to eat.

BS.187 187.1 187.2 Ninh nqox nyanc aeqv, kuv nyei. 3sg husband eat TOP delicious PRS.ST pn.p n vt part v.st asp 187.1When her husband ate (it), 187.2(it) tasted delicious.

BS.188 188.1 188.2 Ninh auv nyanc aeqv, maiv kuv. 3sg wife eat TOP NEG delicious pn.p n vt part neg v.st 188.1When his wife ate (it), 188.2(it) was not delicious.

BS.189 189.1 189.2 189.3 189.4 Ninh nqox nyanc gaam nyei, ninh auv nyanc niangv im nyei, 3sg husband eat sweet AFM 3sg wife eat very bitter AFM pn.p n vt v.st asp pn.p n vt adv v.st asp 189.5 189.6 yaac sopv nyei, naqv maiv njiec jaang. also rough AFM swallow NEG go_down throat top v.st asp vt neg vi n 189.1(When) her husband ate (it) 189.2(it was) sweet; 189.3(when) his wife ate (it), 189.4(it was) very bitter 189.5and also rough 189.6(so she) could not swallow (it).

Page 336: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

310 BS.190 190.1

further go_out CLS bitter_vegitable big come

Ninh nqox aengx gaeqv daaih zouv nyanc, yaac kuv nyei. 3sg husband further pick come cook eat and delicious PRS.ST

Ninh auv zorqv pietv guangc mingh, 3sg wife take sprinkle throw_away go pn.p n vt vt vt v.dr 190.2 aengx cuotv zungh janx-biau_lai, hlo daaih.

adv v cls n v.st v.asp 190.1His wife took, sprinkled, threw (it) away, 190.2(it) further sprouted a bitter vegitable (and it) grew.

BS.191 191.1 191.2 Ninh nqox aengx gaeqv daaih zouv nyanc yaac kuv nyei gaam nyei. 3sg husband again pick come cook eat and delicious PRS.ST sweetAFM pn.p n adv vt v.dr vt vt top v.st asp v.st asp 191.1Her husband again picked (it) up (and) cooked (and) ate (it), 191.2and (it) was delicious (and) sweet.

BS.192 192.1 192.2 192.3 Ninh auv nyanc yaac maiv kuv, im nyei, 3sg wife eat and_yet NEG delicious bitter PRS.ST pn.p n vt top neg v.st v.st asp 192.4 aengx pietv guangc mingh. further sprinkle throw_away go adv vt vt v.dr 192.1His wife ate 192.2but not delicious, 192.3(it was) bitter; 192.4(she) threw (it) away again.

BS.193 Aengx cuotv zungh lai-jaaix daaih. further come_out CLS mustard_green come adv v cls n v.asp

Again (it) came out as a mustard green.

BS.194 194.1 194.2 194.3

pn.p n adv vt vi vt vt top v.st asp 194.1.Her husband picked (it) again, 194.2cooked (and) ate (it); 194.3and (it) was delicious.

BS.195 195.1 195.2 195.3 Ninh auv nyanc yaac maiv kuv, im nyei. 3sg wife eat and_yet NEG delicious bitter PRS.ST pn.p n vt top neg v.st v.st asp 195.1His wife ate (it) 195.2and (it was) not delicious, 195.3(it was) bitter.

Page 337: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

311 BS.196 196.1 Ninh aengx pietv guangc kamx-buoqv mingh, 3sg again sprinkle throw_away lower_side_of_a_house go pn.p adv vt vt n v.dr 196.2 aengx cuotv diuh hlauv daaih. further come_out CLF bamboo come adv v clf n v.dr 196.1She sprinkled, threw (it) away again to the lower side of the house, 196.2(it) further became a bamboo.

BS.197 Wuov_nzunc_hnoi ninh nqox maiv gaeqv aqv. this-time 3sg husband NEG pick CHG-O-ST adv.tmp pn.p n neg vt part.asp

This time her husband did not pick (it).

BS.198 Ninh nqox ziouc weih jienv aqv. 3sg husband so_then surround DUR PFT pn.p n seq.mk vt asp part.asp

Her husband then enclosed (it).

BS.199 199.1 Nqa'haav nv diuh hlauv hlang daaih, later this CLF bamboo bez_high come adv.tmp pn.dem clf n v.st v.asp 199.2 199.3 hlauv mbu'ndongx duc!-duc_wuov, ninh yaac maiv gaeqv. bamboo middle swollen_up_big 3sg also NEG pick n n v.st pn.p top neg vt 199.1Later (when) this bamboo grew tall, 199.2its middle part became swollen up; 199.3he did not cut it.

BS.200 Wuov deix maaih dauh janx-muangz m'gux. those_people have CLF non-Mienh_woman old_woman pn.dem vt clf n n

Those people had a non-Mienh old woman.

BS.201 Ndorm-ndorm daaih mv dauh loz-benv nyei biauv zorqv douz. morning-morning come this CLF employer POSS house take fire n-n vi pn.dem clf n poss n vt n

Every morning (she) would come to this employer’s house (to) get fire.

BS.202 202.1 202.2 Zorqv taux nqa'haav_laai wuov ndorm nziouv nyei daaih zorqv. take reach last that morning early ADVBLZR come take vt v adj pn.dem n v.st part vi vt 202.1(As she) took (the fire) till the last morning, 202.2(she) came to take very early.

Page 338: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

312 BS.203 Zorqv jienv nzuonx taux naaic diuh hlauv-gorn buo nzunc. take SML.ACT return reach that CLF bamboo-root three time vt asp v v pn.dem clf n-n numb n

(She was) taking (the fire), went back as far as the bamboo root three times.

BS.204 Wuov_nzunc nv dauh loz-benv ziouc gorngv, this_time this CLF employer so_then speak adv.tmp pn.dem clf n seq.mk v "Meih naaic bun douz-nangh, meih aeqv oix daic haic. 2sg TOP let fire-alive 2sg so_then want die very pn.p pn.dem vt n-v.st pn.p top v.mod vi adv Nv_nzunc bun douz-taanx daic, this_time let charcoal die adv.tmp v n vi meih mangc gaax dorh mingh buov duqv zieqc mv zieqc." 2sg look try_and_see take go burn can burn_up NEG burn_up pn.p v part vt vi vt v.aux v neg v

This time, this employer then said, “You! (I) gave (you) the alive charcoal, but you!, you always let (it) die. This time (I will) give you the cold charcoal. You take it with you, try and see if it can burn or not.”

BS.205 Nv dauh m'gux ziouc dorh jienv douz-taanx daic nzuonx aqv. this CLF old_woman so_then take SML.ACT charcoal die return PFT pn.dem clf n seq.mk vt asp n vi v part.asp

So this old woman went back taking the dead charcoal.

BS.206 206.1 206.2 Nzuonx taux nv diuh hlauv-gorn ziouc nangh daaih. return reach this CLF bamboo-root so_then alive come v v pn.dem clf n-n seq.mk v.st vi 206.1(When she) returned as far as this bamboo root, 206.2(the charcoal) became alive again.

BS.207 207.1 207.2 Nv dauh m'gux ziouc dorh nzuonx mingh buov zieqc douz daaih. this CLF old_woman so_then take return go burn burn_up fire come pn.dem clf n seq.mk vt v vi vt v n v.asp 207.1Then this old woman took (it) returned home, 207.2(and) lit the fire.

BS.208 208.1 208.2 208.3 Ninh aengx daaux_nqaang daaih goix nv diuh hlauv dorh nzuonx mingh 3sg again return come hack this CLF bamboo take return go pn.p adv vi vi vt pn.dem clf n vt v vi 208.4 208.5 208.6 paaix nqoi daaih ziouc buatc naaic dauh auv yiem gu'nyuoz. slice_open open come so_then see that CLF wife be_in inside vt vi v.asp seq.mk v pn.dem clf n vi n 208.1She returned again 208.2(and) hacked this bamboo 208.3(and) brought (it) back; 208.4(and) split it open 208.5then saw that the wife 208.6was inside.

Page 339: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

313 BS.209

Then she took (her and) kept (her) in the woven basket.

Buatc dauh sieqv yiem mv dauh m'gux nyei biauv.

v.st vi top neg v 215.1(It became) all dark, et (he) would not go home.

3sg speak want_to beat 3sg INCHOAT pn.p v v.mod vt pn.p part.asp

He said he wanted to beat him.

Ninh ziouc dorh mingh dapv jienv kapv gu'nyuoz aqv. 3sg so_then take go put_into DUR woven_basket_with_a_lid inside PFT pn.p seq.mk vt vi vt asp n n part.asp

BS.210 Ninh nyei douz yietc_liuz mv daic aqv. 3sg POSS fire forever NEG die CHG-O-ST pn.p poss n adv.tmp neg vi part.asp

Her fire would not die permanently.

BS.211 Nqa'haav nv dauh poux-ziouv nyei dorn mingh wuov dauh m'gux nyei biauv a'nziaauc. later this CLF shop_owner POSS son go that CLF old_woman POSS house play adv.tmp pn.dem clf n poss n vi pn.dem clf n poss n v

Later a son of the shop owner went to that old woman's house for a visit.

BS.212

see CLF daughter live this CLF old_woman POSS house v clf n vi pn.dem clf n poss n

(He) saw a daughter live(ing) in this old woman’s house.

BS.213 213.1 213.2 Mv_baac ninh nziaauc nziaauc yaac nzuonx nyei. but 3sg visit visit and return AFM conj pn.p vi vi top v asp 213.1But (though) he visited her repeatedly, 213.2(he) still would go home.

BS.214 Nqa'haav aeqv, maiv nzuonx aqv. later TOP NEG return CHG-O-ST adv.tmp part neg v part.asp

Later, he wouldn't go home.

BS.215 215.1 215.2 Hmuangx nzengc yaac maiv nzuonx. dark consumed and_yet NEG return

215.2y

BS.216 Ninh die mingh lorz nzuonx daaih. 3sg father go look_for return come pn.p n vi vt v vi

His father went (to) look for (and brought him) came back.

BS.217 Ninh gorngv, oix mborqv ninh aqv.

Page 340: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

314 BS.218 Ninh gorngv, 3sg speak pn.p v "Mv_dungx mborqv. Wuov dauh m'gux nyei biauv maaih dauh don't beat that CLF old_woman POSS house have CLF neg vt pn.dem clf n poss n vt clf m'sieqv_dorn nzueic haic, hnangv yie nyei maa nor. woman beautiful very like 1sg RELT mother as n v.st adv adv pn.p relt n top Yiem wuov dauh m'gux wuov ndatv ndie. be_there that CLF old_woman that_place weave cloth vi pn.dem clf n pn.dem vt n Ninh heuc yie tengx ninh zorqv ga'naaiv bun ninh, 3sg order 1sg help 3sg take thing give 3sg pn.p v pn.p vt pn.p vt n vt pn.p zorqv wuom bun ninh hopv. take water give 3sg drink vt n vt pn.p vt Ninh ndatv ndie ninh nyei suix aah_loc ndortv. 3sg weave cloth 3sg POSS thread frequently fall pn.p vt n pn.p poss n adv vi Ninh heuc yie tengx ninh zorqv suix lo_haaix." 3sg call 1sg help 3sg take thread and_so_forth pn.p v pn.p vt pn.p vt n n

He said, “Don’t beat (me). In that old woman’s house there is a beautiful woman just like my mother. She weaves clothes in that old woman’s place. She tells me to pick up things to hand over to her, or give water for her to drink. When she weaves cloth her thread often falls [on the floor]. She would tell me to pick it up to give her and to help other things.”

BS.219 "Wuov nyungc, meih buatc hnangv nyei?" that kind 2sg see resemble q pn.dem n pn.p v vi part

“Then, do you see that she looks like your mother?”

BS.220 Ninh gorngv, "Hnangv nyei." 3sg speak resemble PRS.ST pn.p v vi asp

He said, “Looks like.”

BS.221 221.1 221.2 Wuov dauh m'gux maiv maaih haaix nyungc, maaih dauh dungz-gong hnangv. that CLF old_woman NEG have what kind have CLF castrated_pig only pn.dem clf n neg vt q n vt clf n adv 221.1That old woman did not have anything, 221.2(she) only had a castrated pig.

Page 341: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

315 BS.222 222.1 Wuov dauh loz-benv ziouc daav za'eix that CLF employer so_then plan plan pn.dem clf n seq.mk vt n 222.2 222.3 gorngv mbuox wuov dauh m'gux daix dungz nyanc sih_nyietv-hnaangx. speak tell that CLF old_woman kill pig eat birthday_meal v vt pn.dem clf n vt n vt n 222.1So the employer made a plan 222.2(and) told the old woman 222.3(to) kill a pig (to) eat a birthday meal.

BS.223 Mv_zei_aeqv, maiv maaih za'eix aqv. otherwise NEG have plan CHG-O-ST adv neg vt n part.asp

Otherwise, there was no way.

BS.224 Ninh ziouc gorngv mbuox wuov dauh m'gux daix dungz nyanc sih_nyietc. 3sg so speak tell that CLF old_woman kill pig eat birthday_meal pn.p seq.mk v vt pn.dem clf n vt n vt n

He then told that old woman to kill a pig to have a birthday feast.

BS.225 225.1 Wuov dauh m'gux mv_haih_fih_nyungc_zoux, that CLF old_woman can't_do_anything pn.dem clf n idm 225.2 225.3 oix_zuqc muangx ninh nyei waac, ziouc daix mbenc hnaangx nyanc. must listen 3sg POSS word so_then kill prepare meal eat aux v pn.p poss n seq.mk vt vt n vt 225.1The old woman had nothing to do (about it), 225.2(she) had to listen to his words, 225.3then killed (the pig and) prepared meal (to) eat.

BS.226 226.1 Mbenc ziangx kuv lai kuv hnaangx, prepare finish delicious vegetable delicious rice vt vi v.st n v.st n 226.2 yiem dieh gorn ziouc daaih heuc ninh mingh nyanc hnaangx. be_in table base so_then come call 3sg go eat meal vi n n seq.mk vi v pn.p vi vt n 226.1(When she) finished preparing the delicious vegetables (and) delicious rice, [i.e. finished preparing a delicious meal or a feast] 226.2being at the base of the table (she) told him (to) come to have a meal.

BS.227 Ninh ziouc mingh aqv. 3sg so_then go PFT pn.p seq.mk vi part.asp

So he went.

Page 342: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

316

BS.228 228.1 Mingh taux nyei ziangh_hoc, go reach RELT time vi v relt n 228.2 228.3 wuov dauh m'gux, weic_zuqc dieh dueiv maaih norm kapv, that CLF old_woman because table tail have CLF woven_basket_with_a_lid pn.dem clf n conj n n vt clf n 228.4 ninh nyei cien auv yiem naaic kapv gu'nyuoz. 3sg POSS kin wife be_in that woven_basket_with_a_lid inside pn.p poss n n vi pn.dem n n 228.1When (he) went there, 228.2the old woman, 228.3because (she) had a woven basket at the tail of the table, 228.4his real wife is inside the woven basket.

BS.229 Ninh cingx_daaih oix zueiz nitv naaic norm kapv. 3sg therefore want_to sit close_to that CLF woven_basket_with_a_lid pn.p conj v.mod vi vi pn.dem clf n

That’s why he wanted to sit close to that woven basket.

BS.230 230.1 Wuov dauh m'gux gorngv mv bun ninh zueiz dieh dueiv, that CLF old_woman speak NEG let 3sg sit table tail pn.dem clf n v neg v pn.p vi n n 230.2 ninh yaac maiv muangx. 3sg and_yet NEG listen pn.p top neg v 230.1The old woman said (that she would) not let him sit at the tail of the table, 230.2and yet he would not listen (to her).

BS.231 231.1 231.2 Wuov dauh m'gux maiv_haih_fih_nyungc_zoux, oix_zuqc bun ninh zueiz. that CLF old_woman can't_do_anything must let 3sg sit pn.dem clf n idm aux v pn.p vi 231.1The old woman could not do anything, 231.2(she) had to let him sit (there).

Page 343: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

317 BS.232 232.1 Ninh ziouc zueiz nitv jienv nv norm kapv, 3sg so_then sit be_close_to DUR this CLF woven_basket_with_a_lid pn.p seq.mk vi vi asp pn.dem clf n 232.2 232.3 yiem wuov nyanc hnaangx, yietc binc nyanc hnaangx, be_there there eat meal one time eat meal vi pn.dem vt n nmb vt vt n 232.4 232.5 232.6 wuov dauh auv yietc binc longc sim baqv ninh. there CLF wife one time use needle prick 3sg pn.dem clf n nmb vt vt n vt pn.p 232.1So he sat close to this woven basket, 232.2(he) ate food there; 232.3one time (he) ate food, 232.4one time the wife 232.5used a needle 232.6(and) pricked him [i.e. each time he ate food, the wife pricked him with a needle].

BS.233 233.1 233.2 Baqv da'yietv nzunc ninh yaac maiv kolo. prick first time 3sg and_yet NEG make_a_noise vt nmb n pn.p top neg vi 233.1(When she) pricked (him) first time, 233.2he didn’t say anything.

BS.234 234.1 234.2 Baqv da'nyeic nzunc ninh ziouc huin nqaang mangc. prick second time 3sg so_then turn_around behind look vt n n pn.p seq.mk vi n vt 234.1(She) pricked (him) second time, 234.2then he turned around (and) looked.

BS.235 235.1 235.2 Mangc deix ziouc gorngv, look some so_then speak v adv seq.mk v "Nv norm kapv maaih haaix nyungc ga'naaiv mv_bei, danx yie." this CLF woven_basket have what kind thing I_wander sting 1sg pn.dem clf n vt q n n part.f vt pn.p 235.1(He) looked (at it) for a while, 235.2then said, “I wonder what this basket has (in it); it stung me.”

BS.236 Wuov dauh m'gux gorngv, that CLF old_woman speak pn.dem clf n v "Mv maaih haaix nyungc, kapv kungx hnangv." NEG have what kind woven_basket_with_a_lid empty only neg vt q n n v.st adv

The old woman said, “Nothing. The woven basket is just empty.”

BS.237 237.1 237.2 Ninh maiv gorngv haaix nyungc, aengx nyanc hnaangx aqv. 3sg NEG speak what kind again eat meal CHG-O-ST pn.p neg v q n adv vt n part.asp 237.1He didn't say anything 237.2(but) started to eat the meal again.

Page 344: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

318 BS.238 238.1 238.2

BS.242

neg vt q n pn top vt adv BS.243 243.1 243.2

243.3 243.4

Da'faam nzunc haiz baqv ninh ziouc maiv muangx, third time feel prick 3sg so_then NEG listen adj n vi vt pn.p seq.mk neg v 238.3 238.4 jiez sin daaih oix koi kapv mangc. raise body come want_to open woven_basket_with_a_lid look vt n v.dr v.mod vt n vt 238.1The third time (he) felt (being) pricked, 238.2then he did not listen (to the old woman), 238.3stood up 238.4(and) wanted (to) open the woven basket (to) look (inside).

BS.239 Wuov dauh m'gux gorngv, that CLF old_woman speak pn.dem clf n v "Maiv maaih haaix nyungc." NEG have what kind neg vt q n

The old woman said, “There is nothing in it.”

BS.240 Wuov dauh m'gux ziouc qam jienv, mv bun koi. that CLF old_woman so_then hold DUR NEG let open pn.dem clf n seq.mk vt v.st neg vt vt

Then the old woman embraced (the basket), not letting (him) open.

BS.241 Ninh yaac maiv muangx. 3sg also NEG listen pn.p top neg v

He also would not listen.

Ninh gorngv, 3sg speak pn.p v "Mv maaih haaix nyungc ninh yaac mangc dangh." NEG have what kind 3sg also look awhile

Ninh ziouc koi nqoi daaih ziouc buatc ninh auv yiem gu'nyuoz, 3sg so_then open open come so_then see 3sg wife be_in inside pn.p seq.mk vt vi v.asp seq.mk vt pn.p n vi n

ninh ziouc dorh jienv nzuonx aqv, maiv nyanc hnaangx aqv. 3sg so_then take SML.ACT return INCHOAT NEG eat meal CHG-O-ST pn.p seq.mk vt asp v part.asp neg vt n part.asp 243.1So he opened (it) 243.2then saw his wife was inside; 243.3then he (decided to) take her (to) go home, 243.4(he) didn’t eat the meal any more.

Page 345: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

319 BS.244 244.1 Nzuonx taux biauv nyei ziangh_hoc,

Ninh mv maaih ih feix.

BS.249

return reach house RELT time v v n relt n 244.2 ninh dorc buatc ninh nziez nzueic haic. 3sg elder_sister see 3sg younger_sister beautiful very pn.p n v pn.p n v.st adv 244.1When (they) returned home, 244.2her elder sister saw (that) her younger sister was very beautiful.

BS.245 Ninh buatc ninh nziez torqv_saeng nzuonx daaih. 3sg see 3sg younger_sister be_reincarnated return come pn.p vt pn.p n vi v v.asp

She saw her younger sister had been reincarnated (to) come back.

BS.246

3sg NEG have excuse pn.p neg vt n

She didn’t have an excuse.

BS.247 Ninh gorngv, 3sg speak pn.p v "O M'Faam aac, Oh the_third_daughter listen_please intj n.prp part.voc

meih naaic nzueic haic niaa. 2sg TOP beautiful very I-am-thinking pn.p pn.dem v.st adv part.f

Meih hnangv_haaix_nor noic duqv ndongc naaic nzueic?" 2sg how possible can as_much_as that beautiful pn.p adv vi v.aux adv pn.dem v.st

She said, “Oh, M’Faam, you are very beautiful. How could you become as beautiful as this?”

BS.248 Ninh nziez-M'Faam ziouc gorngv, 3sg the_third_sister so_then speak pn.p n seq.mk v "Meih buatc baeqc nyei nzueic nyei?" 2sg see white PRS.ST beautiful is_it? pn.p vt v.st asp v.st q

Her younger sister, the third one, then said, “You see I am white; do you think I am beautiful?”

Ninh dorc gorngv, "Nzueic nyei." 3sg elder_sister speak beautiful PRS.ST pn.p n v v.st asp

Her sister said, “You are beautiful.”

Page 346: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

320 BS.250 Ninh nziez gorngv, 3sg younger_sister speak pn.p n v "Meih oix_longc baeqc nzueic aeqv, 2sg want white beautiful TOP pn.p v.cmp v.st v.st top hnyutv buo pan wuom mbueix aqv." boil three CLS water boil CAS.CMD vt numb cls n vi part.mod

Her younger sister said, “If you want (to be) white (and) beautiful, boil three buckets of water.”

gamh_nziex tongv biortc nyei mv_bei."

pn.p n vt n-vi vt v.dr

BS.251 Ziouc mbuox ninh nziez gorngv, so_then tell 3sg younger_sister speak seq.mk vt pn.p n v "Ziangx aqv. finish PFT vi part.asp

Then she told her younger sister, “Finished.”

BS.252 Ninh nziez ziouc mbuox ninh dorc, 3sg younger_sister so_then tell 3sg elder_sister pn.p n seq.mk vt pn.p n "Meih bieqc tongv gu'nyuoz mingh mangc gaax, 2sg enter barrel inside go look try_and_see pn.p vi n n vi v part

maybe barrel break_open PRS.ST I_guess v.cmp n vi asp part.f

Then her younger sister told her elder sister, “You get into the barrel to check it; Perhaps the barrel might be broken, I wonder.”

BS.253 Ninh dorc ziouc bieqc gu'nyuoz mingh mangc, gorngv, 3sg elder_sister so_then enter inside go look speak pn.p n seq.mk vi n v.dr v v "Maiv biortc oc, maiv haih cuotv wuom." NEG break_open PLT NEG be_possible_to let_go_out water neg vi part adv aux vt n

So her elder sister got inside (and) saw, said, “It’s not broken. Water won’t leak.”

BS.254 Ninh nziez ndamv wuom-mbueix dox njiec. 3sg younger_sister ladle_out water-boil pour let_go_down

Her younger sister ladled boiled water (and) poured into (the barrel).

BS.255 255.1 255.2 Ninh dorc ziouc yiem wuom-tongv gu'nyuoz tinx daic mi'aqv. 3sg elder_sister so_then be_there water-barrel inside thrash_about die RSLT pn.p n seq.mk vi n-n n vi vi asp 255.1Then her elder sister, in the water barrel, 255.2thrashed about (and) died.

Page 347: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

321 BS.256 256.1 256.2

BS.258

The story of the Big Snake has finished.

Ninh dorc aeqv, hluqv daic, 3sg elder_sister TOP scald die pn.p n part vt vi 256.3 maiv haih benx yietc nyungc aqv. NEG be_possible_to be one kind PFT neg aux vi nmb n part.asp 256.1As for her elder sister, 256.2(she) was boiled to death; 256.3(she) didn't become anything.

BS.257 257.1 257.2 Ninh nziez daic buo nzunc, aengx haic panh_nangh daaih. 3sg younger_sister die three time further be_possible come_back_to_life come pn.p n vi numb n adv vt vi v.dr 257.1Her younger sister died three times 257.2and could come back to life again.

258.1 258.2 Ninh dorc daic yietc nzunc, yaac maiv haih benx yietc nyungc gaeng-gueiv. 3sg elder_sister die one time and_yet NEG be_able_to be one kind insect pn.p n vi nmb n top neg aux vi nmb n n 258.1Her elder sister died once, 258.2but could not become (even) one kind of bug.

BS.259 Domh naang gouv liuz aqv. big snake story pass PFT adj n n vi part.asp

Page 348: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

BIBLIOGRAPHY103

Aarts, Bas. 2003. Conceptions of gradience in the history of linguistics. Language

Sciences 26.343-89.

Abbott, H. Porter. 2002. The Cambridge introduction to narrative. Cambridge:

Cambridge University Press.

Achard, Michel. 1996. French modals and speaker control. Conceptual structure,

discourse and language, ed. by Adele E. Goldberg, 1-30. Stanford, CA:

CSLI Publications.

Aikhenvald, Alexandra Y. 2005. Serial verb constructions. Oxford: Oxford

University Press.

Almeida, Michael Joseph. 1988. Reasoning about the temporal structure of

narratives. Ph.D. dissertation, State University of New York at Buffalo.

__________. 1998. Representation of cognitive structures: Syntax and semantics

of French sentential complements. Berlin: Mouton de Gruyter.

__________. 2002. The meaning and distribution of French mood inflections.

Grounding: The epistemic footing of deixis and reference, ed. by Frank

Brisard, 197-249. Berlin: Mouton de Gruyter.

Adams, Marilyn Jager, and Allan Collins. 1979. A schema-theoretic view of

reading. New directions in discourse processing, Advances in discourse

processes II, 1-22. Norwood, New Jersey: ABLEX Publishing

Corporation.

Aikhenvald, Alexandra Y. and R.M.W. Dixon (eds.) 2001. Areal diffusion and

genetic inheritance: Problems in comparative linguistics. Oxford: Oxford

University Press.

103 Thai authors are alphabetized by the first name occurring in the entry. Chinese and Japanese authors are

alphabetized by the family name. Materials cited from websites follow the format of Xia and Crane. 1996. Electronic styles: A handbook for citing electronic information. 2nd ed. Medford, N.J.: Information Today, Inc.

322

Page 349: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

323

Andersen, Francis I. 1974. The sentence in Biblical Hebrew. The Hague: Mouton.

__________. 1994. Salience, implicature, ambiguity, and redundancy in clause—

clause relationships in Biblical Hebrew. Biblical Hebrew and discourse

linguistics, ed. by Robert D. Bergen, 99-116. Dallas: SIL/ Winona Lake:

Eisenbrauns.

Andersen, Gisle. 1998. The pragmatic marker like from a Relevance-theoretic

perspective. Discourse markers: Descriptions and theory, ed. by Andreas

H. Jucker and Yael Ziv, 14770. Amsterdam: John Benjamins.

Anderson, John R. 1995. Cognitive psychology and its implications, 4th ed. New

York: W. H. Freeman and Company.

Aumann, Greg, Misako Aumann, and Pienh Muih Kua. 2002. Hezhou Iu-Mienh

English Chinese dictionary with English—Hezhou Iu-Mienh index. 欧瑞

文, 石井美纱子, 盘美花. 2002 年,《贺洲优勉英汉小词典》, Nanning

City, Guangxi: privately published.

Ariel, Mira. 1998. Markers and form-function correlations. Discourse markers:

Descriptions and theory, ed. by Andreas H. Jucker and Yael Ziv, 223-21.

Amsterdam: John Benjamins.

Arisawa, T. Daniel. 2006. Textlinguistics storyline analysis and its implications to

translation of Genesis 1:1. n.p.

Arnold, Jennifer E., and Thomas Wasow. 2000. Heaviness vs. newness: The

effects of structural complexity and discourse status on constituent

ordering. Language 76: 1. 28-55.

Aumann, Greg, and Paul Sidwell. 2001. Subgrouping of Mienic languages: some

observations. Paper read at SEALS XI, Bangkok, May 2001.

Aziz, Yowell Y. 1988. Cohesion in spoken Arabic texts. Pragmatics, discourse

and text: Some systemically-inspired approaches, ed. by Erich H. Steiner

and Robert Veltman, 148-57. London: Pinter Publishers.

Page 350: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

324

Ballard, W. L. 1985. The linguistic history of south China: Miao-Yao and

southern dialects. Linguistics of the Sino-Tibetan area: The state of the

art—Papers presented to Paul K. Benedict for his 71st birthday, ed. by

Graham Thurgood, James A. Matisoff, and David Bradley, 58-84.

Canberra: Department of Linguistics, Research School of Pacific Studies,

The Australian National University.

Beard, Tim; Betsey Warrick; and Kao Cho Saefong (eds.) 1993. Loz-hnoi, loz-

hnoi uov…: In the old, old days… Traditional stories of the Iu Mien, Vol.

1. Berkeley, CA: Iu Mien Story Project, Laotian Handcraft Project.

__________ (eds.) 1995. In the old, old days… Loz-hnoi, loz-hnoi uov…:

Traditional stories of the Iu Mien, Vol. 2. Berkeley, CA: Iu Mien Story

Project, Laotian Handcraft Project.

Beaugrande, Robert. A. de. 1980. Text, discourse, and process: Toward a

multidisciplinary science of texts, Advances in discourse processes IV.

Norwood, New Jersey: ABLEX Publishing Corporation.

__________. 1985. Text linguistics in discourse studies. Handbook of discourse

analysis 1: disciplines of discourse, ed. by Teun A. van Dijk, 41-70.

London: Academic Press.

__________. 2001. Interpreting the discourse of H. G. Widdowson: a corpus-

based critical discourse analysis. Applied linguistics, vol. 22, 1: 104-121.

Oxford: Oxford University Press.

Beaugrande, Robert. A. de. and Wolfgang Ulrich Dressler. 1981. Introduction to

text linguistics. Essex: Longman Group.

Beekman, John; John Callow; and Michael Kpesec. 1981. The semantic structure

of written communication, 5th edition. Dallas: Summer Institute of

Linguistics.

Benedict, Paul K. 1975. Austro-Thai language and culture, with a glossary of

roots, with a foreword by W. H. Goodenough. New Haven: HRAF Press.

Page 351: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

325

__________. 1976. Austro-Thai and Austroasiatic. Austroasiatic studies, Part I,

ed. by Philip N. Jenner, Laurence C. Thompson, and Stanley Starosta, 1-

36. Honolulu: The University Press of Hawaii.

Bennett, Michael. 1981. Of tense and aspect: One analysis. Tense and aspect,

Syntax and semantics 14 (eds.) by Philip J. Tedeschi and Annie Zaenen,

13-29. New York: Academic Press.

Bergen, Robert D. ed. 1994. Biblical Hebrew and discourse linguistics. Dallas:

SIL and / Winona Lake: Eisenbrauns.

Berry, Margaret. 1981. Systemic linguistics and discourse analysis: A multi-

layered approach to exchange structure. Studies in discourse analysis, ed.

by Malcom Coulthard and Martin Montgomery, 120-45. London:

Routledge.

Bishop, Ruth G. 1979. Tense-aspect in Totonac narrative discourse. Discourse

studies in Mesoamerican languages, Vol. 1 discussion, ed. by Linda K.

Jones, 31-68. Dallas: The Summer Institute of Linguistics and the

University of Texas at Arlington.

__________. 1998b. Marginalia semitica I. Topics in Hebrew and Semitic

linguistics, ed. by Joshua Blau, 185-220. Jerusalem: The Magnes Press.

Blass, Regina. 1990. Relevance relations in discourse: A study with special

reference to Sissala. Cambridge: Cambridge University Press.

Blau, Joshua. 1998a. On the history and structure of Hebrew. Topics in Hebrew

and Semitic linguistics, ed. by Joshua Blau, 11-19. Jerusalem: The Magnes

Press.

Block, Karen Louise. 1994. Discourse grammar of first person narrative in Plang.

MA thesis, University of Texas at Arlington.

Bodine, Walter R. ed. 1995. Discourse analysis of Biblical literature: What it is

and what it offers. Georgia: The Society of Biblical Literature/Scholars

Press.

Page 352: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

326

Bower, Gordon H. and Randolph K. Cirilo. 1985. Cognitive psychology and text

processing. Handbook of discourse analysis, vol. 1: disciplines of

discourse, ed. by Teun A. van Dijk, 71-105. London: Academic Press.

Bowling, Andrew C. 1997. Another brief overview of the Hebrew verb. JOTT

9.48-73. Dallas: SIL.

Bradley, David. 1991. Language policy for minority languages in Thailand and

China. Pan-Asiatic linguistics, Proceedings of the Third International

Symposium on Language and Linguistics. Chulalongkorn University,

Bangkok, Thailand. January 8-10, 1992, vol. II, ed. by Amara

Prasithrathsint, 921-28. Bangkok: Chulalongkorn University.

Brisard, Frank. 2002. The English present. Grounding: The epistemic footing of

deixis and reference, ed. by Frank Brisard, 251-97. Berlin: Mouton de

Gruyter.

__________. ed. 2002. Grounding: The epistemic footing of deixis and reference.

Berlin: Mouton de Gruyter.

Brown, Gillian. 1995. Speakers, listeners and communication: Explorations in

discourse analysis. Cambridge: Cambridge University Press.

Brown, Gillian and George Yule. 1983. Discourse analysis. Cambridge:

Cambridge University Press.

Bryant, Brooke; Karen Block; and John Bryant. 1999. A preliminary discourse

analysis of Wetern Red Karen Narratives. PYU Working Papers in

Linguistics, 3.90-122. Chiang Mai: Payap University Linguistics

Department.

Burquest, Donald A. 1992. An introduction to the use of aspect in Huasa

narrative. Language in context: essays for Robert E. Longacre, ed. by Shin

Ja J. Hwang and William R. Merrifield, 393-417. Arlington: The Summer

Institute of Linguistics and The University of Texas at Arlington.

Page 353: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

327

Bussmann, Hadumod. 1996. Routledge dictionary of language and linguistics,

translated and ed. by Gregory P. Trauth and Kerstin Kazzazi from Lexikon

der Sprachwissenschaft, 2nd ed. London: Routledge.

Buth, Randall. 1994. Methodological collision between source criticism and

discourse analysis: The problem of “unmarked temporal overlay” and the

pluperfect/nonsequential wayyiqtol. Biblical Hebrew and discourse

linguistics, ed. by Robert D. Bergen, 138-54. Dallas: SIL / Winona Lake:

Eisenbrauns.

__________. 1995. Functional Grammar, Hebrew and Aramaic: An integrated,

textlinguistic approach to syntax. Discourse analysis of Biblical literature:

What it is and what it offers, ed. by Walter R. Bodine, 77-102. Atlanta,

Georgia: Scholars Press.

Bybee, Joan. 1995. Modality in grammar and discourse: An introductory essay.

Modality in grammar and discourse, ed. by Joan Bybee and Suzanne

Fleischman, 1-14. Amsterdam: John Benjamins.

Bybee, Joan, and Suzanne Fleischman (eds.) 1995. Modality in grammar and

discourse. Amsterdam: John Benjamins.

Bybee, Joan; Revere Perkins; and William Pagliuca. 1994. The evolution of

grammar: Tense, aspect, and modality in the languages of the world.

Chicago: The University of Chicago Press.

Caenepeel, Mimo, and Marc Moens. 1994. Temporal structure and discourse

structure. Tense and aspect in discourse, ed. by Co Vet and Carl Vetters,

5-20. Berlin: Mouton de Gruyter.

Callaway, Lois, and C. W. Callaway. 1976. Mien (Yao). Phonemes and

orghography: Language planning in the minority langugages of Thailand.

(Pacific Linguistics. Series C. No. 43) Canbera: Department of Linguistics,

Research School of pacific Studies. The Australian National University.

Page 354: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

328

Callow, Kathleen. 194. Discourse considerations in translating the Word of God.

Grandrapids: Zondervan.

Chang, Claire Hsun-huei. 1990. On serial verbs in Mandarin Chinese: VV

compounds and co-verbial phrases. When verbs collide: Papers from the

1990 Ohio State Mini-Conference on Serial Verbs, ed by Brian D. Joseph

and Arnold M. Zwicky, 288-315. Columbus, OH: The Ohio State

University, Department of Linguistics.

Casad, Eugene. 1995. Seeing it in more than one way. Language and the cognitive

construal of the world, ed. by John R. Taylor and Robert E. MacLaury, 23-

49. Berlin: Mouton de Gruyter.

Chafe, Wallace L. 1973. Language and memory. Language 49:2. 261-81.

__________. 1974. Language and consciousness. Language 50:1. 111-33.

__________. 1979. The flow of thought and the flow of language. Discourse and

syntax, Syntax and semantics 12, 159-81. New York: Academic Press.

__________. 1987. Cognitive constraints on information flow. Coherence and

grounding in discourse, ed. by Russel S. Tomlin, 21-51. Amsterdam: John

Benjamins.

__________. 1994. Discourse, consciousness, and time. Chicago: The University

of Chicago Press.

__________. 1996. Beyond beads on a string and branches in a tree. Conceptual

structure, discourse and language, ed. by Adele E. Goldberg, 49-65.

Stanford, CA: CSLI Publications.

Chang, Kun. 1953. On the tone system of the Miao-Yao languges. Language 29:

3. 374-8.

__________. 1966. A comparative study of the Yao tone system. Language 42: 2.

303-10.

Chen Bin and Li Minghui. 1995. Azaleas in the woods: The Yaos. (Women’s

culture series: Nationalities in Yunnan) Kunming: Yunnan Education

Page 355: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

329

Publisher. 阵斌 1995 年,《云南民族女性文化丛书·瑶族 丛林中的映山

红》, 昆明:云南教育出版社.

Chen, Matthew Y. 1991. Competing sound changes: Evidence from Kam-Tai,

Miao-Yao and Tibeto-Burman. Pan-Asiatic linguistics: Proceedings of the

Third International Symposium on Language and Linguistics,

Chulalongkorn University, Bangkok, Thailand, January 8-10, 1992, ed. by

Amara Prasithrathsint and Sudaporn Luksaneeyanawin, 16-27. Bangkok:

Chulalongkorn University Printing House.

Chimombo, Moria, and Robert L. Roseberry. 1998. The power of discourse: An

introduction to discourse analysis. Mahwah, NJ: Lawrence Erfaum

Associates, Publishers.

Clark, Marybeth. 1976. Coverbs and case in Vietnamese. Canberra: Department of

Linguistics, Research School of Pacific Studies, The Australian National

University.

Chisholm Jr., Robert B. 1998. From exegesis to exposition: A practical guide to

using Biblical Hebrew. Grand Rapids, Michigan: Baker.

[Chob Kacha-Ananda. 1980. Conversation lessons in Yao language (Bot

sonthanaa Phasaa Yao). Chiang Mai: Tribal Research Center, Tribal

Welfare Division, Public Welfare Department, Ministry of Defence.] ชอบ

คชาอนันต. 2523. บทสนทนาภาษาเยา. เชียงใหม : ปศูนยวจิัยขาวเขา

กองสงเคราะหชาวเขา กรมประชาสงเคราะห กระทรวงมหาดไทย

__________. 1997. Thailand Yao: Past, present, and future. Tokyo: Institute for

the Study of Languages and Cultures of Asia and Africa (ILCCA).

Clark, Herbert H., and Susan E. Haviland. 1977. Comprehension and the given-

new contract. Discourse production and comprehension, Vo. 1 in the

series Discourse precess: Advances in research and theory, ed. by Roy O.

Freedle, 1-40. Norwood, NJ: Ablex Publishing Corporation

Page 356: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

330

__________. 1988. An inchoative conjunction in Hmong: Extra-sentential topic

marker? Paper presented at the International Symposium on Language and

Linguistics, Faculty of Liberal Arts, Thammasat University, Bangkok,

Thailand, 9-11 August 1988, ed. by Cholticha Bamroongraks et al., 93-

106. Bangkok: Faculty of Liberal Arts, Thammasat University.

Cloran, Carmel. 1995. Defining and relating text segments: subject and theme in

discourse. On subject and theme: A discourse functional perspective, ed.

by Ruqaiya Hasan and Peter H. Fries, 361-403. Amsterdam: John

Benjamins.

Coats, George W. ed. 1985. Saga, legend, tale, novella, fable: narrative forms in

Old Testament literature. (Journal for the Study of the Old Testament

Supplement Series 35.) Sheffield: Sheffield Academic Press.

Comrie, Bernard. 1976. Aspect: An introduction to the study of verbal aspect and

related problems. Cambridge: Cambridge University Press.

__________. 1989. Hmong and areal South-East Asia. Pacific linguistics series A-

No. 77, Papers in South-East Asian linguistics No. 11: South-East Asian

syntax, ed. by David Bradley, 175-230. Canberra: Department of

Linguistics, Research school of Pacific Studies, The Australian National

University.

__________. 1992. Conjunctions as topicalizers: More on Southeast Asian

languages. Papers from the First Annual Meeting of the Southeast Asian

Linguistics Society 1991, ed. by Martha Ratliff and Eric Schiller, 87-107.

Tempe: Arizona State University.

Clifford, John E. 1975. Tense and tense logic. The Hague: Mouton Publishers.

__________. 1985. Tense. Cambridge: Cambridge University Press.

Cooke, Joseph R. 1979. Forms and meanings of the Thai particle si. South-East

Asian linguistic studies, Vol. 4, ed. by Nguyen Dan Liem, 61-98. (Pacific

Page 357: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

331

Linguistics, Series C-No. 49.) Canberra: Department of Linguistics,

Research School of Pacific Studies, The Australian National University.

Coulthard, Malcolm. 1987. Evaluative text analysis. Language topics: Essays in

honour of Michael Halliday Vol. II, ed. by Ross Steele and Terry

Threadgold, 181-90. Amsterdam: John Benjamins.

__________. ed. 1992. Advances in spoken discourse analysis. London:

Routledge.

__________. ed. 1994. Advances in written text analysis. London: Routledge.

__________. 1999. Some contributions of typology to cognitive linguistics, and

vice versa. Cognitive linguistics: Foundations, scope, and methodology,

ed. by Theo Janssen and Gisela Redeker, 61-93. Berlin: Mouton de

Gruyter.

Court, Christopher Anthony Forbes. 1986. Fundamentals of Iu Mien (Yao)

grammar. Ph.D. dissertation, University of California, Berkeley.

__________. 1991. Mien. Thai hill tribes phrasebook, ed. by Sally Steward, 140-

65. Hawthom, Victoria: Lonely Planet Publications Pty Ltd.

Croese, Robert. 1983. Classification of narrative types. Notes on Linguistics 26.

23-6.

Croft, William. 1991. Syntactic categories and grammatical relations: The

cognitive organization of information. Chicago: The University of Chicago

Pres.

Croft, William, and D. Alan Cruse. 2004. Cognitive Linguistics. Cambridge:

Cambridge University Press.

Cumming, Susanna. 1995. Multifunctionality and the realization problem in

modeling discourse production. Discourse grammar and typology: papers

in honor of John W. M. Verhaar, ed. by Werner Abraham, T. Givón, and

Sandra A. Thompson. 248-74. Amsterdam: John Benjamins.

Page 358: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

332

Dahl, Östen. 1981. On the definition of the telic-atelic (bound-nonbound)

distinction. Tense and aspect, Syntax and semantics 14, ed. by Philip J.

Tedeschi and Annie Zaenen, 79-90. New York: Academic Press.

Dai, John Xiang-ling. 1990. Syntactic constructions in serial verb expressions in

Chinese. When verbs collide: Papers from the 1990 Ohio State Mini-

Conference on Serial Verbs, ed by Brian D. Joseph and Arnold M.

Zwicky, 316-39. Columbus, OH: The Ohio State University, Department

of Linguistics.

Dang, Nghiem Van; Son Thai Chu; and Hung Luu. 2000. Ethnic minoritires in

Vietnam. Hanoi: Thegioi Publisher.

DeLancey, Scott. 1987. Transitivity in grammar and cognition. Coherence and

grounding in discourse, ed. by Russel S. Tomlin, 53-68. Amsterdam: John

Benjamins.

de Stadler, Leon, and Christoph Eyrich (eds.) Issues in cognitive linguistics: 1993

proceedings of the International Cognitive Linguistics Conference. Berlin:

Mouton de Gruyter.

[Department of Social and Welfare Development, Ministry of Social Development

and Human Security. 2002. Highland communities within 20 provinces of

Thailand, 2002. Bangkok: Department of Social and Welfare

Development, Ministry of Social Development and Human Security and

UNICEF] กรมพัฒนาสังคมและความมั่นคงของมนุษย. 2545.ทําเนียบชุมชนบนพื้น

ที่สูง ๒๐ จังหวัดในประเทศไทย พ.ศ. ๒๕๔๕. กรุงเทพฯ : กรมพัฒนาสังคมและ

สวัสดิการ กระทรวงการพัฒนาสังคมและความมั่นคงของมนุษยและองคการทุน

สําหรับเด็กแหงสหประชาชาต

Diessel, Holger. 2001. The ordering distribution of main and adverbial clauses: A

typological study. Language 77:2. 433-455.

Page 359: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

333

Dik, Simon C. 1980. Seventeen sentences: Basic principles and application of

Functional Grammar. Syntax and semantics, Vol. 13: Current approaches

to syntax, ed. by Edith A. Moravcsik and Jessica R. Wirth, 45-75. New

York: Academic Press.

__________. 1997. The theory of functional grammar---Part 2: Complex and

derived construction, ed. by Kees Hengeveld. Berlin: Mouton de Gruyter.

Dirven, René. 2004. Cognitive linguistics. The linguistics encyclopedia, 2 ed.,

ed. by Kirsten Malmkjær, 76-82. London: Routledge.

nd

Dixon, R. M. W. 1977. Where have all the adjectives gone? Studies in Language

1:1.19-80.

__________. 1997. The rise and fall of languages. Cambridge: Cambridge

University Press.

Downer, G. B. 1963. Chinese, Thai, and Miao-Yao. Collected papers in Oriental

and African Studies: Linguistic comparison in South East Asia and the

Pacific, ed. by H. L. Shorto, 133-9. London: School of Oriental and

African Studies, University of London.

Doiz-Bienzobas, Aintzane. 2002. The preterit and the imperfect as grounding

predications. Grounding: The epistemic footing of deixis and reference, ed.

by Frank Brisard, 299-347. Berlin: Mouton de Gruyter.

Dooley, Robert A., and Stephen H. Levinsohn. 2001. Analyzing discourse: A

manual of basic concepts. Dallas: SIL International.

Dry, Helen Arister. 1983. The movement of narrative time. The journal of literary

semantics 12(2):19-54.

__________. 1992. Foregrounding: an assessment. Language in context: essays

for Robert E. Longacre, ed. by Shin Ja J. Hwang and William R.

Merrifield, 435-50. Arlington: The Summer Institute of Linguistics and

The University of Texas at Arlington.

Page 360: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

334

Eggins, Suzanne. 1994. An introduction to systemic functional linguistics.

London: Pinter.

Emmott, Catherine. 1999. Embodied in a constructed world: Narrative processing,

knowledge representation, and indirect anaphora. Discourse studies in

cognitive linguistics: Selected papers from the Fifth International

Cognitive Linguistics Conference Amsterdam, July 1997, 5-27.

Amsterdam: John Benjamins Publishing Company.

Endo, Yoshinobu. 1996. The verbal system of Classical Hebrew in Joseph story:

An approach from discourse analysis. Assen: Van Gorcum & Comp.

[__________. 2001. Running away from God’s presence: Hesed (everlasting love)

of God revealed in the Book of Jonah (Expository commentary) (Shu-no

mikao-o sakete: Yonasho-ni shimesareta kami-no 「eienn-no ai (hesed)」

), Tokyo: Word of Life Press.] 遠藤嘉信、(2001)『主の御顔を避けて—ヨ

ナ書に示された神の「永遠の愛(ヘセド)」』、東京:いのちのことば

社.

__________. 2003. Problems in the ‘Grounding-theory’ of R. E. Longacre and

others in the study of Hebrew narrative discourse. Exegetica: Studies in

Biblical exegesis 14: 77-90. Tokyo: Biblical Exegesis Study Group in

Japan.

Enfield, N. J. 2001. On generic and areal linguistics in Maninland Sout-East Asia:

Parallel polyfunctionality of ‘aquire.’ Areal diffusion and genetic

inheritance, ed. by Alexandra Y. Aikhenvald and R. M. W. Dixon, 255-90.

Oxford: Oxford University Press.

Enkvist, Nils Erik. 1987. Text strategies: Single, dual, multiple. Language topics:

Essays in honour of Michael Halliday Vol. II, ed. by Ross Steele and Terry

Threadgold, 191-211. Amsterdam: John Benjamins.

Epstein, Richard. 1996. Viewpoint and the definite article. Conceptual structure,

discourse and language, ed. by Adele E. Goldberg, 99-112. Stanford, CA:

CSLI Publications.

Page 361: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

335

__________. 1999. Roles, frames and definiteness. Discourse studies in cognitive

linguistics: Selected papers from the Fifth International Cognitive

Linguistics Conference Amsterdam, July 1997, ed. by Karen van Hoek,

Andrej A. Kibrik, and Leo Noordman, 53-74. Amsterdam: John

Benjamins.

__________. 2002. The definite article, accessibility, and the construction of

discourse referents. Cognitive Linguistics 12-4. 333-78.

__________. 2002. Grounding, subjectivity and definite descriptions. Grounding:

The epistemic footing of deixis and reference, ed. by Frank Brisard, 41-82.

Berlin: Mouton de Gruyter.

Everett, Daniel L. 2001. Monolingual field research. Linguistic fieldwork, ed. by

Paul Newman and Martha Ratliff, 166-88. Cambridge: Cambridge

University Press.

Fang, Yan; Edward McDonald; and Cheng Musheng. 1995. On theme in Chinese:

From clause to discourse. On subject and theme: A discourse functional

perspective, ed. by Ruqaiya Hasan and Peter H. Fries, 235-73. Amsterdam:

John Benjamins.

Fasold, Ralph. 1990. The sociolinguistics of language. Oxford: Blackwell.

Fauconnier, Gilles. 1994. Mental spaces: Aspects of meaning construction in

natural language. Cambridge: Cambridge University Press.

__________. 1997. Mappings in thought and language. Cambridge: Cambridge

University Press.

Fei Xiaotong. 1991. Fifty years investigation in the Yao mountains. The Yao of

South China: Recent international studies, ed. by Jacques Lemoine and

Chiao Chien, 17-36. Paris: Pangu, Editions de l’A.F.E.Y.

Filbeck, David. 1975. A grammar of verb serialization in Thai. Studies in Tai

Linguistics, ed. by Jimmy G. Harris and James R. Chamberlai, 112-29.

Page 362: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

336

Fillmore, Charles J. 1985. Linguistics as a tool for discourse analysis. Handbook

of discourse analysis 1: disciplines of discourse, ed. by Teun A. van Dijk,

11-39. London: Academic Press.

Fleischman, Suzanne. 1995. Imperfective and irrealis. Modality in grammar and

discourse, ed. by Joan Bybee and Suzanne Fleischman, 519-51.

Amsterdam: John Benjamins.

Flik, Eva. 1978. Dan tense-aspect and Discourse. Papers on discourse, ed. by

Joseph Grimes, 46-62. Dallas: SIL.

Fokkelman, Jan P. 1996. Genesis 37 and 38 at the interface of structural analysis

and hermeneutics. Literary structure and rhetorical strategies in the

Hebrew Bible, ed. by L. J. de Regt, J. de Waard and J. P. Fokkelman, 152-

87. Assen, The Netherlands: Van Gorcum.

Foley, William A. 2004. The notion of ‘event’ and serial verb constructions:

Arguments from New Guinea. A paper presented at The 14th Annual

International Conference of the Southeast Asian Linguistics Society

(SEALS 14), 19-21 May 2004.

Foley, William A. and Mike Olson. 1985. Clausehood and verb serialization.

Grammar inside and outside the clause: Some approaches to theory from

the field, ed. by Johanna Nichols and Anthony C. Woodburg, 17-60.

Cambridge: Cambridge University Press.

Foley, William A. and Robert D. Van Valin, Jr. 1980. Role and Reference

Grammar. Syntax and semantics, Vol. 13: Current approaches to syntax,

329-52. New York: Academic Press.

__________. 1984. Functional syntax and universal grammar. Cambridge:

Cambridge University Press.

__________. 1985. Information packaging in the clause. Language typology and

syntactic description, Vol. 1: Clause structure, ed. by Timothy Shopen,

282-364. Cambridge: Cambridge University Press.

Page 363: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

337

Follingstad, Carl M. 1994. Thematic development and prominence in Tyap

discourse. Discourse features of ten languages of West-Central Africa, ed.

by Stephen H. Levinsohn, 151-89. Dallas: The Summer Institute of

Linguistics and The University of Texas at Arlington.

__________. 2001. Deictic viewpoint in Biblical Hebrew text: A syntagmatic and

paradigmatic analysis of the particle yk (kî). Dallas: SIL International.

Forrest, Linda B. 1996. Discourse goals and attentional processes in sentence

production: The dynamic construal of events. Conceptual structure,

discourse and language, ed. by Adele E. Goldberg, 149-61. Stanford, CA:

CSLI Publications.

Fox, Barbara A. 1987. Discourse structure and anaphora: Written and

conversational English. Cambridge: Cambridge University Press.

Frawley, William. 1992. Linguistic semantics. New Jersey: Lawrence Erlbaum

Associate, Publishers.

Freedle, Roy O. ed. 1979. New directions in discourse processing, Advances in

discourse processes II. Norwood, New Jersey: ABLEX Publishing

Corporation.

Fries, Peter H. 1992. The structure of information in written English text.

Language Science 14:4.461-88.

__________. 1995. Themes, methods of development, and texts. On subject and

theme: A discourse functional perspective, ed. by Ruqaiya Hasan and Peter

H. Fries, 317-59. Amsterdam: John Benjamins.

__________. 2002. The flow of information in a written English text. Relations

and functions within and around language, ed. by Peter H. Fries, et al.,

117-55. London: Continuum.

Fuller, Judith Wheaton. 1985. Topic and comment in Hmong. Ph.D. dissertation:

The University of Minnesota.

Page 364: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

338

Gärdenfors, Peter. 1999. Some tenets of cognitive semantics. Cognitive semantics:

Meaning and cognition, ed. by Jens Allwood and Peter Gärdenfors, 19-36.

Amsterdam: John Benjamins.

Gee, James Paul. 1999. An introduction to discourse analysis: Theory and method.

London: Routledge.

Geeraerts, Dirk. 1997. Diachronic prototype semantics: A contribution to

historical lexicology. Oxford: Clarendon Press.

Geluykens, Ronald. 1993. Syntactic, semantic and interactional prototypes: The

case of left-dislocation. Conceptualizations and mental processing in

language, ed. by Richard A. Geiger and Brygida Rudzka-Ostyn, 709-30.

Berlin: Mouton de Gruyter.

Georgakopoulou, Alexandra, and Dionysis Goutsos. 1997. Discourse analysis: An

introduction. Edinburgh: Edinburgh University Press.

Gesenius, W., and E. Kautzsch. 1910. Gesenius’ Hebrew grammar. Translated,

revised and ed. by A. E. Cowley. Oxford: Clarendon Press.

Givón, Talmy. 1979. From discourse to syntax: Grammar as a processing strategy.

Discourse and syntax, Syntax and semantics Vol. 12, 81-112. New York:

Academic Press.

__________. ed. 1979. Discourse and syntax, Syntax and semantics Vol. 12. New

York: Academic Press.

__________. 1983. Topic continuity in discourse: An introduction. Topic

continuity in discourse, ed. by Talmy Givón, 1-41. Amsterdam: John

Benjamins.

__________. 1984. Syntax: A functional-typological introduction, Vol. 1.

Amsterdam: John Benjamins.

__________. 1987. Beyond foreground and background. Typological studies in

language (TSL) Vol. 11, Ceherence and grounding in discourse, ed. by

Russell R. Tomlin, 175-188. Amsterdan: John Benjamins.

Page 365: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

339

__________. 1990. Syntax: A functional-typological introduction, Vol. 2.

Amsterdam: John Benjamins.

__________. 1994. Isomorphism in the grammatical code: Cognitive and

biological considerations. Iconicity in language, ed. by Raffaele Simone,

47-76. Amsterdam: John Benjamins.

__________. 1995. Coherence in text vs. coherence in mind. Coherence in

spontaneous text, ed. by Morton Ann Cernsbacher and T. Givón, 59-115.

Amsterdam: John Benjamins.

Goldberg, Adele E. ed. 1996. Conceptual structure, discourse and language.

Stanford, CA: CSLI Publications.

Goldman, Ann Yarwood. 1995. Lao Mien embroidery: Migration and change.

Bangkok: White Lotus.

Gordon, Raymond G., Jr. (ed.), 2005. Ethnologue: Languages of the World,

Fifteenth edition. Dallas, Tex.: SIL International. Online

version: http://www.ethnologue.com/.

Grimes, Joseph E. 1971. Kinds of information in discourse. Kivung---Journal of

the Linguistic Society of Papua and New Guinea, 4:2.64-74.

__________. 1972. Outlines and overlays. Language 48:3.513-24.

__________. 1975. The thread of discourse. Mouton: The Hague.

__________. 1977. Narrative studies in oral texts. Current trends in texlinguistics

(Research in text theory, vol. 2), ed. by Wolfgang U. Dressler, 123-32.

Berlin: Walter de Gruyter.

Grimes, Joseph E., and Naomi Glock. 1970. A Saramaccan narrative patern.

Language 46:2.408-425.

Gudschinsky, Sarah C. 1959. Discourse analysis of a Mazatec text. International

Journal of American Linguistics 25:3. 139-46.

Page 366: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

340

Gundel, Jeanette K.; Nancy Hedberg; and Ron Zacharski. 1993. Cognitive status

and the form of referring expressions in discourse. Language 69:2. 274-

307.

Guo, Jiansheng. 1995. The interactional basis of the Mandarin modal néng

‘CAN.’ Modality in grammar and discourse, ed. by Joan Bybee and

Suzanne Fleischman, 203-38. Amsterdam: John Benjamins.

Gutt, Ernst-August. 1992. Relevance Theory: A guide to successful

communication in translation, Lectures delivered at the triennial

translation workshop of UBS Zimbabwe, 1991. Dallas: Summer Institute

of Linguistics, and New York: United Bible Societies.

Gutwinski, Waldemar. 1976. Cohesion in literary texts. The Hague: Mouton.

Hajičová, Eva. 1994. Topic/focus and related research. The Prague School of

struactural and functional linguistics: a short introduction, ed. by Philip

A. Luelsdorff, 245-75. Amsterdam: John Benjamins.

Halliday, Michael A. K. 1967-8. Notes on transitivity and theme in English, parts

1-3, Journal of Linguistics 3.1, 3.2, 4.2.

__________. 1970. Language structure and language function. New Horizons in

Linguistics, ed. by John Lyons, 140-65. Middlesex: Penguin Books.

__________. 1994. An introduction to functional grammar, 2nd ed. London:

Edward Arnold.

Halliday, Michael A. K., and Ruqaiya Hasan. 1976. Cohesion in English. London:

Longman.

__________. 1989. Language, context, and text: Aspects of language in a social-

semiotic perspective (2nd edition). Oxford: Oxford University Press.

Harris, Zellig. 1952a. Discourse analysis. Language 28:1.1-30.

__________. 1952b. Discourse analysis: A sample text. Language 28:4

(Part1).474-94.

Page 367: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

341

Hatim, Basil. 1997. Communication across cultures: Translation theory and

contrastive text linguistics. Devon, EX: University of Exeter Press.

Hasan, Ruqaiya; Carmel Cloran; and David Butt (eds.) 1996. Functional

descriptions: Theory in practice. Amsterdam: John Benjamins.

Hasan, Ruqaiya, and Peter H. Fries (eds.) 1995. On subject and theme: A

discourse functional perspective. Amsterdam: John Benjamins.

Hatim, Basil. 1997. Communication across cultures: Translation theory and

contrastive text linguistics. Exeter: University of Exeter Press.

Haudricourt, André G. 1966. The limits and connections of Austroasiatic in the

Northeast (translated from French in 1961). Studies in comparative

Austroasiatic linguistics, ed. by Norman H. Zide et al., 44-56. The Hague:

Mouton.

Heckert, Jakob K. 1996. Discourse function of conjoiners in the Pastoral Epistles.

Dallas: Summer Institute of Linguistics.

Heimerdinger, Jean-Marc. 1999. Topic, focus and foreground in ancient Hebrew

narratives. Sheffield: Sheffield Academic Press.

Higbie, James, and Snea Thinsan. 2002. Thai reference grammar: The structure of

spoken Thai. Bangkok: Orchid Press.

Hirsch, Richard. 1997. Semantic content and depth of intention: A study in

cognitive semantics. Discourse and perspective in cognitive linguistics, ed.

by Walf-Andreas Liebert, Gisela Redeker, and Linda Waugh, 61-83.

Amsterdam: John Benjamins Publishing Company.

Holisky, Dee Ann. 1981. Aspect theory and Georgian aspect. Tense and aspect,

Syntax and semantics Vol. 14, ed. by Philip J. Tedeschi and Annie Zaenen,

127-144. New York: Academic Press.

Holmqvist, Kenneth, and Jana Holšánová. 1997. Focus movements and the

internal images of spoken discourse. Discourse and perspective in

Page 368: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

342

cognitive linguistics, ed. by Wolf-Andreas Liebert, Gisela Redeker, and

Linda Waugh, 223-46. Amsterdam: John Benjamins.

Hopper, Paul J. 1979. Aspect and foregrounding in discourse. Discourse and

syntax, Syntax and semantics Vol. 12, 213-241. New York: Academic

Press.

__________. 1982. Aspect between discourse and grammar: an introductory essay

for the volume. Tense-aspect: between semantics and pragmatics:

containing the contributions to a Symposium on Tense and Aspect, held at

UCLA, May 1979, ed. by Paul J. Hopper. 3-18. Amsterdam: John

Benjamins.

__________. 1995. The category ‘event’ in natural discourse and logic. Discourse

grammar and typology: Papers in Honor of John W. M. Verhaar, ed. by

Werner Abraham, T. Givón, and Sandra A. Thompson, 139-50.

Amsterdam: John Benjamins.

__________. 1997. Dispersed verbal predicates in vernacular written narrative.

Directions in Functional Linguistics, ed. by Akio Kamio, 1-18.

Amsterdam: John Benjamins.

__________. 1998. Emergent grammar. The new psychology of language:

Cognitive and functional approaches, ed. by Michael Tomasello, 155-175.

New York: Lawrence Erlbaum.

Hopper, Paul J. and Sandra A. Thompson. 1980. Transitivity in grammar and

discourse. Language 56:2. 251-99.

__________ (eds.) 1982. Studies in transitivity. Syntax and semantics, Vol. 15.

New York: Academic Press.

__________. 1984. The discourse basis for lexical categories in universal

grammar. Language 60:4. 703-52.

Horie, Kaoru, and Shigeru Sato. 2001. Cognitive-Functional Linguistics in an East

Asian context. Tokyo: Kurosio Publishers.

Page 369: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

343

[Huáng Yù, ed. and comments. 1989. The Yao history book: Collection of King

Ping’s Charter (Yáozú shĭshū Píng Huáng Quàndié jíbiān). Nanning Shi:

Guangxi Renmin Chubanshe.] 黄钰, 辑注. 1989 年,《瑶族史书 评皇券牒

集编》, 南宁市:广西人民出版社.

Huang Yu. 1991. Preliminary study of the Yao “King Ping’s Charter.” The Yao of

South China: Recent international studies, ed. by Jacques Lemoine and

Chiao Chien, 89-123. Paris: Pangu, Editions de l’A.F.E.Y.

Hwang, Shin Ja J. 1992. The functions of negation in narration. Language in

context: Essays for Robert E. Longacre, ed. by Shin Ja J. Hwang and

William R. Merrifield, 321-37. Dallas: The Summer Institute of

Linguistics and The University of Texas at Arlington.

[IMPECT Association (Inter Mountain Peoples Education and Culture in Thailand

Association). 2002. Essential knowledge of the Mien (Yao) localities.

Chiang Mai: IMPECT Association, The Primary Education Office,

Phayao, and The Primary Education Office in Pong district, Phayao.]

สมาคมศูนยรวมการศกึษาและวฒันธรรมของชาวไทยภเูขาในประเทศไทย (ศ.ว.ท.).

2545.สาระองคความรูภูมิปญญาทองถ่ินอิ้วเมี่ยน(เยา). เชียงใหม: IMPECT

Association, สํานักงานการประถมศึกษาจังหวัดพะเยา, สํานักงานการประถมศึกษา

อําเภอปง จังหวัดพะเยา.

Inglis, Douglas. 1997. Salience hierarchy in a Molbog text. JOTT 9. 1-14

__________. 1999. Some discourse evidence for analyzing Molbog as an ergative

language. PYU Working Papers in linguistics 3: 60-89. Chiang Mai: Payap

University Graduate Linguistics Department.

Inoue, Kazuko. 1982. An interface of syntax, semantics, and discourse structures.

Lingua 57: 259-300.

Page 370: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

344

__________. 1983. [Reported speech and its discourse functions in Japanese

(Nihongo-no denbun hyougen-to sono danwakinou). Gengo (language)

12.113-21]. 井上和子(1983)「日本語の伝聞表現とその談話機能」 『言

語』12.113-21.

Iwasaki, Shoichi. 1989. Clausehood and verb serialization in Thai narratives.

[Language and Linguistics (Phaasaa lae phaasaasaat) 7:2.84-130.

Bangkok: Thammasat University.] (ภาษาและภาษาศาสตร ปที่ 7 ฉบับที่ 2

มกราคม-มถุินายน 2532 สาขาวิชา ภาษาศาสตร คณะศลิปศาสตร

มหาวิทยาลัยธรรมศาสตร )

Jacobsen, Wesley M. 1997. Agentivity and aspect in Japanese. Directions in

functional linguistics, ed. by Akio Kamio, 83-116. Amsterdam: John

Benjamins.

Janssen, Theo A. J. M. 1993. Tenses and demonstratives: Conspecific categories.

Conceptualizations and mental processing in language, ed. by Richard A.

Geiger and Brygida Rudzka-Ostyn, 741-83. Berlin: Mouton de Gruyter.

__________. 1994. Preterit and perfect in Dutch. Tense and aspect in discourse,

ed. by Co Vet and Carl Vetters, 115-46. Berlin: Mouton de Gruyter.

__________. 2002. Deictic principles of pronominals, demonstratives, and tenses.

Grounding: The epistemic footing of deixis and reference, ed. by Frank

Brisard, 151-93. Berlin: Mouton de Gruyter.

Janssen, Theo, and Gisela Redeker. 1999. Cognitive linguistics: Foundations,

scope, and methodology. Berlin: Mouton de Gruyter.

Jaworski, Adam, and Nikolas Coupland (eds.) 1999. The discourse reader.

London: Routledge.

Jennings, Wanda Jean. 1998. Iu-Mienh literacy lessons: Applying learning styles

and perceived needs using familiar texts. Chiang Mai, Thailand: Payap

Page 371: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

345

Research and Development Institute and The Summer Institute of

Linguistics.

Johnson, Marion R. 1981. A unified temporal theory of tense and aspect. Tense

and aspect, Syntax and semantics Vol. 14, ed. by Philip J. Tedeschi and

Annie Zaenen, 145-175. New York: Academic Press.

Johnstone, Barbara. 2002. Discourse analysis. Oxford: Blackwell.

Jones, Larry B., and Linda K. Jones. 1984. Verb morphology and discourse

structure in Mesoamerican languages. Theory and application in

processing texts in non-Indoeuropean languages, ed. by Robert E.

Longacre, 25-48. Hamburg: Helmut Buske Verlag Hamburg.

Jones, Linda Kay. 1977. Theme in English expository discourse. Lake Bluff,

Illinois: Jupiter Press.

__________. ed. 1979. Discourse studies in Mesoamerican languages Vol. 1

discussion. Dallas: The Summer Institute of Linguistics and the University

of Texas at Arlington.

Jones, Linda K., and Ned R. Coleman. 1979. Towards a discourse perspective of

modes and tenses in Kickapoo narratives. Discourse studies in

Mesoamerican languages Vol. 1 discussion, ed. by Linda K. Jones, 69-95.

Dallas: The Summer Institute of Linguistics and the University of Texas at

Arlington.

Jones, Robert B., and Anthony V. Diller. 1976. Discourse analysis of Thai and

Japanese---Part I: Thai discourse. New York: Cornell University.

Jonsson, Hjorleifur. 1996. Shifting social landscape: Mien (Yao) upland

communities and histories in state-client settings. Ph.D. dissertation:

Cornell University.

Joseph, Brian D., and Arnold M. Zwicky (eds.) 1990. When verbs collide: Papers

from the 1990 Ohio State Mini-Conference on Serial Verbs. Columbus,

OH: The Ohio State University, Department of Linguistics.

Page 372: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

346

Jordan, Dean L. 1978. Nafaara tense-aspect in the folk tale. Papers on discourse,

ed. by Joseph Grimes, 84-90. Dallas: SIL.

Joüon, Paul., and Takamitsu Muraoka. 1991. A grammar of Biblical Hebrew, Vol.

2, Part 3: Syntax, translated and revised by Takamitsu Muraoka.

(Originally published as Grammarie de l’Hébreu biblique by the Pontifical

Biblical Institute in 1923) Roma: Editrice Pontificio Instituto Biblico.

Jucker, Andreas H., and Sara W. Smith. 1998. And people just you know like

‘wow’: Discourse markers as negotiating strategies. Discourse markers:

Descriptions and theory, ed. by Andreas H. Jucker and Yael Ziv, 171-201.

Amsterdam: John Benjamins.

Kamio, Akio, ed. 1997. Directions in functional linguistics. Amsterdam: John

Benjamins.

Kamio, Akio. 1997. Evidentiality and some discourse characteristics in Japanese.

Directions in functional linguistics, ed. by Akio Kamio, 145-71.

Amsterdam: John Benjamins.

[Kamontham Cheunphan. 1996. The lexical study of community-forest and

environmental conservation in Mien. MA thesis, Mahidol University.]

กทลธรรม ช่ืนพันธุ. 2539. การศึกษาคําศัพทเกี่ยวกับปาชุมชนและการอนุรักษ

ส่ิงแวดลอม ของชาวเมีย่น. วทิยานพิลธศิลปศาสตรมหาบัณฑติ (ภาษาศาสตร)

มหาวิทยาลัยมหิดล. กรุงเทพฯ : มหาวิทยาลัยมหิดล

[Kasamaphorn Maneekhao. 1989. Referents for participants in mien narrative

discourse. MA Thesis, Chulalongkorn University.] กษมาภรญ มญีขาว. 2533.

รปูภาษาที่ใชอางถึงตัวแสดงบทบาทในปริจเฉทเรื่องเลาภาษาเมี่ยน. วิทยานิพลธ

ปริญญาอักษรศาสตรมหาบัณฑิต ภาควิชาภาษาศาสตร บัณฑิตวิทยาลัย จุฬาลงกรณ

มหาวิทยาลัย. กรุงเทพฯ : จุฬาลงกรณมหาวิทยาลัย

Page 373: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

347

Khong, Dien. 2002. Population and ethno-demography in Vietnam. Chiang Mai:

Silkworm Books.

Kibrik, Andrej A. 1999. Reference and working memory: Cognitive inferences

from discourse observations. Discourse studies in cognitive linguistics:

Selected papers from the Fifth International Cognitive Linguistics

Conference Amsterdam, July 1997, ed. by Karen van Hoek, Andrej A.

Kibrik, and Leo Noordman, 29-52. Amsterdam: John Benjamins

Publishing Company.

Kimura, Shigeru. 1997. Patterns of migration in rural areas of Northern Thailand.

Research papers on interrelationship between population growth in

developing countries and global environment, Volume II: No. 290, March

1997. Tokyo: National Institute of Population and Social Security

Research, Ministry of Health and Welfare.

Kingkarn Thepkanjana. 1986. Serial verb constructions in Thai. Ph.D.

dissertation, The University of Michigan.

Kochańska, Agata. 2002. A cognitive grammar analysis of Polish nonpast

perfectives and imperfectives: How virtual events differ from actual ones.

Grounding: The epistemic footing of deixis and reference, ed. by Frank

Brisard, 349-90. Berlin: Mouton de Gruyter.

Komatsu, Eisuke (ed.) and Roy Harris (tr.). 1993. F. de. Saussure, Troisieme cours

de linguistique general (1910-1911): d’après les cahiers d’Emile

Constantin, Saussure’s third course of lectures on general linguistics

(1910-1911): From the notebooks of Emile Constantin. Oxford: Pergamon.

Kunstadter, Peter. 1967. Autonomy and integration of social systems: The Iu Mien

(“Yao” or “Man”) mountain population and their neighbors. Southeast

Asian tribes, minorities, and nations, Vol. 2, ed. by Peter Kunstadter, 583-

638. Princeton, New Jersey: Princeton University Press.

Page 374: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

348

Lakoff, George. 1972. Hedges: A study in meaning criteria and the logic of fuzzy

concepts. Papers from the Eighth Regional Meeting, Chicago Linguistic

Society, 183-228. Chicago: Chicago Linguistic Society.

__________. 1987. Women, fire, and dangerous things: What categories reveal

about the mind. Chicago: The University of Chicago Press.

Lambrecht, Knud. 1994. Information structure and sentence form: Topic, focus,

and the mental representations of discourse referents. Cambridge:

Cambridge University Press.

Langacker, Ronald W. 1987a. Foundations of cognitive grammar. Vol. I:

Theoretical prerequisites. Stanford, CA: Stanford University Press.

__________. 1987b. Nouns and verbs. Language 63: 53-94.

__________. 1991a. Cognitive Grammar. Linguistic theory and grammatical

description, ed. by Flip G. Droste and John E. Joseph, 274-306.

Amsterdam: John Benjamins.

__________. 1991b. Concept, image, and symbol: The cognitive basis of

grammar. Berlin: Mouton de Gruyter.

__________. 1991c. Foundations of cognitive grammar. Vol. II: Descriptive

application. Stanford, CA: Stanford University Press.

__________. 1994a. The limits of continuity: Discreteness in cognitive semantics.

Continuity in linguistic semantics, ed. by Catherine Fuchs and Bernard

Victorri, 9-20. Amsterdam: John Benjamins.

__________. 1994b. Cognitive Grammar. The encyclopedia of language and

linguistics, Vol. 2, ed. by R. E. Asher and J. M. Y. Simpson, 590-3.

Oxford: Pergamon Press.

__________. 1996. A constraint on progressive generics. Conceptual structure,

discourse and language, ed. by Adele E. Goldberg, 289-302. Stanford,

CA: CSLI Publications.

Page 375: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

349

__________. 1997. The contextual basis of cognitive semantics. Language and

conceptualization, ed. by Jan Nuyts and Eric Pederson, 229-52.

Cambridge: Cambridge University Press.

__________. 1999. Assessing the cognitive linguistic enterprise. Cognitive

linguistics: Foundations, scope, and methodology, ed. by Theo Janssen

and Gisela Redeker, 13-59. Berlin: Mouton de Gruyter.

__________. 2000. Grammar and conceptualization. Berlin: Mouton de Gruyter.

__________. 2001a. Topic, subject, and POSS. A cognitive approach to the verb:

Morphological and constructional perspectives, ed. by Hanne Gram

Simonsen and rolf Theil Endresen, 11-48. Berlin: Mouton de Gruyter.

__________. 2001b. Discourse in cognitive grammar. Cognitive Linguistics 12-2.

143-88.

__________. 2002. Deixis and subjectivity. Grounding: The epistemic footing of

deixis and reference, ed. by Frank Brisard, 1-28. Berlin: Mouton de

Gruyter.

__________. 2002. Remarks on the English groundings systems. Grounding: The

epistemic footing of deixis and reference, ed. by Frank Brisard, 29-38.

Berlin: Mouton de Gruyter.

LaPolla, Randy. 2001. The role of migration and language contact in the

development of the Sino-Tibetan language family. Areal diffusion and

genetic inheritance: Problems in comparative linguistics, ed. by Alexandra

Y. Aikhenvald and R.M.W. Dixon, 225-54. Oxford: Oxford University

Press.

Larsen, Helen. 1974. Some grammatical features of legendary narrative in Ancash

Quechua. Advances in tagmemics, ed. by Ruth M. Brend, 419-40.

Amsterdam: North-Holland Publishing Company.

Page 376: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

350

Larson, Mildred Lucille. 1978. The function of reported speech in discourse.

Arlington, TX: The Summer Institute of Linguistics and the University of

Texas at Arlington.

__________. 1984. The structure of Aguaruna (JIVARO) Texts. Theory and

application in processing texts in non-Indoeuropean languages, ed. by

Robert E. Longacre, 153-209. Hamburg: Helmut Buske Verlag Hamburg.

Lauerbach, Gerda. 1993. Interaction and cognition: Speech act schemata with but

and their interrelation with discourse type. Conceptualizations and mental

processing in language, ed. by Richard A. Geiger and Brygida Rudzka-

Ostyn, 679-708. Berlin: Mouton de Gruyter.

Lemke, J. L. 1988. Text structure and text semantics. Pragmatics, discourse and

text: Some systemically-inspired approaches, ed. by Erich H. Steiner and

Robert Veltman, 158-70. London: Pinter Publishers.

Lemoine, Jacques, and Chiao Chien (eds.) 1991. The Yao of south China: Recent

international studies. Paris: Pangu.

Leška, Oldřich. 1996. The meaning of Meaning in the Prague School. Discourse

and meaning: Papers in honor of Eva Hajičová, ed. by Barbara H. Partee

and Peter Sgall, 127-40. Amsterdam: John Benjamins.

Levelt, W. J. M. 1989. Speaking. Cambridge, MA: MIT Press. [not accessed yet]

Levinsohn, Stephen H. 1992a. The historic present and speech margins in

Matthew. Language in context: essays for Robert E. Longacre, ed. by Shin

Ja J. Hwang and William R. Merrifield, 451-73. Arlington: The Summer

Institute of Linguistics and The University of Texas at Arlington.

__________. 1992b. Discourse features of New Testament Greek: A coursebook.

Dallas: Summer Institute of Linguistics.

__________. 1992c. Participant reference in Koine Greek narrative. Linguistics

and New Testament interpretation: Essays on discourse analysis, ed. by

David Alan Black, 31-44. Nashville: Broadman Press.

Page 377: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

351

__________. 1994. Discontinuities in coherent texts. Discourse features of ten

languages of West-Central Africa, ed. by Stephen H. Levinsohn, 3-14.

Dallas: The Summer Institute of Linguistics and The University of Texas

at Arlington.

__________. ed. 1994. Discourse features of ten languages of West-Central

Africa. Dallas: The Summer Institute of Linguistics and The University of

Texas at Arlington.

__________. 2000. Aspect and backgrounding in Hebrew: lecture notes. n.p.

__________. 2002. Review of topic, focus and foreground in ancient Hebrew

narratives. Journal for the Study of the Old Testament: Supplement Series

295, by Jean-Marc Heimerdinger. 1999. Sheffield, England: Sheffield

Academic Press. JOTT 14:126-47.

Lewis, Paul and Elaine Lewis. 1984. Peoples of the Golden Triangle: Six tribes in

Thailand. London: Thames and Hudson.

Li, Charles N. 1991. The aspectual system of Hmong. Studies in languages

15:1.25-58.

Liang Tao. 1996. Topic discontinuity and zero anaphora in Chinese discourse:

Cognitive strategies in discourse processing. Studies in anaphora, ed. by

Barbara Fox, 487-513. Amsterdam: John Benjamins.

Lichtenberk, Frantisek. 1995. Apprehensional epistemics. Modality in grammar

and discourse, ed. by Joan Bybee and Suzanne Fleischman, 293-327.

Amsterdam: John Benjamins.

Liebert, Walf-Andreas; Gisela Redeker; and Linda Waugh (eds.) 1997. Discourse

and perspective in cognitive linguistics. Amsterdam: John Benjamins.

Liu, Mei-chun. 1996. Conceptual manipulation and semantic distinction: The case

of Mandarin postverbal ‘De’- complements. Conceptual structure,

discourse and language, ed. by Adele E. Goldberg, 333-46. Stanford, CA:

CSLI Publications.

Page 378: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

352

Lode, Lars. 1994. A discourse perspective on the significance of the Masoratic

accents. Biblical Hebrew and discourse linguistics, ed. by Robert D.

Bergen, 155-72. Dallas: SIL/ Winona Lake: Eisenbrauns.

Lombard, Sylvia. (n.d., 1st ed.) Yao language texts: Materials tape-recorded and

transcribed between 1967-1970 in Maechan, Chiang Rai, Thailand. New

York: Privately published.

__________. (n.d., 2nd ed.) Yao language texts: Materials tape-recorded and

transcribed between 1967-1970 in Maechan, Chiang Rai, Thailand. New

York: Privately published.

Lombard, Sylvia J. (compiler), and Herbert C. Purnell, Jr. (editor). 1968. Yao-

English dictionary. New York: Southeast Asia Program Department of

Asian Studies Cornell University.

Longacre, Robert E. 1960. String constituent analysis. Language 36:1.63-88.

__________. 1961. From tagma to tagmeme in Biblical Hebrew. A William

Cameron Townsend en el vigîsimoquinto aniversario del Instituto

Lingúéstico de Verano, 563-91. Mexico: The Summer Institute of

Linguistics.

__________. 1964. Grammar discovery procedures: A field manual. (Janua

Linguarum, Series Minor 33) The Hague: Mouton.

__________. 1967. The notion of sentence. Report of the Eighteenth Annual

Round Table Meeting on Linguistics and Language Studies, ed. by Edward

L. Bansitt, Jr., 15-25. Washington, D.C.: Georgetown University Press.

__________. 1968. Philippine languages: discourse, paragraph and sentence

structure. Santa Ana, Calf.: Summer Institute of Linguistics.

__________. 1970. Sentence structure as a statement calculus. Language

46:4.783-815.

__________. 1972a. Narrative versus other discourse genre. From sound stream to

discourse: Papers from the 1971 Mid-America Linguistics Conference,

Page 379: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

353

Novermber 12-13, 1971, ed. by D. G. Hays and D. M. Lance, 167-85.

Columbia: University of Missouri. Republished in 1974 in Ruth M. Brend

(ed.), Advances in tagmemics, 357-76. Amsterdam: North-Holland.

__________. 1972b. Hierarchy and universality of discourse constituents in New

Guinea languages: discussion. Washington D.C.: SIL and Georgetown

University Press.

__________. 1976. An anatomy of speech notions. Lisse: The Peter de Ridder

Press.

__________. 1977. Discourse genre. Proceedings from the 12th International

Congress of Linguistics. August 28-September 2. 1977. Vienna, 551-54.

__________. 1979a. The structure of the flood narrative. Journal of the American

Academy of Religion 47(1): 89-133.

__________. 1979b. The paragraph as a grammatical unit. Syntax and semantics

Vol. 12: discourse and syntax, ed. by Talmy Givón, 115-34. New York:

Academic Press.

__________. 1981. A spectrum and profile approach to discourse analysis. Text

1:4.337-59.

__________. 1982a. Verb ranking and the constituent structure of discourse.

Journal of the Linguistic Association of the Southwest 5:177-202.

__________. 1982b. Discourse typology in relation to language typology. Text

processing: Text analysis and generation, text typology and attribution—

Proceedings of Nobel Symposium 51, ed. by Sture Allén, 457-86.

Stockholm: Almqvist & Wiksell International.

__________. 1983a. The grammar of discourse. 1st ed. New York: Plenum Press.

__________. 1983b. Discourse analysis and dejargonized tagmemics. n.p., 1-30.

__________. 1983c. Spectrum, profile and constituency structure in text analysis.

Proceedings of the XIIIth International Congress of Linguistics, August

Page 380: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

354

29-September 4, 1982, Tokyo, ed. by Shirô Hattori and Kazuko Inoue,

1024-27. Tokyo: Published under the auspices of the CIPL (The Hague).

__________. ed. 1984. Theory and application in processing texts in non-

Indoeuropean languages. Hamburg: Helmut Buske Verlag Hamburg.

__________. 1985a. Interpreting Biblical stories. Discourse and literature: New

approaches to the analysis of literary genres, ed. by Teun A. van Dijk,

169-85. Amsterdam: John Benjamins.

__________. 1985b. Sentence as combinations of clauses. Language typology and

syntactic description, Vol. II Complex constructions, ed. by Timothy

Shopen, 235-86. Cambridge: Cambridge University Press.

__________. 1987. The semantics of the storyline in East and West Africa.

Journal of Semantics 5:51-64.

__________. 1989a. Joseph: a story of divine providence—a text theoretical and

textlinguistic analysis of Genesis 37 and 39-48. Winona Lake:

Eisenbrauns.

__________. 1989b. Two hypotheses regarding text generation and analysis.

Discourse processes 12: 413-60.

__________. 1990. Storyline concerns and word order typology in East and West

Africa. Supplement 10, Studies in African linguistics. Los Angeles: The

James S. Coleman African Studies Center and the Department of

Linguistics, University of California at Los Angeles.

__________. 1992. [originally 1989] Discourse perspective on the Hebrew verb:

Affirmation and restatement. Linguistics and Biblical Hebrew, ed. by

Walter R. Bodine, 177-89. Winona Lake: Eisenbrauns.

__________. 1994a. Weqatal forms in Biblical Hebrew prose: A discourse-

modular approach. Biblical Hebrew and discourse linguistics, ed. by

Robert D. Bergen, 50-98. Dallas: SIL/ Winona Lake: Eisenbrauns.

Page 381: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

355

__________. 1994b. The dynamics of reported dialogue in narrative. Word

45:2.125-43.

__________. 1995a. Building for the worship of God: Exodus 25:1-30:10.

Discourse analysis of Biblical literature, ed. by Walter R. Bodine, 21-49.

Atlanta, Georgia: Scholars Press.

__________. 1995b. Left shifts in strongly VSO languages. Word order in

discourse (Typological studies in language, 30), ed. by Pamela Downing

and Michael Noonan, 331-54. Amsterdam: John Benjamins.

__________. 1995c. Salience schemes for narrative discourse: Some broader

concerns. Twenty-second LACUS forum, ed. by Bates Hoffer, 5-14. Chapel

Hill: Linguistic Association of Canada and the United States.

__________. 1996. The grammar of discourse, 2nd ed. New York: Plenum Press.

__________. 1998. Review of the verbal system of classical Hebrew in the Joseph

story: An approach from discourse analysis, by Yoshinobu Endo. Hebrew

Studies 39.216-8.

__________. 2003a. Joseph: a story of divine providence—a text theoretical and

textlinguistic analysis of Genesis 37 and 39-48, 2nd ed. Winona Lake:

Eisenbrauns.

__________. 2003b. Holistic textlinguistics. SIL Electronic working papers. n.p.:

SIL International.

Longacre, Robert E., and Shin Ja J. Hwang. 1994. A textlinguistic approach to the

Biblical Hebrew narrative of Jonah. Biblical Hebrew and discourse

linguistics, ed. by Robert D. Bergen, 336-58. Dallas: SIL and Winona

Lake: Eisenbrauns.

Longacre, Robert E., and Stephen Levinsohn. 1978. Field analysis of discourse.

Current trends in texlinguistics (Research in text theory, vol. 2), ed. by

Wolfgang U. Dressler, 104-22. Berlin: Walter de Gruyter.

Page 382: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

356

Longacre, Robert E., and Frances Woods (eds.) 1976. Discourse grammar: Studies

in indigenous languages of Colombia, Panama, and Ecuador, Part 1.

Arlington: Summer Institute of Linguistics and University of Texas at

Arlington.

__________. 1977. Discourse grammar: Studies in indigenous languages of

Colombia, Panama, and Ecuador, Part 2. Arlington: Summer Institute of

Linguistics and University of Texas at Arlington.

Louw, J. P. 1992. Reading a text as discourse. Linguistics and New Testament

interpretation: essays on discourse analysis, ed. by David Alan Black. 17-

30. Nashville: Broadman.

Luchjenbroers, June. 1996. Schematic representation of discourse structure.

Conceptual structure, discourse and language, ed. by Adele E. Goldberg,

347-58. Stanford, CA: CSLI Publications.

Lyons, John. 1977. Semantics, vol. 2. Cambridge: Cambridge University Press.

Maes, Alfons, and Leonoor Oversteegen. 1999. Nominal vs. temporal

interpretation in discourse. Issues in cognitive linguistics: 1993

Proceedings of the International Cognitive Linguistics Conference, ed. by

Leon de Stadler and Christoph Eyrich, 549-66. Berlin: Mouton de Gruyter.

Magonet, Rabbi Jonathan. 1996. Character/author/reader: the problem of

perspective in Biblical narrative. Literary structure and rhetorical

strategies in the Hebrew Bible, ed. by L. J. de Regt, J. de Waard and J. P.

Fokkelman, 3-13. Assen, The Netherlands: Van Gorcum.

[Makino, Seiichi. 1983. Tense variation in narrative (Monogatari no bunsho

niokeru jisei no tenkan). Gengo (language) 12.109-17.] 牧野成一 1983「物

語の文章における時制の転換」『言語』12.109-17.

Mandelblit, Nili. 2000. The grammatical marking of conceptual integration: From

syntax to morphology. Cognitive Linguistics 11-3/4. 197-251.

Page 383: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

357

[Mao, Zong-Wu. 1992. Consice Chinese-Yao dictionary according to semantic

domain (Mien language). (Hàn-Yáo jiănmíng fēnlèi cídiăn) Chengdu:

Sichuan Minzu Chubanshe] 毛宗武. 1992 年,《汉瑶 简明分类词典 (勉

语)》, 中国社会科学院民族研究所主编, 成都: 四川民族出版社.

[Mao, Zong-Wu. 2004. Mien Dialectology of the Yao (Yáozú miănyǔ fāngyán

yánjiū). Beijing: Minzu Press.] 毛宗武. 2004 年. 《瑶族勉语方言研究》,

北京: 民族出版社

Mao, Tsung-wu, and Chou Tsu-yao. 1972. A brief description of the Yao

language. Miao and Yao linguistic studies: Selected articles in Chinese,

translated by Chang Yü-hung and Chu Kwo-ray, ed. by Herbert C. Purnell,

Jr., 239-55. New York: Southeast Asia Program Department of Asian

Studies, Cornell University.

Marchese, Lynell. 1978. Time reference in Godié. Papers on discourse, ed. by

Joseph Grimes, 63-75. Dallas: SIL.

Maschler, Yael. 1998. Rotsè lishmoa kéta? ‘wanna hear something weird/funny

[lit. ‘a segment’]?’: The discourse markers segmenting Israeli Hebrew

talk-in-interaction. Discourse markers: Descriptions and theory, ed. by

Andreas H. Jucker and Yael Ziv, 13-59. Amsterdam: John Benjamins.

Matisoff, Jamaes A. 1973. Tonogenesis in Southeast Asia. Consonant types and

tone, ed. by L. Hyman, 71-95. Southern California Occasional Papers in

Linguistics, No. 1. Los Angeles: UCLA.

__________. 1983. Linguistic diversity and language contact. Highlanders of

Thailand, ed. by John McKinnon and Wanat Bhruksasri, 56-86. Singapore:

Oxford University Press.

__________. 1991. Sino-Tibetan linguistics: Present state and future prospects.

Annual review of anthropology 20:469-504.

Page 384: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

358

__________. 2001. Genetic versus contact relationship: Prosodic diffusibility in

South-East Asian languages. Areal diffusion and generic inheritance:

Problems in comparative linguistics, ed. by Alexandra Y. Aikhenvald and

R. M. W. Dixon, 291-327. Oxford: Oxford University Press.

Matsumoto, Yo. 1996. How abstract is subjective motion? A comparison of

coverage path expressions and access path expressions. Conceptual

structure, discourse and language, ed. by Adele E. Goldberg, 359-73.

Stanford, CA: CSLI Publications.

McArthur, Harry S. 1979. The role of aspect in distinguishing Aguacatec

discourse types. Discourse studies in Mesoamerican languages Vol. 1

discussion, ed. by Linda K. Jones, 97-121. Dallas: The Summer Institute of

Linguistics and the University of Texas at Arlington.

McKinnon, John, and Wanat Bhruksasri (eds.) 1983. Highlanders of Thailand.

Oxford: Oxford University Press.

Miao Language Team. 1972. The role of the Chinese language in the enrichment

and development of the Yao language. Miao and Yao linguistic studies:

Selected articles in Chinese, translated by Chang Yü-hung and Chu Kwo-

ray, ed. by Herbert C. Purnell, Jr., 257-82. New York: Southeast Asia

Program Department of Asian Studies, Cornell University.

Miller, Cynthia L. 1994. Introducing direct discourse in Biblical Hebrew

narrative. Biblical Hebrew and discourse linguistics, ed. by Robert D.

Bergen, 199-241. Dallas: SIL and Winona Lake: Eisenbrauns.

Miller, George A., and Philip N. Johnson-Laird. 1976. Language and perception.

Cambridge, Massachusetts: The Belknap Press of Harvard University

Press.

Mongkol Chanbamrung and Thaworn Foofuang (eds.) 1995. The hill tribes of

Thailand. Chiang Mai, Thailand: Technical Service Club, Tribal Research

Institute, Chiang Mai University.

Page 385: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

359

Morgan, David. 1994. Semantic constraints on relevance in Lobala discourse.

Discourse features on ten languages of West-Central Africa, ed. by

Stephen H. Levinsohn, 125-49. Dallas: The Summer Institute of

Linguistics and The University of Texas at Arlington.

Mourelatos, Alexander P. 1981. Events, processes, and states. Tense and aspect,

Syntax and semantics Vol. 14, ed. by Philip J. Tedeschi and Annie Zaenen,

191-212. New York: Academic Press.

Myhill, John. 1992. Typological discourse analysis. Oxford: Blackwell.

__________. 2001. Typology and discourse analysis. The handbook of discourse

analysis, ed. by Deborah Schiffrin, Deborah Tannen, and Heidi E.

Hamilton, 161-74. Oxford: Blackwell.

Myhill, John, and Laura A. Smith. 1995. The discourse and interactive functions

of obligation expressions. Modality in grammar and discourse, ed. by Joan

Bybee and Suzanne Fleischman, 239-92. Amsterdam/Philadelphia: John

Benjamins.

Neubert, Albrecht. 1983. Discourse analysis of translation. Proceedings of the

XIIIth International Congress of Linguistics, August 29-September 4,

1982, Tokyo, ed. by Shirô Hattori and Kazuko Inoue, 1032-35. Tokyo:

Published under the auspices of the CIPL (The Hague).

Newman, Bonnie. 1978. The Longuda verb. Papers on discourse, ed. by Joseph

Grimes, 25-45. Dallas: SIL.

Newton, Dennis. 1992. The function of time words in Guarayu. Language in

context: essays for Robert E. Longacre, ed. by Shin Ja J. Hwang and

William R. Merrifield, 499-502. Arlington: The Summer Institute of

Linguistics and The University of Texas at Arlington.

Newmeyer, Frederick J. 1980. Linguistic theory in America: The first quarter-

century of Transformational Generative Grammar. New York: Academic

Press.

Page 386: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

360

Niccacci, Alviero. 1990. The syntax of the verb in Classical Hebrew prose.

(Journal for the Study of the Old Testament Supplement Series 86)

Sheffield: Sheffield Academic Press. (Translated by W. G. E. Watson

from the Italian Sintassi del verbo ebraico nella prosa biblica classica.

Jerusalem: Franciscan Printing Press, 1986)

__________. 1994a. On the Hebrew verbal system. Biblical Hebrew and

discourse linguistics, ed. by Robert D. Bergen, 117-37. Dallas:

SIL/Winona Lake: Eisenbrauns.

__________. 1994b. Analysis of Biblical narrative. Biblical Hebrew and

discourse linguistics, ed. by Robert D. Bergen, 175-97. Dallas:

SIL/Winona Lake: Eisenbrauns.

Niederer, Barbara. 1994. Problems in the reconstruction of Proto-Hmong-Mjen

rimes. Paper presented at the 27th International Conference on Sino-

Tibetan Languages and Linguistics. Paris, October 12-16, 1994. Paris:

n.publisher.

Ning, Yu. 1999. Spatial conceptualization of time in Chinese. Cultural,

psychological and typological issues in cognitive linguistics: Selected

papers of the bi-annual ICLA meeting in Albuquerque, July 1995, ed. by

Masako K. Hiraga, Chris Sinha, and Sherman Wilcox, 69-84. Amsterdam:

John Benjamins.

Nishida, Tatsuo. 1989. [Sino-Tibetan] The Sanseido encyclopaedia of linguistics,

Vol. 2 languages of the world, Part Two, ed. by Takashi Kamei, Rokuro

Kono, and Eiichi Chino, 166-87. Tokyo: Sanseido. 西田龍雄(1989)「シ

ナ・チベット語族」(亀井孝・河野六郎・千野栄一篇『言語学大辞典』

第2巻,中,166-87.東京:三省堂)

Nuyts, Jan, and Eric Pederson. 1997. Language and conceptualization.

Cambridge: Cambridge University Press.

Ono, Tsuyoshi, and Sandra A. Thompson. 1996. The dynamic nature of

conceptual structure building: Evidence from conversation. Conceptual

Page 387: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

361

structure, discourse and language, ed. by Adele E. Goldberg, 391-99.

Stanford, CA: CSLI Publications.

Östman, Jan-Ola, and Tuija Virtanen. 1999. Theme, comment, and newness as

figures in information structuring. Discourse studies in cognitive

linguistics: Selected papers from the Fifth International Cognitive

Linguistics Conference Amsterdam, July 1997, ed. by Karen van Hoek,

Andrej A. Kibrik, and Leo Noordman, 91-110. Amsterdam: John

Benjamins.

Palmer, Gary B. 1996. Toward a theory of cultural linguistics. Austin: The

University of Texas Press.

Pan Chengqian. 1991. Yao dialectology. The Yao of south China: Recent

international studies, ed. by Jacques Lemoine and Chiao Chien, 47-70.

Paris: Pangu.

[Pán Chéngqián (compiler) and Shū Huālóng (editor). 1988. Yao language primer.

(Yáowén yánjiū). Nanning Shi: Guangxi Minzu Chubanshe] 盘承乾, 舒化

龙 1988 年,《瑶文课本》, 广西壮族自治区少数民族语言文字工作委员

会编, 南宁市: 广西民族出版社.

Panh, Smith (Aka Koueifo Saephan). 1995. Mienh In-Wuonh Dimv Nangc Sou---

Mien-English Everyday Language Dictionary. 1st edition. Visalia, CA:

privately published.

Parret, Herman. 1971. Language and discourse. The Hague: Mouton.

Payne, David. 1984. Activity as the encoding of foregrounding in narrative: A

case study of an Asheninca legend. Theory and application in processing

texts in non-Indoeuropean languages, ed. by Robert E. Longacre, 49-91.

Hamburg: Helmut Buske Verlag Hamburg.

Payne, Thomas E. 1997. Describing morphosyntax: A guide for field linguistics.

Cambridge: Cambridge University Press.

Page 388: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

362

Peansiri Vongvipanond. 1998. Phonological variation and attitudinal meaning: A

case study of discourse markers in Thai. Journal of language and

linguistics, Vol. 17, No. 1, July-December 1998 (Phaasaa lae phaasaasaat

ภาษาและภาษาศาสตร ปที่ 17 ฉบับที่ 1 กรกฎาคม-ธันวาคม 2541 สาขาวิชา

ภาษาศาสตร คณะศลิปศาสตร มหาวทิยาลัยธรรมศาสตร), 69-90. Bangkok:

Department of Linguistics, Faculty of Liberal Arts, Thammasat University.

Peiros, Ilia. 1998. Comparative linguistics in Southeast Asia. Camberra: Research

School of Pacific and Asian Studies, The Australian National University.

Peregrin, Jaroslav. 1996. Topic and focus in a formal framework. Discourse and

meaning: Papers in honor of Eva Hajičová, ed. by Barbara H. Partee and

Peter Sgall, 235-54. Amsterdam: John Benjamins.

Person, Kirk Roger. 1993. Discourse considerations in the hortatory speech of

Phra Phayom. Chiang Mai: MA thesis, Payap University.

__________. 1996. Thailand’s “Straight-Talking” Monk: a discourse analysis of

the Hortatory speech of Phra Phayom Kalayano. Pan-Asiatic linguistics:

Proceedings of the fourth International Symposium of languages and

linguistics Jan. 8-10, ed. by Suwilai Premsrirat, 2. 767-92. Bangkok:

Instutue of Language and Culture for Rural Development, Mahidol

University at Salaya, Thailand.

Persons, David. 1979. Plot structure in Lachixio Zapotec discourse. Discourse

studies in Mesoamerican languages Vol. 1 discussion, ed. by Linda K.

Jones, 123-40. Dallas: The Summer Institute of Linguistics and the

University of Texas at Arlington.

Pickering, Wilbur. 1980. A framework for discourse analysis. Dallas: Summer

Institute of Linguistics.

Pike, Evelyn G. 1983. Grammatical and referential hierarchies in a prose text—

toward its systematic exegesis. Text and tagmeme, by Kenneth L. Pike and

Evelyn G. Pike, 4-73. London: Frances Printer.

Page 389: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

363

__________. 1988. Trees for constituent analysis from discourse to morpheme.

Workpapers concerning Waorani discourse features, ed. by Evelyn G.

Pike and Rachel Saint, 31-46. Dallas: The Summer Institute of Linguistics.

Pike, Kenneth L. 1964. Discourse analysis and tagmeme matrices. Oceanic

Linguist. 3:1.5-25. (Reprint in Advances in tagmemics, ed. by Ruth M.

Brend, Amsterdam: North-Holland Publishing Company, 1974, 285-305.)

__________. 1967. Language in relation to a unified theory of the structure of

human behavior, 2nd edition in one volume. The Hague: Mouton.

__________. 1981. Tagmemics, discourse, and verbal art. Michigan: Michigan

Studies in the Humanities, Horace H. Rackham School of Graduate

Studies, the University of Michigan.

__________. 1982. Linguistic concepts: An introduction to tagmemics. Lincoln:

University of Nebraska Press.

Pike, Kenneth L., and Evelyn G. Pike. 1982. Grammatical analysis (2nd edition).

Dallas: SIL.

[Prasit Leepreecha; Yanyong Trakanthamrong; and Wisut Leksombun. 2004.

Mien: Various lives from mountains to cities. The academic document in

ethnological study and development, Social Capital Series No. 2. Chiang

Mai: The Social Research Institute, Chiang Mai University.] ประสิทธิ์.

ลีปรีชา, ยรรยง ตระการธํารง, วิสุทธิ์ เหล็กสมบูรณ. 2547. เมี่ยน: หลากหลายชีวิตจาก

ขุนเขาสูเมือง. เอกสารวชิาการชาติพันธุศึกษาและการพฒันา ชุดทนุทางสังคม ลําดับ ที่

2, เชียงใหม: สถาบันวิจัยสังคม มหาวิทยาลันเชียงใหม

Preecha Chaturabhand. 1987. People of the hills. Bangkok: Editions Duang

Kamol.

Price, Norman. 1978. Nchimburu narrative events in time. Papers on discourse,

ed. by Joseph Grimes, 76-83. Dallas: SIL.

Page 390: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

364

Prince, Gerald. 1982. Narratology: the form and functioning of narrative. Berlin:

Mouton.

Proctor, Mersedeh. 1996. What marks language as discourse---from prose to

poetry. Solving language problems: from general to applied linguistics, ed.

by R. R. K. Hartmann, 132-68. Read Hall: University of Exeter Press.

Polinsky, Maria. 1996. Situation perspective: On the relations of thematic roles,

discourse categories, and grammatical relations to figure and ground.

Conceptual structure, discourse and language, ed. by Adele E. Goldberg,

401-19. Stanford, CA: CSLI Publications.

Pourret, Jess G. 2002. The Yao: The Mien and Mun Yao in China, Vietnam, Laos

and Thailand. Bangkok: River Books.

Purnell, Herbert Charles, Jr. 1965. Phonology of a Yao dialect spoken in the

province of Chiangrai, Thailand. Hartford, Connecticut: MA thesis, The

Hartford Seminary Foundation.

__________. 1970. Toward a reconstruction of proto-Miao-Yao. Ph.D.

dissertation, Cornell University.

__________. 1972. Toward contrastive analyses between Thai and hill tribe

languages: Some phonetic data. Tai phonetics and phonology, ed. by

Jimmy G. Harris and Richard B. Noss, 123-30. Bangkok: Central Institute

of English Language Office of State Universities, Mahidol University.

__________. ed. 1972. Miao and Yao linguistic studies: Selected articles in

Chinese, translated by Chang Yü-hung and Chu Kwo-ray. New York:

Southeast Asia Program Department of Asian Studies, Cornell University.

__________. 1985. Developing practical orthographies for the Iu Mien (Yao),

1932-1985: A case study. Linguistics of the Tibeto-Burman Area

10(2):128-41. (Paper presented at the 18th International Conference on

Sino-Tibetan Linguistics and Languages, Bangkok, Ausust 1985.

Originally also presented at the 1983 annual meeting of the American

Page 391: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

365

Association for Applied Linguistics. Minneapolis: The American

Association for Applied Linguistics.)

__________. 1991. The metrical structure of Yiu Mien secular songs. The Yao of

south China: Recent international studies, ed. by Jacques Lemoine and

Chiao Chien, 369-96. Paris: Pangu.

__________. 2001. Filling out the lyrics: Observation on padding syllables in Iu

Mien Yao ‘reading style’ songs. Paper presented at SEALS XI, Bangkok,

May 2001.

Purnell, Herbert C.; Zanh Gueix-Fongc; and V. Ann Burgess, (forthcoming).

Mienh – English Dictionary.

Ramsey, S. Robert. 1987. The languages of China. New Jersey: Princeton

University Press.

Ratliff, Martha. 1992a. Meaningful tone: A study of tonal morphology in

compounds, form classes, and expressive phrases in White Hmong. De

Kalb, Ill.: Center for Southeast Asian Studies, Northern Illinois University.

__________. 1992b. Hand/arm in Hmong-Mien. Paper presented at Hmong-Mien

panel at the XXVth International Conference on Sino-Tibetan Languages

and Linguistics, University of California at Berkeley, October 14-18,

1992. Berkeley.

Ricoeur, Paul. 1984. Time and narrative, Vol. 1. Chicago: The University of

Chicago Press. (Translated by Kathleen McLaughlin and David Pellauer

from the French Temps et récit, vol. 1. Editions du Seuil, 1983)

__________. 1985. Time and narrative, Vol. 2. Chicago: The University of

Chicago Press. (Translated by Kathleen McLaughlin and David Pellauer

from the French Temps et récit, vol. 2. Editions du Seuil, 1984)

__________. 1988. Time and narrative, Vol. 3. Chicago: The University of

Chicago Press. (Translated by Kathleen McLaughlin and David Pellauer

from the French Temps et récit, vol. 3. Editions du Seuil, 1985)

Page 392: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

366

Rong Chen. 1992. Logical connectors, grounding and cognition. Language

Sciences 13, 3/4:433-45.

Salanskie, Jean-Michel. 1994. Continuum, cognition and linguistics. Continuity in

linguistic semantics, ed. by Catherine Fuchs and Bernard Victorri, 127-66.

Amsterdam: John Benjamins.

Sanders, José. 1996. Perspective and attribution: The cognitive representation of

biblical narrative. Poetics 24:57-80.

Sang Hwan Seong. 2001. Transitivity parameter and prominence typology: A

cross-linguistic study. Applied cognitive linguistics I: Theory and

language acquisition, ed. by Martin Pütz; Susnne Niemeier; and René

Dirven, 131-63.

Saussure, Ferdinand de. 1910-1911. Troisième Cours de Linguistique Générale

(1910-1911): d’après les cahiers d’Emile Constantin, French text ed. by

Eisuke Komatsu, English translation by Roy Harris as Saussure’s third

course of lectures on general linguistics (1910-1911), 1993. Oxford:

Pergamon.

__________. 1959. Course in general linguistics: Edited by Charles Bally and

Albert Sechehaye in collaboration with Albert Riedlinger, translated by

Wade Baskin. New York: McGraw-Hill. (First published 1915 as Cours de

Linguistique Générale. Paris: Pyot.)

Schäffner, Christina, and Helen Kelly-Holmes (eds.) 1996. Discourse and

ideologies. Clevedon: Multilingual Matters.

Schiffrin, Deborah. 1987. Discourse markers. Cambridge: Cambridge University

Press.

__________. 1994. Approaches to discourse. Oxford: Blackwell.

Schiffrin, Deborah; Deborah Tannen; and Heidi E. Hamilton. 2001. The handbook

of discourse analysis. Oxford: Blackwell.

Page 393: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

367

Schliesinger, Joachim. 1998. Hill tribes of Vietnam, Vol. 2: Profile of the existing

hill tribe groups. Bangkok: White Lotus.

__________. 2003. Ethnic groups of Laos, Vol. 3: Profile of Austro-Thai-

speaking peoples. Bangkok: White Lotus.

Schrock, Joann L. et al. 1970. Minority groups in Thailand (Ethnographic study

series). Washington, D.C.: Department of the Army.

Scollon, Ron, and Suzanne Wong Scollon. 1995. Intercultural communication: A

discourse approach. London: Blackwell.

Seuren, Pieter A. M. 1985. Discourse semantics. Oxford: Blackwell.

Sgall, Petr. 1987. The position of Czech linguistics in theme-focus research.

Language topics: Essays in honour of Michael Halliday Vol. I, ed. by Ross

Steele and Terry Threadgold, 47-55.Amsterdam: John Benjamins.

__________. 1994. Meaning, reference and discourse patterns. The Prague School

of struactural and functional linguistics: a short introduction, ed. by Philip

A. Luelsdorff, 277-309. Amsterdam: John Benjamins.

Shimasaki, Katsuomi. 2002. Focus structure in Biblical Hebrew: A study of word

order and information structure. Bethesda, Maryland: CDL Press.

Shiratori, Yoshiro. 1974. Yao documents. Tokyo: Kodansha.

Shloush, Shelley. 1998. A unified account of Hebrew bekicur ‘in short’:

Relevance Theory and discourse structure considerations. Discourse

markers: Descriptions and theory, ed. by Andreas H. Jucker and Yael Ziv,

62-82. Amsterdam: John Benjamins.

[Shū Huā-lóng. 1992. Modern Yao language study (Xiàndài Yáoyǔ yánjiū).

Nanning Shi: Guangxi Minzu Chubanshe] 舒化龙 1992 年,《现代瑶语研

究》, 南宁市: 广西民族出版社.

Silva-Corvalán, Carmen. 1995. Contextual conditions for the interpretation of

‘poder’ and ‘deber’ in Spanish. Modality in grammar and discourse, ed. by

Page 394: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

368

Joan Bybee and Suzanne Fleischman, 67-105. Amsterdam: John

Benjamins.

Smith, Carlota S. 2003. Modes of discourse: The local structure of texts.

Cambridge: Cambridge University Press.

Somsonge Burusphat. 1990a. The functions of k√/ in oral Kui narrative. Mon-

Khmer Studies 18-19. 223-31.

__________. 1990b. The bipartite structure of information in Thai narrative

discourse. [Language and Linguistics (Phaasaa lae phaasaasaat)] 9:1.65-

78. (ภาษาและภาษาศาสตร ปที่ 9 ฉบับที่ 1 กรกฎาคม-ธันวาคม 2533 สาขาวิชา

ภาษาศาสตร คณะศลิปศาสตร มหาวทิยาลัยธรรมศาสตร ) Bangkok: Thammasat

University.

__________. 1991a. The structure of Thai narrative. Dallas: The Summer Institute

of Linguistics and The University of Texas at Arlington.

__________. 1991b. Kui narrative repetition. Paper presented at the 24th

International Conference on Sino-Tibetan Languages and Linguistics,

Ramkhamhaeng University and Chiang Mai University, October 7-11,

1991.

__________. 1992a. Highlighting through the particle sw in Sgaw Karen

narrative discourse. Papers from the First Annual Meeting of the SEA

Linguistics Society 1991, ed. by Martha Ratliff and Eric Schiller, 47-63.

Tempe: Arizona State University.

__________. 1992b. The identification of storyline in Thai narrative discourse.

Language in context: Essays for Robert E. Longacre, ed. by Shin Ja J.

Hwang and William R. Merrifield, 419-33. Dallas: The Summer Institute

of Linguistics and The University of Texas at Arlington.

Page 395: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

369

__________. 2002. The temporal movement of the Hlai (Li) origin myth.

Collected papers on Southeast Asian and Pacific Languages, ed. by Robert

S. Bauer, 141-51. Canberra: Pacific Linguistics.

__________. n.d. Highlighting through the particle sw in Sgaw Karen narrative

discourse.

Somsonge Burusphat and Chalermchai Chaichomphu. 2000. Theme identification

in Lanna Buddhist sermons. [Collected papers on language and culture for

rural development], 53-81. [Salaya, Nakorn Pathom: The Institute of

Language and Culture for Rural Development, Mahidol University.] 2543.

รวมบทความทางวิชาการ ภาษาและวัฒนธรรมเพื่อพัฒนาชนบท. ตําบลศาลายา อําเภอ

พุทธมฌฑล จังหวัดนครปฐม: สถาบันวิจัยภาษาและวัฒนธรรมเพื่อพัฒนาชนบท

มหาวิทยาลัยมหิดล

Sperber, Dan and Deidre Wilson. 1986. Relevance: Communication and

cognition. Oxford: Blackwell.

Steele, Ross, and Terry Threadgold (eds.) 1987. Language topics: Essays in

honour of Michael Halliday, Volumes I and II. Amsterdam: John

Benjamins.

Steiner, Erich H., and Robert Veltman (eds.) 1988. Pragmatics, discourse and text:

Some Systemically-inspired approaches. London: Pinter.

Stipp, H. J. 1991. Wehaya für nichtiterative Vergangenheit? Zu syntaktischen

Modernisierungen im masoretischen Jeremiabuch. Text, Methode und

Grammatik: Wlfgang Richter zum 65. Geburtstag, ed. by W. Gross; H.

Irsigler; and T. Seidl, 521-547. St. Ottilien: EOS.

Stubbs, Michael. 1983. Discourse analysis: The sociolinguistic analysis of natural

language. Chichago: The University of Chicago Press.

[Suriya Ratanakul. 1988. Language in Southeast Asia Part 1: Austroasiatic and

Sino-Tibetan languages. Salaya, Nakorn Pathom: Institute of Language

Page 396: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

370

and Culture for Rural Development, Mahidol University.] สุริยา รัตนกุล.

2531. นานาภาษาในเอเซยีอาคาเนย ภาคที่ ๑ ภาษาตระกลูออสโตรเอเซียตกิ และ

ตระกูลจีน-ทิเบต. สาลายา อําเภอพุทธมณฑล จังหวัดนครปฐม : สถานันวิจัยภาษาและ

วฒนธรรมเพื่อพัฒนาชนบท มหาวิทยาลัยทหิดล

Suzuki, Satoko. 1998. Pejorative connotation: A case of Japanese. Discourse

markers: Descriptions and theory, ed. by Andreas H. Jucker and Yael Ziv,

261-76. Amsterdam: John Benjamins.

Swales, John M. 1990. Genre analysis: English in academic and research settings.

Cambridge: Cambridge University Press.

Sweetser, Eve. 1990. From etymology to pragmatics: Metaphorical and cultural

aspects of semantic structure. Cambridge: Cambridge University Press.

Technical Service Club. 2004. The hill tribes of Thailand, 4th ed. Chiang Mai:

Tribal Musium.

Takahashi, Kiyoko, and Kingkarn Thepkanjana. 1997. Negation in Thai serial

verb constructions: A pragmatic study. Southeast Asian Linguistic Studies

in honour of Vichin Panupng, ed. by Arthur S. Abramson, 273-82.

Bangkok: Chulalongkorn University Press.

Talmy, Leonard. 2000a. Toward a cognitive semantics, Vol. 1: Concept

structuring systems. Cambridge, Massachusetts: The MIT Press.

__________. 2000b. Toward a cognitive semantics, Vol. 2: Typology and process

in concept structuring. Cambridge, Massachusetts: The MIT Press.

Talstra, Eep. 1992. Text grammar and Biblical Hebrew: The viewpoint of

Wolfgang Schnerder. JOTT 5:269-97.

Tan Chee Beng. 1986. A legendary history of the origin of the Yao people.

Farmers in the hills: Ethnographic notes on the upland peoples of North

Page 397: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

371

Thailand, ed. by Anthony R. Walker, 55-72. Singapore: Suvarnabhumi

Books.

[Tán Guāng-Guăng, et al. ed.1993. Religions and customs of minority peoples in

China (Chuugoku shousuu minzoku-no shinkou to shuuzoku), Vol. 2,

translated by Masako Hayasi, ed. by Seiji Itoh from Chinese The survey of

minority peoples’ religions in China, Vol. 2, (1982). Tokyo: Dai-ichi

Shobo.] 覃光广 编著 1982 年,《中国少数民族宗教概览》, 内部刊物

Taylor, John R. 2003. Linguistic categorization: prototypes in linguistic theory

(3rd ed.). Oxford: Oxford University Press.

Tehan, Thomas M. 2000. A preliminary study of narrative discourse features in

the Greek text fo Matthew 8 and 9: Pre-publication draft. (Technical paper

# 56) Chiang Mai: Payap Research and Development Institute and SIL

International.

[Theraphan L.-Thongkum.1988a. Tones and their phonetic shape in Yao (Mien)

of Baan Huay Mae Saai, Mae Yaau, Müuang District, Chiang Rai

province. Research paper presented to the Faculty of Arts, Chulalongkorn

University] ธีระพันธ ล. ทองคํา. 2531. วรรณยุกตและพิสัยระดับเสียงวรรณยุกตใน

ภาษาเยา (เมีย่น) บานหวยแมซาย ตําบลแมยาว อําเภอเมอืง จงัหวดัเชยีงราย.

รายงานผลการวิจัยเสนอตอคณะอักษรศาสตร จุฬาลงกรณมหาวิทยาลัย. หนวยปฏิบัติ

การวิจัยทางภาษาศาสตร ภาควิชาภาษาศาสตร คณะอ ักษรศาสตร จุฬาลงกรณมหา

วิทยาลัย กรุงเทพฯ :จุฬาลงกรณมหาวิทยาลัย

__________. 1988b. Different dress and times of migration do not always mean

different ways of talking: A case study of Mien-Yao (Iu Mien) tones.

Paper presented at the 21st Sino-Tibetan Conference on Language and

Linguistics, Lund, Sweden, 7-9 October, 1988. (Reprinted in Theraphan

1989:112-29)

Page 398: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

372

[__________. 1989. Tones in Yao. Research report in Rachadaphiseksomphot

Fund, Chulalongkorn University. Bangkok: Chulalongkorn University.]

ธีระพันธ ล. ทองคํา. 2532. วรรณยุกตในภาษาเยา จุฬาลงกรณมหาวิทยาลัย ทุนวิจัย

รัชดาเษกสมโถช รายงานผลการวิจัย กรุงเทพฯ :จุฬาลงกรณมหาวิทยาลัย

[__________. 1991. Guo Shan Bang: Perpetual redaction of the imperial decree of

Emperor Ping Huang for protection when traveling in the hills. A research

project on various aspects of the language and culture of Yao people living

in Thailand and China. Bangkok: Linguistics Research Unit, Faculty of

Arts, Chulalongkorn University.] ธีระพนัธ ล. ทองคํา. 2534.เกีย เวน็ ปอง :

พระราชสาสนพระเจาผิงหวางใชปองกันตัวสําหรับการเดินทางขามภูเขาฉบบถาวร.

โครงการวิจัยภาษาศาและวัฒนะธรรมเยาไทยและเยาจีน หนวยปฏิบัตการวิจัยทาง

ภาษาศาสตร คณะอักษรศาสตร จุฬาลงกรณมหาวิทยาลัย. กรุงเทพฯ : จุฬาลงกรณ

มหาวิทยาลัย

__________. 1993. A view on Proto-Mjuenic (Yao). Mon-Khmer studies: A

Journal of Southeast Asian Languages, 22:163-230.

__________. 1997. Tone change and language contact: A case study of Mien-Yao

and Thai. Southeast Asian linguistic studies in honour of Vichin Panupong,

ed. by Arthur S. Abramson, 153-60. Bangkok: Chulalonggorn University

Press.

Thompson, Geoff. 1996. Introducing Functional Grammar. London: Arnold.

Thompson, Sandra A. 1997. Discourse motivations for the core-oblique

distinction as a language universal, Directions in Functional Linguistics,

ed. by Akio Kamio, 59-82. Amsterdam: John Benjamins.

Thompson, Sandra A., and Robert E. Longacre. 1985. Adverbial clauses.

Language typology and syntactic description, Vol. II Complex

Page 399: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

373

constructions, ed. by Timothy Shopen, 171-234. Cambridge: Cambridge

University Press.

Tomlin, Russell S. 1987. Linguistic reflections of cognitive events. Coherence

and grounding in discourse, ed. by Russel S. Tomlin, 455-79. Amsterdam:

John Benjamins.

Tomlin, Russell S.; Linda Forrest; Ming Ming Pu; and Myung Hee Kim. 1997.

Discourse semantics. Discourse studies: A multidisciplinary introduction,

Vol. 1, ed. by Teun A. van Dijk, 63-111. London: SAGE Publications.

Toolan, Michael. 2001. Narrative: A critical linguistic introduction, 2nd ed.

London: Routledge.

Trask, R. L. 1993. A dictionary of grammatical terms in linguistics. London:

Routledge.

[Tsuji, Yukio, ed. 2002. An encyclopedic dictionary of Cognitive Linguistics.

Tokyo: Kenkyusha.] 辻幸夫『認知言語学キーワード事典』研究社

Tsumura, David Toshio. 1989. The earth and the waters in Genesis 1 and 2: A

linguistic investigation. (Journal for the Study of the Old Testament

Supplement Series 83) Sheffield: JSOT Press.

__________. 2005. Creation and destruction: A reappraisal of the Chaoskampf

theory in the Old Testament. Winona Lake, Indiana: Eisenbrauns.

[Tsunemi, Jun-ichi. 1980. The Yao pathfinder for migration and the establishment

of new villages in North Thailand. The societies and cultures of Southeast

Asia and Indian Subcontinent: Jubilee volume presented to Dr. Tatsurō

Yamamoto on the occasion of his seventieth birthday. Vol. 2. compiled by

the editorial committee, 185-211. Tokyo: Yamakawa Shuppansha.] 常見

純一 1980.「ヤオ族の移住と村落の形成——マーン=ラーン=トン

(「国見」)を中心として——山本達郎博士古希記念『東南アジア・イ

ンドの社会と文化・下』山本達郎博士古希記念論叢篇集委員会篇、185-

211. 東京:山川出版社。

Page 400: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

374

[Uthai Worasuwannarak. 1982. The influence on folktale on teaching and learning

Thai language in the Yao hill tribe at Ban Huai Sai. M.Ed. thesis,

Chulalongkorn University.] อุทัย วรสุวรรณรักษ 2525. อิทธิพลของนิทานพื้น

บานที่มีตอการเรียนการสอนวิชาภาษาไทยของชาวไทยเผาเยา บานหวยแมซาย. วิทยา

นิพลธ ปริญญาครุศาสตรมหาบัณฑิต ภาควิชามัธยมศึกษา บัณฑิตวิทยาลัย จุฬาลงกรณ

มหาวิทยาลัย. กรุงเทพฯ : จุฬาลงกรณมหาวิทยาลัย

van der Merwe, Christo H.J. 2002. A critical analysis of narrative syntactic

approaches, with special attention to their relationship to discourse

analysis. Narrative syntax and the Hebrew Bible: Papers of the Tilburg

Conference 1996, ed. by Ellen van Wolde, 133-56. Boston: Brill Academic

Publishers.

Van Dijk, Teun A. 1972. Some aspects of text grammars: A study in theoretical

linguistics and poetics. The Hague: Mouton.

__________. 1977. Text and context: Explorations in the semantics and

pragmatics of discourse. London: Longman.

__________. 1981. Studies in the pragmatics of discourse. The Hague: Mouton.

__________. ed. 1985. Handbook of discourse analysis, Vol. 1: Disciplines of

discourse. London: Academic Press.

Van Dijk, Teun A. and Walter Kintsch. 1978. Cognitive psychology and

discourse: Recalling and summarizing stories. Current trends in

texlinguistics (Research in text theory, vol. 2), ed. by Wolfgang U.

Dressler, 61-80. Berlin: Walter de Gruyter.

__________. 1983. Strategies of discourse comprehension. New York: Academic

Press.

Van Hoek, Karen. 1997. Anaphora and conceptual structure. Chicago: The

University of Chicago Press.

Page 401: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

375

Van Valin, Robert D., Jr. 2005. Exploring the syntax-semantics interface.

Cambridge: Cambridge University Press.

Van Valin, Robert D., Jr. and Randy J. LaPolla. 1997. Syntax: Structure, meaning

and function. Cambridge: Cambridge University Press.

Wallace, Stephen. 1982. Figure and ground: The interrlationships of linguistic

categories. Tense-aspect: Between semantics & pragmatics, Containing

the contributions to a Symposium on Tense and Aspect, held at UCLA,

May 1979, ed. by Paul J. Hopper, 201-23. Amsterdam: John Benjamins.

Waltke, Bruce K., and M. O’Connor. 1990. An introduction to Biblical Hebrew

syntax. Winona Lake, Indiana: Eisenbrauns.

Whaley, Lindsay J. 1997. Introduction to typology: The unity and diversity of

language. Thousand Oaks: SAGE Publications.

Wharry, Cheryl. 2003. Amen and Halleluja preaching: Discourse functions in

African American sermons. Language in society 32.203-25.

Wierzbicka, Anna. 1988. The semantics of grammar. Amsterdam: John

Benjamins.

Winther-Nielsen, Nicolai. 1992. ‘In the beginning’ of Biblical Hebrew discourse:

Genesis 1:1 and the fronted time expression. Language in context: Essays

for Robert E. Longacre, ed. by Shin Ja J. Hwang and William R.

Merrifield, 67-80. Dallas: The Summer Institute of Linguistics and The

University of Texas at Arlington.

Wise, Mary Ruth. 1971. Identification of participants in discourse: A study of

aspects of form and meaning in Nomatsiguenga [originally submitted as a

Ph.D. dissertation to The University of Michigan in 1968]. Norman: The

Summer Institute of Linguistics/The University of Oklahoma, Norman.

Xia Li and Nancy B. Crane. 1996. Electronic styles: A handbook for citing

electronic information, 2nd ed. Medford, N.J.: Information Today, Inc.

Page 402: A STORYLINE ANALYSIS IN IU-MIENH NARRATIVE DISCOURSE WITH SOME INSIGHTS FROM COGNITIVE LINGUISTICS - Daniel Arisawa - Thesis

376

Yajai Chuwicha. 1985. k3―The multi-purpose marker in oral Thai narrative: A

discourse analysis. A paper presented at the 18th International Conference

on Sino-Tibetan Language and Linguistics, August 28, 1985.

Yip, Po-Ching and Don Rimmington. 1997. Chinese: An essential grammar.

London: Routledge.

Zee Yunyang. 1991. A comparison of Ba Pai Yao and Guoshan Yao tones. The

Yao of South China: Recent international studies, ed. by Jacques Lemoine

and Chiao Chien, 71-86. Paris: Pangu, Editions de l’A.F.E.Y.

[Zhèn Bīn. 1993. A study on Yao culture (Yáozú wénhuàlùn). Kunming: Yunnan

Renmin Press.] 阵斌 1993 年,《瑶族文化论》, 昆明: 云南人民出版社.

Ziv, Yael. 1998. Hebrew kaze as a discourse marker and lexical hedge:

Conceptual and procedural properties. Discourse markers: Descriptions

and theory, ed. by Andreas H. Jucker and Yael Ziv, 203-21. Amsterdam:

John Benjamins.