Dynamics of functional meanings in discourse

The main body of this thesis is organized as follows; Chapter 1: Introduction considers the philosophical motivations for this research; Chapter 2: Data and Methodology introduces the te

Trang 1

THE DYNAMICS OF FUNCTIONAL MEANINGS IN

DEPARTMENT OF ENGLISH LANGUAGE AND LITERATURE

NATIONAL UNIVERSITY OF SINGAPORE

2010

Trang 2

ACKNOWLEDGEMENTS

The opportunity to undertake truly invigorating research does not come easily

I would like to express my heartfelt thanks to my supervisor Kay O’Halloran I am deeply grateful to her for accepting me into her research group, despite my newness

to the field of Systemic Functional Linguistics I am equally indebted to Kevin Judd for believing I was suitable for the challenging research project

I would also like to thank all research staff and fellow postgraduate students

at the Multimodal Analysis Lab, Interactive and Digital Media Institute, National University of Singapore, for all their personal support, fellowship and friendship through the years: Brad, Marissa, Nizah, Yiqiong, Liu Yu, Melany, Monica, Dezheng, Sabine, Thiha, Victor and Bertrand

Just four years ago I was not even aware of the existence of linguistics I owe

it to Tara Mohanan and Kim Chonghyuck for opening up this world for me in their linguistics undergraduate teaching, and writing recommendations for my application into this MA program

This work is part of the Socio-Cultural Modeling Project at the Multimodal Analysis Lab, supported by the US Air Force Office of Scientific Research (AFOSR) through the Asian Office of Aerospace Research and Development (AOARD) under Research Grant FA2386-09-1-4008 AOARD 094008

Trang 3

TABLE OF CONTENTS

Volume I

TITLE PAGE i

ACKNOWLEDGEMENTS ii

TABLE OF CONTENTS iii

SUMMARY vi

LIST OF TABLES vii

LIST OF FIGURES xi

Chapter 1:

INTRODUCTION 1

Chapter 2:

DATA AND METHODOLOGY 4

2.1 The CNN Story 4

2.2 Systemics (The Software) 10

Chapter 3:

SYSTEMIC FUNCTIONAL ANALYSES 18

3.1 Logical Complexes 18

3.1.1 Concise Complexes 19

3.1.2 Detailed Complexes 24

3.2 Theme 32

3.2.1 Wave Metaphor for the Textual Metafunction 32

3.2.2 Short Range Theme - Rheme Dynamics 37

3.2.3 Thematic Progression (Daneš) 48

3.2.4 Marked Themes 55

3.2.5 Interpersonal Themes 56

3.3 Transitivity 59

3.3.1 Relational Clauses 59

3.3.2 Material Clauses 67

3.3.3 Minority Type Clauses 71

3.3.4 Some Individual Clauses 73

3.4 Ergativity 80

3.4.1 Medium 80

Trang 4

3.4.2 Agency 88

3.5 Mood 92

3.5.1 Mood and Speech Function 92

3.5.2 Modality and Mood Adjuncts 97

3.5.3 Modality in Transitivity and Grammatical Metaphor 100

3.5.4 Tense and Polarity 101

3.5.5 Rankshifting 105

3.6 Grammatical Metaphors 107

3.6.1 Experiential Shifts 110

3.6.2 Logical, Interpersonal and Textual Shifts 116

3.6.3 High Density Clauses 119

3.6.4 Distribution of Grammatical Metaphors 122

3.6.5 ‘Void’ of Grammatical Metaphors 125

3.6.6 Parallel Patterns with Rankshifted and Interrupting clauses 128

Chapter 4:

RECURRENCE PLOTS 130

4.1 Clauses as Vectors 130

4.2 Types of Recurrence Plots 134

4.2.1 XOR count: Difference in Meaning 134

4.2.2 AND count: Similarity in Meaning 135

4.2.3 OR count: Sum of Meanings 136

4.2.4 Reduced Vectors 137

4.2.5 Tag Decomposition 137

4.3 Plots of Individual Metafunctions 139

4.3.1 Mood 139

4.3.2 Theme 145

4.3.3 Transitivity 154

4.4 Three Level Investigation 161

4.4.1 First Level of Delicacy 162

4.4.2 Second Level of Delicacy 165

4.4.3 Third Level of Delicacy 168

4.5 Combined Metafunctions 172

Chapter 5:

SINGULAR VALUE DECOMPOSITION (SVD) 178

Trang 5

5.1 Experiment 1: Main Story 180

5.1.1 Feature F0 184

5.2 Experiment 2: Comments 219

Chapter 6:

CONCLUSION 221

REFERENCES 224

Trang 6

SUMMARY

This thesis explores the dynamics of meaning in discourse, based on the

theoretical framework of Systemic Functional (SF) Grammar Meaning is a

multidimensional concept, spanning the textual, ideational and interpersonal

metafunctions (Halliday 1978) In this thesis, meaning is described not just

qualitatively in words, but also quantitatively as well, in discrete and non-discrete numbers, vectors and matrices, and also presented visually in the form of colour diagrams of the clauses, the grammar trees, and a variety of plots inspired from

research in Dynamical Systems (Judd 2009a, b, c) The goal of this research is to arrive at a deeper understanding of language and its meaning in our world

The main body of this thesis is organized as follows; Chapter 1: Introduction considers the philosophical motivations for this research; Chapter 2: Data and

Methodology introduces the text and the computer software (Systemics) for analysis; Chapter 3: Systemic Functional Analyses interprets and explains the linguistic

significance of the completed analysis; Chapter 4: Recurrence Plots and Chapter 5: Singular Value Decomposition (SVD) advances the linguistic analysis by exploring the novel use of mathematical tools to extract and visualize interesting patterns of meaning in the discourse; and Chapter 6: Conclusion provides a summary of the main achievements of this research

Trang 7

LIST OF TABLES

Table 3-1: Theme analysis of first clause 32

Table 3-2: Theme analysis of Set A - clauses with the word “crisis” 33

Table 3-3: Theme analysis of Set B - clauses with the word “credit” 34

Table 3-4: Thematic progression in successive clauses 39

Table 3-5: Example of RHTH: R19 T20 42

Table 3-6: Actual frequencies of short range thematic progression types 42

Table 3-7: Predicted frequencies of short range thematic progression types 45

Table 3-8: Thematic progression with Theme substitution 47

Table 3-9: Short range Thematic Progression type 3 (TP3) with implicit [T] 49

Table 3-10: Short range Thematic Progression type 3 (TP3) with explicit [T] 50

Table 3-11: Long range Thematic Progression type 3 (TP3) 51

Table 3-12: Thematic Progression type 1 (TP1) 52

Table 3-13: Thematic Progression type 2 (TP2) 53

Table 3-14: The consequence of clause 50 & 51 54

Table 3-15: Thematic Progression type 2 (TP2) with complex Theme 54

Table 3-16: Marked Themes functioning as location in time 55

Table 3-17: Interpersonal Themes 56

Table 3-18: Clause 56 (in opposition to clause 55) 56

Table 3-19: Interpersonal Modal-Metaphors as Theme 57

Table 3-20: Transitivity types in the text 59

Table 3-21: All relational attributive clauses in the text (underlined: Attribute) 60

Table 3-22: All negative Finites in the text 61

Table 3-23: All relational identifying clauses in the text (underlined: Value) 62

Table 3-24: Tightness of encoding vs decoding equivalence relations 65

Table 3-25: Relational clauses with rankshifted material processes underlined 66

Table 3-26: All material clauses in the text 68

Table 3-27: Clauses with Goals in non-ranking material processes 69

Table 3-28: Construal of the impact of the crisis in clauses without Goal 70

Table 3-29: Minority transitive process types I – mental (6 clauses) 71

Table 3-30: Minority transitive process types II – behavioural/verbal/existential

(3 clauses) 72

Trang 8

Table 3-31: Metaphorical verbal clause 44 interpreted as relational meaning 72

Table 3-32: A relational clause which ultimately construes material and mental processes 73

Table 3-33: A mental clause which portrays material events in grammatical metaphors 74

Table 3-34: Congruent usage of grammatical metaphors in clause 23 74

Table 3-35: Word group expansion in clause 23 75

Table 3-36: Categories of frequency occurring Mediums in the text 83

Table 3-37: Medium “we” in clauses at/near boundaries of logico complexes 84

Table 3-38: Voice in the text (ranking clauses) 88

Table 3-39: Types of Effective Voice (ranking) clauses in the Text 88

Table 3-40: Material clauses in the text with non-ellipsed Agent 89

Table 3-41: Agency in the salad analogy 89

Table 3-42: Hidden Agent in the salad analogy 90

Table 3-43: Rankshifted processes with government/policymaker as potential Agent 90

Table 3-44: Ellipsed Agent: government 90

Table 3-45: Rankshifted Agent: government 91

Table 3-46: Mood of clauses in the text 92

Table 3-47: Speech Function in the text 92

Table 3-48: Clauses with non-declarative mood in the text 93

Table 3-49: Nexus of clause 35 and 36 95

Table 3-50: Clauses with Speech Function = Command 96

Table 3-51: Clause with Speech Function = Offer 96

Table 3-52: Modality Orientation in the text 97

Table 3-53: Clauses with Modal Finites (examples) 98

Table 3-54: Relational clauses that construe modality 100

Table 3-55: Some modalised versions of statements in Table 3-54 101

Table 3-56: Tense in the text 101

Table 3-57: Clauses with present, past, future tenses (examples) 102

Table 3-58: Clauses with future tense at the start of the text 103

Table 3-59: Polarity in the text 104

Table 3-60: Clauses with positive/negative polarity (examples) 104

Table 3-61: Clauses with double embedding in the text 105

Table 3-62: Alternate version of clause 17, with no embedding 105

Trang 9

Table 3-63: Grammatical metaphor examples demonstrating all the

grammatical functions 107

Table 3-64: Types of grammatical metaphors in the text 108

Table 3-65: Metaphorical and congruent use of the word “crisis” 109

Table 3-66: Statistics of Experiential grammatical metaphors in the text 110

Table 3-67: Experiential grammatical metaphors shifting Process to Entity 111

Table 3-68: Examples of clauses with experiential grammatical metaphors shifting Process to Entity 111

Table 3-69: Experiential grammatical metaphors shifting Entity to Modifier 112

Table 3-70: Examples of clauses with experiential grammatical metaphors shifting Entity to Modifier 112

Table 3-71: Experiential grammatical metaphors shifting Process to Quality 113

Table 3-72: Examples of clauses with experiential grammatical metaphors shifting Process to Quality 113

Table 3-73: Experiential grammatical metaphors with shifts from Circumstance 114

Table 3-74: Clause showing shifts from Circumstance 114

Table 3-75: Experiential grammatical metaphors with shifts from Quality 114

Table 3-76: Clause showing shift from Quality 114

Table 3-77: Metaphors representing the least frequent experiential shifts 115

Table 3-78: Clause showing infrequent type experiential shifts 115

Table 3-79: Clauses with logical grammatical metaphors (underlined) 116

Table 3-80: Clauses with interpersonal grammatical metaphors (underlined) 117

Table 3-81: Clauses with textual grammatical metaphors (underlined) 118

Table 3-82: Clauses with the most grammatical metaphors 119

Table 3-83: Unpacking grammatical metaphors in clause 58 120

Table 3-84: Experiential grammatical metaphors shifting Entity to Modifier, at the start of the discourse 124

Table 3-85: Grammatical metaphor ‘void’ in red 126

Table 3-86: Grammatical metaphors in the second major logico complex leading to the ‘void’ 127

Table 3-87: Occurrences of “salad” in the grammatical metaphor ‘void’ 127

Table 3-88: Grammatical metaphors versus rankshifted and interrupting clauses 129

Table 4-1: All possible ergativity tags in a clause 131

Table 4-2: XOR Definition 135

Table 4-3: AND Definition 135

Table 4-4: OR Definition 136

Trang 10

Table 4-5: Decomposition of a tag representing intensive relational attributive

process 138

Table 4-6: Tags (35) used in the analysis of Figure 4-10 140

Table 4-7: Tags (21) used in the analysis of Figure 4-14 and Figure 4-22 145

Table 4-8: Tags (61) used in the analysis of Figure 4-24 and Figure 4-27 154

Table 4-9: < First Level > tags (9) analysed 162

Table 4-10: < Second Level > tags (24) analysed 165

Table 4-11: < Third Level > tags (32) analysed 168

Table 4-12: Combined metafunctions - tags (118) analysed 172

Table 5-1: Overview of transitive processes in the main story 181

Table 5-2: Frequency distribution of all T1 process types in reduced matrix 182

Table 5-3: Frequency distribution of all T1 tags in reduced matrix -77 tags 182

Table 5-4: Clauses in the three tightest clusters of clause neighbours in T1 188

Table 5-5: Stretches of text that continuously exhibit similar F1 values 194

Table 5-6: All Mann/Qual occurrences in the text 196

Table 5-7: Clauses of the three deep red regions of F1 in Figure A-7 199

Table 5-8: Clauses of the three deep blue regions of F1 in Figure A-7 202

Table 5-9: Clauses of the faint blue cluster in Figure A-7 202

Table 5-10: Clauses of the two faint red clusters in Figure A-7 204

Table 5-11: The outliers: very faint blue clauses in Figure A-7 205

Table 5-12: Clauses with metaphorical material processes in the text (T1m/Proc/Mat) 212

Table 5-13: Frequencies of tags not related to mental processes but significant in F3 (in bold), with other associated tags not significant in F3 (not in bold) 213

Table 5-14: All Attribute/Circumstance in the text (bold) 213

Table 5-15: All Location/Time in the text (bold) 213

Table 5-16: All circumstances of transitivity, T1 or T1m – 25 tags 214

Trang 11

LIST OF FIGURES

Figure 1-1: Meaning in text as arising from a web of functional relations

(examples of each relation in italics) 3

Figure 2-1: Website layout of CNN article 5

Figure 2-2: Main story with ten comments 9

Figure 2-3: Clausal analysis with the Systemics software (Clause Page) 12

Figure 2-4: Grammatical metaphor analysis in Systemics 14

Figure 2-5: Analysis of metaphors of transitivity in Systemics 15

Figure 2-6: System network for Theme analysis (TH), with selections for clause 48 highlighted (Grammar Page) 16

Figure 3-1: Overview of logical complexes 20

Figure 3-2: Logical complexes of the discourse (concise) 24

Figure 3-3: Logical complexes in detail – part 1 25

Figure 3-5: Elaborations in the second major complex (Figure 3-4), highlighted 28

Figure 3-8: String 1 and String 2 – lexical strings of “crisis” and “credit”, respectively 35

Figure 3-9: The salad analogy (Theme underlined) 41

Figure 3-10: Thematic progression possibilities for k = 2 45

Figure 3-11: Three major patterns of Thematic Progression (Daneš 1974): 48

Figure 3-12: Identification in clause 34 63

Figure 3-13: Logical structure leading to clause 23 – a “downward” progression 76

Figure 3-14: Lexical string of ups and downs – all words associated with

rising/falling in the text 77

Figure 3-15: All Mediums in the text, and their clausal context 82

Figure 3-16: Reference chains 1, 2, 5, 6 corresponding to all participants in the

discourse belonging to categories 1, 2, 5, 6 of Table 3-36 86

Figure 3-17: Items in Reference chains 1, 2, 5, 6 highlighted in the text 87

Figure 3-18: First major logico complex in the text (underlined in red:

non-declarative mood) 94

Figure 3-19: Second major logico complex in the text (underlined in red:

Trang 12

Figure 3-20: Third major logico complex in the text (underlined in red:

Figure 3-21: Mood Adjunct analysis of clause 55 98

Figure 3-22: Mood Adjunct analysis of clause 50 99

Figure 3-23: Piano roll for Tense 102

Figure 3-24: Piano roll for Tense (enlarged) 103

Figure 3-25: Grammatical metaphor analysis of clause 58 120

Figure 3-26: Opposition between clause 55 and clause 58 121

Figure 3-27: Colour code for Figure 3-28 122

Figure 3-28: Grammatical metaphor density against logical structure 123

Figure 3-29: Grammatical metaphor ‘void’ in second major logico-complex 125

Figure 4-1: A vector, in row and column form 130

Figure 4-2: Mapping a clause into tags (clause 7) 130

Figure 4-3: A vector representing the ergativity analysis of clause 7 131

Figure 4-4: Ergativity analysis of clause 8 134

Figure 4-5: A vector representing the ergativity analysis of clause 8 134

Figure 4-6: Difference in ergativity meaning between clause 7 and 8 134

Figure 4-7: Similarity in ergativity meaning between clause 7 and 8 136

Figure 4-8: Sum of ergativity meaning between clause 7 and 8 136

Figure 4-9: Transitivity analysis of clause 7 137

Figure 4-10: Differences in Mood between all clauses (M1 tags –XOR) 141

Figure 4-11: Scale numbering in Figure 4-10 142

Figure 4-12: Mood analysis of clause 55 143

Figure 4-13: Logical complexing context around clause 55 144

Figure 4-14: Differences in Theme between all clauses (TH1 tags –XOR) 146

Figure 4-15: Patterns in thematic differences (from Figure 4-14) 147

Figure 4-16: “Red Block 1”: clause 28-35 – consistent TH1 choices 147

Figure 4-19: “Blue Strip 1”: clause 24 – unique TH1 choices 150

Figure 4-20: “Blue Strip 2”: clause 42, 43 – unique TH1 choices 150

Figure 4-21: “Blue Strip 3”: clause 71 – unique TH1 choices 150

Figure 4-22: Similarities in Theme between all clauses (TH1 tags –AND) 152

Trang 13

Figure 4-23: “Blue Square” clauses in Figure 4-22 153

Figure 4-24: Differences in Transitivity between all clauses (T1 tags –XOR) 155

Figure 4-25: Transitivity analysis of clause 50 155

Figure 4-26: Lexical string of words amplifying scale 157

Figure 4-27: Sum of similarities and differences in Transitivity between all

clauses (T1 tags –OR) 158

Figure 4-28: Two Phases in Transitivity (T1 tags –OR) 159

Figure 4-29: < First Level > Speech Function, Mood, Mood Metaphor tags –XOR plot 163

Figure 4-30: Analysis of clauses with greatest changes in First Level 164

Figure 4-31: < Second Level > Int1 tags –XOR plot 166

Figure 4-32: Analysis of clauses in “Red Square” 166

Figure 4-33: Logico-complexing context of “Red Square” 167

Figure 4-34: < Third Level > Int1 and TH1 tags –XOR plot 169

Figure 4-35: Change in meaning at clause 42, 43 compared across three levels of delicacy 170

Figure 4-36: Clause 42 & 43 TH Analysis 171

Figure 4-37: Combined Metafunctions XOR plot 173

Figure 4-38: Patterns in Combined Metafunctions XOR plot 174

Figure 4-39: Clauses in the “Light Band” 174

Figure 4-40: Clauses in the “Dark Band” 176

Figure 5-1: Visualization of SVD in two dimensions 179

Figure 5-2: Tightest clusters in Figure A-3 187

Figure 5-3: Meaning of F0 and F1 in SVD 190

Figure 5-4 a, b: Contribution of presence and absence of tags to F1 192

Figure 5-5: Stretches of constant F1 values (in the clause list of Figure A-4) 194

Figure 5-6: Deep red regions of F1 in Figure A-7 198

Figure 5-7: Transitivity of some clauses in red region (2) in Figure A-7 200

Figure 5-8: Transitivity of a pair of clauses in red region (3) in Figure A-7 200

Figure 5-9: Deep blue regions of F1 in Figure A-7 201

Figure 5-10: Faint regions of F1 in Figure A-7 203

Figure 5-11: Clauses with metaphors of transitivity in the “tail” of red

cluster (2) in Figure 5-6 206

Figure 5-12: Doublets and triplets of blue relational identifying clauses

in Figure A-9 209

Trang 14

Figure 5-13: Tints for Reln/Attr tags in tag list in Figure A-12 212

Figure 5-14: Examples of mixed red & blue tinted clauses in Figure A-13 212

Figure 5-15: gmakens using Smick’s style 220

Figures in Volume II Figure A-1: Main story T1 F0 tag wheel 391

Figure A-2: Main story T1 F0 tag neighbours 392

Figure A-3: Main story T1 F0 clause neighbours 393

Figure A-4: Main story T1 F1 tag wheel 394

Figure A-5: Main story T1 F1 text 395

Figure A-14: Main story T1 F3 tag neighbourhood 404

Figure A-15: Main story T1 F3 clause neighbourhood 405

Figure A-20: Comments T1 F1 tag wheel 410

Trang 15

As physical beings in a physical world, we are surrounded by physical

phenomena on many scales in time and space: the ripples on a water surface, the circulation of air, and the movements of the sun and moon These physical events are not random but follow fixed laws In recent centuries, mankind has made tremendous progress towards understanding the laws of the physical world, and can now predict and control the evolution of physical systems with unprecedented precision

Our physical existence, however, is merely a platform for our social existence

As social beings in a social world, we are surrounded by social phenomena on many scales, and there exist deeper forces and principles behind social events, which we seek to comprehend But presently, we cannot predict social events with as much success as physical events; hence, we do not understand the social world as well as the physical world The analogy is not perfect, however; we are not merely

surrounded by social phenomena, but we actively engage in them, playing a part in the construction of their rules, and their evolution Language is the instrument with which we define our social identity and conduct our daily social interactions

Linguistic phenomena, be it a spoken/written text or a multimodal website/video presentation, are a reflection of social-cultural events and trends This thesis explores the linguistic dynamics of one textual discourse, affording us a glimpse of the grand scale of the universe of social-cultural dynamics

Trang 16

Meaning is described under the metafunctional principle proposed by Halliday (1978) which is described using system networks from Halliday & Matthiessen (2004) By

collecting and analysing large linguistic datasets reflecting real events in society with Systemic Functional Theory, and combining this with computerised mathematical analysis from the framework of Dynamical Systems Theory (Judd 2009a, b, c), it is hoped that we will be able to better describe and model these events This is the aim

of the Socio Cultural Modelling Project led by Kay O’Halloran and Kevin Judd at the Multimodal Analysis Lab, Interactive Digital Media Institute (IDMI), and this thesis represents the beginnings of the project

The study of meanings in language1 is a complex problem that challenges theoreticians on all fronts To begin with, we can think of meaning in text as arising from meaning in words This is what dictionaries attempt to do However, while this can be helpful in many situations, it fails to cover the meaning derived from the

1

The term “language” in this thesis refers to natural language (as opposed to “artificial language” such

as those created for programming machines), unless stated otherwise

Trang 17

patterns of words within sentences and patterns of sentences within a discourse

Meaning in text is not confined in words, nor is it confined in sentences; it is an

on-going dynamic process of orchestration of functional relations between words,

sentences, discourse and social context, see Figure 1-1:

Figure 1-1: Meaning in text as arising from a web of functional relations

(examples of each relation in italics)

There is often the philosophical question of whether linguistics is like a snake

trying to swallow itself: language was created to describe the world, so how can it be

used to describe meanings within itself (Firth 1948)? A prudent way to proceed might

be to rely heavily upon a kind of language very different from natural language –

mathematics – to describe natural language, thus invoking a new meaning potential

Meaning in natural language is fuzzy (Halliday 1995) and its grammar is flexible,

while mathematics is precise and its rules are strict Two chapters in this thesis,

Chapter 4: Recurrence Plots and Chapter 5: Singular Value Decomposition (SVD),

are devoted to the application of mathematical methods from linear algebra to analyse

the units of systemic functional meaning The next chapter, Chapter 2: Data and

Methodology, will explain how data is collected and analysed with software, and how

the analysis is stored in a format suitable for mathematical analysis

Trang 18

The subject for analysis is a news article2 on the Financial Crisis of 2008, from

an international news website, CNN.com, dated 10 October 2008 A screenshot of the article, in its original form and layout on the website, is shown in Figure 2-1 This text

is used to demonstrate how the results of systemic functional analysis can be

interpreted dynamically in terms of shifts in functionality, the relations and changes between clauses, logogenetic patterns, and the meanings that are generated as a result

The topic of financial crisis was chosen as it was a global phenomenon that impacted all levels of society - affecting many countries, people, and institutions, and

it was an event that extended significantly forwards and backwards in time The scale and complexity of the event would mean that any representation of it would stretch the potential of meaning-making resources in language This particular website was chosen because it allowed internet readers to write in their comments There were a total of 117 comments In addition to the main story, the ten latest comments are analysed, but the focus of the analysis is on the main story

2

The complete link for the article is:

http://edition.cnn.com/2008/POLITICS/10/09/smick.crisis/index.html

Trang 19

Figure 2-1: Website l

The figures which follow

from the website showing

comments

layout of CNN article

which follow are excerpts (preserving the original formatting)from the website showing the full text of the main story, followed by ten

s (preserving the original formatting)

ory, followed by ten reader

Trang 23

Figure 2-2: Main story with ten comments

This article appeared on 10 October 2008, at the very height of the Financial Crisis In the previous month, Lehman Brothers (4th largest investment bank in US) filed for bankruptcy, Merrill Lynch (3rd biggest investment bank in US) sold itself and

US insurer AIG came to the brink of collapse, triggering shockwaves worldwide These events in the US impacted on financial markets in Europe, Asia and all corners

of the globe, and plunged economies into recession In the week before 10 October, global markets lost U$6 trillion in panic selling, and governments injected U$6

trillion into the financial system in attempts to salvage the situation, with little effect These events were disseminated by media worldwide, and people everywhere feared for their jobs and financial investments This CNN story and commentary captures a minuscule snapshot of the spectacular scale of events and interactions that unfolded

Trang 24

Appendix 1.1 (p 227) displays the text of the CNN main story organised into

74 ranking clauses, numbered from “clause 7” to “clause 80” The ten comments have also been organised giving a total of 103 clauses, in Appendix 1.2 (p 230)

2.2 SYSTEMICS (THE SOFTWARE)

Analysis of the text in the CNN article is performed with advanced versions of Systemics, a software for systemic functional analysis developed by Kevin Judd and Kay O’Halloran (O'Halloran 2003) The interface of Systemics is organized into

“Pages” which fulfill different aspects of the analysis

• Text Page: allows decomposition of the text into ranking clauses and addition

of symbols to further organize the text in preparation for analysis

• Clause Page: zooms into each ranking clause, allows the user to annotate the

textual, interpersonal and ideational functionality of words, word groups, and

embedded clauses, as tags in a “Clause Table”, and further interpretation of

the analysis to be encoded in a “Analysis Table”

Trang 25

• Interclausal Page: allows clause complexes to be drawn as links between

clauses, revealing logical meaning

• Discourse Page: allows reference chains and lexical strings to be drawn

between the clauses, revealing cohesion

• Search Page: allows user to selectively retrieve any combination of types of

tags that have been input in the Clause Page, and display their distribution and statistics

• Grammar Page: contains a (expandable) comprehensive database of SF

Grammar, principally compiled from Halliday (1994), Halliday (1998b) and Martin (1992) This represents Halliday’s description of the meaning potential

of language (English written text) encoded in the form of system networks Appendix 7: The Grammar, p 413, shows all system networks used in this thesis, extracted from the Grammar Page Comprehensive descriptions of Halliday’s Systemic Functional systems are found elsewhere, e.g Halliday & Matthiessen (2004), thus are not reproduced here

For a complete description of all the Pages of Systemics, please refer to O'Halloran (2003) The most important Page in concern for this thesis is the Clause Page (shown

in Figure 2-3 next page) because it shows how tags are input, organized and named, and these tag names will be used for advanced mathematical analysis later in Chapter

4 and Chapter 5

Trang 26

Figure 2-3: Clausal analysis with the Systemics software (Clause Page)

Analysis done at the Clause Page is entered into two tables: a Clause Table, which represents direct labeling of clause constituents, and an Analysis Table, which

is based on the interpretation of the clause table analysis and the discourse context For the Clause Table, columns match words/parts of the clause, but for the Analysis

Trang 27

table, columns are arbitrarily arranged For both tables, the tags are organized

metafunctionally into Keys (rows):

The tags3 are colour-coded in the Clause Page according to their Key The Key

numberings (e.g “2” in “M2”) generally (but not always) describe the level of

embedding For example, in Figure 2-3, “M2”, “T2”, “E2” and “Int2”, “Exp2”

correspond to the analysis of the embedded clause “of forcing the banks to start lending” (for this thesis, Themes in embedded clauses are not analysed) The

exception to this Key numbering convention is “TH#” The Key “TH2” is not used to encode embedded clauses, but to encode the ranking clause analysis in the event that the ranking clause is part of a marked clause complex construction In these rare cases, “TH1” will encode the Theme/Rheme at the level of clause complex (see the analysis of clauses 79-80 in Appendix 2: SF Analysis of Clauses, p 329 for an

3

Tag names used in Systemics shall appear in a different font (Calibri) in this thesis In tag names the wildcard character * denotes any letter, number, or slash; and # denotes any number

Trang 28

example)

Grammatical metaphors (Halliday 1998b; Martin 1992) are analysed in

Systemics according to the conventions shown in Figure 2-4 below:

Figure 2-4: Grammatical metaphor analysis in Systemics

The Key “MET1” labels the entire functional element (mood/transitivity

element) in which the grammatical metaphor resides The grammatical metaphor itself

is labeled at the “MET1m” Key For example, “bank” in “a private/global bank

clearing facility” (clause 76) is a grammatical metaphor with a mainly experiential function (hence “Exp”) and congruently represents an Entity (“Ent”) but has been shifted to operate as a Modifier (“Mod”) of the Entity “clearing facility” For an introduction of the grammatical functions (Modifier, Entity, etc) describing the

grammatical metaphors, see section 3.6: Grammatical Metaphors, p 107

In some clauses, there are “T#m” tags together with “T#” These are cases of

clauses with ideational metaphors of transitivity (Halliday 1994: 344) where there is

a discrepancy between realized and interpreted transitivity process type The clause is double coded: “T#m” encodes the metaphorically realized meaning, and “T#” encodes the congruent interpreted meaning For an example, see Figure 2-5 next page The

“m” suffix is also used in the “E#” and “Exp#” Keys

Trang 29

Figure 2-5: Analysis of metaphors of transitivity in Systemics

Having introduced all the Keys in the tables, we now explain the theoretical significance of the tags The tags in the Clause Table and Analysis Table represent, for the instance of the particular clause/word group/word, the choices made by the writer according to system networks stored in the Grammar Page In Systemic

Functional Theory, the choices represent meaning created in the text, and the system network represents the meaning potential For example, consider the “TH1” analysis

of clause 48 in Figure 2-3 (p 12), which has two tags:

Topic/Theme

Rheme

The slash “ / ” in the tag name “Topic/Theme” is denoted in Figure 2-3 (p 12) as new line spacing between “Topic” and “Theme”, and represents hierarchical ordering of the tags, i.e sub-branching of nodes in a system network The two tags

“Topic/Theme” and “Rheme” represent end nodes from the “TH” system network, shown in Figure 2-6 next page, which is an excerpt from the Grammar Page

In Figure 2-6, the main branches of “TH” (in blue) are: “Text” (Textual), “Int” (Interpersonal), “Topic”, and “Rheme” All these branches, except “Rheme”,

eventually end with a “Theme” end node Note that tag hierarchy usually but not

Trang 30

Appendix 7, p 413) and hence, are:

Trang 31

The Clause and Analysis Tables in the Clause Page can be printed in

Systemics The complete set of printouts representing the comprehensive SF analysis

of the main story and the comments are compiled in Appendix 2: SF Analysis of Clauses, p 234 The SF analysis of the main story is also presented in the form of matrices in Appendix 3: SF Analysis Matrices, p 368 Printouts from other Pages of Systemics (Interclausal, Discourse) are also used in this thesis

The main story analysis is comprehensively explained and interpreted in the next chapter

Trang 32

3.1 LOGICAL COMPLEXES

Clause complexing and cohesion between clauses in a text can be described in terms of the type of logico-semantic relation (Expansion/Projection) The basic

systems described by Halliday (1994: 216-221) are reproduced in Appendix 7.2 p

418 To extend Halliday’s terminology, a group of clauses linked by logico-semantic

relations (whether structural or non-structural) shall be called a logico-semantic

complex, or logical complex/logico-complex4 It is useful to begin the SF analysis of the CNN article by examining the logico-complexes in the article, for this generates a large-scale map of meaning in the entire discourse in the logical metafunction, within which we can contextualize further analyses in the textual, interpersonal and

experiential metafunctions, which create more localized meanings (at the level of the clause) The logical complexing encodings were performed in Systemics, in the Interclausal Page, with the following conventions:

4

Not to be confused with the term “clause complex”, which only refers to structurally linked clauses

Trang 33

• Taxis (structural links)

• Cohesive (non-structural links)

Hypotactic links have arrows which point towards the dependent clause

Text Colour Key

• Conjunctive Adjunct (Theme)

• Structural Conjunction (Theme)

• Interrupting Clause or Interrupting Clause Complex

The clause complexing between an interrupting clause and its ranking clause, and between clauses within a rankshifted/ interrupting clause complex, are also analysed, and such internal clauses are numbered in oval rather than rectangular boxes

Trang 34

• 8 Interrupting/Included clauses

The overall logical organization of the ranking clauses is shown in Figure 3-1:

Figure 3-1: Overview of logical complexes

Over the following pages, Figure 3-2 displays the logical analysis of the

ranking, interrupting and embedded clauses in concise format: for Expansion, only the basic divisions into elaborating, extending and enhancing relations are made Note that clause 2 and clause 3 are not ranking clauses, but represent a rankshifted

complex, forming the title of the article

Trang 38

Figure 3-2: Logical complexes of the discourse (concise)

The concise complex analysis provides the basic large-scale logical map of the discourse To provide a finer and more grounded sense of the logical development of the text, a more delicate analysis of all the complexes is presented in sequence in the next section 3.1.2

3.1.2 Detailed Complexes

In the figures of logico-complexes that follow: Figure 3-3 to Figure 3-7, the level of detail of analysis is beyond Figure 3-2, which only subdivided the logico-semantic relation into elaborating, extending, enhancing types For Figure 3-3 to Figure 3-7, the complete grammar for analysing Expansion is used, as explained in Halliday (1994: 225-241, 328-329), and compiled in Appendix 7.2 p 419

Trang 39

Figure 3-3: Logical complexes in detail – part 1 5

The article begins with a small logico-complex of clause 7-10 (see Figure 3-3

above) where the author humorously points out that amidst the chaos and confusion of

the crisis, there is something which is clear: that our leaders don’t how to lead us

through this troubled times: “…driving policy looking through the rearview mirror”

(clause 9, 10) By ridiculing the policymakers, the author appears authoritative and to

have complete knowledge of the situation

After clause 10 in Figure 3-3 is the first major logico complex of the article,

5

In Figure 3-3 and similar figures that follow, the link label “positive” is ambiguous:

1) */enhancement/condition/positive (for clause 13, 18 & << >>)

2) */extension/additive/positive (for all other instances)

Trang 40

spanning clause 11-23, describing the failure of the bailout package and the origins of the crisis Clause 11-14 describe the promises of the bailout package; here again, the author injects a fair amount of derision by exaggerating the leaders’ expectations of the bailout: “…some magic pill, which, if gulped down…” (clause 12, 13) An

extending replacive relation, marked by “instead”, follows, with clause 15-16

describing the failure of the bailout: “Instead, stock markets collapsed and credit markets remained frozen” The remainder of the first major complex, clauses 17-23, forms an enhancing causal relation with clause 15-16, providing the historical

background leading to the financial crisis The cause of the crisis, according to David Smick, seems to center upon securitization by the banks, or as he puts it, the creation

of “paper instruments” This entire logico complex proceeds logically, describing events in causal and temporal sequence, and ends dramatically with the statement that

we are experiencing a “downward spiral”

The second major logico complex, shown in Figure 3-4 next page, elaborates

on the cause of the crisis, by explaining what exactly the paper instruments were, how they were created, and their effects on the financial system The complex begins with

a rhetorical question “So what are these paper instruments…?” (clause 24), the

answer to which constitutes the rest of the complex The first relation in the complex

is elaborating clarification, with clause 25: “I like to use a salad analogy” which construes the paper instruments as salad The author seems to have a habit of using lexical metaphors to deride that what he aims to debunk (bailout package = “magic pill”; asset-backed securities = “paper instrument” = “salad”) Then he explains what was the standard form of loans in the past (“syndicated loans”) in clause 26, and uses

re-an adversative “but” (clause 27) to begin the account of the creation of paper

Định dạng
Số trang	240
Dung lượng	13,46 MB