Top Banner
Aspects of a Systemic-Functional Grammar of Finnish Susanna Shore Thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy in the School of English and Linguistics, Macquarie University, Sydney, N.S.W. July 1992
446

systematic functional linguistic

Dec 29, 2015

Download

Documents

Amee Hanifah

functional grammar
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: systematic functional linguistic

Aspects of a Systemic-Functional Grammar of Finnish

Susanna Shore

Thesis submitted in fulfilment of the requirements for the

degree of Doctor of Philosophy in the School of English

and Linguistics, Macquarie University, Sydney, N.S.W.

July 1992

Page 2: systematic functional linguistic

Contents

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

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

1.1. Purpose and Scope of this Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2. Outline of the Chapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.3. Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1.4. Glosses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

1.5. Theory and Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Chapter 2: An Outline of Systemic-Functional Theory . . . . . . . . . . . . . . . . . . . . . 13

2.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.2. The London School of Linguistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.2.1. General Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.2.2. System and Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

2.2.3. Restricted Languages and Speech Fellowships . . . . . . . . . . . . . . . . . 15

2.2.4. A General Linguistic Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

2.2.5. Meaning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

2.2.6. Rejection of Saussurean Structuralism . . . . . . . . . . . . . . . . . . . . . . . . 20

2.2.7. Firth and SF Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

2.3. SF Theory: General Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

2.3.1. Language as a Linguistic Behaviour Potential . . . . . . . . . . . . . . . . . . 22

2.3.2. The SF Interpretation of Langue . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2.3.3. The Individual and the Social Aspect of Language . . . . . . . . . . . . . . 25

2.3.4. Knowledge of Language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

2.3.5. A Reality Construction View of Language . . . . . . . . . . . . . . . . . . . . 30

2.3.6. Language and Mind . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

2.3.7. Grammar, Semantics and the Context of Situation . . . . . . . . . . . . . . 34

2.3.8. The Notion of Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

2.4. Central Notions in SF Grammar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

2.4.1. Introductory Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

2.4.2. A Brief Historical Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

2.4.3. The Triplanar Organization of Language . . . . . . . . . . . . . . . . . . . . . . 44

2.4.4. Units and the Rank Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

2.4.5. Types of Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

2.4.6. System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

2.4.7. Delicacy and Realization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

2.4.8. Metafunctions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

2.4.9. An Integrated Lexicogrammar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

2.4.10. The Principle of Grammaticalization . . . . . . . . . . . . . . . . . . . . . . . . 63

2.4.11. The Ineffability of Grammatical Categories . . . . . . . . . . . . . . . . . . . 63

Page 3: systematic functional linguistic

ii

2.4.12. Prototypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

2.4.13. Grammatical Proportionalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

2.4.14. Synoptic vs. Dynamic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Chapter 3: The Finnish Language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

3.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

3.2. Background Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

3.3. Some General Characteristics of Finnish . . . . . . . . . . . . . . . . . . . . . . . . . 74

3.3.1. General Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

3.3.2. Verb Inflexions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

3.3.3. Finnish Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

3.4. Issues in the Received Description of Finnish . . . . . . . . . . . . . . . . . . . . . . 87

3.4.1. The Problem of the “Accusative” . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

3.4.2. Boundedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

3.4.3. Traditionally Defined Grammatical Subject in Finnish . . . . . . . . . . . 99

Chapter 4: Constituency and Dependency in Finnish . . . . . . . . . . . . . . . . . . . . . . 105

4.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

4.2. Ranked Constituency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

4.3. Some Problems with the Rank Hypothesis . . . . . . . . . . . . . . . . . . . . . . . . 110

4.3.1. Discontinuous Constituents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

4.3.2. Inclusion of Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

4.4. Phrases in Finnish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

4.4.1. Nominal Phrases (NP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

4.4.2. Pre- and Postpositional Phrases (PP) . . . . . . . . . . . . . . . . . . . . . . . . . 123

4.4.3. Verb Phrases (VP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

4.4.4. Adverbial Phrases (AdvP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

4.4.5. A Further Note on Unmodified P-Positions . . . . . . . . . . . . . . . . . . . . 134

4.5. Preliminary Analysis of Clause Complexes . . . . . . . . . . . . . . . . . . . . . . . . 137

4.5.1. General Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

4.5.2. Interdependency: Parataxis and Hypotaxis . . . . . . . . . . . . . . . . . . . . 139

4.5.3. The Function of a Clause Complex . . . . . . . . . . . . . . . . . . . . . . . . . . 146

4.5.4. Relativization and Embedding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

4.5.5. Projection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

Chapter 5: Interactional Structure in the Finnish Clause . . . . . . . . . . . . . . . . . . 167

5.1. Preliminary Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

5.2. From Rhetorical Functions to Grammatical Categories . . . . . . . . . . . . . . 168

5.2.1. The Clause as an Exchange or Interactive Event . . . . . . . . . . . . . . . . 172

5.2.2. Congruent and Metaphorical Realization . . . . . . . . . . . . . . . . . . . . . . 177

5.2.3. Problems with Halliday’s Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 183

5.2.4. Alternative Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188

5.3. Interactional Options in Finnish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

Page 4: systematic functional linguistic

iii

5.4. Interactional Functions in Finnish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

5.4.1. Finite, Mood Marker, and Residue . . . . . . . . . . . . . . . . . . . . . . . . . . 197

5.4.2. The Grammatical Subject . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201

Chapter 6: Experiential Structures in the Finnish Clause . . . . . . . . . . . . . . . . . . 205

6.1. Preliminary Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

6.1.1. General Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

6.1.2. Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210

6.1.3. Inherent and Non-Inherent Roles . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

6.2. Process Types in Finnish: An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . 213

6.3. Relational Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

6.3.1. Intensive Relational Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

6.3.1.(i) Attributive Intensive Processes . . . . . . . . . . . . . . . . . . . . . . . . . 215

6.3.1.(ii) Identifying Intensive Processes . . . . . . . . . . . . . . . . . . . . . . . . . 219

6.3.1.(iii) Other Intensive Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231

6.3.2. Inclusive Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237

6.3.3. Ambient Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240

6.3.4. Circumstantial (Relational) Processes . . . . . . . . . . . . . . . . . . . . . . . . 244

6.3.4.(i) Introductory Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244

6.3.4.(ii) General Circumstantial Processes . . . . . . . . . . . . . . . . . . . . . . . 244

6.3.4.(iii) Possessive Circumstantial Processes . . . . . . . . . . . . . . . . . . . . 249

6.3.5. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256

6.4. Material Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256

6.4.1. Introductory Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256

6.4.2. General Features of Material Processes . . . . . . . . . . . . . . . . . . . . . . . 257

6.4.3. Subtypes of Material Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265

6.4.3.(i) Meteorological Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266

6.4.3.(ii) Experiencer Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267

6.4.3.(iii) Behavioural Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272

6.5. Mental Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274

6.5.1. Internal and External, Verbalized Consciousness . . . . . . . . . . . . . . . 274

6.5.2. Defined by Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276

6.5.4. Human Consciousness and Projection . . . . . . . . . . . . . . . . . . . . . . . . 278

6.6. Experiential Metaphors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284

6.7. Macro-Roles in Finnish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288

6.7.1. Medium and Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288

6.7.2. The Problem of the “Existential Subject” . . . . . . . . . . . . . . . . . . . . . . 292

6.7.3. A Note on Derivational Affixes and External Causation . . . . . . . . . . 301

6.8. Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306

Chapter 7: Textual Structures in the Finnish Clause . . . . . . . . . . . . . . . . . . . . . . 307

7.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307

7.2. Given and New . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308

Page 5: systematic functional linguistic

iv

7.2.1. Brown and Yule’s Approach to Given and New . . . . . . . . . . . . . . . . 308

7.2.2. Halliday’s Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310

7.2.3. Segmental Markers of Information Structure in Finnish . . . . . . . . . . 315

7.3. Theme and Rheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319

7.3.1. Introductory Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319

7.3.2. Topical Theme (Topic) in Finnish . . . . . . . . . . . . . . . . . . . . . . . . . . . 320

7.3.3. Subsidiary Topical Themes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327

7.3.4 Topical Themes in Post-Verbal Position . . . . . . . . . . . . . . . . . . . . . . 334

7.3.5 Theme and Topic-Worthiness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336

7.3.6. Clause-Initial Interpersonal and Textual Elements . . . . . . . . . . . . . . 339

7.4. Theme in Clause Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341

7.4.1. Relative Clauses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341

7.4.2. Clause as Theme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343

7.4.3. Theme-Rheme in Hypotactic Complexes . . . . . . . . . . . . . . . . . . . . . . 344

7.5. Additional Topics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345

7.5.1. Non-realization of Inherent Human Participant . . . . . . . . . . . . . . . . . 346

7.5.2. Non-Realization of Subject . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348

7.5.3. Habitive and Manifestation Clauses . . . . . . . . . . . . . . . . . . . . . . . . . . 360

7.6. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365

Chapter 8: Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367

8.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367

8.2. The Grammatical Analysis of a Text . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368

8.3. Alongside and Beyond this Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372

8.4. Recurring Themes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376

Appendix 1: Analysis of the “Cat Text” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381

Appendix 2: Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405

Appendix 3: Form Glosses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409

Appendix 4: Function Glosses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410

Appendix 5: Notational Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411

Appendix 6: System Network Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

Maps and Figures

Map 1: Finnish Dialect Groupings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

Figure 2-1: Planes in Language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

Figure 2-2: Levels in Language (Halliday 1961) . . . . . . . . . . . . . . . . . . . . . . . . . . 46

Figure 2-3: Simultaneous Paradigmatic Options in English . . . . . . . . . . . . . . . . . 52

Figure 2-4: Metafunctions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Page 6: systematic functional linguistic

ix

Figure 2-5: Transitivity Options in English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Figure 2-6: Action Options in English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

Figure 3-1: Consonant Gradation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

Figure 3-2: Vowel Harmony . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

Figure 3-3: Present Indicative Forms of asua ‘to live/dwell’) . . . . . . . . . . . . . . . 78

Figure 3-4: Imperative Forms of ottaa ‘(to) take’ . . . . . . . . . . . . . . . . . . . . . . . . . 80

Figure 3-5: Negative Forms of ottaa ‘(to) take’ . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Figure 3-6: Common Case-Forms for Nominals . . . . . . . . . . . . . . . . . . . . . . . . . . 83

Figure 3-7: Convenient Groupings of Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

Figure 3-8: Non-Productive or Semantically Restricted Case-Forms . . . . . . . . . . 85

Figure 3-9: Infinitives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

Figure 3-10: Common Case Forms for Non-Finite Verb Stems . . . . . . . . . . . . . . 86

Figure 4-1: The Rank Scale in Finnish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

Figure 4-2: Constituency and Dependency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

Figure 4-3: Tactic Relationships in a Clause Complex . . . . . . . . . . . . . . . . . . . . 139

Figure 5-1: Variables in an Interactive Event . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

Figure 5-2: Primary Speech Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

Figure 5-3: Mood Options in English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

Figure 5-4: Semantics as an Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

Figure 5-5: Congruent Realizations of Speech Functions . . . . . . . . . . . . . . . . . . . 177

Figure 5-6: Halliday’s Symmetrical Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188

Figure 5-7: Alternative Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

Figure 5-8: Mood Options in Finnish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

Figure 6-1: Major Process Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213

Page 7: systematic functional linguistic

x

Acknowledgements

This text ) like any other text ) is pervaded by the intertextual sources that

engendered it. These intertextual sources are a reflection of my ties with both

Australia and Finland. I am deeply indebted to all of my teachers, colleagues and

friends ) at Macquarie University, the University of Sydney, the University of

Helsinki, and elsewhere ) for the critical and constructive dialogue without which

this study would not have materialized.

I am especially grateful to Professor Ruqaiya Hasan, my supervisor at

Macquarie University, for her advice and for her critical comments. Her question-

ing of me at various stages of writing this study has been an important catalyst in

the development of my thinking. I should also like to express my thanks to

Professor Pentti Leino of the Department of Finnish at the University of Helsinki

for reading and commenting on an earlier version of this thesis and for giving me

the opportunity to work and to do research in an academic environment. I am

indebted to Dr. Maria Vilkuna for her lengthy and thoughtful comments and

criticisms on an earlier draft of this thesis. Marja-Liisa Helasvuo, Ilona Herlin,

Associate Professor Esa Itkonen, Anne Thwaite, and Dr. Eija Ventola have also

read earlier versions of this thesis or parts of it. I am deeply grateful to all of them

for their comments and criticisms.

This thesis would not have been possible without financial support from the

Kone Foundation of Finland and the Australian Government Postgraduate Awards

Scheme.

Last, but not least, I wish to thank my family for their encouragement and

support over the years. I dedicate this study to my mother, Saara Rantamäki, who,

through her example, taught me the meaning of the Finnish expression kestää kuin

nainen ‘endure ) like a woman’.

Page 8: systematic functional linguistic

xi

Abstract

This study is a functional analysis of Finnish based on systemic-functional theory

as developed by Michael Halliday and other linguists. The focus is on the description of

clause types in Finnish in terms of a number of simultaneous grammatical structures,

which are fused together in the process of realization. While the description of Finnish is

not based on any particular text or set of texts, the majority of examples are attested

examples from written Finnish or Finnish as it is spoken in the Helsinki area.

There are three introductory chapters to the study. The first chapter serves as an

introduction proper to the thesis and discusses some general issues concerned with theory

and data and the status of intuition in linguistic description. The second chapter is an

introduction to systemic-functional theory and it also discusses a number of issues (e.g.

knowledge of language) from a wider metatheoretical perspective. The third presents an

ethnographic account and a general outline of Finnish.

Chapters 4 to 7 constitute the main body of this thesis. Chapter 4 defines grammati-

cal units in terms of ranked constituency and gives a preliminary outline of clause

complexing in Finnish. Chapter 5 is concerned with structures in the Finnish clause that

reflect its internal organization as an exchange. A slightly modified version of Halliday’s

model is proposed: the parameters involved are whether the orientation of the exchange

is primarily linguistic (i.e. the exchange of words) or non-linguistic (i.e. directed at

action). The sixth chapter is concerned with the analysis of clauses into process types and

their concomitant participant and circumstantial roles. As in English, it is argued that in

Finnish there is also a basic division into relational processes, material processes and

processes of human consciousness (mental and verbal). The establishment of these process

types is based on Prototype Theory. Chapter 7 deals with textual structures. It presents a

preliminary discussion of Given and New and analyses the Theme-Rheme structure of the

Finnish clause. It is argued that the topical Theme in Finnish is realized by the experiential

element in the position preceding the finite verb, and that another subsidiary experiential

Theme needs to be recognized for Finnish. The Theme-Rheme structure is illustrated with

the analysis of a complete text reproduced in Appendix 1.

The study concludes by bringing together a number of recurring themes and by

presenting an analysis of a small fragment of text. The purpose of the text analysis is to

show how the various structures described in the main body of the study are intertwined

and conflated in a text and how the analysis presented in this study fits into a wider and

more comprehensive framework.

Page 9: systematic functional linguistic

Chapter 1

Introduction

1.1. Purpose and Scope of this Study

This study is a dialogue with linguists from two different traditions in linguistics.

It is a dialogue with Finnish linguists, particularly those working within the

framework of traditional linguistics and in applied areas of linguistics, and it is a

dialogue with systemic-functional linguists, since the theory of language on which

it is based is systemic-functional theory.

Although based on systemic-functional theory, the study also relies on

insights that have been gained from other linguistic theories and from traditional

grammars of Finnish. Traditional linguistics in Finland (generally referred to as

Fennistics) has tended to be theoretically eclectic, which ) from a positive

perspective ) can be seen as not being bound to any particular theory. From a

negative perspective, however, Fennistics could be seen as a mixture of various

theories, in which many of the critical and basic assumptions are not explicated.

This study offers Finnish linguists a way into systemic-functional theory, a way

into a theory that is concerned with explicating basic assumptions about language

and the study of language and with developing a coherent and comprehensive

theory of language.

It also offers Finnish linguists a new perspective on the analysis of Finnish as

it presents a functional description of Finnish grammar. The essence of a functional

grammar ) at least a systemic-functional grammar ) is that it relates the linguistic

system to texts, either spoken or written texts, and these texts, in turn, can be

related to human contexts of living. Moreover, in a systemic-functional perspec-

tive, the relation between language and human contexts of living is seen in both

dialectic and symbiotic terms. This has far-reaching repercussions for the way in

which the role of language in human activity is conceived ) for example, its role

in socialization, in education, in issues of ideology, race, gender (see, for example,

Page 10: systematic functional linguistic

2

1 The last time I heard words to this effect the speaker was, in fact, a Finnish linguist, and hewas not being facetious.

2 This is not meant to imply that grammatical and phonological phenomena are identical, butI think it is useful ) particularly for those who are unacquainted with SF theory ) to drawsome sort of a parallel.

Halliday 1991, forthcoming). While this study focuses on grammar, it is a grammar

that will hopefully be of particular relevance to the study of spoken and written

texts and their function in Finnish society.

What is being offered to systemic-functional linguists in this study is an

insight into the grammatical description of a non-Indo-European language and the

problems that it poses. With a few exceptions, systemic-functional linguists ) like

many other linguists ) have tended to concentrate on the description of English.

There is often an implicit assumption in theoretical linguistics that a theory of

language need only be tested with the description of English: all too often one

hears or reads the words “Take a language, for example, English”.1 This study in

itself is “an act of meaning” (Halliday forthcoming), an act of meaning which may

change the meaning of saying something like “Take a language, for example,

Finnish” or “Take a language, for example, Gooniyandi”.

The purpose of the study is to provide a broad outline of a particular type of

functional grammar, systemic-functional grammar, as applied to Finnish. The focus

is on the description of clause types in Finnish in terms of a number of simul-

taneous grammatical structures. What this means is that the representation of a

clause is not seen as a single structural pattern, but as a number of patterns, which

are “conflated”, i.e. mapped onto each other or fused together, in the process of

realization. This approach to grammatical structure can be seen as a development

and an extension to grammatical phenomena of Firth’s analysis of phonological

structure (see e.g. 1957, Ch.9).2 A similar approach is taken in autosegmental

phonology. According to Halle and Vergnaud (1982: 65), for example:

Page 11: systematic functional linguistic

3

The major insight lying at the base of autosegmental phonology is that the phonologicalrepresentation is composed not of a single sequence of entities roughly corresponding to a lineof type, but rather that the phonological representation is made up of several parallel sequencesof entities, resembling thus more a score for a musical ensemble, than a single line of type.

The musical analogy that is used in this quotation echoes a similar analogy made

by Halliday, who compares grammatical structure to polyphonic music (e.g. 1978:

56). As is obvious from the above, the term grammatical structure as used here

does not only refer to constituency structure but to any kind of (non-random)

organization that is (more or less) restricted in scope to the boundaries of a clause.

The grammatical scope of this study is quite broad. In view of this, the degree

of detail (or “delicacy” in systemic-functional terms) is restricted: only the most

general type of organization is presented. A number of other restrictions on the

scope of the study are presented in the next section, in which there is an outline of

each chapter. One important restriction needs to be mentioned immediately: the

description presented here is not concerned with the “systemic” side of systemic-

functional theory, i.e. the setting up of systems networks to model the paradigmatic

options available in a language. The label “systemic-functional” (henceforth: SF)

is, nevertheless, used throughout to distinguish the approach taken here from other

brands of functionalism (see, e.g., Dirven & Fried (eds.) 1987, Matthiessen &

Halliday (forthcoming)). The systemic side is, nevertheless, implicit in the analysis

that is presented: in SF theory, the grammatical organization of a language is

modelled as a number of parallel complexes of networks, which represent various

kinds of meaningful (paradigmatic) options available in the language being inves-

tigated. This model of grammar iconically reflects the notion that the grammatical

system of a language is a “potential” for expressing and making meanings. A

particular combination of meanings ) at the grammatical level ) will result in a

number of simultaneous grammatical structures.

Page 12: systematic functional linguistic

4

1.2. Outline of the Chapters

This chapter serves as an introduction and discusses some general issues concerned

with theory and data. The next two chapters give background information: as SF

theory is relatively unknown in Finland and as Finnish is relatively unknown to the

majority of linguists outside Finland, I have tried to accommodate possible readers

of this study by giving some background to SF theory in Chapter 2 and by giving

an ethnographic account and a general outline of Finnish in Chapter 3.

Chapters 4 to 7 constitute the main body of this study. Chapter 4 is primarily

concerned with one particular type of grammatical structure in Finnish )

constituency structure. Constituency refers to the part-whole relationships which

extend from the smallest grammatical unit, the morpheme, to the largest unit, the

clause. The chapter will also contain some discussion of clause complexing, which

can be regarded as straddling the area between grammatical and textual organiza-

tion.

The fifth chapter is concerned with structures in the Finnish clause that reflect

its internal organization as an exchange or interactive event. This is reflected in

mood structures, which are regarded as being related to the way in which the role

of the interactants is construed in language. Another aspect of interactive or

interpersonal clause-internal organization in SF theory is concerned with what

Halliday refers to as modality (assessments of probability and usuality) and

modulation (assessments of obligation or inclination) (Halliday 1985a: 86, 334 ff.;

Halliday in Kress ed. 1976: 189-213); however, this area of Finnish grammar will

not be dealt with in this study.

Chapter 6 is concerned with the way in which the clause in Finnish provides

a model of reality, i.e., a linguistic representation of the world around us, of the

world inside us, and of the world of our imagination. Central to the structure of the

clause as a model of reality is the analysis of clauses into process types and their

concomitant participant and circumstantial roles. For example, Akira Kurosawa

Page 13: systematic functional linguistic

5

ohjasi tämän elokuvan ‘Akira Kurosawa directed this movie’ construes (encodes

and constructs) a material process, a happening or an action, whereas Akira

Kurosawa on japanilainen elokuvaohjaaja ‘Akira Kurosawa is a Japanese film

director’ construes a relational process in which the Carrier, realized by the proper

noun Akira Kurosawa is assigned an Attribute.

In the seventh chapter, the clause in Finnish is discussed in terms of the way

it is structured as a message, i.e., in terms of what Prague School linguists refer to

as the Functional Sentence Perspective (Halliday in Kress ed. 1976: 26-31; and e.g.

Daneš (ed.) 1974, Daneš 1987). From the point of view of the message, the

structure of the clause can be looked at in terms of its Theme-Rheme and Given-

New structure. According to Halliday (1985a: Chapter 8), the Given-New structure

in English is realized in intonational patterns (pitch movements) in the tone group.

While there is some preliminary discussion of Given and New in Finnish, a

phonetic analysis of the intonational patterns in Finnish is beyond the scope of this

study. The discussion of the Theme-Rheme structure of the Finnish clause focuses

on the topical Theme. The analysis is illustrated by a complete text, which is repro-

duced in Appendix 1.

The study concludes with the analysis of a small fragment of text in Chapter

8. This analysis is intended to show how the various structures described in the

main body of the study are intertwined and conflated in a text and how the analysis

presented here fits into a wider and more comprehensive framework. Chapter 8 also

brings together a number of recurring themes in the study.

Page 14: systematic functional linguistic

6

1.3. Data

This study is based primarily on the Finnish that is spoken in the Helsinki area and

on standardized written Finnish, as this is the Finnish that I am familiar with.

However, the study is at such a primary degree of delicacy that is seems to me that

the analysis ) for the most part ) is also valid for at least those varieties of Finnish

that are spoken in Finland (see 3.2).

While the analysis and description here is not based on any particular text or

set of texts, the majority of examples in the main body of the text (Chapters 4 to 7)

are attested examples from either spoken or written Finnish. Each attested example

is followed by a reference to its source in square brackets. The sources are listed

in Appendix 2. In some instances, I have “tidied up” the original examples, e.g.

omitted false starts and stutters, since these are irrelevant to the functionally charac-

terized grammatical structures that are being explicated. For the same reason, I

have also standardized certain phonological features of dialect forms. In some

cases, I have also omitted parts of the clause if they are irrelevant to the point being

made.

Since the majority of examples are authentic examples, instead of giving just

one or two examples to illustrate a grammatical phenomenon, a number of

examples are given. This is done to clearly illustrate the kind of phenomenon that

is being described. With constructed examples, it is relatively easy to construct an

example that clearly and unambiguously illustrates a grammatical point. An

intransitive material process, for example, could be illustrated by the example

Lapsi juoksee ‘The child runs/is running’. However, examples like this are rare in

actual text. An authentic example often contains constituents that are irrelevant to

the point being made and this “extraneous matter” cannot be omitted without

sacrificing the authenticity of the example.

Some of the examples are also constructed, i.e. based on my knowledge of

Finnish. This has been done for two reasons: 1) to illustrate a grammatical point by

Page 15: systematic functional linguistic

7

comparing an authentic example with a possible variant and 2) in cases where a

particular type of example ) although commonly used and uncontroversial ) is

unlikely to occur in the corpora available to me, or it seemed that to search for an

example like mua palelee ‘I’m freezing’ would be taking the use of authentic

examples to unnecessary extremes (see below). To avoid taxing the non-Finnish

reader, I have also used simple, constructed examples in Chapter 3, which presents

some of the main features of Finnish in a nutshell.

I have not based this grammatical description on a particular corpus, because

what I am investigating is grammatical phenomena, phenomena that are not

restricted to a particular genre, a particular text or discourse type. While SF theory

recognizes both a semantic and a (lexico)grammatical plane in language,

grammatical categories are also seen as categories of meaning: they are

grammaticalizations of highly generalized semantic categories (Matthiessen 1989:

4).

Grammatical meaning is, thus, distinguished from meaning in a broader sense:

how we understand what someone else has said or written can depend on a whole

range of factors. A particular language enables (and predisposes) us to express

certain general or particular semantic distinctions through its lexicogrammatical

resources; but when we begin to analyse the meanings that are created in a

particular text ) whether spoken or written, monologue or dialogue, informal or

institutionalized, and so on ) then we need to widen our perspective. We not only

need to augment a theory of the lexico-grammatical resources of a language (some

aspects of which are illustrated by this study) with a theory of cohesion (of the way

in which grammatical structures relate to the co-text and context (see e.g. Halliday

& Hasan (1976)), but, as Lemke (1988, 1989, 1990) points out, we need incorpo-

rate a wider text-semantic theory. A text-semantic theory would need to encompass

such notions as genre, intertextuality, and heteroglossia and would need to address

such issues as the social and ideological positions, the values, beliefs, attitudes etc.

of the interactants.

Page 16: systematic functional linguistic

8

1 The notion of a language as a whole is, of course, problematic: Firth (see section 2.2.3)maintained that there was no such thing; needless to say, a grammatical description based onthis notion of language can only be based on regularities and tendencies, and not on hard andfast rules.

If one regards the grammar of a language as a resource that underlies the use

of language in a variety of different contexts, then to base one’s grammatical

description on a particular corpus would unavoidably distort the picture. The notion

of “grammar” can only make sense if it is applied to a language as a whole. The

only access we have to a language ) more or less as a whole1 ) is through our

knowledge of the way in which it is used in a multiplicity of contexts (cf. section

2.3.4, which discusses E. Itkonen’s (1983a, 1983b) view of the precedence of

intuition, i.e. knowledge, in grammatical description). It would be pointless to base

an analysis of the grammatical options available in Finnish on a corpus of casual

conversation, for example, or even on the large computer corpora of written

Finnish that are now available. In any of the available corpora, it would be difficult

to find many examples of imperatives or examples like mua palelee ‘I’m freezing’.

If the example is common and uncontroversial, then it is pointless to ignore it

because for some reason it does not occur in the particular corpus that has been

selected for analysis.

However, I have avoided the use of intuited or invented examples in the

analysis of Finnish presented in this study (Chapters 4 ) 7) because they smack of

a decontextualized view of language. Moreover, often our so-called “intuitions”

about language are influenced by the view of language that we have received

through schooling and an educational system that goes back for centuries. Our

intuitions may also be influenced by standardizing pressures on the language: what

we consider to be acceptable may, in fact, be unduly influenced by what we have

been told is correct.

Page 17: systematic functional linguistic

9

1.4. Glosses

In Appendix 3 and Appendix 4, there are keys for the form and function glosses

used in the Finnish examples. Other notation conventions are listed in Appendix

5 and Appendix 6. There are two copies of each of the appendices, one of which

is detachable so that the reader can easily refer to it while reading the text. (The

loose copy is enclosed in a pocket on the inside cover of the study.) The non-

Finnish reader is advised to look at the glosses and not at the translation, as the

translation is an attempt to convey the meaning and not the grammatical relations

within the Finnish clause.

1.5. Theory and Description

This section is a brief, general discussion on theory per se, as it seems to me that

there is not necessarily a consensus on what is meant by theory and a theoretical

approach to the study of language. In my view, a theoretical approach is, in the first

instance, an attempt to “put one’s cards on the table”. It is an attempt to make

explicit the fundamental assumptions we have about language and the study of it.

Regardless of whether they are explicated or not, we all have assumptions

about language. The problem with linguistic descriptions that are not based on a

particular theory is that one is never sure what is meant by a particular term or

notion, and how the theoretical terms that are used in the description are related to

each other. If we take almost any article or book written about Finnish ) or any

other language for that matter ) regardless of whether the writer has avowedly

taken a theoretical approach, we are likely to find a plethora of linguistic terms.

This is true of traditional Fennistics. It is also true of ethnomethodological

conversation analysis, a more recent approach to linguistic analysis in Finland and

elsewhere, which has its roots in sociology. In Finland, at least,

ethnomethodological conversation analysis has been critical of theory and

preconceived theoretical categories (see A. Hakulinen ed. 1989, Ch. 1-2). The

Page 18: systematic functional linguistic

10

following terms are a random selection from some articles on ethnomethodological

conversation analysis published in Finland (Hakulinen & Sorjonen 1986; A.

Hakulinen ed. 1989; A. Hakulinen 1991; Tainio et al. 1991; Sorjonen & Heritage

1991):

syntaksi ‘syntax’, kielioppi ‘grammar’, kieliopillinen kokonaisuus ‘a grammatical entirety’,rakenne ‘structure’, syntaktinen kokonaisuus ‘a syntactic entirety’, morfeemi ‘morpheme’,lauseke ‘phrase’, lause ‘clause/sentence’, virke ‘sentence’, partikkeli ‘particle’, idiomi‘idiom’, kysymys ‘question’, leksikaalinen aines ‘lexical substance’, modaalinen aines‘modal substance’, modaaliverbi ‘modal verb’, käskylause ‘command/imperative clause’,imperatiivi ‘imperative’, NP, fonologis-morfologinen virhe ‘phonological-morphologicalmistake’, leksikaalinen virhe ‘lexical mistake’, funktio ‘function’, käyttö ‘use’, viittaus ‘refer-ence’, viitata ‘refer’, pronomini ‘pronoun’, nomini ‘noun/nominal’, verbi ‘verb’, topikaali-nen koheesio ‘topical cohesion’, suora lainaus ‘direct quote’, ajatusreferaatti ‘reporting ofthought’, prosodinen kokonaisuus ‘prosodic entirety’, adverbiaali ‘adverbial’, subjekti‘subject’, konteksti ‘context’, diskurssimaailma ‘universe of discourse’, semantiikka‘semantics’, semanttis-pragmaattisesti ‘semantico-pragmatic’, lisämerkitys ‘secondary/addedmeaning’, implikaatio ‘implicature’, spesifinen ‘specific’, geneerinen ‘generic’, kieliopilli-nen kuvaus ‘grammatical description’, ellipsis, co-referentiality, topical focus etc.

Similar terms are, of course, found in traditional descriptions of Finnish. Surely it

is of relevance to ask, for example, “What is grammar? What is syntax? What is

lexis? Are they related to each other? If so, how? What is the nature of the

relationship between them? What is a grammatical entirety? What is a syntactical

entirety? What is a prosodic entirety? Are they related to each other? If so, how?

Is a virke ‘sentence’ a grammatical unit?” etc. etc. Even if linguistic terms such as

the ones that are listed above are not used in a particular article, the fact remains

that they are part and parcel of the implicit assumptions underlying any analysis:

the study of language is theory-laden.

However, making one’s assumptions explicit is not always an easy thing to

do: there are various difficulties involved in the explication of theories and

theoretical concepts. One of the problems in presenting SF theory to linguists who

have been schooled in or have been influenced by other theories is that it quite

clearly belongs to a different paradigm, in Kuhn’s (1970) sense of the word. Kuhn

was of the opinion that different paradigms are incommensurable, and, thus, mutual

understanding is impossible, or near impossible. Moreover, the language of theory,

Page 19: systematic functional linguistic

11

like language itself, is not monoglossic: it, too, is inherently intertextual and

heteroglossic.

In spite of the difficulties involved, at least it can be said of a theoretical

approach is that it explicitly rejects what Chalmers (1982) has referred to as a

“naive inductivist” position. According to the naive inductivist, we start by

carefully observing and faithfully recording what we see and hear, and we do this

with an unprejudiced mind. Then we make statements about our observations, and

on the basis of these statements, we begin to develop our theories. The inductivist

works on the assumption that a general law or principle can be inferred by

observing particular instances. He or she assumes that observation precedes theory

and that we can make observation statements that are independent of theory, that

we can have direct, unadulterated and atheoretical access to what is happening in

the world around us. Such a view has long been questioned by philosophers of

science: no-one is born into a vacuum; our observations, and the statements we

make about them, depend on our education, our expectations, our knowledge, our

experiences and our culture, all of which can be referred to as “low-level” theory

(Chalmers 1982: 28). Observation does not precede theory, it presupposes theory.

The above is true of any of our observations, but when it comes to the

analysis of language, then any educated linguist has had at least twelve years of

exposure to a western European notion of language as it is enshrined in the

educational system (see e.g. Harris 1981, Mühlhausler 1987) and at least some

exposure to so-called traditional grammar, ie. grammatical description in which the

underlying theoretical assumptions are not explicated. Moreover, there are many

folk linguistic notions about language, which are part of the received view of

language in any society. Statements made by linguists about language are bound

to be theory-laden, the assumptions are simply not explicated (see also Joseph &

Taylor (eds.) 1990).

However, with any description of language, whether explicitly theoretical or

not, there is always the danger of making the data fit the categories that one set up

Page 20: systematic functional linguistic

12

or even making it fit the unexplicated assumptions one has about the data. With

linguistic theories, all too often the theory becomes a self-sufficient end in itself,

without regard to the nature of the phenomenon that it purports to describe (cf. A.

Hakulinen 1989: 45 ff.). The linguistic categories that are originally set up to

account for similarities and oppositions in a language become reified (cf. Halliday

1988a: 27-28), and the dialectic between theory and description that should be the

basis of linguistic analysis is lost.

The importance of a dialectic between theory and description has been

stressed from the earliest systemic work. Firth, whose ideas laid the foundation for

SF theory, repeatedly refers to the somewhat mystical sounding “renewal of

connection” (e.g. Firth 1957: 24; in Palmer ed. 1968: 19,175-176). By this he

meant the testing of a theory or hypothesis with data. This topic is taken up by

Fawcett in the foreword of a recent volume of systemic papers, which echoes

earlier statements to a similar effect (e.g. Halliday, McIntosh, & Strevens 1964: 32;

Halliday in McIntosh & Halliday 1966: 41):

The theory that [this volume] discusses is always theory that arises out of the actual textual dataof languages, and that leads back to further description ) thus completing the cycle of the‘renewal of connection’, which J.R. Firth wisely advised us to remember to make. One mighteven propose as a guiding principle: No theory without description, and no descriptionwithout a theory ) the theory, of course, often turning out to be inadequate. (Halliday &Fawcett (eds.) 1987: ix.) [Emphasis added.]

Page 21: systematic functional linguistic

13

Chapter 2

An Outline of Systemic-Functional Theory

2.1. Overview

This chapter provides a short introduction to SF theory, from the early scale and

category grammar to what was known first as systemic theory and later as

systemic-functional theory. There is particular emphasis on those aspects of the

theory that are relevant to an understanding of the grammatical analysis that is

presented in this study and the assumptions that underlie it. It also discusses a

number of issues (e.g. knowledge of language) from a wider metatheoretical

perspective. The chapter is divided into three sections. Section 2.2 is a brief

discussion of the London School of Linguistics, which is important because it

provided the base from which SF theory has developed. Section 2.3 is concerned

with general theoretical issues and some of the basic assumptions of SF theory.

Section 2.4 deals with more specific features of SF grammar. The focus in section

2.4 is on the present position, although there is some discussion of developments

that have occurred since the earliest work. As not all SF linguists would concur on

all of the theoretical issues mentioned, an attempt will be made to bring together

what is common to those working within a SF framework, and take a critical look

at some of the assumptions.

2.2. The London School of Linguistics

2.2.1. General Remarks

SF theory has developed out of the London School of linguistics (see Robins 1967:

213-220; Butler 1985: 1-13; Sampson 1980: 212-235), a loosely framed school of

thought influenced by the ideas of J.R. Firth (1890 ) 1960). Firth’s theory of

language, if it can be referred to as a theory, was never fully and coherently

explicated. His ideas about language are sketchily presented in two early works

Page 22: systematic functional linguistic

14

1 For a short overview of Firth’s ideas, see Henderson 1987.

meant for the non-specialist (Firth 1930, 1937; republished in one volume in 1964)

and two collections of articles, one of them posthumously edited by F.R. Palmer

(Firth 1957; Palmer (ed.) 1968).1

Firth’s main areas of interest were 1) phonology, in particular, prosodic

analysis and 2) meaning, where linguistic meaning was seen as being contextually

determined, both in terms of a) the linguistic context (or, in Saussurean terms, the

linguistic value) of an item at a particular level or stratum and b) the extralinguistic

context of situation. Firth (1957: 27) reserved the word “semantics” for meaning

in this latter sense, i.e., meaning considered in terms of the extralinguistic context.

This has also been referred to as both “contextual meaning” (Firth 1957: 195) and

“situational meaning” (Robins 1961: 195).

2.2.2. System and Structure

Out of Firth’s work in phonology came two sets of notions which were to be of

significance in SF theory: a) system and structure and b) a “multi-structural and

polysystemic” approach to language (in Palmer (ed.) 1968: 200). Firth’s notions of

system and structure are based on the structuralist notions of paradigmatic and

syntagmatic relations, paradigmatic relations (first used by Hjelmslev 1959 [1938]:

152) being a development of Saussure’s psychologically oriented associative

relations (Saussure 1983 [1916]: 121-125). However, in contrast to Saussure, Firth

did not regard these notions as being applicable to the language as a whole (see

next section).

For Firth, a system and a structure are complementary: a structure is formed

by elements in syntagmatic relation at a particular level of analysis, while a system

is made up of the mutually exclusive paradigmatic options that come into play at

Page 23: systematic functional linguistic

15

a particular place in a structure (in Palmer (ed.) 1968: 103). This relationship has

been mnemonically illustrated by the following diagram (Dinneen 1967: 305):

sys

s t r u c t u r eem

Firth’s approach is, however, both “polysystemic” and “multi-structural” (in

Palmer (ed.) 1968: 200). The term “multi-structural” refers to the fact that any

structure may be the result of the integration of two or more co-existing structures.

For Firth (1957: 121-123, 137), a given language is polysystemic in that it involves

a “plurality of systems”. The notion of polysystemicity seems to be a general

principle which can be applied in various ways (see Firth 1957: 121 ff., 136; in

Palmer (ed.) 1968: 43; Robins 1964: 167; 1967: 219).

2.2.3. Restricted Languages and Speech Fellowships

Because of the vastness, complexity and diversity of language, Firth (in Palmer

(ed.) 1968: 97-98,110,112; cf. Bakhtin 1981) saw the task of describing it

exhaustively an impossible one. He insisted that the techniques of description

should be applied to “restricted languages”, not to the language as a whole. A

restricted language is seen as a delimited or circumscribed sub-language within the

general language with its own grammar and dictionary (in Palmer (ed.) 1968: 29-

30,87). It provides data that is already “fenced off”, as Firth put it, for the linguist.

The examples given by Firth would seem to indicate that the manner in which a

restricted language is “fenced off” was not important for Firth: his examples

suggest that a restricted language corresponds not only to a register, but also to the

language of a particular text or set of texts. Examples given by Firth (in Palmer

(ed.) 1968: 29,87,98, 106,112,118-119) include the language of modern Arabic

headlines, of politics or meteorology, or of a particular text or a particular writer.

Page 24: systematic functional linguistic

16

1 Personality, in the Firthian sense, can thus be seen at the intersection of the various socialnetworks (in the sense of Milroy 1987) to which a person belongs.

As well as making a distinction between the general language and a restricted

language, Firth also made a distinction between the general language community

and what he referred to as a “speech fellowship”. In a speech fellowship, the speech

of a group reflects a bond of fellowship based on the “sharing of a truly common

experience” (Firth 1957: 186). Speech fellowships are reflected in the study of

persons or what Firth referred to as “personalities”, i.e. social (or socially

constructed) entities who actively participate in the creation and maintenance of a

particular culture or sub-culture. While Firth maintained that he was interested in

the specific person, he qualified this by saying that those that are to be studied are

“representative” or “usually typical of an important speech fellowship in a wider

speech community” (1957: 143,226; in Palmer (ed.) 1968: 32,187). Such persons

are seen as being in command of “a constellation of restricted languages” which

are “governed by personality in social life and the general language of the

community” (in Palmer (ed.) 1968: 207-208).

Thus, for Firth what was linguistically salient was the social entity, the

“personality”, rather than the biological or natural entity, which Firth referred to

as the “individual”. Firth (1957: 28,184) saw the social person as the product of the

various social roles she or he has to play in a particular society and likened a

person to an actor in a play with various roles to play.1 Social roles are learnt

almost from birth as a person is incorporated into various speech fellowships in a

particular society.

Members of a speech fellowship belong to a wider language community, and

the widest community based on English is the English-speaking world. This wider

community is not homogeneous, but based on diversity; it is not founded on a

standardization which neutralizes the various accents, registers and dialects, but on

the diversity of linguistic “personalities”:

Page 25: systematic functional linguistic

17

Members of various speech fellowships may, however, belong to larger speech or languagecommunities without conflict of values. Both sets of values deserve respect. The vastenterprises of the English-speaking world, operated by English, go on without standardizationof accent. You may estimate the relative values of what is called an Oxford accent, anAberdonian accent, a Boston, a New York, or an Australian accent, but the main thing is awider language community with room for diversity of personality. (Firth 1957: 186.)[Emphasis added.]

As suggested earlier, Firth was aware of the problems inherent in describing

a monolithic system which purported to reflect the unity of language:

The multiplicity of social roles we have to play as members of a race, nation, class, family,school, club, as sons, brothers, lovers, fathers, workers, ... public speakers, involves also acertain degree of linguistic specialization. Unity is the last concept that should be applied tolanguage... There is no such thing as une langue une and there never has been. (Firth1957: 29.) [Emphasis added.]

Thus, Firth’s notion of a general language, such as English or Swahili or Finnish,

can be seen as the union of a vast number of subsets of restricted languages. Firth’s

response to the complexity and diversity of language phenomena was to confine

the analysis to restricted subsets of the language.

Firth’s approach to the vastness of language phenomena can be contrasted

with the approach taken by Chomsky. Chomsky’s (1965: 3-4) response was to set

up a dichotomy between competence and performance, between the speaker-

listener’s knowledge of a language and the actual use of language (as an individual

activity) in concrete situations, and, furthermore, to regard linguistic theory as

being concerned with “an ideal speaker-listener in a perfectly homogenous speech-

community, who knows its language perfectly”.

Thus, while Firth at least purported to be concerned with abstracting and

making generalizations from the speech of specific persons typical of a speech

fellowship, Chomsky was concerned with the representation of the grammatical

knowledge of a collective uniformity. The ideal speaker-listener is a fiction: “it”

is a member of society in which there are no divisions according to race, sex, class

Page 26: systematic functional linguistic

18

or social position. Chomsky’s approach could be seen as focusing on the

intersection of the restricted languages within a given language, or, given his

emphasis on what is common to all languages, on the intersection of the intersec-

tions of the restricted languages of all languages. Consequently, Chomsky’s

approach has led to a view of grammar where all variation is neutralized and

grammatical analysis is restricted to a small set of phenomena which are possible

candidates for language universals. No doubt, Firth would have felt that

Chomskyan linguistics had very little to do with the “living of life” (in Palmer (ed.)

1968: 169). Regardless of their theoretical manifestos, however, in practice both

Firth and Chomsky have based their grammatical descriptions on middle-to-upper

class standard varieties. Firth’s phonetic and phonological analyses of English (in

Firth 1957), for example, are clearly based on a standard middle-to-upper class

variety.

2.2.4. A General Linguistic Theory

Firth (1957: 144; in Palmer (ed.) 1968: 152, 190-202) was interested in a general

linguistic theory, a theoretical framework with which to approach the description

of any given language, but not in a theory of linguistic universals in the sense that

it has become familiar in Chomskyan and post-Chomskyan linguistics:

What is here being sketched is a general linguistic theory applicable to particular linguisticdescriptions, not a theory of universals for general linguistic description. (Firth in Palmer (ed.)1968: 190.)

In response to universalist statements such as “there are no real adjectives in

Swahili”, Firth warned against the “hypostatization”, or reification, of grammatical

categories. The grammatical categories for a particular language are abstractions

which are determined by the interrelations in the systems set up by the linguist for

that language, not realities which are either present or absent in a language. (See

also Hasan 1971.) Moreover, a grammatical category is regarded by Firth (in

Palmer (ed.) 1968: 39) as ineffable: it eludes our conscious attempts to define it.

Page 27: systematic functional linguistic

19

(This problem has been taken up by Halliday (1988a) and will be discussed in

section 2.4.11 (p. 63).)

2.2.5. Meaning

Firth saw meaning as the cornerstone of linguistic theory: the study of language

is the study of linguistic meaning (1957: 190; in Palmer (ed.) 1968, Ch.1).

Moreover, linguistic meaning could only be understood by appreciating the

intimate relationship between language and society. As Firth (1957: 226; in Palmer

(ed.) 1968: 12-13) points out, words are not isolates which somehow have meaning

in and by themselves, as logicians and some linguists would have us believe; they

have meaning because they function in the particular society in which the speakers

happen to live. Thus, language is seen not in terms of an individual mental activity

or as an abstract construct divorced from reality, but as an integral part of the

physical and social world in which we live. Meanings are created in society:

As we know so little about mind and as our study is essentially social, I shall cease to respectthe duality of mind and body, thought and word, and be satisfied with the whole man, thinkingand acting as a whole in association with his fellows. I do not therefore follow Ogden andRichards in regarding meaning as relations in a hidden mental process, but chiefly as situationalrelations in a context of situation and in that kind of language which disturbs the air and otherpeople’s ears, as modes of behaviour in relation to other elements in the context ofsituation. (Firth 1957: 19.) [Emphasis added.]

Firth extended Malinowski’s (1923) view of meaning as “function in context”

to incorporate linguistic contexts at all levels: meaning originates not only in the

social context but in successive linguistic contexts. Linguistic meaning in its

entirety was thus seen as a “complex of contextual relations”:

Meaning ... is to be regarded as a complex of contextual relations, and phonetics, grammar,lexicography, and semantics each handles its own components of the complex in its appropriateenvironment. (Firth 1957: 19.)

Page 28: systematic functional linguistic

20

Each successive linguistic context is, thus, seen as providing a step in the access

to the total meaning of an utterance. Firth, pace Harris (1987a), advocated a two-

way approach: one can either work from the context of situation to phonology or

from phonology to the context of situation (1957: 192).

In discussing his approach to meaning, Firth (e.g. Firth 1957: 19; in Palmer

(ed.) 1968: 200) often used the analogy of the dispersion of light waves into a

spectrum: just as white light is the fusion of a number of colours of differing

wavelengths, linguistic meaning is the fusion of a number of different “modes of

meaning”. This fusion of meaning is impossible to analyse until it is dispersed (or

deconstructed) into various modes of meaning.

It is necessary ... to split up the problem of meaning into its components or elements. Theprocess may be compared, metaphorically speaking, to the dispersion of white light into aspectrum by means of a prism. The prism in our case is descriptive linguistics and the spectrumis the multiple statements of meaning at various levels. (Firth in Palmer (ed.) 1968: 108.)

2.2.6. Rejection of Saussurean Structuralism

While Firth’s notions of a system and a structure are based on the Saussurean

notions of syntagmatic and associative relations, and, in some respects, his

approach to meaning relies on Saussure’s notion of value, Firth did not consider

himself a Saussurean and explicitly rejected many of Saussure’s ideas (Firth 1957:

36,179-181; in Palmer (ed.) 1968: 127-129; Halliday 1978: 51). What is central to

the Saussurean model of language is langue, a static synchronic system in which

there are no positive terms and everything is defined negatively in terms of abstract

relations of opposition (Saussure 1983 [1916]: 118). Such a conception of

language, as Firth (1957: 180-181) points out, excludes not only actual words and

sounds, but also the actual speakers of the language. Thus, Saussurean structural-

ism leads to a reification of langue, and the concrete dialogic nature of language

is ignored (cf. Voloshinov 1973 [1930]). If we ignore the constant dialectic

between language as system (or as multiple systems) and language as “speech and

... texts related to the living of, and therefore to the ‘meaning’ of life” (in Palmer

Page 29: systematic functional linguistic

21

1 See section 2.4.14, for a further distinction between a synoptic and a dynamic perspectiveon language.

(ed.) 1968: 169), then we also ignore inter alia the means of exploring linguistic

variation and change. This distinction between language as system and language

as speech and text is comparable to Hjelmslev’s (1953: 5) notion of language as

system and language as process, which, in later SF terms could be glossed as

language as system and language as (spoken or written) text.1 For Firth, the

dialectic between system and text is central:

Renewal of connection with the processes and patterns of life in the instances of experience isthe final justification of abstract linguistics. (Firth 1957: 24.)

2.2.7. Firth and SF Theory

Many of the assumptions underlying Firth’s approach to linguistic analysis were

carried over into SF linguistics. While there have been many changes of emphasis

and direction, Firth’s influence and input is still evident. SF theory has retained

Firth’s focus on text and has built upon and developed his ideas on restricted

languages, the context of situation, and on the notions of system and structure.

More importantly, however, Firth’s influence is evident in the multifaceted and

wide-ranging approach to meaning: SF grammar is about meaning, about the

resources that are available in a language that allow us to say and do meaningful

things. It is about the lexicogrammatical resources that allow us to make meanings.

Page 30: systematic functional linguistic

22

2.3. SF Theory: General Considerations

2.3.1. Language as a Linguistic Behaviour Potential

For systemic-functional linguists, linguistic behaviour is a form of social behaviour

(Halliday 1978: 36-39): people are socialized into a particular culture and have

learnt to interact in meaningful ways. For example, ways of coming into actual

physical contact with another person (e.g. shaking hands, touching another person’s

shoulder or leg, hugging, kissing, slapping, and hitting) are meaningful and the

meaning of a particular action can vary from one culture (or sub-culture) to the next

and can co-vary with any of a multitude of factors such as gender, social status,

class, profession etc. A particular culture thus embodies a number of choices in its

socially sanctioned forms of interaction. These choices can be seen as making up

a resource or a potential of socially significant ways of behaving. This resource or

potential is referred to by Halliday as a “behaviour potential”.

Although much of our social behaviour, i.e. the behaviour we learn as we

become socialized into a particular group or society, is expressed in non-linguistic

ways, more significant, at least from the point of view of linguistics, is the way in

which we behave linguistically. If, for example, A and B are both speakers of

English, and A wants B to close the door, it is conceivable that A could point in the

direction of the door with appropriate gestures or glare at both B and the open door,

but there are a number of ways of achieving this linguistically, e.g. “How many

times have I told you to close the door behind you”, “Please close the door”,

“Would you please be so kind as to close the door behind you”, “Were you born

in a tent?” etc. From this perspective, a language can be seen as a form of

behaviour which finds expression linguistically. It is a resource or a potential for

doing meaningful things, enabling us to achieve certain ends in certain ways. It can

be represented as a system of choices or options. The use of the term “choice” or

“option”, however, is not meant to suggest that choices in language are simply a

matter of the speaking making a rational decision (see next section and Thibault

1987: 604,607).

Page 31: systematic functional linguistic

23

A language can be seen as a “meaning potential” or a “semantic potential”

(see e.g. Halliday 1973: 51,72); or, perhaps, more appropriately, as a “meaning-

making” potential ) the notion of a potential should not be misunderstood as a

static inventory of predetermined meanings. The semantic potential of a language

is realized in terms of the linguistic resources that are available, i.e. in terms of the

lexicogrammatical system of the language. The semantic potential and the (non-

linguistic) behaviour potential together contribute to the total semiotic potential of

a society.

2.3.2. The SF Interpretation of Langue

To see language as a resource, rather than as a structure or as a system of rules, has

important repercussions for the way in which Saussure’s notions of langue and

parole are interpreted. In SF linguistics, this distinction is seen in terms of

potential vs. actual (Berry 1975: 24; Halliday 1978: 37-38; Fawcett 1980: 55). A

potential can be regarded as what it is possible for people to say in a particular

language, it can be regarded as a system of lexical and grammatical choices ) the

meaning-making resources ) that are available in any language. Actual linguistic

behaviour can be seen as what is said at a particular time and place, as the

particular configuration of choices that is made.

In contrast to Firth, who did not think it was feasible to look at language as

a whole and wanted to restrict linguistic analysis to a subset of a language as

reflected in his “restricted languages”, systemic-functional linguists see language

as the union of all of these subsets. Like Firth, however, they would still reject the

notion of language as a unity. Language is not unified, but inherently variable.

While there is variation in language, it is not random, but can be correlated with

various factors. These factors can be roughly grouped under a number of headings:

Page 32: systematic functional linguistic

24

(1) Variation that can be associated with social phenomena such as gender,

class, etc. (see e.g. Hasan 1987a, 1988, 1989; Hasan & Cloran 1990).

(2) Variation related to differences in the contexts of language use (see 2.3.7).

(3) Variation related to regional (or geographical) differences in language.

An important consequence of regarding langue as a potential is that it allows

us to say that not all individuals (or, more correctly, groups of individuals) have the

same access to this potential. (This also applies to the linguist; see next section.)

An obvious example, is the differential access to legal institutions: many of us are

unfamiliar with legal procedures and courts of law, and, consequently, have to rely

on the services of a lawyer. However, there are also more subtle ways in which

some social groups do not have access to potential forms of linguistic interaction,

and this may put them at a social or educational disadvantage. For example,

Bernstein has attempted to show that the ways in which it is natural for lower

working class pupils to interact is different from the ways in which middle class

children interact, and, because the educational system is based on middle class

assumptions, working class children are at a disadvantage. (For an overview, see

Bernstein 1987.)

Saussure (1983[1916]) saw langue as basically a property of the speech

community, and, in this sense, the SF notion of langue as a potentiality is closer

to the Saussurean notion than Chomsky’s (1957) notion of competence, which is

a property of the individual. For Saussure (1983: 19), however, while langue was

a distillation of social behaviour, it “takes the form of a totality of imprints in

everyone’s brain, rather like a dictionary of which each individual has an identical

copy”, and it is this position which has led to an internalization of the object of

inquiry, as taken up by Chomsky and others. With Chomsky, this has led to an

extreme psycho-biological approach to language. Many other linguists, while they

may not share Chomsky’s views, still tend to view linguistics as a branch of

cognitive psychology.

Page 33: systematic functional linguistic

25

2.3.3. The Individual and the Social Aspect of Language

Systemic-functional linguists stress the social and interactive nature of language

(Halliday 1978: 12-13; 38-39; 56-57; Fawcett 1980: 6; Berry 1975: 22-23).

However, language as a social and interactive phenomenon can be viewed from

two perspectives. Halliday (1978: 12-16, 56-57) contrasts what he refers to as

“inter-organism” and “intra-organism” perspectives, which he sees as complemen-

tary. This is similar to Saussure’s (1983: 8-9) view of the complementarity of the

individual and social aspects of language. According to Saussure, the individual

and the social are complementary in the sense that they are dependent on each

other: “one is not conceivable without the other”.

The inter-organism perspective sees language externally: as something that

is happening between people. The intra-organism perspective looks at it internally:

as something that is going on in a person’s head. Halliday’s main interest is in an

inter-organism perspective on language; there are others within SF linguistics,

notably Fawcett (1980), who subscribe to an intra-organism perspective.

The intra-organism perspective on language is further complicated by the fact

that it can be interpreted in various ways. When Halliday first made the distinction

it was made at a time when linguistics was dominated by Chomsky’s assumptions

about the nature of language and the purpose of linguistics. In Chomsky’s (1968,

1976) approach, knowledge of language refers to abstract knowledge of the rules,

principles, and conditions that characterize all human languages. This knowledge

is abstracted by the linguist from “context-free” sentences, and related to innate,

but as yet unknown, mechanisms in the human brain. However, knowledge of

language involves a lot more than abstract rules about the organization of language:

We do not simply ‘know’ our mother tongue as an abstract system of vocal signals, or as if itwas some sort of grammar book with a dictionary attached. We know it in the sense ofknowing how to use it; we know how to communicate with other people, how to chooseforms of language that are appropriate to the type of situation we find ourselves in, and so on.All this can be expressed as a form of knowledge: we know how to behave linguistically.(Halliday 1978: 13.) [Emphasis added.]

Page 34: systematic functional linguistic

26

It is in this wider sense that an intra-organism perspective is seen as being

complementary to an inter-organism perspective.

An intra-organism perspective on language in use is somewhat different to

Hymes’ (1967, 1986) notion of “communicative competence”. As Halliday (1978:

92) points out, the very notion of communicative competence assumes that there

is a distinct kind of competence (i.e. Chomsky’s grammatical competence) that is

based on one’s knowledge of language in vacuo. From a SF perspective, the only

kind of competence is communicative competence:

If our goal is the pursuit of system-in-language (Fishman 1971: 7), this is surely linguistics,and linguistics always has ... accepted what Hymes (1967) calls the “socio-cultural dimensionsof its subject-matter”, the link between language and the social factors that must be adducedto explain observed linguistic phenomena. By the same token, however, we do not need“communicative competence”, which has to be adduced only if the system has first beenisolated from its social context. If we are concerned with “what the speaker-hearer knows”, asdistinct from what he can do, and we call this his “competence”, then competence iscommunicative competence; there is no other kind. (Halliday 1978: 92.) [Emphasisadded.]

2.3.4. Knowledge of Language

On the other hand, from the perspective of the philosophy of science, it has been

cogently and convincingly argued by E. Itkonen (1983a, 1983b) that knowledge

is epistemologically prior in any investigation, whether the object of investigation

is language or electricity. Any investigation of language must be based on

assumptions that a linguist makes about the regularities in that language, and this

is true whether the language in question is the investigator’s native language, a

foreign language, or even a extinct language. Regularities or tendencies or rules are

not concrete entities that can be observed, they can only be intuited or abstracted

on the basis of our analytical reasoning processes.

While Itkonen talks of “knowledge of language”, he does not give it a

mystical status, but insists on the primarily social nature of knowledge and

language, and on the intersubjectivity of the rules of language (E. Itkonen 1978).

Page 35: systematic functional linguistic

27

1 An idealist might want to deny this. The fact remains, however, that while there isdifferentiation according to class, race, ethnicity, education, age or sex in any society andregional variation in nationally recognized languages, a privileged status is only afforded tocertain varieties: it would be difficult to imagine the president of Finland giving a speech inSavo dialect or a black American judge handing down a sentence in black English vernacular,except in a comedy programme on television.

Moreover, when he describes it as “intuitive”, the implication is not that a native

speaker’s knowledge of a language is innate, but that it is internalized, and, in order

to be explicated, involves introspection. The analysis of a linguist’s knowledge of

the rules of a language is referred to by Itkonen as “autonomous linguistics” (i.e.

this is how Itkonen (1978) defines autonomous linguistics). According to Itkonen

(1983a: 1), autonomous linguistics “investigates ‘language in itself’”, abstracting

from either the social or psychological mechanisms that sustain it. It is “the

conceptual precondition of linguistic investigations dealing with external evidence”

(Itkonen 1983a: 10). As such, autonomous linguistics does not make use of

external evidence, and the use of a corpus is justified only in the analysis of unclear

cases and in the testing of descriptions.

While it seems clear that Itkonen is correct in maintaining that a linguist must

always proceed from (intuitive) knowledge of a language, and this is an important

consideration in the development of a theory, 1) one needs to carefully consider the

implications of what it means to say that knowledge of language is primarily social,

and 2) one needs to look more closely at the status of a corpus in linguistic theory

and description. The first point will be discussed first.

If knowledge of language is social, then it is learnt from experience, through

interaction with other socialized beings. Our knowledge of a language, thus,

depends on our capacity to abstract from our experiences with a particular language

and language community. If this is accepted, then it follows that a linguist’s

knowledge of language will depend not only on her or his social positioning but

also on the types of language that are legitimized through social and educational

institutions.1

Page 36: systematic functional linguistic

28

Moreover, when we attempt to articulate this knowledge, we cannot assume

that we have direct or unadulterated access to it (cf. 1.5): we cannot overlook the

fact that many of our supposed “intuitions” about language depend on a received

view of what language is and how it works. Harris (1981), for example, has

referred to the received view of language in Western societies as “the language

myth”, a myth which he sees as having been perpetuated by modern theoretical

linguistics. One facet of the language myth is the fixed code fallacy, a fallacy

which has been institutionalized by an educational and political system intent on

standardizing the linguistic behaviour of pupils and which is based on the

sacrosanctity of the dictionary and grammar book.

Furthermore, our view of language is also influenced by the language we

speak, since any language incorporates theories about reality in its very structure

and organization (Whorf in Carroll 1956; Popper 1972: 165). Language itself is

part of that reality, and as Reddy (1979) has demonstrated, the English language,

for example, incorporates certain ways of talking about language, and, conse-

quently, predisposes us to talk about language in a particular way. The situation is

further complicated when we look at the position of the linguist as compared with

that of the non-linguist. While even the theoretically naive native speaker is

exposed to esoteric knowledge about language, i.e. the kind of knowledge that is

constructed in educational institutions, the position of the linguist is more complex

in that she or he, like any scientist, is schooled into a particular way of seeing

things (cf. Kuhn 1970).

To turn to the second point concerning the status of a corpus in linguistic

theory and description, as stated earlier, Itkonen (1983a, 1983b) sees the use of a

corpus as being justified only in the analysis of unclear cases and in the testing of

descriptions. While Itkonen (1983a: 10) does not claim that the use of a corpus is

incompatible with autonomous linguistics (as he defines it), he nevertheless gives

a prior and privileged status to the use of intuited examples. More recently,

however, there seems to be a slight shift in Itkonen’s position (1990: 354-355):

Page 37: systematic functional linguistic

29

instead of minimizing the role of “external evidence”, he stresses the need for a

symbiotic relationship between a corpus, on the one hand, and, on the other hand,

one’s knowledge of language as manifest in the use of intuited examples and

grammatical tests (such as the deletion test used in determining, for example, the

core or obligatory actants in a process (see Chapter 6)).

One the other hand, one cannot simplistically assume that a corpus is

objective simply because it consists of text or fragments of text that have not been

intuited by the linguist. As Itkonen (1983a: 8) points out, an act of observation is

not necessarily any less subjective than an act of intuition. The choice of one

example rather than another is a subjective choice, as too is the choice of the text

from which the example is taken. Thus, the choice of a particular text representa-

tive of a particular genre ) e.g. newspaper editorials, dialects interviews (or

fragments from them), interpersonal (casual) conversations ) are not objective

choices, but necessarily involve theoretical assumptions and skew the description

towards a particular genre. In all of the examples just given, for instance, language

is constitutive rather than ancillary ) as in a service encounter or in a game of ice

hockey ) and this is true of the vast majority of linguistic corpora that I am aware

of.

As pointed out in Chapter 1, I have avoided the use of intuited examples

because 1) they smack of a decontextualized view of language and 2) one’s

knowledge of (the use of) language is often unduly influenced by received views

of language. This study, however, is not a corpus-based study in that it is not based

on a particular corpus. The text examples have been selected on the basis of my

knowledge of the way in which Finnish is used in Finland: they are used to

illustrate my knowledge of Finnish. This, as Itkonen’s points out, must be the

starting point.

Page 38: systematic functional linguistic

30

2.3.5. A Reality Construction View of Language

The view of language assumed in SF theory and in this study is what Grace (1987)

refers to as a “reality construction view” of language. Grace is not a systemic-

functional linguist, but his views are presented here as it seems to me that they are

more accessible to those unfamiliar with systemic-functional linguistics than the

writings of systemic-functional linguists. While the views of systemic-functional

linguists may be more developed than those of Grace (1987), they are also more

complex and, possibly, more esoteric (see, e.g. Halliday 1973, 1974, 1978, 1984,

1987; Hasan 1984a, 1988, 1989).

Grace distinguishes his “construction view” of language from a “mapping of

a common reality view”. In the “mapping view”, languages are analogous to maps

of a common reality; in the construction view, language is seen as one of the

essential means by which reality is socially constructed. Grace subscribes to a

construction view of language, as would many systemic-functional linguists. The

discussion that follows is based mainly on Grace (1987); however, as I do not

entirely agree with his interpretation of a reality construction view, I depart from

it in some respects. First, I shall explore some of the implications of the mapping

view.

The mapping view is seen as the predominant view in linguistics; it is the

view of language which underlies “normal science” (Kuhn 1970). It assumes there

is a pre-existing language-independent reality, which can be talked about through

language. Each language is like a map of this reality. Although each language does,

in fact, provide a somewhat different mapping of reality, the differences are merely

differences in classification: languages divide reality up in different ways, they

provide different maps of the same content. The key assumption of the mapping

view, according to Grace, is the intertranslatability postulate, i.e. anything can be

said in any language. This assumption can be made because languages are seen as

empty codes which mediate a universally-shared reality.

Page 39: systematic functional linguistic

31

1 To say that reality is unknowable is not, of course, the same thing as saying that it does notexist.

Because languages are like maps of a language-independent reality, the

semantics of language is seen in terms of truth-conditions. This truth-conditional

approach is criticized by Grace and others (e.g. Harris 1987b) for seriously

misrepresenting the way in which language really works. In Harris’s view, truth-

conditional semantics is part of a view of language which divorces language from

human activities and truth from human morality and values, it is “a semantics for

robots, not human beings” (1987b: 159). (See also Shore 1991a.)

Whereas the mapping view is based on objectivist assumptions about reality,

the unknowability of reality is central in a reality construction view. Because

reality is unknowable,1 the languages we speak cannot represent it; instead, they

provide models of it (Grace 1987: 6; Halliday 1987). Language is assumed to

represent a reality which has been created by human beings rather than reflecting

some objective external reality. (Grace 1987: 118.) The constructionist view

emphasises the social and cultural shaping of our effective environment, i.e. the

world in which we live and conduct our daily affairs. It is a world which, to a large

extent, has been created by language and knowledge of which is maintained and

transmitted through language.

In a constructionist view (Grace 1987: 10,121-124), there is no clear line

between thinking, bringing a thought into being, and encoding the thought by

putting it into words. Grace sees thoughts as being dependent on words. I would

expand this to say that thoughts are dependent on our semiotic systems, and the

most important of our semiotic systems is the language that we use in everyday

life. (Another important consideration is how one understands the notion

“thought”. Some linguists appear to see thought in terms of problem-solving rather

than as a semiotic (meaning-making) process.)

A constructionist view rejects a truth-conditional approach to semantics.

Grace (1987: 49) discusses semantics in terms of “goodness of fit”: “instead of a

Page 40: systematic functional linguistic

32

1 It is clear from the context, that Halliday is referring to truth in the truth-conditional sense.

sharp line between truth and falsity there are degrees of goodness of fit”. Thus,

when someone makes a statement such as “The dog bit the man” or, to use Austin’s

example, “France is hexagonal”, according to Grace, the statement is judged on the

basis of the goodness of fit between the model and the actual reality. It seems to

me, however, that Grace is still tied to a tradition that sees semantics as being

primarily concerned with truth and falsity, in Grace’s case, with degrees of truth

or falsity. Halliday (1985: 76), on the other hand, maintains that “semantics has

nothing to do with truth”1, and this is reflected, for example, in his analysis of

polarity and modality. Semantics is seen as being concerned with how we use

language as a resource for meaning, how we can do meaningful things with

language, either truthfully or falsely. (See also Harris’ (1987b: 158 ff.) discussion

of truth-conditional semantics, truth and morality.)

The rejection of a truth-conditional approach has repercussions inter alia for

the way in which clause types are analysed in this study. An equative clause, for

example, has been defined in truth conditional terms as one in which the (“real-

world”) referent of two NPs is the same, i.e. the NPs are co-extensive (see

6.4.1.(ii)). Thus, the clause The barn is now the garage is not an equative clause

as the NPs are not really co-extensive, but become co-extensive once the sentence

has been uttered. This logical approach is based on the relating of clauses to real-

world referents, of seeing bits of language as parts of a map of the real world, as

providing labels or names for phenomena in the real world. As Hasan (1988: 48ff.)

points out, this approach ignores an essential feature of language: language is a

meaning-making resource and it has an active role in the creation of meaning.

While Grace acknowledges the importance of the social and cultural in vague

general terms, he looks at the constructionist view from the point of view of the

individual. When he discusses the role played by the speaker and the speech event

in mediating between linguistic expressions and reality, both speaker and speech

Page 41: systematic functional linguistic

33

event are discussed as though they were asocial constructs. Grace, in fact,

downplays the social and looks at language as though it were entirely a mental

phenomenon and the property of the individual. This is where he departs from the

systemic-functional view, which stresses the social, and explores how socio-

contextual factors shape language (see, e.g. Halliday 1973, 1974, 1978, 1984,

1987; Hasan 1984a, 1988, 1989; Hasan & Cloran 1990).

2.3.6. Language and Mind

If language is looked at as a mental phenomenon, then the focus, from a SF point

of view, is on those aspects of the mental that are a reflection of the social. This is

not to deny that there are features of the way in which language works that are not

dependent on social factors (for example, analogical processes in language (see e.g.

Anttila 1977, E. Itkonen 1991) and the actual mechanisms of speech), or that there

are universal features of language which may be shaped or constrained by the

nature of human perception or by features of human abilities.

For the systemic-functional linguist, to see the individual as largely a product

of the social involves a number of premisses. Firstly, along with Vygotsky (Wertch

& Stone 1985) and the Soviet semioticians of the Bakhtin circle (Voloshinov 1973

[1930]), systemic-functional linguists would argue that individual consciousness

emerges out of socially organized experience, i.e. out of the social life into which

an individual is born and into which she or he is integrated (see, for example,

Hasan 1988). Moreover, mediation between the social (external) and the individual

(internal) is based on semiotic processes, and especially on language. Thus,

according to the Soviet semioticians of the Bakhtin circle:

Consciousness becomes consciousness only once it has been filled with ideological (semiotic)

content, consequently, only in the process of social interaction ...

If we deprive consciousness of its semiotic, ideological content, it would have absolutely

nothing left. Consciousness can harbour only in the image, the word, the meaningful gesture,

Page 42: systematic functional linguistic

34

and so forth. Outside such material, there remains the sheer physiological act unilluminated by

consciousness, i.e., without having light shed on it, without having meaning given to it, by

signs.” (Voloshinov 1973[1930]: 11-13). [Emphasis added.]

Hasan (1985a: 32) has argued against Leech’s (1974: 66) view that a

mentalist theory of meaning could be seen as a recognition of a common-sense

reality that “meaning is a mental phenomenon and it is useless to pretend

otherwise”. In Hasan’s view, the fact that meanings are stored in the human brain

does not make meaning a mental phenomenon:

There is a very obvious sense in which every piece of knowledge is mental. Whatever thechild’s or adult’s understanding of the linguistic sign dog, this understanding is surely storedin the brain; further, it is only because of the structure of the brain that it is possible for humansto arrive at understandings of this sort. But in a rather important sense it does not make themeaning of the sign dog a mental phenomenon; the dictionary may be located in the brain butthe specific details relating to each entry in the dictionary originate not in the brain but in thesocial human milieu. Meaning and mind are created in a social environment, throughsocial agencies ... [Emphasis added.]

This does not, of course, mean that there is a simple one-to-one correspondence or

isomorphism between the external and the internal, as Vygotsky (Wertsch & Stone

1985: 166-167) pointed out. This is evidenced not only by the fact that both society

and language change over time, but also by the fact that at any given time, there are

co-existing variants of each (Hasan 1988: 45).

2.3.7. Grammar, Semantics and the Context of Situation

SF theory has always rejected the notion of a “context-free” sentence. A basic

assumption, which underlies any SF analysis, is that people not only talk, they talk

to each other in socially recognizable forms of interaction. As Halliday (1978: 28-

29) has put it:

We do not experience language in isolation ... but always in relation to a scenario, somebackground of persons and actions and events from which the things are said to derive theirmeaning. This is referred to as the “situation”, so language is said to function in “contexts of

Page 43: systematic functional linguistic

35

1 Thus the SF notion of genre covers not only written genres but also includes what Bakhtin(1986: 78-79) refers to as “speech genres”.

situation” and any account of language which fails to build in the situation as an essentialingredient is likely to be artificial and unrewarding.

This is similar to Searle’s (1979: 117) argument against the context-free interpreta-

tion of sentences:

There is no such thing as the zero or null context for the interpretation of sentences ... Weunderstand the meaning of such sentences only against a set of background assumptions aboutthe contexts in which the sentence could be appropriately used. (Searle 1979: 117.)

SF theory clearly belongs to what Harris (1987a: 131) has called the

integrational paradigm: one which sees language as “manifested in a network of

human abilities and activities complexly integrated in social interaction”. It is from

this position that systemic-functional linguists give prominence to the context of

situation and generic subcategories of speech events (i.e. to genres such as medical

consultations, scientific reports, casual conversation, service encounters etc.).1

However, such notions as context of situation and genre can be conceived of as

quite separate from grammatical analysis. As Harris (1987a) points out, the context

of situation can be seen as something which can be tacked on in order to complete

the description of decontextualized sentences that have already been analysed in

phonological and grammatical terms. It can be thought of as an optional extra,

which has no real bearing on the grammatical analysis. Harris suggests that a true

grasping of the full theoretical implications of the linguistic significance of the

context of situation requires that it is integrated into our grammatical analysis: the

analysis of grammatical organization of a language cannot be divorced from the

description of language in use.

In SF theory, the context of situation is built into the grammatical analysis of

a clause. The link between the grammatical form of language and the situation is

Page 44: systematic functional linguistic

36

1 Thus, the grammatical (morpho-syntactical) form of a language is not a semantically-emptyframe into which lexical items are slotted. See 2.4.7 for a discussion of realization.

made possible by a hypothesis, a “metafunctional” hypothesis, about the relation-

ship between them (Halliday in Thibault 1987: 608-610). A metafunctional hypo-

thesis is not a hypothesis about the functions or uses of text, as reflected in

functional theories of language put forward by scholars such as Malinowski and

Bühler who were working in other disciplines (Halliday 1981a: 33-35; Halliday &

Hasan 1985: 16ff.). It is a hypothesis about the internal grammatical organization

of the clause, about those functions of language that are built into the very structure

and organization of language itself. These functions of language can be seen as

generalized types of meaning, which are realized in the lexicogrammatical form of

a language.1 The types of meaning recognized by Halliday are interpersonal,

ideational (which is subdivided into experiential and logical) and textual (see also

2.4.5 and 2.4.8).

Interpersonal meaning is the meaning associated with language as a way

of getting things done, as a way of acting upon the world in which we live. It

reflects Malinowski’s (1923) and Austin’s (1962) view of language as a mode of

action. Interpersonal meaning can be subdivided into interactional and attitudinal

meaning. Interactional meaning has to do with the interactional roles that are

created in the speech situation (e.g. giver/demander of information). Attitudinal

meaning is concerned with the way in which the speaker (or listener) relates to

what is being said. This is concerned with the way in which speakers assess truth

or falsity, probability or improbability, frequency or rarity of occurrence, obligation

and willingness. Interpersonal meaning is mostly realized in the mood and

modality systems of a language.

Ideational meaning is concerned with what has traditionally been referred

to as semantics. It is concerned with the way in which language mediates about the

reality which we assume to be in us and around us, both real and imagined; it is

Page 45: systematic functional linguistic

37

concerned with the way in which language serves as a model of reality. It can be

subdivided into experiential and logical meaning. Experiential meaning is

concerned with the “things” that we can talk about, and logical meaning with the

interdependency relationships that language allows us to construct between these

things. It is important to note that logical meaning is concerned with the logic of

a language and not with formal or mathematical logic, which has been based on

and has developed from logical expressions in language. Ideational meaning is

realized in what Halliday refers to as the transitivity system and in systems

concerned with the interdependency relations between units in a language.

Textual meaning relates to the way in which language makes links with

itself and with the situations in which it is used. Speech (and writing) is not just a

random collection of words that are unrelated 1) to what the speaker (or someone

else) has just said (or written) or 2) to the context in which they are uttered. Textual

meaning is made manifest in the resources that allow the speaker to create a

coherent text, one that makes sense in the context in which it is being uttered and

in the context of what has been uttered and will be uttered. It is mostly realized in

the theme and information focus system.

These types of meaning are referred to as “metafunctions” because they are

a result of a number of very broad and abstract uses that language has evolved to

serve. The metafunctions can be seen as the “interface” which links language to

other semiotic systems, they provide a link between language and what is outside

language (Thibault 1987: 608). They link up with another interface, the context of

situation, which is defined by Halliday (Thibault 1987: 610) as a “generalized

semiotic construct”:

The context of situation is a generalized semiotic construct deriving from the culture )something that is recognized by the members as a form of social activity that they engage in.

Contexts of situation are characterized by Halliday (e.g. 1978: 61-62, 142-

145, 221-230; Halliday & Hasan 1985: 12-14, 56-69) in terms of three dimensions:

1) field, 2) tenor, and 3) mode, each of which can be simply glossed as 1) what is

Page 46: systematic functional linguistic

38

happening, 2) who is taking part, and 3) the role of language in the linguistic

contact. These dimensions will be discussed in more detail below. However, as the

notion of a situation type and various developments of Halliday’s ideas are not

central to this study, the discussion below is brief and superficial. For a more

comprehensive discussion, see Gregory 1967, 1988; Gregory & Carroll 1978,

Halliday 1978, Halliday & Hasan 1985, Ventola 1987, 1988.

Field characterizes the situation in terms of the social activity that is taking

place, e.g. making an appointment, visiting someone in hospital, playing a game

of ice hockey, discussing a game that has been played etc. In some cases, it may

be necessary to make a distinction between first-order and second-order fields

(Halliday 1978: 144). In a discussion about a game of ice hockey, for example, the

discussion itself, the verbal interaction, constitutes the social activity that is taking

place. It is the first-order field. The game that is being talked about constitutes the

second-order field.

Tenor is concerned with the various kinds of relationships that hold between

the participants who are involved in the situation either directly (e.g. two people

speaking or writing to each other) or indirectly (e.g. a writer and her or his

audience). Tenor, thus, encompasses such variables as the relative statuses of the

participants, i.e. the power relationships and the social distance between them, their

frequency of contact, the emotional relationship between them etc.

Mode is concerned with the place that is assigned to the text in the situation

and encompasses a number of variables, which gives rise to a distinction between

the spoken and written medium. For example, the semiotic distance between the

text and the social activity in which it is embedded can either be constitutive (e.g.

a lecture, discussion or scientific article) or ancillary (e.g. playing a game of

basketball or cards). The relationship between the text and the participants can be

seen in terms of whether the producer of a text can edit it before it is exposed to an

addressee (e.g. as in most written texts and in texts that are written to be spoken)

and whether the addressee can share in the process of text creation (as in casual

conversation). A third variable within mode is channel, which can either be phonic

or graphic.

Page 47: systematic functional linguistic

39

In a framework in which grammatical analysis is based on (abstractions from)

the situation, there is no need for a distinction between semantics and pragmatics

(Hasan MS; Halliday in Thibault 1987: 611). This distinction evolved in linguistics

from the late 1960s, when the generative semanticists began to assail Chomsky’s

standard theory (Levinson 1983: 4). Pragmatics came to refer to “those linguistic

investigations that make necessary reference to aspects of the context” (Levinson

1983: 5). Pragmatics was part of a tradition that tacked the situational context on

to a context-independent notion of language. It assumes that there are some kinds

of meanings that are not dependent on context, and these are handled under

semantics, and there are other kinds of meaning that relate these context-

independent meanings to the context in which language is used, and these

meanings are dealt with under pragmatics. In SF theory, the grammatical system

is based on the options that are available in language in use. These options involve

not only experiential meanings, but also interpersonal and textual meanings. This

means that, pace Leech (1983), experiential meanings are 1) context-dependent and

2) that they are no more integral to language as a system than interpersonal and

textual meanings (Hasan MS).

2.3.8. The Notion of Reference

A central notion in any linguistic analysis that needs to be discussed is the notion

of reference. In a tradition that goes back to Ogden and Richards (Lyons 1977a:

98), the term “referent” is applied to an object in the outside world to which a word

refers and “reference” to the “picking out” of something in an objective external

world. This notion of reference is clearly tied to a mapping of reality view of

language (see 2.3.5). In a reality construction view, on the other hand, the things

that we talk about when we use language are constructs of the language and the

culture that have made these things meaningful for us. I shall elaborate on this

shortly.

Page 48: systematic functional linguistic

40

In SF theory, however, while reference can be seen as a relation between

language and the social and material reality that makes our words meaningful, part

of this reality is language itself. A linguistic item can not only refer exophorically

to something outside language, it can also refer endophorically to another linguistic

item ) to a bit of text (cf. Larjavaara 1990, Chapter 3). Hence Halliday & Hasan

(1976), talk of reference as a cohesive relation in a text. Reference as a cohesive

relation will not be discussed; this section will briefly focus on reference as a

relation between language and the extralinguistic.

Rather than see reference in logical terms of the “things” that we refer to in

the external world, the notion of reference can be seen in terms of Saussure’s

signifié, the signified part of the linguistic sign (Saussure 1983 [1916], translation

by Roy Harris). A similar approach is taken by Hudson (1984: 138-139):

I am using the term [i.e. referent] to denote de Saussure’s ‘concept’ which is the signified partof a linguistic sign ... Of all the obviously available words for referring to this mental entity,the word referent seems the most suitable in spite of its traditional links with the outside world.However, the connections between the real world and linguistic expressions are quite indirectand not of any particular importance to a linguistic theory, so it seems preferable to adopt theterm “referent” for an area where we do need it, rather than to leave it for use on a hypotheticaloccasion.

In order to explicate Saussure’s notion of the signifié, the main points of his

approach will be outlined. Saussure saw a language as a system of arbitrary signs.

A sign consists of a relationship between two mutually dependent elements: a

sound pattern (image acoustique) and a concept (concept), which Saussure

referred to as the signifiant (‘signification’) and the signifié (‘signal’), respectively.

A linguistic sign is arbitrary in that a signal (signifiant) has no natural connection

with its signification (signifié), the signal which is written in Finnish as maa [ma:]

has no natural connection with what it signifies, and, consequently, a foreigner

could not guess its signification. (I shall give some translation equivalents shortly.)

Since the connection between a signal and its signification is not natural, it can

only be established by convention.

Page 49: systematic functional linguistic

41

To say this much about the arbitrariness of the linguistic sign, however, is

nothing new, and was nothing new even in Saussure’s time. This aspect of the

arbitrariness of the sign was accepted well before Saussure, and can be traced back

to the debate in ancient Greece between the naturalists and conventionalists

(Robins 1967: 17ff.). Saussure’s insistence on the arbitrariness had more far-

reaching repercussions: if a sign such as maa or [ma:] in Finnish, Chinese,

Gooniyandi, or English is arbitrary, then its signification, the concept with which

this sound pattern is associated, can only be determined by reference to other signs,

by its linguistic value (valeur):

In a language, there are only differences, and no positive terms. Whether we take thesignification or the signal, the language includes neither ideas nor sounds existing prior to thelinguistic system, but only conceptual and phonetic differences arising out of that system.(Saussure 1983: 118).

The conceptual part of linguistic value is determined solely by relations and differences withother signs in the language... (Saussure 1983: 116).

Signs are thus relationally defined. They cannot be determined by reference to pre-

existing ideas; they are not labels for pre-existing, naturally occurring categories.

One area in which we may need to invoke a philosophico-logical notion of

reference is when we begin to consider how a novice, an infant or a second-

language learner, breaks into a linguistic system. As Hasan (1985a: 24) points out,

Saussure’s treatment of signification and value is circular. The signification of a

sign with the sound pattern maa [ma:] in Finnish, for example, can only be

determined by relating it to other signs, for example, taivas and kaupunki, whose

significations can only be determined by their values. Similarly the value of maa

in Gooniyandi is determined by its value. But the value of the sign is determined

by relating it to other signs, for example, manyi and gawi, whose significations can

only be determined by their values. Anyone who does not know Finnish or

Gooniyandi is none the wiser. The Finnish or Gooniyandi infant learning the word

maa, however, does not learn it in a vacuum, but in a process that involves “an

active experience of the word in conjunction with reality within a culturally

recognizable context of situation” (Hasan 1985a: 31). This is how a novice breaks

Page 50: systematic functional linguistic

42

into a linguistic system: by making connections with other semiotic systems, which

include the concrete objects in the environment.

As for the second-language learner, if her or his native language is English,

then the impasse begins to be broken once she or he is told that the English

translation equivalents of Finnish maa include ‘earth, property, land, country’ and

in certain expressions ‘countryside (as opposed to the city)’, since the extra-

linguistic things to which these English items can refer are known to a native

speaker of English. The Gooniyandi examples are taken from McGregor (1990a:

594-596). I am not competent to comment on the relationships between the items,

but McGregor glosses maa as ‘meat’ and manyi as ‘vegetable food’ in the semantic

domain “food, cooking and fire” and gawi as ‘fish’ in the semantic domain “marine

life”.

2.4. Central Notions in SF Grammar

2.4.1. Introductory Remarks

The questions addressed in this section are more specific than those addressed in

the previous section. This section gives an overview of the main features and

principles underlying SF grammar, from the earliest work in the sixties to the

present. It is important to realize that there has been a gradual evolution from the

earliest work to the present grammar, and, thus, the grammatical concepts that were

articulated within the early framework have not been abandoned, but have become

recontextualized. Moreover, as the theory has expanded and developed, some

concepts have been given more prominence and new ones have been introduced.

Page 51: systematic functional linguistic

43

2.4.2. A Brief Historical Overview

The earliest version of SF grammar, which came to be known as scale and category

grammar, was developed during the late fifties and early sixties by Halliday (1961;

Halliday et al. 1964), who studied under Firth. Although the first published paper

on scale and category grammar appeared in the mid-fifties (Halliday 1956), the

version published in 1961 will form the basis of comments made in the following

sections, since this later version is more developed and provides a more compre-

hensive account than the earlier one. Scale and category grammar is known as such

because it sets up three “scales” of abstraction (i.e. abstract relations), which are

referred to as rank, delicacy, and exponence. The relata are referred to as

“categories”, and include units, structures, classes, and systems. Thus, units,

structures, classes, and systems are related to each other and to the data in terms of

rank, delicacy and exponence (or realization). These concepts will be discussed in

some detail throughout the following sections.

From about the mid-sixties (Halliday 1966a), more theoretical significance

became attached to the notion of system, to the paradigmatic sets of options that

generate structures. The system became the fundamental organizing concept in the

grammar; and, as a result, systemic grammar emerged out of the scale and category

model. Scale and category grammar had been primarily designed for the analysis

of texts, and while this orientation was not abandoned, the developments that

accompanied the name change brought with them an added goal, that of generative-

ness (Hudson 1974). In order to test the generative potential of the theory, some

systemicists started to look at computer applications of systemic theory (e.g.

Henrici 1981 [1966], Winograd 1972). Research into computer applications has

continued to the present (see e.g. Butler 1985: 206 ff., Matthiessen 1988a, 1988b,

Patten 1988).

Soon afterwards, Halliday (1968: 207 ff.) began to develop the idea that the

systems in the grammar of a language were organized as a number of components.

These components are related to the functions (later “metafunctions”, see 2.4.8)

Page 52: systematic functional linguistic

44

1 I use the term “plane” instead of “level”, because it is less concrete. Moreover, level issometimes confused with “rank” (see 2.4.4 below). Cf. Matthiessen & Halliday(forthcoming), who use the term “stratum” for similar reasons.

that language has evolved to serve. The metafunctions comprise the semantics of

a language, but, because there is a non-arbitrary relationship between semantics

and grammar, they are embodied in the grammatical organization of a language,

i.e. they shape the form of the grammar (Halliday in Thibault 1987: 608). With the

development and explication of the metafunctional organization of language,

systemic grammar also became known as systemic-functional grammar.

2.4.3. The Triplanar Organization of Language

Based on Hjelmslev (1953), Halliday (1978: 187; 1985b: 10,42) points out that

most elementary coding systems have two planes: a content plane and an

expression plane. For example, a cat might express the content or meaning ‘feed

me’ by a particular kind of miaow or an action that its owner recognizes. There is

a simple one-to-one correspondence between content and expression. Language,

however, differs in that it involves a third intermediate plane. This is the plane of

lexicogrammatical form, which includes grammar (syntax and morphology) and

lexis.1

Thus, language involves three planes of organization with a realization

relationship between them: a system of meanings, semantics, is realized as a

lexicogrammatical system, which, in turn, is realized phonologically as a system

of sounds in spoken language or a system of signs in sign language. (It can also be

realized as a system of written characters, graphology, which, in many languages

is based on the phonological system, however, for convenience, graphology will

henceforth be ignored.) This can be represented as follows; the downward-pointing

arrow is a symbol for ‘realizes/is realized by’. The downward pointing arrow used

to symbolize a realization relationship is somewhat misleading as it implies

directionality. A more appropriate symbol might be a double-headed arrow: a.

Page 53: systematic functional linguistic

45

However, directionality is also implied in the way in which we generally talk about

realization: “x is realized by y”. This directionality can be avoided by the rather

awkward expression “x and y are in a realization relationship”.

semanticsú

lexicogrammar- - - - ú - - - - - - -

phonology

The broken line in the above diagram is meant as an indication that the realization

relationship between phonology and the other planes of linguistic organization is

qualitatively different to the relationship between semantics and lexicogrammar.

The line indicates that the relationship is an arbitrary one. The relationship between

semantics and lexicogrammar, on the other hand, is not an arbitrary one: the

lexicogrammatical form of a language constructs the meanings that are available

in a language, and linguistic meanings could not be expressed except through the

lexicogrammatical form of a language (Halliday 1985a: xvii; Hasan & Cloran

1990).

These planes, which comprise the subject matter of linguistics, can be related

in terms of the form-substance distinction to what falls outside the scope of

linguistics and becomes the scope of phonetics: the phonological form of a

language is realized as phonic substance. Thus, there is a two-tier realization

relationship: 1) a realization relationship which relates form to form (i.e. between

the semantics, lexicogrammatical, and phonological form of a language) and 2) a

further realization relationship which relates form to substance, i.e. between

phonological form and phonic substance.

The above account can be combined with Halliday’s (1961) earlier model,

which brings in the form-substance distinction:

Figure 2-1: Planes in Language

Page 54: systematic functional linguistic

46

‹———————— LINGUISTICS ————————›‹——— PHONETICS ———›

SITUATION

extra-textualfeatures

‹————›

context

FORM

grammarlexis

‹————›

phonology

graphology

SUBSTANCE

phonicsubstance

graphicsubstance

This early model needs to be revised. Firstly, the terms semantics or contextual

semantics are now preferred to context. The word context was originally preferred

because semantics tends to be understood in conceptual terms, i.e. as a mental non-

observable phenomenon. Context, on the other hand, is seen as a link between the

lexical and grammatical forms of a language and abstractions from the actual

situations in which language is used (Halliday 1961: 245 footnote 13). Secondly,

systemic-functional linguists generally concur that meanings are realized on all

planes of language (cf. Firth in 2.2.5). Furthermore, the semantic and lexicogram-

matical resources of a language are now regarded as being organized

metafunctionally, the very broad and general functions which language has evolved

to serve are reflected in its semantic and lexicogrammatical organization (see

2.4.8).

2.4.4. Units and the Rank Scale

In scale and category grammar, a unit is a category that is set up to account for

stretches that carry grammatical patterns. The units of a language are related to

each other by reference to the scale of rank. The units originally set up for

English are: sentence, clause, group and phrase, word, and morpheme (Halliday

1961: 253). Later, the sentence was reinterpreted as a clause complex (see below),

which is not a basic unit on the rank scale. In present-day SF theory, the sentence

is not a grammatical unit but the orthographic unit that is “contained between full

stops” (Halliday 1985a: 193). It has become customary to mark a group or phrase

Figure 2-2: Levels in Language (Halliday 1961: 244)

Page 55: systematic functional linguistic

47

boundary by a single vertical line, a clause boundary by double vertical lines, and

a clause complex boundary by three vertical lines, as in the following example:

(1) /// When / Alice / began telling / them / about her adventures, // the White Rabbit /appeared / at the door. ///

The units are hierarchically ordered on a rank scale such that there is a constituency

relationship between a higher and a lower unit: clauses consist of groups or

phrases, which consist of words, which consist of morphemes.

SF grammar is, thus, a constituency grammar, but it differs from other

constituency analyses in that the constituents are ranked: each unit ) a clause, a

phrase or a group, and a word ) is potentially expandable into two or more units

at the rank below. This is generally referred to as the “rank hypothesis”. The

fundamental basis of the rank hypothesis is not that there are constituents that can

be ranked in terms of their size, but that these ranked constituents form a “strict

hierarchy” (Halliday 1985a: 25). The rank hypothesis allows for 1) singulary

branching, 2) rankshifting, and 3) the expansion of a unit at any rank to form

complexes and 4) the exclusion of certain elements from the rank scale.

Singulary branching means that a unit on the rank scale, say a group, can

consist of one unit on the rank below, the word, so that in a tree diagram the node

at the group rank does not branch into two separate branches, but continues as a

single branch. Each node is seen in terms of its potential for branching, and, thus,

an argument for singulary branching is that, in many instances where singulary

branching occurs, the unit is expandable (Lyne 1988: 62). For example, potatoes

in I’d prefer potatoes is a “simple” nominal group, but it is potentially expandable,

e.g. roast potatoes. While pronouns or proper names such as Alice are not

expandable, they are in paradigmatic contrast to other items (e.g. the little girl) that

are.

The phenomenon known as rankshifting, which is also referred to as

embedding (e.g. Halliday 1985a: 166), allows for a grammatical unit to function

as part of another unit at the same or at a lower rank. For example, a phrase

(example 2 below) or a clause (example 3) can be embedded in the structure of a

Page 56: systematic functional linguistic

48

nominal group. It has become customary to mark an embedded phrase with single

square brackets and an embedded clause with double bar square brackets.

(2) Just then her head struck against the roof [of the hall].(3) They would not remember the simple rules [[their friends had taught them]].

A unit on the rank scale can be expanded (by co-ordination or subordination)

to form a complex. Any unit is, thus, a potential point of expansion. Expanded

units are referred to as supplementary units and include: clause complexes,

group/phrase complexes, word complexes, and morpheme complexes (Huddleston

1965: 57). An example of a clause complex is:

(4) /// She opened it, // and found in it a very small cake, // on which the words ‘EATME’ were beautifully marked in currants. ///

Examples of a co-ordinative nominal group complex and a subordinative verbal

group complex are: the lobsters and the turtles and tried to reach, respectively.

An example of a (co-ordinative) word complex is more and more (in more and

more miserable) and a morpheme complex pro- and anti- (in pro- and anti-

abortionists).

As originally formulated, the theory of rank required that:

Each unit should be fully identifiable in description. This means that, if the description istextual, every item of the text is accounted for at all ranks ... If the description isexemplificatory, exactly the same is implied ... (Halliday 1961: 253). [Emphasis added.]

This became known as the “total accountability requirement”, and it became an

issue that was discussed within the scale and category framework (Halliday 1966a:

111). According to Halliday (1966a: 115-117) in reply to criticisms by Matthews

(1966), some elements were considered to fall outside the total accountability

requirement, i.e. they are excluded from the rank scale. These would include

logical elements such as and and or, which expand units into complexes, but do

not have constituency status. Another class of units which would have to be

excluded from the total accountability requirement are adjuncts like moreover in

Moreover he was late (Lyne 1988: 62). Presumably, these can be treated like

Page 57: systematic functional linguistic

49

logical elements in that they do not directly engage in the constituency structure

of the clause.

McGregor (1990b, 1991), on the other hand, suggests that constituency

relations are relevant only for experiential meanings, i.e. meanings that are related

to the way in which language constructs and reflects the world in which we live.

From this point of view, the ranked constituency hypotheses is valid only for the

experiential meanings. This view is also consistent with Halliday’s most recent

views on the structural organization of language (e.g. Halliday 1978: 188; 1981a:

35-38; Thibault 1987: 612), where constituency structure is seen as being typically

associated with experiential meaning (see 2.3.7 and 2.4.8). This view of the rank

hierarchy hypothesis will be adopted as a starting point for Finnish (in Chapter 4).

2.4.5. Types of Structure

Another “category” that dates from scale and category grammar but has been

progressively redefined as it has become recontextualized is structure. Originally

a structure was set up to account for the abstract syntagmatic patterning of units

(e.g. phrases or words). While it involved ordering and patterning, it did not

necessarily refer to sequentiality. Nevertheless, structure originally seems to have

been equated with discrete segmentable elements and to the way in which these are

organized. More recently, however, it has been used to refer to any clause-rank

organization or patterning, i.e. to phenomena that are not random (Halliday 1978:

188; 1981a: 35-38; 1984: 33; Thibault 1987: 612; Butt 1988).

A structure (i.e. a pattern) is seen as the realization of grammatical choices,

i.e. simultaneous structures realize choices from a number of concurrent systems.

From an experiential perspective, for example, there are a number of options at

clause-rank associated with process type: material, relational etc. Interpersonal

options include those associated with either giving or asking for information that

Page 58: systematic functional linguistic

50

are realized by declarative or interrogative structures. The following clause, for

example, realizes an interrogative and a mental process. (For form glosses, see

Appendix 3.)

(5) Mistä aineista sä pidät? [SIIIM3b:2]which+ELA subject+PL+ELA you(sg)-NOM like+2SG

Phenomenon Senser Process: mental‘Which subjects do you like?’

The experiential functions realized in the clause can be glossed as Senser0Pro-

cess0Phenomenon, where the raised dot indicates a concatenation of elements with

no implication of sequence. (Sequence, on the other hand, can be indicated by the

caret or hat symbol: ^.)

Grammatical structure at the clause level is generally thought of in terms of

constituency, in terms of discrete and easily segmentable units, particularly in

isolating languages such as English and Chinese. As mentioned earlier, Halliday

sees this type of structure as being typically associated with experiential meaning.

The discrete and segmentable nature of experiential structure is seen as a reflection

of the relative discreteness of the phenomena of our experience. If one considers

the clause cows eat grass in terms of its experiential meaning, then the words

cows, eat, and grass can be related to recognizable and relatively discrete things

or events in our experience. This is at least true at a simple and concrete level, but,

as Halliday (1981a: 36) points, many phenomena are, in fact, unbounded and

fuzzy. Following Whorf, he suggests that the discreteness and segmentability of

experiential structures predisposes us to think of the phenomena of our experience

in similar terms.

Logical meaning is seen as being typically characterized by recursion, i.e. by

repetition of the same variable. For example, the clause-initial nominal group in the

previous example can be co-ordinatively extended: cows and horses eat grass,

cows and horses and sheep eat grass. Structures that result from recursion are

referred to as univariate structures. The term “univariate” (i.e. involving one

variable) dates back to the earlier scale and category model (Halliday 1981b

Page 59: systematic functional linguistic

51

[1965]), where a distinction is made between univariate and multivariate structures

(i.e. involving a number of different variables, as with the experiential structure

exemplified in 5 above).

In contrast with experiential meaning, interpersonal and textual meaning do

not correspond to discrete, clearly delineated units. Interpersonal meaning, i.e.

speaker attitudes and assessments and role assignment, is seen as being typically

realized by prosodic patterns that run throughout the clause or by scopal relations

that affect the entire clause (see also 5.4.2). The clearest indication of prosodic

patterning is in intonation contours, which are typically associated with interper-

sonal meaning. (See, for example, Halliday in Kress (ed.) 1976: Ch.14) for English

and Hirvonen 1970 for Finnish.) The scope of a modal element is seldom restricted

to one discrete unit in the clause, and, furthermore, modal elements can be strung

throughout the clause reinforcing the effect of each other.

Textual meaning, according to Halliday, is typically realized by wave-like

movements associated with peaks of prominence in the clause. These peaks

typically occur at either clause-initial or clause-final position, and, consequently,

with a sequence of clause, there is a wave-like movement from one peak to the

next. In Halliday’s view, textual meaning in the English clause is realized by two

independent structures: Theme-Rheme and Given-New. In terms of the peaks of

prominence just discussed, Theme constitutes one peak while New constitutes

another peak, with the rest of the clause rising or falling towards these peaks.

2.4.6. System

The notion of a system was first introduced in 2.2.2 in the discussion on Firth.

Whereas a structure is concerned with syntagmatic relations, amongst elements in

praesentia; system is concerned with paradigmatic relations, amongst elements

in absentia (Halliday 1981b [1965]: 124). A system accounts for “the occurrence

Page 60: systematic functional linguistic

52

1 See appendix 6 for other notational conventions.

of one rather than another from a number of like events” (Halliday 1961: 264).

These paradigmatic choices are represented as networks.

As mentioned earlier, system and structure were given an equal status in early

scale and category grammar. From about the mid-sixties, the system became the

central organizing concept in grammatical description; structures were seen as

secondary, as the output of a system or a number of systems. The grammatical

organization of a language came to be represented as a vast number of intercon-

nected and simultaneous systems. The following network (which is grossly over-

simplified for the purpose of illustration), combines mood, transitivity, theme and

information focus networks. The fact that the options are simultaneous is indicated

in systemic notation by curly brackets.1 (The examples given in the following table

(and in the above table) are meant only as an indication of the type of option

involved. A clause does not realize only one option but conflates a number of

simultaneous options (see also 2.4.8)).

Figure 2-3: Simultaneous Paradigmatic Options in English

Page 61: systematic functional linguistic

53

1 For some discussion of issues related to the formalization of options in systems networks,see Martin 1987 and Fawcett 1988.

These options can be further refined and extended.1 As Figure 2-3 illustrates, at a

primary degree of delicacy, systems networks tend to be taxonomic. It is only when

the degree of delicacy is increased and the networks become more complicated that

their efficacy and application in text generation, for example, becomes more

apparent.

When systemic-functional linguists model the grammatical organization of

a language, they generally give priority to the systems and the grammatical

structures are seen as being generated by these systems. From an ontological

perspective, however, neither system nor structure can be given priority. We cannot

have one without the other:

There is no way in which a structure is first described and then by a separate step brought intoparadigmatic relation with other structures. A description is a statement of paradigmaticrelationships. (Halliday 1984: 6-7.)

This dual perspective is needed from the start: we can only describe a structure by

relating it to other structures. Methodologically, however, it is easier to start by

describing a structure, rather than by modelling the networks that have generated

the structure. Structures are more tangible; networks are more hypothetical. As

Halliday puts it (1985a: xxvii): “structures are less abstract; they are so to speak

nearer the text”).

From the discussion of the notions of system and structure in this section and

the previous section, it should be clear that the grammatical organization of a

language is not seen as a structure in SF theory ) either as a monolithic superstruc-

ture or as a number of clause structures. A language is seen as a resource or

potential, which, from a grammatical perspective, embodies lexical and grammati-

cal choices: a language is a meaning-making resource.

Page 62: systematic functional linguistic

54

1 Finnish: hienojakoisuus, tarkennus.

2.4.7. Delicacy and Realization

In section 2.4.4, the scale of rank was discussed. The other scales are delicacy and

realization (earlier referred to as exponence). The term “scale” is used in a slightly

different sense with reference to each of these. The scale of rank is referred to as

a scale because the units are hierarchically ordered: clause, group/phrase, word, and

morpheme. Delicacy1 refers to the degree of detail or differentiation that is made

at a particular rank. It is a scale in the sense of a cline: one can start by making very

broad and general distinctions in a system and then refine the analysis by making

finer, more specific distinctions. For example, at the primary degree of delicacy at

the rank of clause in Finnish grammar one might want to distinguish between

material and relational processes. At a greater degree of delicacy, one might then

want to distinguish between intensive and circumstantial relational processes, and

then distinguish between identifying and attributive relational processes, and so on.

Realization is referred to as a scale in that it involves successive steps along

a scale of abstraction which relates categories to each other and to the data. For

example, the option interrogative can be realized in English by the structure Finite

^ Subject. The element Subject is generally realized by a nominal group, which is

realized by a (M) H (Q) structure (i.e. the Head, which is possibly preceded by

a Modifier and followed by a Qualifier), which, in turn, is realized by certain

classes of words. The classes are then realized by the lexical items, for example,

“the”, “man”, “outside”. So far, I have referred to what Halliday and Martin (1981:

343) refer to as “intrastratal” realization relation (i.e. between systems, structures,

and units at the lexicogrammatical stratum or plane). As indicated in 2.4.3 (p. 45),

however, it can also refer to an interstratal relation: the lexical items mentioned

above are realized phonologically: /ðM mæn autsaid/. (There is also a realization

relationship between phonology and phonetics, i.e. between form and substance

(see Figure 2-2, p. 46).)

Page 63: systematic functional linguistic

55

1 See Karvonen (1991, 1992) for a discussion of grammatical metaphor in Finnish.

The notion of realization in SF theory, however, must not be misunderstood

in terms of the folk linguistic notion of expression. It is erroneous to equate the

linguistic statement “X is realized by Y” as “X is expressed by Y”. For many, the

latter statement implies that X exists prior to Y. For example, in many studies of

definiteness and indefiniteness or of given and new, it seems to be assumed that

definiteness or givenness are prelinguistic (or extralinguistic) notions (see the

discussion of Brown and Yule in 7.2.). In the SF view of realization, X does not

exist without Y and Y does not exist without X. Thus, while in natural language

there is always a directional metaphor, for example, linguists often talk about

semantics being “above” the grammar, and systemic-functional linguists talk of the

context of situation and culture being “above” the semantics, semantics and

grammar are not related to each other in causal terms, but in dialectic and

realizational terms. An attempt to capture the notion of realization is made by

referring to it in terms of “construal”: X is construed by Y. (For further discussion

of realization as a dialectic, see Halliday (in press), Matthiessen & Halliday

(forthcoming, section 1.2.1).)

This realization relationship cannot, moreover, be seen in terms of a simple

one-to-one correspondence: for example, an interrogative need not necessarily

realize a question, or, it could realize a question at one level, but it functions

rhetorically as a request (e.g. Would you turn the radio on?). Halliday extends the

notion of metaphor to grammatical phenomena, and regards examples like these as

types of grammatical metaphor (Halliday 1985a: 319 ff.; Ravelli 1988).1 A

distinction is made between a “congruent” (or literal) and a “metaphorical”

realization. Congruency is often glossed in terms of typicality, but it seems to me

that it is more appropriate to think of a congruent realization in terms of Halliday’s

(1984: 14) alternative suggestion as “a kind of baseline” or as a “maximally

simple” way of expressing things, for example, as the way in which young children

typically express things. As with lexical metaphors, e.g. the leg of a table,

Page 64: systematic functional linguistic

56

1 While metafunctions are sometimes referred to as part of the lexicogrammatical system ofa language, this is not strictly true; however, statements like this are sometimes made becauseof the non-arbitrary relationship between grammar and semantics (Thibault 1987: 608). Inthis regard, Ellis (1987) makes a consistent distinction between function (i.e. metafunction)and functional component.

grammatical metaphors also retain their literal or congruent meaning. (See also

5.2.2 and 6.6.)

2.4.8. Metafunctions

The notion of a metafunction was introduced 2.3.7. Metafunctions are seen as

being crucial to the semantic organization of language, and, since there is a non-

arbitrary relationship between semantics and grammar, they are reflected in the

lexicogrammatical organization of language.1 The lexicogrammatical plane

involves a “conflation”, a coming together, of different modes of meaning.

However, systemic-functional linguists vary on the number of metafunctions that

they recognize. According to Halliday (1978: 50; 1981a: 34; 1985: xiii), there are

three, with some further subdivisions; Fawcett (1980: 26ff.) recognizes eight major

core metafunctions (which he refers to as “functional components”) and two or

three minor ones. Figure 2-4 on page 58 below is a revised and simplified version

of a diagram which appears in Fawcett (1980: 28).

It is important to note that the examples given in Figure 2-4 are meant as an

indication of the type of meaning that is involved: they should not be misunder-

stood as implying that each clause (or linguistic unit) has only one function. This

position seems to be taken by Traugott (1982), whose model of linguistic change

is based on Halliday & Hasan (1976). Nevertheless, it is clear from Traugott’s

discussion that her model is not Hallidayan. For example, her “propositional

component” is explicitly based on truth-conditional semantics. As Halliday (1985b:

23) points out (in the context of analysing a written text):

Page 65: systematic functional linguistic

57

We cannot pick out one word or one phrase and say this has only experiential meaning, or thishas only interpersonal meaning ... Every sentence in a text is multifunctional; but not in sucha way that you can point to one particular constituent or segment and say this segment has justthis function. The meanings are woven together in a very dense fabric in such a way that, tounderstand them, we do not look separately at its different parts; rather we look at the wholething simultaneously from a number of different angles, each perspective contributing towardsthe total interpretation.

Thus, essential to an SF approach is the analysis of language from different

semantic perspectives, i.e. from the perspective of the various metafunctions.

Moreover, grammatical description is organized on the basis of these metafunc-

tions, not on the basis of forms. This does not, however, mean that formal and

structural properties are ignored (see further 2.4.10).

The main difference between Figure 2-4 and Fawcett’s original diagram

concerns Fawcett’s view of how his minor functional components (marked with the

superscript “m” in Figure 2-4) relate to the metafunctions postulated by Halliday.

Fawcett puts a question mark in the Halliday column indicating that he is not sure

how these would fit into Halliday’s schema. It seems to me that Fawcett’s

inferential (he even fell down) and metalingual (he fell down, as it were) would

be considered interpersonal in Halliday’s schema as they are concerned with

speaker attitudes and judgments. The discourse organizational (firstly, he fell

down) seems to belong to the textual metafunction, as indicated in Figure 2-4. (Cf.

Gregory (1987: 100), who also regards Fawcett’s metalingual as falling within

Halliday’s interpersonal; he sees Fawcett’s discourse organizational as being

related to matters of textual cohesion, textual structure and register.)

Page 66: systematic functional linguistic

58

Halliday’s Metafunctions

ideational(experiential &logical)

interpersonal

textual

Fawcett’s FunctionalComponents

experiential

logical

negativity

interactional

affective

modality

inferentialm

metalingualm

thematic

informational

discourseorganizationalm?

Examples of Realization atClause Rank

he fell downhe was pushed down

he fell downif he fell down

he fell downhe didn’t fall down

he fell downdid he fall down?

he fell downunfortunately he fell down

he fell downhe may have fell down

he fell downhe even fell down

he fell downhe fell down, as it were

he fell downdown he fell

he fell downyes

he fell downfirstly, he fell down

As can be seen from Figure 2-4, Fawcett’s and Halliday’s analyses differ in

that Fawcett separates out each metafunction into a number of distinct functional

Figure 2-4: Metafunctions

Page 67: systematic functional linguistic

59

components. As Gregory (1987: 100) suggests, this could be seen as an increase

in delicacy, which could be useful for certain purposes and preoccupations. From

this point of view, Halliday’s and Fawcett’s positions are not incompatible. How-

ever, there is an essential difference in the way in which Halliday and Fawcett

conceive of their metafunctions or functional components. Fawcett (1980: 26)

sees his functional components as different “types of meaning that are reflected in

the organization of language”. Halliday (1978: 143; Thibault 1987: 608), on the

other hand, sees the metafunctions as an interface between the lexicogrammatical

organization of language and the contexts in which language functions, as out-

lined in section 2.3.7.

Thus, Halliday’s position is based on relating the linguistic system to the

context of situation. However, the hypothetical nature of Halliday’s analysis of

both of these should not be forgotten: the analysis of the linguistic system in

terms of three metafunctions is a hypothesis about the linguistic system and the

analysis of the context of situation in terms of three dimensions is a hypothesis

about the semiotic structure of the situation. While Fawcett has come up with an

alternative metafunctional analysis of the linguistic system, others (e.g. Hymes

1986) have put forward other variables for the context of situation. There is no

apriori reason to assume that there are just three dimensions in the linguistic sys-

tem and just three in the context of situation.

Problems arise when we begin to consider how these hypotheses can be

tested. It would appear that there are three separate hypotheses that need to be

tested:

(1) the metafunctional hypothesis: is it the case that systemic relationships (i.e.

the paradigmatic options available in a language) tend to fall into relatively

independent sets and that each set more or less corresponds to a different func-

tion of language (see Halliday 1973: 110);

(2) the second hypothesis concerns the tripartite analysis of the context of situ-

ation (i.e. in terms of field, tenor, and mode);

Page 68: systematic functional linguistic

60

(3) the hypothesis about the relationship between these two: the contextual

variables of field, tenor and mode are regarded by Halliday (1978: 116-

117,142-145) as being typically associated with a particular metafunction.

Whether or not these hypotheses are feasible has never really been put to the test.

Gregory (1987: 104) suggests that an attempt to prove the existence of the

metafunctions could be seen as falling into the trap of scientism. He suggests that

they are “a means of tackling the manifold complexity of meaning” (cf. Firth in

2.2.5) and that we should continue using them as organizing and investigative

concepts in linguistic analysis.

2.4.9. An Integrated Lexicogrammar

In SF theory, grammar is usually understood as including lexis. Grammar and

lexis are seen as being part of a continuum. As Halliday and Hasan (1976: 5) put

it:

There is no hard-and-fast division between vocabulary and grammar; the guiding principle inlanguage is that the more general meanings are expressed through the grammar, and the morespecific meanings through the vocabulary.

The relationship between grammar and lexis is seen as one of delicacy. This

approach was first suggested within the framework of scale and category gram-

mar by Halliday (1961: 267):

The grammarian’s dream is (and must be, such is the nature of grammar) of constant territo-rial expansion. He would like to turn the whole of linguistic form into grammar, hoping toshow that lexis can be defined as “most delicate grammar”.

However, for work like this to be viable, a vast amount of statistical analysis

needs to be done, and, as Halliday (1961: 267) pointed out over three decades

ago, serious statistical work in linguistics had hardly begun. Because of this and

because of the difficulty of distinguishing lexical items using grammatical crite-

ria, scale and category grammar saw lexis in terms of open sets.

Page 69: systematic functional linguistic

61

1 It is customary to write the name of an option in square brackets when it appears in thebody of a text; brackets are not used when it appears in the systems network.

2 For notational conventions, see Appendix 6.

In spite of the increased use of mainframe computers in linguistics today, no

work in lexis using statistical evidence has been done in SF linguistics. However,

a study that relies on the linguist’s own observations and intuitions has been done

by Hasan (1987b). This will be used to further illustrate the notion of “lexis as

most delicate grammar” and to provide another example of a systems network.

Some of the details of Hasan’s analysis will be simplified for the purposes of il-

lustration.

Hasan (1987b) has described how in a transitivity system for English if the

option [material process]1 is chosen, then a choice between sub-types of [action]

and [behaviour] processes is available. The following diagram, based on Hasan

(1987b: 186), illustrates these choices.2

If [action] is chosen, then there is a simultaneous choice in two systems. There is

a choice in the systems network which Hasan labels [ACT], which is a gloss for

the type of action, and, at the same time, there is a choice that Hasan labels

Figure 2-5: Transitivity Options in English

Page 70: systematic functional linguistic

62

[BENEFACTION], which is a gloss for whether or not someone benefits from the

action. The following diagram (based on Hasan 1987b: 189) follows on from the

systems network above, and focuses on a sub-type of [ACT], which Hasan labels

[disposal], which can be described as actions that result in the gain or loss of ac-

cess to things, i.e. actions that are realized by such lexical items as give, share,

collect etc.

The system [ACCESS] involves a choice between whether someone loses or

gains access to something, i.e. whether she or he is taking or giving something;

the label [CHARACTER] is a gloss for the system which involves a choice between

whether or not an action is inherently iterative. Hasan shows how by taking the

combination [deprivation] and [iterative] and by postulating further sub-systems,

features can be established to distinguish between such lexical items as scatter,

strew, and spill.

It is clear, however, that this type of analysis ) in which we proceed from

general to specific meanings ) is most readily applicable to experiential mean-

ings; it is questionable whether it is viable for the other metafunctions. Another

question that needs to be addressed is whether the above analysis does, in fact,

proceed from grammar to lexis, i.e. from general to specific meanings. Crucial to

the above analysis is the initial classification of action types (e.g disposal, trans-

form, locomotion ... ). It is not altogether clear that these can these be established

on the basis of grammatical criteria. If they cannot be established on the basis of

Figure 2-6: Action Options in English (Continuation of Fig. 2-5 above)

Page 71: systematic functional linguistic

63

grammatical criteria, then the notion of lexis as most delicate grammar is mis-

leading.

2.4.10. The Principle of Grammaticalization

The categories employed in grammatical description are not simply set up to label

differences in meaning. Any semantic distinction that is set up must be systemati-

cally reflected in the grammatical organization of a language:

We do not argue: “these two sets of examples differ in meaning; therefore they must be sys-tematically distinct in the grammar”. They may be; but if there is no lexicogrammatical reflexof the distinction they are not. If we simply took account of differences in meaning, then anyset of clauses and phrases could be classified in all kinds of different ways; there would be noway of preferring one scheme over another. The fact that this is a ‘functional’ grammarmeans that it is based on meaning; but the fact that it is grammar means that it is an interpre-tation of linguistic forms. Every distinction that is recognised in the grammar ... makes somecontribution to the form of the wording. Often it will be a very indirect one, but it will besomewhere in the picture. (Halliday 1985a: xx.)

As Halliday points out, the lexicogrammatical reflex of a semantic distinction is

not to be seen in terms of a simplistic one-to-one correspondence between

meaning and form or that the only formal distinctions that count are overt and

self-evident (such as case-endings in Finnish); meaningful distinctions can be

reflected in the grammatical organization in less overt ways (cf. Whorf’s discus-

sion of gender in English and covert categories in Carroll (ed.) 1956: 68 ff, 89).

2.4.11. The Ineffability of Grammatical Categories

As pointed out in section 2.2.4, Firth talked of the “ineffability” of grammatical

categories. This topic has been taken up again by Halliday (1988a). A category is

ineffable in that it eludes our conscious attempts to define it. Grammatical catego-

ries, whether they are functional or formal (Subject, Actor, Goal, Theme, Given,

noun, passive, present, modal etc.) have not been consciously designed since the

Page 72: systematic functional linguistic

64

1 Cf. Hockett’s (1968) critique of Chomskyan linguistics in which he points out that thegrammar of a language is not well-defined, but ill-defined.

grammar of a language is not a designed system, but an evolved one.1 As Halli-

day (1988a: 42) puts it:

Designed systems are designed so as to be effable; in fact, effability is a necessary conditionof design. You cannot design unless the principle can be made explicit. But a language is anevolved system; and evolved systems rest on principles that are ineffable ) because they donot correspond to any consciously accessible categorization of our experience.

Grammatical categories are ineffable because there are no counterparts to Subject,

Actor, Theme etc. in the outside world. There is no counterpart to a typical gram-

matical category, say Actor, because languages are not maps of a pre-existing

model of reality:

If language was a purely passive partner, ‘expressing’ a ‘reality’ that was already there, itscategories would be eminently glossable. But it is not. Language is an active participant inthe semogenic process. Language creates reality ) and therefore its categories of contentcannot be defined, since we could define them only by relating them to some pre-existingmodel of experience, and there is no model of experience until the linguistic categories arethere to model it. The only meaning of Subject is the meaning that has evolved along with thecategory itself. (Halliday 1988a: 39.) [Emphasis added.]

Similarly, Whorf (in Carroll (ed.) 1956: 92) says that “grammatical categories

represent experience ... but experience seen in terms of a definite linguistic

scheme”.

Functional labels such as Actor or Possessor can be thought of as mnemonic

devices, capturing some core meaning. However, while the labels are meant to

refer to core instances (Halliday 1985a: 106), this does not mean, as Huddleston

(1988: 164) assumes, that the meanings are “intended to apply only to core in-

stances”. The meanings are ineffable. While it may be impossible to find an ade-

quate gloss for a grammatical category, to exhaustively define it, this does not

make a category devoid of meaning.

Page 73: systematic functional linguistic

65

1 Cf. Firth’s discussion (in Palmer (ed.) 1968: 46-47) of indeterminacy in language andlinguistic description.

2.4.12. Prototypes

SF theory (and its predecessor scale and category grammar) has never been based

on the notion of categorial grammatical categories, as is clear from Halliday’s

(1961: 254) earliest account of the theory, where he discusses categories in terms

of likeness, which is seen as a cline ranging from “having everything in common”

to “having nothing in common”.1 In spite of Halliday’s early recognition of the

importance of not seeing linguistic phenomena in terms of categorial categories,

and in spite of the fact that concepts clearly related to the notion of a prototype

have been put forward during the past century (see Karlsson 1983a, Givón 1986),

it was not until the work of cognitive psychologists such as Rosch that other

schools of linguistic thought began to discuss linguistic categories in terms of

prototypes (e.g. Lakoff 1973, 1977, 1982, 1987; Ross 1972, 1973, 1974; Lakoff

& Johnson 1980; Givón 1982, 1986; Karlsson 1983a).

The notion of a prototype is offered as an alternative to the Platonic view

that categories are discrete and absolute, with no overlap between them, and that

the members of a category can be determined on the basis of a certain number of

necessary and sufficient properties. With prototypes, on the other hand, members

of a category are considered to be more or less representative of the particular

category. Rather than having certain necessary and sufficient criterial properties,

a category can be defined in terms of characteristic or typical features or proper-

ties. In this approach, properties may be of unequal importance in determining

category membership.

In linguistics, the notion of a prototype has been extended from the analysis

of lexical items to grammatical phenomena. Analysis in terms of prototypes is

evident in the analysis of process types in Chapter 6. Not all subtypes of rela-

tional processes, for example, possess all of the properties of the superordinate

process type.

Page 74: systematic functional linguistic

66

With Prototype Theory, however, the moot point is the basis that we use for

establishing a prototype. Karlsson (1983a), although aware of this inherent prob-

lem, seems to put a lot of store on the quantitative analysis of computer corpora.

As most of the computer corpora in Finland consists of either various genres of

standardized written Finnish (from newspapers, magazines and novels) or dialect

interviews (i.e. interviews between a linguist and a speaker of a dialect), it is

worth considering whether the available corpora represent prototypical genres of

Finnish. If we are to invoke quantitative empirical evidence in support of a proto-

type, then it seems to me that our data should consist of what we consider to be

prototypical genres, i.e. genres with which all (or almost all) speakers of the lan-

guage come into contact in everyday life (see further Ch. 5, Shore 1991b). Other-

wise we should restrict our claims to the genres that are represented in our analy-

ses.

2.4.13. Grammatical Proportionalities

Any formal or functional label used in linguistic description is a convenient way

of referring to what Halliday refers to as a grammatical proportionality (1985a:

xxxii-xxxiii). A grammatical proportionality is the “main heuristic technique” in

linguistic analysis: proportional relations are postulated between a number of

pairs of items:

a : b :: c : d :: e : f etc.

Thus, a and b differ in some respect, and there is a similar difference between c

and d and between e and f.

The notion of “agnateness” is important in the establishment of

proportionalities. Clauses that differ from each other in a particular respect are

said to be agnate clauses. For example, an active clause in English can be related

to an agnate passive clause. When clauses are said to be agnate, it is not implied

that one is derived from the other or that there is a transformational relationship

between them.

Page 75: systematic functional linguistic

67

The notion of grammatical proportionality underscores E. Itkonen’s point

that knowledge of language is epistemologically prior in linguistic investigations

(2.3.4). A proportionality is not something that can be proved by finding text

examples that vary in some respect, they can only be based on our knowledge of

systematic distinctions between linguistic items. Underlying the notion of gram-

matical proportionality is, of course, the classical notion of analogy, which fell

into disrepute in the 1960s with the onset of transformational-generative grammar

(see Anttila 1977). E. Itkonen (1991) has recently brought the issue of analogy vs.

innate knowledge of language into focus again. Following Anttila (and others),

Itkonen argues that we are not born with innate knowledge of language but rather

with an innate capacity to analogize.

In his discussion of analogy, Anttila does not appear to make a distinction

between analogy and metaphor, and, indeed, as Givón (1986: 100) points out:

These two names for the same phenomenon come into linguistics from two separate tradi-tions. The metaphor term comes from the literary analysis tradition, the analogy term fromthe philosophical tradition, most recently via Kant and Peirce.

While these terms are similar, whether in fact they are identical is questionable. In

SF theory, at least, while analogy and metaphor are both pervasive meaning-mak-

ing resources in language, analogy is more basic, unlike metaphor it cannot be

related to a more basic (linguistic) analogy.

2.4.14. Synoptic vs. Dynamic

A distinction is made in SF theory between dynamic and synoptic perspectives in

linguistic analysis (Halliday 1985b: 97 ff.; Martin 1988: 243-244). From a dy-

namic perspective, grammatical options are seen as a process, produced in a step

by step temporal sequence. From a synoptic perspective, the larger structures that

result from these choices are viewed as a product. Both of these perspectives are

relevant ) and necessary ) to the description of grammar and to the description of

discourse.

Page 76: systematic functional linguistic

68

In SF grammar, multivariate structures (e.g. the experiential structure of the

clause) are analysed synoptically and univariate ones (e.g. clause complexes)

dynamically. While it would also be feasible to analyse multivariate structures

dynamically ) focusing, for example, on how certain choices affect or determine

subsequent choices ) there are theoretical reasons for giving priority to a synoptic

perspective in the analysis of multivariate structures. The clause itself is regarded

as a meaningful unit: it realizes a conglomeration of functions. In the case of ex-

periential functions, for example, the function of a particular NP cannot be deter-

mined until the whole conglomeration of functions is realized.

As described above, the notions “synoptic” and “dynamic” are related to

time and temporal sequence. This is how these notions are employed in this

study. However, from a deeper and more philosophical perspective, linguistic

description can only be synoptic. In the first place, we cannot analyse something

unless we freeze it. As Harris (1980: 16) points out, in his discussion of the status

of spoken and written language in linguistics:

The systematic analysis of spoken languages depends essentially on their conceptualization assystems amenable to representation in a medium other than sound. (Harris 1980: 16.)

In the second place, the analysis of language ) or any other phenomena ) is nec-

essarily synoptic since any analysis must hold steady the dynamic flow of lan-

guage; it necessarily involves a synoptic perspective on the dynamic ) a frozen

representation of it. The linguist has to abstract and generalize from a constant

flux in which ) in the final analysis ) nothing is repeated. As Firth (1957: 190)

put it “each word when used in a new context is a new word”. (See further, Hasan

(in press).)

Page 77: systematic functional linguistic

69

Chapter 3

The Finnish Language

3.1. Overview

This chapter is meant to serve as a brief introduction to the linguistic situation

in Finland and to the Finnish language. In the section 3.2, there is a brief dis-

cussion of languages that are related to Finnish and of other languages that are

spoken in Finland. This is followed by an outline of the dialects of Finnish and

a brief discussion of the development of standard Finnish. The following sec-

tion, section 3.3, gives an outline of some of the formal characteristics of Finn-

ish. This section is meant simply as a convenience for those readers who do not

have access to grammatical outlines of Finnish. The information on case-forms

and verb inflexions is meant simply as a guide: the details can be glossed over

and, if necessary, reference can be made to this chapter when reading later

chapters of this study. The final section (3.4) is devoted to a number of impor-

tant issues in the received view of Finnish grammar. The received notion of the

accusative and the notion of subject will be discussed and criticized. This will

be followed by a sub-section devoted to boundedness, which is an important

semantic and grammatical feature of Finnish.

3.2. Background Information

Although Finland is often referred to as part of Scandinavia, Finnish is not a

Scandinavian language. In fact, it is one of the few languages spoken in Europe

that is not an Indo-European language. It belongs to the Finno-Ugric group of

languages, which are part of the Uralic family of languages. Finnish is closely

related to Estonian and to some other languages spoken in the region of the

Baltic Sea, including Karelian, Vepsian, Izhorian, Votian, and Livonian. Be-

cause of their proximity to the Baltic Sea, these languages are referred to as the

Baltic-Finnic languages. The Baltic-Finnic languages are related to the Saame

(Lappish) languages and more distantly to Mordvinian, Mari, Udmurt, and

Page 78: systematic functional linguistic

70

Komi, which are spoken in Russia. The languages mentioned so far comprise

the Finnic branch of the Finno-Ugric languages. On the Ugric side are Mansi

(also referred to as Vogul) and Khanty (also referred to as Ostyak), spoken on

the eastern side of the Ural Mountains, and Hungarian. The Finno-Ugric lan-

guages and the Samoyed languages, which are spoken in the north-eastern re-

gions of Russia, together form the Uralic family of languages. (For more detail,

see, e.g., Austerlitz 1987, Branch 1987.)

Finnish is spoken by some 4.7 million speakers living in Finland. It is also

spoken in countries to which Finns have emigrated, the most important of

which are Sweden, the United States, Canada, and Australia. There is a small

minority of Finns living in Finland who do not have Finnish as their first lan-

guage. The largest established minority is comprised of Swedish-speaking

Finns, whose linguistic status has been legally reinforced: the Language Edict

of 1922 enforces the status of Swedish as the second official language in Fin-

land. According to the Finnish Bureau of Statistics (quoted by Nuolijärvi 1991),

there are about 300 000 Swedish-speaking Finns (6.2 % of the population) in

Finland. Another significant indigenous minority is made up of speakers of

Sign Language. There are 8 000 deaf people in Finland, and of these approxi-

mately 5 000 use Sign Language, either Finnish Sign Language or signed Finn-

ish. Three Saame (Lappish) languages are spoken in Finland: North Saame,

Inari Saame and Skolt Saame, with about 2 000 speakers altogether. There are

approximately 6 000 Romanies (Gypsies) in Finland, but the active use of the

Romany language in everyday interaction has all but died out. As well as these

indigenous minorities, there are about 53 000 people in Finland who were not

born in Finland. These include descendants of Finnish emigrants who have re-

turned to Finland, immigrants or foreigners temporarily resident in Finland, and

refugees.

As the total number of people who speak a language other than Finnish as

their first language is no more than approximately 7% of the population, and

almost all of the members of the established minority groups speak Finnish (or

signed Finnish) as their second language, it is clear that Finnish clearly domi-

nates the language scene in Finland. The Finnish spoken in Finland comprises a

Page 79: systematic functional linguistic

71

Map 1: Finnish Dialect Groupings

number of dialects, all of which are mutually intelligible. The main dialect

groupings in Finland are shown on the map below.

Page 80: systematic functional linguistic

72

Although these groupings are based on the situation prior to World War II,

they are still used by Finnish linguists today. They are used, for example, in the

Finnish Dialect Dictionary (Suomen murteiden sanakirja). As indicated on the

map, a broad division is generally made between the eastern and western dia-

lects. The so-called far northern dialects are usually grouped with the western

dialects. Swedish is spoken mainly along the coastal areas and Saame in the far

north of Finland.

Both the eastern and western dialects have contributed to standard written

Finnish (kirjakieli literally ‘book language’). Standardized spoken Finnish is

based on kirjakieli: what approximates as a spoken form of kirjakieli is consid-

ered to be standardized spoken Finnish (i.e. the Finnish heard on radio and tele-

vision news and current affairs broadcasts and in similar formal and public

situations). Written Finnish is very phonemic, so kirjakieli serves as a guide to

“correct pronunciation”, and when children learn to read and write Finnish, they

also learn the basis of standardized spoken Finnish.

To understand the dialect basis of standard written Finnish, one has to go

back to the origins of literary Finnish and take a brief look at the history of Fin-

land. Finland was under Swedish rule until 1809, when it became an Autono-

mous Grand Duchy under the Russian Empire. It became an independent re-

public in 1917. While Finland was still under Swedish rule and for some time

afterwards, about the only texts written in Finnish were a handful of religious

and legal ones. The earliest of these texts date back to the middle of the 16th

century: the first translation of the New Testament appeared in 1548, and other

texts soon followed. These early texts and translations were the work of Mikael

Agricola, a Finnish priest who attended university in Wittenberg and was later

appointed bishop of the diocese of Turku. Agricola is considered to be the

founder of literary Finnish. The Finnish in Agricola’s texts is based mainly on

the dialects of the south-west Finland (near Turku). However, in some respects

it is quite stilted: the influence of the languages from which he translated is

apparent.

Page 81: systematic functional linguistic

73

It was not until the 1870s (when Finland was an Autonomous Grand

Duchy in the Russian Empire) that Finnish began to develop as a national and

literary language. Finnish scholars began to consciously develop the resources

of their languages so that Finland could develop as an advanced industrial na-

tion. This meant the development inter alia of various scientific and political

genres, and thus a lot of new words and scientific terms were coined. Some of

these words were taken from dialects other than those used by Agricola in his

writings. As well as this lexical input, other dialects increasingly began to influ-

ence the grammar and morphology of the standard written language. (For more

detail, see Korhonen 1986: 67ff.)

Thus, standard Finnish is historically an amalgam of dialects and not

based exclusively on any one particular dialect. Nowadays, however, it would

be fair to say that it is most strongly influenced by the Finnish spoken in Hel-

sinki and in southern Finland generally, as this is the area in which the majority

of people live. It would also be fair to say that this standard is accepted by most

Finns as it provides access to literacy and education and, moreover, it is not as

associated, at least overtly, with such factors as class or race, as is the case with

some other standard languages in the world.

The status of standardized Finnish in schools and elsewhere is reinforced

by the fact that language standardization is institutionalized in Finland. This

standardization is carried out by the Language Board (Kielilautakunta) and ad-

vice on what is considered to be correct Finnish is given by a certain sections of

the Research Centre for Domestic Languages (Kotimaisten kielten tutkimuskes-

kus). Unlike the French Academy, however, neither the Language Board nor

the Research Centre for Domestic Languages has legal jurisdiction.

Page 82: systematic functional linguistic

74

3.3. Some General Characteristics of Finnish

3.3.1. General Remarks

As indicated earlier, this section is included simply as a convenience for those

readers who do not have access to reference grammars of Finnish and may be

puzzled by some of the formal features of Finnish that are glossed in the main

body of this study (Chapters 4 ) 7). Recent outlines of Finnish written in Eng-

lish include Karlsson 1983b, Leino 1986: 13-16, and Branch 1987: 593-617.

Karlsson’s grammar is also in Swedish (Karlsson 1978a).

Finnish is generally typologically characterized as an agglutinative lan-

guage: there is a tendency to tack morphemes onto one another to form words:

kirja+ssa+ni+kin (book+in+my+too) ‘in my book too’. However, there are also

fairly strong inflectional tendencies in Finnish, and, thus, there are often com-

plex (morpho)phonological changes in word stems and affixes, e.g., käsi

‘hand’, kädessä ‘hand+INE’, kättä ‘hand+PAR’. Finnish has sixteen cases, i.e.

case-forms, which are discussed in section 3.3.3 below. There is also a rich

derivational morphology in Finnish. Affixes can be added to stems to modify

the meaning of the stem, for example, hypätä ‘(to) jump’, hypähtää ‘to make a

quick jump’, hypellä ‘to jump about (in a desultory fashion)’.

Because case-marking is used to indicate grammatical relations, Finnish

word order is fairly flexible. For the most part, it has a textual function. There

are a number of interpersonal and textual suffixes in Finnish (see e.g. A.

Hakulinen 1976, Vilkuna 1984, Vilppula 1984). These are suffixes which have

no effect on the experiential meaning of the clause, e.g. -pa/pä in on+pa ‘is

(3sg)’ and ei+pä ole ‘is not (3sg)’ typically occurs in assertions and denials.

There are no articles in standardized written Finnish, although se ‘it/that’ and

yks ‘one’ are used in an article-like way in informal spoken Finnish, particularly

in the speech of young people (see e.g. Laury 1991).

As mentioned in the previous section, written Finnish (kirjakieli) is very

phonemic, but this is partly because standardized spoken Finnish is based on

standardized written Finnish; i.e. basing the way in which something is pro-

nounced on the way it is spelt is the norm in Finnish. In the everyday, unself-

conscious speech of some Finns, foreign proper names, apart from those that

Page 83: systematic functional linguistic

75

have a Finnish equivalent (e.g. Lontoo ‘London’, Saksa ‘Deutschland’), are

sometimes pronounced as though the letters correspond to Finnish sounds: thus

the English name Peter Westlake [pi:tM wgstleik] would be pronounced [pe-ter

vest-la-ke] (where the hyphen indicates a syllable boundary and the vowels and

consonants are given Finnish values).

There are eight vowel phonemes in Finnish, which are generally tran-

scribed phonetically in the same way as they are normally written, although for

consistency /y/ is sometimes represented as /ü/, since the umlaut (() is used to

mark front vowels in Finnish. The front vowels are: i, y (ü), e, ö, and ä; the

back vowels are: u, o, and a. There are seventeen consonant phonemes in Finn-

ish: p, b, t, k, g, d, s, š (sometimes written or typed as sh), h, f, v, j, l, r, m, n, õ

(written n, occurs only before k). The consonants p, t, and k are unaspirated.

The consonants b, g, š, and f are found only in recent loan words and are often

replaced by p, k, s, and v in the unselfconscious speech of some Finns. The

consonant d is also a recent addition to the consonant inventory of Finnish but it

is more established. It either occurs in recent loan words or else it occurs as the

counterpart of t in consonant gradation (see below); it is sometimes replaced by

t (unaspirated) or omitted. The use of d as a counterpart to t in consonant grada-

tion is largely the result of spelling pronunciation: it is assumed that the original

counterpart to t was a voiced dental fricative (IPA: [ð]), which Agricola repre-

sented orthographically as d or dh. When this sound began to disappear from

Finnish dialects, the d(h) in written Finnish began to be pronounced as a voiced

dental stop on the model of Swedish where a sound change from [ð] to [d] had

occurred. (See Lehikoinen & Kiuru 1991: 94 ff.)

Compared with English, there are a lot of diphthongs in Finnish (ei, ey, äi,

ui, ai, oi, öi, yi, au, ou, eu, iu, äy, öy, ie, yö, uo); but consonant clusters are

fairly restricted. Length is phonemically distinct in Finnish: short vowels are

distinct from long vowels and consonants are distinct from geminates (double

consonants). In Finnish orthography, a long vowel is written as a double vowel:

e.g. long /a:/ is represented as aa (cf. latu ‘skiing track’, laatu ‘quality’).

Geminates are written as double consonants (cf. muta ‘mud’, mutta ‘but’).

Page 84: systematic functional linguistic

76

Amongst the more pervasive phonological features of Finnish are conso-

nant gradation and vowel harmony. Consonant gradation means that there is an

opposition between what are referred to as strong and weak consonants. Strong

consonants occur in certain phonological environments, weak consonants occur

in others. As a general rule, strong consonants occur before an open syllable,

weak consonants occur in closed syllables. For example:

pp: p tt: t kk: k

loppu: lopusta (end: end+ELA)hattu: hatussa (hat: hat+INE)kukka: kukan (flower: flower+GEN)

p: vt: dk: –

tupa: tuvasta (hut: hut+ELA)pata: padassa (pot: pot+INE)jalka: jalan (foot: foot+GEN)

In some phonological environments there is assimilation of a weak consonant

and the following consonant: e.g. ranta ‘shore’, rannalla ‘on the shore’. Some

recent loan words and proper nouns do not undergo consonant gradation: e.g.

auto ‘car’, autossa ‘in the car’. (See Karlsson (1983b: 30 ff.) for a more com-

prehensive outline of sound alternations in Finnish.)

Vowel harmony means that a word either contains front vowels (ortho-

graphically: ä, ö, and y) or back vowels (a, o and u). The vowels /i/ and /e/ are

neutral and can occur with either front or back vowels.

FRONT BACK

CLOSE y i u

HALF-CLOSE ö e o

OPEN ä a

Figure 3-1: Consonant Gradation

Figure 3-2: Vowel Harmony

Page 85: systematic functional linguistic

77

Because of vowel harmony, many case-endings have front and back variants. For

example, the inessive ending is either -ssa or -ssä, depending on whether the stem

has front vowels or back vowels: solmu ‘knot, tie’, solmussa ‘in a knot’, silmä

‘eye’, silmässä ‘in (one’s) eye’. There is no vowel harmony between the parts of

a compound word: työmaa ‘work place’. Exceptions to vowel harmony are wide-

ly found in recent loan words, e.g. hypoteesi ‘hypothesis’, konduktööri ‘conduc-

tor’.

3.3.2. Verb Inflexions

Finite verbs in Finnish are similar to finite verbs in European languages: the

tenses in Finnish are present (or, more correctly non-past, as it is also used to

express future time), past, perfect, and pluperfect. Finite verbs are marked for

person and number, but not for gender, as gender is not expressed in the morphol-

ogy of Finnish. (Even the third person singular pronoun hän ‘he/she’ can refer to

either male or female; and in spoken Finnish, the third person singular pronoun se

‘he/she/it’ is used to refer to human and non-humans.)

From the point of view of the better-known European languages, Finnish

verbs are different in at least two respects. Firstly, the so-called Finnish passive

form, which I shall refer to as the indefinite, differs in that it incorporates a bound

morpheme which indicates that the process (i.e. the action or state) was brought

about by a human participant, but the identity of this participant is not further

specified (see Shore 1986, 1988). Thus, as well as transitive verbs, intransitive

and modal verbs and the verb olla ‘(to) be’ occur in the indefinite. The second

way in which the Finnish verb system is distinct is in the variety of nominalized

forms of the verb. Finnish is generally said to have four infinitives, with the so-

called first infinitive having a short and long form (Ikola 1977: 52). This first

infinitive in its short form is the only form that occurs without a case-ending. The

other infinitives are really nominalized verb stems which can only occur with a

case-ending (see next section).

Page 86: systematic functional linguistic

78

1 The term “subject” is used here as a formal category, i.e. the nominal (or nominal group)that agrees with the finite verb in number and person.

There is agreement between the finite verb and the subject1: if there is a

finite verb with a personal ending in the clause, and, if the subject is realized in

the clause, then the verb agrees with it in number and person. This agreement

between subject and finite verb is consistent in standard written Finnish, although

it is not entirely consistent in informal spoken Finnish, where the third person

plural form of the verb is the same as the third person singular form and the indef-

inite (passive) form is used as a first person plural. The present indicative forms

of asua ‘to live’ (i.e. ‘to dwell’), for example, are as follows:

STANDARDIZED WRITTEN

SINGULAR:

1. (minä) asu+n ‘I live’2. (sinä) asu+t ‘you live’3. hän asu+u ‘she/he lives’

se asu+u ‘it lives’

PLURAL:

1. (me) asu+mme ‘we live’2. (te) asu+tte ‘you live’3. he asu+vat ‘they [human]

ne asu+vat ‘they [non-human] live’

INDEFINITE (passive):asu+ta+an ‘live (non-specified human participant(s))’

INFORMAL SPOKEN

SINGULAR:

1. mä asu+n ‘I live’2. sä asu+t ‘you live’3. se asu+u ‘she/he/it lives’

PLURAL:

1. me asutaan ‘we live’2. (te) asu+tte ‘you live’3. ne asu+u ‘they [human or

non-human] live’

INDEFINITE (passive):asu+ta+an ‘live (non-specifiedhuman participant(s))’

The variants given illustrate the declarative in standardized written Finnish and in

informal spoken Finnish (i.e. based on the Finnish spoken in the Helsinki area).

Verbs inflect in all tenses. The personal endings are the same for all verbs, al-

though there is morphophonological variation in the stem of different verb types

(for a more detailed account of the different types, see L. Hakulinen 1961, Karls-

son 1983b).

Figure 3-3: Present Indicative Forms of asua ‘to live/dwell)

Page 87: systematic functional linguistic

79

The indefinite (passive) is not marked for number or person as there is no

subject in an indefinite clause: Kaupungeissa asutaan kerrostaloissa [in+cities

live in+apartment+ houses] ‘In cities people/one/we/they/you live(s) in apartment

houses’ (see Shore 1986, 1988). The third person singular form asuu is used with-

out a subject to refer to a hypothetical person: jos asuu lasitalossa ... [if live+3SG

in+glass +house] ‘if one lives in a glass house ...’. A variant of this in spoken

Finnish is a second person singular with the unstressed pronoun sä ‘you (sg)’: jos

sä asut lasitalossa ... [if you live+2SG in+glass+house] ‘if one lives in a glass

house ...’. This is a recent development in Finnish, and appears to have come

about under the influence of English.

The bracketing of some of the pronouns in Figure 3-3 above serves to indi-

cate that a first or second person pronoun would only be realized in standardized

written Finnish in a marked or contrastive environment (see Helasvuo 1988: 67-

68); in spoken Finnish, the pronoun is often realized but there is some variation.

In other instances, it is typical for the pronoun to be realized unless it is presup-

posed by ellipsis, i.e. retrievable from the cotext (see Chapter 7). It should be

noted that the above division into standardized written and informal spoken does

not imply a simple dichotomy: as Halliday (1985b: 32) has pointed out, “there

are all sorts of writing and all sorts of speech, many of which display features

characteristic of the other medium”.

There are also conditional and potential inflexions in the verb. The condi-

tional has the infix -isi- e.g. mä asuisin ‘I would live’. The potential, which is

characterized by the infix -ne- (the n assimilates with the preceding consonant of

the stem in certain types), is rare and is generally confined to formal genres.

Finnish also has imperative inflections in the verb. The imperative forms of ottaa

‘(to) take’ for both spoken and written Finnish are given in Figure 3-4:

Page 88: systematic functional linguistic

80

1.2.3.

SINGULAR

)ota ‘take’otta-koon ‘let X-SG take’

PLURAL

otta-kaa-mme ‘let us take’otta-kaa ‘take’otta-koot ‘let X-PL take’

INDEFINITE: otetta-koon ‘let (them/people in general) take/let it be taken’ (no subject)

The first plural imperative (otta-kaa-mme ‘let us take’) is used mostly in formal

contexts, e.g. in a church service or in a toast. In most genres of Finnish, the pres-

ent tense of the indefinite form (otetaan ‘take-human participant’) in clause-ini-

tial position generally functions as an optative imperative. To some extent, this

form could be regarded as (a grammaticalization) of a 1. person plural imperative,

particularly if accompanied by high initial pitch; however, a clause-initial indefi-

nite can simply be a presentation of an action, rather than an orientation to a non-

linguistic response, e.g. tanssitaan ‘there’ll be dancing/let’s dance (see Shore

1988: 162-63.)

An important feature of Finnish verbs is that the negative element is not a

particle but a verb form which inflects for person and number. Finnish grammari-

ans generally refer to the “negative conjugation” (or inflection) or to the “nega-

tive (auxiliary) verb with an incomplete paradigm” (i.e. inter alia, there is no neg-

ative indefinite and tenses are formed paraphrastically) or to a “verb-like nega-

tive” in Finnish (e.g. L. Hakulinen 1961, 1979; Siro 1964: 91; Penttilä 1963: 250;

Hakulinen & Karlsson 1979: 268; Karlsson 1983b: 67). The negative auxiliary

combines 1) with a stem or a past participle of the lexical verb to form the present

and past tense or 2) with a stem from olla ‘be’ and a participle of the lexical verb

to form the other tenses. Thus, there is no invariant negative form like no or not

as in English: negatives in Finnish are inflected in the imperative (äl- + personal

ending), and in the indicative (e- + personal ending):

Figure 3-4: Imperative Forms of ottaa ‘(to) take’

Page 89: systematic functional linguistic

81

INDICATIVE (PRESENT)

SINGULAR PLURAL

1 en ota emme ota (~ei oteta )2 et ota ette ota3 ei ota eivät ota

INDEFINITE: ei oteta

IMPERATIVE

SINGULAR PLURAL

1 ) älkäämme ottako2 älä ota älkää ottako3 älköön ottako älkööt ottako

As with the positive forms in Figure 3-3 above, however, there is generally

no distinction between the singular and plural in the negative auxiliary of a third

person declarative in informal spoken Finnish, and a first person plural is formed

with the pronoun me and the indefinite form, which also has a third person nega-

tive auxiliary. With se ‘(s)he/it’, ne ‘they’, and me ‘we’, the pronoun would nor-

mally be realized in a negative clause, unless it is presupposed by ellipsis. With

the other personal forms, however, the negative is marked for number, even in a

simple response to a polar interrogative:

(1) <A> oot sä hiihtäny jo?be+2SG you-SG skied+PTC yet/already‘have you been skiing yet?’

<B> en [CA10:2]NEG+1SG

‘No /I haven’t.’

Figure 3-5: Negative Forms of ottaa ‘(to) take’

Page 90: systematic functional linguistic

82

1 The term “group” is not employed in the description of Finnish in this study (see Ch. 4).

3.3.3. Finnish Cases

Modern grammars of Finnish generally recognise 15 or 16 cases (i.e. case forms)

in Finnish. These include the so-called locative cases, similar to locative cases in

American Indian and Australian languages. Like English prepositions, locative

cases are not restricted to concrete expressions of location: they are also used in a

variety of non-concrete ways. Figure 3-6 lists the main case forms in Finnish.

The translations are given only as some sort of guide for those unfamiliar

with Finnish; a Finnish partitive, for example, can be glossed as ‘non-bounded’ in

contrast to a nominative, accusative or genitive, which are ‘bounded’ (see section

3.4.2) and would often be translated by a basic nominal in English (i.e. without an

article). The capital A in some of the case-endings stands for /a/ or /ä/. The capital

V in the illative indicates a lengthening of the stem vowel. Many of the case-

forms listed here have other, phonologically conditioned variants.

The examples given in Figure 3-6 are of an adjective and noun (pieni talo

‘small house’) and a pronoun (sinä ‘you’), but there is agreement or concord

amongst all the words in a nominal phrase,1 so if a Deictic is added, it would be

in the same case-form as the Head:

(2) noissa pienissä taloissathose+PL+INE small+PL+INE house+PL+INE

‘in those old houses’

There are, however, a handful of common adjectives that do not inflect for case or

number: for example, koko ‘whole’ (koko talossa ‘in the whole house’), viime

‘last’ (viime viikolla ‘(during) last week’), eri ‘different’ (eri paikoissa ‘in differ-

ent places’).

Page 91: systematic functional linguistic

83

CASE-FORM

NOMINATIVE)

PARTITIVE-(t)A

GENITIVE-n

ACCUSATIVE-t

ESSIVE-nA

TRANSLATIVE-ksi

INESSIVE-ssA

ILLATIVE-Vn

ELATIVE-stA

ADESSIVE-llA

ALLATIVE-lle

ABLATIVE-ltA

SINGULAR

pieni talo a/the small housesinä you

pientä taloa (of) a/the small housesinua (of) you

pienen talon of a/the small housesinun your

) (genitive)sinut you

pienenä talona as/for a/the smallhousesinuna as you, if I were you

pieneksi taloksi into (a/ the) smallhousesinuksi into you

pienessä talossa in(side) the smallhousesinussa in(side) you

pieneen taloon into (a/ the) smallhousesinuun to you

pienestä talosta from in(side) thesmall housesinusta from you

pienellä talolla by/on /near a/thesmall house, autolla by carsinulla by/on you

pienelle talolle to (by etc.) thesmall housesinulle to/for you

pieneltä talolta from (by etc.) thesmall housesinulta from you

PLURAL-t, -i- (-j-)

pienet talot (the) small houseste you

pieniä taloja (of) (the) small housesteitä (of) you

pienten talojen of (the) small housesteidän your

) (nominative plural)teidät you

pieninä taloina as/for small housesteinä as you, if I were you

pieniksi taloiksi into (the) small housesteiksi into you

pienissä taloissa in(side) (the) small housesteissä in(side) you

pieniin taloihin into (the) small housesteihin to you

pienistä taloista from in(side) the smallhousesteistä from you

pienillä taloilla by/on/near a/the smallhouses, autoilla by (more than one) carteillä by/on you

pienille taloille to (by/on/near) the smallhousesteille to/for you

pieniltä taloilta from (by/on/ near) thesmall housesteiltä from you

Figure 3-6: Common Case-Forms for Nominals

Page 92: systematic functional linguistic

84

The nominative, partitive, genitive and accusative are generally referred to

as the “grammatical cases”: they realize abstract (least specific) meanings. The

inessive, illative, elative, adessive, allative, and ablative as the “locative cases”,

as in concrete instances, at least, their meanings are fairly specific. The intermedi-

ate cases, the translative and the essive, are generally used in abstract ways. In the

following table, I have grouped the elative and the adessive as both intermediate

and locative, as some of their uses are more abstract. They are particularly rele-

vant in the distinction that is made between process types in Chapter 6 of this

study. However, as pointed out earlier, all of these locative cases are used in a

variety of non-concrete ways.

grammatical

intermediate

nominative (talo ‘house’)partitive (taloa ‘of the/a house’)genitive (talon ‘house’s’)accusative (minut ‘me’)

essive (talona ‘as/for a house’)translative (taloksi ‘into a house’)

elative (talosta ‘from/of the/a house’)adessive (talolla ‘by a/the house’)

illative (taloon ‘(in)to the/a house’)ablative (talolta ‘from a/the house’)allative (talolle ‘to (by) a/the house’)inessive (talossa ‘in the house’)

locative

The uses of the following cases are more restricted: i.e. they typically occur

with only certain types of stem or in certain idiomatic expressions. The abessive

and the instructive, however, are commonly used with a verb stem (see Figure 3-

10 below).

Figure 3-7: Convenient Groupings of Cases

Page 93: systematic functional linguistic

85

ABESSIVE

INSTRUCTIVE

rahatta (SG) ‘without money ‘

jalan (SG) ‘by foot’ käsin (PL) ‘by hand’

COMITATIVE

PROLATIVE

lapsineni (SG/PL) ‘with my child/children’

meritse (SG/PL) ‘by sea/seas’;sähköpostitse (SG/PL) ‘by electronic mail’

Plural forms of the abessive and instructive are rare, and there is no distinction

between singular and plural in the comitative and the prolative case.

As mentioned earlier, there are a number of other non-finite verb forms in

Finnish generally referred to as infinitives. Finnish is said to have four infinitives

(see Karlsson 1983b: 155 ff.). The following are examples of non-finite forms

using olla ‘(to) be’ and mennä ‘(to) go’:

1st infinitive (basic form): olla ‘(to) be’, mennä ‘(to) go’ (longer form): ollakseen ‘(in order) to be’

STEM + TRANSLATIVE + POSSESSIVE SUFFIX

2nd infinitive olle-ollessa (INESSIVE) ‘while/in being’ollen (INSTRUCTIVE) ‘thus/so being’

3rd infinitive olema- + case endingmenemä- (see below)

4th infinitive (rare) meneminen, menemistä ‘to go’(NOM/PAR)

Figure 3-8: Non-Productive or Semantically Restricted Case-Forms

Figure 3-9: Infinitives

Page 94: systematic functional linguistic

86

What is called the fourth infinitive occurs in (often slightly archaic sounding)

modal type constructions: sinne ei ole menemistä ‘one must not go there’.

The longer form of the 1st infinitive as well the 2nd and 3rd infinitive al-

ways occurs with a case-ending. The following table lists the case-forms that can

occur with these infinitives and gives some examples.

TRANSLATIVE

INESSIVE

ILLATIVE

ELATIVE

ADESSIVE

ABESSIVE

INSTRUCTIVE

elä+ä+kse+mmelive+INF+TRA+POS/1PL

‘in order to live’

tietä+ä+kse+niknow+INF+TRA+POS/1SG

‘as far as I know’

ui+ma+ssaswim+INF+INE

‘(in) swimming’

uid+e+ssa- (+ poss. suff.)swim+INF+INE

‘while swimming’

ui+ma+answim+INF+ILL

‘(to) swimming’

ui+ma+staswim+INF+ELA

‘(from) swimming’

ui+ma+llaswim+INF+ADE

‘by swimming’

sano+ma+ttasay+INF+ADE

‘without saying’

juost+e+nrun+INF+INS

‘(by) running’

Syömme elääksemme.We eat in order to live.

Tietääkseni se on valmis.As far as I know its ready.

Mika on uimassa.Mika is swimming.

Uidessaan meressä mies sai sydänkohtauksen.While swimming in the sea the manhad a heart attack.

Mika meni uimaan.Mika went swimming.

Mika on tullut uimasta.Mika has been swimming [has come from swimming].

Uimalla kohoat kuntoasi.Swimming will make you fit. [By swimming you’ll get fit.]

Jukka seisoi sanomatta mitään.Jukka stood (there) without saying any-thing.

Mika tuli juosten kotiin.Mika came running home.

Figure 3-10: Common Case Forms for Non-Finite Verb Stems

Page 95: systematic functional linguistic

87

1 Even though the pronoun se ‘(s)he/it’ is used to refer to humans in spoken Finnish (seeFigure 3-6 above), it does not pattern like a human pronoun but like a singular nominal.

Some of these cases also occur with an indefinite (passive) stem (see Karlsson

1983b: 157 ff).

The so-called fourth infinitive is identical in form with a frequently used

deverbal noun (i.e. a noun derived from a verb) ending in -minen, e.g. juoksemi-

nen ‘running’, uiminen ‘swimming’, which is inflected like any other nominal.

Present and past participles also inflect like other nominals when they function

like a nominal: tulevana vuotena [coming+ESS year+ESS] ‘during the coming

year’; kokeneelta opettajalta [experienced+ABL teacher+ABL] ‘from an experi-

enced teacher’.

3.4. Issues in the Received Description of Finnish

3.4.1. The Problem of the “Accusative”

The account that I have given above differs from some traditional accounts of

Finnish cases (e.g. Ikola 1977, Hakulinen & Karlsson 1979, Karlsson 1983b) in

respect of the accusative case. As discussed in this section, the notion of the accu-

sative in many accounts of Finnish is based on a confusion of form and function.

In the ensuing discussion, I am not, of course, suggesting that Finnish linguists

are not aware of this confusion: the accusative in the received grammatical tradi-

tion has developed as a convenience in order to explain the case-marking of what

is traditionally referred to as the object.

In Finnish, there is no special case-form for a noun that ) in traditional

terms ) is considered to be the object in a clause. It is either in the nominative,

genitive or partitive. A pronoun that refers to a human,1 on the other hand, has an

accusative ending (-t) in structures where other nominals would be either in the

nominative or genitive. Like a noun, a pronoun object can also be in the partitive.

Page 96: systematic functional linguistic

88

1 In fact, it might be more appropriate to refer to the nominative in Finnish as an absolutive.

Leaving aside the partitive for the moment, a noun (or a nominalized verb form)

that is the object in a clause is said to be in the accusative. In this view, there are

two types of “accusative” forms for nominals: a) without an ending, the

“nominative-like” accusative, and b) with an ending, the “genitive-like” accusa-

tive. In other words, if an object is in the nominative or genitive, it has been re-

ferred to as an “accusative”.

“NOMINATIVE-LIKE ACCUSATIVE”:

(3) Liisa osti molemmat talot.Liisa+NOM bought+3SG both+NOM+PL houses+NOM+PL

‘Liisa bought both houses.’

(4) Osta talo maaseudulta!buy+2SG+IMP house+NOM countryside+ABL

‘Buy a house in the countryside!”

(5) Talo myytiin eilen.house+NOM sold+INDE yesterday.‘The house was sold yesterday.’

“GENITIVE-LIKE ACCUSATIVE”:

(6) Liisa osti talon.Liisa+NOM bought+3SG house+GEN

‘Liisa bought a/the house.’

As illustrated by these examples, the nominative in Finnish is best regarded

as a form without a case-ending. It should not be equated with the Latin nomina-

tive as both subject and object can occur in the nominative in Finnish:1 the nomi-

native marks an inherent participant that is “bounded”. The notion of bounded-

ness will be discussed in the next section; for the moment, the distinction between

boundedness and non-boundedness in Finnish can be seen as being roughly

equivalent to both the distinction between perfective and imperfective aspect and

the distinction between definiteness and indefiniteness in Indo-European lan-

guages.

Page 97: systematic functional linguistic

89

The genitive-like accusative occurs only in instances in which the object is

a singular nominal, the polarity of the clause is positive, and the finite verb is

neither imperative nor indefinite (passive). Moreover, the process must be

bounded; if it is not bounded, the nominal is in the partitive:

(7) Liisa osti maata maaseudulta.Liisa+NOM bought+3SG land+PAR countryside+ABL

‘Liisa bought (some) land in the countryside.’

(8) Liisa lukee kirjaa.Liisa+NOM read+3SG book+PAR

‘Liisa is reading a book.’

Moreover, if the polarity is negative, the object is generally in the partitive:

(9) Liisa ei ostanut taloa.Liisa+NOM NEG+3SG buy+PTC house+PAR

‘Liisa didn’t buy a/the house.’

The genitive-like accusative illustrated in 6 above is phonologically identi-

cal to an ordinary singular genitive. However, historical linguists generally as-

sume that it has developed from a distinct form, and this historical evidence ap-

pears to be another factor in assuming that there is a genitive-like accusative in

contemporary Finnish that is distinct from an ordinary genitive. As the evidence

does not appear to be water-tight, I shall give a brief outline of the received view

in historical linguistics as well as a dissident view that has recently been put for-

ward by Décsy (1990).

It is generally assumed that the “accusative” n in the genitive-like accusative

is the result of a sound change *-m > -n. Like the genitive-like accusative in con-

temporary Finnish, this so-called m-accusative of Proto-Uralic appears to have

been restricted to singular nominals (see L. Hakulinen 1979: 98-99). This m-accu-

sative has been “preserved” in Mari, a Finno-Ugric language spoken in the middle

of Russia (Kangasmaa-Minn (1991), Hakulinen (ibid.)). The postulation of an m-

accusative in Proto-Uralic seems to be based on the assumption that a genitive-

like accusative is quite unlike an ordinary genitive. This assumption is put into

Page 98: systematic functional linguistic

90

1 According to one statistical study (Pajunen & Palomäki 1982), the genitive is the secondmost frequent case-form in Finnish.

question if one takes into account the function of the genitive in modern Finnish.

As discussed more fully below, the genitive in Finnish scans a semantic contin-

uum from a somewhat concrete and tangible notion of possession (e.g. Susannan

talo ‘Susanna’s house’) to a more ineffable and abstract notion of subordination

(hypotaxis) (e.g. yllättävän lapsellinen [surprising+GEN childish] ‘surprisingly

childish’). This being the case, one could also argue that there is a subordinative

relationship between the verb and an NP in the genitive-like accusative (as illus-

trated by 6 above). The fact that only certain singular nominals are marked in this

way is no more or no less problematic than assuming that only certain objects

were marked in the accusative.

In Décsy’s (1990: 68-69,81) view, on the other hand, there was no special

accusative marking in Proto-Uralic. Décsy regards the possessive relationship as

the most basic: the m-accusative developed from a 1st person possessive ending

at a later stage. The genitive is considered by Décsy to be secondary. A first per-

son possessive in modern Finnish ends in -ni (e.g. minu+n talo+ni I+GEN

house+1SG/POS ‘my house’), but the possessive n is assumed to have developed

from an m (cf. minä ‘I’). What happened during or after the Proto-Uralic stage is

pure conjecture, but the link that Décsy sets up between the accusative and the

possessive also links the accusative with the genitive, since a possessive relation-

ship in Finnish is marked by a possessive suffix as just indicated, or by a pronoun

or noun in the genitive: minun taloni I+GEN house+1SG/POS ‘my house’) Susan-

nan talo [Susanna+GEN house] ‘Susanna’s house’. This link between the accusa-

tive and the genitive undermines the need to have postulated a separate accusative

form in the first place.

To consider more fully the link between the genitive-like accusative and the

ordinary genitive, it may be helpful to look at ordinary genitives in Finnish. The

genitive is extensively used in modern Finnish,1 and it seems to have a consistent

grammatical function. According to Kangasmaa-Minn (1991), the genitive in

Page 99: systematic functional linguistic

91

Finnish is an indicator that the unit in question is subordinated to (i.e. dependent

on) another unit; in SF terms, the genitive is a marker of hypotaxis: the genitive

forms in (i) talon ‘house+GEN’, (ii) meidän ‘we+GEN’ and (iii) yllättävän ‘surpris-

ing+GEN’, for example, indicate that there is another element on which it is de-

pendent: (i) talon takana [house+GEN behind] ‘behind the house’, talon ikkunat

[house+GEN windows] ‘the windows of the house’ or talon omistajalle

[house+GEN owner+ALL] ‘to the owner of the house’ ; (ii) meidän talomme

[we+GEN house+POS/1PL] ‘our house’ or hän antoi meidän mennä [(s)he gave

we+GEN go+INF] ‘(s)he allowed us to leave’, (iii) yllättävän lapsellinen

[surprising+GEN childish] ‘surprisingly childish’ or yllättävän hitaasti [surpris-

ing+GEN slowly] ‘surprisingly slowly’. If the dependent item is in turn modified

by other items, then the dependent items agree with the sub-head in number and

case, i.e. they are all in the genitive (singular or plural): noiden vanhojen talojen

uudelle omistajalle [those+PL+GEN old+PL+GEN house+PL+GEN new+ALL

owner+ALL] ‘to the new owner of those old houses’.

One might alternatively argue that the genitive marks a rankshifted (em-

bedded) element in Finnish. In fact, while Kangasmaa-Minn refers to the genitive

as a marker of subordination (or dependency), she attempts to throw some light

on clauses containing a constituent in the genitive by comparing them to clauses

that contain an embedded clause. A clause containing a genitive is said to be syn-

tactically complex. The clause Helsingin kaupunki sijaitsee rannikolla [Hel-

sinki+GEN city is-situated coast+ADE] ‘The city of Helsinki is situated on the

coast’ is compared to a matrix clause Kaupunki sijaitsee rannikolla ‘The city is

situated on the coast’ and an embedded clause Kaupunki on Helsinki ‘The city is

Helsinki’. As is obvious from this example, Kangasmaa-Minn’s notion of embed-

ding is based on transformational-generative theory: the surface clause is repre-

sented in a hypothetical deep structure as a clause containing another embedded

clause. Moreover, in Kangasmaa-Minn’s analysis, the genitive itself is not em-

bedded, but the clause containing the genitive is compared to a clause in which

there is another embedded clause.

Page 100: systematic functional linguistic

92

1 The present participle koskevan is considered to be a “genitive-like accusative” by Finnishgrammarians (e.g. Ikola 1977: 153).

The varied environments in which there is a genitive in Finnish can be illus-

trated by the following text example, which shows how the genitive marks an-

other layer of modification within a unit ) or a layer of submodification:

(10) Telakkateollisuuden ongelmat tuntuvatdock(yard)+industry+GEN (1) problems+NOM+PL seem+3PL

pinnallisesti katsoen koskevan vain muutamiasuperficially look+INF+INS affect+PTC+GEN

1 (2) only few+PL+PAR

sellaisia satamakaupunkeja, joissa onsuch+PL+PAR port+town+PL+PAR which+PL+INE be+3SG

laivanrakennustoimintaa. [HKV]ship+GEN (3) + building+activity+PAR

‘Looked at superficially, the problems in the ship-building industry seem to affect only afew (such) ports in which ships are being built.’

The subject ongelmat ‘problems’ is modified by telakkateollisuuden [dock +

industry + GEN], which constitutes a layer of modification within the subject NP.

Similarly the finite verb tuntuvat ‘seems’ is modified by the participle, koskevan

‘affecting’, which in turn is extended by the NP muutamia sellaisia

satamakaupunkeja ‘a few (such) ports’. The third example of the genitive is in

the compound, laivanrakennustoiminta [ship + GEN + building + activity], where

the Head toiminta ‘activity’ is modified by rakennus ‘building’ which, in turn is

submodified by laiva ‘ship’, which is in the genitive.

The extensive use of the genitive underlines the importance of dependency

relations in Finnish. In systemic-functional descriptions of English, the main em-

phasis is generally put on constituency, and it is only in complexing that the no-

tion of dependency is employed. It would seem that in Finnish, one needs to take

into account both constituency and dependency even in the basic grammatical

organization (i.e. just not in complexing). It seems to me that a daughter-depend-

ency model, along the lines that were earlier suggested by Hudson (1987), pro-

Page 101: systematic functional linguistic

93

1 The instructive (also ending in -n) could also be tied into the discussion of the genitive andaccusative, since it has been etymologically related to the genitive (L. Hakulinen 1979: 106).However, in contemporary Finnish, while the singular instructive is identical with the singulargenitive, there is a plural instructive (while rarely used) that differs in form from both thegenitive plural and the nominative plural.

vides a more realistic view of “how language works” (Halliday 1961: 242), or, at

least, of how the Finnish language works.

As indicated earlier, it may also be valid to regard the genitive-like accusa-

tive as a marker of a bounded unit that (sub)modifies the verb. I shall not pursue

this question further as it is not central to my concerns. The point of this discus-

sion has been to show how the notion of a genitive-like accusative is used in

many grammatical descriptions of Finnish and to put into question the need to use

such a category. It is clearly based on a confusion of form and function, and while

this may be seen as a convenient way of discussing case-marking in Finnish, it

nevertheless obscures the grammatical patterning in Finnish and takes for granted

that the notion of an object (case-marked as an accusative) is crucial in the gram-

matical description of Finnish.

Chapters 4 – 7 of this study outline a functional grammar of Finnish, and, a

functional grammar needs to take into account both form and function. However,

to refer to a “genitive-like accusative” is like saying that Steve and Finnish mov-

ies in Sonja likes Steve/Finnish movies are nominative-like accusatives because

him could be regarded as an accusative in Sonja likes him. As it is no more prob-

lematic to regard a singular nominal ending in n as a genitive than it is to regard

it as a “genitive-like accusative”, I shall gloss it as a genitive in this study.1

Page 102: systematic functional linguistic

94

1 The notion of space should not be understood only in concrete terms. Moreover, asdiscussed in this section, a distinction between space and time is not construed in thegrammar of boundedness in Finnish.

2 Finnish grammarians have traditionally referred to a bounded process as a “resultative”process. This can roughly be glossed as perfective or telic (i.e. one that moves towards alogical conclusion (see Comrie 1976: 44-48)).

In contrast to nouns, there is a distinct accusative (-t) for human pronouns in

Finnish. However, this is considered to be a relatively recent development, i.e.

something that has happened during the time that Finnish has developed as a sep-

arate language. According to L. Hakulinen (1979: 98-99)) the accusative in hu-

man pronouns is (historically) a feature of the Eastern dialects; in the Western

Dialects, these pronouns had the same “accusative” ending as singular nominals

(-n): in other words, there were only genitive forms for pronouns. While there are

three possibilities in case-marking for a nominal that is the object in a clause, i.e.

nominative, genitive and partitive, with human pronouns, there is only a two-way

opposition between the accusative and the partitive. Thus, in the examples above

where the nominal in question was in the genitive or nominative (3 ) 6), a human

pronoun would be in the accusative in a similar grammatical environment.

3.4.2. Boundedness

The difference in the case-marking of nominals between the partitive, on the one

hand, and the nominative, accusative and genitive, on the other, is essentially a

difference in boundedness. To simplify things somewhat, the nominative, accusa-

tive, and genitive refer to something that is constructed as being bounded in time

or space1: either to 1) a bounded entity or an entirety or to a set of entities or to

2) a process which is bounded.2 If the entity or process is not bounded, it is real-

ized by an NP in the partitive. The following clauses illustrate some of the ways

in which this distinction is realized; the a-clauses are non-bounded, the b-clauses

bounded:

Page 103: systematic functional linguistic

95

(11) a. Maalasin taloa.paint+PAS+1SG house+PAR

‘I was painting a/the house/I painted part of the house.’

b. Maalasin talon.paint+PAS+1SG house+GEN

‘I painted a/the house (all of it).’

(12) a. Pekka ajoi autoa.Pekka drive+PAST+3SG car+PAR

‘Pekka drove/was driving a/the car.’

b. Pekka ajoi auton talliin.Pekka drive+PAST+3SG car+GEN garage+ILL

‘Pekka drove a/the car into the garage.’

(13) a. Miestä ammuttiin.man+PAR shoot+INDE+PAST

‘A/The man was shot (not fatal).’

b. Mies ammuttiin.man+NOM shot+INDE+PAST

‘A/The man was shot (dead).’

(14) a. Häntä ammuttiin.(s)he+PAR shoot+INDE+PAST

‘She/He was shot (not fatal).’

b. Hänet ammuttiin.(s)he+ACC shot+INDE+PAST

‘She/He was shot (dead).’

As indicated in example 12b, a process can be bounded by an adjunct in the

clause.

A clause with negative polarity is generally non-bounded, thus the relevant

nominals in the b-clauses would be in the partitive if the clauses were negative.

(15) En maalannut taloa.NEG+1SG paint+PAS+1SG house+PAR

‘I didn’t paint a/the house.’

(16) Pekka ei ajanut autoa talliin.Pekka NEG+3SG drive+PTC car+PAR garage+ILL

‘Pekka didn’t drive a/the car into the garage.’

Page 104: systematic functional linguistic

96

(17) Miestä ei ammuttu.man+PAR NEG+3SG shoot+INDE+PTC

‘A/The man wasn’t shot.’

(18) Häntä ei ammuttu.(s)he+PAR NEG+3SG shoot+INDE+PTC

‘She/He wasn’t shot.’

Thus, there is a clear link in Finnish between boundedness and non-boundedness,

on the one hand, and positive and negative polarity, on the other. If one negates a

process in Finnish, then it is construed as not being temporally or materially

bounded. However, even in a negative clause it is also possible for the object to

be in the genitive, if it is a nominal, or in the accusative, if a human pronoun. For

example:

(19) Eikö hänet ammuttu(kin).NEG+3SG+Q he/she+ACC shot+INDE+PTC

‘He was shot, wasn’t he? (Wasn’t it the case that he was shot.)’

The significant factor is not whether the finite verb is formally negative, but

whether the meaning is negative.

The bounded/non-bounded distinction also occurs in what have been tradition-

ally referred to as existential clauses in Finnish. (See also next section and Chap-

ter 6 for further discussion of these.)

(20) Pöydällä on lasi.table+ADE be+3SG glass+NOM

‘There’s a glass on the table.’

(21) Pöydällä on lasia.table+ADE be+3SG glass+PAR

‘There’s glass on the table.’

(22) Pöydällä on laseja.table+ADE be+3SG glass+PL+PAR

‘There are glasses on the table.’

(23) Pöydällä on silmälasit.table+ADE be+3SG eye+glass+NOM+PL

‘There are spectacles [eye+glasses-bounded] on the table.’

Page 105: systematic functional linguistic

97

If the clause is negative, the nominal in the nominative in the first example above

(20) would generally be in the partitive. (See next section (3.4.3) for a discussion

of the effect of negative polarity on example 23.)

An adjective or noun complement in an intensive clause can also be

bounded or non-bounded. In intensive clauses, the boundedness of the comple-

ment depends on 1) whether the subject is bounded and 2) whether the relation-

ship between the subject and complement is bounded in space or time. As a first

rule of thumb, if the subject is regarded as realizing a bounded entity or a

bounded set, then the complement is in the nominative. Such subjects would gen-

erally include singular nominals that realize a concrete bounded entity (e.g. kirves

‘axe’) or a bounded set, which would be in the nominative plural. What is con-

strued as a bounded set in Finnish is problematic, and, in the final analysis, ineffa-

ble (see 2.4.11); however, bounded plurals would generally include, for example,

plurale tantum words (e.g. sakset ‘scissors’), parts of the body (e.g. silmät

‘eyes’), culturally defined sets of activities (e.g. häät (NOM/PL) ‘wedding’,

hautajaiset (NOM/PL) ‘funeral’, hipat ‘party (informal)’, uutiset ‘news (on televi-

sion/radio)’).

(24) Kirves on tylsä.axe+NOM be+3SG blunt+NOM

‘The axe is blunt.’

(25) Uutiset oli(vat) lyhyet.news+NOM/PL be+PAST+3SG(3PL) short+NOM/PL

‘The (television/radio) news were short.’

In these example, kirves ‘axe’ is a concrete entity and uutiset ‘news’ construes a

set of events that is bounded in space and time.

On the other hand, if the relationship between a bounded subject and the

complement is not exhaustive, then complement would be in the partitive.

Page 106: systematic functional linguistic

98

1 Cf. Kirves on kultainen ‘The axe is golden’, where there is an exhaustive relationshipbetween the adjective kultainen ‘golden’ and the subject kirves ‘axe’, i.e. the axe isexhaustively characterized as being golden. In example 26, on the other hand, the substancerealized by the mass noun kulta ‘gold’ is not exhausted by the axe, and thus it is in thepartitive as illustrated.

2 See Halliday (1985a: 160) for a discussion of these terms with reference to English.

(26) Kirves on kultaa.1

axe+NOM be+3SG gold+PAR

‘The axe is (of) gold.’

With the uutiset ‘news’ example above (25), if the complement were in the plural

partitive (i.e. uutiset ovat lyhyitä) it would mean something like ‘(television or

radio) news are short (in all countries and at all times)’.

If the subject realizes a non-bounded entity, e.g. a mass or non-count noun

(puuro ‘porridge’), or an abstract noun (suru ‘sorrow’), then the complement

would generally be in the partitive.

(27) Puuro on terveellistä.porridge+NOM be+3SG healthy+PAR

‘Porridge is healthy.’

(28) Suru on tarttuvaa.sorrow+NOM be+3SG contagious+PAR

‘Sorrow is contagious.’

However, it is also possible for the complement to be in the nominative. This

appears to most naturally occur in contexts where both 1) the noun realizing a

non-bounded entity such as suru ‘sorrow’ is further specified by a demonstrative

or possessive deictic,2 and 2) the complement realizes an attribute that can be

regarded as all-encompassing (and thus exhaustive):

(29) Hänen surunsa oli ääretön.(s)he+GEN grief+POS/3 be+3SG boundless+PAR

‘Her/His grief was boundless.’

Ordinary plural subjects (e.g. pojat ‘(the) boys’) are generally non-bounded:

Page 107: systematic functional linguistic

99

1 As indicated in Figure 3-3 above, in spoken Finnish, a third person plural subject does notalways agree with the verb in number and a first person plural subject occurs with a verb formthat is not marked for number or person (see Shore 1988: 155, 164). However, in theseinstances, there is agreement in standardized written (and formal spoken) Finnish.

(30) Pojat olivat iloisia. boy+NOM/PL be+PAS+3PL happy+PL+PAR

‘The boys were happy/glad.’

Although the partitive seems to be far more frequent, in certain instances the com-

plement can be in the nominative. Yli-Vakkuri (1969) suggests that it is most

natural for the complement to be in the nominative if the relationship between the

subject and complement is cotextually or contextually bounded. The following

example is from Yli-Vakkuri (1969: 263):

(31) Saamastaan lahjasta pojat olivat hurjanreceive+ELA+POS/3 gift+ELA boy+NOM/PL be+PAS+3PL wild+GEN

iloiset ja pyytävät välittämään kiitoksensa.happy+PL+PAR and ask extend their thanks

‘The boys were wildly happy about the gift and would like to extend their thanks.’

The discussion presented here is meant only as an outline. For a more com-

prehensive discussion, see, for example, Setälä (1891) or earlier or later editions,

Sadeniemi 1950, Denison 1957, Yli-Vakkuri 1969, T. Itkonen 1974a, 1979, Brig-

den 1984, Heinämäki 1984, Toivonen 1986, Larjavaara 1991 and Leino 1991; see

also Carlson 1981, Dahl 1981 and Langacker 1987 for a general discussion of

boundedness with examples from English. The notion of boundedness as realized

in Finnish can also be related to the notion of definiteness as realized in English

(see Chesterman 1991).

3.4.3. Traditionally Defined Grammatical Subject in Finnish

In this study, I reject the received notion of grammatical subject in Finnish. I con-

fine the subject in Finnish to NPs with the following characteristics:

) The subject is an NP in the nominative case.) The subject agrees with the finite verb in number and person.1

Page 108: systematic functional linguistic

100

) The subject is unaffected by the polarity of the verb, i.e. if we compare the clause in ques-tion with an agnate clause with negative polarity, the relevant element would remain in thenominative case.

A similar view is taken by Vilkuna (1989: 46,150). This definition of the gram-

matical subject in Finnish will be adopted in Chapters 5 and 6 of this study. Un-

like English, however, the grammatical subject in Finnish does not realize an

interpersonal function: it will be argued that it realizes an experiential function.

The received notion of subject in Finnish (see Hakulinen & Karlsson 1979:

158-172), is a mixture of grammatical and semantic properties, some of which

contradict each other. It seems to me that the reason for this is that the notion of

subject ) as it has evolved in grammatical descriptions of Finnish ) is based on

the Indo-European notion of subject. The first models for grammars of Finnish

were grammars of Latin, and, subsequently, Finnish grammarians have been in-

fluenced by grammatical descriptions of Indo-European languages, and, in recent

years, more specifically by descriptions of English.

The traditional view of subject in Finnish is derived from the logically-based

notion of a subject, which has its roots in a tradition stemming from Aristotle to

grammars of Latin, and then to Latin-based grammars of other languages. As

Halliday (1977: 36) has pointed out, in the Aristotlean tradition the (main) com-

ponents of a sentence, the subject and predicate, are logical concepts: they are the

components of a proposition, the functions in a premiss. As these logical concepts

have to be distinguished from the linguistic elements that enter into logical rela-

tions, the grammatical concepts of subject and predicate are treated as purely for-

mal items. Consequently, the subject is grammatically defined, for example, in

terms of position in the clause, case (e.g. nominative case), and/or marking on the

verb (e.g. person, number or gender agreement.) During the second half of the

nineteenth century, grammarians started to differentiate the psychological and

logical subject from the grammatical subject (see, e.g., Matthews 1981: 102;

Halliday 1985a: 33); and, more recently, Keenan (1976) has introduced the notion

Page 109: systematic functional linguistic

101

1 They are similar to the impersonal verb constructions in Australian languages as described,for example, by Walsh (1987). See further Chapter 6.

of a prototypical subject, which ) not surprisingly ) more or less corresponds to

the prototypical semantic features of what is considered to be the grammatical

subject in English.

The logical basis of subject and predicate in Finnish is evident in the follow-

ing quotation from Siro (1964: 16):

There are different components in a clause, the main components being the subject and thepredicate, for example Aurinko paistaa ‘The sun shines/is shining’, in which the expressionaurinko ‘sun’ is referred to as the subject and paistaa ‘shines’ the predicate. [Translation:S.S.]

Siro, however, goes on to point out that many clauses types in Finnish do not

have a subject. Some indication of the proportion of subjectless clauses in Finnish

is given in a corpus-based statistical analysis of written Finnish: 19% of all

clauses examined were subjectless (Hakulinen, Karlsson & Vilkuna 1980: 29-31).

Moreover, the notion of a “grammatical subject” in this analysis is elastic: it in-

cludes not only NPs in the nominative case that agree with the finite verb in num-

ber and person but also NPs in the genitive or partitive case that do not agree with

the verb.

A statistical analysis of written Finnish ) i.e. based on formal, written gen-

res of Finnish ) fails to capture the fact that subjectless clauses are in some sense

very basic and central clause types, characteristic of everyday spoken Finnish.

One class of subjectless clauses, which have been referred to as causatives of

feeling (see Vilkuna 1989: 45-46), for example, consists of what is typically con-

strued in Finnish as an agentless physical or mental process that is not self-engen-

dered:1

(32) mua paleleeI+PAR freeze+3SG

‘I’m freezing.’

Page 110: systematic functional linguistic

102

(33) mua ottaa päähän.I+PAR take+3SG head+ILL

‘I’m feel annoyed/irritated.’

Regardless of their statistical frequency in a written corpus, clauses like these are

by no means unusual in everyday interaction ) they are in some sense very fun-

damental to intimate, everyday interaction as expressions of personal drives, feel-

ings and emotions. This is reflected in the fact that they occur in child language at

a very early age (see Toivainen 1980: 126-27; 1986: 457).

Other clause-types which are generally regarded as subjectless by Finnish

grammarians include clauses that refer to temporal or meteorological states or

conditions, e.g. sataa [lit. rains] ‘it’s raining’, on tiistai [is Tuesday] ‘it’s Tues-

day’, on myöhä [is late] ‘It’s late’) and indefinite clauses (see Shore 1986, 1988),

e.g. maalla tehtiin työtä [countryside+ADE did+INDE work+PAR] ‘country folk

worked/people in the countryside worked’, ohjelma pantiin hyllylle [pro-

gramme+NOM put+INDE+PAST shelf+ALL] ‘the programme was shelved’. Thus,

even in the received view, there are a significant number of subjectless clauses in

Finnish.

On the other hand, there are a number of other clause types that contain an

NP which has been regarded as a subject, but this subject does not meet the crite-

ria listed at the beginning of this section, i.e. that it is in the nominative case,

agrees with the verb in number and person, and it is unaffected by the polarity of

the verb. Furthermore, there are no other grammatical criteria that would link

these so-called subjects to the subject as defined in this study. Clauses with dubi-

ous subjects include existential clauses (which includes a subset of what have

been referred to as “possessive constructions”) and necessitative clauses. The

“subject” in an existential clause is in the nominative or partitive case, and the

verb (olla ‘be’ or an intransitive verb) is always in the third person singular form,

i.e. if the “existential subject” is plural, it does not agree with it in number:

(34) Pöydällä on lasi.table+ADE be+3SG glass+NOM

‘There’s a glass on the table.’

Page 111: systematic functional linguistic

103

(35) Pöydällä on lasia.table+ADE be+3SG glass+PAR

‘There’s glass on the table.’

(36) Pöydällä on laseja.table+ADE be+3SG glass+PL+PAR

‘There are glasses on the table.’

Whereas the subject (as defined at the beginning of this section) is always in

the nominative, an “existential subject”, on the other hand, may be in the nomina-

tive in a clause with positive polarity, but would be in the partitive in an agnate

clause in the negative. Thus, if the polarity of clause 34 above is changed, the

“existential subject” would be in the partitive:

(37) Pöydällä ei ole lasia.table+ADE NEG+3SG be glass+PAR

‘There isn’t a glass on the table ~ There’s no glass on the table.’

The traditional notion of an existential clause in Finnish also includes

clauses in which the NP in the grammatical case is in the nominative plural, for

example:

(38) Pöydällä on silmälasit.table+ADE be+3SG eye+glass+NOM+PL

‘There are spectacles [eye+glasses-bounded] on the table.’

These inherently bounded NPs are not necessarily affected by the polarity of the

verb. In line with Vilkuna’s (1989: 159) suggestion, however, it seems to me that,

in instances where the NP in question is unaffected by the polarity of the verb, the

clause is not an existential clause but a normal circumstantial relational process.

Instances in which the NP would be affected by the polarity of the verb, on the

other hand, could be considered existential.

An NP in the genitive case in a so-called “necessitative clause” has also

been regarded as a subject (e.g. Hakulinen and Karlsson 1979: 172; Karlsson

1983b: 90) in spite of the fact that the verb is always in the third person singular.

Page 112: systematic functional linguistic

104

(39) Minun pitää lähteä kotiin.I+GEN must+3SG leave-INF home+ILL

‘I’ve got to/I must go home.’

The genitive NP in clauses like these can be characterized as a primary participant

and as the unmarked Theme (see Chapter 7). It clearly contrasts with clauses in

which there is a subject that agrees with the verb in number and person:

(40) Minä lähden kotiin.I-NOM leave+1SG home+ILL

‘I’m going home.’

To summarize the main points in this section: the traditionally defined gram-

matical subject in Finnish has evolved from a logically based notion of subject

imported from the Indo-European grammatical tradition. As a consequence it

involves a confusion of form and function. It is generally in the nominative, but

sometimes in the partitive or even in the genitive. It is generally unaffected by the

polarity of the verb, but in some instances it is affected. In this study, the gram-

matical subject is defined as a formal category: an NP in the nominative that (po-

tentially) agrees with the verb in number and person. In later chapters, I shall

argue that it is one way in which the experiential function Medium is realized.

Page 113: systematic functional linguistic

Chapter 4

Constituency and Dependency in Finnish

4.1. Overview

In the discussion of grammatical structure (or patterning) in 2.4.5 (pp. 49 ) 51), it

was pointed out that structure in SF theory is seen as a number of different types

of pattern that are conflated or mapped onto each other in the process of realiza-

tion. Thus, the term structure when applied at the rank of clause does not refer

only to constituency structure, with which this chapter will be concerned, but to

any kind of (non-random) organization that is restricted in scope to the bound-

aries of the clause. As discussed in 2.4.4 (p. 49), constituency relations are re-

garded as being primarily relevant to the experiential organization of a language.

This chapter focuses on constituency and dependency in the grammatical organi-

zation of Finnish. Sections 4.2 and 4.3 are concerned with the rank hierarchy hy-

pothesis and some of the problems that arise when this hypothesis is applied to

Finnish. Section 4.4 gives an overview of phrases in Finnish and section 4.5 dis-

cusses clause complexing in Finnish.

4.2. Ranked Constituency

The rank hierarchy hypothesis assumes that there are certain basic units, which

form a strict constituency hierarchy. While this hierarchical constituency organi-

zation might be assumed to be universal, the units themselves are language spe-

cific. If, for example, one assumes that the basic units in Finnish are the same as

in English (as, for example, Hakulinen & Karlsson (1979: 65-66) assume), then in

the basic grammatical organization of Finnish the highest ranking unit on the

constituency structure hierarchy is a clause, which consists of one or more

phrases or groups, which consist of one or more words, which consist of one or

Page 114: systematic functional linguistic

106

1 There are often complex morphophonological changes in word stems and affixes; thus,dividing a word into its constituent morphemes is not always a straightforward procedure.

more morphemes.1 While, as pointed out in 2.4.4 (pp. 48 ) 49), non-experiential

elements can be excluded from the rank scale, and varied amounts of singulary

branching and rank-shifting (downranking, embedding) have to be allowed for,

this model is meant to account for the basic experiential organization of Finnish.

The dynamics of grammar can then be accounted for by complexing, which is

seen in terms of (inter)dependency between units.

The term phrase (Finnish: lauseke) was introduced into grammatical de-

scriptions of Finnish by linguists influenced by transformational-generative gram-

mar. I shall use the term phrase is the SF sense, i.e. for the rank intermediate be-

tween clause and word. The term group will not be employed: while there are

differences between nominal phrases (NPs) and pre- and postpositional phrases

(PPs) in Finnish, these do not correspond to the difference between phrase and

group in English. Moreover, verbal and adverbial phrases also differ in crucial

ways from NPs and PPs. It seems to me that the use of a Premodifier with the

term phrase ) nominal, prepositional, verb(al) and adverbial ) suffices to distin-

guish between various types of phrases in Finnish. Thus the term VP as used in

this thesis is not to be confused with a VP in the transformational-generative

sense (where VP ÷ V + NP). Huddleston (1984: 112 ff.) also uses VP in the

sense that I am using it here; he suggests that the term “extended VP” is used to

refer to the transformational-generative notion of a VP.

A terminological distinction similar to the one made by Halliday (1985a: 83,

see Chapter 2, p. 46) between “sentence” (Finnish: virke) and “clause” (lause)

has traditionally been made by some Finnish grammarians (e.g. Ikola 1977: 126-

27). This distinction is sometimes lost in translations where there is a tendency to

translate Finnish lause as “sentence”. Throughout this thesis, lause and “clause”

are considered to be translation equivalents; the term “sentence” (virke) will be

used to refer to an orthographical unit, i.e. a unit of the written language bounded

by a capital letter at the beginning and a full stop at the end. A sentence (virke) is

Page 115: systematic functional linguistic

107

not a grammatical unit; it may correspond to one clause or more than one clause,

it may be an elliptical clause, or even a phrase or word.

The rank scale in Finnish can be illustrated by the following example:

1

Clause C * Phrase C ))))))) C ))))))))))))) C ))))))))))))))))))))) C * * * * Word C C C C ))))))))))))))))))))C * * * * * Morpheme C C)))C)))C C))))C C C ))) C ))) C ))))) C * * * * * * * * * * * (1) sä ol + i + t möki+llä viime viiko+n + loppu +na [Tel1:1]

you be+PAST+2SG cottage+ADE last week +GEN + end +ESS

‘You were at the (summer) cottage at the weekend (last weekend).’

Where branching does not occur in the example, it would have been possible for

branching to occur (e.g. mökillä ‘at the cottage’ ÷ firman omistamalla mökillä

[firm+GEN own+INF+ADE] ‘at the cottage owned by the firm’) or else the item is

in paradigmatic contrast with other items that would permit branching (e.g. sä

‘you’ ÷ sun ruotsalaiset ystävät ‘your Swedish friends’, with concomitant

changes in the form of the verb).

With many examples from dialogue, one needs to take into account the fact

that dialogue is collaborative, and what someone has just said is rarely repeated

but presupposed by ellipsis, as illustrated in the following example:

(2) – Oot sä konserteissa käyny? [TIIM3d]be+2SG you concerts+PL+INE go/visit+PTC

‘Have you been to concerts?’

– no joo sillon tällön käyn ainawell yeah now & then go/visit+1SG always.‘Well yeah now & then I always go.’

÷ – minkälaisissa konserteissa?what-kind+PL+INE concerts+PL+INE

‘What kinds of concerts?’

Figure 4-1: The Rank Scale in Finnish

Page 116: systematic functional linguistic

108

The phrase/group minkälaisissa konserteissa? ‘what kinds of concerts?’ realizes

an elliptical clause: it can only be understood as minkälaisissa konserteissa sä

käyt (~ oot käyny)? ‘what kinds of concerts do you go to (have you been to)?’

Thus, even if a turn in a conversation consists of a phrase or a word, it can often

only be interpreted on the basis of or as part of a larger experiential structure.

There are, of course, many elements in a turn in a conversation or entire

turns that cannot be considered as part of the rank scale, for example, the ele-

ments no joo ‘well yeah’ in the example above. In SF theory, these are regarded

as textual phenomena that cannot be described by the same sorts of grammatical

concepts that are being employed in this study. (For a discussion of concepts rele-

vant to text, see Halliday & Hasan 1976). From a formal perspective, the syntagm

no joo can be considered a minor clause. A distinction is made in SF theory

between major and minor clauses. Major clauses are clauses with a finite verb;

they can also be elliptical, as in example 2 above. Minor clauses are complete

syntagms without a finite verb. They include greetings, newspaper and chapter

headings, and telephone openings and closings in Finnish. For example, Leea

täällä hei [Leea here+ADE hi] ‘Hi it’s Leea’) consists of two minor clauses in

Finnish. These minor clauses, however, do not realize all of the types of meaning

discussed in this study, e.g. the option declarative vs. interrogative, and are be-

yond the scope of this study.

The rank hypothesis can be seen as capturing a kind of baseline of grammat-

ical organization as viewed from an experiential perspective. Uncomplicated

examples of the rank scale are readily found in the language of young children, in

children’s literature, and in casual conversation. While the analysis of authentic

examples is often complex and complications are bound to arise, this is only to be

expected. Given the vastness and complexity of natural language, it seems to me

that rather than see something wrong with a ranked constituency model that al-

Page 117: systematic functional linguistic

109

lows singulary branching, rank-shifting, and the exclusion of certain elements

from the rank scale, there is something wrong with a neat and almost mechanical

model of the kind suggested by Matthews (1966: 109) in his early criticism of the

rank hypothesis. However, as Halliday (1966b) and McGregor (1991) suggest, we

need to be able to delimit and define the exceptions and exclusions.

While this chapter focuses on constituency ) part-to-whole ) relationships

in the clause, since this is essential to a synoptic representation of the configura-

tions of experiential functions that are realized in the clause, it may well be that

we need a model of grammatical organization for Finnish that also incorporates a

dependency relationship between the verb and its complements. A dependency

model is more feasible when we begin to look, for example, at the scope of inter-

personal elements. In other words, the combination of dependency and constitu-

ency that Halliday employs in his analysis of English groups and phrases (see

Halliday 1985a: 170 ff.) would also apply to the clause in Finnish. As Hudson

(1987: 250) points out, each approach ) constituency (part-to-whole) and depend-

ency (part-to-part) ) has virtues that the other lacks.

The constituency approach has the virtue of identifying items larger than words ) clauses andphrases ) so that the grammar can make generalizations about these items; for instance, thisallows us to distinguish between interrogative and declarative clauses, or between relative andadverbial clauses, without having to pretend that these properties were properties of someparticular word, such as the verb ... Systemic grammars depend crucially on being able toassign features to higher nodes, ... so it is essential to have higher nodes, as in the constitu-ency approach ....

On the other hand, the dependency approach has the virtue of being able easily to capture thedependency relations between parts ...

Page 118: systematic functional linguistic

110

4.3. Some Problems with the Rank Hypothesis

4.3.1. Discontinuous Constituents

There are at least two problems that immediately arise when the rank hypothesis

is applied to Finnish. The first problem arises when any kind of constituency

analysis ) ranked or otherwise ) is applied to Finnish. It concerns what are gener-

ally referred to as discontinuous constituents, as exemplified by the verbal group

ei oo menny (standardized written: ei ole mennyt) ‘hasn’t gone’ in the following

clause:

(3) ei se varmaan sinne oo tässä jumalanNEG+3SG (s)he-NOM sure+ILL to there be this+INE god+GEN

ilmassa menny [Tel1:1]weather+INE gone+PTC

‘(S)he certainly wouldn’t have gone there in this raging storm.’

Because word order in Finnish is flexible (for the most part it is conditioned by

textual factors), examples like this are not uncommon in either spoken and written

Finnish (for some further examples and discussion, see Vilkuna 1989: 124-127,

197 ff.). Thus, many of the constituency trees in Finnish would contain branches

that crossed over each other. However, because the mode of interpretation in SF

grammar is a functional one, and grammatical structure is explained by reference

to the meaning, there is no problem in recognizing ei oo menny ‘hasn’t gone’ as

a unit that can be interpreted as a Process.

On the other hand, it can be reasonably argued, as Sammallahti (1991) does

for Saame (Lappish), that a constituency-based tree diagram is unsuitable for rep-

resenting the relationships between the elements in the Finnish clause, and that it

would be more appropriate to represent it in terms of dependencies, using arrow

notation. While a tree diagram seems a rather crude representation and fails to

capture the dependency relationships among the elements in ei oo menny ‘hasn’t

gone’, for example, it nevertheless provides a synoptic representation of the

Page 119: systematic functional linguistic

111

1 Halliday’s (1985a: Chapter 6) analysis of groups is based on both constituency anddependency, although he does not develop a way of showing dependency in a tree diagram.

clause. On the other hand, the arrow notation used in dependency grammar to link

a dependent element to its Head would be more suitable for capturing the depend-

ency relationship between elements, particularly between the elements in a dis-

continuous constituent. It would be feasible to develop a representation similar to

Hudson’s (1976, 1987) in which dependency arrows are used in conjunction with

a constituency tree diagram.1

‘(S)he certainly wouldn’t have gone there in this raging storm.’

(The modal NP varmaan ‘for sure’ is not analysed as it is outside the experiential

structure of the clause.)

As indicated in the previous section, the synoptic representation of part-

whole relationships as an entirety as reflected in a constituency tree diagram is

essential in a functional grammar, where the parts realize functions that are inter-

preted by reference to the whole. From an experiential semantic perspective, the

parts in a clause represent different variables, i.e. different experiential semantic

functions. For example, in the material process illustrated in Figure 4-2 above

there is an Actor (se ‘(s)he’), Process (ei oo menny ‘hasn’t gone’) and two non-

inherent roles (circumstances) (sinne ‘there’ and tässä jumalan ilmassa ‘in this

raging storm’). From a synoptic perspective, the other roles form a structural con-

Figure 4-2: Constituency and Dependency

Page 120: systematic functional linguistic

112

1 Ellipsis will be discussed more fully in Chapter 7.

figuration in conjunction with the verb ei oo menny ‘hasn’t gone’. The meaning

of the process realized by the verb mennä ‘(to) go’ depends on the structural en-

tirety of which it is a part, not on the isolated verb form. In other syntagms, the

verb form mennä ‘(to) go’ can realize the process in a relational process: mä

menin ihan kalpeaksi ‘I went/ turned completely pale’.

4.3.2. Inclusion of Morphology

While this study is not concerned with an in-depth analysis of the problems in-

volved in incorporating morphology ) both inflexional and derivational morphol-

ogy ) into a lexicogrammatical description of Finnish, a few general points about

morphology need to be made. As pointed out in 2.4.9 (p. 60), the lexicogrammar

of a language is seen as a continuum in which more general meanings are realized

grammatically and more specific meanings are realized lexically. Grammatical

meanings in Finnish are realized both syntactically and morphologically, and this

presents another problem for the rank hierarchy hypothesis. For example, the

finite verb in Finnish is marked for person and the indefinite verb form contains a

morpheme that realizes an unspecified human participant. This means that an

Agent can be realized morphologically: juoksin ran+1SG ‘I ran’; juostiin

‘ran+INDE (unspecified human participant(s))’. Thus, for Finnish, one needs to

recognize that, in certain instances, the VP can realize two functional roles: the

process and an inherent participant.

(4) Olin Szegedissä. (from Helasvuo 1988: 75)be+PST+1SG Szeged+INE

Process+Participant Circumstance‘I was in Szeged.’

Clauses like this are typical of written Finnish; in spoken Finnish, on the other

hand, it is typical for the pronoun to be realized, unless it is presupposed by ellip-

sis1:

Page 121: systematic functional linguistic

113

(5) Mä olin Szegedissä.I was+PST+1SG Szeged+INE

Participant Process Circumstance‘I was in Szeged.’

This poses a problem for the rank hypothesis since an essential feature of the

rank hypothesis is that it facilitates the making of generalizations about paradig-

matic and syntagmatic relations:

It defines a point of origin for systems and structures, so that the assignment of any item to agiven rank, as also the assignment of the structures and systems themselves, becomes animportant step in generalization (Halliday 1966: 112).

Thus, for example, it is generally assumed that the entry condition for transitivity

(experiential structure) is the major clause and that transitivity functions (such as

Actor, Process, Goal etc.) are realized by phrases. When singulary branching

occurs, a phrase can be made up of a single word or even a single morpheme.

Thus, words consisting of a single morpheme ) Anne, poika ‘(a/the) boy’ and mä

‘I’ ) can realize transitivity functions. Singulary branching in itself is not a prob-

lem. What is problematic in instances such as example 4 above, is the fact that a

VP in Finnish can consist of morphemes that realize different transitivity func-

tions: ol+i+n ‘be (stem) + PAST + 1SG’. In these instances, the 1. person singular

morpheme is an immediate constituent of the clause: it directly realizes a transi-

tivity function.

This means that we may have to acknowledge that upward rankshifting oc-

curs, i.e. the morpheme moves up the rank scale and functions in the way we

would expect a phrase to function. However, the instances in which upward

rankshifting could be said to occur in Finnish are clearly defined: VPs with a 1. or

2. person ending and VPs containing an indefinite morpheme. Moreover, these

exceptions are interpretable from a common sense point of view: 1. and 2. person

pronouns are so-called speech act pronouns that do not necessarily need the ex-

panded semantic potential that is afforded by the phrase. The Finnish indefinite,

on the other hand, is used to refer to (a) non-specified human participant(s), and it

would be semantically anomalous to expand it using the potential of an NP (see

Shore 1988: 160).

Page 122: systematic functional linguistic

114

1 See Karlsson (1983b: 91-93) for an outline and some discussion of possessive suffixes instandardized written Finnish; spoken Finnish is briefly discussed by Karlsson (1983b: 208).

2 I have substituted the term Entity for Halliday’s label Thing because it sounds less concrete(cf. Halliday 1985a: Chapter 6).

Similar problems arise elsewhere in the grammar of Finnish. For example, at

phrase rank in Finnish there are at least two other instances in which one would

have to acknowledge upward rankshifting: 1) with Classifiers, which always form

a compound with the Thing or Entity in an NP (see next section) and 2) with pos-

sessive suffixes, where in some instances the suffix alone realizes the function

Possessive (taloni ‘my house’), whereas in other instances it can be realized by a

pronoun in the genitive or a combination of a pronoun and a suffix (minun

vaimoni [my wife+POS/1SG] ‘his wife’; mun vaimo ‘my wife’.1

4.4. Phrases in Finnish

4.4.1. Nominal Phrases (NP)

Nominal Phrases are phrases that have a nominal as Head, i.e. the Head belongs

to a word class that can be inflected for case (see 3.3.3, pp. 82 ) 87) and number.

NPs have the same functions in a clause as nouns; however, as nouns in Finnish

are also inflected in the so-called locative cases, an NP in Finnish can also func-

tion as an Adjunct in a clause. The following are examples of NPs in Finnish:

(6) tuo upea, iso valkoinenthat-NOM magnificent-NOM big-NOM white-NOM

Deictic Epithet Epithet Epithet

puutalowood+house-NOM

Classifier + Entity2

‘that magnificent, big white wooden house’

Page 123: systematic functional linguistic

115

(7) noissa upeissa, isoissa valkoisissathat+PL+INE magnificent+PL+INE big+PL+INE white+PL+INE

Deictic Epithet Epithet Epithet

puutaloissawood+house+PL+INE

Classifier + Entity

‘in those magnificent, big white wooden houses’

(8) sielläthere+ADE

Deictic

‘there’

These examples and many of the examples used in this section are not authentic,

and may even sound contrived; moreover, I doubt that NPs with a lot of modifiers

occur very frequently in actual (spoken or written) text. However, the point of the

examples is to illustrate the potential of NPs in Finnish.

In this study, an NP is defined by its Head and the type of modifiers that it

can (potentially) have, and not by its function. An NP can function in the clause

as a participant (inherent role) or as a circumstance. This can be compared to the

approach adopted by Hakulinen & Karlsson (1979: Chapters 7 ) 8), which is

influenced by transformational grammar. They refer to Noun Phrases (NPs), with

a noun as Head, and other “nominal phrases” (nominaaliset lausekkeet). Under

the heading of nominal phrases, they include Adjective Phrases, Quantifier

Phrases, Adverbial Phrases and Pre- and Postpositional Phrases. On the other

hand, they distinguish Adverb Phrases, with an adverb as Head, from Adverbial

Phrases, which function in a clause as an adverbial (i.e. a Circumstance). This

means that they postulate seven different types of Adverbial Phrase in Finnish:

Adverb Phrase (which has an adverb as Head), subordinate clauses, Noun Phrases

in which the Head is not in one of the grammatical cases (as in 7 above), P-posi-

tional Phrases, Quantifying Phrases, Adjectival Phrases, and non-finite verb con-

structions (Hakulinen & Karlsson 1979: 151-152). Thus, in Hakulinen & Karls-

son’s analysis, example 7 above, would be an NP functioning as an Adverbial

Phrase.

Page 124: systematic functional linguistic

116

It seems to me the approach adopted in this study is clearer. It is less confus-

ing to talk about Nominal Phrases (NPs) with a nominal as Head and to consider

its function in the clause separately, e.g. as a Circumstance, Actor, Goal etc. (see

Chapter 6). Nominal Phrases (NPs) in this study are also considered to have vari-

ous subtypes. The subtypes considered in this section are roughly equivalent to

what Hakulinen & Karlsson refer to as noun, adjective and quantifier phrases.

As illustrated in the examples above, Premodifiers in an NP agree with the

Head in number and case. A Classifier, however, always forms a compound with

the Head: puutalo (puu ‘wood’ + talo ‘house’) ‘wooden house’, pikajuna (pika

‘express’ + juna ‘train’) ‘express train’. Halliday (1985a: 171) analyses an Eng-

lish NP (e.g. a magnificent ornamental eighteenth-century carved mahogany

mantelpiece) as a Head with Premodifiers. If it is also assumed that an NP in

Finnish can be analysed as a logical structure as well, then we can analyse an NP

in Finnish as follows, where Greek letters are used to mark a hypotactic (depend-

ency) relation between elements:

(9) tuohon upeaan, isoon valkoiseen puu + taloonthat+GEN magnificent+ILL big+ILL white+ILL wood house+ILL

. , * ( $ "

‘to that magnificent, old white wooden house’

However, it could be argued that the relationship amongst the constituents in

the examples given so far is not one of logical dependency. While it could be

argued that the Entity, talo ‘house’ (or the combination of Classifier and Entity,

puutalo ‘woodhouse’) is in some sense the most substantial or tangible from an

experiential semantic point of view ) it is the element in the NP that is most

clearly anchored in the world of our experience ) the grammatical evidence sug-

gests that, in Finnish, the other elements in the NP are not dependent on it. It is

significant that, as illustrated above, each element in an NP inflects independently

for case (in both standardized written and unselfconscious spoken Finnish). This

could be seen as an indication that each element has its own independent status

Page 125: systematic functional linguistic

117

1 I have adopted Huddleston’s (1984) convenient notation of using angle brackets that encloseletters to represent different speakers.

within the NP, i.e. that tuohon ‘(to) there’ is not subordinative to puutaloon ‘to

the wood(en)house’. This is reflected by the fact that if the Entity is omitted, un-

like English, nothing is needed in Finnish to fill its place, as illustrated by the

following dialogues:1

(10) <A> missä Kalle asuu?which+INE Kalle live‘Where does Kalle live?’

<B> tuossa valkoisessa talossathat+INE white+INE house/building+INE

‘in that white house’

(11) <A> missä talossa Kalle asuu?which+INE house/building+INE Kalle live‘In which house does Kalle live?’

<B> tuossa valkoisessathat+INE white+INE

‘in that white one’

<A> vai tuossa se asuuor-Q that+INE he lives‘So that’s where he lives [so there (in that) he lives]’

On the other hand, there are arguments that indicate that the Entity is (typi-

cally) the Head (unless it is presupposed by ellipsis). For example, Huddleston

(1984: 110) in his analysis of English, points out that when phrases combine to

form larger units, there are often restrictions that affect the Head, but not the de-

pendents. Thus, for instance, it is the Entity/Head that determines the case-form

of a Post-Modifier, as in the following example:

Page 126: systematic functional linguistic

118

(12) [Professori Lagerspetz ei ole lähtenyt sotaan iskeäkseen pöytään -

‘Professor Lagerspetz has not taken up the issue [lit. gone to war] in order to establish [puton the table]]-‘

oman suurenmoisen teoriansa ihmisenown+GEN grandiose+GEN theory+POSS/3 person+GEN

aggressiivisuudesta. [HKV]aggressiveness+ELA

‘her own grandiose theory of the aggressiveness of humankind.’

If the Entity had been realized by suhtautuminen ‘relation(ship)’, for example, if

one were talking about Professor’s Lagerspetz’s relationship (the way she relates

to) aggressiveness, then the Post-Modifier would be in the illative suhtautuminen

aggressiivisuuteen ‘relation [the way in which one relates to] aggressiveness’.

However, it seems to me that Post-Modifiers are grammatically distinct from pre-

modifiers, and while it may be valid to talk of a Modifier and a Head in

postmodification, the same does not apply to the elements that precede the Entity.

As discussed in 3.4.1 (p. 89 ff.), however, it is possible for there to be a

Head-Modifier relationship in the nominal phrase, and in these instances, the

modifier is marked by the genitive in Finnish. Thus in the following example,

vanhalle omistajalle ‘to the old owner’ is modified by tuo valtavan iso talo ‘that

extremely big house’, in which iso ‘big’, in turn, is modified by valtavan ‘ex-

tremely’.

(13) tuon valtavan ison talon vanhalle omistajallethat+GEN huge+GEN big+GEN house+GEN old+ALL owner+ALL

__________ ______$ "SUB-MODIFIER SUB-HEAD

_________________________________________________ _____________________

$ "MODIFIER HEAD

‘to the old owner of that extremely big house’

Page 127: systematic functional linguistic

119

The submodification of iso ‘big’ by valtava ‘huge’ in this example is not obvious

as the Sub-Head iso ‘big’ is part of another Modifier. If it were part of an inde-

pendent phrase, however, the submodification would be clear: tuo valtavan iso

talo ‘that extremely big house/building’.

Pace Huddleston (1988: 144), SF theory does recognize a phrase (or in

Halliday’s terms, a group) with an adjective as Head, i.e. an Adjectival Phrase.

However, an Adjectival Phrase is considered a subtype of nominal phrase. In

other words, the difference between the two is a matter of delicacy. I shall not

discuss Adjectival Phrases in Finnish. However, it needs to be pointed out that

while adjectives in Finnish can be distinguished from nouns in that they have

comparative and superlative forms, the distinction between nouns and adjectives

in Finnish is less clear-cut than it is in English. Both nouns and adjectives are

inflected for number and case. The use of the comparative is not restricted to ad-

jectives in Finnish. In some instances, nouns can occur in the comparative: e.g.

ranta ‘shore’, rannemmaksi [shore+comp+tra] ‘more towards the shore’; syksy

‘autumn’, syksympänä [autumn+comp+ess] ‘(during) later on in the autumn’ (see

L. Hakulinen 1978: 115-116).

There are various infinitives, gerunds and participles in Finnish (see 3.3.3, p.

85 ff.), and these can function as the Head or as a Modifier in an NP. For exam-

ple:

(14) tuossa palaneessa talossa (ei ollut ihmisiä.)that+INE burn+PST/PTC+INE house+INE (were not people)‘in that burnt-down house (there were no people)’

(15) hyväksytyt (saavat aloittaa ensi viikolla).accept+INDE+PST/PTC-NOM/PL

‘those accepted (can begin next week).’

Non-finite verb forms are Janus-like (double-faced) in that, on the one hand, they

function like other types of constituents in an NP, but, on the other hand, they can

occur with elements realizing Participants and Circumstances that are generally

associated with the Process in a clause. Because of this, it seems to me that it is

valid to analyse these non-finite verb forms in two ways: those instances in which

Page 128: systematic functional linguistic

120

the non-finite form occurs alone can be treated like ordinary Heads or Modifiers

in an NP, whereas those instances in which the non-finite form is accompanied by

a Participant or a Circumstance can be treated as instances of embedding, i.e. it is

a non-finite clause that functions within the structure of the host NP. For example,

the participle and the circumstance that occurs with it in the following NP can be

regarded as a non-finite clause embedded within the host NP:

(16) [[ viime syksynä palaneessa ]] talossalast autumn+ESS burn+PST/PTC+INE house+INE

‘in the house that burnt down last autumn’

Particularly in written Finnish, the various types of dependencies and

embeddings within an NP can become quite complex:

(17) monissa [[ hänen Islannin matkansa tuloksinamany+PL+INE her+GEN Iceland+GEN trip+POS/3 result+PL+ESS

syntyneissä ]] töissään [HKV]born+PAS/PTC+PL+INE work+PL+INE+POS/3

‘in many of her works (of art) that came about as a result of her trip to Iceland’

In this example, there is a dependency relationship between hänen ‘her’ and

Islannin matkansa ‘Iceland trip’ and the whole NP hänen Islannin matkansa

‘her Iceland trip’ is dependent on tuloksina ‘result’. The NP hänen Islannin

matkansa tuloksina ‘the results of her Iceland trip’, in turn, is part of an embed-

ded non-finite clause hänen Islannin matkansa tuloksina syntyneissä ‘born as

the results of her Iceland trip’ in the host NP with töissään ‘in her works (of art)’

as Head.

Page 129: systematic functional linguistic

121

1 As discussed in Chapter 7, I make a distinction between non-realization, which iscontextually conditioned, and ellipsis, which is a (co)textual phenomena.

In many NPs, it is the Entity (` nominal) that functions as the Head. How-

ever, as mentioned above, an adjective is the Head in an Adjectival Phrase and

there are also NPs, in which an Epithet, Numerative or Deictic is the Head. These

are generally either elliptic, i.e. the Entity can be retrieved from the preceding

(spoken or written) text, or it is dependent on the (extralinguistic) context for its

interpretation.1

There is, however, a fairly clearly defined subtype of NP in Finnish where a

numerative or a measure is the Head. These could be referred to as Quantifying

(Nominal) Phrases: for example, many phrases that contain a numerative greater

than one (e.g. nuo kaksi upeaa valkoista taloa ‘those two magnificent white

houses’), a measure (kuppi kahvia ‘a cup of coffee’), or certain other quantifying

elements (paljon työtä ‘a lot of work’). Hakulinen & Karlsson (1979: 144 ff.)

have referred to a similar set of phrases as Quantifier Phrases (kvanttorilauseke).

However, Hakulinen and Karlsson’s Quantifier Phrases do not include phrases

with a measure as Head (kuppi kahvia ‘a cup of coffee’), since their definition of

a Quantifier Phrase is based on the logical notion of a quantifier, and a word like

kuppi ‘cup’ or litra ‘litre’ “does not have scope” (Hakulinen & Karlsson 1979:

144). The Quantifying Nominal Phrase being proposed here is one in which 1) the

Head is a quantitative expression, i.e. a word expressing an exact or inexact quan-

tity, or a measure expression (see Halliday 1985a: 163,174) and 2) the Modifier

is realized by a nominal in the partitive or elative case.

Page 130: systematic functional linguistic

122

(18) puolet heistähalf-NOM/PL they+PL+ELA

Head ModifierNumerative Entity

‘half of them’

(19) kuppi kahviacup-NOM coffee+PAR

Head ModifierMeasure Entity

‘a cup of coffee’

(20) kymmenen kilometriäten-NOM kilometre+PAR

Head ModifierNumerative Entity

‘ten kilometres’

As illustrated by these examples, while the Measure or the Numerative is the

Head, the Modifier in the partitive realizes the Entity, which is the most salient

from an experiential semantic perspective. Thus, a Quantifier Phrase is a subtype

of NP in which the Head and Entity do not coincide. This is one of the many in-

stances in this study which illustrates how different functional structures contrib-

ute to the interpretation of linguistic phenomena (cf. Halliday’s analogy of poly-

phonic music in 1.1 (p. 3) and Firth’s analogy of the prism in 2.2.5. (p. 20)). A

similar kind of thing happens with pre- and postpositional phrases, discussed in

the next section.

Page 131: systematic functional linguistic

123

4.4.2. Pre- and Postpositional Phrases (PP)

Pre- and postpositional phrases (PPs) could also be regarded as a subtype of Nom-

inal Phrase. As Finnish has both pre- and postpositions, I shall use the cover term

“p-position” and refer to pre- and postpositional phrases as PPs. A preposition

precedes the nominal (e.g. kohti Singaporea ‘towards Singapore’) whereas a

postposition follows the nominal (e.g. Singaporen jälkeen ‘after Singapore’).

Nominals that are case-marked in one of the locative cases (e.g. Singaporessa ‘in

Singapore) are NPs in Finnish.

Historically, the vast majority of pre- and postpositions in Finnish have de-

veloped from inflected nominal forms (L. Hakulinen 1979: 501), and, this is evi-

denced by a number of features in contemporary Finnish. The case-ending of a p-

position is often evident, although in some instances the cases are used with the

meanings they had at earlier stages in the development of Finnish: the translative

(-ksi) , the essive (-na/nä) and the partitive (-(t)a/(t)a) originally had locative

meanings in Finnish (see, e.g., L. Hakulinen 1979, Leino 1990). The nominal-like

nature of p-positions is also reflected in the fact that a possessive suffix can be

attached to many of them, e.g. (minun) edessä+ni [my in-front-of+POS1] ‘in front

of me’, (sinun) takanasi [your(sg) behind+POS/2SG] ‘behind you’ (cf. nominals

with a possessive suffix (minun) talo+ni ‘my house’).

PPs in Finnish are hypotactic (Head-Modifier) structures: the p-position is

the Head and the other constituents are either in the (singular or plural) genitive

or in the partitive case. As illustrated in the next example, with some p-positions,

the grammar of Finnish requires the speaker to use a locative case-ending to spec-

ify location at a position or direction towards or away from a position.

Page 132: systematic functional linguistic

124

1 takana ‘[at] behind’, takaa ‘from behind’, taakse ‘to behind’.

2 For a list of the most common Finnish pre- and postpositions, see Karlsson 1983b.

(21) LOCATION AT A POSITION:

a. tuon ison pöydän alla/ edessäthat+GEN big+GEN table+GEN under+ADE/ in front of+INE

‘under/in front of that big table’ [Cf. 13 above]

MOVEMENT TO(WARDS) A POSITION:

b. tuon ison pöydän alle/ eteenthat+GEN big+GEN table+GEN under+ALL to front+ILL

‘(to) under/in front of that big house’

MOVEMENT (AWAY) FROM A POSITION

c. tuon ison pöydän alta/ edestäthat+GEN big+GEN table+GEN under+ABL/ from front+ELA

‘from the front of/from behind that big table’

(22) noiden talojen taakse1

that+PL+GEN house+PL+GEN to behind‘(to) behind those houses’

(23) noita isoja taloja kohtithat+PL+PAR big+PL+PAR house+PL+PAR towards‘towards those big houses’

As illustrated in these examples, all of the submodifying elements agree with each

other in number and case. Prepositions are not as numerous as postpositions.

While postpositions tend to occur with the genitive, prepositions tend to occur

with the partitive.2

(24) ennen toista maailmansotaabefore second+PAR World+War+PAR

‘before World War II’

(25) alle normaalipainonbelow normal+weight+GEN

‘below normal weight’

Page 133: systematic functional linguistic

125

1 There is a slight inconsistency in Hakulinen & Karlsson’s (1979) analysis in that on p. 107a p-positional phrase is said to be one in which a pre- or a postposition is the Head.

As stated earlier, it seems to me that the p-position is the Head in a PP: 1)

the genitive in Finnish is a clear indicator that the unit in question is subordinated

to another unit (see 3.4.1 p. 89 ff.) and 2) a possessive suffix can be added to p-

positions. However, this evidence only relates to those p-positions in which the

modifier is in the genitive. On the other hand, the criteria that Huddleston (1984:

109-111) appeals to in establishing the Head of a phrase would indicate that it is

also the p-position that is the Head when the modifier is in the partitive. For ex-

ample, it is the p-position that determines the case-form of the dependent NP: in

24 above, for example, ennen toista maailmansotaa ‘before World War II’ it is

the preposition ennen that determines that the Modifier is in the partitive. There is

nothing inherent in the NP toinen maailmansota ‘World War II’ that indicates

that it must be in the partitive; with the postposition jälkeen ‘after’ the NP would

be in the genitive.

However, there is disagreement amongst Finnish linguists as to which ele-

ment is the Head in a Finnish PP. For example, L. Hakulinen (1979: 501), regards

what he call the genitive attribute (i.e. what I have referred to as the dependent

NP in the genitive) the Head; although historically the postposition was the Head.

Hakulinen & Karlsson (1979: 153) regard PPs as exocentric constructions (see

Bloomfield 1933: 194) in which neither element is Head:1 if either element is

omitted the syntactic status of the construction is changed. As Hakulinen &

Karlsson point out, this does not apply to instances in which a possessive suffix is

attached to a p-position ((minun) edessäni [(my) front+INE+POS/1SG] ‘in front of

me’); but they simply relate these instances to other constructions in which a first

person suffix occurs: (minun) koirani [(my) dog+POS/1SG] ‘my dog’, (minä)

nukun [(I) sleep+1SG] ‘I’m sleeping’. Simply relating these instances to other

Page 134: systematic functional linguistic

126

1 Cf. Salmi (1990: 14), who in terms of Cognitive Grammar, regards examples like this asinstances in which the landmark is unspecified. Salmi, however, also refers to these asadverbs.

2 This, of course, corresponds to the traditional analysis of words like before in English in thecontexts before the altar and She had never been to Finland before. The former is consideredto be a preposition, the latter an adverb.

similar phenomena does not explain what is a clear exception to their analysis,

moreover it involves a contradiction in that it is clear that in their analysis of NPs,

NPs such as (minun) koirani ‘my dog’ would be regarded as endocentric (Head-

Modifier constructions).

Furthermore, it is quite common for the p-position to occur alone; Hakulinen

& Karlsson (1979: 85) say that instances are plentiful (“runsaasti”) in Finnish,

and cite a number of examples, but they regard these as instances of homonymy:

if, for example, vieressä ‘beside, at the side of’ occurs without a nominal in the

genitive, it is regarded as an “adverb”. The following examples are from Hakuli-

nen & Karlsson (1979: 85):

(26) a. Koira kulkee sokean vieressä.dog move+3SG blind+GEN side/beside+INE

‘The dog is walking beside the blind person.’

b. Koira kulkee vieressä.dog move/go+3SG side/beside+INE

A translation has not been provided for 26b because something like “the dog is

walking beside/at the side” would be strange and unidiomatic in English. A

translation that captures the completeness of the Finnish would be “the dog is

walking nearby”, but this translation is closer to the p-position with the stem

lähe- ‘near’ in Finnish. There is nothing odd about 26b in Finnish: the entity that

the dog is walking beside is simply not specified.1 Hakulinen & Karlsson’s analy-

sis of vieressä in 26b as an adverb is the traditionally accepted analysis in Fin-

land, and it is adopted by Nykysuomen sanakirja (Contemporary Finnish Dictio-

nary).2

Page 135: systematic functional linguistic

127

The fact that p-positions without a modifier have been referred to as adverbs

seems to stem from the fact that word-classes or parts of speech in Finnish have

been based on traditionally recognized Indo-European parts of speech, which

were first established for Greek and Latin. In Priscian’s grammar (Robins (1967:

57-58, 64), for example, an adverb is an unchangeable form used with a verb and

a preposition is a separate word that occurs before case-inflected words. Stem-

ming from this Latinate model is the generally accepted traditional view that ad-

verbs are a word class used to qualify verbs, adjectives and other adverbs and

prepositions are a word class used together with a noun or NP (e.g. Hartmann &

Stork 1972). Thus, the class of adverbs in Finnish is made up of a heterogenous

set of forms that modify a verb or an adjective: Karlsson (1983b: 184-186), for

example, includes adverbs like hitaasti ‘slowly’ that end in -sti and have compar-

ative and superlative forms (hitaammin, hitaimmin) and focus particles (see

7.2.3) like aika ‘quite, rather’ and juuri ‘just’ as well as a number of words that I

refer to as p-positions without a modifier.

If, on the other hand, one takes the view that grammatical categories are

abstractions based on the interrelations and oppositions within a language (see

2.2.4), then there is nothing anomalous about referring to a p-position without a

modifier. A p-position is not an Indo-European preposition that can also come

after the modifying word; it constitutes a class of words in Finnish that can be

defined by certain formal and functional characteristics, and this word class con-

trasts, for example, with bound morphemes such as -ssa/ssä (e.g. talossa in the

house) and adverbs that end in -sti like hitaasti ‘slowly’. I shall take up the issue

of p-positions without modifiers again later on in this section.

In contrast to Hakulinen & Karlsson (1979), Kangasmaa-Minn (1991: 6)

refers to a postposition (with a nominal in the genitive) as the Head, but she says

that the semantic information is concentrated in the Modifier. Kangasmaa-Minn’s

analysis comes closest to the analysis that is proposed in this study. Moreover, as

I shall illustrate shortly, a functional interpretation explicates this dual interpreta-

tion of PPs. It should be clear from the discussion so far that PPs in Finnish are

reminiscent of what Halliday (1985a: 174) refers to as Facet expressions in Eng-

Page 136: systematic functional linguistic

128

lish (the front of/the side of the house), which can function as complex preposi-

tions (in the front of/by the side of the house). This is obviously a reflection of

the nominal origin of many p-positions in Finnish. However, as the term Facet is

rather too concrete for many of the temporal and other non-spatial uses of p-posi-

tions, I use the term Orientation, in other words, a p-position expresses orienta-

tion towards an Entity, e.g. a physical object or a temporal event. The label Orien-

tation is based on Leino’s (1990: 133 ff.) analysis of p-positions in Finnish. Leino

employs concepts from Cognitive Grammar in his analysis. He shows how loca-

tive case-endings express spatial relations in “basic physical space”; whereas

p-positions express relations in “oriented physical space”.

With NPs in Finnish, the Entity and the Head are generally conflated;

whereas in PPs the nominal in the genitive/partitive is the Entity, and the most

salient from an experiential semantic perspective, but it is the p-position that is

the Head.

(27) ennen toista maailmansotaabefore second+PAR World+War+PAR

Orientation Numerative EntityHead Modifier

‘before World War II’

(28) noiden talojen taaksethat+PL+GEN house+PL+GEN to behind

Deictic Entity Orientation (Facet), Modifier , Head

‘(to) behind those houses’

This last example (28) can be given a clause context such as the following: Vanki

juoksi noiden talojen taakse ‘The prisoner ran behind those houses’. The p-posi-

tion taakse ‘(to) behind’ has a translative ending (in this instance, an archaic -kse

and not -ksi), which expresses what in Finnish is a basic location or direction (‘to

inside (something)’). However, the destination is an abstract physical space (‘be-

hind/at the back of’), which is oriented towards something else, i.e. nuo talot

‘those houses’.

Page 137: systematic functional linguistic

129

In those instances in which the p-position occurs without a modifier (e.g.

26b above Koira kulkee vieressä ‘The dog is walking beside [i.e. beside the blind

person]’), the Entity that the p-position is a facet of or the Entity that is oriented

towards is not explicated. P-positions without modifiers are reminiscent of what

Hasan (1984a) refers to as implicit semantic style in language: the interpretative

source is either in the co-text or in the situation. The Entity does not need to be

explicated because it is taken for granted that it is interpretable.

(29) [ ) Minä näin uunien luota kun raahasit Jylhän tänne. Huomasiko Taavetti?

‘From beside the furnaces I could see you dragging Jylhä here. Did Taavetti noticeanything?’

) Ei.‘No.’]

Salin huokasi ja meni taakse. [TSM:19]Salin sighed+PST/3SG & went+PST/3SG to behind/back‘Salin sighed and went to the back/rear.’

This example is from a novel. It is clear from the co-text that Salin went to the

back of the storeroom where Jylhä had been taken.

As pointed out above, p-positions used alone are common in Finnish, but

they have been classed as adverbs. Thus ennen ‘before’ is classed by Nykysuo-

men sanakirja (Contemporary Finnish Dictionary) as an adverb in the context

Oli ennen ukko ja akka [was before old-man and old-woman] ‘There once was an

old man and an old woman’ but as a preposition in ennen joulua ‘before Christ-

mas’. Similarly takana ‘(at) behind/at the back of’ and edessä ‘before/in front of’

are classed as a postpositions in 30a, but are adverbs in 30b:

Page 138: systematic functional linguistic

130

(30) “POSTPOSITION”:

a. Edessämme oli satama,before/front+INE+POS/1PL was+PST/3SG harbour

takanamme aava meri. [NS]behind/back+ESS+POS/1PL open sea

‘Before us was the harbour, behind us the open sea.’

“ADVERB”:

b. Edessä oli vihainen sonni,before/front+INE was+PST/3SG angry bull

takana jyrkkä kallio. [NS]behind/back+ESS steep rock/cliff‘There was an angry bull in front, and a steep cliff behind.’

However, these unmodified p-positions are quite unlike adverbs like

nopeasti ‘quickly’ or words like heti ‘immediately’ or joskus ‘sometimes’. Apart

from the fact that they do not have a genitive or partitive Modifier, they pattern in

the same way as p-positions. As discussed earlier, many p-positions have locative

case-endings which express direction or location in basic physical space; this is

also true of unmodified p-positions. Adverbs formed from adjectives, on the other

hand, have comparative and superlative forms. Thus the adjective selvä ‘clear’

has the comparative selvempi ‘clearer’ and the superlative selvin ‘clearest’. The

adverb selvästi ‘clearly’ has the comparative selvemmin ‘more clearly’ and the

superlative selvimmin ‘most clearly’. While some unmodified p-positions also

have comparative and superlative forms, some nouns in Finnish (see footnote on

p. 119) can also occur in the comparative. Moreover, like these nouns, p-positions

in the comparative must have a locative case-ending (edempänä ‘more/further

towards the front’, kevää+mpä+nä ‘more/further into the spring’), whereas an

adverb never has a locative case-ending. As illustrated in the example above,

unmodified p-positions function in the same way in a clause as modified p-posi-

tions: they both function as Circumstances. Moreover, it would be possible to

expand the “adverb” in 30b to veljesten edessä ‘in front of the brothers’ (the ex-

Page 139: systematic functional linguistic

131

ample obviously refers to an episode from the first novel written in Finnish

Seitsemän veljestä ‘The Seven Brothers’).

Like p-positions, unmodified p-positions are interpretable in terms of a refer-

ence point in time or space, i.e. in terms of what is referred to as a “landmark” in

Cognitive Grammar. Interestingly enough, a similar interpretation is implied in at

least one traditional grammar of Finnish (Ikola ed. 1977: 66), which gives

suhdesanat ‘relation words’ (or possibly ‘orientation words’) as a cover term for

postpositions and prepositions. With unmodified p-positions the point of refer-

ence is implicit: it is taken for granted that it is interpretable from the co-text or

the situation. The re-analysis of p-positions presented in this section underscores

the point that the study of language must always be based on the study of lan-

guage in context, and this applies even when the focus of study is an abstract

word class. This has always been stressed by Halliday (1978: 28-29), in whose

view any account of language must build in the situation, otherwise it will be

artificial and unrewarding (see 2.3.7, p. 34 ff.).

4.4.3. Verb Phrases (VP)

An in-depth analysis of VPs in Finnish is beyond the scope of this thesis: accord-

ing to Karlsson (1983c: 357), for example, a verb that has a regular inflection in

Finnish can have 528 finite forms. Much of the structure of the VP in Finnish is

realized in bound morphemes. From the point of view of the functional analysis

of the clause that is presented in subsequent chapters, suffice to say that a verb

phrase (VP) functions as the process in a clause. It is either a single word, a verb,

or it is expanded phrase that functions in the clause in the same way as a verb. A

VP can be illustrated by the following example:

Page 140: systematic functional linguistic

132

(31) en olisi uskonutNEG+1SG be+CON+3SG believe+PST/PTC

Polarity: Aux: Eventnegative Tense: past/

Modality: conditional

‘I wouldn’t have believed (it)’

As illustrated by this example, and as discussed earlier in section 4.3.2, an inher-

ent participant can be realized as a bound morpheme in the VP.

According to Karlsson (ibid.), a verb with a regular inflection also has 324

infinitive forms and 11 000 participial forms. The vast majority of these forms do

not function in the structure of the VP but function like ordinary nominals (see

3.3.3, p. 84 ff.).

(32) a. lähden uimaanleave+1SG swim+3INF+ILL

‘I’m going swimming.’

Cf. b. lähden kauppaanleave+1SG shop+ILL

‘I’m going to the shop’

(33) a. tulen uimastacome+1SG swim+3INF+ELA

‘I’ve been swimming (I’ve come from swimming).’

Cf. b. tulen kaupastacome+1SG shop+ELA

‘I’ve been to the shop (I’ve come from the shop).’

(34) a. olin uimassabe+PAS+1SG swim+3INF+INE

‘I was swimming.’

Cf. b. olin kaupassabe+PAS+1SG shop+INE

‘I was at the shop’

Page 141: systematic functional linguistic

133

1 In Finnish, this means that it is not inflected in one of the grammatical cases (see Chapter3). There is a problem in Finnish, however, with what I refer to as the (spatial or temporal)Extent of the process (see 6.5 and 6.8), which is case-marked like an inherent role orparticipant but functions more like a Circumstance.

4.4.4. Adverbial Phrases (AdvP)

An adverbial phrase has an adverb as Head. Thus an Adverbial Phrase corres-

ponds to what Hakulinen & Karlsson (1979: 83,151-152) refer to as an Adverb

Phrase (adverbilauseke), whereas for them an Adverbial Phrase (adverbiaalilau-

seke) is one that functions as an Adverbial in a clause (see p. 115 above). Accord-

ing to Hakulinen & Karlsson (1979: 83), there are two criteria by which adverbs

can be distinguished from nominals: 1) nominals can have premodifiers that agree

with it in case and number (näissä taloissa ‘in these houses’) whereas an adverb

cannot (hitaasti ‘slowly’) and 2) adverbs can be modified by what Hakulinen &

Karlsson refer to as an “ad-adjective”, i.e. an adverb like melko ‘quite’ that can-

not occur alone but only as a submodifier of another adverb (melko hitaasti ‘quite

~ fairly slowly’, melko usein ‘quite often’).

The following are examples of Adverbial Phrases in Finnish:

(35) yllättävän hitaastisurprising+GEN slowly‘surprisingly slowly’

(36) yllättävän useinsurprising+GEN often‘surprisingly often/frequently’

An Adverbial Phrase functions in a clause as a Circumstance. Circumstances

in Finnish can also be realized by Pre- and Postpositional Phrases and by a Nomi-

nal Phrase in which the Head is inflected in one of the oblique cases.1

Page 142: systematic functional linguistic

134

4.4.5. A Further Note on Unmodified P-Positions

As discussed in 4.4.2, I do not consider unmodified p-positions to be adverbs.

There is another small set of words that have traditionally been regarded as ad-

verbs, but, it could be argued, are more like unmodified p-positions (“rela-

tion/orientation words”). These are forms which, like unmodified p-positions,

have limited case-inflection. The following examples are listed by Karlsson

(1983b: 156) as “adverbs of place”.

kotona ‘at home’, kotiin ‘[to] home’, and kotoa ‘from home’;ulkona ‘outside’, ulos ‘out, [to] outside’, ulkoa ‘from outside’;alhaalla ‘down, below’, alhaalle ‘[to] down, below’, alhaalta ‘from down, below’;täällä ‘here’, täältä ‘(from) here’ tänne ‘[to] here’;tuolla ‘there’, tuonne ‘[to] there’ tuolta ‘(from) there’;siellä ‘there (unspecified place; place beyond view)’, sinne ‘[to] there’, sieltä ‘(from) there’kaikkialla ‘everywhere’, kaikkialle ‘[to] everywhere’, kaikkialta ‘from everywhere’

It seems to me that the traditional classification of these as adverbs distorts the

grammatical organization of Finnish. The consistent three-way case-inflection

that is characteristic of the vast majority of p-positions in Finnish distinguishes

these from the other adverbs listed by Karlsson.

As suggested in 4.4.2 above, the forms above seem to have been classified

as adverbs because p-positions in Finnish have been related to prepositions in

Indo-European languages. If, on the other hand, we define p-positions language

internally as (fossilized) nominal forms with limited case-inflection which, in

concrete instances at least, locate something in physical space relative to some-

thing else, then it seems more feasible to regard these as p-positions in which the

reference point is not explicated since it is taken for granted that it is interpret-

able. Thus, it would be highly unlikely that most of the words listed above could

ever occur with a genitive or partitive Modifier: e.g. *talon ulkona [house+GEN

outside], *kaupungin kaikkialla [town+GEN everywhere].

Furthermore, if we compare the following sets of examples, then it is clear

that modified p-positions are specifically anchored (in time or space) by the modi-

fier:

Page 143: systematic functional linguistic

135

1 Both modified and unmodified p-positions could be anchored by NP functioning as aCircumstance of Place e.g. kaupungissa ‘in (the) town’, but this is irrelevant to the pointbeing made here.

(37) UNMODIFIED P-POSITION:

kaikkialla ~ ulkona ~ alhaallaeverywhere outside down below

(38) MODIFIED P-POSITION

sillan alla ~ sillan kohdalla ~ siltaa päin 1

bridge+GEN below bridge+GEN point bridge+PAR towards

The unmodified p-positions are not anchored by a genitive, but interpreted

deictically. They are similar to what Halliday & Hasan (1975) refer to as demon-

strative reference items or what in logically oriented linguistic descriptions are

referred to as demonstrative deictics (Larjavaara 1990) or place-deictic words

(Levinson 1983 : 79 ff.). The reference point for these words is a deictic centre

that is never explicated, because it is always taken for granted. From this perspec-

tive, it no longer seems incongruous to refer to these as (unmodified) p-positions

(“orientation words”). The exception to this is kotona ‘at home’, which can be

expanded , e.g. kotonamme ‘at our home’, (hänen) vanhempiensa kotona ‘at

her/his parent’s home’. The fact that it is an exception seems natural since the

stem koto- ‘home’ is lexically more specific than the stems of the other words

listed above. It refers to something concrete, rather than to a location relative to

something else. However, the essentially deictic character of kotona ‘at home’ is

evident from the fact that the declarative Lähden kotiin ‘I’m going home’ refers

to the speaker’s home and the interrogative Lähdetkö kotiin ‘Are you going

home’ refers to the addressee’s home. In this respect it is different from other

NPs. For example, with lähdetkö autolle ‘are you going to the car’ and lähden

autolle ‘I’m going to the car’, in either instance it could be my car, your car or

someone else’s car, depending on the situation.

In other words, what I am proposing is that the “adverbs” listed above are a

subclass of p-positions not a subclass of adverbs. It also seems to me that other

Page 144: systematic functional linguistic

136

traditionally recognized subtypes of the category “adverb” in Finnish need to be

reconsidered. An obvious subset that differs grammatically (morphosyntactically)

and semantically from other “adverbs” are what Iivonen et al (1987: 234) refer to

as “focus particles”, e.g. juuri ‘just, precisely’, jopa ‘even’. The textual function

of a focus particle is discussed in 7.2.3.

The discussion above does not pertain to ordinary NPs like alapuoli ‘lower

~ under side’ that have a similar directional meaning when they are inflected e.g.

alapuolella. These are generally not classified as adverbs but as nominals. They

can occur with a genitive modifier kosken alapuolella ‘(in the area) below the

falls’. They can also occur without a modifier, if so the modifier is a deictic centre

in the situation that is not realized in the linguistic structure: alapuolella ‘(in the

area) below (someone ~ something)’. On the other hand, the distinction between

this kind of nominal and a postposition is not a clear-cut one. This seems to be a

reflection of the fact that, as pointed out in 4.4.2, postpositions appear to have

developed from nominal forms.

Page 145: systematic functional linguistic

137

4.5. Preliminary Analysis of Clause Complexes

4.5.1. General Comments

The complexing of clauses ) the way in which a speaker or writer combines

clauses ) straddles the area between grammatical organization and the organiza-

tion of a text. Thus we can approach the way in which clauses can be combined

from two perspectives. We can approach it from the point of view of text seman-

tics, i.e. how parts of the text are related to each other in meaningful ways. The

other approach is to build upon the constituency relationships discussed in the

previous sections of this chapter. This is the approach that will be taken here.

The organization of grammatical units in terms of a rank hierarchy serves to

define the upper limit of grammatical organization. Thus, grammatical units can

be ranked from the lowest unit on the rank scale, the morpheme, to the highest

unit, the clause. A constituency tree diagram showing the relationships between

the units on the rank scale could be drawn for the following clauses, both of

which were taken from an actual text. The first from a written text, the second

from a spoken one.

(39) hänen kesäkotinsa nimi on Leporello [HKV]his/her+GEN summer+home+POS/3 name-NOM is+3SG Leporello-NOM ‘the name of his/her summer home is Leporello (Restarello)’

(40) mä oon ollu niin semmonen rapakuntonen ... [Tel1: 9]I-NOM be+1SG be+PTC so that/such kind-NOM weathered+condition-NOM

‘I’ve been in such poor shape.’

A clause, like other units on the rank scale, can be expanded into a complex.

When clauses are combined into complexes, this is done in a meaningful way. As

Halliday (1985b: 82) points out, if something is “being represented as a complex

phenomenon, or as a set of interrelated phenomena”, then this relationship also

has to be brought out.

Page 146: systematic functional linguistic

138

The examples above, were, in fact, parts of a larger complex. Both were part

of a larger set of interrelated phenomena. Clause boundaries are marked by dou-

ble slashes, as follows:

(41) 2 Sanoitko, 2 että hänen kesäkotinsa nimisay+PST+2SG+Q that his/her+GEN summer+home+POS/3 name-NOM

on Leporello, 2 hekotin. 2 [HKV]is+3SG Leporello-NOM laugh/chortle+PST+1SG

‘”Did you say that the name of his/her summer home is Leporello (Restarello)”, Ichortled.’

(42) 2 kyllä se vois olla kauheeta 2yes/certainly it-NOM can+CON+3SG be+INF awful+PAR

mutta ku mä oon ollu niin semmonenbut as I-NOM be+1SG be+PTC so that/such kind-NOM

rapakuntonen ja huono-olonen kaiken .. kesääweathered+condition-NOM & poor+state of being-NOM all+GEN summer+PAR

oikeastaan keväästä lähtien 2 ni mulla siinäin fact spring+ELA leave+INF+GEN so I+ADE in it/that

on ollu joitakin asioita Û jotka onbe+3SG be+PTC some+PL+PAR matters/things+PL+PAR which-NOM/PL be+3SG

käyny mulle niinku rasitukseks. â 2 [Tel1:9]go+PTC I+ALL like/kind of burden/strain+TRA

‘sure it could be awful but since I’ve been in such poor shape and not feeling well all sum-mer, in fact, since spring, (so) I’ve got a few things there that have become a bit of a strainfor me.’

The notion of projection (which, for the moment can be seen as being more or

less equivalent to the notion of quoted and reported speech) as illustrated in ex-

ample 41 will be discussed in section 4.5.5 below, and relative clauses, as illus-

trated by the clause marked with square brackets in 42, will be discussed in sec-

tion 4.5.4. The following section will concentrate on the kind of interdependency

relations between the clauses marked by double slashes in example 42.

Page 147: systematic functional linguistic

139

Figure 4-3: Tactic Relationships in a Clause Complex

4.5.2. Interdependency: Parataxis and Hypotaxis

There are two notions needed in the analysis of complexing ) parataxis and hypo-

taxis. Parataxis refers to the linking of units of equal status at any rank, and hypo-

taxis refers to the linking of units of unequal status at any rank. As Martin (1988:

241) points out, the distinction between paratactic and hypotactic structures more

or less corresponds to Bloomfield’s (1933: 195) distinction between co-ordinative

and subordinative constructions, with the proviso that rankshifting or embedding

(to be discussed in 4.5.4) is distinct from subordination. As the Latin-based terms

co-ordination and subordination are more familiar than Halliday’s Greek-based

ones, they will be also be employed in this study. The Latin-based term “co-

ordinator” is used to refer to a conjunction in paratactic complex and the term

“subordinator” to one in a hypotactic complex. Numerals will be used to symbol-

ize co-ordination (1, 2, 3, ...), and Greek letters (", $, (, *, ...) for subordination.

The following diagram (reproduced from Halliday 1985a: 195) provides a

useful summary of the tactic relationships between clauses and the labels that are

used to refer to clauses in an tactic (interdependency) relationship with each

other:

Parataxis(co-ordination)

Hypotaxis(subordination)

Primary Clause

1 initiating

" dominant

Secondary Clause

2 continuing

$ dependent

This diagram is not concerned with relativization, embedding, or projection,

which, as argued in sections 4.5.4 and 4.5.5 below, do not involve (inter)depen-

dency relationships between clauses.

Page 148: systematic functional linguistic

140

Parataxis and hypotaxis can be illustrated by the following examples:

PARATAXIS (CO-ORDINATION)

(43) <A> sit ne oliki aika railakkaat juhlat jathen they-NOM be+3SG+TIS quite lively-NOM/PL party-NOM/PL &‘then it was quite a lively party after all and’

1 (initiating)

<B> ne päätty neljältä [CA3:13; two speakers]they-NOM end-(3SG/PAS) four+ABL

‘it ended at four o’clock’2 (continuing)

(44) Sari on ollu ilmarisella töissä Sari-NOM be+3SG be+PTC Ilmarinen+ADE work+PL+INE

‘Sari has been working for Ilmarinen’s’1 (initiating)

tai on tietysti vieläor be+3SG of course still‘or still is of course’

2 (continuing)

mut on nyt -- lomalla [CA3:25]but be+3SG now ... leave+ADE

‘but is now on leave’3 (continuing)

HYPOTAXIS (SUBORDINATION)

(45) mä pingottaisin varmaan hillittömästi [CA3:16]I-NOM feel nervous+CON+1SG for sure uncontrollably‘I’d be sure to feel uncontrollably nervous’

" (dominant)

jos mun pitäis esittää ison yleisön edessäif I+GEN have to+CON+3SG present+INF large+GEN audience+GEN front+INE

‘if I had to talk in front of a large audience’$ (dependent)

(46) ne nukkuu [CA3:11]they-PL sleep+3SG

‘they fall asleep’" (dominant)

kun ne painaa yläluomen alaluomen päällewhen they-PL press+3SG top+lid+GEN bottom+lid+GEN top+ALL

‘as soon as they press their eye-lids together’$ (dependent)

Page 149: systematic functional linguistic

141

In both parataxis and hypotaxis, one clause expands on another. For English,

at least, Halliday (1985a: 202 ff.) considers that there are three basic types of

expansion. Extension, where one clause is simply added to another, is symbol-

ized by a plus sign (1 +2, " +$). Elaboration, symbolized by an equals sign (1

=2, " =$), is a relationship between clauses in which one clause restates, exem-

plifies, specifies or describes another. With enhancement, one clause adds some

kind of circumstantial qualification to another; this is symbolized by a multiplica-

tion sign (1 x2, " x$). It is beyond the scope of this study to consider whether

these types of expansion are applicable in Finnish (see Kalliokoski 1989: 128 ff.

for a discussion with reference to Finnish ja ‘and’). What I would like to briefly

consider, however, are some of the grammatical reflexes of this basic distinction

between parataxis and hypotaxis.

Before proceeding, however, it should also be noted that a distinction is

being made between a conjunction and a conjunctive (or discourse) adjunct (see

Halliday 1985a: 50-51; Halliday & Hasan 1976: 226 ff. for a discussion of this

distinction in English). A conjunctive adjunct ) for example, siksi (the translative

form of se ‘it/that’) ‘for this reason’, kuitenkin ‘nevertheless’, esimerkiksi ‘for

example’ ) functions as a Circumstance (of time, cause etc.) in the clause. Con-

junctive adjuncts often come at the beginning of the clause, but they can also

come elsewhere in the clause. If there is a conjunction in the clause, the conjunc-

tion precedes the conjunctive adjunct: ja siksi ‘and for this reason’, mutta

esimerkiksi ‘but, for example’.

Conjunctions such as ja ‘and’, mutta ‘but’, and koska ‘because’, on the

other hand, are outside the structure of the clause; they do not have a role in the

experiential semantic organization of the clause. They almost always occur at the

beginning of a clause, although Finnish kun ‘when’ is sometimes exceptional in

this respect in that it can occur after the initial nominal, as with the second kun in

example 47, which is about a ring:

Page 150: systematic functional linguistic

142

(47) (i) 2 nyt ku sen panee mun sormee 2now when it+GEN put+3SG I+GEN finger+ILL

‘now when I put it [the ring] on my finger’

÷ (ii) ni se ku on hetken siellä 2so it-NOM when be+3SG moment there+ADE

‘and then when it’s there for a moment’

(iii) ni nousee tämmönen patti 2so rise+3SG this kind+NOM swelling+NOM

‘then a kind of swelling occurs’

(iv) et se ei lähe pois 2 [AA1: 3].that it+NOM NEG+3SG leave/come off‘so that it won’t come off’

The reason for this is not apparent, and the phenomena requires further research.

(For further examples and some discussion, see also Vilkuna 1989: 27).

The distinction between paratactic and hypotactic clauses ) like all gram-

matical distinctions ) is not a clear-cut one (see 2.4.12, p. 65), but can be seen in

terms of a continuum, with clauses in a hypotactic relationship sharing certain

features and clauses in a paratactic relationship sharing other features (cf. Kytö-

mäki 1985, who implicitly takes this approach in the analysis of sillä ‘for, as’ in

Finnish). However, in discussions of parataxis and hypotaxis (or co-ordination

and subordination) and of clause complexes in general, it is often assumed apriori

that certain conjunctions are paratactic and others are classed as hypotactic, and

no attempt is made to grammatically and semantically motivate the different ways

in which we can combine clauses. In what follows, an attempt is made to ground

the distinctions that are made on a consistent grammatical and semantic basis.

In parataxis, the order of the clauses cannot be changed; whereas this is (at

least hypothetically possible) in hypotaxis. Thus the clause that starts with a para-

tactic (co-ordinative) conjunction, i.e. a co-ordinator, such as ja ‘and’, mutta

‘but’, tai ‘or’, vai ‘or’ cannot be the initiating clause in a complex:

Page 151: systematic functional linguistic

143

1 See Kytömäki 1985: 52 for a discussion and some examples of cohesive ja in Finnish and,for example, Sorjonen & Heritage 1991 for a discussion of cohesive and in English in aninformal medical encounter, where it constructs an agenda-based nextness of a question. Cf.Halliday & Hasan (1976: 236) “it [and] often links a series of questions, meaning “the nextthing I want to know is ... “.

(48) ?? ja ne päätty neljältä [reversal of 43 above]‘and it ended at four o’clock’

(sit) ne oliki aika railakkaat juhlat‘then it was quite a lively party after all’

This is, of course, simply a grammatical reflex of the semantic relationship be-

tween the clauses: one is initiating and the other is continuing. It would be seman-

tically anomalous to make a continuing clause an initiating clause.

However, paratactic forms can also be cohesive (see Halliday & Hasan

(1976: Chapter 5) for a discussion of the cohesive relation of conjunction).

Thus, if one looks at spoken or written text, one is bound to find orthographic

sentences or turns in conversation that start with ja ‘and’, for example. A cohe-

sive ja does not link a clause to the following clause, but links what follows to the

previous text.1 The complexing discussed here is a different kind of phenomenon,

it is a structural relationship between clauses (though these structural relations

may also hold across turns in a conversation as illustrated by 43 above).

With hypotaxis, on the other hand, it is possible to change the order in a

hypotactic clause complex and produce a grammatically acceptable complex. (As

discussed below, this change in order is purely hypothetical.) The following com-

plexes are identical in form to 45 and 46 above except that the order of the

clauses is reversed:

(49) jos mun pitäis esittää ison yleisön edessä‘if I had to talk in front of a large audience’

$ (dependent)

mä pingottaisin varmaan hillittömästi‘I’d be sure to feel uncontrollably nervous’

" (dominant)

Page 152: systematic functional linguistic

144

(50) kun ne painaa yläluomen alaluomen päälle‘as soon as they press their eye-lids together’

$ (dependent)

ne nukkuu‘they fall asleep’

" (dominant)

This again is simply a grammatical reflex of the semantic relationship between

the clauses: it does not matter whether a dependent clause precedes or follows the

dominant clause. A dependency relationship holds irrespective of sequence.

The distinction that is made here is an abstraction and a generalization based

on my knowledge of Finnish, and, the discussion here ) as in many other places

in this study ) reaffirms Itkonen’s view on the precedence of intuition in gram-

matical description (see 2.3.4, p. 26 ff.). The order of clauses in a hypotactic com-

plex cannot be changed in a real text, as any change in order brings about a

change in textual meaning. Moreover, the experiential meanings realized in the

clauses may make it more natural for one to precede and the other to follow.(In

other words, there may be what are sometimes referred to as “semantic con-

straints” on the ordering of clauses.)

Thus, in order to apply the distinction between hypotaxis and parataxis to

examples from a text, cohesive elements have to be ignored or changes have to be

made in order to produce an acceptable complex. For example, changes would

have to be made if there is an anaphoric element or an element presupposed by

ellipsis in the second clause in a hypotactic complex. Cohesive elements like niin

in pairs such as kun ) niin ‘when ) then’ and jos ) niin ‘if ) so’ would have to

be omitted if the order is reversed.

(51) jos sul . on viis markkaa [CA2:22] if you(sg)+A(DE) be+3SG five-NOM mark+PAR

‘if you’ve got 5 marks’$ (dependent)

ni sä voit avata pankkitilinthen/so you-NOM can+2SG open+INF bank+account+GEN

‘then you can open a bank account’" (dominant)

Page 153: systematic functional linguistic

145

In order to reverse the order of these clauses, we would have to omit the niin at

the beginning of the second clause:

(52) sä voit avata pankkitilin‘you can open a bank account’

" (dominant)

jos sul . on viis markkaa [CA2:22] ‘if you’ve got 5 marks’

$ (dependent)

Another feature that distinguishes paratactic and hypotactic complexes is the

fact that there can be more than two clauses in paratactic relation to each other

(e.g. more than two clauses can be joined by ja ‘and’) but only two clauses (or

complexes) can be hypotactically related. A feature that distinguishes co-ordina-

tors from subordinators is the different ways in which they can combine with each

other. For example, in a combination of co-ordinator and subordinator, the co-

ordinator comes first: ja kun ... ‘and when ...’ (?? kun ja ‘when and’ ...); mutta

koska ... ‘but because ...’ (?? koska mutta ‘because but ...’). These are not thus

random features; the way in which these conjunctions can combine with each

other is related to their meaning. Like other meanings in language, we would not

normally expect to find semantically anomalous sequences.

Clauses in paratactic relation differ from hypotactic complexes in the way in

which mood options are available in the second clause. As illustrated by 53, the

second clause in a paratactic complex can also be realized as an interrogative.

(53) kuinka päin nää on kiinnihow-Q direction+INS these-NOM/PL be+3SG closed‘which way are these closed/which way do you close these’

1 (initiating)

ja kuinka päin nää on auki? [CA3:4]& how-Q direction+INS these-NOM/PL be+3SG open‘and which way are they open/and which way do you open them?’

2 (continuing)

Page 154: systematic functional linguistic

146

In a hypotactic complex, if the dominant clause is realized as an interrogative, the

dependent clause falls within the scope of this interrogative:

(54) no tuliko hyväwell come+PAS+3SG+Q good-NOM

‘Well did it turn well’" (dominant)

ku hän oli mukana [CA3:22]when/as (s)he-NOM be+PAS+3SG along/with/involved‘now that/since he was involved?’

$ (dependent)

The distinction between parataxis and hypotaxis that is being made here

does not distinguish between hypotaxis, where one clause is dependent on an-

other, i.e. the relationship is between clauses in a complex, and relativization

and embedding, which will be discussed in section 4.5.4. Before going on to dis-

cuss relativization, embedding and projection, however, I shall discuss the func-

tion of clause complexing with particular reference to spoken language, as it

sometimes seems to be assumed that clause complexing is more properly a feature

of written language.

4.5.3. The Function of a Clause Complex

As pointed out in 4.5.1, the way in which we combine clauses is meaningful.

Halliday (1985b: 82) regards a complex as representing a complex phenomenon

or as a set of interrelated phenomena. According to Halliday (1985b: 87 ff.),

while the way in which speakers exploit the resources of the clause complex in

unselfconscious conversation is different to the way in which they are exploited

in written language, both are just as complex, but they are complex in different

ways. As Halliday points out:

Of course, much conversation is fragmentary, with speakers taking very short turns; and herethe potential for creating these dynamic patterns does not get fully exploited. But the differ-ence is not so great as it might seem, because what happens in dialogue is that the speakersshare in the production of discourse; so that although the grammar does not show the paratac-

Page 155: systematic functional linguistic

147

tic and hypotactic patterns of the clause complex in the way that these appear when the samespeaker holds the floor, some of the same semantic relations may be present across turns.(Halliday 1985b: 87.)

While some clauses may be rather loosely combined into complexes in spoken

language, there is ample evidence of complexing in spoken Finnish.

Example 47 above (repeated below as 55) is a good illustration of the intri-

cate complexity of spoken language. It is about a ring, which is introduced in the

first unfinished clause. The rest of the turn is made up of a clause complex, in

which all of the constituent clauses are part of a whole. As intonation is important

in arguing for the way in which I analyse it, the syllable in bold type is what I

perceived as being the tonic syllable.

(55) se sormus ei milläänit/that-NOM ring-NOM NEG+3SG any+ADE

‘that ring won’t in any

(i) 2 nyt ku sen panee mun sormee 2now when it+GEN put+3SG I+GEN finger+ILL

‘now when it’s put on my finger’

(ii) ni se ku on hetken siellä 2so it-NOM when be+3SG moment there+ADE

‘and then when it’s there for a moment’

(iii) ni nousee tämmönen patti 2so rise+3SG this kind+NOM swelling+NOM

‘then a kind of swelling occurs

(iv) et se ei lähe pois 2 [AA1: 3].that it+NOM NEG+3SG leave/come off‘so it won’t come off’

The text in which this complex occurred was spoken quickly and without any

hesitation. While the example is complex, it has a characteristically spoken fla-

vour about it.

There is no doubt that complexes like this are an important resource in both

spoken and written language. This is not to deny, of course, that there are differ-

ences between spoken and written language (see Halliday 1985b, particularly 76-

Page 156: systematic functional linguistic

148

101). Moreover, an in-depth analysis of complexing in spoken language would

also need to incorporate an analysis of intonational patterns in complexes.

The overall relationship in the complex could be given as follows:

(56) (i) nyt ku sen panee mun sormee $‘now when it’s put on my finger’

(ii) ni se ku on hetken siellä " 1 $‘and then when it’s there for a moment’

(iii) ni nousee tämmönen patti " 1 "‘then a kind of swelling occurs

(iv) et se ei lähe pois " 2‘so it won’t come off’

The complex begins with a dependent clause (i), which is followed by the domi-

nant clause. However, this dominant clause contains the conjunction kun ‘when’,

which indicates that it, in turn, is part of a complex. Thus, the dominant clause is,

in fact, a complex. Clause (ii) is dependent on the following clause (iii). The into-

nation does not fall at the end of the third clause: the speaker goes straight on

without pausing to the last clause (iv). The last clause realizes the effect in a

cause $ effect relationship, and the cause is realized by the previous two clauses.

The fact that the effect is realized in two clause (ii ) iii) is evidenced by the into-

nation: clause ii realizes given information and the tonic syllable comes at the end

of iii. The conjunction et(tä) ‘so’ in this kind of context is a co-ordinator, since

the order of clauses (ii ) iii) and (iv) cannot be changed.

Halliday (1985b: 87) uses a behavioural analogy to describe the way in

which clauses in spoken language form complexes: the organization of spoken

language is choreographic, like steps in a dance. Of course, the original text is

represented here in a transcribed form, and this gives us a synoptic perspective

(see 2.4.14) on the whole complex: what originally was a dynamic flow becomes

something that is static and frozen. This synoptic perspective is also important: it

allows us to attend to the text (or part of it) as a whole.

Page 157: systematic functional linguistic

149

1 A possible exception to this is found in Finnish sayings and proverbs, where it is not unusualfor a relative clause to refer forward. The following example is from the Bible (Proverbs26:27):

Joka kuopan kaivaa, se kaatuu siihen.who/which pit+GEN dig+3SG it/(s)he storm+GEN it+ILL

‘Whoso diggeth a pit, shall fall therein.’

In other genres, it would be more natural to have the following word order: Se, joka kaivaakuopan, kaatuu siihen ‘He who digs a pit will fall into it’.

4.5.4. Relativization and Embedding

According to Halliday (1985a: 219 ff.), embedding or rankshifting occurs when

one clause functions as part of another clause. The terms “embedding” and

“rankshifting” are probably not the best terms to describe the kind of complexing

that is at issue in Finnish. Firstly, they have traditionally implied a synoptic, static

view of a complex, and when one looks at complexing in spoken Finnish, then it

is clear that an embedded clause unfolds dynamically just like any other complex.

Secondly, Halliday regards a defining relative clause in English as being embed-

ded in the host clause, whereas a non-defining relative clause is considered to be

a dependent clause in a hypotactic complex. As I shall argue below, with Finnish

relative clauses, on the other hand, there is no grammatical or prosodic distinction

that can be equated with the defining/non-defining distinction that is made in

English. However, relative clauses are different from hypotactic clauses ) the

kind of criteria invoked in section 4.5.2 above (p. 142 ff.) to distinguish clauses in

a hypotactic (subordinative) relationship from clauses in a paratactic (co-ordina-

tive) relationship does not apply to relative clauses. For example, the order of the

clauses cannot be changed, since ) to use the traditional terms ) the relative

clause cannot precede its antecedent.1

For want of a better term, I shall regard as instances of embedding both 1)

relative clauses and 2) clauses which function directly in the structure of a host

clause. The important thing is to distinguish these complexes, where the relation-

ship is not between clauses at the rank of clause, from paratactic and hypotactic

Page 158: systematic functional linguistic

150

complexes. Firstly, I shall briefly consider clauses that are directly embedded and

then discuss the problem of relative clauses in Finnish. By directly embedding, I

refer to a clause that is rankshifted to function in the structure of another clause

without a nominal as an intermediary. The most common type of clause that is

directly embedded in Finnish is a non-finite clause. These non-finite embedded

clauses were discussed briefly in 4.4.1 above (p. 119 ff.). It seems to be less com-

mon for finite clauses to be directly embedded in a host clause in Finnish. A like-

ly environment for this kind of embedding is a relational clause, as illustrated by

the following example, where the embedded clause is in fact a rather long clause

complex, which is italicized in the example:

(57) “Hyvä esimerkki suomalaisesta heinäkenkäasenteesta oli, kun oltiin joskus Valkeassafestivaaleilla, ja Ian Gillan tuli meidän bussiin, nappas Herba-pullon, joi ja sanoi, ettäfantastic. Me luultiin, että se tarkoitti bändiä, mutta se tarkoittikin sitä viinaa”, Sakke kertoo.

[HS 5.11.91: D10]

“A good example of a Finnish bumpkin attitude was when we were once at a festival inValkea and Ian Gillan came into our bus, grabbed a bottle of Herba, drank from it and said‘fantastic’. We thought he was referring to the band, but it was the liquor that he was referringto”, said Sakke.

From the point of view of traditional grammar, the embedded complex is the

complement of the host clause. From a functional perspective (see Chapter 6), the

embedded complex can be regarded as the Identified in a relational process.

Relative clauses, on the other hand, involve a different kind of embedding:

the clause expands on an NP in the host clause by describing or defining it. In

contrast to a clause that is directly embedded in the host clause, a relative clause

is, thus, tied to a phrase functioning in the structure of the host clause. The fol-

lowing example illustrates the kind of phenomenon that is usually understood as

embedding:

Page 159: systematic functional linguistic

151

1 The reference is to the first novel written in Finnish: Seitsemän veljestä (The SevenBrothers) by Aleksis Kivi.

(58) Monet, Û jotka vilpittömästi pitävätmany+NOM/PL who+NOM/PL sincerely hold/regard+3PL

itseään eläinystävinä â vieroksuvat kissaaself+PAR+POS/3 animal+friend+PL+ESS shy away from/shun+3PL cat+PAR

sen oletetun petomaisuuden takia. [HKV]it+GEN suppose+INDEF/PTC+GEN predatory nature+GEN because

‘Many who sincerely regard themselves as animal lovers shy away from cats because oftheir supposed predatory nature.’

In this example the relative clause beginning with an inflected form of joka ‘who,

which’ functions as part of an NP, which is the subject of the clause. This differs

from the hypotactic (subordinative) complex illustrated in the previous section: in

a hypotactic complex the relationship is at the rank of clause and this is reflected

in the fact that the order of the clauses can be changed.

The example above (58) looks like an English defining or restrictive relative

clause. While there are clauses, like the one illustrated above, that can be trans-

lated by a defining clause in English, there are also clauses, such as the following,

that can be translated by a non-defining clause in English.

(59) Olihan seitsemällä veljekselläkin oma temppujabe+PAS+3SG+TIS seven+ADE brothers+ADE+TIS own+NOM tricks+PL+PAR

tekevä Matti-kissa, Û jolla olimake/do+PTC/NOM Matti-cat+NOM who+ADE be+PAS+3SG

turvattu asema perhepiirissä. â [HKV]secure+INDE/PTC position+NOM family+circle+INE

‘After all even the Seven Brothers1 had their own trick-performing cat called Matti, whohad a secure position in the family circle.’

What is significant in Finnish is not that a clause is defining or non-defining, but

that the speaker expands an NP by adding something that is seen as being integral

Page 160: systematic functional linguistic

152

1 It may be necessary to point out to a Finnish reader that the use of commas to mark off arelative clause in English is a meaningful option that reflects intonational differences betweendefining and non-defining relative clauses: (a) Pilots whose minds are dull do not live long /=(b) Pilots, whose minds are dull, do not live long. According to clause a, only certain pilotsdo not live long, namely those who are not always on the alert. According to clause b, pilotscan be characterized by the fact that they have dull minds.

or significant. While we can translate Finnish clauses as either defining or non-

defining in English, grammatical distinctions in Finnish cannot be based on trans-

lation equivalence.

Following Helasvuo (in press), it seems to me that there is no defining vs.

non-defining distinction in Finnish, in spite of the fact that this distinction is often

made by Finnish grammarians (e.g. Hakulinen & Karlsson 1979: 125). What is

significant in Finnish is that a speaker expands on an NP by adding something

that she or he sees as being important or relevant, and whether this actually de-

fines or restricts the clause is not a grammatical or semantic feature that can be

associated with the clause itself. As Helasvuo points out, there appear to be no

recognizable intonational differences amongst joka clauses. Moreover, as Helas-

vuo points out, the orthographic conventions of standardized written Finnish ap-

pear to reflect this fact: all relative clauses are marked off with commas from the

host clause.1 The following example is instructive:

(60) Inhoan feminismiä, joka nojautuu Derridaanloathe+1SG feminism+PAR which+NOM lean/depend+3SG Derrida+ILL

tai Lacaniin. [HSkl 22.1.92 p. 43]or Lacan+ILL

‘I loathe feminism (that is) based on Derrida or Lacan (~ I loathe feminism, which is basedon Derrida or Lacan.)’

From the point of the defining/non-defining distinction, this example is ambigu-

ous: as indicated, it could be translated as either defining or non-defining in Eng-

lish. In this instance it can be inferred that the meaning given by the first transla-

tion is what is intended, but this is not a grammatical distinction in Finnish.

Page 161: systematic functional linguistic

153

The claim that there is no grammatical distinction in relative clauses be-

tween defining and non-defining clauses does not deny the fact that other gram-

matical features may combine to produce a similar sort of distinction. The initial

NP in tuo mies on Aki Kaurismäki ‘that man is Aki Kaurismäki’ contains a (se-

lective) demonstrative, which indicates that the item in question (tuo mies ‘that

man’) can be identified in the environment ) in the text or the situation (see Halli-

day & Hasan 1976). If we further specify the NP by adding a relative clause ) tuo

mies, joka kävelee meitä vastaan, on Aki Kaurismäki ‘that man who is walking

towards us is Aki Kaurismäki’ ) then the relative clause presents the criteria by

which tuo mies ‘that man’ can be identified. However, as indicated above, all that

a relative clause does is add something that is constructed as being significant to

a host NP. Whether or not this serves to identify the item is not a grammatical

feature of the relative clause.

As for the grammatical status of relative clause in clause complexes, Helas-

vuo (in press) does not regard them as being embedded. However, whether or not

they are considered to be embedded in Finnish depends on what we understand

by embedding. Halliday (1985a: 219) characterizes embedding in (defining) rela-

tive clauses in English as follows:

Embedding is a mechanism whereby a clause or phrase comes to function as a constituentWITHIN the structure of a group, which itself is a constituent of a clause. Hence there is nodirect relationship between an embedded clause and the clause within which it was embed-ded; the relationship of an embedded clause to the ‘outer’ clause is an indirect one, with agroup as intermediary.

In this sense, relative clauses in Finnish are embedded: the link between the

clauses is not at the rank of clause, but at the rank of phrase.

Moreover, to say that a clause in Finnish is embedded does not diminish

its textual importance. When one looks at text, there are many instances where

the host clause introduces an NP, which then becomes the topical Theme of the

embedded clause. For example:

Page 162: systematic functional linguistic

154

1 Cf. milloin ‘when’, which is an interrogative, and silloin ‘when’ (from se ‘it, that’), whichfunctions as a conjunctive (discourse) adjunct.

(61) se on tyttö Û joka on(s)he/it/that-NOM be+3SG girl-NOM who-NOM be+3SG

matkustanu â [CA3:3]travel+PTC

‘She’s a girl who’s travelled.’

(62) Tuskin on olemassa ihmistä, scarcely is+3SG being/existence+INE person+PAR

Û joka suhtautuu välinpitämättömästi kissaan. â [HKV]who+NOM relate+PASS-REF+3SG indifferently cat+ILL

‘It’s rare for someone to be indifferent to cats.’

(This topic is taken up again in Chapter 7.)

As Helasvuo points out, a joka clause in spoken Finnish typically comes at

the end of the host clause. This can result in a chain of embeddings. The follow-

ing example from Helasvuo is from the Finnish pear story data (see Chafe 1980):

(63) ja sitten oli kolme poikaa& then be+PAS+3SG three-NOM boy+PAR

‘and then there were three boys’

Û jolla yhellä oli semmonen .. pingispallo âwho+ADE one+ADE be+PAS+3SG kind of-NOM ping-pong+ball-NOM

‘one of whom had a kind of ping pong ball’

Û joka oli kiinnitetty semmoseen mailaan âwhich-NOM be+PAS/3SG attach+INDE+PTC kind of+ILL racket/bat+ILL

‘which was attached to a kind of racket’

Û jota se pompotti. âwhich+PAR he-NOM palpitate/bounce+PAS/3SG

‘which he bounced’

As illustrated in this example, and in all of the examples quoted by Helasvuo, the

relative pronoun joka consistently inflects for number and case. Moreover, a time

affix can also be attached to it: jolloin ‘when (at the time when)’.1 This is, of

Page 163: systematic functional linguistic

155

1 In fact, onnellinen ‘happy’ in Finnish is an adjective, but as pointed out in the section onNPs, Adjectival Phrases are considered to be a subtype of NP.

course, clear evidence that the link between the clauses is not at the rank of

clause, but at the rank of phrase.

There is another relative pronoun mikä ‘which’, which is typically used

when the antecedent is non-human. However, unlike joka, mikä (or the partitive

form mitä) can also function as a conjunction linking clauses in paratactic rela-

tionship with each other (e.g. Hän tuli ajoissa, mitä ei ole ennen sattunut ‘(S)he

came on time, which hasn’t happened before.’ In other words, it can be used

rather like a non-defining English relative clause whose domain is the entire pre-

vious clause.

Another important relative element that figures in embedded expansions is

että ‘that’. This form also occurs in taxis and projection, but when että introduces

an embedded clause, like joka above, it expands on an NP. The kind of example

that immediately comes to mind is the construction niin ‘so’ + ADJECTIVE + että

‘that’.

(64) nähtävästi miehet on niin onnellisia Û et neapparently man-NOM/PL be+3SG so fortunate/happy that they

nukkuu kun ne painaa yläluomen alaluomen päälle â [CA3:11]sleep when they press their lids together

‘Apparently men are so fortunate that they fall asleep as soon as they press their eye-lidstogether.’

The clause beginning with et ‘that’ expands on the NP niin onnellisia ‘so

happy’.1

Often an embedded että clause is related to a cataphoric demonstrative, for

example, se ‘it/that’ (which can be case-inflected) or what might be termed a

characterizing demonstrative, sellainen ~ semmoinen (from se ‘it/that’) ‘the/that

kind of, such’, tällainen ~ tämmöinen (from tämä ‘this’) ‘this kind of’. Cataphor-

Page 164: systematic functional linguistic

156

ic demonstratives that are followed by an embedded että clause are typical of

Finnish:

(65) [The discussion is about a piggy bank (säästöpossu).]

mul. on sellaneI+(ADE) be+3SG the kind-NOM

Û et siihe pääsee avaimella â [CA2:22]that it+ILL get to+3SG key+ADE

‘I’ve got the kind (that) you can open with a key.’

(66) sitä vartenhan ystävät ovat Û että heitäit/that+PAR reason(PP)+TIS friends-NOM/PL be+3PL that they+PL+PAR

pyydetään auttamaan â [CA3:14]ask+INDE help+INF+ILL

‘The reason for having friends is that you can ask them for help.’

(67) Sigfridsin kannalle Hannus päätyi kuitenkinSigfrids+GEN support/behind+ALL Hannus end up+PAS+3SG nevertheless

siksi, Û että tämä on kotoisin Vaasanit+TRA that this-NOM be+3SG by home Vaasa+GEN

läänistä ... âprovince+ELA [HKV]

‘Hannus ended up supporting Sigfrid because he was from the province of Vaasa ...’

Another typical environment for embedding is an attributive relational pro-

cess. The host clause realizes the Process and the Attribute (see Chapter 6), e.g.

on totta ‘ (it) is true’, and the Carrier of this Attribute is a fact realized as a finite

clause, or clause complex, as in the following example:

Page 165: systematic functional linguistic

157

(68) On käsittämätöntä,be+3SG incomprehensible+PAR

Û että joidenkin mielestä ainoa todellinen kissa onthat some+PL+GEN only real+NOM cat+NOM be+3SG

tappeluissa parkkiintunut, repalekorvainen kolli 2fight+PL+INE toughen+PASS+PTC/NOM shred+eared+NOM tom+NOM

ja siksi he itsepintaisesti kieltäytyvät& it/that-TRA they+NOM stubbornly refuse+PASS-REF+3SG

leikkauttamasta sitä. â [HKV]have neutered+INF+ELA it+PAR

‘It’s incredible that in the opinion of some people the only real cat is a tomcat who hasbeen toughened in fights and whose ears are in shreds and for this reason they stub-bornly refuse to have it neutered.’

This analysis may not appear to be very satisfying from a textual perspective, but

this is not a problem since, in an SF approach, different kinds of structure contrib-

ute to the total meaning. Clauses like this will be discussed from a textual per-

spective in Chapter 7.

While there is no “dummy subject” in instances exemplified by 68 in stan-

dardized written Finnish, this is not the case in spoken Finnish where the pronoun

se ‘it, that’ can function in much the same way as English it in the translation

equivalents.

(69) se oli vähä harmi et siitit/that-NOM be+PAS+3SG little pity-NOM that it+ELA

piti tulla sähkösauna [CA3:19]have to+PAS+3SG come+INF electricity+sauna-NOM

‘It’s a bit of a pity that it had to be an electric sauna.’

However, there are also many instances without a pronoun in spoken Finnish.

As pointed out earlier, however, the notion of embedding as used here is

conceived of as a dynamic resource in Finnish. This can be illustrated by the fol-

lowing rather complex examples:

Page 166: systematic functional linguistic

158

(70) [Esahan on sen kerran tehny onnellisena -- nukkunu lentokoneen lähdön ohi --‘Esa has done it once blissfully .. slept through the departure of his plane’]

÷ oli semmonen tilanne Û että Esa ensinbe+PAS/3SG such kind-NOM situation-NOM that Esa-NOM first

sano(i) et ku meiät pyydettiin illalla kyläänsay(+PAS/3SG) that as we-ACC ask+INDE+PAS evening+ADE visit+ILL

ja on nii aikanen että, et ehkei hän lähde& be+3SG so early that that perhaps+NEG/3SG he-NOM go

mut sit se sano(i) et no! kyllähänbut then he-NOM say(+PAS/3SG) that well/so for sure+TIS

mä ny herään â [CA3: 13]I-NOM now wake+1SG

‘The situation was such that Esa first said that since we’d been asked to visit someone inthe evening and it’s so early that perhaps he wouldn’t go but then he said “what the heck!I’m sure to wake up”.’

(71) niill . on joku tämmönen systeemithey(+ADE) be+3SG some this kind-NOM system-NOM

Û et jos on koulumusiikkilinjalla nii pitäs pysyythat if be+3SG school+music+stream+ADE then/so must+CON+3SG stay

siellä et ne ei mielellään päästä ihmisiä ... âthere+ADE that they-NOM NEG+3SG by preference let go of+INF person+PL+PAR

[CA3:26]

‘They’ve got a kind of system like this whereby if you’re in the music teacher stream thenyou’ve got to stay put, they don’t like to let go of people ..’

In the first example the embedded complex expands on semmonen tilanne ‘such

a situation’ and in the second example there is an embedded complex expanding

on joku tämmönen systeemi ‘a kind of system like this’.

In this section, I have concentrated on joka and että. There are, of course,

other conjunctions and relative pronouns that figure in embedding, as illustrated

by kun ‘when’ and mitä ‘which, what’ in the following examples:

Page 167: systematic functional linguistic

159

(72) mullaki on seI+ADE+TIS be+3SG it/that-NOM

Û et mä yleensä virkistyn siinä vaiheessathat I-NOM generally become alert+1SG it/that+ESS stage+INE

Û ku pitäs mennä nukkumaanâ â [CA3:10]when must+COND+3SG go+INF sleep+INF+ILL

‘I’ve got that habit/characteristic too that I’m generally start being alert at the stage whenone/you should be going to sleep.’

(73) Kysymys filosofian tieteellisyydestä on luonnollisestiquestion-NOM philosophy+GEN scientificness+ELA be+3SG naturally

suhteellinen siihen, [[ mitä tieteellä tarkoitetaan. ]] [IN]relative-NOM it+ILL which+PAR science+ADE mean+INDE

‘The question of the status of philosophy as a science is naturally relative to the way inwhich we define science.’

The analysis of these other relative pronouns and conjunctions that figure in em-

bedding needs to be incorporated into an analysis of embedded complexes in

Finnish. The analysis here also needs to be supplemented with an analysis of co-

hesive relations in Finnish so that examples such as those illustrated in 65 ) 67

are analysed from the perspective of the cohesive relations in a text.

4.5.5. Projection

The notion of projection in SF theory can be seen as being roughly equivalent to

some of the kinds of phenomena that are analysed under the heading of direct,

indirect and free indirect speech (and thought), if these traditional notions are not

understood in simplistic terms of reporting something that was said (or thought)

in another situation. Moreover, as discussed in this section, it also includes other

related phenomena. What is essential in projection is that a clause (either finite or

non-finite) ) instead of being a direct representation (or construction) of non-lin-

guistic reality ) is at a further remove from this reality (Halliday 1985a: 227 ff.).

Page 168: systematic functional linguistic

160

Clause complexes involving projection can be distinguished from complexes

in a tactic relationship by the fact that the second clause (which is generally the

projected clause) can either immediately follow the finite verb in the first clause

or else the boundary between the clauses is marked by the conjunction et(tä)

‘that, so, thus’. In spoken Finnish, the projected clause is typically introduced by

et(tä) ‘that, so, thus’ even if it is a quoted locution (“direct speech”), as in exam-

ple 74:

(74) mä sanoin et mul pitäs ollaI-NOM say+PAS+1SG that I+ADE must+CON+(3SG) be+INF

täällä tili [CA2:23-24]here+ADE account+NOM

‘I said “I should have an account here”.’

However, if the projected clause is a polar interrogative, there is no conjunction

between the clauses:

(75) mä en oikeen tiedäI-NOM NEG+1SG really know

kuuluuks se Viherlaaksoon vai Lähderantaan. [CA3: 23]belong+Q it+NOM Viherlaakso+ILL or Lähderanta+ILL

‘I don’t really know whether it’s part of Viherlaakso or Lähderanta’.

As is obvious from this example, the conjunction et(tä) is not necessarily the

mark of “indirect speech” in Finnish. In a projecting complex, että means some-

thing like “thus, in this way, like so” (see Kuiri 1984: 136 ff.). While examples

like 74 above are typical of spoken Finnish, one is unlikely to come across an

example like this in traditional descriptions of Finnish grammar: the analysis of

projection in Finnish is complicated by the fact that there is a distinction between

“direct and indirect speech” in standardized written Finnish that does not accord

with the way in which clauses are projected in unselfconscious, spoken Finnish.

As Kuiri (1984: 3 ff.) points out, the distinction between direct and indirect

speech has come from Latin-based models of grammar. Thus, in this received

view, the first example below illustrates direct speech and the second example

indirect speech:

Page 169: systematic functional linguistic

161

(76) a. Hän sanoi: “Nyt minä menen levolle.”(s)he said+PAS/3SG now I-NOM go+1SG rest+ALL

‘She/he said, “Now I’m going to rest.”’

b. Hän sanoi, että hän menee levolle.(s)he said+PAS/3SG that (s)he-NOM go+3SG rest+ALL

‘She/he said that she/he’s going to rest.’

According to the conventions of standardized written Finnish, a comma is put

before the et(tä). However, in spoken Finnish, if there is a pause, it often comes

after the että: se sanoi että (pause) ... ‘(s)he said that/thus (pause)’ From the

perspective of English, this may seem strange, as in English if there is a pause, it

would most probably come before that: he said (pause) that. The pause in Finn-

ish reflects the essentially different character of et(tä) (‘thus, like so’).

While, as I shall argue below, it is not feasible to analyse projecting com-

plexes in terms of interdependency, as Halliday (1985a) does for English, there

are a number of other distinctions made by Halliday (1985a) that are useful in the

analysis of projection in Finnish. Halliday distinguishes three types of projected

clauses: 1) locutions, 2) ideas and 3) facts. In a locution, the projecting clause is

a verbal process (e.g. sanoa ‘say’), a process of externalized human conscious-

ness. Example 74 above is an example of a locution, as are the following exam-

ples, which further illustrate the use of et(tä) ‘thus, like so’ in spoken Finnish:

(77) kysyin et millä sä kuljet töihin [CA3:8]ask+1SG that/so which+ADE you travel+2SG work+PL+ILL

‘I asked “how do you get to work?”.’

(78) mä sanoin et älä välitä [CA3:17]I-NOM say+PAS+1SG that NEG+2SG/IMP worry‘I said “Don’t worry”’.

In spite of the conjunction, in both of these examples the projected clause could

stand alone: millä sä kuljet töihin ‘how do you get to work?’, älä välitä ‘don’t

worry’.

Page 170: systematic functional linguistic

162

In what Halliday calls an idea, the projecting clause is a mental process (e.g.

tietää ‘know’), i.e. a process of internal human consciousness:

(79) mä tiedän et mun täytyy aamuyöstä noustaI-NOM know-1SG that I+GEN must+3SG morning+night+ELA get up+INF

‘I know (that) I have to get up in the early hours of the morning.’ [CA3:10]

(80) se -- suunnitteli viimeks et se ottaashe .. plan+PAS+3SG last+TRA that she+NOM take+3SG

kamat mukaan ja tulee suoraan [CA12:5]thing+NOM/PL along and come+3SG directly

‘She planned to take her things with her and come here directly.’

A fact, on the other hand, is something that comes “already packaged” as

something that is a further remove from reality, so there is no projecting clause.

Facts are realized as clauses embedded within a host clause:

(81) onks nii et Irma jäi siihe, Munkkivuoreen?be+3SG+Q so that Irma+NOM stay+PAS+3SG there+ILL Munkkivuori+ILL

‘Is it true/so that Irma stayed there in Munkkivuori?’ [CA3: 23]

Halliday also distinguishes 1) quotes (“direct speech”), reports (“indirect

speech”), and the various intermediate forms between reporting and quoting

(“free indirect speech”). A quote can be distinguished from a report in that there

is nothing in the wording that shows that it is projected: it could occur alone as a

direct representation of non-linguistic reality rather than being at a further remove

(Halliday 1985a: 228).

(82) mä sanoin et mul pitäs ollaI-NOM say+PAS+1SG that I+ADE must+CON+(3SG) be+INF

täällä tili [CA2:23-24]here+ADE account+NOM

‘I said “I should have an account here”.’

Page 171: systematic functional linguistic

163

÷ mul pitäs olla täällä tili‘I should have an account here’

Quoted locutions are typical of natural, unselfconscious spoken Finnish, as indi-

cated above. However, whether a locution is realized as a quote or a report seems

dependent on a number of factors. The following example illustrates a typical

environment where a verbal process is followed by a report:

(83) hän tunnusti että hän oli juonut(s)he confess+PAS/3SG that (s)he be+PAS/3SG drink+PTC

puol(i) pulloo viskiä ennen sitä [CA3: 17]half-NOM bottle+PAR whisky+PAR before it/that+PAR

‘He confessed that he had drunk half a bottle of whisky before that [giving a paper].’

As illustrated by this example, the deictic elements in a report shift from reference

to the speech situation: personal pronouns, demonstratives and tenses change.

The choice of the verb in this example is clearly significant: tunnustaa ‘confess’

rather than sanoa ‘say’ or kysyä ‘ask’. The verb chosen by the speaker indicates

an interpretation by the speaker; she has interpreted what was said as a confes-

sion. Thus, it would be somewhat anomalous for the projected clause to be real-

ized as a quote. Of course, even with verbal processes like tunnustaa ‘confess’ or

myöntää ‘admit’, which explicitly indicate an interpretation of a saying, a quote

is possible. However, it seems to me that a quote would be more typical of a liter-

ary style.

As pointed out above, however, the distinction between quoted and reported

locutions (sayings) in Finnish is complicated by the fact that the distinction in

standardized written Finnish is based on Latinate models. This does not, however,

affect the fact that it is possible to distinguish between a quoted and a reported

locution in Finnish. However, while it can be argued that the distinction between

a quote and a report is relevant for locutions, as indicated above, this does not

imply that there is a simplistic (“transformational”) relationship between them,

i.e. that quotes and reports are interchangeable. On the other hand, when the pro-

jected clause is an idea (i.e. the projecting clause realizes a mental process), it

seems more natural for it to be a report. However, quotes are also possible: they

Page 172: systematic functional linguistic

164

appear to be particularly relevant in literature and in narrative episodes in inter-

personal conversation.

(84) sit mä aattelin et kuule me ollaan karaatemiehiä [Tel2:2]then I thought that listen we are karate+men‘Then I thought ‘hey listen we’re karate men’‘

(85) – Tässäkö se kaikki on? mä ajattelin. [AR:50]‘Is that all there is to it?’ I thought.

(86) – Saapa nähdä tapahtuuko nyt jotain erikoista, Muumipeikko ajatteli. [TH: 119]“I wonder if something unusual is about to happen”, thought Moomin Troll.

Characteristic of the unselfconscious spoken Finnish of young Finns is the

quoting of an idea with the verb olla ‘(to) be’ in the projecting clause. Projection

with olla often appears to put a reaction that someone has had into words.

(87) [siellä oli yks semmone japanilainen poika -- sil oli hirmu siistit vaatteet semmone ihananahkatakki oikeen rahakkaan näkönen kello se oli ihan ku kultane

‘there was a sort of Japanese boy there ... he had really neat clothes, a sort of marvellousleather jacket ) a really expensive looking watch, it was really like made of gold’]

mä olin et kyllä japanilaisetI-NOM be+1SG that indeed/certainly Japanese-NOM/PL

on varakkaita [CA2:21]be+3SG rich+PL+PAR

‘I’m like “Gosh Japanese people are rich”.’

The use of quotes in examples like this gives a dramatic flavour to a narrative

episode. This reinforces Halliday’s (1985a: 233) point that quoting and reporting

are not simply formal variants; they differ in meaning. According to Halliday,

quoting “is more immediate and lifelike, and this effect is enhanced by the orien-

tation of the deixis, which is that of drama not that of narrative”.

Page 173: systematic functional linguistic

165

Projected clauses (both locutions and ideas) can also be embedded. In in-

stances of embedding, the embedded projection is linked to the host clause by a

nominal that is a label for a metaphenomenon (e.g. väite ‘claim’, tosiasia ‘fact’,

ajatus ‘thought’, etc.), as illustrated by the following example:

(88) Väite, Û että kissa kiintyy paikkaan,claim+NOM that cat+NOM attach+PASS+3SG place+ILL

ei ihmisiin, â pitää paikkansa, jos .. [HKV]NEG+3SG person+PL/ILL hold+3SG place+POS/3

‘The claim/allegation that a cat gets attached to a place and not to people is true if ...’ (seeAppendix 1).

While embedding is relevant in projection, it seems to me that taxis is not a

relevant concept. As mentioned above, projection and expansion are two very

different phenomena and it is questionable to analyse the relationship between the

clauses in a projecting clause complex in Finnish in terms of interdependency, as

Halliday (1985a) does for English. According to Halliday (1985a), there is a para-

tactic (co-ordinative) relationship between a quoting and a quoted clause and a

hypotactic (subordinative) relationship between a reporting and a reported clause.

McGregor (MS) also criticizes Halliday’s view that projection involves

hypotaxis and parataxis. He points out a number of fundamental differences be-

tween projection and expansion in English, some of which apply to Finnish as

well. One of the criteria used to distinguish between parataxis and hypotaxis in

section 4.5.2 above was the fact that in parataxis, where one clause is initiating

and the other is continuing, the order of the clauses cannot be changed. This does

not apply to a quoted clause, which Halliday regards as paratactic: either the pro-

jected or the projecting clause can come first. While this variation in order seems

to me to be more characteristic of written Finnish, the fact that a change of order

is possible is a reflex of the fact that the relationship between the clauses in a

projected complex is different to the relationship between clauses in a paratactic

expansion. Furthermore, hypotaxis was characterized by the fact that the order of

Page 174: systematic functional linguistic

166

1 This analogy was used by Esa Itkonen in a lecture given at the University of Helsinki.

the clauses can be changed, whereas with a reported locution or idea the project-

ing clause always precedes the projected clause. Another distinguishing factor is

that the projecting clause can be omitted in a quoted locution in spoken Finnish.

In these instances, the quote is indicated by a change in voice quality. This is not

possible with paratactic expansion. A feature, which applies to Finnish but not to

English, is the availability of mood options in the projected clause. A projected

interrogative remains in the interrogative form in both quotes and reports in Finn-

ish. With hypotaxis, on the other hand, the dependent clause cannot be in the

interrogative.

The analysis of clause complexes presented in this chapter is meant to serve

as an outline. Future research will need to make more delicate distinctions, and

the analysis presented here needs to be supplemented by a phonetic analysis of

complexes ) both acoustic and auditory. There are also various types of infinitive

(nominalized) forms in Finnish that need to be incorporated into an account of

projection in Finnish. (Some of these infinitive forms will be discussed briefly in

Chapter 6 with reference to mental and verbal processes.) It also needs to be

stressed that the distinctions that have been made in this chapter should not been

seen as categorial distinctions (see 2.4.12, p. 65). As Halliday (1985a: 219-220)

points out, the fact that two categories are distinct in principle does not mean that

every instance is clear-cut: there will always be anomalous and borderline cases.

Robinson (1975: 21) has expressed this idea very clearly:

For if we cannot say at what precise moment day becomes night does that mean we don’tknow day from night?

A similar and more appropriate analogy relates to a social rather than a natural

phenomenon: although we may not be able to precisely define the cut-off point

between being rich and being poor, this does not mean that there is no distinction

between the rich and the poor in the world today.1

Page 175: systematic functional linguistic

167

1 The terms modalization and modulation can be seen as being roughly equivalent to thelogically based notions of epistemic and deontic modality.

Chapter 5

Interactional Structure in the Finnish Clause

5.1. Preliminary Remarks

In SF theory, one of the types of meaning realized in the clause is interpersonal

meaning, one aspect of which is concerned with the interactional role of the

speaker. This chapter is concerned with interactional structures, i.e. mood struc-

tures, in the Finnish clause. Other aspects of interpersonal meaning, e.g. modality

and modulation, will not be considered except where they bear on the discussion

of mood. There are at least two reasons for this exclusion. Firstly, some modal

expressions are dependent on mood options: for example, many modality and

modulation options are not available in the imperative mood. Thus, modality and

modulation constitute more delicate options in an interpersonal network, and thus

fall beyond the scope of this study. Secondly, the interpretation and analysis of

modality options is not so much dependent on formal mood categories, but on

what Halliday (1985a: 342) refers to as rhetorical functions. For example, the

interpretation of clauses with täytyy ‘must’ (e.g. Sinun täytyy olla huolima-

ton/Sinun täytyy olla huolellinen ‘You must be careless/You must be careful’) in

terms of modalization (probability) or modulation (obligation)1 does not depend

on the mood of the clause in which it occurs (i.e. declarative, interrogative, imper-

ative), but on whether the speaker is trying to get someone to do something (i.e.

whether the clause is oriented to action) or whether it is oriented to the exchange

of information.

This chapter discusses how systemic-functional theory attempts to bridge the

gap between the vast range of rhetorical functions that a discourse, text or conver-

sational analyst might want to distinguish in spoken or written text and the small

number of grammatical categories ) mood categories ) through which these func-

tions are realized. While the analysis presented here relies heavily on the insights

Page 176: systematic functional linguistic

168

1 The notion of a rhetorical function can obviously be linked to the logico-philosophicalnotion of a speech act, as discussed for example, in philosophical discourse by Austin (1962)and Searle (1969) and, from a linguistic perspective, by Levinson (1983), and to the notionof a “speech act verb”, as discussed by philosophically oriented linguists (e.g. Verschueren1980, Wierzbicka 1987).

of Halliday’s (1984, 1985a) analysis of mood in English, it differs from Halli-

day’s analysis in important respects. While this alternative model arose when I

started looking at mood in Finnish, it seems to me that it is also applicable to

English. As I shall argue, there are some problems inherent in Halliday’s analysis

of mood in English. On the other hand, I do not assume that what I am presenting

reflects anything universal about the organization of language. There are lan-

guages such as Gooniyandi (an Australian language spoken in the north of West-

ern Australia) that do not make the same sorts of grammatical distinctions that are

discussed here (McGregor 1990a: 382-383; see also section 5.2.3). However as

McGregor emphasises, this does not mean that speakers of Gooniyandi do not, for

example, ask questions: questions in Gooniyandi are rhetorical functions.

5.2. From Rhetorical Functions to Grammatical Categories

When speakers (or writers) refer to what someone has said (or written), they often

explicitly assess the function of the utterance in the context in which it occurred,

i.e. its rhetorical function (Halliday 1985a: 342). A rhetorical function can be

realized either as a noun (e.g. we can recognize something that someone has said

as “a threat” or “a promise”) or as a verb (e.g. “X threatened Y” or “X promised

to do something”). Nouns and verbs such as (to) threat and (to) promise are,

thus, used to construe verbal (or linguistic) processes or acts. In any language,

there are countless rhetorical functions that speakers of the language recognize.

Halliday (1985a: 342) lists over 60 such functions for English, functions such as

offering, promising, threatening, vowing, undertaking, ordering, requesting, en-

treating, urging, persuading, commanding, instructing ... and adds the proviso that

he has listed just a few. Austin (1962: 150) claimed that there were well over a

thousand expressions like this in English.1

Page 177: systematic functional linguistic

169

1 The translations are very approximate, focusing on one or two of what I consider to be themost central uses. The fact that I have given a translation does not imply that there is a one-to-one correspondence between rhetorical functions in English and Finnish or that the set ofrhetorical functions are universal.

Similarly, in Finnish, we can recognize a rich array of rhetorical functions,

as reflected, for example, in the following verbs:1

väittää ‘(to) claim’, huomauttaa ‘(to) point out/remind’, vastata ‘reply’, valittaa ‘com-

plain’, kannella ‘complain (to a higher authority)’, tokaista ‘to speak out brusquely/snap’,

tarjota ‘(to) offer’, tyrkyttää ‘force something upon someone’, käskeä ‘command/order/tell

someone to do something’, määrätä ‘command/order (often given by an official body)’,

painostaa ‘force/pressurize’, pakottaa ‘force (someone to do something)’, pyytää ‘request’,

vaatia ‘demand’, luvata ‘promise’, vannoa ‘swear’, sitoutua ‘undertake/bind oneself’,

haukkua ‘speak negatively about someone/abuse’, herjata ‘abuse/insult’, moittia ‘blame’,

syyttää ‘accuse’, paheksua ‘disapprove of’, arvostella ‘criticize’, pilkata ‘make fun of’,

uhata ‘threaten’, leuhkia ‘boast’, ilmoittaa ‘announce’, julistaa ‘proclaim’, paljastaa ‘re-

veal’, vakuuttaa ‘assure’, inttää ‘insistently assert or raise objections’, kysyä ‘ask’, udella

‘ask (pryingly)’, etc. etc.

A clause such as mulla on päänsärky ‘I’ve got a headache’, however, is not

grammatically marked as an instance of a particular rhetorical function. It is real-

ized as a declarative, but there is nothing that tells us whether it is, for example, a

reply (to a question), a report (to a doctor), a complaint (to a friend), or a rebuff

(to a lover). The fact that it grammatically realized as a declarative does not tell

us very much.

As Halliday (1985a: 342-343) points out, rhetorical function is made mani-

fest by a variety of factors. These include linguistic features such as intonation

and the co-text, i.e. the utterances that precede or follow the utterance in question.

Rhetorical function can also be made manifest in paralinguistic and behavioural

features (e.g. voice quality, facial expression, and gesture) and features of the

context of situation and the context of culture (e.g. the relative status of the partic-

ipants may be a significant factor in interpreting something as a command or a

request, and status is something that is culturally defined). Similar observations

are made by Hakulinen & Karlsson (1979: 261).

Page 178: systematic functional linguistic

170

1 Given that there are likely to be well over a thousand rhetorical functions in Finnish, itseems to me that the functions that we might want to recognize depends on the type of datathat we have and the purpose for which the analysis is being proposed. Another problem isthat a rhetorical function is a textual phenomenon, and, as such, it is embedded in the fluxof “social heteroglossia” (Bakhtin 1981: 263). Thus, the interpretation of a rhetorical functionis variable: what one person regards as questioning, for example, could be interpreted byanother as prying.

I am not concerned with trying to establish a set of rhetorical functions for

Finnish, if indeed this is a feasible thing to do.1 What I am concerned with is the

relationship between these rhetorical functions and the grammatical categories

through which they are realized. The question I am addressing is why is it that we

can distinguish mood categories such as declarative, interrogative, and imperative

in Finnish, English and many other languages, and what is the relation between

these categories and the rich array of rhetorical functions in spoken or written

text. The following Finnish clauses, for example, could realize a number of rhe-

torical functions.

DECLARATIVE:

(1) Sulla on toinen. [TSM (Novel): 316]you+ADE be+3SG other-NOM

‘You’ve got another (one).’

(2) nyt se on mun vuoro [CA6:10]now it-NOM be+3SG my turn-NOM

‘Now it’s my turn.’

(3) sä olit mökilla viime viikonloppuna [Tel1:1]you+2SG be+2SG cottage+ADE last weekend+ESS

‘You were at the summer cottage last weekend.’

(4) Emma! Sinne et mene! [OH: mother to toddler]Emma! to-there NEG+2SG go‘Emma! You’re not to go in there!’

Page 179: systematic functional linguistic

171

1 I have translated the Finnish imperative with an interrogative in English. An imperative withhigh clause-initial pitch is not blunt or impolite in Finnish. See also Shore (1991b: 244 ff.).

(5) Et ihan totta viittis huutaa kovaaNEG+2SG quite true+PAR bother/incline+CON yell hard/loud+PAR

meitä tuijoitetaan [AR: 126]we+PAR stare+INDEF

‘You really don’t have to talk so loud/Do you really have to talk so loud, people are star-ing at us.’

(5) Joensuu on Pohjois-Karjalan läänin pääkaupunki. [TB: 1]Joensuu be+3SG North-Karelia+GEN province+GEN main+town-NOM

‘Joensuu is the capital of the province/county of North Karelia.’

INTERROGATIVE:

(6) Olisiko Pekka tavattavissa? [typical telephonebe+CON(3SG)+Q Pekka-NOM meet+INDE+PTC+PL+INE (available) request]‘Is Pekka available?/Could I speak to Pekka please?’

(7) juotko kahvia?drink+2SG+Q coffee+PAR

‘Will you have some coffee?/Do you drink coffee?’

(8) tiedätsä mikä on bussi tampereeks? [Tel2: 14]know+2SG+you what/which-NOM is+3SG bus-NOM Tampere+TRA

‘Do you know what they call a bus in Tampere (dialect)?’

(9) käviskö sulle että sä pitäisit puheen?go+CON(3SG)+Q you(2SG)+ALL that you-NOM hold+CON+2SG speech+GEN

‘Would it be convenient for you to give/Would you give a speech?’ [Tel1: 4]

(10) mikäs ikkuna on auki? which-NOM+TIS window-NOM is+3SG open

pitäisikö se laittaa kiinni? [OH: Teacher to students]must+CON+3SG+Q it-NOM put+3SG closed

‘Is there a window open? Should it be closed?’

IMPERATIVE:

(11) Osta mulle purkki raejuustoa.1 [OH: exchangebuy+2SG/IMP I+ALL carton-NOM cottage+cheese+PAR between colleagues in‘Would you get (buy) me a carton of cottage cheese (please)?’ a hairdressers]

(12) Susanna paa se radio auki. [OH: exchangeSusanna put+2SG/IMP it/that-NOM radio-NOM open between friends]‘Susanna, would you turn on the radio (please)?’ (See previous footnote.)

Page 180: systematic functional linguistic

172

(13) Ota kakkua.take+2SG/IMP cake+PAR

‘(Do) have some cake.’

(14) Sano sitten Pekalle terveisiä. [Tel3: 13]say+2SG then Pekka+ALL regards+PL+PAR

‘(Please) give my regards to Pekka then/Say hello to Pekka for me then, won’t you?’

(15) Ole hyvin peloissasi. [Ad1 (advertisement forBe+2SG/IMP very afraid+PL+INE+POS/2SG. Kärpänen II (The Fly II)]‘Be very afraid.’

(16) Puhuttakoon mitä puhutaan. [Leskinen 1970: 71-72]speak+INDE+IMP which/what+PAR speak+INDE

‘Let them/people say what they like.’

(17) Älköön kukaan julkisesti käyttäytymiselläänNEG/IMP+IMP/3 anyone/no-one publicly behaviour+ADE+POS/3

häiritkö yleistä järjestystä tai loukatko säädyllisyyttä.disturb+IMP general order/peace or offend+IMP propriety/decency

‘Let no-one behave publicly in such a way as to disturb the public peace or offend theproprieties.’ [Helsinki City Ordinance 3 §]

Even on the basis of these examples removed from their context, we could recog-

nize an array of rhetorical functions that may have been realized in the original

text. From a grammatical perspective, however, we can distinguish only three

mood options. This chapter is concerned with the meaning of these mood options

in Finnish.

5.2.1. The Clause as an Exchange or Interactive Event

According to Halliday (1984: 11-16; 1985a: 69), an exchange or an interactive

event involves two interactants and it can be viewed in terms of two parameters:

1) the “commodity” exchanged and 2) the role of the initiator in the exchange.

From a linguistic viewpoint, the commodity exchanged can either be a) linguistic

(i.e. information) or b) non-linguistic (i.e. “goods-&-services”). In other words,

when we interact with each other, we can either simply talk or write to another

person (i.e. we exchange language) or we can get another person to obtain some-

thing or do something (i.e. we exchange goods-&-services). The role of the initia-

tor in an exchange can either be the i) giver or ii) demander (of information or

goods and services).

Page 181: systematic functional linguistic

173

On the basis of these parameters, Halliday defines four primary speech func-

tions: STATEMENT, QUESTION, OFFER and COMMAND:

STATEMENT: give language

QUESTION: demand language

OFFER: give goods-&-services

COMMAND: demand goods-&-services

Halliday (1985a: 70-71) refers to STATEMENTS and QUESTIONS as propositions

and to OFFERS and COMMANDS as proposals. It needs to be stressed that these are

abstract speech functions, not rhetorical functions. (To make this clear, I have

written them in capital letters.) Speech functions have been abstracted from the

rhetorical functions discussed earlier, they constitute a semantic interface between

rhetorical functions and the mood options, declarative, interrogative and impera-

tive. One of the examples cited earlier, Sulla on toinen, occurs as what might be

considered an accusation in the following text (from the novel Tavallisen

suomalainen mies by Alpo Ruuth (pp. 315-316).

[Ossi is having an affair with Suvi; he is waiting for her in her flat.]

Ossi istui koko illan tuijottamassa televisiota. Tunnelmat vaihtuivat tunneittain, hän vihasi japelkäsi vuoron perään. Eniten hän pelkäsi sitä että Suvi toisi jonkin kaverin mukanaan ...

‘Ossi sat all evening with his eyes glued to the television. His feelings changed by the hour,he felt anger and fear by turns. Most of all he was frightened that Suvi would bring someoneelse home ...’

Figure 5-1: Variables in an Interactive Event

Figure 5-2: Primary Speech Functions

Page 182: systematic functional linguistic

174

1 In the original diagram, Halliday locates ellipsis in the mood network. Although ellipsis maybe regarded as textual, it is apparently included in mood because the potential for ellipsis hasto be specified interpersonally (Martin 1981: 55). However, this part of the network need notconcern us here.

– Sulla on toinen.‘You’ve got someone else/another (lover).’

– Ei ole.‘No I haven’t.’

– Varmasti on.‘You must have.’

The clause sulla on toinen ‘you’ve got another (lover)’ has the speech function

STATEMENT, not because it functions rhetorically as a statement, but because

what is being exchanged is linguistic and the initiator’s role is that of giver.

In order to consider the grammatical system, Halliday then looks at the rele-

vant grammatical options available in a particular language, which, in Halliday’s

analysis, is English. According to Halliday, the grammatical options in English

can be represented by the following (very basic and simplified) mood network:1

A similar network could be drawn for the basic mood options in Finnish. As men-

tioned in 4.2, minor clauses are those in which mood options are not available.

Figure 5-3: Mood Options in English

Page 183: systematic functional linguistic

175

To recapitulate, contextual (or situational) options (e.g. [give: information])

realize semantic options (e.g. [STATEMENT]), which in turn realize grammatical

options (e.g. [indicative]). In spite of the way in which this has just been phrased,

realization is not a series of steps that proceed in a certain direction, at least, this

is not so ontologically. Methodologically, however, what has happened is that the

starting point is the context of situation, and on the basis of this, the semantic

options are defined. Then, the analysis turns to the lexicogrammatical form of

English and the options that are available in English. This analytical procedure

most clearly reveals itself when one considers the category OFFER, which has not

been established on the basis of the primary mood options in English, but rather,

on the basis of the parameters described above. Halliday sees this procedure as

“shunting” (see Halliday 1961: 254, 286), i.e. moving back and forth between

linguistic levels, or, to put it in another way, linguistic analysis at one level neces-

sarily involves keeping an eye on other levels.

Thus, for Halliday, the semantic level, which is analysed from an interac-

tional perspective in terms of the abstract speech functions defined above, acts as

an interface between the context and the lexicogrammatical form of a language.

The downward pointing arrow in the following diagram is the symbol for realiza-

tion in systemic-functional theory. In the light of what was said above, a more

appropriate symbol would be a double-headed arrow (a), as pointed out in 2.4.3

(p. 44).

RHETORICAL FUNCTIONS IN SPOKEN OR WRITTEN TEXT

`SPEECH FUNCTIONS

`MOOD OPTIONS

One of the problems inherent in this model is the mismatch between the

situationally defined semantic notions (or contextual semantic notions) and the

Figure 5-4: Semantics as an Interface

Page 184: systematic functional linguistic

176

mood options in English. In order to bridge the gap, Halliday invokes the notion

of “congruency” (to be discussed in the next section).

The SF notion of a speech function should not be equated with the tradi-

tional notion of a speech function found in some grammar books (e.g. Hakulinen

and Karlsson 1979: 276-293). Traditional notions such as statement, question,

command and exclamation are semantic relabellings of grammatical categories

and are not explicitly intended to make a link with (abstractions from) interaction-

al options in the context of situation. On the other hand, Halliday’s speech func-

tion should not be confused with the way in which the notion is used in discourse

analysis. It is not a discourse notion, but a grammatical notion. A. Hakulinen (A.

Hakulinen ed. 1989: 45-46), for example, has obviously misunderstood these

speech functions to be rhetorical functions; she makes no mention of the rhetori-

cal functions listed by Halliday in the chapter concerned with phenomena that are

“beyond the clause” (1985a: 342). She criticizes Halliday for postulating the four

primary speech functions “with little or no empirical evidence”. The grammatical

evidence on which Halliday’s analysis is based are the grammatical categories

declarative, interrogative and imperative, which are also assumed in

ethnomethodological conversation analysis (see list in 1.5, p. 10). Hakulinen,

however, seems to assume that Halliday’s An Introduction to Functional Gram-

mar is discourse analysis not grammar: she says Hallidayn ongelma on kaikille

diskurssintutkijoille yhteinen ‘Halliday’s problem is common to all discourse

analysts’. Hakulinen’s criticism is ironic in that it is concerned with “precon-

ceived categories”, yet she criticizes Halliday on the basis of her own precon-

ceived categories (cf. the discussion of explicit and implicit assumptions in sec-

tion 1.5).

As can be seen from the above, Halliday’s model reflects most appropriately

face-to-face interaction, which can be regarded as the paradigm case of linguistic

interaction. This view is shared by other linguists (including Lyons 1977a: 589ff,

Fillmore 1981: 152, Levinson 1983: 54). For example, according to Fillmore:

Page 185: systematic functional linguistic

177

The language of face-to-face conversation is the basic and primary use of language, all othersbeing best described in terms of their manner of deviation from that base.

However, Halliday’s model possibly goes a step further: the basic and primary

use of language is not simply face-to-face conversation, which, after all is primar-

ily concerned with the exchange of language (i.e. language is constitutive). The

basic and primary use of language, as reflected in Halliday’s model, is face-to-

face language-in-action: for example, the ordinary everyday exchanges of lan-

guage, action and goods-&-services that take place in the home or in the play-

ground or when people are shopping or working.

5.2.2. Congruent and Metaphorical Realization

Halliday relates the speech functions defined above to the grammatical options of

English by postulating what he refers to as a congruent relationship between

them. A congruent realization relationship is sometimes seen in terms of typical-

ity (e.g. Halliday 1984: 14; Halliday 1985a: 320; cf. below however): a STATE-

MENT is congruently (or typically) realized as a declarative, a QUESTION as an

interrogative and a COMMAND as an imperative. With OFFERS, there is no congru-

ent realization: they can be realized in various ways.

SPEECH FUNCTION CONGRUENT REALIZATION

statement declarative

question interrogative

command imperative

offer —

It seems to me, however, that congruency cannot not be equated with typi-

cality. According to Berry (1987: 58-59), typicality can be understood either 1) as

a statistical norm or 2) as what is perceived as being typical by speakers of a lan-

Figure 5-5: Congruent Realizations of Speech Functions

Page 186: systematic functional linguistic

178

1 See 2.3.4 and 2.4.12.

2 As Halliday (1985a: 329-30) points out, however, “the concept of plain and simple is veryfar from being plain and simple”.

guage. As Halliday (1984: 14) himself suggests, it is not entirely self-evident that

asking someone to do or get something for you is typically realized as an impera-

tive in English. (See also Levinson 1983: 264,275.) It also seems questionable to

equate congruency with typicality in the sense of what is perceived as being typi-

cal by language users. If one were to ask a speaker of English to consider a situa-

tion in which she or he were trying to get someone else to do something such as

close a door or window (i.e. demand goods-&-services), then it is unlikely that

she or he would suggest an imperative. It seems to me that, in the majority of

instances, 2. person imperatives realizing the option [demand: goods-&-services]

tend1 to be used in English if the speaker is in a position of power (e.g. parent or

teacher to child, sergeant to private). Exceptions to this include certain anony-

mous written genres (such as operating instructions and recipe books), certain

requests for a linguistic action e.g. Give my love to Nigel, and actions that are

regarded as benefiting the addressee e.g. Take a holiday.

Rather than relate congruency to typicality, it seems more appropriate to

think of congruent realization in terms of Halliday’s (1984: 14) alternative sug-

gestion as “a kind of baseline” or as a “maximally simple” way of expressing

things.2 In English, for example, this could be thought of as the way in which

young children typically express things or the way in which something would be

expressed in an emergency (e.g. Quick! Get a doctor!). The way in which a base-

line is characterized, however, is language-specific: in Finnish, for example, ask-

ing for goods-&-services from a peer or an equal is commonly realized as an

imperative if no imposition is involved, as indicated by examples 11 and 12

above (page 171), and by the following example, which was addressed to a col-

league/friend heading for the photocopying machine to take a copy of a short

article:

Page 187: systematic functional linguistic

179

1 See Karvonen (1991, 1992) for a discussion of grammatical metaphor in Finnish and itsapplication in the analysis of text.

2 Cf. however quote from Halliday (1985a: 321) on page 181 below.

(18) Ota mullekin kopio. [OH: exchange between colleagues.]take+IMP/2SG I+ADE+TIS copy-NOM

‘Would you (please) take a copy for me too?’

Crucial to Halliday’s analysis of mood is the idea that the notion of meta-

phor can also be extended to grammatical phenomena, (see 2.4.7 (p. 54), 2.4.13

(p. 67); Halliday 1985a: 319-345, Ravelli 1988).1 As Halliday (1985a: 320-321)

points out, the term metaphor is usually used to refer to the “non-literal” use of

words: a word is said to be used with transferred meaning, e.g. flood in the fol-

lowing example:

(19) A flood of protests poured in following the announcement.

However, instead of a form-based approach to metaphor, e.g. instead of looking

at a lexical item such as flood and analysing the ways in which it is used with

transferred meaning, it is equally valid to take a meaning-based approach to meta-

phor. Thus, the example above (19), expresses a meaning that could have been

expressed in a more simple way.2 For example:

(20) A lot of people wrote in and protested after they had heard the announcement.

If metaphor is regarded as a variation in meaning, then example 19 can be seen as

a metaphorical variation of 20. Once we move away from a form-based lexical

approach to a wider meaning-based approach, then the notion of metaphor can be

extended to apply to the meanings realized at the rank of clause: metaphorical

variation in systemic-functional theory is considered to be a grammatical as well

as a lexical phenomenon.

If the realization relationship between a speech function and a mood option

is not a congruent one (as in Figure 5-5 on p. 177 above), if, for example, a de-

Page 188: systematic functional linguistic

180

1 The person addressed is about to withdraw some money from an automatic teller. The unitreferred to is 100 marks (approx. $30 Australian).

2 It can, of course, be even more simply realized in a minor clause, as, for example, in a bankrobbery in Finnish: rahat tänne [money+NOM/PL here] cf. “your money or your life”.

mand for goods-&-services is not realized as an imperative but as an interroga-

tive, then the relationship is metaphorical. Both of the following examples were

demands for goods-&-services (a loan) from a close friend.

(21) Lainaa kymppi. [OH: exchange between friends]lend+IMP/2SG ten+NOM

‘(Would you) lend me 10 marks (approx. $3 Aust.)?’

(22) Olisko liikaa pyydetty et [OH: exchangebeCON(3SG)+Q too much+PAR request+INDE+PTC that between friends]

sä heittäisit mulle yhen yksikön?you-NOM throw+CON+2SG I+ADE one+GEN unit+GEN

‘Would it be too much to ask for you to lend me a unit.’1

Both examples express more or less the same meaning, i.e. they are both requests

for a loan. The meaning is not, of course, exactly the same, but we can recognize

that one is a variation of the other. The request is realized in the simplest possible

way in example 21: it is realized congruently as an imperative.2 In example 22,

on the other hand, the meaning is expressed in a different way: here the request

for a loan is realized metaphorically as an interrogative.

As with any metaphor (cf. fossilized lexical metaphors such as pöydän

jalka ‘the leg of the table’), a grammatical metaphor retains its literal or congru-

ent meaning — otherwise we would not be able to recognize it as a metaphorical

expression. Consequently, with grammatical metaphor, there is no anomaly in-

volved in responding to the congruent as well as the metaphorical meaning, as in

the following example from a telephone conversation:

Page 189: systematic functional linguistic

181

(23) – onks Pekka kotona be+3SG+Q Pekka-NOM home+ESS

‘Is Pekka at home?’

– joo hetkinenyes moment‘Yes just a moment.’

Thus, something can be congruent and metaphorical at the same time, i.e., as

Levinson (1983: 269) puts it: “both readings are simultaneously available and

utilized”.

The example above illustrates another important point about metaphor. In

some genres, the metaphorical mode of expression has become the norm: a re-

quest for someone to come to the telephone, for example, is generally realized by

an interrogative in Finnish. But anyone who has asked a small child who has an-

swered the telephone Onks sun äiti kotona? ‘Is your mother at home?’ will be

aware that the interrogative is metaphorical. In my experience, the child generally

responds to the question, and will not automatically call for her or his mother to

come to the telephone but needs to be prompted by a follow-up such as Pyytäisit-

kö hänet puhelimeen ‘Would you ask her to come to the telephone?’.

Another general point about metaphor needs to be underscored: metaphor is

neither inherently good or bad, it is a natural semiotic (meaning-making) re-

source in language. According to Halliday (1985a: 321):

For any given semantic configuration there is (at least) one congruent realization in thelexicogrammar. There may be others that are in some respect transferred, or METAPHORICAL.

This is not to say that the congruent realization is better, or that it is more frequent, or eventhat it functions as a norm; there are many instances where a metaphorical representation hasbecome the norm, and this is in fact a natural process of linguistic change. Nor is it to suggestthat a set of variants of this kind will be totally synonymous; the selection of metaphor isitself a meaningful choice, and the particular metaphor selected adds further semantic fea-tures. But they will be systematically related in meaning, and therefore synonymous in certainrespects.

Page 190: systematic functional linguistic

182

Analysing the relationship between speech functions and mood in terms of

metaphor has clear advantages over other approaches. It avoids the inherent con-

tradiction involved in a speech act approach: while it is somewhat anomalous to

regard a speech act as being both direct and indirect at the same time (cf. Levin-

son 1983: 268 ff.), something can be congruent and metaphorical at the same

time. Moreover, when contrasted with the notion of a hedge used in

ethnomethodological conversation analysis (see, e.g., A. Hakulinen ed. 1989: 118

ff.) the notion of grammatical metaphor goes linguistically a lot deeper since it

attempts to bridge the gap between grammatical categories and spoken or written

text. It also brings together related phenomena: e.g., it allows us to treat examples

such as the following example as mood metaphors:

(23) Sä tuut tähän nyt! [OH: parent to child]you come+2SG here+ILL now/at once‘You will come here at once.’

This was addressed by a parent to an unruly child in a tube train. I have also heard

a parent say the following in a similar situation:

(24) Tuu istumaan. [OH: parent to child]come+2SG/IMP sit+INF+ILL

‘Come and sit down.’

Presumably neither of these would be treated as a hedge by conversation analysts,

yet its seems to me that linguistically, at least, they fall under the same general

phenomenon as examples 21 and 22 above (Lainaa mulle kymppi ‘(Would you)

lend me ten marks’ and Olisiko liikaa pyydetty et sä heittäisit mulle yhen yksi-

kön? ‘Would it be too much to ask for you to lend me a unit?’).

Treating instances like this in terms of grammatical metaphor also allows us

to treat the mood metaphors I have discussed as one facet of a wider, pervasive

meaning-making process in language: under the same general phenomenon, Halli-

day (1985a: 319-345) discusses metaphors in modality and in transitivity struc-

tures in the clause. While the notions of congruency and metaphor are problem-

Page 191: systematic functional linguistic

183

1 As pointed out in Chapter 1, Bakhtin’s (1981: 263) notion of “social heteroglossia” pertainsnot only to the phenomenon being observed, i.e. language, but also to the observer, thelinguist.

atic in that they are susceptible to variable interpretation,1 so too is the notion of

a hedge. At least, a systemic-functional approach is an attempt to ground the in-

terpretation of congruency and metaphor in an explicit and linguistically moti-

vated way.

5.2.3. Problems with Halliday’s Analysis

As mentioned earlier, it seems to me that there are some problems with Halliday’s

analysis. The problems are connected with the postulation of the speech function

OFFER. Halliday (1985a: 342) regards a threat such as I’ll shoot the pianist as an

OFFER, i.e. as the giving of goods-&-services. It is not entirely clear to me that a

clause that functions rhetorically as a threat actually involves any exchange of

goods-&-services.

Halliday (1984: 19-20) comments on the fact that exchanges of goods-&-

services are problematic from a grammatical point of view. He notes that ex-

changes of information are more likely to be congruent than exchanges of goods-

&-services. This can be explained ) at least partly ) by the fact that linguistic

exchanges are essentially different from exchanges of goods-&-services: a pro-

cess or action can take place independently of language. For this reason, accord-

ing to Halliday (1984: 20), languages do not display clear-cut grammatical cate-

gories corresponding to OFFERS and COMMANDS. Halliday refers to the imperative

as a “fringe category, teetering between finite and non-finite ... having either no

distinct clause or verb form or else one that is only minimally distinguished”.

This characterization of the imperative is surprising given that Halliday’s

analysis of mood is based on language in face-to-face interaction, a central aspect

of which is language as a means of getting things done. If our analysis of mood is

based on language-in-action, then it seems incongruous to regard the imperative

Page 192: systematic functional linguistic

184

1 This example and example 15 on p. 172 (Ole hyvin peloissasi ‘Be very afraid’), which alsohas an imperative form of olla ‘be’, may be considered strange by some speakers of Finnish.They are all, nevertheless, authentic examples; and, if the imperative in Finnish is interpreted“orientation to a non-linguistic (action) response”, as I shall suggest, then there is nothing oddabout these examples.

as a fringe category. In fact, a “verb form ... that is only minimally distinguished”

could be regarded as the most basic and fundamental verb category. A 2. person

singular imperative, for example, is morphologically the simplest verb form in

both Finnish and English; it is also amongst the first verb forms learnt by a child

(Toivainen 1980: 32). In contrast to English, however, verbs in both standardized

Finnish and in all Finnish dialects are inflected for person and number in the im-

perative (see figures 3-4 (p. 80) and 3-5 (p. 81) in Chapter 3).

The examples of the imperative quoted earlier (page 171) and the following

examples illustrate some of the uses of the imperative in Finnish:

(25) Tulkaa tänne.come+IMP+2SG here‘Come on over here/(Why don’t you) Come here.’

(26) Painu helvettiin!descend+2SG/IMP hell+ILL

‘Go to hell!’

(27) Ajattele nyt vähän.think+2SG/IMP now a little‘Now think about it a bit.’

(28) Älä uneksi siitä, ole sitä.1 [HS 11.8.1991 p. B1]NEG/IMP+2SG dream it+ELA be+2SG it+PAR

‘Don’t dream (about) it, be it!’

(29) Pelkää rauhassa! Halua lisää! [Ad2]fear+IMP/2SG peace+INE want+IMP/2SG more+PAR

Näytä tunteesi! Naura enemmän!show+IMP/2SG feeling+POS/2SG laugh+IMP/2SG more+COM+GEN

‘Be afraid in peace! Want more! Show your feelings! Laugh more!’

Page 193: systematic functional linguistic

185

1 The example is from Edna O’Brien, Johnny I Hardly Knew You (Weidenfield & Nicolson,London, 1977) p. 21.

It is difficult to see many of these examples in terms of an exchange of

goods-&-services. This is also true of many English imperatives:

(30) I was wearing very high heeled shoes. He said take them off.1

(31) Take a look at this.

It seems to me that what is at issue in imperatives in Finnish and English is orien-

tation to action, where action need not necessarily be understood in physical

terms but is simply a label for a response that is not information-oriented, i.e. not

directed to the exchange of information. The action may, of course, be a linguistic

action, as in Sano sitten Juhalle terveisiä or its English equivalent Say hello to

Peter for me, where “saying” can be seen as a mode of doing. (At a deeper philo-

sophical level, all language is a mode of action ) this is the point of this chapter.

The reference here to saying as a mode of doing is on a more concrete level.)

The alternative analysis that will be proposed in the next section departs

from Halliday’s analysis in that the distinction between giving and demanding is

not applied in the case of goods-&-services. This analysis of Finnish is similar to

both Fawcett’s (1980: 104 ff.) and Butler’s (1988) analysis of English. Butler

(1988: 150), for example, refers to “action-seeking acts”. Counter-arguments to

this re-analysis will be discussed in the next section. Before addressing these,

however, I shall look at another related problem with Halliday’s analysis. The

other problem with Halliday’s analysis of mood is connected with the function

Subject in an OFFER. According to Halliday, the grammatical subject in English

realizes an interpersonal function. Because the traditional definition of grammati-

cal subject (number and person agreement with the finite verb) is of limited appli-

cability in English, Halliday (1985a: 71-73) regards the subject in English as the

element that is picked up if a tag is added to a clause: you won’t give it away,

will you? In Halliday’s view, the grammatical subject in English is a meaningful

category, not simply an empty “surface” phenomena.

Page 194: systematic functional linguistic

186

1 In Halliday’s discussion of English, he uses Subject (with a capital letter) to refer to boththe element that is picked up in the tag and the function that it realizes. I prefer to make adistinction between the formal category (subject) and the functional category (Subject). Thereason for this is that Halliday’s (1985a) description is of English, a language in which it canbe argued that the grammatical subject realizes an interactional function. The notion of agrammatical subject has also been applied to languages that do not have a mood tag, forexample, Finnish, where there is an element in the nominative case that agrees with the verbin person and number. As argued in Chapter 6, this element realizes an experiential functionin Finnish. It is also conceivable that what has been referred to as the subject in otherlanguages does not realize an interactional function.

The Subject is a function in the clause as an exchange. It is the element that is held responsi-ble: in which is vested the success of the clause whatever its particular speech function.(Halliday 1985a: 36-37.)1

The function of Subject (meaning ‘responsibility in an interactive event’) is

seen as being relevant to both propositions and proposals. According to Halliday,

with proposals (i.e. commands or offers), the meaning of “responsibility” can be

seen in fairly concrete terms:

It is perhaps easier to see this principle of responsibility in a proposal (a ‘goods-&-services’clause), where the Subject specifies the one that is actually responsible for realizing (i.e. inthis case, for carrying out) the offer or command. For example, in I’ll open the gate, shall I?(offer) the opening depends on me; in Stop shouting, you over there! (command) it is foryou to desist or otherwise. Hence the typical Subject of an offer is the speaker, and that of acommand is the person being addressed. (Halliday 1985a: 76.)

The problematic status of OFFER can be illustrated by the fact that in some offers

the entity actually responsible for carrying out the offer is not always the same as

the one picked up in the tag. For example, there are various grammatical realiza-

tions of offering someone a drink in English. These include:

(32) You’ll have a drink, ) Won’t you? (33) Would you like a drink? ) (Would you?)(34) There’s a drink on the table for you, if you want one.(35) Do have a drink. ) (Won’t you?)(36) What you need is a stiff drink. ) ?(Don’t you?/Isn’t it?)(37) A stiff drink is what you need. ) ?(Isn’t it?)(38) I could fix you a drink. ) (Could I?)

Page 195: systematic functional linguistic

187

While in many of the examples the acceptability of a mood tag is debatable ) the

only example that I feel completely confident about adding a mood tag to is the

first one ) it is quite clear that to the extent that it is acceptable, it does not pick

up the speaker (I) in most of the instances above. Presumably, as Halliday (1985a:

76) puts it, “the one that is actually responsible for realizing (i.e. in this case, for

carrying out) the offer” is the same in all cases: the person who is doing the offer-

ing.

The reason why offers are not grammatically coded is obvious: as pointed

out above, while the exchange of information necessarily involves language (or

some other semiotic system), many exchanges of goods-&-services can occur

without language, particularly the giving of goods-&-services: you can give

someone a kiss or a flower without saying anything. As Halliday points out

(1984: 11-12), if language is used in the exchange of goods-&-services, then its

role is often ancillary and two distinct processes occur, i.e. 1) the exchange of

language and 2) the exchange of goods-&-services.

English and Finnish are not unusual in not having a congruent way of realiz-

ing an offer, there appears to be no language in which an offer is congruently

encoded. The appeal of the category seems to stem from the fact that it fits nicely

in the symmetrical analysis postulated by Halliday (see e.g. Martin 1981:60). The

variables postulated by Halliday (give/demand; language/goods-&-services) may

in fact be facets of a European way of viewing information. It was mentioned

earlier that in Gooniyandi (McGregor 1990a: 382-83), there are no formal distinc-

tions in the clause that can be related to the parameters giving vs. demanding and

information vs. goods-&-services, i.e. there are no mood distinctions in Gooni-

yandi corresponding to the distinction between declarative and interrogative or

between indicative and imperative. McGregor suggests that this may be a reflex

of the different way in which Aboriginal societies view (linguistic) interaction

(see also Eades 1982).

As pointed out in Chapter 2, a grammar is “an interpretation of linguistic

forms” (Halliday 1985: xx) and, thus, any semantic distinction that is set up must

Page 196: systematic functional linguistic

188

be systematically reflected in the grammatical organization of language. While

Halliday’s approach provides an insightful basis for the analysis of interactional

options in the clause, the semantic categories it establishes are not based on

lexicogrammatical patterns. If it were, 1) the option [offer] would not have been

postulated and, moreover, 2) the grammatical option [imperative] would not be

considered as the realization of the option [demand: goods-&-services], as I have

argued above. If speech functions are established on the basis of grammatical

patterning of options related to the clause as an interactive event, we would end

up with a different set of speech functions. The approach that I am advocating is

presumably similar to the approach advocated by Martin (1981: 52) when he says

we should “let the grammar of English decide” ) or the grammar of Finnish, as

the case may be.

5.2.4. Alternative Analysis

Halliday analyses a typical interactive event in terms of a symmetrical system:

COMMODITY EXCHANGED

LINGUISTIC GOODS-&-SERVICES

GIVE statement offer

DEMAND question command

As indicated above, from a linguistic point of view, it is questionable to assume

this kind of symmetry. The alternative analysis that I have proposed for English

and Finnish is asymmetrical:

Figure 5-6: Halliday’s Symmetrical Model

Page 197: systematic functional linguistic

189

1 The distinction between modalization and modulation is comparable to the distinctionbetween epistemic and deontic modality made in philosophical semantics, except thatassessments of usuality or frequency (e.g. Mary usually knows the answer) are included undermodalization and assessments of inclination (e.g. I felt like going) under modulation.However, SF theory is not based on truth-conditional semantics, and, thus, modality andmodulation refer to the semantic space between positive (“yes”, “it is so”, “do it”) andnegative (“no”, “it isn’t so”, “don’t do it”). See further, Halliday 1985a: 85 ff., 324 ff.;Halliday in Kress (ed.) 1976: 189 ff.

ORIENTATION

LINGUISTIC ACTION

GIVE statementproposal

DEMAND question

There are at least two possible counter-arguments to the above re-analysis

that need to be addressed before proceeding to a more detailed discussion of

mood in Finnish. Halliday’s (1985a: 235-237, 334-342) analyses of modality and

projection (see 4.5.5, p. 159 ff.) could be used as evidence for the grammatical

realization of the semantic category of offer since both analyses are based on a

basic division between propositions (statements and questions) and proposals

(commands and offers). One of the relevant points in Halliday’s (1985a: 334 ff.)

analysis of modality, for example, is that the type of modality involved in a clause

will depend on its underlying speech function: modalization (expressions of prob-

ability and usuality) is relevant for statements and questions, and modulation (ex-

pressions of inclination and obligation)1 is relevant for offers and commands, as

reflected in the following examples from Halliday:

(39) Mary will probably know.

(40) Fred should tell them.

The first example is a modalized statement and the second a modulated command

(i.e. demand for goods-&-services). However, there is nothing in Halliday’s anal-

ysis which gives support to an independent speech function of “offer”. The differ-

Figure 5-7: Alternative Model

Page 198: systematic functional linguistic

190

ences could just as well be explained by a semantic distinction between orienta-

tion to a linguistic exchange and orientation to action.

Secondly, according to Halliday (1985a: 235-37), propositions (STATE-

MENTS and QUESTIONS) and proposals (COMMANDS and OFFERS) are projected in

different ways. For example, to simplify things somewhat, the reported clause in

a reported proposition is generally finite, e.g. Mary told Tom that she was happy

(reported clause: I’m happy) whereas with proposals, it can also be non-finite,

e.g. He told me to take off my shoes (reported clause, e.g. Take them off). How-

ever, once again there is nothing in this analysis that presupposes a distinction

between COMMANDS and OFFERS. Halliday’s analysis is based simply on the ba-

sic distinction between propositions and proposals, which in the alternative analy-

sis that is being put forward here corresponds to the difference between orienta-

tion to a linguistic exchange and orientation to action. Moreover, as Halliday

points out, reported proposals like He told me to take off my shoes shade into

structures of a causative kind (He got me to take off my shoes); with this alterna-

tive analysis, it is perhaps easier to understand the fuzzy area between reported

proposals and causatives. It seems to me that a continuum that moves from an

action-orientated exchange that is being reported to a causative structure is more

natural that a continuum that moves from the demanding of goods-&-services to

a causative. Furthermore, there are clauses like I’ll shoot the pianist, which, as

mentioned above, are difficult to see in terms of the exchange of goods-&-ser-

vices; yet they are projected like proposals (cf. He threatened to shoot the pia-

nist): the relevant semantic factor can be glossed as “orientation to action”.

Page 199: systematic functional linguistic

191

1 Finnish: sävy.

2 The only research that has been done on the tone system of Finnish is Hirvonen (1970).Unlike Halliday (1967c), however, Hirvonen does not base his analysis on naturally occurringconversation but on constructed dialogues read by informants. Hirvonen’s analysis focuseson the unmarked tones of mood options in Finnish (cf. Halliday 1985a: 284).

5.3. Interactional Options in Finnish

From an interactional perspective, the primary grammatical options in Finnish are

indicative and imperative. This is reflected by the fact that the indicative and im-

perative are distinguished by different inflections in the finite verb (see Chapter 3

for examples of these inflections). Moreover, there are various tense and modality

distinctions available in the indicative that are not available in the imperative.

However, the term “indicative” as it is being used here is somewhat misleading,

as its use here differs from the way the term is used in traditional, formal gram-

mars of Finnish (e.g. Ikola 1977:54-55; Karlsson 1983b: 138), where the indica-

tive contrasts with conditional and potential inflections in the verb. In this analy-

sis, the term indicative says nothing about options in modality (i.e. declarative vs.

conditional or potential).

As discussed in 5.1, modality and modulation will not be considered in this

study. Another important system related to the system of Mood but beyond the

scope of this study is that of Key (see Halliday 1967c; 1985a: 281 ff.; Matthies-

sen & Halliday (forthcoming)). The term “key”,1 which has been borrowed from

music terminology, refers to the meanings associated with melodic contours in

the tone group. Earlier (footnote 1, page 171) it was suggested that an imperative

clause that is realized as a tone group characterized by high initial pitch is gener-

ally not impolite in Finnish (when directed at a peer); but an imperative can also

be realized by low initial pitch. These falls and rises in the tone group are mean-

ingful, and are significant in the determination of rhetorical function.2

Page 200: systematic functional linguistic

192

The indicative, in turn, is realized by declarative or interrogative. Interroga-

tive clauses are of two main types: polar and content interrogatives. Polar inter-

rogatives ask for information about the polarity of the clause. In standardized

Finnish, a polar interrogative is realized by the interrogative suffix, -ko/-kö,

which is added to the finite verb in clause-initial position.

(41) onko sulla hirvee(n) kiire [Tel1:1]is+3SG+Q you+ADE terrible+GEN busy‘Are you terribly busy?’

(42) ootko hyvin kiireinen [Tel3:1]is+2SG+Q very busy‘Are you terribly busy?’

A typical phonological variant of this in informal (spoken) genres of Finnish is -

ks (sometimes simply -k):

(43) soittaaks muut sun kaverit [SIIIM3b:6]play+3SG+Q others+NOM/PL you+GEN mates/pals+NOM/PL

‘Do your other mates play [the guitar]?’

Another variant of a polar interrogative in informal (spoken) Finnish is a

change in word order so that the finite verb precedes the subject. This is confined

to clauses where the subject is a second person singular pronoun:

(44) käyt sä siellä usein [SIIIM3b:3]go/visit+2SG you(SG)-NOM there+ADE often‘Do you go there often?’

Whether examples like this are indicative of stable variation or whether they are

indicative of language change remains to be seen. While the patterning here is

similar to an English or Swedish polar interrogative, its use is more restricted: it

occurs only in informal genres of (spoken) Finnish in interrogatives where there is

a second person singular pronoun and only in instances where this pronoun is the

Page 201: systematic functional linguistic

193

1 A similar phenomena occurs in a small number of common expressions such as: paljo(n)kello on? (many/much clock is) ‘What’s the time?’. (In written Finnish, paljon would befollowed by the interrogative suffix: Paljonko kello on?). Unlike the examples with secondperson singular pronoun subjects, however, the subject (kello ‘clock’) and finite verb (on ‘is’)are not crucially involved in the forming of the interrogative.

grammatical subject.1 In the following example, there is a second person singular

pronoun (sul(la) [you-ade] ‘at/by you’) in the first clause below, but it is not the

subject. Unlike the previous variant (example 44), there is always an interrogative

suffix attached to the finite verb in examples where the second person pronoun is

not in the nominative:

(45) onks sul uus? [CA3:1]is+3SG+Q you+ADE new-NOM

‘Have you got a new one?’

It is important to note, however, that variation in the way that an interrogative is

realized in Finnish ) either in the examples illustrated above or, even in dialectal

variation, ) is not, in principle, a problem in SF theory. As pointed out in Chapter

2 (p. 23, 53), a language is seen as a system of meanings in SF theory, as a system

of paradigmatic options, not as a monolithic formal structure. Thus, in the gram-

matical description of Finnish, it is assumed that there is a grammatical distinc-

tion between interrogative and declarative. This distinction holds regardless of

whether it is realized in formal written or informal spoken genres of Finnish, as

spoken in Turku, Helsinki or in the far north of Finland.

A content interrogative is a request for information concerning a participant

or circumstance. It is realized by an interrogative pronoun, kuka ‘who’, mikä

‘which, what’, kumpi ‘which (of two)’ and millainen/minkälainen ‘what kind of’

(inflected for number and case, see, for example, Karlsson 1983b: 123-24), in

clause-initial position.

(46) mistä sä sait niitä? [SIIN3c:23]which/what+ELA you(SG)+NOM get+PAS+2SG they+PAR

‘Where did you get them from?’

Page 202: systematic functional linguistic

194

1 Alternatively, this could be translated as a declarative with rising (tone 2) intonation: Shelives in Iisalmi?. It is difficult to say which is the most appropriate; and, needless to say, theoptions available in Finnish cannot be equated with the those available in English (see 2.2.4).

(47) minkälaisia ihmisiä siellä käy? [SIIIM3b:3]which+GEN+kind/sort person+PL+PAR there+ADE go/visit+3SG

‘What kind of people go there?’

Content interrogatives also include adverbials, which are not inflected, for exam-

ple, koska/milloin ‘when, kauan(ko) ‘how long [with reference to time]’, kuinka

‘how’ (for example, kuinka usein/pitkä ‘how often/long’) and paljon(ko) ‘how

much’.

The third type of interrogative is a content check interrogative. In Finnish,

this is realized by the interrogative suffix -ko/-kö (-ks/-k) attached to the word

being checked.

(48) <A> – Iisalmessaks se muuten asuu? [CA2:18]Iisalmi+INE+Q she by the way live+3SG

‘Was it in Iisalmi that she by the way lives ?’1

<B> – mm, kyllä.mm, yes.

(49) <A> – miss on se paperijuttu [CA7:1]what+INE is+3SG the/that paper-thing‘Where’s that paper thing?’

<B> – on tääl paperiis+3SG here paper‘There’s paper here.’

ai nii sitä ohjepaperia,oh yea the/that instruction-paper,‘oh yea the instruction paper (paper with instructions on it),’

÷÷ sitäksä tarkoitit?that+PAR+Q+you(2SG) mean+PAS+2SG

‘Is that what you meant?’

A variant of the content check ko/kö-interrogative is the particle vai, whose

use is mostly confined to spoken genres of Finnish. The difference between these

variants is difficult to ascertain, and any interpretation based on only a limited

Page 203: systematic functional linguistic

195

1 The variant tai ‘or’ typically occurs in declarative clauses. One could answer by sayingkahvi tai tee, ihan sama ‘coffee or tea, it’s all the same to me’.

number of examples may be skewed by the particular tone contour of the utter-

ance. A tentative interpretation is that the vai-interrogative is less direct than the

ko/kö content check. Rather than being a direct content check, it seems to serve as

an offer of something to comment on (cf. Eades 1982):

(50) <A> – hän on asunu kans Espoossa [CA3:24]she has lived+3SG too Espoo+INE

‘she’s lived in Espoo too’

<B> – kerrostalossa vaiapartment-house+INE Q

‘in a unit/flat?’

<A> – ei vaan rivitalossaNEG+3SG but row+house+INE

‘no in a terraced house (row house)’

This interrogative vai is clearly related to the conjunction vai ‘or’, which typi-

cally occurs in interrogative clauses e.g. Otatko kahvia vai teetä? ‘Will you have

coffee or tea?’1 The vai at the end of a turn is like asking someone for a possible

alternative or contradiction: “or?”.

Thus, a characteristic context for vai is following the finite verb, where it

functions as what might be referred to as a forward channel, i.e. a back channel

that gives the turn back to the first speaker.

(51) <A> – mä olin Toron tykö tiistaiI+NOM be+PAS+1SG Toro+GEN place Tuesday‘I was at Toro’s place Tuesday.’

<B> – olit vaibe+PAS+2SG Q

‘you were?’

<A> – joo [TP1:11]‘yeah’

Page 204: systematic functional linguistic

196

1 See Appendix 6 for system network conventions. This network ) as with any network in SFtheory ) indicates the options that are available. It does not imply that the options are equi-probable. I have included the check question although this option is typically available onlyin a non-initial move in a conversation.

The grammatical choices at a primary level of delicacy in Finnish can be

summarized in the following simplified network:1

The X in the network refers to the item being checked or the item that is picked

up and offered back as something to be commented on. Particularly in the case of

vai, the notion of a check interrogative should not be taken too literally. As seen

in the examples above, it is a means of taking the conversation forward. This is

also true of the ko/kö-check: e.g. niinkö [so + kö] ‘really?’. Nevertheless, it is

clearly an interrogative in the sense outlined above: it is oriented to the exchange

of language and the speaker’s role is that of “demander”.

Figure 5-8: Mood Options in Finnish

Page 205: systematic functional linguistic

197

5.4. Interactional Functions in Finnish

5.4.1. Finite, Mood Marker, and Residue

As pointed out in 2.4.5 (p. 51), interpersonal structure is generally characterized

as either prosodic or scopal. McGregor (1988) suggests, that an element with a

scopal relation holds the entire clause in its scope, i.e. the scope is clause-internal

(or extended and restricted to the boundaries of the clause). According to

McGregor, the scopal element can be compared to a logical operator like | (|(P)

‘it is asserted that P’). McGregor’s approach is reminiscent of Searle’s (1969:31)

analysis of illocutionary acts and illocutionary force indicating devices:

The general forms of (very many kinds of) illocutionary acts is F(p), where the variable “F”takes illocutionary force indicating devices as values and “p” takes expressions for proposi-tions. We can then symbolize different kinds of illocutionary acts in the form, e.g.,

| (p) for assertions ! (p) for requestsPr (p) for promises W (p) for warnings ? (p) for yes-no questions

And so on.

There are, however, important differences. Searle is a philosopher, not a gram-

marian: he is taking a logico-philosophical perspective on what would be consid-

ered to be rhetorical functions in SF theory (see 5.2, p. 168). This is apparent in

his discussion of promises, particularly when he states that they are not realized

grammatically and need not necessarily be accompanied by an explicit illocution-

ary force indicating devices (such as the projecting clause I promise) (Searle

1969:31,64,68).

A declarative could be considered analogous to Searle’s |(p); Searle’s excla-

mation mark in the formulation !(p) could be appropriated for imperatives and

question mark for interrogatives ?(p). However, this logical or quasi-logical anal-

ogy cannot be pushed too far. With content interrogatives, for example, the scope

of the interrogative ) if the notion of scope is used in its logical sense (see, e.g.,

Allwood, Andersson & Dahl (1980: 57,78)) ) is not the entire clause, but part of

it. From the point of view that is presented in this chapter, what is common to all

interrogative clauses, at least when used congruently, is that the clause as a whole

Page 206: systematic functional linguistic

198

1 An anomaly acknowledged by Halliday in his analysis is the fact that an Attribute in arelational clause is regarded as the Complement in spite of the fact that it does not have thispotential. While the conflation of Attribute and Theme may be quite marked in English, tothe extent that it is possible, it is not picked up in a tag and it does not agree with the finiteverb in number and person: Happy/A poet am I, (am I not)? Here the pronoun I remains theSubject.

is oriented to the seeking of information, and it is this interactional and interper-

sonal orientation rather than a strict logical notion of scope that is significant in

the analysis of the clause as an exchange.

As illustrated by the discussion in this chapter, the finite verb and/or what

could be termed a “mood marker” ) an imperative ending, an interrogative suffix,

particle or pronoun ) holds the rest of the clause in its “scope”. Thus, it is the

Finite and/or a Mood Marker that “carries the burden of the clause as an interac-

tive event” (Halliday 1985a: 77). The rest of the clause can be referred to as the

Residue.

(52) onko sulla hirvee(n) kiire [Tel1:1]is+3SG+Q you+ADE terrible+GEN busyFinite+Mood Marker Residue ------------------------->‘Are you terribly busy?’

The residue is described by Halliday (1985a: 74) as what is left over, “the remain-

der of the clause”. In his analysis, Halliday (1985a: 78) nevertheless analyses the

Residue in terms of the “functional elements”, Predicator, Complement and Ad-

junct. Why these elements should be given functional status in a mood analysis is

unclear. A possible response for the analysis of Complement and Adjunct is that

they are defined in terms of their potential to become Subject in English: a Com-

plement has the potential of being Subject, an Adjunct does not have this poten-

tial (Halliday 1985: 79).1 However, it would seem to me that the basis for this

potential is textual rather than interactional. At any rate, the argument would not

apply to Finnish, in which the subject is not an interactional function and word-

order is extremely flexible, mostly conditioned by textual factors (see Vilkuna

1989 and Chapter 7).

Page 207: systematic functional linguistic

199

According to Halliday (1985a: 75), the Finite relates what is being said to

the context of the interactive event, i.e. ties it to the “here and now”. As its name

implies, the Finite has the function of making the proposition finite. That is to

say, it circumscribes it; it brings it down to earth, so that it is something that can

be argued about. A good way to make something arguable is to give it a point of

reference in the here and now; and this is what the finite does. It relates the propo-

sition to its context in the speech event. Halliday (1985a: 75) sees the

contextualization of the Finite in English is in terms of 1) primary tense and 2)

modality. Here the term “modality” also subsumes polarity: both are regarded by

Halliday as interpersonal. Expressions of modality and modulation are seen as

realizing meanings intermediate between positive and negative polarity (see

Halliday 1985a: 75,86,335); and polarity is interpersonal in that it represents

whether or not the speaker denies or affirms a proposition or proposal (Matthies-

sen 1988: 161).

There is a problem, however, with assuming that modality and modulation

(as opposed to polarity) are subsumed under the interactive function Finite, i.e.

that they are involved in relating the proposition to its context in the speech

event. As in English, modality in Finnish ) the speaker’s assessment of the proba-

bilities or the obligations involved in what she or he is saying (or, in a question, a

request for the listener’s opinion) ) can be realized in the finite verb, but it can

also be realized by a modal adjunct. This creates a problem: while on the one

hand, finiteness is said to be realized by a temporal or modal verbal operator

which, combined with the specification of polarity, constitutes the verbal compo-

nent in the mood element (Halliday 1985a: 75-76), finiteness can also be ex-

pressed in a range of modal adjuncts, which need not necessarily be tied to the

verbal component, e.g. to my mind, as expected (see Halliday 1985a: 50).

The problem seems to stem from the fact that primary tense and modality are

related to the here and now of the speech event in different ways. In the philo-

sophical linguistic tradition, for example, links between language and the speech

situation are covered by deixis, which roughly corresponds to areas of both inter-

Page 208: systematic functional linguistic

200

personal and textual meaning in SF theory. Temporal deixis (realized by primary

tense) and demonstrative and personal deixis have traditionally been included in

deixis, but modality has not (see e.g. Lyons 1977b: 636-37; cf. Fawcett (1980:

29-30), who distinguishes between interactional meanings, on the one hand, and

affective and modal meanings, on the other).

Grammatically, the contextualization of the Finite in Finnish ) the way it

relates the utterance to its spatiotemporal context ) is done in terms of 1) primary

tense, 2) polarity, and 3) personal reference (as illustrated in Chapter 3 the finite

verb in Finnish incorporates a personal ending). Thus, in this study, polarity is

regarded as an essential concomitant of finiteness, but modality is not. While,

following Halliday (1985a: 75,86,335) polarity and modality can both be re-

garded as aspects of interpersonal meaning, there are grammatical reasons for

excluding modality from a role in the contextualization of the Finite.

Firstly, polarity features can be linked more clearly than modality features to

the here and now, to the spatiotemporal context of the utterance. While SF theory

is not concerned with truth from the point of view of truth-conditional semantics

(see 2.3.5 p. 30 ff.), polarity features can be linked to (a commitment on the part

of the speaker) as to whether or not a state of affairs holds. Through polarity fea-

tures a speaker makes a commitment as to the truth, or otherwise, of what she or

he is saying. Thus, for example, if we consider a situation in which someone says

Mari ei ollut kotona ‘Mari was not at home’ (or Mari oli kotona ‘Mari was at

home’), then the speaker is making a commitment to the truth of what (s)he is

saying. If, in fact, the statement were not true, then the speaker could, depending

on the circumstances, be accused of telling a lie or a white lie or giving false evi-

dence in a court of law. If, on the other hand, the speaker had said Mari olisi voi-

nut olla kotona ‘Mari could have been at home’, with the verb in the conditional

(-isi-) for example, then the same sort of commitment is not involved, as re-

flected by the fact that it would hardly count as evidence in a court of law. Thus,

the notion of truth, as it is being used here, is far from the notion of truth as it is

understood in truth-conditional semantics, where truth is seen in terms of a simple

Page 209: systematic functional linguistic

201

one-to-one correspondence between sentences and the way things are in the real

world ) or any possible world (cf. Harris 1987b: 157ff.). Moreover, to regard

polarity as an essential concomitant of finiteness is of particular relevance in

Finnish, where the negative element is not a particle but a verb form which in-

flects for person and number (see Figure 3-5, p. 81).

5.4.2. The Grammatical Subject

As I shall argue in Chapter 6, the grammatical subject in Finnish realizes the ex-

periential macro-role of MEDIUM. The question that arises from an interactional

perspective is whether the grammatical subject also realizes the interactional

function of Subject, i.e. whether MEDIUM and Subject are necessarily conflated in

Finnish.

As discussed in 3.4.3 (p. 99 ff.), there is a significant number of subjectless

clause-types in Finnish. For example:

(53) a. yleensä tanssipaikoille on ikärajoja [TIIIN3c:4]generally dance+place+PL+ALL be+3SG age+limit+PL+PAR

‘Generally dance places have age limits/there are age limits to dance places.’

(54) a. mua nukuttaaI+PAR sleep+3SG

‘I feel/I’m feeling sleepy.’

(55) a. Ulkona sataa.outside+ESS rain+3SG

‘It’s raining outside.’

(56) a. Hänet todettiin kuolleeksi.he/she+ACC proclaim/pronounce+INDE dead+TRA

‘She/he was pronounced dead.’

With all of the subjectless clause-types the option [INTERROGATIVE] is available:

Page 210: systematic functional linguistic

202

(53) b. onko yleensä tanssipaikoille ikärajojabe+3SG+Q generally dance+place+PL+ALL age+limit+PL+PAR

‘Are there generally age limits to dance places/Do dance places generally have age lim-its?’

(54) b. nukuttaako suasleep+3SG+Q you+PAR

‘Are you feeling/Do you feel sleepy?’

(55) b. Sataako ulkona.rain+3SG+Q outside+ESS

‘Is it raining outside?’

(56) b. Todettiinko hänet jo silloin kuolleeksi. proclaim/pronounce+INDE+Q she/he+ACC already then dead+TRA

‘Was she/he pronounced dead (already) at that time?’

The fact that a QUESTION can be realized by a clause without a subject is a clear

indication that the subject in Finnish is not crucially involved in the clause as an

interactive event.

Of relevance to the discussion of the mood organization of the clause is the

fact that the verb can occur alone in Finnish. The following are variants of 54a

and 54b above:

(57) nukuttaasleep+3SG

‘Feeling sleepy. (I’m feeling sleepy.)’

(58) nukuttaakosleep+3SG+Q

‘Feel sleepy? (Are you feeling sleepy?)’

As I shall discuss more fully in 7.4.2, the participant that is not realized in the

linguistic structure is the speaker in a STATEMENT and the addressee in a QUES-

TION. Thus, this participant ) which, for convenience, could be referred to as the

“primary inherent participant (or role)” ) need not be realized if it is the speaker

or addressee.

While one might want to argue that this primary inherent participant is cru-

cially involved in the clause as interactive event, the fact remains that there is no

Page 211: systematic functional linguistic

203

single grammatical realization of this participant: it can be realized (i) as the sub-

ject (i.e. it is in the nominative and agrees with the verb in number and person) or

(ii) as what has traditionally been referred to as the object, as in 54a and 54b

above, where it is in the partitive and the verb is always in the third person singu-

lar, or (iii) in the adessive in a possessive clause, where the verb is always in the

third person singular or (iv) in the elative in what is generally referred to as a

resultative clause. In these resultative clauses, the verb ) once again ) is always

in the third person singular. These can be illustrated with the following examples

from A. Hakulinen (1983: 246), who also argues that the subject is not central in

the grammatical organization of Finnish. Among the evidence cited by Hakulinen

is the fact that a possessive or reflexive suffix is not “controlled” by the subject.

In the terminology used here, a possessive or reflexive suffix is co-referential with

the primary inherent participant (which is not necessarily the subject).

(59) Harri tuntee itsensä. (subject)Harri-NOM know+3SG self+POS/3

‘Harry knows himself.’

(60) Harrilla on vain itsensä. (adessive, verb: 3sg)Harri+ADE be+3SG only self+POS/3

‘Harry has only got himself.’

(61) Harri ei pärjää lastensa kanssa. (subject)Harri-NOM NEG/3SG cope child+PL/GEN+POS/3 with‘Harry can’t cope with his children.’

(62) Harrista ei ole lastensa vartijaksi. (elative,Harri+ELA NEG/3SG be child+PL/GEN+POS/3 keeper+TRA verb: 3sg)‘Harry won’t do as the custodian of his children.’

(63) Harria harmittaa epäonnistumisensa. (partitive,Harri-PAR vex+3SG non+success+POS/3 verb: 3sg)‘Harry is vexed by his failure.’

While this primary inherent participant is important in the grammatical organiza-

tion of Finnish, there is no evidence to suggest that it is particularly relevant in

the realization of mood.

A third factor in the consideration of the modal status of the subject is the

mood tag. Like many other languages, Finnish does not have a mood tag that

picks up the Subject and Finite as in English. Possible tags in Finnish are eikö

Page 212: systematic functional linguistic

204

vaan ~ eikö niin (eiks niin)? [not so?] ‘isn’t that so’, which refer to the whole

proposition and do not pick up the subject:

(64) hän on matemaatikko eikö vaanshe/he-NOM is+3SG mathematician NEG+3SG+Q so‘She/he is a mathematician, isn’t she/he [isn’t that so]?’

The subject here is hän ‘she/he’, which is not picked up in the tag. Furthermore,

only the finite verb need be repeated in a confirmation or a denial:

(65) <A> Meneks tää bussi keskustaan?go+3SG+Q this-NOM bus-NOM centre+ILL

‘Does this bus go to the centre?’

<B> Menee.goes+3SG

‘Yes, it does.’

(66) Sulla on toinen. [TSM (Novel): 316]you+ADE be+3SG other-NOM

‘You’ve got someone else/another (lover).’

– Ei ole.NEG+3SG be‘No I haven’t.’

– Varmasti on.surely be+3SG

‘You must have.’

On the other hand, the subjectless clauses exemplified earlier (53 ) 56), do

not occur in the imperative, or at least not in the 2nd person imperative: *nukuta!

‘feel sleepy!’. This may be seen as an indication that the grammatical subject is in

some way tied in with the interactional options in the clause. However, as argued

above, the imperative mood involves orientation to action, and an important fea-

ture of a proposal (in Halliday’s terms a command or an offer) is a participant that

makes the action or process possible, i.e. the Medium. This would again indicate,

as will be argued in the next chapter, that the subject in Finnish realizes an experi-

ential function.

Page 213: systematic functional linguistic

205

Chapter 6

Experiential Structures in the Finnish Clause

6.1. Preliminary Considerations

6.1.1. General Remarks

This chapter focuses on experiential structures in the Finnish clause, i.e. those

meanings that are concerned with the way in which language serves as a model of

reality, with the way in which it is used to represent and construct the world in

which we live. Since the description presented is based on SF theory, reference is

made, in particular, to the writings of Halliday. Halliday’s (1985a) description of

English makes no claims about the experiential organization of Finnish, and, thus,

in referring to Halliday, I am taking those aspects of his description of English

that appear to be relevant in the application of SF theory to Finnish.

Throughout this chapter, reference is also made to Hakulinen & Karlsson’s

(1979: Ch. 6, 301 ff.) analysis of basic and marginal clause types in Finnish.

Hakulinen & Karlsson’s analysis is based on Lyons’ (1977: 469ff.) analysis of

“kernel-sentences”, which distinguishes intransitive, transitive, equative, ascrip-

tive, locative and possessive sentences. The description presented here and

Hakulinen & Karlsson’s analysis are done for different purposes and within dif-

ferent theoretical frameworks, and, thus, they could be regarded as incommensu-

rable. I refer to Hakulinen & Karlsson’s analysis for two reasons. Firstly, their

analysis is the only general description that has been done of clause types in Finn-

ish. Secondly, their typology of clause types is widely known in Finland, and thus

I can engage in a dialogue with Finnish linguists by relating my own description

to an analysis that is familiar to a Finnish linguist and focus on what I see as the

advantages of a theory-based grammar that is metafunctionally organized. How-

ever, as is obvious from the discussion, I rely heavily on the insights of Hakulinen

& Karlsson, and their insights, in turn, rely on the insights of other Finnish gram-

marians.

Page 214: systematic functional linguistic

206

As pointed out in 2.4.7 (p. 54 ff.) and 2.4.11 (p. 63 ff.), while grammatical

analysis in SF theory is based on meaning, these meanings must be related to

distinctions that are reflected in some way in the organization of the clause. Thus,

the analysis addresses two questions simultaneously:

1) What grammatical meanings are realized in a particular clause type?

2) How are these meanings construed in the organization of the clause?

In this respect, systemic-functional grammar differs from other grammars, for

example, case-role analysis as developed by Fillmore (1968) and Anderson (1971,

1977). While Fillmore originally based his case analysis on grammatical distinc-

tions in English, the cases that he postulated are generally assumed to be univer-

sally applicable. Hakulinen & Karlsson (1979: 101 ff.) have applied these case-

roles to Finnish. In their analysis, for example, the following underlined nominals

are assigned different roles, since an agentive is animate whereas an instrument is

inanimate:

(1) Mies joi. AGENTIVE

man-NOM drink+PAS/3SG

‘The man drank.’

(2) Traktori perkasi ojan. INSTRUMENTAL

tractor-NOM clear/dig+PAS/3SG ditch+GEN

‘A/The tractor dug the ditch.’

On the other hand, if traktori ‘tractor’ in this last example is replaced by mies

‘a/the man’, then the subject would undoubtedly be regarded as an Agentive:

(3) Mies perkasi ojan. AGENTIVE

man-NOM clear/dig+PAS/3SG ditch+GEN

‘A/The man dug the ditch.’

The grammatical pattern in clause 2 and 3, however, is the same:

X verb Y+GEN.

In the approach adopted in this study, the underlined NP in each of these clauses

realizes an Agent or Actor and the clause-final NP in 2 and 3 realizes the Goal.

The clauses differ in meaning, of course, but this because of different lexical

choices.

Page 215: systematic functional linguistic

207

1 The inverted commas are Laitinen’s: it is clear that she is using the term informally.

Thus, the meanings that are being focused on in this chapter ) and in this

study ) are those that are construed in the clause. Meaning-making is not, of

course, restricted to lexical items or clauses. There are meanings, such as censure

or irony, for example, that cannot be related to specific lexicogrammatical fea-

tures. Following Lemke (1988, 1989, 1990) I make a methodological distinction

between grammatical meaning and meanings that are made in a text. While gram-

mar and text are seen to be in a dialectic and symbiotic relationship in SF theory

and text semantics is regarded as subsuming grammatical meaning, in order to

have some principled way of analysing the way in which we make meaning, we

need to be able to separate out those meanings that are realized

lexicogrammatically.

To illustrate what I mean by this distinction, I shall briefly discuss what is

sometimes informally referred to as a “censorious passive” in Finnish (e.g. Laiti-

nen 1988). Laitinen uses this term in an analysis of letters to the editor in Helsin-

gin Sanomat (the leading newspaper in Finland); the letters were part of a debate

about the role of working mothers vs. mothers who are not employed outside the

home. Laitinen (1988: 164-165) does not make a distinction between grammatical

meaning and text semantics, and thus the status of what she refers to a “censori-

ous passive” (“moittimispassiivi”)1 is unclear in her analysis. As Laitinen herself

is no doubt aware, there is no form or clause type in Finnish that can be referred

to as a “censorious passive”. In the following example from Laitinen, this “cen-

sorious passive” (glossed INDE) is in cursive script:

(4) Nyt ei kotiäitiys ole muodissa. now NEG+3SG home+motherhood-NOM be+PTC fashion+INE

‘It’s no longer fashionable for mothers to stay at home.’

Nyt luodaan uraa, toteutetaan itseään janow create+INDE career+PAR fulfil+INDE self+PAR+3POS &

tasa-arvoillaan.equal+status ~ opportunity +FREQUENTATIVE VERBAL AFFIX + INDE

‘These days you forge a career, fulfil yourself and you equalize your opportunities.’

Page 216: systematic functional linguistic

208

The censure that can be read (or read into) this example is not a grammatical fea-

ture of Finnish: taken by itself nyt luodaan uraa, for example, is not necessarily

censorious. While, as Laitinen points out, the frequentative verb tasa-arvoilla

‘(to) equalize your opportunities’ is not established in usage in Finnish and has a

slightly mocking effect in this context, it is not the frequentative or the indefinite

(passive) per se, but a combination of features that gives this bit of text a mock-

ing effect. (This can be compared to the indefinite use of “they” in English, which

is sometimes used in contexts expressing disapproval e.g. That’s what happened

when they let buses go up St James’s Street [Graham Green (1978), The Human

Factor, p. 34].)

Censure is not a grammatical phenomenon: its interpretation can depend on

a wide range of factors, both intratextual features (such as the combination of

indefinite, present tense, frequentative, the repetition of the clause-initial

continuative-like nyt ‘now’ etc.) and factors such as who is reading the text and

whom the reader is siding with in the debate. The following example of the Finn-

ish indefinite is taken from the first paragraph of the actual text of Laitinen’s arti-

cle:

(5) Puhuttiin naisten rooleista ... [Laitinen 1988: 159]speak+INDE woman+PL+GEN role+PL+ELA

‘The discussion was concerned with the roles of women ..’

It would be difficult to detect any note of censure in the paragraph in which this

indefinite occurred, i.e. as part of an academic text in a collection of articles pub-

lished by a group interested in women’s studies and feminist issues in Finland. If

this clause had occurred in a text written, for example, by a man who was known

to be a misogynist and anti-feminist, it would be read in a different way, it could

even be interpreted as being censorious.

This study is concerned with grammar and grammatical meaning. These

form the basis for the exchange and the negotiation of meanings when people

engage in the process of semiosis. Without this basis, without a shared grammati-

cal system, it would be extremely difficult to exchange meanings and negotiate

about them.

Page 217: systematic functional linguistic

209

Experiential structures at the rank of clause are generally referred to as tran-

sitivity structures in SF theory. The use of the term “transitivity” in this sense is

an extension of its traditional meaning. Traditionally, verbs are divided into tran-

sitive and intransitive according to whether an action or a process is seen as 1)

being limited to one participant or 2) extending from one participant to another.

In most modern linguistic theories, transitivity is seen as a feature of the clause.

Moreover, the notion of extension on which it is based is essentially an experien-

tial semantic notion (i.e. it relates features of the clause with features of the

extralinguistic world). Thus, the term transitivity when applied to the clause in SF

theory is used to refer to the experiential semantic structure of the clause in its

entirety (Halliday in Kress ed. 1976: 159). Consequently, in SF theory, the transi-

tivity structure of a clause such as “The lion chased the tourist” refers not to the

notion of the extension of a process from one participant to another, but to the

configuration of experiential semantic functions that are realized in the clause: for

example, Actor, Process, and Goal.

Thus, while transitivity in SF theory can be seen as being somewhat similar

to the case role analysis of a Fillmorean case grammar, as well as to the semantic

level of Dik’s (1978, 1980, 1987) functional grammar, and to the semantic struc-

ture analysis of Foley and van Valin’s (1984: 27-74) functional grammar, as indi-

cated above, there are important differences. SF theory is not concerned with es-

tablishing a universal experiential semantic characterization of the clause at the

specific level of case roles. Any grammatical category that is established in the SF

description of a language is, of necessity, language-specific, since it is an abstrac-

tion based on the interrelations and oppositions found in the grammatical organi-

zation of the language being described, and the grammatical organization of no

two languages can be regarded as equivalent: the sorts of grammatical distinctions

and oppositions that are made in any language are unique to that language (Hasan

1971). While a transitivity function such as Actor, for example, may typically

refer to the “doer of an action” in a number of languages, functions are not

referentially but linguistically defined in terms of grammatical contrasts and

oppositions. A label such as Actor can be best thought of as a mnemonic device

capturing a grammatical generalization; the label is often an attempt to cap-

Page 218: systematic functional linguistic

210

ture the meaning of central or concrete instances, but the meaning of the category

as a whole is ineffable (see Chapter 2).

Moreover, as discussed in 2.4.8 (p. 56 ff. and elsewhere in Chapter 2), while

grammatical description in SF theory is organized on the basis of meaning, expe-

riential meaning is only one type of meaning realized in the clause: a number of

different types of meaning are simultaneously realized and conflated in the

clause. In contrast with these other grammars, SF theory does not restrict seman-

tics to experiential semantics.

The aim of this chapter is to analyse the way in which experiential meaning

is realized at the rank of clause. It will be concerned with grammatically and se-

mantically motivating basic clause types in Finnish ) clause types that are de-

fined experientially. Only basic options will be dealt with; the analysis needs to

be extended in delicacy for each clause type. I shall, however, analyse one partic-

ular process type ) relational intensive processes ) at a somewhat greater degree

of delicacy than the other process types in order to illustrate how the analysis

proceeds.

6.1.2. Processes

The term “process” is used in two senses in SF theory and throughout this study :

1) in its narrow sense, it refers to the function that is typically realized by a verb

in a language and 2), in its wider sense, it can be seen as the SF equivalent of the

term “state of affairs”, if a state of affairs is understood as being a linguistic con-

struct and not a logical notion. A process in this wider sense (i.e. a linguistically

construed state of affairs) could also be referred to as a “representation”.

A process ) as a representation or a linguistically construed state of affairs )

consists potentially of three components: 1) the process itself, 2) the inherent or

core roles involved in the process, i.e. what Halliday (e.g. 1985a) refers to as the

participants, and 3) the non-inherent or peripheral roles associated with the pro-

Page 219: systematic functional linguistic

211

1 Halliday’s use of the term “participant” in this quotation corresponds to the what I amreferring to as a “role”. He says, for example, that you are required by the regulations to paythe full fee to the examining board “involves four participants”.

cess, i.e. what Halliday refers to as the circumstances. These experiential seman-

tic functions provide a basic framework for organizing and interpreting the world

of our experience, whether real or imagined. They can be subclassified into more

specific functions such as Actor, Goal, Source etc.

6.1.3. Inherent and Non-Inherent Roles

A broad distinction is generally made between inherent and non-inherent experi-

ential functions (other than the Process) or between participants and circum-

stances in SF grammars. A similar distinction is made in other grammars: for ex-

ample, dependency grammar distinguishes between actants and circumstances

(Tarvainen 1977: 21), Foley & Van Valin (1984: 77) refer to core and peripheral

arguments and Dik (e.g. 1978, 1987) to nuclear and satellite arguments. Within

SF theory, McGregor (1990a), for example, distinguishes between inherent and

non-inherent roles in his analysis of Gooniyandi, and then makes further subdivi-

sions within non-inherent roles. He defines inherent roles as those that are inher-

ent in the structure of the clause: they define minimal clauses (McGregor 1990a:

329), or, perhaps it would be more to the point to say that inherent roles and mini-

mal clause types are mutually defining since the minimal clauses have been clas-

sified on the basis of inherent roles. In other words, inherent roles are those that

can be used to characterize a particular process type. Similarly, Fawcett (1980:

155) distinguishes between inherent and circumstantial roles.

Thus, what is essential in the analysis of experiential structures in the clause

is the determination of the experiential functions that can be used to classify and

define process types. As Halliday (in Kress (ed.) 1976: 159) puts it: “in most

types of process it is possible to bring in participants1 other than those that are

essential to the process; and conversely one of those that are essential to it may

not actually be expressed in the structure”. This means that we need to determine

Page 220: systematic functional linguistic

212

a minimal configuration of functions associated with each process type, the pro-

cess itself and the other inherent or core functions associated with it. As E. Itko-

nen (1990: 354) points out (see 2.3.4), this can only be done on the basis of our

linguistic intuitions, i.e. on the basis our knowledge of what counts as a minimal

- yet complete - structure, since minimal configurations are unlikely to occur fre-

quently in actual texts. A non-inherent or peripheral function can be determined

on the basis of a deletion test: if it is omitted, a complete structure remains.

I shall follow McGregor’s (1990a) terminology and refer to inherent and

non-inherent roles in Finnish. The term experiential function will be used as a

cover term for processes and roles. As pointed out in 4.3.2 (pp. 112-114), an in-

herent role in Finnish can be realized as a bound morpheme. For example:

(6) Ol + i + n Szegedissä.be + PAS + 1SG Szeged+INE

‘I was in Szeged.’

(7) Käv + i + t + kö Szegedissä.go/visit + PAS + 2SG + Q Szeged+INE

‘Did you visit Szeged?’

The fact that an experiential function can be realized as a morpheme that is part of

the verb is an indicator par excellence of its status as inherent role. Its inherent-

ness is evidenced by the fact that it realized by a morpheme that is actually part of

the finite verb. The representation of a first or second person participant as a sepa-

rate pronoun is textually conditioned in Finnish. Matthiessen & Halliday (forth-

coming, section 3.3.4) suggest that in instances like this, the pronoun has only

textual value and does not have a function in the transitivity structure of the

clause. In the glosses in this chapter, where there is both a 1st or 2nd person pro-

noun and a 1st or 2nd person morpheme (“personal ending”) in the linguistic

structure this is indicated by a subscript: e.g. mä lähde+n ‘I’m leaving’ would be

glossed as: Actori Process+Actori.

Page 221: systematic functional linguistic

213

6.2. Process Types in Finnish: An Overview

The major process types in Finnish are: relational, material, and mental processes

(following Matthiessen’s (1989) description of English, I shall use mental as a

cover term for mental, perceptive, reactive, verbal etc.). Once again, it needs to be

stressed that while I use roughly the same labels that are used in SF descriptions

of English, this does not mean that Finnish and English categories are equivalent

(see 6.1 and Chapter 2). The macrofunctions that I postulate for Finnish, Medium

and Domain, will be discussed in section 6.7.

In line with the theory of prototypes (see 2.4.12, p. 65), these major process

types are not seen as discrete and absolute categories, but rather as overlapping

categories:

RELATIONAL

MATERIAL

MENTAL

Within each process type there are a number of subtypes, which can be grammati-

cally and semantically linked with each other. These subtypes are also analysed in

terms of being more or less representative of the subtype.

Figure 6-1: Major Process Types

Page 222: systematic functional linguistic

214

1 The use of these English translations is simply a convenience. It is not meant to imply thatthere is a universal set of meanings.

6.3. Relational Processes

Relational processes in Finnish are either intensive (X is (like) Y)1 or circumstan-

tial (at X is Y, Y is at X). Within intensive processes a further distinction can gen-

erally be made between identifying and attributive processes. A peculiarity of

Finnish, when compared to English as least, are ambient processes (at X is Y-like

~ like Y). These are distinct from circumstantial processes, which are concerned

with the location of something in space or time. An important subtype of circum-

stantial process in Finnish is a possessive process (at Y is X = ‘Y is in the posses-

sion of X’).

6.3.1. Intensive Relational Processes

Intensive processes are realized by a clause which encodes a relationship of same-

ness or similarity between two constituents. Halliday (1985a:xx ff.) makes a fur-

ther distinction in his description of English between attributive and identifying

intensive processes. If two constituents in a clause are designated “X” and “Y”,

then in the attributive mode, Y functions as a qualitative attribute of X or X is a

member of the class Y; whereas in the identifying mode, Y serves to define the

identity of X, i.e. a relation of identity is set up between X and Y. (See Appendix

4 for glosses.)

(8) a. Sarah is wise. ATTRIBUTIVECar Pro:int Att

b. Tim is a dancer.Car Pro:int Attribute

(9) Steve is the tall one. IDENTIFYINGId/Tk Pro:int Ir/Vl

Halliday (1985a: 124) points out that the distinction between attribution and

identification is not clear-cut in English. This distinction, however, is grammati-

Page 223: systematic functional linguistic

215

cally realized by the fact that if the positions of the nominal constituents in an

identifying process are reversed (e.g. The tall one is Steve), the functions are also

reversed: the tall one is the Identified, i.e. the one to be identified. Whereas in an

attributive process in English, reversing the positions of the nominal constituents

is generally highly marked, but when it does occur (e.g. in proverbs) the functions

are not reversed:

(10) Faithful are the wounds of a friend;Att Pro:int Car

but the kisses of an enemy are deceitful. (From The Bible.)Car Pro: int Att

The distinction between attributive and identifying that is made in SF gram-

mar is somewhat different to a distinction that has been made in descriptions of

Finnish between attributive and equative clauses (e.g. Hakulinen & Karlsson

1979: 94-95), which is based on a truth-conditional approach to semantics: an

equative clause is one in which two NPs have an identical referent in the logical

sense (see 2.3.5 p. 30 ff.). From an SF perspective, on the other hand, language is

used to reflect or create a relationship of similarity in the attributive mode and a

relationship of equality, identity or sameness in the identifying mode. As illus-

trated below, this alternative approach has repercussions for what counts as an

attributive or identifying clause.

6.3.1.(i) Attributive Intensive Processes

The most common type of intensive process in Finnish is the attributive. Attribu-

tive intensive processes can generally be analysed as a configuration of the fol-

lowing functions:

Carrier @ Process:intensive @ Attribute

The raised stop ( @ ) indicates that the functions are not necessarily realized in this

order, since word order in Finnish is textually conditioned. However, to the extent

to which word-order variation is possible, as in English, what is significant about

Page 224: systematic functional linguistic

216

attributive processes in Finnish is that the experiential functions are unaffected by

changes in word order (cf. identifying processes below).

The following are examples of intensive attributive processes in Finnish.

The abbreviation (NR) stands for a non-inherent role. A discontinuous constituent

(e.g. on ollut ‘has been’ in on÷ viime aikoina ²ollut ‘has in recent times been’)

is marked with an arrow.

(11) AkiraKurosawa -- on japanilainen elokuvaohjaaja. [W5:3374]AkiraKurosawa+NOM be+3SG Japanese movie+director-NOM

Car Pro:int Att‘Akira Kurosawa is a Japanese motion picture director.’

(12) [mul ois ens viikon torstaihin aikaa kirjoittaa se puhtaaksi, ] [Tel2:13]‘I’ve got until next Thursday to type up a final copy’

joka on mulle helvetin hidasta hommaawhich-NOM be+3SG I+ALL hell+GEN slow+PAR work+PAR

Car Pro:int (NR) Att‘which is hellishly/damned slow work for me’

(13) ootko hyvin kiireinen [Tel3:1]is+2SG+Q very busy-NOM

Pro:int+CarAtt‘Are you terribly busy?’

(14) Aurinkoenergian tutkimus on÷ viime aikoinasun+energy+GEN research-NOM be+3SG÷ recent time+PL+ESS

Car Pro:int÷ (NR)

²ollut erittäin vilkasta. [T3/83:14]²be+PTC extremely vigorous+PAR

²Pro Att

‘Research on solar energy has been extremely vigorous in recent times.’

As illustrated in these examples, an attributive process in Finnish is typically real-

ized by the verb olla ‘(to) be’ (but see section 6.3.1.(iii) below). The Attribute is

realized by an NP, with either a noun (11 ) 12) or an adjective (13 ) 14) as Head.

The Head of the NP realizing the Attribute can either be in the nominative or in

the partitive, and this is true whether it is a noun or adjective. The choice between

nominative and partitive is basically dependent on boundedness, some details of

which are presented in section 3.4.2 (pp. 94 ) 99). The Carrier is realized by an

NP in the nominative (11 ) 14 or by a bound morpheme (example 13).

Page 225: systematic functional linguistic

217

1 The former is the spoken, the latter the written Finnish variant.

Another important grammatical feature of intensive attributive processes in

Finnish is the fact that the NPs realizing the Attribute and Carrier agree with each

other in number. In examples 11 ) 14 above, the Carrier is realized by an NP in

the singular or by a bound morpheme that is singular. In agnate clauses in which

the Carrier is plural, the Attribute would also be plural:

(15) He ovat japanilaisia elokuvaohjaajia. [cf. 11 above]they-NOM/PL be+3PL Japanese+PL+PAR movie+director+PL+PAR

Car Pro:int Att‘They are Japanese motion picture directors.’

(16) ootteko ~ oletteko1 hyvin kiireisiä [cf. 13 above]is+2PL+Q very busy+PL+PAR

Pro:int+Car Att‘Are you [plural] terribly busy?’

This agreement is consistent in both written and spoken Finnish. Agreement be-

tween the NP realizing the Carrier and the finite verb is more complex. The NP

agrees with the verb in person, i.e. it is what I have referred to as the grammatical

subject in Chapter 3. However, while in standardized Finnish, the Carrier NP

would also agree with the verb in number, this is not the case in spoken Finnish

(see 3.3.2 p. 78). The spoken Finnish variant of 15 above is as follows:

(17) ne on japanilaisia elokuvaohjaajia.they-NOM/PL be+3SG Japanese+PL+PAR movie+director+PL+PAR

Car Pro:int Att‘They’re Japanese motion picture directors.’

As indicated in 17, however, number agreement between Carrier and Attribute is

consistent even in spoken Finnish. This number agreement between the NPs in an

intensive process is an important feature of Finnish intensive processes in general,

and it will be mentioned again in the discussion of other subtypes of intensive

process in this section.

Page 226: systematic functional linguistic

218

1 See 4.5.5 (p. 149 ff.) for a discussion of embedding.

As pointed out above, the experiential functions in an attributive process (as

opposed to an identifying process) are unaffected by changes in word order.

Of course, claims about the possibility of changing word order have to be quali-

fied when examples are taken from a text, since any change in word order has an

effect on textual meaning. Moreover, with actual text examples there are more

obvious constraints, e.g., it is highly unlikely that an anaphoric pronoun could be

moved to the end of a clause and a clause embedded1 in the structure of an Attrib-

ute would be unlikely to be found at the beginning of a clause. A common varia-

tion of the word order in an attributive clause is one in which the Attribute pre-

cedes the Carrier (Attribute ^ Carrier ^ Process): example 18b, thus, is a likely

variant of example 18a, an example taken from a text:

(18) a. mähän oisin -- pelkkä nukke sielä [Tel2:9]I-NOM+TIS be+CON+1SG mere doll-NOM thereCari Pro:int+Cari Att (NR)‘I’d be nothing but a doll there.’

(18) b. pelkkä nukke mä oisin sielämere doll I would be thereAtt Cari Pro:int+Cari (NR)‘A mere doll is what I’d be there.’

In this particular instance, actual reversal of the positions of the inherent roles is

unlikely because the pronoun mä ‘I’ is unstressed, and, moreover, since it is a

deictic element, it is typically Given, and thus is most likely to occur at the begin-

ning of a tone group (see 7.2.2). If the pronoun occurs after the verb, i.e. pelkkä

nukke olisin minä siellä ‘A mere doll would I be there’, the resulting clause

sounds biblical in Finnish. Reversal is more likely with other NPs:

(19) Varmaa on vain muutos. [TT6/91: 1]certain+PAR be+3SG only change-NOM

Att Pro:int Car‘The only thing that is certain is change.’

Page 227: systematic functional linguistic

219

(20) Tapa! Kaunis on sota. [From a poster]kill! beautiful-NOM be+3SG war-NOM

Att Pro:int Car‘Kill! War is beautiful.’

However, the point that is being made here is that regardless of word order, the

functions of the Attribute and Carrier do not change.

The use of this criterion ) i.e. the effect of a change in word order on the

experiential functions ) may be questionable to some grammarians, who may ask

“What kind of evidence is this and what evidence is there that the roles have or

have not changed?” In a functional grammar, however, one does not pretend that

one can make judgments about grammaticality that are independent of meaning.

This criterion is no less or no more questionable than a traditional grammarian

saying that häntä ‘(s)he+PAR’ is the object in both of the following clauses, which

vary only in word order:

(21) Minä rakastan häntä.I-NOM love+ISG (s)he+PAR

‘I love her/him.’

(22) Häntä minä rakastan.(s)he+PAR I-NOM love+ISG

‘It’s her/him I love.’

In SF terms, a speaker of Finnish knows that häntä ‘(s)he+PAR’ has the same ex-

periential function because she or he has an understanding of both clauses.

6.3.1.(ii) Identifying Intensive Processes

Identifying intensive processes, on the other hand, are distinguished by the rever-

sal test mentioned above: if the positions of the nominal constituents in an inten-

sive process are reversed, the functions are also reversed. The following text is

taken from a review of a play by the eighteenth-century French playwright Pierre

de Marivaux, who ) it can be assumed ) would be unknown to many Finnish

theatre-goers. After a long introduction to Marivaux and his work, the reviewer

Page 228: systematic functional linguistic

220

compares Marivaux’s plays to Mozart’s music and then introduces the leading

female role using an identifying process:

(23) [Harhatunteissa ne aariat keskeisesti laulaa vasta leskeksi jäänyt nuori kaunis Markiisitar...‘In La Seconde Surprise de l’Amour these arias are centrally sung by the young, beauti-ful marchioness, who has just become a widow ....’]

Kansallisteatterin Markiisitar on Eeva-Kaarinanational+theatre+GEN marchioness-NOM be+3SG Eeva-Kaarina-NOM

Volanen,Volanen-NOM

‘The National Theatre’s marchioness is Eeva-Kaarina Volanen,’

[meikäläisittäin suvereeni komedienne, jonka häikäisevä henkilökohtainen panos tekeeHarhatunteista esityksellisen juhlan.] [HKV: 57]

‘by our standards a sovereign/outstanding comedienne, whose dazzling personal input makesLa Seconde Surprise de l’Amour a performance festival.’

In this example, the two NPs realize functions that are equated: the first is lexi-

cally realized by the character in the play, Kansallisteatterin Markiisitar ‘the

National Theatre’s marchioness’, the second by the proper noun that refers to the

actress Eeva-Kaarina Volanen.

One of the sets of inherent roles that Halliday (1985a: 115 ff.) postulates for

intensive processes in English is Identified and Identifier. According to Halliday

(1985a: 118), as a general rule, the Identified in the experiential structure of the

clause conflates with the Given element in the information structure and the Iden-

tifier conflates with the New element. I assume that this is also true for Finnish.

Hence, in the example above, Kansallisteatterin Markiisitar ‘the National Thea-

tre’s marchioness’ realizes the Given element in the context above, i.e. it is the

information that is presented by the writer as recoverable, it is not news (Halliday

1985a: 277; see also Chapter 7). It is also the Identified, i.e. the element to be

identified. The New element is the element that is presented as non-recoverable:

it is the newsworthy bit of the clause. In the clause above, it is Eeva-Kaarina

Volanen. This element also realizes the Identifier.

Page 229: systematic functional linguistic

221

1 A problem that ensues from this discussion is whether Identified and Identifier areredundant functions that can be replaced by Given and New (cf. Halliday (1985a: 118), whomaintains that this is not the case in English). In other words, can we simply say that in aFinnish identifying clause Given and New are equated? Since Given and New are assumedto be prosodically realized (see Chapter 7), this question is beyond the scope of this study.

(24) Kansallisteatterin Markiisitar on Eeva-Kaarina Volanen.Identified (Given) Pro:int Identifier (New)‘The National Theatre’s marchioness is Eeva-Kaarina Volanen ...’

Another test for determining which element realizes the Identifier is that if, on the

basis of an identifying clause we formulate an implied content interrogative

(Kuka/mikä on X? ‘who/which is X?’), then the interrogative pronoun (or the

answer to the question) refers to the Identifier. In the example above, the implied

question would be: Kuka Kansallisteatterin Markiisitar on? (Or: Kuka on

Kansallisteatterin Markiisitar?) ‘Who is the National Theatre’s Marchioness?’;

and the response, i.e. the Identifier, would be Eeva-Kaarina Volanen.1

As mentioned above, however, the distinguishing feature of an identifying

process is that if we reverse the positions of the NPs we also reverse the func-

tions. In principle, the following clause is possible:

(25) Eeva-Kaarina Volanen on Kansallisteatterin Markiisitar.Identified (Given) Pro:int Identifier (New)‘Eeva-Kaarina Volanen is the National Theatre’s marchioness.’

In an example like this, Eeva-Kaarina Volanen would be treated as recoverable

by virtue of the fact that she is known to Finnish theatre-goers. In this instance,

she is the Identified, i.e. the one to be identified by her role in the play being re-

viewed. In fact, in the same review, three of the five other actors are identified in

just this way.

Page 230: systematic functional linguistic

222

(26) Marjukka Halttunen on ilakoivan ketterä,Marjukka Halttunen is light-hearted+GEN agile

* Id (G) * Pro:int * Ir (N)

teräväpäinen kamarineitsyt Lisette,sharp-witted lady’s maid Lisette, *

‘Marjukka Halttunen is the light-heartedly agile and sharp-witted lady’s maid Lisette,’

(27) Heikki Nousiainen irtonaisesti hyväntyylinen kamaripalvelija Lubin,Heikki Nousiainen loosely good-humoured valet Lubin,

* Id (G) * Id (N) *

‘Heikki Nousiainen the loosely good-humoured valet Lubin,

(28) Esa Saario arvovaltaisesti yksivakainen,Esa Saario authoritatively grave

* Id (G) * Ir (N) ÷

nenästä vedettäväksi syntynyt Kreivi.nose+ELA pull+INDE+PTC+TRA born Count

*

‘Esa Saario the authoritatively grave born-to-be-duped Count.’ [HKV:57]

The last two examples are elliptical: the finite verb on ‘is’ is retrievable from the

first clause in the complex (26).

Thus, a clause realizing an identifying process sets up an identity relation

between two NPs, and if the positions of the NPs is reversed, the roles are re-

versed. Of course, the category of process being discussed here is called identify-

ing because it so happens that this is the meaning construed by the

lexicogrammatical pattern: mnemonically the label “identifying” serves a pur-

pose. However, in using this label, there is no implication that whenever there is

identification, the linguistic pattern necessarily counts as an identifying process.

As pointed in 2.4.10 (p. 63) and at the beginning of this chapter, the categories in

a functional grammar are not recognized simply on the basis of meaning. In other

words, it is not enough to define an identifying process simply on the basis of

meaning. The response to the question Kuka on Eeva-Kaarina Volanen? ‘Who

is Eeva-Kaarina Volanen?’ could be realized as a material process, e.g. Siinä hän

kävelee Pentti Siimeksen kanssa kadun toisella puolella ‘She’s walking over

there with Pentti Siimes on the other side of the street’. Even attributive processes

Page 231: systematic functional linguistic

223

1 It should be noted that while the approach taken by Kelomäki (1988) is logico-philosophi-cal, one of Kelomäki’s central themes is the inadequacy of a strictly logical approach to theanalysis of language.

can be used to identify someone: if Eeva-Kaarina Volanen were amongst a small

crowd of people, she could be identified by saying Hän on pieni ja tummatukkai-

nen ‘She’s small and dark-haired’.

A clear example of an identifying clause being used to identify is in the fol-

lowing fragment of text. The example is from a telephone call from a person who

was familiar to me but whose name I did not know:

(29) Hei Mirja Niemi täällähi Mirja Niemi here‘Hi this is Mirja Niemi,’

÷ mä olen se, [[ joka kysyi sultaI be+1SG it-NOM which/that-NOM ask+PAS/3SG you+ADE

Ir Pro:rel:int Id

eilen siitä Hallidayn modaalisuudesta ]].yesterday it+ELA Halliday+GEN modality+ELA

‘I’m the one who asked you yesterday about Halliday’s (notion) of modality.’

Here the NP se ‘it/that’ + embedded relative clause serves to identify the caller

mä ‘I’.

Identifying processes also include processes in which the function of one NP

is to define, exemplify or name the other. Thus, identifying processes in this

study include examples like the following, which from a logical perspective is

rejected by Kelomäki (1988: 38)1 as identifying:

Page 232: systematic functional linguistic

224

(30) Eräs ongelmien syntyyn vaikuttanut tekijäcertain-NOM problem+PL+GEN birth+ILL influence+PTC factor-NOM

Id

on se, [[ että suomalaiset laivanvarustajat ovat osoittaneetbe+3SG it/that-NOM [[ that Finnish ship owners have displayedPro:int Ir ÷

“isänmaallisuuttaan” tilaamalla aluksensa ulkomailta]]. [HKV]their “patriotism” by ordering their ships from overseas]].

‘One of the factors (/A certain factor) that has caused problems is the fact that Finnish shipowners have displayed their “patriotism” by ordering their ships from overseas.’

This example is from an article in which the writer first discusses the crisis that is

facing Finnish dockyards. In this particular clause, the writer identifies one of the

factors that has contributed to the crisis. The Identifier (realized by the NP with a

clause embedded in it) could be seen as a response to the implicit question “What

factors have/What has contributed to the birth of these problems?”. In the context

of the article, it is not feasible to reverse the positions of the NPs, but, in princi-

ple, this is possible. In which case the NP that has been translated as ‘the fact that

Finnish ship owners have displayed their “patriotism” by ordering their ships

from overseas’ would be the Identified (Given) and ‘one of the factors that has

caused problems (in the dockyards)’ would be the Identifier. The reason the ex-

ample is rejected by Kelomäki is that the indefinite pronoun eräs ‘a certain’ ex-

plicitly indicates that there are other factors involved. While this is true, it is irrel-

evant from the point of view of identification as it is understood in this study.

What is relevant is that one of the factors is being identified. (See Halliday (1968:

190) for his analysis of a similar type of clause in English: Gladstone would be

an example).

With a first or second person subject, the Identified (Given) may be realized

by a bound morpheme:

(31) Olen Henna Partanen. [PL 1]be+1SG Henna Partanen-NOM

Pro:int + Id Ir‘I’m Henna Partanen.’

Page 233: systematic functional linguistic

225

The morpheme -n (referring to the speaker) realizes the Identified (Given). This is

recoverable because a speaker and an addressee are assumed in an interactive

situation. Examples such as this can be referred to as “naming (identifying)

clauses”.

Thus, in naming clauses, the Identifier NP is lexically realized by a name )

a linguistic tag ) by which the speaker can be identified. This resource is often

employed in dialogues with children learning to speak or in dialogues with adults

learning a foreign language. The following examples are from a textbook for

Finnish as a second language. The first example is accompanied by a picture of a

black cat to which the hand is pointing.

(32) Mikä tuo L on?what/which that-NOM be-3SG

Ir [/Interrogative] Id Pro:int‘What’s that?’

Se on musta kissa. [SS1:17]it-NOM be-3SG black-NOM cat-NOM

Id Pro:int Ir‘It’s a black cat.’

(33) – Anteeksi, mikä katu tämä on?excuse me what/which-NOM street-NOM this-NOM be-3SG

Ir Id Pro:int‘Excuse me, what is (the name of) this street?’

– Tämä on Puistokatu. [SS1:28]this-NOM be-3SG park+street-NOM

Id Pro:int Ir‘This is Park Street.’

Halliday postulates another set of functions for identifying clauses in Eng-

lish, Token and Value, with Token conflating with either Identifier and Identi-

fied. These functions are also relevant for Finnish, and are a reflection of the fact

that, if we have two roles X and Y, then Y can identify X in two ways: “either by

specifying its form, how it is recognized, or by specifying its function, how it is

valued” (Halliday 1985a: 115). Thus, one NP realizes the Token, i.e. the outward

sign, the name, the form, the holder, or the occupant, to which is given a Value,

Page 234: systematic functional linguistic

226

i.e. a meaning, a referent, a function, a status, or a role (Halliday 1985a: 115).

However, as Halliday (1985a: 123), points out, intensive processes involve a

great deal of multivalence and ambiguity, and, thus, it is not always completely

obvious which is the Token and which is the Value. The theatre example above

(23), repeated below as 34, is a fairly clear-cut example: the Identified

Kansallisteatterin Markiisitar ‘the National Theatre’s marchioness’ is the Value,

the function or role, and Eeva-Kaarina Volanen is the Token, the occupant of

this role:

(34) Kansallisteatterin Markiisitar on Eeva-Kaarina Volanen.National Theatre’s marchioness is Eeva-Kaarina VolanenId/Value Pro:int Identifier/Token‘The National Theatre’s marchioness is Eeva-Kaarina Volanen.’

Indisputable examples of identifying processes are more difficult to find

than examples of attributive processes, and indeed the distinction between identi-

fication and attribution in both Finnish and English (see 6.3.1, p. 214) is not a

clear-cut one. While identifying clauses occur far less frequently than attributive

clauses, this study is not based on frequency of occurrence, but on the grammati-

cal options that are available in Finnish. As in English, identifying processes in

Finnish are an important semiotic resource in scientific discourse, where the expe-

riential world is ordered and classified by setting up a system of technical terms

that are related to each other and to the subject matter or the data that is being

studied (see Eggins, Martin, & Wignell 1987 and Martin 1991 for an analysis of

how this is done in geography and history textbooks). Scientific statements relat-

ing phenomena or metaphenomena to each other are often realized by verbal pro-

cesses (e.g. “X is called/referred to as Y”) but identifying processes are also com-

mon (“X is Y” or “X is an example of Y”).

(35) Tärkeimmät puhe-elimet ovat keuhkot,important+SUP+NOM/PL speech-organs+NOM/PL are+3PL lungs+NOM/PL,Id/Vl Pro:int Ir/Tk

kurkunpää, kitapurje, kieli ja huulet. [FP: 12]larynx+NOM soft palate+NOM, tongue+NOM and lips+NOM/PL

‘The most important speech organs are the lungs, the larynx, the soft palate, the tongueand the lips.’

Page 235: systematic functional linguistic

227

The following are from the entries for Oidipus ‘Oedipus’ and Laatokka ‘(Lake)

Ladoga’ in a Finnish encyclopaedia:

(36) Oidipus (‘Paksujalka’) on kreikkalaisessa mytologiassaOedipus+NOM (‘fat/thickfoot’) is+3SG Greek+INE mythology+INE

Id/Tk Pro:int (NR)

thebalaisen taruston päähenkilö. [OSE: 6/4825]Theban+GEN set of tales/legends+GEN main character+NOM

Ir/Vl

‘Oedipus (‘Thickfoot’) is the main character of the Theban legends in Greek mythology.’

(37) Laatokka (ven. Ladožkoje ozero) on Euroopan suurinLadoga+NOM (Russ. Ladožkoje ozero)is+3SG Europe+GEN largest+SUP+NOM

Id/Tk Pro:int Ir/Vl

ja Neuvostoliiton neljänneksi suurin järvi. [OSE: 5/3478]& Soviet Union+GEN fourth+TRA large+SUP+NOM lake+NOM

‘Lake Ladoga (Russian name: Ladožkoje ozero) is the largest lake in Europe and thefourth largest lake in the Soviet Union.’

In the examples so far, both the Identifier and the Identified have been real-

ized by an NP in the nominative case. It is relevant to ask whether this is a neces-

sary requirement. According to Hakulinen and Karlsson (1979: 95) and Kelomäki

(1988: 2), this is indeed the case. In fact, where Kelomäki is concerned the re-

quirement is self-evidently logical, so much so that in his view any statement

explicitly asserting that both NPs in an equative clause must both be in the nomi-

native is superfluous. The logical reasoning behind this assumption is as follows.

An equative clause is defined in terms of co-extension. It follows that the two

terms to be equated in an equative clause must have the same extension. Two NPs

cannot not have the same extension if one of the NPs is in the nominative, which

realizes a bounded entity or a bounded set (see Chapter 3), and the other is in the

partitive, since that partitive indicates that the NP is unbounded. The acceptance

of this argument raises problems in the description of many naming and exempli-

fying clauses in Finnish. The following example is taken from a section of an

introductory linguistics textbook in which bound morphemes (affixes) are classi-

fied and named:

Page 236: systematic functional linguistic

228

(38) [Affiksit voidaan ryhmittää kolmia sen mukaan miten ne sijoittuvat suhteessa vartaloon.‘Affixes can be divided into three groups according to where they are placed in relation tothe stem.’]

÷ Vartaloon eteen sijoittuvat affiksitstem+ILL front+ILL place+REF/PAS+PTC+NOM/PL affixes+NOM/PL

Id/Vl

ovat prefiksejä. [JYK:124]be+3PL prefixes+PL+PAR

Pro:int Ir/Tk

‘Affixes that come before the stem are prefixes.’

... Vartalon sisään sijoittuvia affikseja sanotaan infikseiksi ... Vartalon jälkeistä affikseistakäytetään termiä suffiksi.

‘... Affixes that are placed inside the stem are called infixes ... For affixes that come afterthe stem we use the term suffix.’

From the point of view of their semantics, clauses of this type are like other iden-

tifying clauses, since it is clear to any Finn reading the text that the set of prefixes

is exhausted by the set of affixes that come before the stem and that the set of

affixes that come before the stem exhausts the set of prefixes. The second NP,

however, is in the partitive, which, as Kelomäki points out, generally means that

the NP is unbounded.

The example above, however, could be seen as carrying the crucial property

of an identifying structure as defined in this section, but in a covert form. It is

possible to construct an agnate clause in which the position and function of the

two NPs is reversed, such that the original clause and the agnate clauses are syn-

onymous (from an experiential perspective):

Page 237: systematic functional linguistic

229

1 To demonstrate this, one often needs to construct an agnate clause in which the Token andValue are in the plural.

(39) (a = b)

a. Vartaloon eteen sijoittuvat affiksit ovat prefiksejä.stem+ILL front+ILL situated+NOM/PL affixes+NOM/PL be+3PL prefix+PL+PAR

‘Affixes that come before the stem are prefixes.’

b. Prefiksit ovat vartaloon eteen sijoittuvia affikseja.Prefix+NOM/PL be+3PL stem+ILL front+ILL situated+PL+PAR affixes+PL+PAR

‘Prefixes are affixes that come before the stem.’

Here 39a is from the original text (JYK:124), while 39b displays reversal of the

NPs. If simply on the basis of the different case-forms of the two NPs (i.e. nomi-

native and partitive), we were to exclude such clauses from the category of identi-

fying clause, then it would be necessary to argue that this criterion by itself car-

ries more weight than the criterion of reversibilty of function together with the

semantics of identification, which a speaker of Finnish responds to so readily. For

these reasons, it is suggested that clauses of the type exemplified by 39 above fall

into the class of identifying process.

Grammatically and semantically there is a kind of symmetry between 39a

and 39b. The permutation illustrated here is reminiscent of what logicians refer to

as a symmetrical relation. With attributive clauses, on the other hand, symmetri-

cal clauses are not synonymous:1

(40) (a =/ b)a. Suomalaiset ovat [geeniensä perusteella

Finns+NOM/PL be+3PL genes+PL+GEN+POS/3 basis+ADE

Car Pro:int (NR)

selvästi] europpalaisia.clearly European+PL+PAR

(NR) Att [T 8-9/85: 41]

‘[On the basis of their genes, it is clear that] Finns are Europeans.

b. Eurooppalaiset ovat suomalaisia.Europeans+NOM/PL be+3PL Finns‘Europeans are Finns.’

Page 238: systematic functional linguistic

230

A similar kind of symmetry is also found in another subtype of identifying

processes, which I shall refer to as an exemplifying identifying process. The fol-

lowing is an example of an exemplifying process:

(41) a. [Julkisia tiedotuksia esitetään radiossa, televisiossa ja lehdissä.‘Public/official bulletins/reports are given on radio and television and in newspapers.’]

Ne voivat olla esimerkiksi säätiedotuksia,they+NOM/PL can+3PL be+INF example+TRA weather+reports+PL+PAR

Id/Tk Pro: int (NR) Ir/Vl

jäätiedotuksia, tiedotuksia merenkulkijoille,ice+reports+PL+PAR reports+PL+PAR seafarer+PL+PAR

poliisitiedotuksia, henkilökohtaisia tiedotuksia. [ÄA 3-4: 45]police+bulletin+PL+PAR personal+PL+PAR bulletins+PL+PAR

‘They can be, for example, weather reports, bulletins on ice conditions, maritime/seafaringreports, police bulletins, and missing persons announcements.’

As with the examples above there is a kind of symmetry if the positions and func-

tions are reversed:

(41) b. Esimerkiksisäätiedotukset ja poliisitiedotuksetexample+TRA weather+reports+PL+PAR & police+bulletin+PL+PAR

ovat julkisia tiedotuksia.be+3PL public+PL+PAR bulletin+PL+PAR

‘For example, weather reports and police bulletins are public bulletins.’

This distinguishes them from attributive processes. However, as with any gram-

matical phenomena, there is bound to be a fuzzy boundary between the two.

Page 239: systematic functional linguistic

231

1 Another verb that can occur in a resultative clause is kasvaa ‘grow, develop, mature’.

6.3.1.(iii) Other Intensive Processes

Intensive clauses include a number of subtypes, whose lexicogrammatical charac-

teristics differ from those outlined in the previous section. I shall briefly discuss

three fairly obvious and important subtypes. I first discuss resultative clauses.

These are what are generally referred to as “resultative constructions”

(tulosrakenne) in Finnish. This is followed by a discussion of what I refer to as “a

temporary intensive clause”. In the final part of this subsection, I discuss what I

have labelled “sensory intensive processes”.

Whereas intensive clauses encode or construct a relationship of sameness or

similarity, resultative clauses encode or construct a change which results in a

relationship of sameness or similarity. The process in a resultative clause is typi-

cally realized by the verb tulla ‘(to) come’,1 which is always in the third person

singular. The one that has undergone a change (or a change in status or has ac-

quired a new Attribute etc.) is realized by an NP in the elative case. The result of

the change is realized by an NP in the nominative or partitive.

(42) [Vladimir Jashtshenko, (18) ei varmaankaan aavistanut, mitä tapahtuu jos hyppääensimmäisenä maailmassa 233 senttiä korkeutta. ‘There’s no doubt that VladimirYashtshenko, 18, had little idea of what would happen when one is the first in the world toclear the 2.33 metre high jump.’]

a. Hullunmyllyhän siitä tulee.mad+mill+NOM+TIS it+ELA comes+3SG

Att Car Pro:int:res‘Pandemonium is what results.’

[Hypättyään Richmondissa tuon 233‘Having jumped that 2.33 at Richmond,’]

Page 240: systematic functional linguistic

232

1 The same problem exists in English with clauses like Vladimir became this summer’s mostinteresting sportsman.

b. Vladimirista tuli kesän kiinnostavinVladimir+ELA come+PAS+3SG summer+GEN interest+PTC+SUP+NOM

Tk/Id Pro:int:res Vl/Ir

urheilija. [HKV: HS 26.8.77]sportsperson+NOM

‘Vladimir became this summer’s most interesting sportsman.’

(43) Pekasta tulee opettaja. [IL 3.10.91:24]Pekka+ELA come+3SG teacher+NOM

Car Pro:int:res Att‘Pekka is going to be a teacher.’

As indicated in these examples, a resultative process can either be attributive

(examples 42a, 43) or identifying (example 42b). However, it is highly unlikely

for the roles in an resultative process to be actually reversed. The typical word

order is Carrier Process: resultative Attribute, and the most common variation

of this word order is that exemplified by 42a: Attribute ^ Carrier ^ Process. The

sequence of Carrier ^ Process in a resultative clause is one of the few instances in

which word order appears to be fairly fixed in Finnish. This poses a problem in

distinguishing identifying from attributive resultative clauses in Finnish.1 The

only recourse is to compare a resultative clause with an agnate intensive clause

that is not resultative.

In section 6.3.1.(i) above, it was pointed out that a crucial feature of almost

all intensive processes is the fact that the NPs agree in number. This is the gram-

matical link between resultative clauses and other intensive clauses. Thus, exam-

ple 43 contrasts with the following example, in which both NPs are in the plural:

(44) Heistä tulee opettajia. [IL 3.10.91:24]they+PL+ELA come+3SG teacher+PL+PAR

Car Pro:int:res Att‘They are going to be teachers.’

Page 241: systematic functional linguistic

233

A feature that distinguishes resultative clauses from other intensive pro-

cesses is the fact that they are affected by the polarity of the clause. If the polarity

is negative, the second inherent role is in the partitive. The following clause with

negative polarity contrasts with the second clause in example 42 above:

(45) Vladimirista ei tullut urheilijaa.Vladimir+ELA NEG+3SG come+PTC sportsperson+PAR

Car Pro:int:res Att‘Vladimir didn’t become/turn out to be a sportsman.’

The effect of negative polarity is a feature that resultative clauses share with ma-

terial processes and some mental and verbal processes.

There are other intensive processes in which one of the inherent roles is in

one of the other intermediate cases (see Chapter 2), i.e. the translative or the

essive. These will be referred to as “temporary intensive processes” (abbreviated:

Pro:int:temp). These correspond to some of the clause types in which there is

what has been traditionally referred to as a “complement adverbial”

(predikatiiviadverbiaali) (see, e.g. Hakulinen & Karlsson 1979: 211 ff.). The

process in a temporary intensive process is typically realized by verbs such as

olla ‘(to) be’, tulla ‘to (come)’, jäädä ‘(to) remain’, toimia ‘(to) act/function [as

something]’, pysyä ‘to stay’ etc. .

(46) Silloin tosin jäivät työn tulokset heikoiksi.then to be sure remain+3PL work+GEN result+NOM/PL weak+PL+TRA

(NR) (Modal) Pro:int:temp Car Att [HKV]‘Then to be sure work results remained meagre.’

(47) a. Fagerholm tuli 1926 Arbetarbladetin toimittajaksiFagerholm came 1926 Arbetarbladet+GEN journalistCar & Tk/Id Pro:int:temp (NR) Att

b. ja oli 1934-37 lehden päätoimittajana. [HKV]& be+3SG 1934-37 paper+GEN head+journalist+ESS

Pro:int:temp (NR) Vl/Ir

‘Fagerholm became a journalist for Arbetarbladet in 1926 and was editor-in-chief from1934 to 1937.’

Page 242: systematic functional linguistic

234

(48) miksi ihmiset ovat äkkiä tulleetwhy person+NOM/PL be+3PL all of a sudden come+PTC+NOM/PL

(NR) Car Pro:int:temp-> (NR) <-Pro

aggressiivisiksi [HKV]aggressive+PL+TRA

Att

‘Why have people become aggressive all of a sudden?’

As with resultative processes, a temporary intensive process can either be attribu-

tive (examples 46, 47a, 48) or identifying (example 47b).

Temporary intensive processes are grammatically linked to other intensive

clauses by the fact that the NPs agree with each other in number, as illustrated by

examples 46 and 48, where the translative forms of the adjectives heikoiksi

‘weak’ and aggressiivisiksi ‘aggressive’ are in the plural, agreeing with the sub-

ject in number. However, there is some variation in Finnish in this respect: if the

NP realizing the Carrier is plural, the NP realizing the Attribute can either be sin-

gular or plural. This seems to be particularly true of an Attribute realized by an

NP in the essive.

Like other intensive processes, processes of this type typically construe a

relationship in which a participant is assigned a certain quality (e.g. being weak

or happy) or a certain (social) role (e.g. being a journalist or a teacher). Tempo-

rary processes can be generally characterized by the fact that the Attribute or

Identifier is explicitly marked as something that is not a permanent or an inherent

feature of the Carrier or Identified: the Attribute or Identifier is seen as a tempo-

rary feature or the result of a particular circumstance or set of circumstances (cf.

Kelomäki 1988: 144).

However, the label “temporary” is slightly misleading in the case of some

clauses in which the second NP is in the translative (-ksi), since, in some in-

stances the NP in the translative refers to a permanent feature of the Carrier. For

example, it would have been possible to continue after the first clause in 47 above

by saying ‘and worked as a journalist for the rest of his life’. If we contrast a tem-

Page 243: systematic functional linguistic

235

porary clause (NP + tulla ‘come’ + NP-translative) to a resultative clause (NP-

elative + tulla ‘come’ + NP), the focus in a temporary clause is on the transition

rather than on the result. Thus, the label “temporary” includes both temporariness

and transition. When the second NP in a temporary clause is in the essive

(-na/nä), on the other hand, the focus is typically on temporariness of the rela-

tionship between the NPs. However, the essive can also be used, for example,

when the focus is on role or function rather than on the assignment of an Attrib-

ute:

(49) Pekka on opettaja.Pekka-NOM be+3SG teacher+NOM

Car Pro:int Att‘Pekka is a teacher.’

(50) Pekka on opettajana Turussa.Pekka-NOM be+3SG teacher+ESS Turku+INE

Car Pro:int:temp Att Circ‘Pekka is (working as) a teacher in Turku.’

Example 50 focuses on the capacity in which Pekka is functioning, rather than on

the assignment of the Attribute ‘teacher’ to Pekka as in 49.

The fact that intensive clauses with an NP in the essive typically realize

temporariness is reflected in a comment made by Hakulinen & Karlsson (1979:

213). In their view, if the lexical meaning of an Attribute includes temporariness,

then the essive is not possible. For example, epävakainen ‘unsettled’ is an attrib-

ute that is not a permanent or inherent feature of something:

(51) a. Sää on epävakainen. (Example from Hakulinen &weather-NOM is-3SG unsettled-NOM Karlsson 1979: 213.)‘The weather is unsettled’

b. ?? Sää on epävakaisena. weather-NOM is-3SG unsettled-ESS

‘The weather is being unsettled.’

To the extent that this is true, it can be explained by the fact that it is unnecessary

to explicitly encode a clause as a temporary process if it is something that is al-

ready understood as being temporary.

Page 244: systematic functional linguistic

236

The third type of relational intensive process to be discussed in this section

is what I refer to as a sensory process. Sensory processes are always attributive.

The significant thing about these processes is that the attribute is realized by an

NP (typically with an adjective as the Head) in the ablative (or, less frequently,

the allative). The process is realized by one of a small set of verbs that refer to

sense perception or impression, e.g., maistuu ‘taste’, näyttää ‘looks/seems/-

appears’, vaikuttaa ‘seems’, haista/haiskahtaa/tuoksua ‘smell’, kuulostaa

‘sound’, tuntua ‘feel’.

(52) Se kuulosti niin ihanalta. [Tel1:5]it+NOM sound+PAS+3SG so wonderful+ABL

Car Pro:int:sen Att‘It sounded so wonderful.’

(53) Maallikon silmään juttu näyttäälayperson+GEN eye+ILL matter/affair+NOM seem+3SG

(NR) Car Pro:int:sen

hullulta. [HKV]mad/crazy/idiotic+ABL

Att‘In the eyes of a layperson, the affair seems idiotic.’

Again, with attributive sensory processes as with other intensive processes,

the NPs realizing the Carrier and Attribute agree with each other in number:

(54) Ne vaikutti(vat) tyytyväisiltä.they+NOM/PL appear/seem+PAS+3(PL) content/satisfied+PL+ABL

Car Pro:int:sen Att‘They seemed satisfied.’

(55) [Puiston pyöreisiin lyhtyihin syttyi oranssinväriset valot.‘The round lamps in the park were lit by orange-coloured lights.’]

Ne näytti valtavilta appelsiineilta. [AR: 25]they-NOM/PL appear/look-3SG enormous+PL+ABL orange+PL+ABL

Car Pro:int:sen Att‘They looked like enormous oranges.’

This agreement is consistent in both spoken and written Finnish.

Page 245: systematic functional linguistic

237

6.3.2. Inclusive Processes

As pointed in each of the sections above, an important feature of intensive pro-

cesses is the fact that there is number agreement between the NPs. Intensive pro-

cesses are semantically characterized by the fact that they set up a relation of

sameness or similarity: ‘X is (like) Y’. On these grounds, what Hakulinen &

Karlsson (1979: 98) refer to as inclusive clauses constitute a distinct process type

in Finnish. In an inclusive clause in Finnish the relationship that is set up is one of

inclusion: X belongs to the set Y:

(56) Paita on ihmiskunnan vanhimpiashirt-NOM/SG be-3SG humankind+GEN oldest+PL+PAR

vaatekappaleita. [HKV]clothing+item+PL+PAR

‘The/A shirt is amongst the oldest clothing items of humankind.’

(57) vain murto-osa on valkoisia [WALTARI]only break/fraction+part-NOM/SG be-3SG white+PL+PAR

‘only a fraction are white’

In this particular type of inclusive clause, one of the nominals is in the singular

nominative and the other is in the plural partitive.

In the other type of inclusive clause discussed by Hakulinen & Karlsson

(1979: 98), one of the nominals is in the nominative and the other is in the parti-

tive. In Hakulinen & Karlsson’s examples, both nominals are in the singular:

(58) Tämä kylä on Hämettä.this-NOM village-NOM be+3SG Häme+PAR/SG

‘This village is part of Häme.’

(59) Tämä on sitä uuttathis-NOM be+3SG it+PAR/SG new+PAR/SG

poronjuustoa.reindeer+GEN+cheese+PAR/SG

‘This is that new Reindeer Cheese.’

Page 246: systematic functional linguistic

238

Based on Sadeniemi (1970), Hakulinen & Karlsson define inclusive clauses as

those in which the complement (i.e. the Attribute) realizes the original function of

the partitive, i.e. ‘part of something’. According to Hakulinen & Karlsson, the

relationship between the NPs above could be seen in terms of set theory:

tämä kylä ‘this village’ 0 Häme (a province in Finland)

tämä ‘this’ 0 uutta poronjuustoa ‘new Reindeer Cheese’.

Similarly, with 56 above:

paita ‘shirt’ 0 ihmiskunnan vanhimpia vaatekappaleita

‘oldest clothing items of humankind’

However, if we simply use this kind of criteria, then it is difficult to distin-

guish inclusive processes from other attributive processes, for example 12 above,

repeated below as 60:

(60) [mul ois ens viikon torstaihin aikaa kirjoittaa se puhtaaksi, [Tel2:13]‘I’ve got until next Thursday to type up a final copy’

joka on mulle helvetin hidasta hommaawhich-NOM be+3SG I+ALL hell+GEN slow+PAR work+PAR

Car Pro:int (NR) Att‘which is hellishly/damned slow work for me’

One could equally argue that tämä (homma) ‘this job’ belongs to the set of slow

jobs (i.e. tämä (homma) ‘this (job)’ 0 helvetin hidasta hommaa ‘hellishly slow

work/job(s)’).

What links Hakulinen & Karlsson’s examples (58 ) 59) to the other exam-

ples above (56 ) 57) is the lack of agreement between the NPs realizing the Car-

rier and Attribute. Thus, if 58 above, for example, is changed so that the initial

NP is plural, the NP at the end of the clause remains in the singular:

(61) Nämä kylät ovat [~ spoken: on] Hämettä.these-NOM village-NOM/PL be+3PL be+3PL Häme+PAR/SG

‘These villages are part of Häme.’

Page 247: systematic functional linguistic

239

Hakulinen and Karlsson (1979: 98) regard inclusive clauses as a separate

clause type but relate them to what are traditionally called “complement clauses”,

i.e. clauses with a subject and complement. On the basis of some grammatical

features, these are like attributive processes: the process is realized by olla ‘be’,

the NPs are in the nominative or partitive, the functions are unaffected by word-

order changes. However, it seems to me that they are semantically and grammati-

cally distinct from attributive processes, and, thus, in line with Hakulinen &

Karlsson, I shall regard them as a separate type of relational process. Grammati-

cally inclusive clauses can be distinguished by the lack of agreement between the

NPs; semantically they are distinguished by the fact that the labels Carrier and

Attribute are clearly inappropriate. More appropriate functional labels would ei-

ther be Member and Set or Part and Entirety.

It should be noted that an example such as 62 below is an ordinary attribu-

tive process in Finnish. There appears to be no clear grammatical distinction be-

tween examples 62 and other attributive processes in Finnish.

(62) Suomalaiset ovat geeniensä perusteellaFinns+NOM/PL be+3PL genes+PL+GEN+POS/3 basis+ADE

Car Pro:int (NR)

selvästi europpalaisia.clearly European+PL+PAR

(NR) Att [T 8-9/85: 41]

‘On the basis of their genes, it is clear that Finns are Europeans.’

Thus, what are referred to as inclusive processes are distinct from what could be

seen as a classifying attributive process. While on semantic grounds, one could

argue that in example 62 that suomalaiset ‘Finns’ belongs to the class of Europe-

ans, this clause is grammatically similar to the clause Suomalaiset ovat vaaleita

‘Finns are blonde’.

Page 248: systematic functional linguistic

240

6.3.3. Ambient Processes

This section deals with what could be considered a subtype of intensive process.

What I refer to as an “ambient process” in Finnish corresponds to some of the

clauses in English that have a “dummy subject”, for example, it was cold on Sun-

day, it was late, it’s Thursday today. The translation equivalents of these in

Finnish would correspond to “on Sunday was cold” and “was late” and “on this

day (today) is Thursday”. Thus, Finnish is one of those languages in which one

says “In London is cold” rather than “London is cold” (although “London is cold”

is also possible) (see Lyons 1977b: 476).

An ambient process characterizes or assigns an Attribute to a spatial or tem-

poral situation. The process is typically realized by the verb olla ‘be’ and one of

the inherent roles is realized by an NP in the nominative or partitive. These fea-

tures relate ambient processes to intensive processes. The other participant is gen-

erally in one of the locative cases and it realizes a Circumstance of time or place.

The circumstantial element is not obligatory:

(63) sunnuntaina oli sit niin hienoa [TEL1:1]Sunday+ESS be+PAS+3SG then so wonderful/fineCirc Pro:int:amb (NR) Att‘Sunday was then so wonderful.’

(64) Jossakin täytyy olla parempaa ja väljempää. [HKV]somewhere+INE must be+INF better+PAR & spacious+COMP+PAR

Circ (Mod) Pro:int:amb Att Att‘There must be somewhere where it is better and more spacious [“somewhere must bebetter and more spacious”]’

(65) meillä on rauhatonta [HKV]we+ADE (at our place) be+3SG restless+PAR

Circ Pro:int:amb Att‘It’s restless (noisy, rowdy) at our place/Our place is restless.’

(66) (Ulkona) on kylmä. (from Hakulinen & Karlsson 1979: 301)outside+ESS be-3SG cold-NOM

Circ Pro:int:amb Att‘It’s cold (outside).’

(67) (Nyt) on myöhä ~ yö. (Hakulinen & Karlsson 1979: 93)(now) is late ~ nightCirc Pro:int:amb Att‘It’s (now) late ~ night (night-time).’

Page 249: systematic functional linguistic

241

Hakulinen & Karlsson (1979: 93,301) refer to clauses like this as ‘situation

clauses’ (Finnish: tilalause). According to them, a situation clause is expressed

by the structure (A) V or (A) V C (where ‘A’ stands for Adverbial, V for Verb and

C for Complement). By this criterion, they also regard the following as a situation

clause:

(68) (Nyt ~ täällä) sataa. (Hakulinen & Karlsson 1979: 93)(now ~ here+ADE) rain+3SG

‘It’s raining now ~ here.’

Thus, what appear to be two distinct grammatical structures are subsumed

by Hakulinen and Karlsson under the one general clause type. What is common to

the two is the fact that there is no grammatical subject and the initial Adverbial is

optional. However, there are also crucial differences between an example such as

68 and ambient processes. For these reasons, example 68 is considered to be a

subtype of material process referred to as a meteorological process in this study

(for further discussion of material and meteorological processes, see 6.4 and

6.4.3.(i)).

As defined in this study, ambient processes are a subtype of relational pro-

cess: they set up a relation between X and Y. They differ from intensive processes

in that an Attribute is assigned to a Circumstance rather than to an Entity (i.e. the

Circumstance is the Carrier). A clause such as 68, on the other hand, constructs a

material change. Unlike the verb olla ‘be’, the verb in 68 can be qualified by an

adverb of manner, e.g. satoi rankasti ‘it rained hard’. If an NP is added to 68, its

function is different to the function of an NP in an ambient process:

(69) Täällä sataa lunta.here+ADE) rain+3SG snow+PAR

‘It’s snowing (raining snow) here.’

The final NP lunta ‘snow’ does not assign an attribute to the Circumstance täällä

‘here’; it qualifies the process of raining. One can also say in Finnish: sataa vettä

‘it’s raining water’.

Page 250: systematic functional linguistic

242

Hakulinen & Karlsson’s motivation for subsuming ambient and meteoro-

logical processes under the one general clause type rests on two factors. Firstly,

these clauses are subjectless: there is nothing corresponding to a “dummy sub-

ject” in Finnish (cf. English It’s cold, It’s raining). Secondly, if something is

added to the beginning of the clause, it would be a Circumstance. Without any

co-text to indicate otherwise, if someone says to me on kylmä “is cold” or sataa

“rains”, I would generally assume that they mean that it is cold or it is raining

‘now ~ at the moment ~ here ~ outside etc.’. In other words, the Attribute ‘cold’

qualifies the temporal or spatial domain of discourse, similarly the process of ‘rain-

ing’ is located in the temporal or spatial domain of discourse. This is, of course,

also valid for the English translation equivalents. What is different about Finnish

is that the unmarked place for the Circumstance is at the beginning of the clause.

Moreover, in the case of ambient processes as defined in this study, it is this Cir-

cumstance that is being assigned an Attribute.

The optionality of the initial Circumstance is a textual feature. Process types

as defined in this study are not based on textual variation, as textual variation is a

realization of the textual metafunction. However, to regard the variation that is at

issue here as textual means re-assessing what is meant by textual. For some, tex-

tual phenomena seems to be restricted to considerations of the way in which lan-

guage make links with itself, i.e. to the co-text. Equally important, however, is the

way in which language makes links with the situations in which it is used, i.e. the

context. In SF theory, the textual metafunction also encompasses those features

that are related to the way in which language makes links with the situations in

which it is used. This is what is at issue in ambient and meteorological processes.

Hakulinen & Karlsson seem to treat this variation as syntactic. Their analy-

sis is partly inspired by transformational-generative grammar. Given the assump-

tion that S ÷ NP + VP in transformational-generative theory, the absence of an

NP is problematic. A further problem ensues from the fact that if an NP is added

to the clauses in Finnish, the NP is not the subject. Although not actually made

explicit, this appears to be the reasoning in Hakulinen & Karlsson’s analysis.

From a Finnish perspective, of course, the original assumption (S ÷ NP + VP) is

questionable.

Page 251: systematic functional linguistic

243

By treating the optionality of the initial Circumstance as a syntactic feature,

Hakulinen & Karlsson’s analysis obscures the textual patterning in Finnish. This

can be illustrated by one of Hakulinen & Karlsson’s examples:

(70) (Täällä ~ Hämeessä) on kylmä. (Hakulinen & Karlsson 1979: 93)here+ADE here+ADE be-3SG cold-NOM

‘It’s cold (here ~ in Häme).’[Häme is a province in Finland.]

It seems to me that this example is misleading: I would argue that the initial ad-

verbial is not optional in the clause Hämeessä on kylmä. It can only be omitted if

it is presupposed by ellipsis, i.e. there has to be some reference to Häme in the

preceding text. Thus, the relationship between on kylmä [is cold] and täällä

[here] is different to the relationship between on kylmä and Hämeessä [in Häme].

As indicated earlier, without any co-text to indicate otherwise, on kylmä ‘it’s

cold’ would be interpreted as täällä on kylmä ‘it’s cold here’ or ulkona on kylmä

‘it’s cold outside’, i.e. if Pekka says to Jussi “on kylmä”, the interpretative source

is in the here and now of the speech situation. If on the other hand, Pekka is tell-

ing Jussi about his trip to Häme, and says oli kylmä [was cold], then it would be

interpreted as Hämeessä oli kylmä [in Häme was cold], i.e. the interpretative

source is in the co-text.

Ambient processes like on kylmä [is cold] are part of a wider, general phe-

nomenon in Finnish: there are other declarative clauses in Finnish that begin with

the verb. If something is added to a verb-initial declarative, it would be a Circum-

stance and its unmarked place would be at the beginning of the clause. Clauses

like this include example 5 from the beginning of this chapter: Puhuttiin naisten

rooleista ‘The discussion was concerned with the roles of women’ and examples

such as Syttyi sota [break out+PAS+3SG war-NOM] ‘There was a war/A war broke

out’. As discussed in Chapter 7, these clauses too are about the temporal or spatial

domain of discourse.

Page 252: systematic functional linguistic

244

6.3.4. Circumstantial (Relational) Processes

6.3.4.(i) Introductory Comments

Circumstantial relational processes also set up a relation but the relation is be-

tween an Entity and a Circumstance: something is located somewhere in space or

time. The space in which something is located can also be abstract, for example,

a physical or mental state or an activity. This type of circumstantial process will

be referred to as a general circumstantial process. A subtype of circumstantial

process in Finnish is a possessive process.

6.3.4.(ii) General Circumstantial Processes

General circumstantial processes can be grammatically characterized as processes

in which one of the inherent roles (X) is realized by an NP in either the nomina-

tive or partitive, the verb is typically realized by olla ‘(to) be’, and the other role

(Y) is realized by an NP in a locative case-form or by a PP. Thus role X is located

in space or time at Y. I shall refer to role X as the Positioned (Posit:ed): it realizes

the entity that is given a location or positioned in some (set of) circumstances.

Role Y will be referred to as the Circumstance (Circ). In the following examples,

I shall only distinguish between circumstances of place and time; in a more deli-

cate analysis, different types of Circumstances would need to be distinguished.

(71) sä olit mökillä viikonloppuna [TEL1:1]you(SG)-NOM be+PAS+2SG cottage+ADE weekend+ESS

Posit:edi Pro:rel + Posit:edi Circ:place Circ:time‘You were at the (summer) cottage at the weekend.’

(72) a. Ritvahan on naimisissaRitva+TIS be+3SG marriage+INE

Posit:ed Pro:rel Circ‘Ritva’s married of course’

b. -- Ritva on ollut ilmarisella töissäRitva+NOM be+3SG be+PTC Ilmarinen+ADE work+PL+INE

Posit:ed Pro:rel Circ Circ‘Ritva has been at Ilmarinen’s (a firm)/has been working for Ilmarinen’

Page 253: systematic functional linguistic

245

c. -- mut on nyt lomallabut be+3SG now leave/holiday+ADE

Pro:rel (NR) Circ‘but is now on leave.’

d. -- Jussi on -- jossain pankissa [CA3:24-25]Jussi be+3SG some bank+INE

Posit:ed Pro:rel Circ‘Jussi is at some bank/working for some bank.’

(73) me olemme lakossa [CA8:5]we+NOM be+1PL strike+INE

Posit:edi Pro:rel+Posit:edi Circ‘We’re on strike.’

(74) hän on myös samallahe+NOM is+3SG also same+ADE

Posit:ed Pro:rel (NR) Circ

irtisanottujen listalla [CA8: 5]sack/give notice+INDE+PL+GEN list+ADE

‘He is also on the same list of retrenchments.’

(75) siinä sä oot perkeleen oikeassathere+ESS you-SG be+2SG devil+GEN right+INE

* (NR) * Posit:edi * Pro:rel + Posit:edi * Circ *‘You’re damn right there.’ [AR:160]

As illustrated by the examples, the positioning need not be concrete or spatial: X

on naimisissa ‘X is married [in (the institution of) marriage]’, X on lakossa ‘X is

on strike’, X on listalla ‘X is on the list’. They all have a similar grammatical

patterning in Finnish. They differ in meaning, of course, but these differences are

handled at a greater degree of delicacy in the approach adopted here.

In the examples above, the Positioned is the Theme of the clause, but it is

also possible for the Circumstance to be the Theme:

Page 254: systematic functional linguistic

246

(76) [From an encyclopaedia entry on Erik XIV of Sweden and Finland.]

Veljesten välinen riita leimahti ilmiliekkiin 1563, jolloin alkoi ns. pohjoismainenseitsenvuotinen sota lähinnä Baltian kauppaetujen vuoksi.

‘The conflict between the brothers [i.e. Erik & his brother] broke out in full flame in 1563with the start of the so-called seven-year Nordic War, which was waged mostly for traderights in the Baltic region.’

Eerikiä vastassa olivat Lyypekki,Erik+PAR against/opposition+INE be+PAS+3PL Lübeck+NOM,

* Circ * Pro:rel * Posit:ed

Puola ja Tanska [HKV]Poland+NOM & Denmark+NOM

*

‘Opposing Erik [in opposition to Erik] were Lübeck, Poland and Denmark.’

The basic difference between example 76 and examples 71 ) 75 above is textual.

With Circumstantial relational processes in Finnish, a problem is posed by

clauses in which there is a nominalized verb form, the so-called 3rd infinitive (see

Figure 3-9 (p. 85)):

(77) Anna on uimassa.Anna is swim+INF+INE

‘Anna is swimming’

A clause like this could be a response to the question ‘Where is Anna?’. Simi-

larly, one could have Anna on kävelemässä ‘Anna is walking’ from the verb

kävellä ‘(to) walk’, Anna on nukkumassa ‘Anna is sleeping’ from the verb nuk-

kua ‘(to) sleep’, Anna on syömässä ‘Anna is eating (having lunch)’ from the verb

syödä ‘(to) eat’.

Clause 77 differs from its English translation equivalent in that it patterns

more like a circumstantial relational clause than a material clause. Example 77

has a similar structure to Anna on Suomessa ‘Anna is in Finland’, except that the

Circumstance uimassa has a stem (ui-) that is derived from the verb uida. The

form uimassa consists of a verb stem followed by a nominalizing affix, the so-

called 3rd infinitive, and the inessive case-ending. It could, in fact, be translated

as something like ‘in the process/activity of swimming’. Following Leino et al.

Page 255: systematic functional linguistic

247

1 Leino et al. (1990) is an analysis of Finnish locative cases based on Jackendoff’s (1983)conceptual semantics.

(1990),1 I regard a form like this as being analogous to a concrete spatial circum-

stance.

(78) Anna on uimassa.Anna is swim+INF+INE

Posit:ed Pro:rel Circ

‘Anna is swimming’

In clauses like this, the Circumstance construes an activity or a state in which an

entity is located. However, in Finnish, the activity cannot be modified by some-

thing like hyvin ‘well’. This can be compared with English where the modifier

“well” is possible: Anna is swimming well, as said, for example, by Anna’s coach

watching Anna as she trains for the Olympics.

The difference between these Finnish clauses and their English translation

equivalents can be also illustrated by the following example:

(79) Jumala on olemassa.God is be+INF+INE

Posit:ed Pro:rel Circ

‘God exists.’

The verb (to) exist can only be translated into Finnish by this kind of structure,

which corresponds to something like “is in the state of being”.

The analysis, however, is complicated by the fact that the process realized

by the stem of the nominal realizing the Circumstance can be extended to a Goal:

(80) Anna on maalaamassa taloa.Anna is paint+INF+INE house+PAR

Posit:ed Pro:rel Circ Goal

‘Anna is painting a/the house.’

Page 256: systematic functional linguistic

248

This is a reflection of the fact that the stem of the Circumstance is indeed a verb

and not a noun.

In the examples given so far, the process in a circumstantial process is real-

ized by olla ‘be’. There are doubtless other verbs that can realize a circumstantial

relational process. An obvious example is sijaita ‘to be situated/ located’:

(81) Hasanniemen kesäteatteri sijaitseeHasanniemi+GEN summer+theatre+NOM situate+3SG

Loc:ed Pro:rel

Pyhäselän rannalla. [TB: 9]Pyhäselkä+GEN shore+ADE

Circ

‘The Hasanniemi summer (open-air) theatre is situated on the shores of Lake Pyhäselkä.’

(82) Puheen perustana olevat hermomekanismit sijaitsevat vasemmassa aivopuoliskossa.[HKV]

‘The nerve mechanisms that form the basis of speech are located in the left hemisphere ofthe brain.’

The verb sijaita ‘be situated/located’ could be regarded as a stylistic variant of

olla ‘be’.

The category of circumstantial process that is recognized here partly in-

cludes what traditional grammars of Finnish have referred to as existential clauses

(eksistentiaalilause). The following examples would be regarded by many Finn-

ish linguists as existential clauses.

(83) tääl on autoja [CA2:28]here(+ADE) be+3SG car+PL+PAR

Circ Pro:rel Posit:ed‘There are cars here.’

(84) torstaina oli ihan hirveän kova myrsky [O]Thursday+ESS be+PAS+3SG really terrible+GEN severe+NOM storm+NOM

Circ:time Pro:rel Posit:ed‘There was a (really) terribly severe storm on Thursday.’

As indicated by the function glosses, examples like this are considered to be cir-

cumstantial processes in this study, since they construct a relation whereby some-

thing is located somewhere in space or time. As I see it, the grammatical and se-

Page 257: systematic functional linguistic

249

mantic features associated with the Finnish eksistentiaalilause constitute simulta-

neous sets of options cutting across the various process types. These features will

be discussed more fully in section 6.7.

6.3.4.(iii) Possessive Circumstantial Processes

While possessive processes can be regarded as a subtype of circumstantial pro-

cesses in Finnish, the analysis presented here is rather lengthy. This is because it

is generally assumed that there are possessive clauses in Finnish, but there seem

to be various ways of defining and delineating them. Thus this section will com-

ment fairly extensively on other views of what constitutes a possessive clause in

Finnish.

In his analysis of English, Halliday (1985a: 121-123) regards possessive

processes as realizing a relation of ownership: in concrete instances, at least, one

entity possesses another. The Possessor is typically human. Following Halliday, I

shall label the functions in a possessive process as Possessor (Poss:er) and Pos-

sessed (Poss:ed). Thus, the first example below is a circumstantial process while

the second example is a possessive one.

(85) Kadulla on useampia autoja.street+ADE be+3SG several+PL+PAR car+PL+PAR

Circ Pro:rel Positioned‘There are several cars in the street.’

(86) Elviksellä on useampia autoja.Elvis+ADE be+3SG several+PL+PAR car+PL+PAR

Poss:er Pro:rel Poss:ed‘Elvis has several cars.’

While a possessive process in Finnish is structurally similar to a circumstantial

process, a grammatical feature that distinguishes it from circumstantial processes

is the fact that the sequence is typically fixed: Possessor ^ Process: relational ^

Possessed. Semantically, possessives differs in that the relationship between the

entities need not be one of location or positioning in a concrete sense: in 86

above, the cars in question are not necessarily located near or in the vicinity of

Elvis. Elvis could be in Europe, while his cars could be in America. With 85, the

Page 258: systematic functional linguistic

250

1 This is generally referred to as the “subject” of an existential clause by Finnish grammarians(e.g. T. Itkonen (1979)).

cars are necessarily in the street. This appears to be a reflection of the fact that a

possessive relationship is conventionalized in Finnish. Thus, the structure NOMI-

NALADESSIVE + olla ‘be + NOMINAL exemplified by 86, where the first nominal

refers to a human and the second to an Entity, is generally interpreted in terms of

possession rather than location.

As indicated in example 86, the Possessor in a possessive process in Finnish

is realized by an NP in the adessive: Elvikse+llä ‘at/on/by Elvis’. The finite verb

is always a third person singular form of the verb olla ‘be’; if the Possessor is

plural, the verb is still in the singular.

(87) Filmitähdillä on useampia autoja.film+star+PL+ADE be+3SG several+PL+PAR car+PL+PAR

Poss:er Pro:rel Poss:ed‘Film stars have several cars.’

Possessive processes have features that are typically associated with what Finnish

linguists call existential clauses. It is possible for one of the inherent participants,1

in this instance, the Possessed, to be in the partitive. A feature peculiar to posses-

sive processes is that while a singular, bounded entity in a clause with positive

polarity is in the nominative singular (e.g. Minulla on uusi auto ‘I’ve got a new

car’), a personal pronoun is in the accusative: Minulla on sinut [I+ADE is

you(SG)+ACC] ‘I’ve got you’.

The following is a text example of possessive processes in Finnish. It is ap-

parent from the co-text that the object referred to (the uus ‘new (one)’) is a (radio-

)cassette player, which could be regarded as a typical possession in western soci-

ety:

Page 259: systematic functional linguistic

251

(88) a. <A> onks sul uusbe+3SG+Q you(SG)+ADE new-NOM

Pro:rel Poss:er Poss:ed‘Have you got a new one

b. vai onks toi meiän?or be+3SG+Q that-NOM our+GEN

Pro:rel Tk/Id Vl/Ir (possessor)‘or is that ours?’

<B> .. ei tää on tämmönen laina homma ..‘no this is a kind of borrowed thing ..’

Halliday regards clauses such as The piano is Peter’s as a possessive identifying

clause in English, and thus the translation equivalent given for 88b would be re-

garded as a possessive identifying clause. I shall not pursue arguments for or

against analysing clauses like The piano is Peter’s in this way in English. In

Finnish, on the other hand, it is clear that clauses 88a and 88b are grammatically

very different from each other. Example 88b represents a subtype of intensive

identifying processes in Finnish: the Token/Identified is in the nominative case

(the partitive is not possible) and it agrees with the finite verb in number and per-

son.

The label “possessive process” is problematic in that examples of possessive

clauses in actual text appear to be seldom concerned with ownership, as sug-

gested by Halliday, or with a “permanent possessive relationship”, as suggested

by Kangasmaa-Minn (1971: 258). As is clear from Halliday’s writing and from

the context in which Kangasmaa-Minn refers to a permanent possessive relation-

ship (in contrast to a temporary circumstantial one), the problem inheres in using

language “turned back on itself” (Firth 1957: 190) and in labelling linguistic cate-

gories that are, in fact, ineffable (see 2.4.11 (p. 63) and Halliday 1988). Although,

as indicated by the co-text of example 88a above, ownership can be at issue in a

possessive process, the meaning of a possessive process is more ineffable, and, in

the final analysis, it can only be defined by the oppositions between it and other

process types. Sometimes the relationship is a socially defined “belonging” or

“being in possession of”, as in the following examples:

Page 260: systematic functional linguistic

252

(89) onhan mulla mies [CA8:5]be-3SG+TIS I+ADE man/husband+NOM

Pro:rel Poss:er Poss:ed‘After all I’ve got a husband.’

(90) kenellä on eniten seiskoja? [CA6:5 (children playing cards)]who+ADE be+3SG most seven+PL+PAR

Poss:er Pro:rel (NR) Poss:ed‘Who’s got the most sevens?’

minulla, mull on ainakin seittemäntoista seiskaaI+ADE I+ADE be+3SG at least seventeen+NOM seven+PAR

Poss:er Poss:er Pro:rel Poss:ed‘I have, I’ve got at least seventeen sevens.’

There are, of course, many other, more abstract relationships between the Pos-

sessor and the Possessed, as in the following example:

(91) heill on ... ihan omat kuvionsa [CA3:4]they+ADE be+3SG entirely own+NOM/PL figure/pattern+POS/3

Poss:er Pro:rel Poss:ed

‘They’ve got their own life styles [Finnish kuviot [pattern-NOM/PL] refers here to a moreor less fixed set of things that one does and set of people that one meets].’

In example 91, the meaning is more like ‘X pertains to Y’ or ‘Y is associated with

X’.

A number of bodily states can also be realized as possessive processes. For

example:

(92) M(in)ulla on jano ~ kylmä ~ päänsärky.I+ADE be+3SG thirst-NOM cold-NOM headache-NOM

Poss:er Pro:rel Poss:ed‘I’m thirsty/I’m cold/I’ve got a headache.’

(93) Mulla on paha olo [AR:159]I+ADE be+3SG bad-NOM feeling-NOM

Poss:er Pro:rel Poss:ed‘I feel bad/I feel sick.’

(94) Hänellä on syöpä.(s)he+ADE be+3SG cancer-NOM

Poss:er Pro:rel Poss:ed‘She/he’s got cancer.’

Page 261: systematic functional linguistic

253

These can be contrasted with material processes, such as minua janottaa

[me+PAR thirsts+3SG] ‘I feel thirsty’; päätäni särkee [my head aches] ‘I’ve got a

headache’ (see 6.4.3.(ii) below).

The claim that instances like 92 ) 94 are possessive processes rests on the

fact that the initial nominal is not interpreted as a location. Example 94 above can

be contrasted with the following example, where the initial nominal is in the in-

essive case:

(95) Hänessä on syöpä. (Cf. 94 above.)(s)he+INE be+3SG cancer‘She’s/he’s got cancer’.

This can be regarded as a circumstantial clause ‘in X is Y’. Example 95 focuses

on the fact that ) in some sense ) the disease “is contained” in the person; where-

as the possessive clause (94) sets up a relation that associates the disease with the

person. Of relevance here is Leino’s (1990: 129-130) analysis of the way in

which Finnish cases realize basic spatial relations. According to Leino, the rela-

tionship between what he refers to as the trajector and the landmark ) which in

this context can be translated as the relationship between the Circumstance and

the Positioned ) is one of inclusion when the Circumstance is in the inessive

(pullo+ssa ‘in(side) the bottle; stuck (fast) to ~ intimately attached to the bottle’),

but one of association when it is in the adessive (pöydä+llä ‘on/at/near/in the

vicinity of the table’). It seems to me that this distinction also applies to these

more abstract instances.

While there is a fuzzy line between possessive processes and general cir-

cumstantial processes, i.e. a clause of the form X:llä on Y [X-adessive is Y] can

mean ‘at X is Y’ or ‘X has Y’, there seems to be a more clear-cut line between

possessives, on the one hand, and some other clause types which some Finnish

linguists have regarded as possessive. Hakulinen and Karlsson (1979: 96-97, 209)

refer to clauses such as the following as inanimate possessives constructions:

(96) Veneessä on kaksi mastoa. [Hakulinen & Karlsson 1979: 97]boat+INE be+3SG two-NOM mast+PAR

‘There are two masts on the boat/The boat has two masts.’

Page 262: systematic functional linguistic

254

The possessive translation in English has the advantage of retaining the textual

structure of the Finnish clause, but translating a Finnish clause by an English pos-

sessive does not make it a possessive in Finnish: the meanings that are made in

Finnish are dependent on the interrelations and oppositions that are available in

Finnish, they are not based on English meanings.

As suggested by Leino (1989: 196) there is a fairly clear distinction between

these clauses and possessive clauses in the grammatical organization of Finnish,

and this is reflected in the different case-endings: X:ssä on Y ‘in/at X is Y’ and

X:llä on Y ‘X has Y’. It is generally only when the first role is in the adessive that

it is natural to attach a possessive suffix to the second role:

(97) a. Laivalla on kapteeninsa.ship+ADE be+3SG captain+POS/3

‘A/the ship has her captain.’

b. ?? Laivassa on kapteeninsa.ship+INE be+3SG captain+POS/3

‘On the ship is her captain.’

The inessive example (97b) is from Nikanne (1990: 82), whose analysis is based

on Jackendoff’s theory of conceptual semantics and attempts to assess the limits

of grammaticality in language. I have put question marks in front of the example,

because it is problematic. Nikanne finds the example acceptable, and, moreover,

his translation differs from mine. He translates this as ‘the ship has her captain’

and claims that a circumstantial interpretation is not possible.

Since Nikanne’s example (97b) is problematic, I asked others whether they

thought it was acceptable Finnish. Many of my informants rejected it and said

that laiva ‘ship’ should be in the adessive. This reflects the intuition that if a pos-

sessive suffix is attached to the second nominal, one expects a possessive rela-

tionship, and if there is a possessive relationship the first nominal should be in the

adessive. However, Nikanne is correct in that a possessive suffix can be attached

to what I have called the Positioned. Hakanen (1972: 48) cites a number of exam-

ples taken from novels. Pace Nikanne, however, it seems to me that if a posses-

sive suffix is attached to a Positioned, the interpretation is circumstantial, i.e.

Page 263: systematic functional linguistic

255

something is located somewhere or is an integral part of something. This is evi-

denced by Hakanen’s examples:

(98) miehessä on muutakin kuin ulkopuolensa.man+INE be+3SG more+PAR+TIS than exterior+POS/3

‘There is more to a man than his exterior.’

(99) Ja keltaisessa postivaunussa oli oma miehistönsä.& yellow+INE post+van+INE be+PAS+3SG own+NOM crew+POS/3

‘And in the yellow mail van was its own crew.’

There is another argument that has been appealed to in claiming that exam-

ples like 96 above (repeated below as 100) are possessive.

(100) Veneessä on kaksi mastoa. [Hakulinen & Karlsson 1979: 97]boat+INE be+3SG two-NOM mast+PAR

‘There are two masts on the boat/The boat has two masts.’

The argument rests on the fact that vene ‘boat’ can occur as a genitive modifier in

an NP:

(101) veneen kaksi mastoaboat+GEN two-NOM mast+PAR

‘the two masts on/of the boat/the boat’s two masts’

The view has been put forward by Vähämäki (1980: 38; 1984: 310-330). As dis-

cussed in Chapter 3, however, the genitive is extensively used in Finnish (to mark

a dependent element) and it cannot be equated with a possessive (see 3.4.1 p. 89

ff.). Moreover, if we accept this argument then by the same token the following

clause is possessive.

(102) Helsingissä on kaksi katolista kirkkoa.Helsinki+INE be-3SG two-NOM Catholic+PAR church+PAR

‘There are two Catholic churches in Helsinki.’

By Vähämäki’s criterion this is possessive because we can say Helsingin kaksi

katolista kirkkoa ‘Helsinki’s two Catholic churches/The two Catholic churches in

Helsinki’.

Page 264: systematic functional linguistic

256

1 Material processes were earlier referred to by Halliday (in Kress (ed.) 1976:161) as actionprocesses; this earlier term has been retained by Fawcett (1980:138).

6.3.5. Summary

The process types discussed in section 6.3 under the heading of relational pro-

cesses have all been concerned with the setting up of a relationship: a relationship

between two entities, between an entity and a circumstance or between an entity

and a attribute. Intensive relational processes set up a relation of the kind ‘X is

(like) Y’. Inclusive relational processes set up an inclusive relationship ‘X 0 Y’.

Ambient relational processes set up a relation in which a temporal or spatial Cir-

cumstance is characterized: ‘at X is like Y’.

Finally, in circumstantial processes something is located in space or time: ‘X is at

Y’ or ‘at Y is X’, and, on this basis, circumstantial processes are considered to

belong to the superordinate category of relational processes. Possessive processes

are a subtype of circumstantial process in Finnish.

6.4. Material Processes

6.4.1. Introductory Comments

While a relational process constructs a relation, a material process1 constructs “a

happening” or “a doing”, which typically realize some kind of material change,

transition, or activity in the external world or in the world of our imagination. The

set of verbs capable of realizing material processes are far more varied and com-

plex than those that realize a relational process. I shall indicate only a few fairly

clear-cut subtypes that have been discussed in other descriptions of Finnish.

Material processes approximately subsume what other grammars have re-

ferred to as transitive and intransitive clauses. Hakulinen & Karlsson (1979: 93-

94) regard transitive and intransitive clauses as separate clause types in Finnish.

The transitive-intransitive dichotomy that is made by Hakulinen & Karlsson and

in many other grammars of Finnish stems from an ancient tradition that goes back

Page 265: systematic functional linguistic

257

to Apollonius (Robins 1967: 37). However, the validity of making this primary

division in clause types is not ensured by this long history. As Hakulinen &

Karlsson (1979: 94) themselves point out, the distinction is not clear-cut, and they

suggest that transitive and intransitive clauses could be combined into one basic

clause type (see also Leino 1991: 21 ff.). This suggestion will be adopted in this

study: transitive and intransitive clauses are conflated under the general category

of material process. However, despite this general agreement with Hakulinen &

Karlsson’s suggestion, there is a crucial difference between the description pro-

posed here and that suggested by Hakulinen and Karlsson. For them, a clause

such as Lapset ovat puistossa ‘The children are in the park’ is an intransitive

clause. As discussed in section 6.3.4, this is considered to be a circumstantial

relational clause in this study, and the transitive/intransitive distinction cannot be

applied to it. Moreover, processes related to human consciousness ) mental and

verbal processes (to be discussed in section 6.5) ) are distinguished from material

processes.

6.4.2. General Features of Material Processes

A (transitive or intransitive) material process in Finnish is minimally realized by

the following configuration of functions:

Actor @ Process:material (@ Goal)

The round brackets indicate that the Goal is optional. The Actor is an inherent

(and obligatory) role in all material process except in meteorological and

experiencer process (which will be discussed below). Material processes typically

construe a doing or a happening. They can be probed by the questions mitä

tapahtui? ‘what happened’, mitä X teki (Y:lle) ‘what did X do (to Y)?’.

Page 266: systematic functional linguistic

258

Before considering some of the grammatical features that clauses with mate-

rial processes have in common, it is important to note that material processes are

also defined in contrast to the other major process types. The over-riding feature

common to material processes is the fact that they differ from the relational pro-

cesses discussed in 6.3 and they differ from the mental and verbal processes that

are discussed in 6.5. Thus, for example, if there are two NPs in a material process,

the NPs would not agree in number, as is the case with intensive relational pro-

cesses. A feature that is shared by many mental processes is the fact that they

project: tiedän, että Eija lähtee ‘I know that Eija is leaving’. Projection is not a

typical feature of a material process.

The following clauses exemplify a material process without a Goal.

(103) juoksin .. hullun lailla [CA12:4]run+PAS+1SG .. mad+GEN like/way+ADE

Pro:mat+Ac (NR)‘I ran like a madman/like crazy.’

(104) Nainen ... kuoli välittömästi. [HS 11.10.91:A9]woman+NOM die+PAS+3SG immediatelyAc Pro:mat (NR)‘The woman died immediately.’

(105) kone lähtee kello seittemältä [CA3:13]plane+NOM leave+3SG clock-NOM seven+ABL

Ac Pro:mat Circ‘The plane leaves at 7 o’clock.’

(106) mä en bussil oo kertaakaa(n) menny [CA3:8]I-NOM NEG+1SG buss+A(DE) be+PTC time+TIS go+PTC

Aci Pro:÷+Aci (NR) ²Pro:÷ (NR) ²Pro:mat‘Not once have I gone by bus.’

Clauses 103 ) 106 exemplify the structure Actor @ Process:material, in spite

of the fact that none of the examples above are actually realized as a minimal

configuration. As discussed in 6.1.3, a minimal configuration cannot be deter-

mined on the basis of text examples. However, one can think of situations in

which a minimal configuration is possible with the material process exemplified

Page 267: systematic functional linguistic

259

in 103 ) 106, and it is only on the basis of our knowledge of what can count as a

minimal configuration that we can define a process type. For example:

(107) Mitä tapahtui?‘What happened?’

Hän kuoli. [Cf. 104 above.]‘She died.’

(108) Äkkiä. Kone lähtee. [Cf. 105 above.]‘Quick. The plane is leaving/about to leave.’

In some instances, nevertheless, a minimal configuration may seem far-fetched.

The following exemplify a material process with a Goal:

(109) mä avasin ulko-oven [AR:10]I-NOM open+1SG outside+door+GEN

Aci Pro:mat + Aci Go‘I opened the outside door.’

(110) mä sain yhen tosi surkee[n]I+NOM receive/get+PAS+1SG one+GEN real sad/wretched[GEN]

Aci Pro:mat+Aci Go

kuntosen rolfin [CA6:3]condition+ADJECTIVAL AFFIX+GEN Rolf+GEN

‘I got a Rolf (apparently a toy of some kind) in really appalling condition.’

(111) kaikki vois tuoda vähä safkaa mukanaall-NOM can+CON+3SG bring+INF some grub+PAR alongAc [Modality] Pro:mat Go (NR)‘Everyone could bring some grub (food) along.’ [Tel2:4]

(112) me ... rakennettiin lumilinnoja [SIIN3c:21]we build+INDE+PAS snow+castle+PL+PAR

‘we built snow castles (castles in the snow)’

(113) firma tekee kymmenien miljoonienfirm+NOM make/do+3SG ten+PL+GEN million+PL+GEN

Ac Pro:mat Go

tappiot [HS 19.11.89: A32]loss+NOM/PL

‘The/A firm makes losses running into tens of millions.’

Page 268: systematic functional linguistic

260

As indicated in these examples, if the Goal is realized by a nominal (or NP), it is

in the genitive (109, 110), partitive (111, 112) or nominative (113). The nomina-

tive and genitive refer to something that is bounded in space or time,

whereas with the partitive, it is non-bounded (for a fuller discussion, see

3.4.2 (pp. 94 ) 99)). A human pronoun is either in the accusative (bounded) or

partitive (non-bounded):

(114) sä voit vaikka tappaa mutyou-NOM can++2SG even kill I+ACC

Aci [Modality]+Aci Pro:mat Go‘you can even kill me’

tapa mut joku päivä [AR:53]kill-2SG/IMP I+ACC some+NOM day+NOM

Pro:mat Go Circ‘kill me some day’

The distinctions in boundedness outlined above are an important feature

shared by material processes. These boundedness distinctions are available in

material processes because they construct change, activity or transition. As indi-

cated above, the distinctions are realized in the case-marking of the NP realizing

the Goal, i.e. what is traditionally referred to as the object. Clauses with a mental

process are also considered to have an object, and, similarly the object can be in

the genitive nominative, partitive and accusative. However, as I shall argue in the

section on mental processes, the same semantic options are not available.

Another feature that is characteristic of almost all clauses realizing material

processes in Finnish ) whether transitive or intransitive ) is also related to the

semantics of change and activity. A clause with a material process can be ex-

tended by an NP that realizes spatial or temporal Extent. The NP realizing the

Extent patterns in a similar way to a Goal in Finnish, i.e. it is similar to an object:

Page 269: systematic functional linguistic

261

(115) Yön kuningatar kukkiinight+GEN queen+NOM blossom+3SG

Ac Pro:mat

yhden yön. [Example from Penttilä 1963:600]one+GEN night+GEN

Ex: temp

‘The midnight lily [a plant] blooms for one night.’

(116) (mä ..) juoksin vielä loppumatkan [TP1:1](I+NOM ..) ran+1SG still/yet/besides end+trip/way+GEN

Ac Pro:mat (NR) Ex: spat‘and what’s more I ran the rest of the way’

In example 115, the process of ‘blooming’ is given a temporal extent (‘one night’)

and in example 116 the process of ‘running’ a spatial extent (‘the rest of the

way’).

The NP exemplified by the underlined constituent in 115 and 116 is either

referred to as an “object-like modifier” (Finnish: objektin sukuinen määrite

(Penttilä 1963: 660; Siro 1964: 24; Ikola 1977: 142) or as “an adverbial of quan-

tity case-marked like an object” (objektinsijainen määrän adverbiaali) (see e.g.

Hakulinen & Karlsson 1979: 179, 216ff., Leino 1991: 181ff.). This is commonly

abbreviated as “OSMA” (from the Finnish (objektinsijainen määrän adver-

biaali). As indicated above and as pointed out by Leino (1991: 181), an OSMA

has a similar kind of case-marking (or variation in case-marking) as an object. For

example, a characteristic feature of an object in Finnish is that it is affected by the

polarity of the verb: if the polarity is negative, the object is typically in the parti-

tive (see 3.4.2, pp. 95 ) 96). This is also true of an OSMA:

(117) mä juoksin loppumatkanI+NOM ran+1SG end+trip/way+GEN

Ac Pro:mat Ex: spat‘I ran the rest of the way’

(118) mä en juossut loppumatkaaI+NOM NEG+1SG run+PTC end+trip/way+PAR

Ac Pro:mat Ex: spat‘I didn’t run the rest of the way’

Page 270: systematic functional linguistic

262

While there are some differences between the case-marking of objects and

OSMAs, the similarities are striking, and this point is often emphasised by Finn-

ish linguists (Leino 1991: 181).

Because of these similarities, the term OSMA (where the “A” at the end

stands for “adverbial”) is misleading in that an OSMA in Finnish is not really a

subtype of adverbial, but more like an object. In traditional Latin-based grammars

of Indo-European languages, the object is seen as the entity to which an action is

“passed over” or extended (see e.g. Robins 1967: 37). An “object-like modifier”

also realizes an extension of the action ) its extension in space or time. The fact

that an NP realizing the Extent of the process is related to the NP realizing the

Goal (i.e. the object in a material process) is also evidenced by the fact that there

is a fuzzy line between the two. This is particularly true of NPs realizing the spa-

tial Extent of a process. The Extent in example 117 is similar the underlined NP is

the following example, which could be considered a Goal:

(119) se on kolunnu ton kaukoidän(s)he+NOM be+3SG go through/over+PTC that+GEN Far-East+GEN

Ac Pro:mat Go

aika hyvin [CA3:3]quite well(NR)

‘She has gone over (travelled through) the Far-East quite well’

The fact that both Extent and Goal can be realized in the one clause is an

indication that they are separate functions:

(120) mä oon mussuttanu noita dominokeksejäI+NOM be+1SG overeat/stuff+PTC those+PL+PAR domino+biscuit+PL+PAR

Ac Pro:mat Goal

täs koko kesän ja syksyn [Tel2:6]here/this+INE whole (UF) summer+GEN & autumn+GEN

(NR) Extent

‘I’ve been scoffing those domino biscuits here all summer & autumn.’

Page 271: systematic functional linguistic

263

1 If a spatial Extent occurs with a verb that is typically associated with a relational or mentalprocess, I would argue that a material process is being construed.

Hakulinen & Karlsson (1979: 179) make a similar point: based on Siro (1964),

they argue that OSMAs and objects are distinct.

The function Extent can also be realized in a relational or mental process,

but in these instances, the process can only be extended temporally.1 The same

kind of case variation occurs. For example:

(121) a. Jussi oli siellä yhden yön.Jussi-NOM be+3SG/PAS there+ADE one+GEN night+GEN

‘Jussi was there for a night/Jussi spent a night there.’

b. Jussi ei ollut siellä yhtäkään yötä.Jussi-NOM NEG/3SG be+PTC there+ADE one+PAR+TIS night+PAR

‘Jussi wasn’t there for even a night.’

(122) Anna on ollut opettajana neljä vuotta.Anna-NOM be-3SG be+PTC teacher+ESS four-NOM year+PAR

‘Anna has been a teacher for four years.’

Thus, I am not claiming that a temporal Extent is specific to a material process,

but the notion of temporal extension is typical of a material process. When a tem-

poral Extent is added to a relational processes, the relation is construed as some-

thing with a temporal dimension.

Another feature shared by the majority of material processes is that they can

take a -sti adverb of manner that modifies the temporal dimension of the process:

X lähti/meni/tuli/käveli/kävi kiireisesti/nopeasti.

X left/went/came/walked/visited hurriedly/quickly

X kuoli/söi hitaasti

X died/ate slowly

These adverbs are of course possible with a relational process that constructs a

change, i.e. a resultative process such as meistä tulee nopeasti kuuluisia ‘we

Page 272: systematic functional linguistic

264

shall quickly become famous’, but it is certainly not typical of a stative intensive

process such as he ovat kuuluisia ‘they’re famous’. Mental processes, on the

other hand, commonly occur with a -sti adverb and this is a reflection of the fact

that many mental processes have a material aspect to them.

A final feature to be discussed concerns the use of the so-called 3rd infini-

tive in the inessive. The stem of a verb that typically realizes a material process

can occur with a nominal affix in the inessive in response to the question Missä X

on? ‘Where is X?’:

(123) Joku / jokin on X+ma+ssasomeone / something is VERB STEM + NOM. + INESSIVE

‘Someone/something is X:ing’

As mentioned in 6.3.4.(ii), a structure such as this realizes a circumstantial rela-

tional processes, and the meaning can generally be glossed as ‘some-

one/something is in the process of doing something’. For example, X on

ui+ma+ssa ~ juokse+ma+ssa jne. ‘X is [in the process of] swimming ~ running

etc.’ With transitive verbs, the process realized by the verb stem can be extended

by a Goal.

(124) Joku / jokin on X+ma+ssa Y + (t)asomeone / something is VERB STEM + NOM. + INESSIVE noun + PAR

‘Someone/something is X:ing someone/something.’

For example, X on maalaa+ma+ssa taulua ~ rakenta+ma+ssa taloa ~ os-

ta+ma+ssa autoa ‘X is [in the process of] painting a picture/building a

house/buying a car’ etc.

The structures given in 123 and 124 are unlikely if the stem is a mental pro-

cess: Missä X on? ‘Where is X?’ ) ??X on ajattelemassa ‘X is [in the process of]

thinking’, ??X on tykkäämässä Sailasta ‘X is [in the process of] liking Saila’, or

??X on vihaamassa Sailaa ‘X is [in the process of] hating Saila’. There are, of

Page 273: systematic functional linguistic

265

1 On the other hand, I would consider a verb like mietiskellä ‘to meditate’ to be a materialprocess. While it refers to a mental activity, the emphasis is on the activity, and for thisreason, it is construed like a material process.

2 Although this example has been used by Finnish linguists, it is perhaps more natural to sayPommi on räjähtämäisillään ‘The bomb is about to explode’.

course, mental verbs like pohtia ‘ponder’ that can be construed as a material

process: He ovat pohtimassa asiaa ‘They’re deliberating over the matter’.1

The meaning of the clause type exemplified in 123 above is also dependent

on the meaning of 1) the verb stem, 2) other derivational affixes attached to the

verb and 3) the subject. In SF terms ) where lexicogrammar is seen as a contin-

uum from highly abstract to more specific meanings ) the clause-type 123 above

can be given a general and abstract meaning, as indicated above, but more deli-

cate morphological and lexical choices determine its more precise meaning. Thus,

for example, with Pommi on räjähtämässä2 ‘The bomb is [in the process of]

exploding (i.e. it is about to explode)’ which realizes a material processes in

which the verb is an “achievement” verb ) in the sense of Dowty (1979) (based

on Vendler (1967)) ) i.e. it refers to something that happens instantaneously, the

clause refers to a process that is about to happen. (For further discussion, see

Heinämäki 1981, Brigden 1984: 193 ff., Maamies 1990: 85 ff.)

6.4.3. Subtypes of Material Process

I shall not attempt an exhaustive analysis of the various subtypes of material pro-

cesses. However, I shall consider in some detail a number of basic subtypes of

material process, which I shall refer to as (i) meteorological, (ii) experiencer, and

(iii) behavioural processes. These particular subtypes are not more important or

more frequent that other subtypes of material processes, but (a) meteorological

and experiencer processes represent clause types that have been discussed by

Finnish linguists and (b) experiencer and behavioural processes represent a fuzzy

area between material and mental processes. Both meteorological and experiencer

processes differ from other material processes in that an Actor is not obligatory.

Page 274: systematic functional linguistic

266

6.4.3.(i) Meteorological Processes

The verb in a meteorological process is always in the third person singular. There

is no Actor in the clause, and, thus, a meteorological process can be minimally

realized by the Process alone:

(125) Sataa.rain+3SG

‘It’s raining.’

However, like other material processes, they can be modified by a -sti adverb,

(e.g. sataa rankasti ‘It’s raining hard’) and they can be given a temporal Extent

(e.g. satoi koko viikon ‘It rained all week’).

It is also possible for a clause realizing a meteorological process to be ex-

tended by an NP, which is generally in the partitive (e.g. sataa lunta ‘rains

snow’). This NP has traditionally been referred to as the object. However, it dif-

fers from other Goals in material processes. According to Halliday (1985a: 104),

there are two types of material process with a Goal. One type is the dispositive,

i.e. something is done to another entity (as in sä voit vaikka tappaa mut ‘you can

even kill me’). The other type is the creative, i.e. the other entity is brought into

being by the process (e.g. me rakennettiin lumilinnoja ‘we built snow castles’).

The NP in a meteorological clause is neither an entity that something is done to

(dispositive) nor an entity created by the process (creative). It further specifies the

process: sataa ‘it’s raining’, sataa vettä ‘it’s raining (water), sataa lunta ‘it’s

snowing [snow is falling]’, sataa rakeita ‘it’s hailing [hail is falling]’, sataa rän-

tää ‘it’s sleeting [sleet is falling]’. I shall refer to them as Specifiers (of the pro-

cess).

(126) (Ulkona) sataa (lunta/vettä).outside+ESS rain+3SG snow+PAR/water+PAR

Circ Pro:mat:met Specifier‘It’s raining/snowing outside.’

Page 275: systematic functional linguistic

267

(127) Tuulee (kovasti). [NS]wind+3SG hard/strongPro:mat:met Circ:manner‘It’s windy/There’s a strong wind.’

(128) On pyryttänyt jo kolme päivää. [NS]be+3SG snowstorm+PTC already three day+PAR

Pro:mat:met (NR) Ext‘There’s been a snowstorm for three days now.’

These meteorological processes contrast with ambient processes discussed

in 6.3.3 (e.g. (Merellä) oli tuulista [(at sea) was windy] ‘It was windy at sea’).

Arguments against the conflation of ambient and meteorological processes were

presented in 6.3.3 and need not be repeated here.

6.4.3.(ii) Experiencer Processes

What I refer to as experiencer processes are also referred to as “causatives of feel-

ing” by Finnish linguists (e.g. Vilkuna 1989: 45). The finite verb in an experien-

cer clause is always in the third person, either singular or plural. The NP realizing

the Experiencer typically refers to a human being or to a part of the human body.

This NP is always in the partitive, and does not agree with the finite verb in num-

ber or person. In traditional Finnish linguistics, it is referred to as the object.

(129) mua ärsytti [TP1:35]I+PAR annoy+PAS+3SG

Experiencer Pro:mat:expI’m felt irritated/I felt annoyed.’

(130) a. polvea särkee, ...knee+PAR hurt+3SG

Experiencer Pro:mat:exp‘My knee hurts/aches’

b. päänahkaa kutittaa [HS 27.11.90:D10]scalp+PAR itch+3SG

Experiencer Pro:mat:exp‘My scalp itches/is itching.’

The NP realizing the Experiencer often (though by no means always) precedes the

verb.

Page 276: systematic functional linguistic

268

In spite of the fact that the finite verb often contains a causative affix

(-tta/ttä-) ) as in 129 and 130 above ) there is no Instigator in the clause, i.e.

nothing in the clause to indicate what is bringing about the process (of being an-

noyed, itching etc.) in the clauses above. However it is possible to add one, in

which case it has traditionally been referred to as the grammatical subject:

(131) mua uuvutti jo ajatuskin [AR:12]I+PAR exhaust/wear out+3SG already/even thought-NOM+TIS

Exp Pro:mat:exp Instigator

‘Just thinking about it wore me out/I was worn out by the thought of it.’

(132) mua ihmetyttää presidentti Koiviston käyttäytyminenI+PAR wonder+3SG president+NOM Koivisto+GEN behaviour+NOM

Exp Pro:mat:exp Instigator‘I wonder at/I’m surprised by President Koivisto’s behaviour.’ [S3/91: 61]

The fact that the final NP is the grammatical subject is evidenced by the fact that

in standardized written Finnish, there is agreement between a plural NP and the

finite verb.

(133) Minua ihmetyttävät monenlaiset asiatI+PAR wonder+3PL many+GEN+kind+NOM/PL thing+NOM/PL

Exp Pro:mat:exp Instigator

tässä maassa.this+INE country+INE

(NR)

‘I wonder at/I’m surprised by many things in this country.’

Experiencer processes are on the borderline between mental and material

processes. Many of the verbs that realize an experiencer process refer to human

feelings and responses, as indicated in the previous example. Other examples

include: (m(in)ua ‘I-PARTITIVE’) oudoksuttaa ‘to be astonished’, vituttaa ‘to be

pissed off’, huvittaa ‘to be amused’, hävettää ‘to be ashamed’ etc. Thus, expe-

riencer processes have some grammatical features of mental processes, as dis-

cussed in 6.5 below (i.e. variation in the case-marking of the NP realizing the

Experiencer is not possible) and it could be argued that they are a subtype of men-

tal process. However, I have grouped them with material processes as they are

Page 277: systematic functional linguistic

269

construed like internal activities, as evidenced by the fact that the finite verb often

contains a causative affix.

In the examples given so far, the NP realizing the Experiencer has preceded

the verb. It is also possible for it to follow the verb, e.g. Presidentti Koiviston

käyttäytyminen ihmetytti Moskovaa ‘President Koivisto’s behaviour surprised

Moscow’. This brings in the question of whether experiencer processes are simply

textual variants of material processes. While it may be valid to argue along these

lines, it seems to me that they can be regarded as a distinct sub-type on the basis

of the fact that the realization of an Instigator is not textually conditioned. The

fact that this process can be realized with only one inherent role is an inherent

characteristic of an experiencer process.

Hakulinen & Karlsson (1979: 98-99), in their analysis of what they also

refer to as “experiencer clauses” (Finnish: kokijalause), include not only causa-

tives of feeling, but also a number of subtypes that are grammatically quite differ-

ent from each other, but linked by the fact that the initial constituent is seen as

some kind of experiencer. Firstly, their analysis refers to the following as an

experiencer clause.

(134) Minulla on jano. [Hakulinen & Karlsson 1979: 98]I+ADE be+3SG thirst+NOM.Poss:er Pro:rel Poss:ed‘I’m thirsty [I’ve got a thirst]’.

In this study, clauses of this type are considered to be relational processes (pos-

sessive processes) (see 6.3.4.(iii)). Secondly, they include a number of fossilized,

idiom-like clause-types. In these the initial nominal is in the genitive and the verb

is always in the third person singular:

Page 278: systematic functional linguistic

270

(135) Minun käy sääliksi häntä.I+GEN go+3SG pity+TRA him/her+PAR

‘I feel (I’m inclined to feel) sorry for him/her.’

(136) Minun tekee mieli (syödä) jäätelöä ~ lähteäI+GEN do+3SG mind-NOM (eat+INF) icecream+PAR leave+INF

‘I feel like (eating) some icecream ~ leaving.’

The analysis of the these clauses with the initial nominal in the genitive is

problematic. As Hakulinen & Karlsson point out in a footnote, it is also possible

to regard some modal constructions in Finnish as experiencer clauses. In a partic-

ular type of modal clause referred to as a “necessitative construction” in tradi-

tional grammars of Finnish, the initial nominal is also in the genitive and the verb

is always in the third person singular:

(137) Minun täytyy aivastaa. [Hakulinen & Karlsson 1979: 99]I+GEN must+3SG sneeze+INF

‘I’ve got to sneeze.’

(138) Meidän täytyy kirjoittaa raportti.we+GEN must+3SG write+INF report+NOM

‘We’ve got to write a/the report.’

Hakulinen & Karlsson regard modal/necessitative constructions as having a

“genitive subject”. However, they point out that from the perspective of tradi-

tional grammar the initial NP could be analysed as a dative adverbial: it realizes

the entity at which the “having to” is directed.

In some respects, the genitive NP is like an Experiencer since, as Laitinen

(1988: 162-163) points out, it is not in control of the situation. However, this gen-

itive NP differs from the NP realizing the function Experiencer in at least two res-

pects. First of all, the case-marking is different: one is always in the partitive, the

other is always in the genitive. Secondly, the NP realizing the Experiencer is tra-

ditionally referred to as the object, and, as indicated above, a subject can be added

to the clause. A genitive NP, on the other hand, is not like an object; clauses with

Page 279: systematic functional linguistic

271

1 The case-marking of an object in a clause in which there is a genitive NP differs from“ordinary” clauses with a grammatical subject. As illustrated by 102 above, the object is inthe nominative rather than in the genitive.

2 This comparison is not meant to imply some kind of simple transformation. Moreover, thereare certain meanings that are generally always construed as non-volitional e.g. m(in)un täytyyaivastaa ‘I’ve got to sneeze’.

a genitive NP commonly contain an object, as illustrated by 135, 136 and 138

above.1

Examples such as 135 ) 137 are construed as non-volitional events that af-

fect the genitive participant. The participant is “controlled” by force of circum-

stance or by an automatic or instinctive response. The clauses contrast with the

following clauses:2

(139) (Minä) säälin häntä.I+NOM pity+1SG him/her+PAR

‘I pity him/her.’

(140) (Minä) syön jäätelöä.I+NOM eat+1SG icecream+PAR

‘I’ll eat icecream ~ I’ll have icecream’.

(141) (Me) kirjoitamme raportin.we+NOM write+1PL report+GEN

‘We’ll write a/the report.’

Rather than regard examples such as 135 and 136 as another subtype of experien-

cer clause, they can be regarded as expressions of modality in Finnish, if modality

is understood in the sense that it is used by Halliday (1985a: 85 ff., 334 ff; Halli-

day in Kress (ed.) 1976: 189-213) to refer to the semantics of personal participa-

tion. Halliday sees modality in terms of a number of scales: probability, usuality,

obligation and inclination. Of these scales, usuality and obligation are not usually

included in logically based analyses of modality (e.g. Lyons 1977b: 787 ff.).

Halliday’s scale of inclination appears to be particularly relevant to the grammar

of personal participation as realized in the examples above, although the scale

itself and its application would have to be re-thought for Finnish. Both examples

135 and 136 above could be seen in terms of inclination.

Page 280: systematic functional linguistic

272

In Hakulinen & Karlsson’s view, experiencer clauses also include the clause

type Minun on jano ‘I’m thirsty’, a slightly archaic variant of the possessive

clause in which the initial nominal is in the genitive. There are a number of other

common expressions like this which do not have an archaic flavour about them,

for example, M(in)un on hyvä olla ‘I feel good’, Mikä Leilan on? [what/which

Leila+GEN be-3SG] ‘What’s up with Leila?’. These do not display the same kind

of variation indicated above (minä) lähden ‘I’m leaving’ ~ Minun täytyy lähteä

‘I’ve got to leave’ and can be considered variants of possessive processes (see

6.3.4.(iii)).

6.4.3.(iii) Behavioural Processes

A sub-category of material process is a behavioural process, a doing or a happen-

ing that is a typically human activity, for example, puhua ‘to speak’, surra ‘to

grieve’, hymyillä ‘(to) smile’ or pelata ‘(to) play [a game]’, soittaa ‘(to) play (a

musical instrument)’, leikkiä ‘to play [in the sense of children’s play]’. Like

experiencer processes, behavioural processes can be regarded as being on the

borderline between mental and material processes.

The NP that is traditionally regarded as an object, however, is unlike a Goal

in an ordinary material process: if one says, for example, Anne puhui suomea

‘Anne spoke Finnish’ or Anne pelasi tennistä ‘Anne played/was playing tennis’,

then while it could be argued that Anne did something, the final NP is not some-

thing that is acted upon or created by the process. Halliday (1985a: 134 ff.) postu-

lates the function Range for English: “the Range is the element that specifies the

range or scope of the process”. As Halliday puts it, the Range defines the co-ordi-

nates of the process, so to speak. Thus, in the clause Anne pelasi tennistä ‘Anne

played tennis’, it is not a question of what Anne did to tennis, but what is indi-

cated is the scope or range of her playing. The following are examples of behav-

ioural processes in Finnish:

Page 281: systematic functional linguistic

273

(142) sä soitat kitaraa [SIIIM3b:5]you(SG) play+2SG guitar+PAR

Ac Pro:mat:beh Range‘You play the guitar.’

(143) mut kyl seki oli semmosta et -- asiakas tuli puhuun mulle ruotsia [TIIN3b:8]‘but then it was like -- a customer would come and speak to me in Swedish’

mä puhuin suomeaI-NOM speak+PAS+1SG Finnish+PAR

Ac Pro:mat:beh Range‘I spoke (in) Finnish.’

hän puhui ruotsia takas [SIIN3c:19](s)he-NOM speak+PAS+3SG Swedish+PAR backAc Pro:mat:beh Range (NR)‘(S)he’d speak back in Swedish.’

Behavioural processes are more like material processes in that they construe

activities in the external world, and like other material activities, they can be

probed by mitä X teki? ‘What did X do?’. In example 143 above, it would have

been possible to respond to asiakas tuli puhuun mulle ruotsia ‘a customer would

come and speak to me in Swedish’ by asking mitä sä sitten teit? ‘what did you do

then?’. Behavioural processes can be modified by an adverb of manner like

nopeasti ‘quickly’ or hitaasti ‘slowly’: puhua nopeasti ‘speak quickly’, soittaa

nopeasti ‘play [an instrument] quickly’ and it is not unusual for the verb stem to

occur in the inessive hän on puhumassa Annen kanssa ‘(s)he’s speaking with

Anne’.

On the other hand, behavioural processes have some features in common

with mental processes. As discussed in section 6.4.2, material processes display

distinctions in boundedness that are realized by the case-marking of the NP real-

izing the Goal. In the next section on mental processes (6.5), I shall argue that the

same options are not available in mental processes. This is also true of many

behavioural processes: the NP is typically in the partitive and, thus, there is no

variation in case-marking to distinguish between bounded and non-bounded

events: puhua suomea ‘to speak Finnish’, pelata tennistä ‘to play tennis’, surra

kuollutta ystävää ‘to grieve for one’s dead friend’ etc.

Page 282: systematic functional linguistic

274

6.5. Mental Processes

6.5.1. Internal and External, Verbalized Consciousness

Mental processes are concerned with the internal world of human consciousness,

and, in fact, another possible label for this process type is “processes of human

consciousness”. From a more philosophical perspective, of course, all of language

can be seen as being intimately linked with human consciousness. This section,

however, is concerned with the way in which human consciousness is realized in

language, with the way in which processes of saying, thinking, sensing, feeling,

perceiving, and reacting are construed in Finnish. Mental processes in this study

constitute a superordinate category that covers processes concerned with both the

internal consciousness of the human mind and external verbalized consciousness.

A similar superordinate category appears in Matthiessen’s (1989) description of

English. At a greater degree of delicacy, one would need to distinguish various

subtypes. For example, Matthiessen (1989) distinguishes mental, perceptive, reac-

tive and verbal processes in English.

As stressed throughout this study, however, the setting up of a category,

such as a mental process, means that we need to be able to distinguish it gram-

matically. At first sight, mental processes may seem to comprise a completely

heterogeneous set of process types that have nothing in common grammatically,

which puts into question the postulation of a superordinate process type. The fol-

lowing are examples of what I consider to be mental processes in Finnish. In the

glosses, I make only a broad distinction between verbal, on the one hand, and

mental, reactive and perceptive, on the other. The inherent (human) role is re-

ferred to as a Sayer in a verbal process and a Senser in a mental process. Since

mental processes construe human consciousness, the Senser or Sayer is typically

realized by an NP that refers to a conscious being. Following Halliday’s (1985a)

analysis of English, the role of the NP that refers to what is said will be labelled

the Verbiage and the role of the NP that realizes what is sensed, perceived or re-

acted to will be referred to as the Phenomenon.

Page 283: systematic functional linguistic

275

(144) kuulet satakielen laulun pihapuustahear+2SG nightingale+GEN song+GEN yard/garden+tree+ELA

* Pro:men + Sen* Phen *

‘you can hear the song of a nightingale in (/coming from) a tree in the garden’

ja näet ikkunasta hirviperheen uivansee+2SG window+ELA elk+family+GEN swim+GEN

& * Pro:men + Sen * Circ * Phen

lahden poikki [HKV]bay/inlet+GEN across

*

‘and from your window you can see a family of elk swimming across the inlet.’

(145) Mä rakastan sua. [AR:38]I-NOM love+1SG you(sg)+PAR

Seni Pro:men + Seni Phen‘I love you.’

(146) en mäkä(än) semmosia talojaNEG+1SG I-NOM+TIS such kind+PL+PAR house+PL+PAR

Pro÷ + Seni * Seni * Phen *

ole nähny [TP1:14]be+PTC see+PTC

²Pro÷ ²Pro:men

‘I haven’t seen houses like that either.’

(147) mä tykkään Haikaran pesästä [TIIIN3d:7]I+NOM like+1SG Stork+GEN Nest+ELA

Seni Pro:men + Seni Phen

‘I like Stork’s Nest (a restaurant).’

(148) ja sit mä sanoin ** kui moneltand then I-NOM say+PAS+1SG how many+ABL

Sayi Pro:men + Sayi ** Projection

mä pääsen [TP1:12]I-NOM get out/into+1SG

(Material process)**

‘and then I said when (at what time) I’d be getting out/finishing’

Page 284: systematic functional linguistic

276

6.5.2. Defined by Differences

The grammatical features shared by mental processes could be regarded as being

more covert than the grammatical features that are shared by the other major pro-

cess types ) even if we still work with the notion of a prototype. However, recent

discussions of randomness and order in chaos theory have made us aware that

order and organization can not be seen in simplistic terms. As Butt (1988) points

out, chaos theory has sensitized us to the subtlety of possible orders. What links

many of the processes that I refer to here as mental processes is the fact they are

grammatically different from material processes.

Hakulinen & Karlsson (1979: 94) refer to a set of “semi-transitive verbs” in

Finnish. These are verbs that realize a process in which the second inherent role,

unlike the Goal discussed above, is not realized by an NP in the nominative, par-

titive or genitive, but in one of the locative cases (see Figure 3-7, p. 84). Some of

the verbs mentioned by Hakulinen & Karlsson realize what are being referred to

here as mental processes:

(i) pitää ‘(to) like’ + jostakin ‘someone/something+ELA’varmistua ‘be certain of’ + jostakin ‘someone/something+ELA’

(ii) kyllästyä ‘get/be fed up’ + johonkin ‘someone/something+ILL’rakastua ‘fall in love with’ + johonkin ‘someone/something+ILL’

(iii) säästyä ‘be spared’ + joltakin ‘someone/something+ABL’välttyä ‘avoid, get out of’ + joltakin ‘someone/something+ABL’

These semi-transitive verbs can be characterized by the fact that they are different

from transitive verbs: in traditional terms, they do not take an object:

(149) mä pidän ElviksestäI-NOM like+1SG Elvis+ELA

Seni Pro:men + Seni Phen‘I like Elvis’

(150) mä rakastuin ElvikseenI-NOM love+PAS/AUT+1SG Elvis+ILL

Seni Pro:men + Seni Phen‘I fell in love with Elvis’

Page 285: systematic functional linguistic

277

1 While many reactive verbs (e.g. rakastua ‘to fall in love’, kyllästyä ‘get/be fed up’) can beregarded as involving a change in the state of the person undergoing the process, they are notconstrued as material changes in the external world, i.e. changes to which aspectualdistinctions and distinctions in total and partial quantity can normally be applied.

Thus, the choice between the partitive and the other grammatical cases is not

available and distinctions in boundedness cannot be made. The fact that these

verbs do not take an object seems highly significant: mental processes are not

constructed as processes that involve a change or an activity in the external, ma-

terial world.1

However, there are many verbs that refer to saying, thinking, sensing, feel-

ing, perceiving, and reacting in Finnish, and yet they “take an object”. These in-

clude verbs like tietää ‘(to) know [a fact]’, tuntea ‘(to) know [a person]’, rakas-

taa ‘(to) love’, nähdä ‘(to) see’, ajatella ‘(to) think’, unohtaa ‘(to) forget’, muis-

taa ‘(to) remember’, haluta ‘(to) want’ etc. Many of these differ from material

processes in much the same way as these processes in which there is a “semi-tran-

sitive” verb. The Phenomenon in these instances is typically (and often almost

invariably) either in the partitive or in the nominative/genitive/accusative:

(151) Mä ajattelen ElvistäI-NOM think+1SG Elvis+PAR

Seni Pro:men + Seni Phen‘I’m thinking about Elvis

(152) Mä muistan Elviksen / hänet.I-NOM remember+1SG Elvis+GEN / he+ACC

Seni Pro:men + Seni Phen‘I remember Elvis/him’.

Nevertheless, there are other verbs that have features of both mental and

material processes in Finnish. The verb rakastaa ‘(to) love’, for example, typi-

cally construes a non-bounded process: the Phenomenon is almost always real-

ized by an NP in the partitive.

Page 286: systematic functional linguistic

278

(153) Mä rakastan sua. [AR:38]I-NOM love+1SG you(sg)+PAR

Sen Pro:men + Sen Phen‘I love you.’

However, the verb rakastaa ‘(to) love’ can also realize a material process in the

following ) rather unusual ) type of context:

(154) rakastan sinut kuoliaaksilove+1SG you(sg)+ACC to death‘I love you to death’ (the name of an American film)

(155) rakastin sinut rappiollelove+1SG you(sg)+ACC destruction+ALL

‘I loved you to destruction’ (from the Finnish poet Eino Leino).

Thus mental processes can be distinguished by the fact that the Phenomenon

or Verbiage is not case-marked like an object in a material process: either the NP

does not display the same kind of variation in case-marking as an NP realizing the

Goal in a material process or it is realized by an NP in one of the locative cases.

Mental process can also be negatively defined with respect to some features of

material processes: for example, the stem of a verb that typically realizes a mental

process is unlikely to occur in the syntagm Hän on ___ massa/mässä ‘(S)he is

(in the process of) ___ ing’ in response to the question Mitä hän tekee ‘What is

(s)he doing?’. Another significant feature associated with mental processes,

which distinguishes them from material processes, is that they can project. Projec-

tion will be discussed in the next section.

6.5.3.

6.5.4. Human Consciousness and Projection

As discussed in 4.5.5 (p. 159 ff.), projection in SF theory does not only refer to

what is traditionally labelled direct and indirect speech. It refers to the fact that a

clause or a non-finite clause ) instead of being a direct representation of reality )

is at a further remove from the reality. It seems natural that a category of mental

processes ) processes of human consciousness ) allow us to reflect on reality

through the resources of projection.

Page 287: systematic functional linguistic

279

From the point of view of distinguishing mental processes as a superordinate

category, it is not so much the distinction between Quotes and Reports or be-

tween Ideas and Locutions that is important in Finnish, since this kind of projec-

tion is only part of the picture. The discussion on projection in Chapter 4 focused

on projection as a relationship between clauses, but another important notion that

was also introduced in Chapter 4 is the notion of embedding, where one clause

functions in the structure of a host clause. Thus projection can either be a relation-

ship between clauses at the rank of clause, or it can involve a relationship, where-

by one clause functions in the structure of another clause.

Before discussing embedded projection in Finnish, a set of distinctions use-

ful in the analysis of mental processes will be discussed. Matthiessen (1989: 85)

distinguishes various types of Phenomena in mental processes: a simple phenom-

enon, a macrophenomenon and a metaphenomenon. These different classes of

realizing units will be exemplified in turn. A simple Phenomenon is an NP, as in

the first clause of example 144 and in 147 above:

PHENOMENON:

kuulet satakielen laulun (See example 144 above.)Pro:men+Sen Phen‘you can hear the song of a nightingale’

mä tykkään Haikaran pesästä (See example 147 above.)Seni Pro:men+Seni Phen‘I like Stork’s Nest’

A macrophenomenon (or a composite phenomenon) is non-finite clause function-

ing as phenomenon. This is illustrated by the second clause in 144 above:

Page 288: systematic functional linguistic

280

MACROPHENOMENON:

näet hirviperheen uivan lahden poikkisee+2SG elk+family+GEN swim+GEN bay/inlet+GEN acrossPro:men+Sen Phen‘you can see a family of elk swimming across the inlet’

These macrophenomena are important in Finnish and correspond to what Finnish

linguists refer to as “clause equivalents” (lauseenvastike). They will be discussed

more fully below.

By contrast, a metaphenomenon is a projected clause. With metaphenomena,

the clause can either be 1) part of a clause complex where the mental process is

realized in the projecting clause and the projected clause realizes the phenome-

non, or else 2) the clause is embedded. The first instance can be illustrated by the

following examples:

METAPHENOMENON:

(156) mä tiedän et mun täytyyI+NOM know+1SG that I+GEN must+3SG

Seni Pro:men+Seni Phen2 PROJECTING CLAUSE 2 PROJECTED CLAUSE

aamuyöstä nousta [CA3:10]morning+night+ELA rise/get up+INF

2

‘I know I have to get up in the early hours of the morning.’

(157) mä sanoin sille sano munI-NOM say+1SG (s)he/it+ALL say+IMP+2SG I+GEN

Sayer Pro: verbal Receiver Pro:verbal Ver

2 PROJECTING CLAUSE 2 PROJECTED CLAUSE

sanoneen, sulta loppuu pohjavedet [OH]say+PTC+GEN you+ABL finish/end+3SG ground+water+NOM/PL

2 2

‘I told him “Remember my saying [say that I’ve told you] ‘your ground water will dry upon you’”.

Page 289: systematic functional linguistic

281

This last example is further complicated by the fact that the second clause con-

tains a nonfinite clause, a macrophenomenon, mun sanoneen ‘my saying’, which

contains a non-finite form of the verbal process sano- ‘say’, which, in turn, pro-

jects another clause, sulta loppuu pohjavedet ‘your ground water will dry up on

you’.

As mentioned above, a projected clause can also be embedded. This is an

important resource in Finnish, and here the embedded clause is often part of an

NP that realizes a Fact (see 4.5.5, p. 161 ff.).

(158) mä ... unohdin [... interpolation ...] [Tel3:9]I-NOM forget+PAS+1SG

Senser Pro:men

sen Û että mun piti tuoda se puhelinkone sinne âit/that+GEN that I was supposed to take the telephone receiver therePhen (Fact)

‘I forgot that I was supposed to take the telephone receiver there.’

(159) mä oon ajatellu just sitä et se I-NOM be+1SG think+PTC precisely it+PAR that it+NOM

Senser Pro:men Phen (fact)

ois kauheeta kun nukkus ohi [CA3:12]be+CON(3SG) awful+PAR when sleep+CON(3SG) past

‘I’ve thought precisely that ) that it would be awful to sleep in’ (The context here is: itwould be awful to sleep in if you had to go on a trip.)

The English translations of these obscure the fact that the Phenomenon is realized

by a pronoun followed by an embedded clause.

As mentioned above, so-called clause equivalents are a significant feature of

mental processes in Finnish. The Phenomenon (or Verbiage) is realized as a

macrophenomenon, a composite phenomenon with a non-finite verb form as

Head. One type of macrophenomena is realized by what Ikola (1974) has referred

Page 290: systematic functional linguistic

282

1 This is a somewhat misleading term, as many of the verbs that occur in the “matrix clause”are difficult to imagine in terms of a report or a quote.

to as a “reporting clause [or sentence] equivalent” (referatiivinen lauseenvas-

tike).1 The following are example of macrophenomena in Finnish:

(160) sano mun sanoneen .. [see 157 above]say+IMP+2SG I+GEN say+PTC+GEN

Pro:Ver Verbiage (macrophenomenon ` non-finite clause)‘Remember my saying .. ‘

(161) jos yhä vain pelkää kuntonsa rapistuvanif evermore frighten+3SG condition+POS/3 deteriorate+PTC+GEN

Pro:men Phen (macrophenomenon ` non-finite clause)

‘if you’re increasingly afraid that your condition is deteriorating (that you’re not gettinginto better shape)’

kannattaa mennä kalliomaastoon sieneen. [HKV]it’s worth your while to go picking mushrooms in rocky terrain.’

(162) he kokevat olevansa täälläthey+NOM/PL experience+3PL be+PTC+GEN here+ADE

Sen Pro:men Phen (macrophenomenon ` non-finite clause)

vähän parempia kuin muut [HKV]a bit good+COMP+PL+PAR than rest+NOM/PL

‘Here (in Finland) they feel that they’re better than the rest [they experience a feeling ofbeing better than the rest].’

(163) Telegin huomasi heti törmänneensäTelegin+NOM notice+PAS+3SG immediately bump into+PTC+GEN+POS/3

Sen Pro:men (NR) Phen (macrophenomenon `

lahjakkuuteen. [HKV]giftedness/talent+ILL

non-finite clause)

‘Telegin immediately realized that he had bumped into an exceptionally gifted person.’

These “clause equivalents” are also referred to as “participial constructions”,

which is a more appropriate label.

These macrophenomena also include impersonal projections (i.e. where the

Senser is not explicated):

Page 291: systematic functional linguistic

283

(164) ohjelmaa tuntuu aina olevan niin paljon [CA3:5]programme+PAR seem+3SG always be+PTC+GEN so muchPhen ÷ Pro:men ²Phen (macrophenomenon ` non-finite clause)‘There always seem to be so many things to do (too much on one’s agenda).’

The finite verb here tuntua ‘to seem/feel’ was one of the verbs included among

sensory attributive processes. Like näyttää ‘look/seem/appear’ and vaikuttaa

‘seem’, which were also included among sensory attributive processes, it often

realizes a mental process. In fact, it could be argued that the sensory processes

outlined in 6.3.1.(iii) are, in some respects, like relational processes and in other

respects like mental processes.

As examples 160 ) 163 illustrate, the projecting verb includes not only

verbs of saying and thinking, but also other verbs related to human experience,

e.g. kokea ‘to experience’, pelätä ‘to be afraid’. It need not necessarily refer to

what “someone has said or thought”; as pointed out in 4.5.5 (p. 159), projection

in general should not be understood in simplistic terms as referring to what

“someone has said or thought”. The following verbs occurred with a participial

construction in the HKV corpus (see appendix 2):

ajatella ‘to think’, arvata ‘to guess’, arvioida ‘to guess/estimate’, haluta ‘to want’, huoma-ta ‘to notice’, ilmoittaa ‘to announce’, julistaa ‘to proclaim’, katsoa ‘to see/regard’, kertoa‘to tell/relate’, kiistää ‘to deny/argue against’, korostaa ‘to stress’, kuvitella ‘to imagine’,luulla ‘to think/feel’, kokea ‘to experience’, kuulla ‘to hear’, muistaa ‘to remember’, myön-tää ‘admit/concede’, naureksia ‘to laugh’, nähdä ‘to see’, odottaa ‘to expect’, olettaa ‘toassume’, osoittaa ‘to indicate’, otaksua ‘to assume, suppose’, painottaa ‘to stress’, pelätä‘to be afraid’, povata ‘to forecast’, sanoa ‘to say’, sopia ‘to agree’, suunnitella ‘to plan’,tietää ‘to know’, todeta ‘to confirm, verify’, todistaa ‘to prove’, tulkita ‘to interpret’,tunnustaa ‘to confess’, tuntea ‘to feel’, uskoa ‘to believe’, väittää ‘to claim’.

I am not suggesting that these are the only verbs that realize mental processes, nor

that these verbs always realize a mental process. This list is simply to give some

indication of the kind of verb that is likely to realize a mental process.

Page 292: systematic functional linguistic

284

6.6. Experiential Metaphors

The major process types outlined above present the bare outlines of experiential

structures in the Finnish clause: each process type needs to be analysed in much

greater delicacy. However, the outline given above provides some indication of

the experiential meaning potential at clause rank in Finnish. While the

lexicogrammatical resources of a language enable us to make certain meanings,

language as a system is not a static reservoir of pre-determined meanings. A per-

vasive semiotic resource in language is metaphor.

Metaphor in experiential structures can either be lexical or grammatical. The

notion of lexical metaphor is familiar from literary studies. Grammatical meta-

phor, on the other hand, is a far more subtle form of metaphor, which has only

recently begun to be noticed and described. As argued in Chapter 5, it is based on

the notion of a “baseline”, a baseline for realizing interpersonal and experiential

functions. With experiential functions, this means inter alia that processes (and

relations) at clause-rank are realized by verbs or VPs and that inherent and non-

inherent roles (participants and circumstances) are realized by NPs, PPs or AdvPs

in Finnish. These baseline realizations are referred to as congruent realizations.

Like lexical metaphors, grammatical metaphors gradually lose their meta-

phorical force, and become unmarked forms of encoding in a language. Even if

metaphorical encodings are unmarked, if congruent forms exist in the language,

then the choice between a congruent and a metaphorical realization represents

another set of options in the language in question. Examples of fairly unmarked

grammatical metaphor in Finnish are found when verbs like käydä ‘(to) go, func-

tion etc.’ and käyttää ‘(to) use’ occur with a Range that refers to a verbal act: e.g.

käydä neuvotteluja [verb + negotiation+PL+PAR] ‘to negotiate, to have negotia-

tions’; käyttää puheenvuoro [verb + speech+turn] ‘to speak (publicly on a

topic)’. As Halliday points out (1985a: 327), in expressions like these, the “verb

simply expresses the fact that some process takes place, and carries the verbal

Page 293: systematic functional linguistic

285

categories of tense, polarity and so on, while the process itself is encoded as a

nominal group functioning as a Range”.

Often grammatical metaphor is tied up with the textual “packaging” of infor-

mation:

(165) Opiskelija Jussi Virtanen, 21, katosi lauantaina vastaisena yönä 7. syyskuuta VikingLinen risteilijä Mariellalta.

Jussi Virtanen, a student, 21, disappeared during the early hours of Saturday morning,7. September, on the Viking Line cruiser Mariella.

÷ Katoamisen on arvioitu tapahtuneendisappearance-GEN estimate-PER/INDEF happen+PTC+GEN

kello kolmen ja neljän välillä. [HS 11.9.91:A9]between three and four o’clock.

‘It is estimated that the disappearance happened between three and four o’clock.’

This example is further complicated by an embedded projection: there is a fact

projected by the verb arvioida ‘to estimate’. In an agnate finite clause, the fact

would have been katoaminen tapahtui kello kolmen ja neljän välillä ‘the dis-

appearance happened between three and four o’clock’. In unselfconscious spoken

Finnish, if someone has disappeared, it is highly unlikely that one would say

katoaminen tapahtui kello kolmen ja neljän välillä ‘the disappearance happened

between three and four o’clock’. A more congruent way of saying this would be,

for example: (Ne arvioi että,) Jussi katos kolmen ja neljän välillä ‘(They esti-

mate that) Jussi disappeared between three and four o’clock’. However, the above

text appears to be from a police report, and police reports are not concerned with

particular people but with events such as murders, disappearances and robberies.

Grammatical metaphor also has the effect of objectifying and distancing the

author from the event (cf. Eggins et al. 1987, Martin 1991). Something like a

murder or a disappearance, which in the extralinguistic world is a process )

something that happens in time rather than a concrete entity ) is constructed or

Page 294: systematic functional linguistic

286

packaged as a nominal. For this reason, grammatical metaphor is common in sci-

entific texts: in the western understanding of science, at least, in order to study

something, one has to objectify and reify it (cf. Keller 1985). The above example

of metaphor is rather complex since it also involves a projected fact. The follow-

ing examples of grammatical metaphor in Finnish illustrate how metaphorical

expressions need not be grammatically complex. Example 166 (a ) c) is a frag-

ment of text from a newspaper article:

(166) a. Konservatiivisten voimien vaikutus NL:nconservative+PL+GEN force+PL+GEN influence USSR+GEN

* Ac

talouselämässä heikkeneeeconomy+life+INE weaken+3SG

* Pro:mat *

‘The influence of conservative forces in economic circles in the USSR is weakening’

b. ja reformiajattelu voimistuu.& reform+thinking+NOM strengthen+3SG

Ac * Pro:mat *

‘and revisionist thinking is gaining ground.’

c. Murrosvaiheen kivikot ovat kuitenkincrisis/transition+stage+GEN stony patch+NOM/PL be+3PL neverthelessPositioned Pro:rel (NR)

vielä edessä [[ talousuudistuksen tietästill in front+INE economy+renewal+GEN road+PAR

(NR) Circ:temp (NR)

kuljettaessa. ]]travel+INDEF+INF+INE

[HS 1.9.91:A2]

‘The roughest patches in the transitional stage are, nevertheless, still ahead as theytravel along the road to economic reform (~ the road to economic reform is travelledalong).’

Page 295: systematic functional linguistic

287

(167) Antropologian ja naistutkimuksen välinenanthropology+GEN & female+research+GEN intermediary+NOM

* Ac

luova jännite syntyy molempiencreate+PTC/NOM tension+NOM be born+3SG both+PL+GEN

* Pro:mat * Circ

kiinostuksesta toisaalta erilaisuutta,interest+ELA the one side+ABL differentness+PAR

toisaalta samanlaisuutta kohtaan. [Y 14/91:9]the other side+ABL sameness/similarity+PAR toward

*

‘The creative tension between anthropology and feminist studies is created by an interestin both disciplines in sameness (similarity), on the one hand, and differentness (differ-ence), on the other hand.’

Examples 166a, 166b and 167 have a fairly uncomplicated grammatical pattern in

which the finite verb realizes a material process. Yet, to unpackage the metaphors

and put them in a more congruent form that could be understood by a child would

be quite a feat. The second sentence in example 166 (clause c) continues in the

same metaphorical vein; however, it is not an example of grammatical metaphor

but of lexical metaphor. It is slightly more complicated in that it includes an em-

bedded non-finite clause.

Grammatical metaphor as illustrated here is, of course, typical of written

language. As Halliday (1987: 149) has pointed out, written language gives us

another perspective on experience: a synoptic one. According to Halliday (1987:

148):

Writing puts language in chains; it freezes it, so that it becomes a thing to be reflected on.Hence it changes the ways that language is used for meaning with.

Writing allows us to package information concisely, and this is important in the

accumulation and development of knowledge. Crucial in packaging are

nominalization and grammatical metaphor. As Halliday puts it:

Page 296: systematic functional linguistic

288

1 However, when complex grammatical metaphor occurs in a text that is meant for the non-initiate, then it is self-defeating, if this is used as the principal means of initiation into asubject (see, for example, Karvonen’s (1991, 1992) analysis of Finnish textbooks for school-children).

Until information can be organized and packaged in this way ) so that only the initiate under-stands it ) knowledge cannot accumulate, since there is no way one discourse can start wherethe other ones left off. When I can say

the random fluctuations in the spin components of one of the two particles

I am packaging the knowledge that has developed over a long series of preceding argumentsand presenting it as “to be taken for granted ) now we can proceed to the next step”. If Icannot do this, but have to say every time that particles spin, that they spin in three dimen-sions, that a pair of particles can spin in association with one another, that each one of the pairfluctuates randomly as it is spinning, and so on, then it is clear that I will never get very far.

Nominalization and grammatical metaphor are resources that allow us to take

information for granted, so that we can build on the knowledge that has already

been accumulated. As Halliday (1987: 151) says, this “shuts the layman out”.

Shutting the layman out is a necessary part of specialist discourse.1

6.7. Macro-Roles in Finnish

6.7.1. Medium and Domain

Halliday (1985a: 144 ff.) discusses English in terms of the generalized functions

(or macro-roles) Medium and Agent. The macro-roles that I postulate for Finnish

are Medium and Domain. I use cursive script to underscore the fact that these

macro-roles are at a higher level of abstraction than the roles discussed earlier in

this chapter. Thus both Medium and Domain can be seen as generalized catego-

ries that are further specified in a particular lexicogrammatical environment. The

Medium in Finnish subsumes the Carrier or Identified in an intensive relational

process, the Located in a circumstantial relational process, the Actor in a material

Page 297: systematic functional linguistic

289

1 In an identifying clause where both NPs are in the nominative (e.g. KansallisteatterinMarkiisitar on Eeva-Kaarina Volanen ‘The National Theatre’s marchioness is Eeva-KaarinaVolanen’) there are no grammatical criteria that can be appealed to in determining which NPis the subject. On the other hand the question of which NP is the subject (and thus which NPrealizes the Medium is immaterial in such a clause since the point of the clause is to equatethe NPs. Cf. Kelomäki (1988: 67 ff.).

process, and the Sayer or Senser in a mental process. The realization of the Me-

dium is complex in Finnish, and will be discussed more fully in Chapter 7. Suf-

fice to say here that it can be realized by what has traditionally been referred to as

the grammatical subject or by a bound morpheme (personal ending) attached to

the verb. For example:

(168) AkiraKurosawa -- on japanilainen elokuvaohjaaja. [W5:3374]AkiraKurosawa-NOM be+3SG Japanese movie+director-NOM

Car (Medium) Pro:int Att‘Akira Kurosawa is a Japanese motion picture director.’

(169) Olen Henna Partanen.1 [PL 1]be+1SG Henna Partanen-NOM

Pro:int + Id (Medium) Ir‘I’m Henna Partanen.’

(170) Ritvahan on naimisissaRitva+TIS be+3SG marriage+INE

Posit:ed (Medium) Pro:rel Circ‘Ritva’s married of course’

(171) juoksin .. hullun lailla [CA12:4]run+PAS + 1SG .. mad+GEN like/way+ADE

Pro:mat + Ac (Medium) (NR)‘I ran like a madman/like crazy.’

(172) mä tykkään Haikaran pesästä [TIIIN3d:7]I+NOM like + 1SG Stork+GEN Nest+ELA

Seni (Medium)Pro:men + Seni (Medium) Phen‘I like Stork’s Nest (a restaurant).’

(173) syötkö hernekeittoa [CA2:27]eat+2SG+Q pea+soup+PAR

Pro:mat + Ac (Medium) Range ‘Do you eat pea soup?’

According to Halliday (1985a: 146) in his analysis of English, the Medium

is the “participant that is the key figure in that process; .. the one through which

the process is actualized, and without which there would be no process at all”. In

Page 298: systematic functional linguistic

290

1 Finnish: ‘piiri, ulottuvuus’.

English, the Medium is realized either by the grammatical subject of an intransi-

tive verb or by the object. As discussed below, this kind of division is not central

in the grammatical organization of Finnish. If we apply Halliday’s characteriza-

tion of Medium to Finnish, then the most likely candidate would be the experien-

tial function that is realized by the grammatical subject or by a bound morpheme

(personal ending) on the finite verb. If the Medium in Finnish is characterized as

“the participant that is a key figure .. without which there would be no process at

all”, then there is no better grammatical evidence for this function than the fact

that it can either be realized by a morpheme that is part of the finite verb or else it

is cross-referenced on the verb: mä tykkää+n ‘I like’, kone lähte+e ‘the plane is

leaving’, vieraat lähte+e/lähte+vät ‘the guests are leaving’.

As discussed in 3.4.3 (pp. 99 ) 100), the grammatical subject as defined in

this study is the NP that agrees with the verb in person (and generally also in

number). What I am proposing, thus, is that the grammatical subject is one real-

ization of the function Medium. The Medium can be realized by a grammatical

subject (m(in)ä ‘I’, kone ‘plane’, vieraat ‘guests’) or it can be realized as an affix

(personal ending) attached to the verb stem: -n (1sg), -t (2sg), -mme (1pl), -tte

(2pl). These first and second person morphemes are not subjects. In fact, the

grammatical subject is not a particularly central category in Finnish. As discussed

in 5.4.2, there are many subjectless clause types in Finnish and grammatical fea-

tures that are associated with the grammatical subject in English at least (e.g.

control of a reflexive suffix) are not confined to the subject in Finnish (see A.

Hakulinen 1983).

The other macro-role that is proposed for Finnish will be referred to as the

Domain.1 The Domain subsumes the Goal in a material process, the Experiencer

in an experiencer process, the Range in a behavioural process, the Qualifier of the

process in a meteorological process, the Phenomenon in a mental process, and the

Verbiage in a verbal process:

Page 299: systematic functional linguistic

291

1 This form of mä ‘I’ is Häme dialect.

(174) syötkö hernekeittoa [CA2:27]eat+2SG+Q pea+soup+PAR

Pro:mat + Ac GoalProcess + Medium Domain‘Do you eat pea soup?/Are you eating pea soup?’

(175) mua ärsytti [TP1:35]I+PAR annoy+PAS+3SG

Experiencer Pro:mat:expDomain Process‘I felt irritated/I felt annoyed.’

(176) Mää1 ymmärrän sua. [AR:188]I-NOM understand+1SG you+PAR

Seni Pro:men + Seni PhenMedium Process + Medium Domain‘I understand you.’

The Domain refers to the sphere of influence of the process: the entity that is

affected by the process or the entity to which the process is extended. The NP

realizing the Domain does not agree with the verb in number or person. It is in

one of the grammatical cases (see Figure 3-7, p. 84): the nominative, genitive or

partitive case. Human pronouns are either accusative or partitive.

The Domain as outlined so far covers what is traditionally referred to as the

object. It could also be argued that what is traditionally referred to as the “exis-

tential subject” in Finnish also realizes the Domain of the process. While this

“existential subject” is referred to as a subject, it patterns more like an object.

This point is discussed in greater detail in the following section.

Page 300: systematic functional linguistic

292

1 It appears to be prototypical in the sense that it corresponds to an existential clause inEnglish, Swedish and French.

6.7.2. The Problem of the “Existential Subject”

The discussion in this section is based on an extended debate on what Finnish

linguists have referred to as existential clauses in Finnish (eksistentiaalilause)

(see e.g. Ikola 1954; Siro 1964: 49 ff.; Hakanen 1972, 1973; Karlsson 1978b;

Hakulinen & Karlsson 1979; T. Itkonen 1974a, 1979, 1980; Toivainen 1986;

Chesterman 1977, 1991). On the outset, it once again needs to be stressed that the

term existential, as used in grammatical descriptions of Finnish, should not be

equated with the way in which it is used in English and other European lan-

guages. The term “existential clause (or sentence)” was first used by Jespersen

(1924: 155) to refer to “sentences corresponding to English sentences with there

is or there are, in which the existence of something is asserted or denied”. In

Halliday’s view (1985a: 130-131), existential processes in English are those that

represent that “something exists or happens”.

The Finnish term eksistentiaalilause has obviously been borrowed from

Jespersen, but ) with the exception of Hakulinen & Karlsson’s (1979) analysis )

it is generally used to refer to a different set of grammatical phenomena. Hakuli-

nen & Karlsson (1979: 95-96) look at an eksistentiaalilause from the perspective

of English, Swedish and French existential clauses. For them, a “prototypical

existential clause”1 in Finnish (e.g. Autotallissa oli vettä [garage+INE be+PAS+3SG

water+PAR] corresponds to an existential (or presentative) clause in Indo-Euro-

pean languages: Det fanns vatten i bilstallet, There was water in the garage, Il

y avait d’eau dans la garage.)

While it may be possible to treat existential clauses in Finnish as a separate

process type along with the other process types mentioned earlier in this chapter,

it seems to me that it is more revealing to treat the grammatical features of an

existential clause as a set of parallel options. The options that are relevant to the

Page 301: systematic functional linguistic

293

1 Chesterman uses the term “expression”. The SF term realization (2.4.7) is more appropriatein that it means that there cannot be “what is being expressed” without “expression” (cf.Saussure’s signifiant and signifié). If, for example, in a fictitious world where English andthe English-speaking world were replaced by languages in which there were no definite andindefinite articles, then it would be unlikely that definiteness would be an issue in Finnishlinguistics.

existential clause in Finnish are concerned with 1) the information (Given-New)

structure of the clause, and 2) notions of boundedness. As pointed out in Chapter

3, boundedness in Finnish is realized by the distinction between the partitive and

the other grammatical cases (nominative, accusative, genitive). While bounded-

ness can be connected with the notion of definiteness in English (see Chesterman

1991), the two cannot be equated. For instance, in Hakulinen & Karlsson’s exam-

ple of a prototypical existential Autotallissa oli vettä ‘There was water in the ga-

rage’, the nominal vettä ‘water+PAR’ is non-bounded and new information, and

from this it can be inferred that it is indefinite. However, as Chesterman (1991:

132) points out, the fact that “it is possible to infer” should not be confused with

realization, i.e. distinctions in definiteness are not grammatically realized in Finn-

ish.1

In what follows, I shall briefly outline some of the features of those clauses

that are generally referred to as existential clauses by Finnish linguists. I shall

then discuss the features that existential subjects and objects have in common.

The “subject” of this existential clause is unlike the constituent that has been re-

ferred to as the grammatical subject in this study. The grammatical subject is al-

ways in the nominative case and it (generally) agrees with the verb in number and

person. The “existential subject”, on the other hand, can be in the nominative but

it can also be in the partitive. The following existential clauses are reminiscent of

English existential clauses. I have underlined the “existential subject”.

(177) ai ai tuoll on poliisi [CA3:5]that(+ADE) be+3SG police+NOM

‘Oh dear, there’s a policeman over there.’

(178) tääl on autoja [CA2:28]here(+ADE) be+3SG car+PL+PAR

‘There are cars here.’

Page 302: systematic functional linguistic

294

While the grammatical subject in a non-existential clause is unaffected by the

polarity of the verb, if the clause is existential, the so-called subject is always in

the partitive if the polarity is negative:

(179) tuolla ei ole poliisia (Cf. example 177)that(+ADE) NEG+3SG be police+PAR

‘There isn’t a policeman over there.’

In the examples above, the existential subject is a noun. However, as posses-

sive clauses in Finnish are considered to be existential, it is also possible for the

existential subject to be a human pronoun in the accusative case. As discussed in

6.3.4.(iii), possessives in Finnish are characterized by the fact that a human pro-

noun realizing a Possessed is in the accusative case.

(180) Niin kauan kuin minulla on sinut [From T. Itkonen 1979: 83.]as long as I+ADE is+3SG you+ACC

‘As long as I’ve got you,

tunnen itseni onnelliseksi.‘I feel happy.’

As in other existential clauses, the pronoun would be in the partitive if the polar-

ity is negative.

However, existential clauses in Finnish comprise a far more heterogeneous

set. Firstly, the verb need not necessarily be the verb olla ‘to be’, as in examples

177 ) 180 above. There is a wide range of verbs that typically realize a material

process that can occur in an existential clause. The most typical seem to be verbs

expressing movement or change (of state) (cf. T. Itkonen 1979: 81):

(181) nyt kun tähän on muuttanu uusii ihmisii now that here+ILL be+3SG move+PTC new+PL+PAR people+PL+PAR

tähän taloon aika paljon, [niin nepä juttelee]here+ILL house+ILL quite a lot, [then/so they+NOM/PL+TIS chat+3SG]

‘Now that there have been quite a few new people who’ve moved into this (apartment)house, well they certainly chat with me.’ [CA8:12]

Page 303: systematic functional linguistic

295

(182) aina tulee masennuksii [CA8:12]always come+3SG depressions+PL+PAR

‘there are always bound to be periods of depression’

Other verbs can also occur in an existential clause and the existential subject can

also come at the beginning of the clause:

(183) Sienimyrkytyksiä sattuu meillä harvoin.mushroom+poisoning+PL+PAR happen+3SG we+ADE rarely‘Mushroom poisonings seldom occur in Finland/There are few mushroom poisoningsin Finland.’

(184) [Jugoslavian armeija moukaroi Dubrovnikia kolmelta suunnalta.‘The Yugoslavian army bombarded Dubrovnik from three directions’]

Useita kuoli ja kymmeniä loukaantui [HS 3.10.91: C1]several+PL+PAR die+PAS+3SG & ten+PL+PAR hurt+REF/PAS+3SG

‘Several (people) died and dozens were injured/There were several deaths and dozens ofinjured people .’

(185) poikia oli kiivennyt istutusten puihinboy+PL+PAR be-3SG climb+PTC(SG) planted area+GEN/PL tree+PL+ILL

‘boys had climbed into the trees that lined the road (? there were boys who had climbedinto the trees)’ [SE: 412]

In the following example from Hakanen (1972), the verb is even modified by a -

sti adverb:

(186) Uudisraiviolle kasvoi nopeasti orasta. [Hakanen 1972: 54]new+cleared land+ALL grow+PAS+3SG quickly sprout/shoot+PAR

‘On the newly cleared land crops quickly sprouted up.’

Existential clauses in English are sometimes seen in terms of the information

structure (Given ) New structure) of the clause (e.g. Quirk et al. (1985: 1402 ff.).

As the following examples illustrate an existential subject in Finnish is not neces-

sarily New, i.e. an existential clause is not simply a means of introducing a new

element into a text:

Page 304: systematic functional linguistic

296

1 Moreover, the functions Given and New are not only applicable to existential clauses, butto all clauses.

(187) a. aina tulee masennuksiialways come+3SG depressions+PL+PAR

‘there are always bound to be periods of depression’

÷ b. mutta niitä on terveilläkin ihmisillä [CA8:12]but they+PL+PAR be+3SG healthy+PL+ADE+TIS person+PL+ADE

‘but healthy people have them (periods of being depressed) too [they’re had byhealthy people too]’

(188) [Keskiasteen oppilaitoksia ei ole tarpeeksi. ‘There are not enough upper secondary andvocational schools.’]

÷ Niitä ei ole siellä,they+PAR NEG+3SG be+PTC there+ADE

missä niitä eniten tarvittaisiin.[[where+INE they+PAR most need+CON+INDEF]]

‘They are not where they are needed the most.’

Example 188 is from a newspaper and is taken from an article by Karlsson (1978:

297). In both examples, 187 and 188, the pronoun niitä ‘they-(PARTITIVE)’ is an

“existential subject”. It is, however, anaphoric in both examples: in example 187

it is co-referential with masennuksii ‘(periods) of depression’ and in example it is

co-referential with keskiasteen oppilaitoksia ‘middle-stage schools’. Thus, the

existential subject cannot necessarily be equated with the New element in the

clause.1

A significant feature of an existential clause is that it cannot have an object.

Thus existential clauses do not include transitive material processes. This is sig-

nificant because, as indicated at the end of section 6.7.1, it appears that both ob-

jects and “existential subjects” realize the macrorole Domain. They both indicate

the domain or the sphere of influence of the process. The clause lapsi söi ome-

nan/ omenaa [child eat apple+GEN/PAR] ‘The child ate the apple/was eating the

apple ~ ate some of the apple’ construes a material process (‘eating’) and indi-

cates the sphere of influence or the domain of the eating: omena ‘apple’. The

case-marking of the Goal indicates whether or not this was completed in space or

Page 305: systematic functional linguistic

297

1 Itkonen’s analysis seems to have been inspired by Comrie (1975). As Comrie’s analysis isless developed and overlooks some features of Finnish, I shall refer only to Itkonen’s analysis.

time. The clause Puistossa on/leikkii lapsia [park+IN E be+3SG/play+3SG

child+PL+PAR ‘In the park (there are) children/ children playing’ indicate con-

strues a relational process (on ‘be’) or a material process (leikkii ‘plays’), and the

entity which is included in the domain of this process (lapsia ‘children’).

This link between objects and existential subjects is supported by an analy-

sis of formal case-marking patterns in Finnish by T. Itkonen (1979).1 Itkonen

discusses Finnish in terms of the traditional notion of ergativity (i.e. ergativity as

a formal case-marking feature of languages). According to Itkonen, while the

case-marking in Finnish resembles a classical ergative system, the crucial distinc-

tion in Finnish is not between 1) a transitive subject and 2) an intransitive subject

and an object, as in an idealized ergative system. In other words, to use English as

a metalanguage, the distinction is not between subjects illustrated by The child

kicked the toy, on the one hand, and subjects like The child is sleeping and ob-

jects like The child kicked the toy, on the other hand. According to Itkonen:

The crucial border in Finnish ... is one separating the subject of “non-existential” or “normal”sentences from both the subject of “existential” sentences and the object, with concomitantdifferences in the form of the finite verb (T. Itkonen 1979: 79-80).

To use English as a metalanguage, again, the distinction in Finnish is between

The child kicked the toy, on the one hand, and objects like The child kicked the

toy and “existential subjects” like In the cupboard (there) are toys ~ There are

toys in the cupboard.

Itkonen cites a number of grammatical criteria that relate objects in Finnish

with “existential subjects”: 1) they can both be in the partitive, 2) they are both

generally in the partitive if the finite verb is negative, 3) a human pronoun is in

the accusative in both instances (if it is not in the partitive), and 4) neither triggers

agreement with the verb. This all points to the fact that, from a highly generalized

Page 306: systematic functional linguistic

298

1 Itkonen does not consider what is referred to as the complement (predikatiivi) in traditionalgrammar, although it too has some of the features of a Domain. Like an object and anexistential subject, the complement can be in the nominative, partitive or genitive and it doesnot trigger agreement with the verb. The complement, however, is unaffected by the polarityof the verb, and a human pronoun is the grammatical subject in instances such as Se olet sinä‘It’s you’.

semantic viewpoint, what is traditionally referred to as the subject of an existen-

tial clause and the object of a transitive verb are realizations of the one function.

The function Domain seems to be the most appropriate label for this macro-role,

since in these existential clauses, it also refers to the entity to which the process

(of being, becoming, dying, finding, etc.) is extended, as illustrated by the exam-

ples above.1

In order to understand what is happening in an existential clause, the NP

realizing the Domain in an existential clause needs to be contrasted to the NP

realizing the Medium in an agnate non-existential clause. Thus example 185

above (poikia oli kiivennyt puihin ‘boys had climbed into the trees’) can be com-

pared with pojat olivat kiivenneet puihin ‘(The) boys had climbed into the trees’,

where the NP in question is in the nominative and agrees with the verb in person

and number. The example above occurs in a context in which there is no previous

mention of the boys:

(189) At the beginning of a new section of a chapter of the novel Sinuhe, egyptiläinen (Sinuhethe Egyptian) by Mika Waltari:

Niin laskematon on ihmisen mieli ja siinä määrin oli faraon totuus sokaissut minut, ettenaavistanut mitään pahaa, vaikka paahtavassa päivässä viipyi yhä kytevien raunioiden sau-hua ja virrasta nousi ruumiiden haju. Sillä oinasten valtatie oli liputettu kirjavin viirein jaihmisjoukot reunustivat sitä määrättöminä nähdäkseen faraon ja -

‘So incalculable is the human mind and the Pharaoh’s truth had blinded me to such anextent that I had no premonition that all is not well, although in the heat of the day therestill lingered the smoke from smouldering ruins and the smell of corpses rose from theriver. Since the Road of the Rams was decked with colourful flags and crowds of peoplelined the road in order to see the Pharaoh and’

Page 307: systematic functional linguistic

299

÷ poikia oli kiivennyt istutusten puihinboy+PL+PAR be-3SG climb+PTC(SG) planted area+GEN/PL tree+PL+ILL

‘boys had climbed into the trees that lined the road (? there were boys who had climbedinto the trees)’

ja Pepitaton oli varannut lukemattomia kukkakoreja -- [SE: 412]‘and Pepitaton had ordered countless baskets of flowers’

As pointed out earlier, the existential subject cannot simply be equated with New.

In the clause ihmisjoukot reunustivat sitä ‘crowds of people lined it (the road)’

from this example (189) the NP ihmisjoukot ‘crowds of people’ is New; there had

been no mention of ihmisjoukot ‘crowds of people’ earlier in the passage. Yet this

NP is not an “existential subject”, but an ordinary subject: it is in the nominative

and agrees with the verb in person and number.

Finnish grammarians usually explain the meaning of the existential subject

by saying that its “existence is not presupposed”. According to T. Itkonen (1979:

81), for example:

Existential sentence is a label for intransitive sentences which, without foregoing presup-positions, express the existence of the subject, its coming into existence, its cessation ofexistence, or an essential change in state ) usually from the standpoint of location in its wid-est sense. [Emphasis added.]

The characterization in terms of the existence or change of existence applies to

some of the examples above but is not applicable to all instances: it is difficult to

see how poikia oli kiivennyt puihin ‘boys had climbed into the trees’ (185), for

example, expresses the “existence of the subject” or an essential change in state.

It seems to me that it is not so much a question of the existence of an entity, but

of its presence in the situation and whether or not this can be assumed. If we

return to the situation created by the novelist in this example (185), then the scene

that is depicted is of a celebration in the street, and thus one can expect that there

will be crowds of people: oinasten valtatie oli liputettu kirjavin viirein ja

ihmisjoukot reunustivat sitä määrättöminä nähdäkseen faraon ‘the Road of the

Rams was decked with colourful flags and crowds of people lined the road in

Page 308: systematic functional linguistic

300

1 Here I am assuming that the New element normally comes at the end of the tone group inFinnish (see Chapter 7).

order to see the Pharaoh’. In the clause poikia oli kiivennyt puihin ‘boys had

climbed into the trees’, the existential subject poikia ‘boys’, on the other hand, is

construed as something whose presence is not assumed. Thus the partitive NP in

an existential clause construes an entity (or entities) whose presence is not as-

sumed.

As suggested earlier, it seems to me that a better understanding of what is

involved in existential clauses in Finnish is gained by analysing the different

meanings that are conflated separately; to use Firth’s metaphor (see 2.2.5. p. 20),

we need to disperse or split up the meanings, rather like a prism can be used to

disperse white light. The notion of an existential clause is problematic in that it

involves both notions to do with the information structure of the clause (Given-

New) and notions to do with boundedness. If we disperse the meanings, then the

final NPs in examples 190a and 190b below are both nonbounded and New,

whereas the final NPs in examples 191a and 191b are both bounded and New:1

(190) a. Autotallissa oli vettägarage+INE be+3SG water+PAR

‘There was water in the garage’

b. tääl on autoja [example 178 above]‘There are cars here.’

(191) a. tuoll on poliisi [example 177 above]‘There’s a policeman over there.’

b. Hänellä on kauniit silmät.(s)he+ADE be+3SG beautiful+NOM/PL eye+NOM/PL

‘She/he’s got beautiful eyes.’

Example 191b illustrates one of the differences between boundedness in Finnish

and definiteness in English. Parts of the body (a particular person’s body) are

bounded in Finnish.

Page 309: systematic functional linguistic

301

In the following examples, the initial NP is unbounded but Given:

(192) mutta niitä on terveilläkin ihmisillä [Example 187 above]‘but healthy people have them (periods of depression) too [they’re had by healthy peo-ple too]’

(193) Meistä ei ole miljonääreiksi. Me emme kestä rahaa.‘We’re not the millionaire type. We can’t handle money.’

÷ Kun sitä tuliwhen it+PAR come+PAS/3SG

‘when it/some came our way’

se pantiin samaan tien menemään. [Example from Karlsson 1978b: 297]‘it was spent immediately.’

To summarize, it seems to me that the realization of boundedness in Finnish

should be regarded as a distinct option that can be conflated with other options in

the clause. NPs in which distinctions in boundedness can be made are referred to

as the Domain in this study. The Domain is realized by what has traditionally

been referred to as the object and the existential subject. The Medium ) the entity

that the “key figure in a process .. without which there would be no process at all”

is realized by the subject or by a bound morpheme (personal ending) attached to

the verb.

6.7.3. A Note on Derivational Affixes and External Causation

In his analysis of English, Halliday points out that:

Either the process is represented as self-engendering, in which case there is no separateAgent; or it is represented as engendered from the outside, in which case, there is anotherparticipant functioning as Agent. Thus the clauses the glass broke, the baby sat up, the boyran are all structured as Medium + Process. In the real world, there may well have been someexternal agency involved in the breaking of the glass; but in the semantics of English it isrepresented as having been self-caused. For that matter there may have been some externalagency also in the baby’s sitting up, and even in the boy’s running .. We may choose to putthe Agent in, as in the heat broke the glass, Jane sat the baby up, the lion chased the boy(Halliday 1985a: 147.)

Page 310: systematic functional linguistic

302

1 See also Kulonen 1989: 29 ff.

Halliday (1985a: 151) distinguishes between middle clauses (the glass broke, the

baby sat up, the boy ran) and effective clauses, which have the feature ‘agency’.

Effective clauses are either active (the heat broke the glass, Jane sat the baby

up, the lion chased the boy) or passive (the glass was broken (by the cat)). The

kinds of distinctions that Halliday is referring to here are realized in Finnish by

derivational affixes in the verb, although the distinctions that are made in the

grammar of Finnish are, of course, different from those made in English.

A verb in Finnish can contain one or more of a number of derivational af-

fixes. The affix -u-/-y-, for example, is generally referred to as a passive affix in

grammatical descriptions of Finnish; Kulonen-Korhonen (1985)1 refers to it as an

automative, since it refers to a process that does not involve an external agent:

(194) Kaad + u + i + n kadulla.fall + PASSIVE/AUTOMATIVE + PAST + 1SG street+ADE

Pro:mat + Ac Circ ‘I fell over in the street.’

This automative affix -u-/-y- construes a process that is self-engendered. The

verb kaatua ‘to fall over, to spill’ in (194) contrasts with kaataa ‘to pour, to fell

[e.g.] a tree’. An appropriate probe for a clause like 194 is Mitä tapahtui X:lle?

‘What happened to X?’ (rather than ‘What did X do?’ or ‘What happened?’). The

affix -u-/-y- makes the inherent participant more like an undergoer than an Actor.

Other examples include: kukka kuihtui ‘the flower wilted’, Kyllästyin esitykseen

‘I got fed up with the performance’, Lapsi nukkuu ‘The child is sleeping’, Sulta

loppuu pohjavedet [from/on you will end your ground water] (example 157)

‘Your ground water will dry up on you’, Punastuin (from puna- ‘red’) ‘I

blushed’, Sää muuttui kirkkaammaksi ‘The weather became clearer’, where the

verb is muuttua ‘to change, to move’, Lasi särkyi ‘The glass broke’, Hän suuttui

minuun ‘(S)he got angry with me’ (suuttua ‘to get angry’) etc. As indicated in the

quote from Halliday above, there may well have been some external agency in-

volved in the wilting of the flower, the breaking of the glass, in the child going to

Page 311: systematic functional linguistic

303

sleep etc. One can say kukka kuihtui ‘the flower wilted’ even if this happened

because the speaker did not water it. Processes with an automative affix are con-

strued in the semantics of Finnish as having been self-engendered.

An automative process in Finnish is different to what Halliday refers to as a

middle clause in English (e.g. the boy ran, the glass broke). A Finnish automa-

tive process is something that happens by itself or it is, in some sense, inexorable.

In Finnish, the process of ‘running’ is not something that happens by itself, and,

thus, in the clause poika juoksi ‘the boy ran’ the verb juosta ‘run’ is not morpho-

logically marked as an automative. There are other similar processes that are

morphologically unmarked in Finnish: laulaa ‘to sing’, kuolla ‘to die’, lukea ‘to

read’ etc. These processes are not self-engendered in Finnish, but neither are they

morphologically marked as being engendered by an external Agent.

Corresponding to Halliday’s (1985a: 151) “effective” are clauses in which

the process is realized by a verb that has a causative affix. Finnish has a number

of causative-type affixes that construe a process that is not self-engendered, and

in these the Actor is more like an agent. The following are examples of the most

common causative affix (-tta-/-ttä-): Aurinko sulatti lumen ‘The sun melted the

ice’ (sulattaa ‘to melt’, cf. sulaa ‘to melt’), Hän pelotti minua ‘(S)he frightened

me’ (pelottaa ‘to frighten’, cf. pelätä ‘to be afraid ~ to fear’), Ilona muutti maal-

le ‘Ilona moved to the countryside’ (muuttaa ‘to move, change’, cf. muuttua ‘to

change, become’), Hän kirjoittaa väitöskirjaa ‘(S)he is writing a doctoral disserta-

tion’ (kirjoittaa ‘to write’) etc.

Causative affixes are common in experiencer process (see 6.4.3.(ii)), e.g.

mua ihmetyttää ~ suututtaa ~ janottaa ‘I am surprised ~ feel angry ~ feel thirsty’

etc. They can also occur in some meteorological processes (see 6.4.3.(i)), e.g.

pyryttää ‘to snowstorm’. Similar so-called impersonal clauses are common in

Australian languages (see Walsh 1987). Both experiencer and meteorological

processes in Finnish can be characterized by the fact they can occur without an

Actor, and so, at first sight, they appear to be problematic: they incorporate a

Page 312: systematic functional linguistic

304

causative affix, yet there is no external causation. As Halliday points out in the

quotation above (p. 301), from the perspective of English, either a process is self-

engendering, and occurs without an Agent, or it is engendered from the outside,

in which case there is an external agent. Experiencer processes in Finnish are a

clear indication of the ambivalence of many basic physical and emotional human

processes: on the one hand, emotions are construed as not being self-engendered

and, on the other hand, they need not necessarily be caused by an external Agent.

In other words, the Experiencer in an experiencer process is affected by a process

that can take place without a specific instigator. As pointed out in 6.4.3.(ii), how-

ever, an Instigator can be added to an experiencer clause.

Halliday makes a distinction between the active and passive in effective

clauses in English. What is referred to as the passive in grammars of Finnish is

referred to as an indefinite in this study. The indefinite is problematic, and one

cannot make a distinction between indefinite clauses and other clauses based on

the assumption that an indefinite clause has the feature ‘agency’. While the indef-

inite (passive) has been etymologically linked to the causative (Lehtinen 1984), in

contemporary Finnish not only transitive verbs but also intransitive verbs, modal

verbs and the verb olla ‘to be’ can be in the indefinite. It can often be translated

by ‘they’, ‘we’, ‘you’ or ‘people in general’:

(195) Genevessä ollaan ymmärtäväisempiä. [HSkl 1/84: 44]Geneva+INE be+INDE understanding+COMP+PL+PAR

‘They’re more understanding in Geneva.’

(196) Puhutaan suomalaisen muotoilun kriisistä. [SK 14.1.83]speak+INDE Finnish+GEN design+GEN crisis+GEN

‘People are talking about the crisis in Finnish design.’

Within the framework presented here, it could be argued that the indefinite (pas-

sive) morpheme realizes an unspecified human Medium (see Shore 1986, 1988).

Thus, an indefinite clause is not necessarily effective in Finnish. It typically con-

strues a process that involves unspecified participants, and these participants must

Page 313: systematic functional linguistic

305

1 Thus, the English clauses The glass was broken by the cat and The tree was blown down bythe wind would not be translated by an indefinite in Finnish.

be human.1 Of course, an indefinite clause can have the feature of ‘agency’ ) and

many indefinite clauses do have this feature ) but this is not a mandatory feature

of an indefinite clause. The indefinite as a means of introducing a feature of exter-

nal human agency is clearest in those process types that are inherently agentless,

for example, resultative intensive processes. As discussed in 6.3.1.(iii), a resulta-

tive clause has the structure X+ELATIVE tulla ‘come’ Y in Finnish, e.g. Minusta

tulee taitelija ‘I’m going to be an artist’. If the intransitive verb tulla ‘come’ is

replaced by an effective indefinite verb, an additional feature of external human

agency is added to the clause:

(197) Lars-Gunnar Nordströmistä tehtiin vuodenLars-Gunnar Nordström+ELA make+INDE year+GEN

Car Pro:int:res Att+ Agent (human, unspecified)

taitelija. [HSkl 1/84: 52]artist+NOM

‘Lars-Gunnar Nordström was made artist of the year.’

Similarly, in the following example, the resultative process X+ELATIVE tulla

‘come’ Y is construed as being brought about by human forces:

(198) [Koulussa mä joskus leuhkin poikakavereista, mutta paskat ei mulla semmoista ole.‘At school I sometimes brag about my boyfriends, but shit I haven’t got any.’]

Mustahan rakennetaan kitkerän katkeraa vanhapiikaa,I+ELA+TIS build+INDE bitter+GEN bitter+PAR spinster+PAR

Car Pro:int:res + Agent Att(unspecified)

‘I’m being made into a really embittered spinster

joka kiljuu pihalla leikkiville lapsille. [HS 28.1.84: 21]who yells at children playing in the yard.’

As this brief discussion indicates, the distinctions made in the morphology

of the verb are complex in Finnish. There is not a simple two-way distinction

between middle and effective in Finnish, and the active/passive distinction in

Page 314: systematic functional linguistic

306

English does not correspond to the distinction between indefinite and non-indefi-

nite clauses in Finnish. There are also other derivational affixes that need to be

considered in a more delicate description, e.g. the reflexive (-utu-/-yty-,

kirjoi+tta+utu+a ‘to write oneself, i.e. to register’, cf. kirjoittaa ‘to write’). Re-

cent discussions by Finnish linguists of automative, causative and reflexive

derivational affixes include Koivisto 1991, Kulonen-Korhonen 1985, Kytömäki

1989 and Räisänen 1988.

6.8. Concluding Remarks

This chapter has been concerned with the experiential meanings that are con-

strued at the rank of clause in Finnish. Finnish process types were grouped under

three superordinate process types: relational, material and mental (where mental

refers to processes of human consciousness). The process types were not seen as

discrete and absolute, but as overlapping, and an attempt was made to show the

prototypical grammatical and semantic features shared by each process type. The

notion of grammatical metaphor was also discussed and applied to experiential

meanings at clause rank. The macro-roles Medium and Domain were postulated

on the basis of grammatical criteria that has been put forward by T. Itkonen. In

the final section, some indication was given of how the semantics of the engen-

dering of a process is construed through derivational affixes in Finnish. What is

referred to as the passive in Finnish, on the other hand, does not construe distinc-

tions relating to whether the process is self-engendered or externally engendered,

but is seen as a parallel option realizing an unspecified, human Medium.

Page 315: systematic functional linguistic

307

Chapter 7

Textual Structures in the Finnish Clause

7.1. Overview

The chapter is concerned with the organization of the clause as a message, with

the way in which textual meanings are realized in the clause. According to

Matthiessen & Halliday (forthcoming, section 2.1):

The role of the textual metafunction is an enabling one. It serves to enable the presentation ofideational and interpersonal meaning as text in context. While the ideational and interpersonalmetafunctions orient towards first-order reality ) reality that exists independently of language) the textual metafunction orients towards the reality brought into existence by languageitself, viz. second-order, symbolic reality.

As Matthiessen & Halliday emphasise, however, this does not mean that the tex-

tual metafunction comes into operation only after ideational and interpersonal

meanings have been created. These different strands of meaning are simulta-

neously realized in the clause.

SF theory assumes that there are two kinds of textual meaning that are real-

ized in any language: Given-New and Theme-Rheme. These two kinds of mean-

ing are independently variable: although Given and Theme may be conflated,

they need not necessarily be conflated. Given and New are functions in the infor-

mation unit, which Halliday (1967c, 1985a) assumes is realized in the tone group.

While a detailed analysis of information structure in Finnish is beyond the scope

of this study, since this would entail a comprehensive phonetic and phonological

analysis of Finnish, there will be some preliminary discussion of Given and New

in the next section (7.2) since in the interest of clarity, it is worthwhile making

explicit what is meant by these terms since the discussion of Theme-Rheme struc-

tures presupposes some familiarity with the way in which Given and New is de-

fined in SF theory. The section on Given and New includes a short section (7.2.3)

Page 316: systematic functional linguistic

308

on the way in which segmental elements ) e.g. word order, focusing particles etc.

) serve as indicators of information structure in Finnish.

Section 7.3 discusses the Theme-Rheme organization of the clause in Finn-

ish with particular emphasis on the topical Theme. This analysis is illustrated

with a complete text (the “cat text”) reproduced in Appendix 1. Theme-Rheme

structures in Finnish (as in English) are realized by the sequencing of constitu-

ents, although, in Finnish, where word-order is far more flexible than it is in Eng-

lish, the situation is more complex. I shall not be concerned with the non-struc-

tural cohesive resources that can cut across clause (or clause complex) boundaries

(see Halliday 1985a: Chapter 9; Halliday & Hasan 1976; Halliday & Hasan 1985:

Chapter 5.), except where they bear on the discussion of Theme.

Section 7.4 looks at the textual structure of clause complexes complement-

ing the analysis of clause complexes presented in Chapter 4. Section 7.5 discusses

some other central topics in the textual organization of the Finnish clause. Section

7.5.1 discusses the non-realization (or “ellipsis”) of inherent participants in Finn-

ish. Section 7.5.2 deals with the realization of the grammatical subject in Finnish,

and the problems that it poses for the recognition of topical Theme. The final

section 7.5.3 is concerned with clause types proposed by Finnish linguists but

ignored in Chapter 6 of this study since the patterning in these clauses is consid-

ered to be textual.

7.2. Given and New

7.2.1. Brown and Yule’s Approach to Given and New

As discussed in 2.4.10, and elsewhere in Chapter 2, in SF theory, one does not

assume that there are meanings that simply exist, i.e. that they are independent of

a semiotic system in which they are realized. In much of the literature on Given

and New, linguists adopt this alternative approach to meaning. Brown & Yule

Page 317: systematic functional linguistic

309

(1983: 182), for example, define Given information as “whatever knowledge

speakers and hearers share”. Knowledge that is or is not shared then becomes the

starting point for looking at what is Given and what is New, or alternative terms

with intermediate categories, such as in Prince’s (1981) taxonomy of information

status, which is discussed by Brown and Yule. Prince’s taxonomy ranges from

entities that are “brand new” to entities that are either situationally or textually

“evoked”, i.e. entities that are salient in the context (such as ‘I’ and ‘you’) or enti-

ties which have already been introduced into the discourse. In-between are enti-

ties that are inferable, i.e. entities that can be inferred from some other entity that

has already been introduced into the discourse (e.g. mention of a car, in traffic at

least, would permit the inference of a driver).

Brown and Yule then apply the categories of information status thus defined

to the analysis of data and look at how, for example, a brand new entity is real-

ized linguistically. Their data is drawn from artificial experiments (see Yule

1981) in which participant A has a drawing (of lines, squares, circles etc.), which

participant B cannot see. Participant B has a paper and a red and black pen, and

through interaction with A, B is expected to reproduce the drawing. Since in an

artificial situation like this ) which is rarely, if ever, repeated in real life ) one

can, to a large extent, determine the information status of entities, one can then

look at how these categories of entity are expressed linguistically. Brown and

Yule correlate the information status of these entities with 1) linguistic features

associated with the phrase in English (definite or indefinite article, pronominal

reference, ellipsis) and 2) phonological prominence.

Since Given is defined by Brown and Yule as shared knowledge, a problem

arises in their experiments with inferable entities such as the corner (of a square).

In spite of the fact that participant A (the speaker) knows that both the square and

the corner are physically present in the context since they have just been drawn

by participant B (the hearer), these inferables are referred to by expressions that

are phonologically prominent. This leads Brown & Yule (1983: 187.) to the fol-

lowing conclusion:

Page 318: systematic functional linguistic

310

If the speaker has no reason to believe that the hearer is paying attention to these particularentities or properties of entities [e.g. to the corner of a square], he mentions them with phono-logical prominence.

Another problem that Brown and Yule (1983: 187) illustrate with another text, a

recipe, is that “brand new” entities can be expressed by a nominal with either an

indefinite or a definite article:

(1) a. Slice the onion finely, brown in the butter and then place in a small dishb. ... add to the fat in the panc. ... and boil for twenty minutes with the lid on the pan.

Even if these were to be treated as inferable rather than brand new entities, ac-

cording to Brown and Yule, the problem remains that they are not treated in the

same way. They conclude that the only way of assessing the information status

attributed by a writer or speaker is in the form of the expression produced.

The problem with Brown & Yule’s analysis is that they postulate meanings

associated with information status that are independent of language. These pre-

determined meanings are expressed by definite or indefinite NPs (a/the square),

pronominal reference and ellipsis (it, 0/ ), and phonological prominence. When one

begins to look at the formal distinctions that are made in a language, then prob-

lems begin to arise, as in the case of the definite and indefinite articles above.

Either one assumes that speakers or writers are inconsistent in the way in which

they express these predetermined meanings, which is the only recourse that

Brown and Yule have, or one abandons the approach that leads to these inconsis-

tencies.

7.2.2. Halliday’s Approach

Ironically an alternative non-objectivist approach taken by Halliday (1967c,

1985a: 59, 271-286, 346-365) is criticized and rejected by Brown & Yule. How-

ever, Halliday’s approach is later appealed to again when their own objectivist

position begins to break down: citing Halliday, Brown and Yule (1983: 189) con-

Page 319: systematic functional linguistic

311

1 For notational conventions, see Appendix 5. The numbers refer to the tones of Englishdistinguished by Halliday: tone 1 is a falling tone, the unmarked statement tone, and tone 4is a falling-rising tone, often associated with contrasts, reservations or conditions.

clude that information status is determined by the speaker. In Halliday’s

approach, meanings do not exist independently of language but are realized, or in

Halliday’s (1967c) earlier terminology, expounded in language: meanings such as

‘definiteness’ or ‘newness’ are construed by the lexicogrammatical and phonolog-

ical resources of a language. The meanings associated with the English articles in

the recipe example above are to do with definiteness, what is “knowable” (cf.

Chafe’s (1987: 50) “identifiable”): the article the, for example, can be glossed as

indicating that the speaker expects the listener to know what she or he is talking

about: 1) it is something that can be taken for granted either in the universe or

some culturally defined part of it (e.g. the sun, the government, the doorman, the

cat) or 2) it has already been mentioned in a (spoken or written) text or it is com-

ing up in the text (structural cataphora). The meaning of definiteness in English,

however, is inseparable from the article system that realizes it, and, thus, in the

final analysis, its meaning is ineffable.

As for Given and New, Halliday does not assume we can come up with a

“rigorous definition”, as Yule (1981: 42) seems to assume, or that they can be

defined extralinguistically in terms of shared knowledge. For Halliday (1985a:

275), Given and New, and the meanings of Given and New, are inextricably tied

up with prosodic features of the tone group in English. In each tone group there is

a foot or syllable that carries the main pitch movement (rise or fall or change in

direction). This foot or syllable is referred to as the tonic element, and this

prosodic feature (i.e. pitch movement) is referred to as tonic prominence. The

tonic element is said to carry the information focus. In Halliday’s view, it marks

where the New element ends.1

(2) a. // 4 v in / this job / Anne we’re // 1 working with / silver / v //b. // 1 v now / silver / needs to have / love / v // (Halliday 1985a: 283)

(3) // 4 v well they could / make the / credit mark fi / fifty // (Halliday 1967: 49.)

Page 320: systematic functional linguistic

312

The meaning of Given is “this is not news”, and the meaning of New is “attend to

this, this is news” (Halliday 1985a: 277). This is in accord with Brown & Yule’s

(1983: 164) characterization: “phonological prominence has a general Watch this!

function”).

Thus, in Halliday’s approach, Given and New cannot be glossed as “previ-

ously mentioned” (or shared information) and “not previously mentioned”. Given

information is “not New”: it is information that is assigned by the speaker as

something (s)he assumes that the listener can take for granted. According to

Halliday (1985a: 277), what is treated as Given may be something that has been

mentioned before or it may be something that is in the situation, such as I and

you; it can be something that is taken for granted or is “in the air”, or something

that is not around at all but is presented as Given for rhetorical purposes. What is

presented as New, may be something that has not been mentioned before, but it

may be something that has been mentioned before, but the speaker wants the lis-

tener to attend to it. The text in example 2 above is taken from the speech of a

manageress initiating a new salesgirl, Anne, to a silverware department in a large

department store. The demonstrative this in in this job is New because it is

contrastive (i.e. in this job as opposed to other jobs). As for silver in we’re work-

ing with silver, it is evoked in Prince’s sense since Anne and the manageress are

in a silverware department: Anne knows perfectly well that they are working with

silver. However, it is prominent because this is what Anne is being asked to at-

tend to. The Newness of love in now silver needs to have love (in the second

tone group of example 2) as well as fifty in example 3 (well they could make the

credit mark fi fifty) can be related to the fact that they have not been mentioned

before.

I shall not go further into Halliday’s analysis of English (see Halliday

1967c, 1985a: 59, 271-286, 346-365; Kress (ed.) Chapter 14; see also Tench

1991). The discussion above is simply to illustrate that meanings such as Given

and New in English are tied in with the form of the language, with formal phono-

logical distinctions. The investigation of tone groups in Finnish and the way in

which they realize Given-New structures in Finnish would require a comprehen-

Page 321: systematic functional linguistic

313

sive phonetic and phonological analysis of conversational Finnish, along the lines

suggested by Halliday (1967c). While, as already mentioned, such a phonetic

analysis is beyond the scope of this study, I shall make a few preliminary com-

ments and observations.

There are basic prosodic differences between Finnish and English. In terms

of the broad ) and somewhat over-simplified ) division made between syllable-

timed and foot-timed (or stress-timed) languages, Finnish has traditionally been

regarded as having syllable-timed rhythm (Hackman 1978, Nevalainen 1990:

238). The tones in Finnish are also bound to be different. For example, the un-

marked realization of a polar interrogative (functioning as a question) in English

is a rising tone (tone 2; see Halliday 1985a: 284), whereas, in Finnish, a polar

interrogative (functioning as a question) is generally realized with a higher initial

pitch level than a corresponding declarative (functioning as a statement) (see

Hirvonen 1970; Iivonen et al. 1987: 239 ff.).

On the other hand, while Halliday (1985a) makes no claims about the uni-

versality of the way in which Given-New structures are realized in the tone group

in his analysis of English, very similar kinds of comments concerning the func-

tion of tonic prominence in Finnish have been made by Finnish phoneticians, e.g.

Iivonen et al. (1987: 229 ff.), Nevalainen (1990). In a study that contrasts English

and Finnish prosody, Nevalainen (1990: 243) states:

We can distinguish three kinds of information focus that are signalled by prosodic means:focus for new information, focus for contrast and focus for emphasis. In both languages, anew information focus is usually placed on the last lexical item of the intonation unit, andcontains what in British literature is called the nuclear or tonic syllable. Contrastive and em-phatic focus may vary positionally, and show more extreme prosodic contrasts in both lan-guages.

However, rather than talk about variation in pitch, Finnish phoneticians generally

refer to stress or accentuation, i.e. a combination of higher than normal pitch,

increased intensity, extended duration and clearer articulation, not all of which

need necessarily be present (Iivonen 1976: 39; Iivonen et al. 1987: 229 ff.;

Nevalainen 1990). Moreover, in accordance with traditional terminology, they

Page 322: systematic functional linguistic

314

1 Cf. however, Halliday (1985a: 276) who shows how the pre-tonic and variation in rhythmin English can be an index of the Given-New structure before the tonic element.

refer to sentence (or clause) stress (lausepaino), which ) from a systemic-func-

tional perspective ) confuses grammatical and prosodic phenomena.

In Halliday’s approach, the three kinds of information focus mentioned by

Nevalainen can be subsumed under a general attend to this (this is news)! ~

watch this! function. However, rather than confine the focus of a New informa-

tion focus to the word that is realized with a syllable carrying tonic prominence,

Halliday (1985a: 275) regards the tonic foot as defining “the culmination of what

is New”. If this comes at the end of the tone group, then what this means is that )

on phonological grounds ) we cannot tell whether the first part of the tone group

is also New or whether it is Given.1 Taken out of context, example 2b above re-

peated below as 4, indicates that at least love is New, but we cannot tell where the

New element begins: whether the whole clause is New or whether silver or silver

needs is Given.

(4) // .. silver / needs to have / love // (Halliday 1985a: 276, 283, 346)(i) -----------------------------------> New(ii) Given --------------------------> New(iii) Given -----------> * --------> New

However, it is usually only in linguistics and philosophy that one comes across

bits of text out of context; in real life, there is other evidence for determining the

information structure. In the context of this example (see 2 above), we know that

silver is Given as it had just been mentioned.

Thus, according to Halliday (1985a: 276-277), the unmarked position for the

New element is at the end of the information unit. It is possible, however, for

New material to come at the beginning of the information unit, in which case it is

contrastive. It seems to me that these comments are also valid for Finnish, al-

though the position of a contrastive focus seems to be more variable in Finnish.

This is reflected in the fact that the focusing particle -kin (see next section) can

occur in various positions in the tone group. Another point made by Halliday that

Page 323: systematic functional linguistic

315

1 Even in instances where clause and tone group more or less correspond to each other, theboundaries do not always exactly match because a tone group, as defined by Halliday, beginswith a salient beat (or a silent one).

is also important in the analysis of Finnish is that there are a number of elements

in language that are typically Given: 1) anaphoric elements, i.e. those that refer to

things previously mentioned and 2) deictic elements, i.e. those that refer to the

here and now of the speech situation. Thus items like I, you, here and tomorrow

would only carry information focus if contrastive. In example 2a, repeated below

as 5, the demonstrative this is focused because it is contrastive:

(5) // 4 v in / this job / Anne we’re // 1 working with / silver / v //

The tonic prominence on this implies a contrast with other jobs.

The discussion of Given and New, so far, has been confined to their realiza-

tion in the tone group. A tone group is not the same thing as a clause. A clause,

being a unit of lexicogrammar, consists of phrases, words and morphemes, while

a tone group, a phonological unit, consists of feet and syllables. These categories

represent different orders of abstraction. However, there are many instances in

which a clause realized in spoken text conflates with a tone group, and this pro-

vides us with a link between grammatical structures in the clause and Given-New

structures in the tone group. The conflation ) or near conflation1 ) of clause and

tone group can, thus, be regarded as unmarked (Halliday 1985a: 59), i.e. as dis-

cussed in Chapter 5, this can be regarded as “a baseline”.

7.2.3. Segmental Markers of Information Structure in Finnish

If we assume that the tone group conflates with a clause in unmarked instances,

then the unmarked order in the tone group also serves as a base for the unmarked

order in a written text. Given Nevalainen’s (1990: 243) observation quoted on

page 313 above about the similarity between Finnish and English, then we would

Page 324: systematic functional linguistic

316

1 Iivonen et al. (1987: 233), however, claim that they are “common” in spoken Finnish.

normally expect New material to come at the end of a clause in a written text in

both languages. Without any indication to the contrary, we would expect the New

element to come at the end in the following example.

(6) Mä lähen lenkilleI-NOM leave+1SG jog+ALL

-----------------------> New‘I’m going jogging’

If this were part of a script for a play, we would assume that the tonic prominence

falls on the last word: mä lähen lenkille. This would be the unmarked “reading”

of the information structure of this written fragment.

However, there may be indications in a written text that point to a marked

information focus. Halliday (1985a: 59-61) discusses how predicated Theme

structures (referred to as “cleft sentences” in formal grammars) serve both the

Theme and information structure of English. Predicated themes of the type se oli

Diana joka ... ‘It was Diana who’ are considered rather marginal in Finnish and

they are not used in standardized written Finnish.1 One indication of marked in-

formation focus in written Finnish is the presence of what is aptly referred to by

Iivonen et al. (1987: 234) as a focus particle: these include the enclitic particle -

kin ‘too, also, even’(or -kaan/kään in a negative context) and a number of other

particles that are traditionally referred to as adverbs: myös ‘too, as well’, jopa

‘even’, edes ‘even (in a negative context)’, juuri ‘precisely’, and vain ‘only’. A

focus particle usually precedes or follows the element that it focuses on. As Iivo-

nen et al. (1987: 234) point out that these focus particles are associated with

contrastive stress; however, no phonetic analyses of natural conversation have

been carried out. Whether the item to which a focus particle such as -kin is at-

tached is actually stressed is debatable; it seems nearer to the truth to say that it

is attached to a word containing an element with tonic prominence (pitch varia-

tion). Example 6, above, can be contrasted with the following transcribed exam-

ple:

Page 325: systematic functional linguistic

317

1 There are also many instances where -kin is not a focus particle but an inseparable part ofa word, e.g. kuitenkin ‘nevertheless’, tuskin ‘scarcely’. These were not counted.

(7) <A> - mä oon lähös(sä) lenk(ille) [Tel2: 1]I+NOM be+1SG leave+INE jog+ALL

‘I’m going for a jog’

<B> - no mäkin lähen lenkillewell I+NOM+kin leave+1SG jog+ALL

‘well I’ll go jogging too.’

In this example, the first syllable of the word to which -kin was attached was

marked as being stressed or prominent by the transcriber, which, as noted above

(p. 313) can refer to any or all of a number of features (pitch, intensity, duration,

clearer articulation) in the Finnish tradition. I did not perceive the syllables in

question as stressed, but I did perceive pitch variation. In Speaker B’s turn in this

example, lähen lenkille is Given since it repeats what Speaker A has said. The

first person pronoun is deictic, and, therefore typically Given, but it is contrastive,

it is what the listener is being asked to attend to: “you’re going for a jog and so

am I”.

Like other focus particles, the suffix -kin directs us to “read” the information

structure in a certain way. Thus if the clause mäkin lähen lenkille had occurred in

a written text, the suffix -kin would have directed us to “read” the information

structure in the way indicated in 7 above. The suffix -kin is particularly important

in written Finnish ) in the “cat text” analysed in Appendix 1, for example, there

are 83 orthographical sentences and 16 occurrences of -kin or -kaan.1 Because

Finnish linguists and phoneticians generally talk about stress, rather than tonic

prominence, the suffix -kin is not always associated with prosodic features of the

tone group. For example, Vilppula (1984: 58) in an otherwise insightful discus-

sion of -kin assumes that the examples he discusses (all of which are in written or

transcribed form) are with “normal sentence stress”. However, it is difficult to

imagine any of his examples without tonic prominence (at least, some kind of

pitch variation that distinguishes it from the rest of the tone group). The following

Page 326: systematic functional linguistic

318

example from Vilppula is taken from a poem that is about two crows sitting on a

fence. One of the crows opens the conversation with the following words:

(8) Kurkikin jo lähti. (Lauri Pohjanpää, Syksy (Autumn).)crane+NOM+kin already left+3SG/PAS

‘even the crane has left already [or, if read, the crane has left already].’

It seems to me that it would be almost impossible to read this with the tonic

prominence coming at the end of the tone group, i.e. with the same kind of into-

nation contour as Pekka lähti Turkuun ‘Pekka left for Turku’.

Another important indication of marked information focus in Finnish is vari-

ation in word order. Word order variation is also apparent in a written text and it

also directs the reader to read the information structure in a particular way. The

word order in the following example, from a published collection of radio plays

(quoted by Vilkuna 1989: 104), indicates that the first constituent would have a

marked (contrastive) information focus if read aloud:

(9) [ – Mennään katsomaan mitä sille papan huoneelle tehdään. ‘Let’s go and see what has to be done to Grandpa’s room.’

– Tietää tuon katsomattakin.‘That’s obvious without looking.’]

Remontti siinä on tehtävä.renovation/repairs-NOM it/that-ESS be-3SG do/make-INDEF+PTC

‘Repairs are what’s needed/A renovation is what must be done’

The unmarked word order would be: Siinä on tehtävä remontti ‘It needs to be

renovated’ and the tonic prominence would come at the end of the tone group if

the clause were spoken. If example 9 were read aloud, the tonic prominence

would fall on remontti ‘renovations’.

Page 327: systematic functional linguistic

319

7.3. Theme and Rheme

7.3.1. Introductory Remarks

The analysis of Theme and Rheme presented here is illustrated by a text, which is

analysed and reproduced in full in Appendix 1. Since the text is about cats, for

convenience I shall refer to it as the “cat text”. This text was chosen because it

illustrates a number of problems that have to be addressed in the analysis of the

Theme-Rheme structure of Finnish. The textual resources used by the writer are

varied, yet the text is, nevertheless, a fairly simple text and it is relatively easy to

ascertain what the text as a whole is about: it is a text about cats and people’s

attitudes to cats. While the majority of my examples come from this particular

text, the generalizations that I make are based on my knowledge of Finnish and

are intended ) unless otherwise stated ) to be applicable to the central genres of

spoken and written Finnish. Where necessary, I shall quote other, isolated exam-

ples and bits of text.

According to Halliday (1985a: 38), the Theme is the element that “serves as

the point of departure of the message; it is that with which the clause is con-

cerned”. The Rheme is the remainder of the message. The Theme in English is

realized by initial position. However, Halliday sees the Theme as having an inter-

nal metafunctional structure of its own: a clause can have a number of themes.

The experiential Theme is referred to by Halliday as the topical Theme. The

topical Theme basically corresponds to what is regarded as the theme or topic in

other models. This section will concentrate on the topical Theme in Finnish and

some of the problems associated with determining which clause-initial constituent

realizes the topical Theme.

Page 328: systematic functional linguistic

320

7.3.2. Topical Theme (Topic) in Finnish

The topical Theme is an experiential notion; it is characterized either as “a point

of departure” or it is “what the clause is about”. It seems to me that a point of

departure seen from an experiential perspective can only be seen in terms of what

the clause is about, otherwise the definition of topical Theme is confused with

notions that are related to its realization. In other words, if an experiential point of

departure does not mean “what the clause is about”, then it can only mean the

experiential element that comes first in the temporal unfolding of the clause. The

conflation of definition and recognition criteria would mean that the definition of

topical Theme in any language is tied to the way in which it is realized in Eng-

lish.

If the topical Theme is “what the clause is about”, then the question that

needs to be addressed is: what is meant by “aboutness”. From an experiential

perspective, aboutness is concerned with the things that we talk about, and the

things that we talk about, the topics of our conversations, are phenomena that

have been construed as entities. The following example is from a conversation

between two women. They are discussing the fact that the daughter of one of

them (speaker A) is abroad. Speaker B comments on what a marvellous opportu-

nity it is for a young person to go abroad. Speaker A replies that it is, and contin-

ues:

(10) vaikka kamalasti se siellä itkialthough terribly she+NOM there+ADE cry+PAS/3SG

kun pari enimmäistä kertaa soitin [Tel1: 5]when couple first+PAR time+PAR ring+PAS+1SG

‘Although she cried terribly the first few times I rang.’

It seems unlikely that vaikka ‘although’ or kamalasti ‘terribly’ would be regarded

as the topical Theme of this clause. Similarly, in the following example from the

Page 329: systematic functional linguistic

321

1 Examples given in this chapter do not include form glosses. The glosses given in thischapter are intended to make it easier for a non-Finnish reader to read the examples andconcentrate on the ordering of the constituents. Form glosses for the cat text are given in thegloss of the complete text in Appendix 1. The number of the (orthographical) sentence fromwhich the example is taken is given in square brackets after the English translation.

cat text (from Appendix 1), it is unlikely that kummallisen sitkeästi ‘strangely

persistently’ would be regarded as the topical Theme.1

(11) Kummallisen sitkeästi elää käsitys,strangely persistently live view

että kissa saattaa repiä silmät päästäthat cat might scratch eyes from head

‘How strangely persistent is the view that a cat might scratch a person’s eyes out ..’ [4]

The circumstances of manner at the beginning of 10 and 11 have a function in the

experiential structure of the clause, but they are not construed as something that

might be an experiential point of departure in any other sense than in the sense

that they come first. The constituents kamalasti ‘terribly’ and kummallisen

sitkeästi ‘strangely persistently’ are not “topic-worthy”, to borrow and extend on

Vilkuna’s (1989: 47) notion of the “T-worthiness” of a constituent. For Vilkuna,

a T is defined positionally as the slot immediately before the verb and it appears

to be most closely related to the textual function of topic.

The finite verb in the following example can also be discounted as what the

clause is about.

(12) Olihan seitsemällä veljekselläkin oma temppujahad after all seven brothers even own tricks

tekevä Matti-kissa ..performing Matti-cat

‘After all even the Seven Brothers had their own trick-performing cat called Matti ...’ [9]

Page 330: systematic functional linguistic

322

The finite verb does not construct an entity but a process (or a relation between

entities). Thus it cannot be regarded as a topical Theme. A topical Theme ) as

Downing (1991: 126) points out ) is typically an “entity”, and, thus, for a process

to be a topical Theme it has to reified, i.e. it is realized by a nominal or nominal-

ized verb form.

The notion of a topic entity should not be taken too literally. What is at issue

is some kind of reification. The reification that is necessarily involved in making

something a topic can be illustrated by the actual English text of this study. In

order to problematize the notion of what the clause is about and focus on the

“about”, which is a preposition, I had to use a noun that is not common in Eng-

lish: “aboutness”. While we can talk about all sorts of experiential phenomena

construed as nouns, verbs, p-positions etc., in order to make something into a

topic we need to reify it. Topic-worthiness can roughly be ascertained by how

naturally an experiential function can figure as a response to the question “What

is this clause about?/Mistä tämä lause kertoo?”. If we ask this question in relation

to examples 10 or 12 above, then it seems to me that it is highly unlikely that

anyone would reply “It’s about ‘terribly’” or even “It’s about ‘having’”. Defined

in terms of aboutness, a topic can also refer to the previous text:

(13) Sama koskee myös pientä ja siroa burmaa, ..same affects also small & elegant Burmese‘The same goes for the small and elegant Burmese ..’ [73]

In this example, the topic sama ‘the same’ refers to the previous text.

Examples like 10, 11 and 12 above, where an adverb or a finite verb is the

first experiential element in the clause, are not particularly frequent in the cat text,

but they have been introduced into the discussion at this point in order to

problematize the notion of aboutness. In many of the clauses in the cat text it is

relatively easy to determine what the clauses is about. In the following example,

it is clear that the first clause is about kissan ystävät ‘people who like cats’ and

Page 331: systematic functional linguistic

323

the second clause is about he ‘they’, which anaphorically refers to kissan ystävät

‘people who like cats’.

(14) Kissan ystävät ovat suurpiirteisempiä,cat’s friends are more tolerant

he hyväksyvät helpommin toisen yksilöllisyyden.they accept more easily another’s individuality

‘People who like cats are more tolerant, they more readily accept the individuality ofanother.’

As for the realization of topical Theme in Finnish, it is clear that ) in the

majority of instances ) the position immediately preceding the finite verb appears

to be the most relevant. According to Hakulinen & Karlsson (1979: 298) the

theme of the clause is realized by the constituent in the pre-verb slot. Similarly,

Vilkuna (1989) postulates a word-order schema for Finnish in which there is a

syntactic slot (referred to as T) that immediately precedes the verb. Vilkuna’s

schema can be illustrated with an example from the cat text:

(15) Omakotitalossa kissa oppii pysymään omassa pihapiirissä ..in detached+house cat learn to stay in own yardK T V-field

‘In a detached house a cat learns to stay within the confines of its own yard ..’ [28]

As mentioned above (p. 321), Vilkuna’s T-position appears to be most closely

related to the textual function of topic. The K-position in Vilkuna’s word-order

schema often (though not always) appears to be most clearly related to what

would be considered a contrastive New element in the approach adopted here.

(See also Shore 1990, for a discussion of Vilkuna 1989.)

In order to make the exposition clearer, I shall oversimplify and maintain as

a first rule of thumb that the topical Theme in a declarative clause in Finnish is

realized by the constituent immediately preceding the finite verb ) with the pro-

viso that this constituent also realizes a function in the experiential structure of

the clause. Thus a modal element preceding the verb is discounted as a topical

Theme. This recognition criterion for topical Theme means that in the temporal

unfolding of the clause the finite verb marks the beginning of the Rheme, or, in

Page 332: systematic functional linguistic

324

the light of what is said later on in this chapter, the finite verb, in fact, marks the

end of the Theme. There are many instances in the cat text to support the claim

that the topical Theme is the experiential element preceding the finite verb. In the

following fragment of text, the topical Theme (marked by double underlining)

can be tracked in this way.

(16) (a) Kissasta ei ole seuraa kävelyretkille,of cat is not company walking trips

(b) mutta toisaalta sitä ei myöskään tarvitse kuljettaa ulos.but on the other hand it not also need take out

‘A cat is no company on a walk but, on the other hand, it does not need to be taken out ei-ther.’ [24]

It should be noted that both of these examples (16a and 16b) are subjectless, and,

thus, the topical Theme cannot be equated with subject in Finnish. In traditional

grammars, clause 16a is considered to have an “existential subject” seuraa ‘com-

pany. In 16b, the verb is in the third person singular and there is no subject in the

clause: this corresponds to English ‘one need not take it out’. However, if there is

a subject in the clause, it is often the topical Theme, as in clauses 17b ) f in the

following fragment, where, once again, the topical Theme can be tracked by its

position immediately preceding the verb:

(17) (a) Kaikenlaista ikävää onkin aikojen saatossa sattunutall kind of sad (things) have during time happened

÷ (b) mutta myös koirat ovat purreet,but also dogs have bitten

÷ (c) hevoset potkineethorses kicked

÷ (d) ja pässit pökkineet.& rams butted

÷ (e) Silti koira on ihmisen paras ystävä,nevertheless dog is person’s best friend

÷ (e) hevonen jalo eläinhorse noble animal

(f) ja pässistäkin voi keksiä mukavaa sanottavaa.& about ram even can think of nice to be said

‘All kinds of distressing things have in fact occurred during the course of time, but it’s alsotrue that dogs have bitten, horses have kicked and rams have butted people. Nevertheless adog is a man’s best friend, a horse (is) a noble animal, and one can even think of somethingnice to say about a ram too.’[5 ) 6]

Page 333: systematic functional linguistic

325

I shall have little to say about what can be considered marked and what can

be considered unmarked in the Theme-Rheme structure, as it seems to me that

generalizations about what is textually marked in Finnish need to be based on a

detailed and comprehensive analysis of number of texts representing various gen-

res, and, moreover, more research needs to be done on information structure in

Finnish in order to ascertain the interaction of Given-New and Theme-Rheme

structures. In 17, for example, there are segmental indicators that information

structure is marked: in 17a the word order is marked and the focusing particle -kin

is attached to the finite verb and in 17f the topical Theme pässistä ‘about a ram’

is marked by the particle -kin. On the basis of these two examples, however, it

would be wrong to draw the conclusion that if the topical Theme is not realized as

the grammatical subject it is marked, the non-subject themes in the previous ex-

ample (16) are in no sense textually marked.

Given the discussion of information structure in the previous section, in the

approach taken here it is possible for the same element to simultaneously realize

both Theme and (contrastive) New. Thus, if Vilkuna’s K and T are given the

value of contrastive New and topical Theme respectively, then in some instances

K and T are conflated. Thus example 17f above, could be analysed as follows:

(18) ja pässistäkin voi keksiä mukavaa sanottavaa.& about ram even can think of nice to be said

Theme RhemeFocus

In this example, the topical Theme and contrastive New element (Focus) are con-

flated; in Vilkuna’s terms, this position could be labelled K/T, where the slash

indicates a conflation. Thus, in the SF approach, both the Given-New structure

and Theme-Rheme structure vary independently and can be conflated in different

ways. They combine with experiential and interpersonal structures to create an

integrated ‘polyphonic’ grammatical structure.

Page 334: systematic functional linguistic

326

An adjective can be the topical Theme in Finnish. In some respects, adjec-

tives in Finnish are grammatically more like nouns: for example, they inflect for

case and number and an adjective can be the Head in a phrase where English, for

example, typically requires a substitution element like one. Thus, the English

clause I’ll take the red one can only be translated into Finnish as Otan punaisen

‘I’ll take the red’. Because an adjective is more “substantive” in Finnish, if it oc-

curs in the pre-verb position is clearly the topical Theme in Finnish:

(19) Tärkeintä on, Û että kiellettyjä asioita ei ole niin monta,most important is that forbidden things not be so many

Û että kissaa täytyy jatkuvasti komennella. â âthat cat must be continually ordered about

‘It’s important that (/what is important is that) there are not so many no-nos that you have tobe continually ordering a cat about.’

In this particular example, the Rheme is an embedded clause; however, this is not

necessarily the case, as in the following example:

(20) Varmaa on vain muutos. [TT6/91:1]certain is only change‘The only thing that is certain is change.’

In the following sections, I shall consider some of the problems associated

with the recognition criterion for Theme stated above and examine some more

contentious cases. The discussion that follows rests on the claim that the finite

verb, in fact, marks the end of the Theme, and that other experiential elements

may precede the topical Theme. Some of the observations presented are only ten-

tative ) further research needs to be done on the analysis of running text and spo-

ken discourse.

Page 335: systematic functional linguistic

327

7.3.3. Subsidiary Topical Themes

If the topical Theme in Finnish is realized by the pre-verb position, then a prob-

lem arises with other experiential elements that come at the beginning of the

clause. In this section, I shall argue that it is necessary to recognize what might be

called a “subsidiary topical Theme”. Subsidiary topical Themes in Finnish may

further need to be classified into various subtypes along the lines suggested by

Downing (1991) for English.

The most common category of experiential element that precedes the topical

Theme in Finnish is a Circumstances of time and place, as illustrated by the fol-

lowing example from the cat text:

(21) Omakotitalossa kissa oppii pysymään omassa pihapiirissä,in detached house cat learns to stay in its own yard

jos se on pienestä pitäen tottunut siihen --if it has from when small accustomed to the fact that

‘In a detached house, on the other hand, a cat learns to stay within the confines of the yard,if, from the time that it is small, it is used to the fact that --‘ [28]

In the English translation I have added “on the other hand” which could be seen

as corresponding to Finnish toisaalta (or taas), which does not appear in the

Finnish text.

Since word order in Finnish is textually conditioned, the ordering of ele-

ments is textually relevant in Finnish. A Circumstance of place or time occurring

before the topical Theme in a clause has a cohesive function: it functions in much

the same way as a cohesive conjunctive such as on the other hand, however, etc.

in English. In Halliday & Hasan’s analysis of cohesive relations in English, the

conjunctive elements “presuppose the presence of other components in the dis-

course” (1976: 226). Of course, any element presupposes the presence of other

components in the discourse, but Halliday & Hasan’s point is that a conjunctive

element explicitly marks a stretch of text as being related to another stretch of

Page 336: systematic functional linguistic

328

text. There are cohesive conjuncts in Finnish, as illustrated by the silti ‘neverthe-

less’ in example 17 above, part of which is repeated as 22 below:

(22) (e) Silti koira on ihmisen paras ystävä,nevertheless dog is person’s best friend

(e) hevonen jalo eläinhorse noble animal

(f) ja pässistäkin voi keksiä mukavaa sanottavaa.& about ram even can think of nice to be said

‘Nevertheless a dog is a man’s best friend, a horse (is) a noble animal, and one can even thinkof something nice to say about a ram too.’[6]

In this example, the conjunct silti ‘nevertheless’ creates a cohesive tie with the

preceding text. It is my general impression that cohesive conjuncts are not as fre-

quent in a Finnish text as they are in English. If this is true, part of the reason for

this is the fact that these initial Circumstances of time and place function like

cohesive conjuncts in that they too “reach out into the preceding or following

text” (Halliday & Hasan 1976: 226).

In order to illustrate how the clause-initial Circumstantial omakotitalossa

‘in a detached house’ in example 21 above functions cohesively, one needs to

look at the preceding context, which was about looking after a cat in an apartment

house. When the constituent kaupunkiasunnossa ‘apartment house’ is introduced,

it is preceded by a focusing particle jopa ‘even’: ‘Its [a cat’s] need for exercise

gets to be satisfied even in an apartment house’. This ‘sets the stage’ for the con-

trast that is later made with taking care of a cat in a detached house:

Liikunnantarpeensa se saa tyydytetyksi jopa kaupunkiasunnossa, tarpeensa se suorittaalehtisilpulla tai muoviritilällä pohjustettuun vatiin. Kun hoito tapahtuu omassa kodissa, eilemmikkinsä takia joudu kiistaan muiden kanssa. Se onkin aikamoinen helpotus, kun tietääkuinka usein koirantaluttajat saavat ulkopuolisten kiukut niskoilleen. Omakotitalossa kissaoppii pysymään omassa pihapiirissä, jos se on pienestä pitäen tottunut siihen .. [25 ) 28]

Page 337: systematic functional linguistic

329

‘Its [a cat’s] need for exercise gets to be satisfied even in an apartment house, its business canbe done in a basin lined with paper shreds or with a piece of perforated plastic at the base.When things are taken care of in your own home, then you’re unlikely to get into a quarrelwith others because of your pet. This is quite a relief when you know how often people outwalking dogs are subject to the abuse of outsiders. In a detached house, on the other hand, acat learns to stay within the confines of the yard, if, from the time that it is small, it is used tothe fact that --‘

The constituent omakotitalossa ‘in a detached house’, however, is not simply a

textual element like silti ‘nevertheless’, which does not have a function in the

experiential structure of the clause. It functions both as a Circumstance of place

in the experiential structure of the clause and it functions cohesively. For this

reason, it seems clear that we need to recognize two themes in Finnish: a subsid-

iary topical Theme (Th: sub, with single underlining) and the main topical Theme

(Th: top, with double underlining).

(23) Omakotitalossa kissa oppii pysymään omassa pihapiirissä,in detached house cat learns to stay in its own yardTh: sub Th: top

‘In a detached house, on the other hand, a cat learns to stay within the confines of its ownyard --‘ [28]

The subsidiary Theme brings in another tier of experiential meaning into the

Theme in Finnish. Downing (1991) makes similar claims for English, and, based

on Chafe (1976), refers to this extra tier of experiential meaning as a “frame-

work”, and distinguishes different kinds of framework for English.

Another example of the cohesive function of a clause-initial Circumstance is

in the following extract from an advertisement that was commissioned by the

Finnish Work Union (Kotimaisen Työn Liitto). The initial Circumstances are

underlined.

Page 338: systematic functional linguistic

330

(24) a. Viime vuosikymmenellä me suomalaiset kulutimme ennätysmääränduring the last decade we Finns consumed record quantityTh: sub Th: top

ulkomaisia tuotteita ja palveluja ....foreign products & services

‘During the last decade we Finns consumed a record quantity of foreign products andservices ...’

b. Tänään ulkomainen nettovelkamme jokaiselle suomalaiselle jaettunatoday our nett foreign debt divided for each FinnTh: sub Th: top

on 25.239 markkaa ....is 25 239 marks

‘Today our net foreign debt divided amongst all Finns (i.e. per capita) is 25 239marks.’

c. Samaan aikaan suomalaisten tuotteiden ja palvelujen kysyntäat the same time the demand for Finnish products & servicesTh: sub Th: top

on laskemassa ...is falling

‘At the same time the demand for Finnish products and services is falling.’

d. Nyt suuntaa pitää muuttaa .... [HS 4.11.91 p. B16]now direction must changeTh: sub Th: top‘The direction/trend must now be changed.’

The advertisement is obviously intended to encourage people to buy Finnish pro-

ducts, and thus the topical Themes in the sentences are related to the topic of the

paragraph: “Finns”, “our net debt per capita”, “the demand for Finnish products

and services” and “(future) directions”. The circumstances of time that I have

translated as “the last decade”, “today”, “the same time” and “now” have both a

textual function ) they “reach out into the preceding or following text” ) and a

function in the experiential structure of the clause. Finnish nyt ‘now’ in the last

sentence does not simply mark a step in the argument, it also means ‘at this time’.

Page 339: systematic functional linguistic

331

There are also experiential elements preceding the topical Theme that are

not circumstances of time or place. In the following example, the first constituent

in both clauses is a Goal and this is followed by an Actor.

(25) Liikunnantarpeensa se saa tyydytetyksi jopa kaupunkiasunnossa,its need for exercise it gets satisfied even in city houseTh: sub Th: top

tarpeensa se suorittaa lehtisilpulla tai muoviritilällä pohjustettuun vatiin.its needs it carries out paper shreds or plastic grill based basinTh: sub Th: top

‘Its need for exercise gets to be satisfied even in an apartment house, its business can be donein a basin lined with paper shreds or with a piece of perforated plastic at the base.’ [25]

Here the subsidiary Theme does not introduce a temporal or spatial frame for the

topic, but highlights some aspect of it, and this is reinforced ) in this particular

instance ) by the possessive suffix (-nsa ‘its’) attached to the subsidiary Theme.

This kind of subsidiary Theme approximates what Downing (1991) refers to as an

“individual framework”.

In example 25 above, there is a contrast between liikunnantarpeensa ‘its

need for exercise’ and tarpeensa ‘its needs’ and this creates a cohesive tie be-

tween the clauses. This is reinforced by the lexical repetition of tarpeensa ‘it

needs’. However, it seems to me that a subsidiary Theme in Finnish is not neces-

sarily cohesive. There are instances where an initial experiential element is more

like an “individual framework” that highlights some aspect of the topical Theme,

and in these instances the element need not have a cohesive function. It appar-

ently occurs in initial position because it presents contrastive New information:

(26) Sen turkista irtoava karva on niin näkymätöntä,its from coat falling hair is so invisibleTh: top

÷ että ylimääräistä siivoustyötä se ei juurikaan aiheuta.that extra cleaning work it not hardly+FOC cause

Th: sub Th: top

‘The hair that falls off its coat is so invisible that it scarcely causes any extra cleaningwork at all.’ [75]

Page 340: systematic functional linguistic

332

The word order in the second clause together with the focusing particle -kaan

(negative of -kin) suggests a marked information structure. The first constituent

ylimääräistä siivoustyötä ‘extra cleaning work’ (realizing the Goal in the experi-

ential structure) is information that is being focused on, information that the

reader is being asked to attend to. It is followed by the topical Theme se ‘it’,

which is Given since it anaphorically refers back to the topical Theme of the pre-

vious clause (‘the falling hair’). The initial position of the Goal (ylimääräistä

siivoustyötä ‘extra cleaning work’) can possibly be explained by the fact that it is

‘in the air’, as Halliday (1985a: 277) puts it, or, it is in some sense evoked (in

Prince’s (1981) terminology). If there is hair falling from a pet that is kept inside

then the notion of ‘extra cleaning work’ would certainly be ‘in the air’ in a Finn-

ish context. Of course, in an implicit sense, a response to something that is

evoked by the previous text could be seen as cohesive, but it is not as explicitly

cohesive as the use of a conjunctive adjunct or the use of the temporal or spatial

subsidiary Themes discussed above.

There are another two similar instances in the cat text where a subsidiary

Theme is not explicitly cohesive, but in initial position because it presents

contrastive New information: they occur in sentences 29 and 31. In both in-

stances, the word order and the use of a focusing element indicate a marked infor-

mation structure. Sentence 31 will not be discussed at this point, as it brings in a

problem that will be addressed in section 7.3.4. In sentence 29, the topical Theme

of the first clause tämä ‘this’ refers to the previous sentence, which is about a cat

learning to keep within the confines of its own yard.

(27) Jos tämä ei onnistu, voi aluksi rakentaa ulkoilutarhan,if this doesn’t succeed can first build outdoor+cage

Th:top

÷ niin vieraalta kuin ajatus voi tuntuakin.as strange as thought can seem+FOC

Th: sub Th: top

‘If this doesn’t succeed, (then) in the beginning you can build an outdoor cage as strange asthe idea may in fact seem.’ [29]

Page 341: systematic functional linguistic

333

1 One cannot discount the possibility that it could be both cohesive and focused.

The topical Theme in the second clause ajatus ‘thought’ refers back to the build-

ing of an outdoor cage. The subsidiary Theme focuses on an aspect of the topical

Theme: an adverse reaction to ‘this thought’.

In the instances discussed, the subsidiary Theme is either cohesive or fo-

cused in the information structure1. Elements that are focused are clearly different

from clause-initial Circumstances that have a cohesive function. More research on

these elements may indicate that we may need to distinguish two (or more) sub-

types of subsidiary Theme, or that we are dealing with two distinct phenomena

that should not be subsumed under the one function “subsidiary Theme”. It is

clear that the research needs to be based on the analysis of running text. If we

simply look at isolated sentences then it is difficult to make any kind of claim

concerning ‘what the clause is about’ or the kind of framework that is being set

up by an initial experiential element. Only by looking at a running text can we

relate the textual structure in the clause to the overall thematic development in the

text.

The discussion of subsidiary Themes that are focused in the information

structure should not be seen as implying that the topical Theme in Finnish is nec-

essarily Given, and that (contrastive) New and topical Theme cannot be conflated

in Finnish. An example of a topical Theme conflating with New was given in 18

above: ja pässistäkin voi keksiä mukavaa sanottavaa, which following the con-

stituent order in Finnish would be: “and even about a ram (one) can think of nice

things to say”. There is no subject in Finnish that corresponds to the English

“one”; the constituent pässistäkin ‘even about a ram’ has a focusing particle at-

tached to it. Other examples were given in section 7.2.3 at the beginning of this

chapter, for example, mäkin lähen lenkille ‘I’ll go for a jog too’ (p. 317), Kurki-

kin jo lähti ‘Even the crane has left’ (p. 318).

Page 342: systematic functional linguistic

334

7.3.4. Topical Themes in Post-Verbal Position

In a polar interrogative, the topical Theme follows the finite verb. As pointed out

in 4.5.4, a projected interrogative remains in the interrogative in both quotes and

reports in Finnish:

(28) Kissa vaistoaa erittäin herkästi,cat feels instinctivelyextremely sensitively

2 onko se kodissa todella hyväksytty. 2is++QUES it in home really accepted

‘A cat has an extremely sensitive instinct concerning whether or not it is really accepted in ahome.’ [56]

The finite verb and the interrogative suffix (-ko) can be considered interpersonal

points of departure. These, and other interpersonal elements at the beginning of

the clause will be discussed more fully in section 7.3.6 below.

More problematic are those instances in which the topical Theme follows

the finite verb in a declarative clause. There are a number of instances in the cat

text where the finite verb precedes other experiential elements in the clause. As it

is more usual for the finite verb to follow an experiential element, this verb-initial

word order is marked in Finnish. It is clearly tied in with the system of key (al-

though the information structure may also play a part). A finite verb in clause-

initial position in a declarative clause in Finnish has the feature ‘assertive’: it

either asserts or repudiates something that has been said, one’s own expectations,

or an implicit assumption (cf. Hakulinen 1989: 59).

(29) Mutta eipä silti, eivät kaikki suomalaiset ole ainabut not nevertheless not (3pl) all Finns have always

ajatelleet, että kissan paikka on vain ulkorakennuksissa hiirenpyynnissä.thought that cat’s place is only in outbuildings in mouse catching

‘But on the other hand (for all that), not all Finns have always thought that the only place fora cat is in outbuildings catching mice.’ [8]

Page 343: systematic functional linguistic

335

(30) Olihan seitsemällä veljekselläkin oma temppujahad seven brothers own tricks

tekevä Matti-kissa, jolla oli turvattu asema perhepiirissä.performing Matti-cat who had a secure position in the family circle

‘After all even the Seven Brothers had their own trick-performing cat called Matti, whohad a secure position in the family circle.’ [9]

Both these sentences come at the beginning of the cat text, where the writer dis-

cusses people’s negative attitudes to cats. She then goes on, first, to disclaim in

part what she herself has written in example 29, and then assert a counterclaim in

30 ) the reference to the ‘seven brothers’ is to the first novel written in Finnish:

“The Seven Brothers” by Aleksis Kivi, which can assumed to be Given in a Finn-

ish cultural context. The counterclaim is, thus, that even people as significant in

Finnish culture as the seven brothers had a cat.

There are also instances in the cat text where there is an experiential element

at the beginning of the clause and this is followed by the verb. However, the finite

and predicator are split by another experiential element. In these instances, the

marked word order seems to be the result of a marked information structure.

(31) Erityisesti pikkulintujen lentoharjoitusten aikaan on kissan liikkumistaespecially during small birds flying practice time is cat’s movements

rajoitettava.to be restricted

‘When small birds are learning to fly is one particular time when a cat’s movements needto be restricted .’ [31]

The constituent that I consider to be the topical Theme, kissan liikkumista ‘a cat’s

movements’ comes between the finite element and a participle. From the point of

view of the thematic development of the text, it is clear that kissan liikkumista ‘a

cat’s movements’ is the topical Theme: the previous sentences were about a cat’s

need for exercise and its movements in general. Moreover, the Rheme is realized

by the participle rajoitettava ‘to be restricted’, and this, in fact, follows the

Theme. There is a marked information focus on erityisesti pikkulintujen

Page 344: systematic functional linguistic

336

lentoharjoitusten aikaan ‘especially when small birds are learning to fly’ (which

is an NP in Finnish) and, this appears to be the reason why the topical Theme gets

shifted to a position between the finite and the predicator.

The following example is in some respects reminiscent of an English struc-

ture where an element is introduced by as for or as regards and, in other respects,

it is like a predicated Theme structure in English (see Halliday 1985a: 60ff.):

(34) Leikkaamattomien kollikissojen hajumerkit eivät nekäänuncastrated tomcats’ scent marks not (3pl) they + focusing particle Th:subi Th: topi

edistä naapurisopua.promote neighbour+harmony

‘The scent marks of uncastrated tomcats do not promote neighbourly love ~ As for thescent marks of uncastrated tomcats they do not promote neighbourly love.’

The preceding sentence is about a cat’s sexual instinct and the fact that its wailing

at night is not the most pleasant thing to listen to. In this context, the first constit-

uent is a New element and it is picked up in an equative-like structure by a co-

referential pronoun that follows the finite verb. It seems to me that a structure

such as this is marked in Finnish, and, moreover, it seems to be restricted to cer-

tain genres. Unlike the English as regards structure, structures such as 34 are

more characteristic of informal registers, and would be unlikely to occur in offi-

cial or scientific prose.

7.3.5. Theme and Topic-Worthiness

Unlike English, Finnish does not employ “dummy elements” like it and there in

clauses like It’s easy to run out of money or There’s a glass on the table. The

Finnish translation equivalents to these could be literally glossed as “Easily hap-

pens thus that money runs out” and “On the table is a glass”:

Page 345: systematic functional linguistic

337

(32) Helposti käy niin, että rahat loppuvat kesken.easily happens so that money ends in the middle‘It’s only too easy to run out of money.’

(33) Pöydällä on lasi.on table is glass‘On the table is a glass ~ There’s a glass on the table.’

This means that a circumstantial element can occur in the pre-verb slot in spite of

the fact that it may not be significant for the thematic progression of text. It be-

comes the Theme by “default”, since its position at the beginning is tied to the

information structure of the clause rather than to thematic progression.

However, not all “default Themes” are topic-worthy. The most likely candi-

dates for topic-worthiness are circumstances of time or place (such as pöydällä

‘on the table’ in 33), since times and places constitute things that we talk about.

The following clauses could be glossed as being about ‘Saturday’ and ‘Sunday’:

(34) Lauantaina on mahdollisuus tutustua pitkäkarvarotuihin,on Saturday is chance to get acquainted with long-haired breeds

sunnuntaina esitellään lyhytkarvakissoja.on Sunday are presented short-haired cats

‘On Saturday it’s your chance to get acquainted with long-haired breeds, on Sunday youcan get introduced to short-haired cats.’

There are a number of instances in the cat text where the element in the pre-

verb position is not particularly topic-worthy, and, moreover, it is difficult to re-

late it to the general topic of the text. In the following example, the issue of he-

redity factors is introduced in the Rheme as new information, and the paljon ‘a

lot, much’ comes at the beginning of the clause.

(35) Paljon riippuu perintötekijästä ..a lot/much depends on heredity factor‘A lot depends on heredity factors’ [60]

If one were to ask someone what this clause is about, it is unlikely that the answer

would be paljon. Since the position of paljon is motivated by the information

Page 346: systematic functional linguistic

338

structure, it seems to me that we can either say that there is no topical Theme in

this clause or that paljon is the topical Theme by default.

With a Circumstances of Manner, such as 32 above (“Easily happens thus

that money runs out”), however, the situation is different: it seems to me that it

can be discounted as a topical Theme, even by default. If the Circumstance of

Manner is immediately followed by the finite verb, the clause often contains an

embedded projection (see Chapter 4):

(36) Kummallisen sitkeästi elää käsitys,strangely persistently lives view

että kissa saattaa repiä silmät päästä ...that cat might scratch eyes from head

‘How strangely persistent is the view that a cat might scratch a person’s eyes out ..’ [4]

Another equally valid translation for this example would be: ‘There is a strangely

persistent view that a cat might scratch a person’s eyes out’, since 36 represents

the most natural way to render this alternative translation back into Finnish.

Since example 36 is not clear-cut, I informally tested the intuitions of others.

I asked them what they thought this sentence was about, and, at the same time, I

asked them about sentence 17 in the cat text, which to me is a clear-cut example.

With the clear-cur example (sentence 17), people came up with the answer I ex-

pected, but with example 36 above, they usually hesitated for a while and then

they said something like “it’s about people’s views (about cats)” or “it’s about

people’s strange views”. In instances like this, it seems to me that we need to look

at the whole complex and the relationship between the host clause and the embed-

ded clause. As discussed more fully in 7.4.2, the host clause can be considered to

be the Theme, while the embedded clause is the Rheme. In the cat text, a clause

Theme is marked as follows:

Page 347: systematic functional linguistic

339

(37) Kummallisen sitkeästi elää käsitys, [Theme in embedded complex]strangely persistently lives view

että kissa saattaa repiä silmät päästä ...that cat might scratch eyes from head

‘How strangely persistent is the view that a cat might scratch a person’s eyes out ..’ [4]

As indicated, the embedded clause can also be analysed for topical Theme.

7.3.6. Clause-Initial Interpersonal and Textual Elements

So far I have ignored interpersonal and textual elements at the beginning of the

clause. For English, Halliday (1985a) considers initial interpersonal and textual

elements to be part of the “theme”, although it is somewhat difficult to conceive

of the majority of these in terms of what the clause is about (from either an inter-

personal or textual viewpoint). Halliday’s alternative characterization as “the

point of departure of the message” seems more applicable (cf. Downing 1991).

However, if it is conceded that clause-initial textual and interpersonal elements

are “points of departure”, then, for Finnish, we need to determine whether these

points of departure are at the beginning of the clause because of the Theme-

Rheme structure, or because of the information structure.

The issue is complicated when there are a number of modal elements strung

throughout the clause. In these instances, it is difficult to say in what sense the

first interpersonal element is the interpersonal point of departure except by virtue

of the fact that it comes first:

(38) [eiks tääl ollu aika vaarallista asua sodan aikana? [SIIIM2d: 16]‘wasn’t it quite dangerous to live here during the war?’]

.. kyllähän se tietysti kai tavallaan vaarallista oli .. sure+TIS it-NOM certainly perhaps in a way dangerous+PAR be+PAS/3SG

‘Sure it certainly perhaps was sort of dangerous.’

Page 348: systematic functional linguistic

340

The same question can be applied to Halliday’s (1981: 26) example from English:

Surely they can’t possibly be serious about it d’you think?

The issue of the textual status of clause-initial textual and interpersonal ele-

ments is best left as an area for future research. In the cat text, a textual or inter-

personal element that comes before the finite verb and a finite verb in clause-ini-

tial position is simply marked in cursive type:

(39) Lue tämä ennen kuin otat kissan.read this before you take a cat.‘Read this before you get yourself a cat.’ [Title of the cat text.]

(40) Koskaan kissaa ei pitäisi ottaa noin vain mistä vain,never cat should be taken from anywhere at all

sillä silloin onnistuminen on aina epävarmaa.as then success is always uncertain

‘as then/in that event success is bound to be uncertain.’ [61]

In content questions, with a mi- or ku- (‘wh’) element, an interpersonal element

can be conflated with the topical Theme:

(41) Mikä muu on yhtä vähäarvoista?which/what else is as little+valued‘What else is of as little value?’ [43]

Page 349: systematic functional linguistic

341

7.4. Theme in Clause Complexes

7.4.1. Relative Clauses

In Chapter 4, it was argued that all joka ‘who/which’ relative clauses are embed-

ded in Finnish, i.e. they are part of a constituent that functions in the structure of

the host clause. The fact that they are embedded does not diminish their textual

significance; on the contrary, it is often the case that the host clause introduces an

element in the Rheme and this becomes the topical Theme of the embedded clau-

se:

(42) Olihan seitsemällä veljekselläkin oma temppuja tekevä Matti-kissa,did have seven brothers own tricks performing Matti-cat

Û jolla oli turvattu asema perhepiirissä. âwho had secure position family+circle

‘After all even the Seven Brothers had their own trick-performing cat called Matti, whohad a secure position in the family circle.’ [9]

(43) Tuskin on olemassa ihmistä,hardly is existing person

Û joka suhtautuu välinpitämättömästi kissaan. âwho relates indifferently to cat

‘It’s rare for someone to be indifferent to cats.’ [1]

In example 42, the host clause introduces Matti-kissa ‘Matti-cat’ in the Rheme,

and this then becomes the topical Theme of the relative clause. In 43, which is the

first sentence in the cat text, there is no topical Theme in the first clause, but the

hypothetical ‘person’ that is introduced in the Rheme becomes the topical Theme

of the second clause.

Sometimes there are chains of relative clauses in which the Rheme of the

previous clause becomes the Theme of the subsequent clause.

Page 350: systematic functional linguistic

342

(44) ja sitten oli kolme poikaa& then was three boys‘and then there were three boys’

Û jolla yhellä oli semmonen .. pingispallo âone of whom (ADE) had kind of ping-pong ball‘one of whom had a kind of ping pong ball’

Û joka oli kiinnitetty semmoseen mailaan âwhich (NOM) was attached to a kind of racket/bat‘which was attached to a kind of racket’

Û jota se pompotti. â [from Helasvuowhich (PAR) he bounced (in press)]‘which he bounced’

In 44, the first clause introduces kolme poikaa ‘three boys’ in the Rheme. The

second picks up on this, but further specifies it: jolla yhellä ‘one of whom (i.e.

one of the three boys)’. In this clause and in the subsequent clauses, something is

introduced in the Rheme and this becomes the Theme of the following clause.

In spoken Finnish, the embedded clause generally comes at the end of the

host clause; the expansion of a nominal at the beginning of a clause, on the other

hand, appears to be more typical of written Finnish (cf. Helasvuo, in press):

(45) Monet, Û jotka vilpittömästi pitävät itseään eläinystävinä âmany who sincerely regard themselves as animal friends

vieroksuvat kissaa sen oletetun petomaisuuden takia.shy away from cat because of its supposed predatory nature

‘Many who sincerely regard themselves as animal lovers shy away from cats because oftheir supposed predatory nature.’

The initial nominal and the embedded clause in its entirety is the Theme of the

clause.

The topical status of the relative pronoun is assured regardless of its role in

the experiential structure since the relative pronoun joka can be inflected in Fin-

nish for number and case. As well as inflected forms of joka, there is a special

form that refers to a temporal antecedent: jolloin ‘when, at which time’ (cf. the

Page 351: systematic functional linguistic

343

interrogative milloin ‘when’). Because of this, it seems to me that a relative pro-

noun reverses the positions of subsidiary and topical Theme. This is evidenced by

many examples in the cat text:

(46) Kuitenkin on aina myös niitä,nevertheless is always also those

Û jotka henkeen ja vereen puolustavat tällaisten kissaparkojen oikeutta ..who with body and soul defend the right of these poor cats ...

‘Nevertheless there are always those who will defend with their body and soul the right ofthese poor cats ...’ [45]

In 46, the topical Theme of the second clause is introduced in the Rheme of the

first clause. As illustrated here, the subsidiary Theme comes between the relative

pronoun and the finite verb.

7.4.2. Clause as Theme

In 7.3.5 (p.339), it was pointed out that a host clause can be the Theme in an em-

bedded complex. In the instance discussed, there was a adverb of manner at the

beginning of the clause: (kummallisen sitkeästi ‘strangely persistently’). There

are similar instances in the cat text where an embedded clause follows a host

clause, but in these instances there is nothing before the verb that could be coun-

ted as topical Theme. Direct translations would render the clauses as “Is incom-

prehensible that ..” and “Is wonderful that ..”. As with the initial adverb of man-

ner, it seems to me that the host clause can be treated as the Theme in the clause

complex:

(47) On käsittämätöntä, [Theme in clause complex]is incomprehensible

Û että joidenkin mielestä ainoa todellinen kissa onthat in the opinion of some only real cat is

tappeluissa parkkiintunut ... kolli âin fights toughened tom

‘It’s incredible that some people think that the only real cat is a tomcat who has beentoughened in fights ...’ [35]

Page 352: systematic functional linguistic

344

1 This does not, of course, exclude other motivation for the sequencing of clauses in a clausecomplex.

(48) On ihanaa, [Theme in clause complex]is wonderful

Û kun kissa kulkee viskellen villiä häntäänsä, .. âwhen cat moves waving about wild its tail

‘It’s wonderful when a cat moves about waving its bushy tail about ..’ [22]

At the clause complex rank, the embedded clause is the Rheme, and it, in turn,

can be analysed for its Theme-Rheme structure. It seems to me that this kind of

many-layered analysis comes closer to capturing the complexity of meaning than

an analysis that looks at clauses and clause complexes as if they involved only

one strand of textual meaning.

7.4.3. Theme-Rheme in Hypotactic Complexes

In clause complexes involving hypotaxis (a dominant ) dependent relationship) it

is also feasible to analyse the clauses from a textual perspective, since, as pointed

out in Chapter 4, a feature of a hypotactic relationship between clauses is variati-

on in the order of the clauses. Either the dominant or the dependent clause can

come first. The fact that variation in word order is possible points to a textual

motivation for the sequencing of the clauses.1

In his analysis of dependent clauses in English, Halliday (1985a: 61 ff.)

assumes that at the clause-complex level the first clause is the Theme and the

second the Rheme. Similarly hypotactic conjunctions like jos ‘if’, kun ‘when’ and

vaikka ‘although’ in Finnish could be seen as setting up a clause-complex seman-

tic relationship that can be analysed thematically. At the clause-complex level, it

could be argued that one clause is what the whole complex is about:

Page 353: systematic functional linguistic

345

(49) Jos eläisimme keskiajallaif we were living in the middle-ages

$ (DEPENDENT CLAUSE) = THEME

sinunlaisesi ihmiset varmaan polttaisivat minut roviolla kissoineni.people like you for sure would burn me at the stake with my cat(s)

" (DOMINANT CLAUSE) = RHEME

‘If we were living in the middle ages, people like you would be sure to burn me at thestake with my cat(s).’ [15]

As this hypothesis needs to be further tested by looking at environments in which

the dependent clause follows the dominant clause and the kind of variation that is

possible, I have not marked hypotactic clauses in the cat text as Theme-Rheme

structures at the clause complex level.

7.5. Additional Topics

The rest of this chapter is concerned with a number of textual issues that have not

yet been addressed. The next two sections are concerned with what is generally

referred to as ellipsis. However, in what follows, a distinction is made between

ellipsis and (non)-realization in the linguistic structure. The following is an exam-

ple of ellipsis:

(50) a. ei se puhunut mitään jaNEG-3SG (s)he/it+NOM speak+PTC anything &

b. käski riisua puseron pois [TP1: 4]command-3SG take off blouse+GEN off

‘He didn’t say anything and told me to take my blouse off.’

In this example, the pronoun se ‘he’ in clause a is presupposed in the subsequent

clause. In the interest of clarity, ellipsis of the kind illustrated in 50 is distinguis-

hed from instances where an element is not realized in the linguistic structure,

since in instances of non-realization one cannot assume 1) that an element of

structure is “missing” and 2) there is nothing in the previous text that could be

said to be presupposed by ellipsis:

Page 354: systematic functional linguistic

346

(51) Väsyttääkö? [Ad3]tired + INTERROGATIVE SUFFIX

‘Tired?’

(52) En omista videoita (Vilkuna 1989: 93)NEG+1SG own videos‘I don’t own any videos’

It seems to me that it is misleading to say that sinua ‘you+PARTITIVE’ or minä ‘I-

NOMINATIVE’ is presupposed by ellipsis, as in the texts in which 51 and 52 occur-

red, there were no pronouns that could be presupposed by ellipsis.

Section 7.5.1 is concerned with the non-realization of an inherent role as

illustrated by 51. Section 7.5.2 discusses the realization of a first or second person

pronoun subject in Finnish, as indicated in 52, a pronoun is not necessarily reali-

zed in Finnish if the verb contains a bound morpheme that refers to the first or

second person. In section 7.5.3, I shall discuss two clause types that have been

discussed by Finnish linguists, but were ignored in the discussion of process types

in Finnish. As these clause types ) habitives and manifestation clauses ) are

clearly based on textual patterning, they are not considered to be process types in

this approach.

7.5.1. Non-realization of Inherent Human Participant

Halliday (1985a: 90) points out, that the non-realization of subject in English is

dependent on speech function. In a question or a command, the listener will un-

derstand the subject to be “you” whereas in a statement or an offer the subject

will be understood as “I”:

(53) a. Carry your bag? (Shall I carry your bag?)b. Met Fred on the way here. (I met Fred on the way here.)

(54) c. Seen Fred? (Have you seen Fred?)d. Play us a tune. (Will you play us a tune?)

Page 355: systematic functional linguistic

347

There are problems with this. As pointed out in Chapter 5 the speech function of

offer is problematic: thus in the offer Like a cup of tea? the assumed subject in

Halliday’s view would be I, which is not the case (Would you like a cup of tea?).

However, in spite of these problems, it is clear that Halliday’s basic insight is

valid: that the non-realization of an inherent role in the experiential structure of

the clause can only be “explicated” by reference to the interpersonal semantics,

i.e. by reference to the speech functions STATEMENT, QUESTION, and PROPOSAL

(see 5.2.4).

In the majority of instances in Finnish there is necessarily reference to ‘I’ or

‘you’ in the linguistic structure since the first and second person is marked as a

bound morpheme attached to the verb. However, in some process types ) e.g.

experiencer processes ) the verb is always in the third person singular in Finnish,

and thus, if the ‘I’ or ‘you’ is not realized as a separate pronoun, there is no indi-

cation of the speech act person in the morphology of the verb.

If an inherent role is not realized in Finnish, then it can be assumed that it is

either “you” or “I”. In a STATEMENT, where the initiator’s role is that of giver, it

can be assumed that the role that is not realized is ‘I’. For example, if someone

says “väsyttää”, which is a third person singular form, then we can assume that

the role that is not realized is an inherent (human) role, and, moreover, that it

refers to the speaker: minua ‘I-PARTITIVE’, which in Chapter 6 was referred to as

an Experiencer. In a QUESTION or a PROPOSAL, on the other hand, we can assume

the role is not realized is an inherent (human) role and that it refers to the addres-

see. Thus, if someone says väsyttääkö? ‘tired?’, it can be assumed that the role

that is not realized is sinua ‘you-PARTITIVE’, and similarly, Painu helvettiin ‘Go

to hell’ indicates that the non-realized role is sinä ‘you’.

The following are text examples in which this principle can be applied. The

first example below is from an advertisement for a pep-up pill, the second from a

comic strip, where the speaker is complaining of symptoms to a doctor:

Page 356: systematic functional linguistic

348

(55) Ärsyttääkö? Pänniikö? .. Väsyttääkö?irritate/annoy+3SG+Q annoy/fed up+3SG+Q tire+3SG+Q

Pätkiikö? .. Masentaako?cut+3SG+Q depress+3SG+Q

‘Irritated? Fed up? Tired? Lapses in memory? Depressed?’

(56) a. polvea särkee,knee+PAR hurt+3SG

‘My knee hurts/aches’

÷ b. viluttaa,make cold/chill+3SG

‘(I) feel chilly’

c. päänahkaa kutittaa [HS 27.11.90:D10]scalp+PAR make to itch+3SG

‘My scalp itches.’

In the first example, it is clear that the inherent role that is not realized is ‘you’,

the hypothetical reader of the advertisement. The second example consists of

STATEMENTS, and, thus, the inherent role that is not realized in 56b refers to the

speaker.

7.5.2. Non-Realization of Subject

In earlier chapters, I have argued that the grammatical subject in Finnish is the

element in the nominative case that agrees with the finite verb in person and, po-

tentially at least, in number, and that this element realizes the Medium. The Me-

dium can be realized as a separate NP (i.e. the subject) or it can be realized by a

bound morpheme (personal ending) attached to the stem of the finite verb. These

personal endings are not regarded as subjects: the subject is a nominal that is se-

parately realized. What I am concerned with in this section are the conditions

under which the Medium is realized morphologically and the conditions under

which it is (also) realized as separate nominal, i.e., the variation illustrated in the

following:

Page 357: systematic functional linguistic

349

(57) minä lähden ~ mä lähden ~ lähdenI-NOM leave+1SG I-NOM leave+1SG leave+1SG

‘I’m leaving’

The differences between these is not self-evident, and will be discussed shortly.

With third person verb forms, on the other hand, the pronoun or nominal subject

is realized (unless it is presupposed by ellipsis, as in the example above). Thus,

with the verb lähteä ‘(to) leave’, the subject would typically be realized in the

vast majority of genres:

(58) Pekka / hän / laiva / se lähteePekka (s)he ship (s)he~it leave-3SG

‘Pekka/he/ship/it is leaving.’

As indicated earlier, a first or second person pronoun is generally realized in

a short unstressed form in unselfconscious spoken Finnish (mä ‘I’, sä ‘you (sg)’,

me ‘we’, te ‘you (pl)’. In standardized written Finnish, the pronoun is generally

not realized. The following example is a quote attributed to a pop singer in an

article in the colour magazine of a national daily. It seems unlikely that speaker

did not begin with the pronoun mä.

(59) “Yritän muutamalla biisillä kuvatatry+1SG several+ADE piece/number+ADE describe

iskumeininkiä” [HSkl 19/91:14]casual sex scene+PAR

‘I try to describe the casual sex scene with a few of the numbers (on the album).’

However, as Helasvuo (1988: 67-68) points out, a pronoun can be realized

in written Finnish in a marked or contrastive environment, i.e. in SF terms, if it is

contrastive New.

Page 358: systematic functional linguistic

350

(60) Jos et sinä usko suomalaiseen työhön,if NEG-3SG you-NOM/SG believe Finnish+ILL work/craftsmanship-ILL

niin kuka sitten? [HS 4.11.91: B16]so/well who-NOM then

‘If you don’t believe in Finnish craftsmanship, then who will?’

There is a clear indication that the information focus is marked in this example:

the word order whereby the subject pronoun follows the negative auxiliary et puts

the focus clearly on sinä ‘you’, which is contrasted with kuka ‘who (else)’ in the

following clause. The following example illustrates how a similar environment

affects the realization of the pronoun minä ‘I’ in a written text. The example is

from a short article written by a well-known Finnish music critic. It’s about wo-

men in the Finnish Radio Symphony Orchestra wearing colourful clothes, instead

of the traditional black and white. The article, written in a light vein, describes

what some of the women were wearing. The writer continues:

(61) Uimapukufinaali käydään Rysän, Hohdin ja Carlsson-Parikan välillä Kaivopuistossa lau-antaina 30.5 kello 14. [HS 24.5.92:C6]

The bathing costume final is to be decided amongst Rysä, Hohti and Carlsson-Parikka andwill be held in Kaivopuisto Park at 2 p.m. on Saturday 30th May.

÷ Minä en siihen konserttiin mene,I-NOM NEG+1SG it+ILL concert+ILL go‘I shall not be going to the concert’

koska siinä on juontaja,radion johtaja Olli Alhobecause it+ESS is compere radio+GEN director-NOM Olli Alho

‘because it is being compered by the director of Finnish radio, Olli Alho.’

÷ Valtuutan täten hänet valitsemaan lopullisen Miss RSO:n.authorize+1SG hereby he+ACC choose+INF+ILL final+GEN Miss RSO+GEN

‘I hereby authorize him to choose the final Miss Radio Symphony Orchestra.’

Here the constrast is between minä ‘I’ and radion johtaja Olli Alho ‘radio direc-

tor Olli Alho’. If read aloud, the first syllable of the pronoun would be given tonic

Page 359: systematic functional linguistic

351

prominence. Note that in the final clause of example 61 there is no first person

pronoun, only a bound morpheme, as there is no contrast or no implied contrast.

In unselfconscious spoken Finnish, on the other hand, the pronouns are ge-

nerally realized. There are unstressed forms of singular first and second person

pronouns: mä (instead of minä) ‘I’ and sä (instead of sinä) ‘you (sg)’. With plural

forms, there are no shortened forms as they are already short: me ‘we’, te

‘you(pl)’.

(62) <A> mä en nyt o(le) pankissa ikinä ollut töissäI-NOM NEG+1SG now be bank+INE ever be+PTC work+PL+INE

mut mä oon nähnyt sitä siinäbut I-NOM be+1SG see+PTC it+PAR there

sen verran läheltäthat-GEN extent-GEN close-ABL

‘While I’ve never worked in a bank, I’ve seen it there from such a close perspecti-ve’

<B> joo .. mä en vois mennä pankkiin töihin ..yeah I-NOM NEG+1SG can+CON go+INF bank+ILL work+PL+ILL

‘Yeah I couldn’t go to work in a bank..’ [CA2:26]

The realization of an unstressed pronoun in spoken Finnish is only a general

tendency; and, moreover, there are some environments that are exceptions to this.

The clearest exception to this is when the finite verb is focused in a non-initial

move in an exchange:

(63) <A> sä oot kauheeyou-NOM/SG be+2SG awful/terrible-NOM

‘You’re awful’

<B> mä oon ihanaI-NOM be+1SG wonderful-NOM

‘I’m wonderful’

÷ <A> nii ootkii, joskus [CA6: 11]so be+2SG+TIS sometimes‘(And) so you are, sometimes’

Page 360: systematic functional linguistic

352

Interrogatives with vai that function as a “forward channel” (see 5.7) are another

clear instance where a second person pronoun is not realized. Here again the mo-

ve is a non-initial move:

(64) <A> mä oon lähös lenk(ille) .. [Tel2:1]I-NOM be+1SG leave(+INE) jog+ALL

‘I’m going for a jog.’

<B> no mäki lähen lenkillewell I+NOM+TIS leave+1SG jog+ALL

‘well I’ll go jogging too.’

÷ <A> lähet vaileave+2SG Q

‘are you?’

However, the distinction between informal spoken and standardized written

does not imply a simple dichotomy: “there are all sorts of writing and all sorts of

speech, many of which display features characteristic of the other medium” (Hal-

liday 1985b: 32). The generalizations above are a reflection of general tendencies

associated with standardized, formal written genres and with unselfconscious,

informal spoken genres. For example, in more formal spoken genres, a first or

second person pronoun need not be realized. The following was addressed by a

student to a professor in an oral examination of a licentiate thesis:

(65) Olet aivan varmasti oikeassa. [OH]be+2SG most certainly right+INE

‘You are most certainly right.’

Moreover, there are formal, standardized written genres for which the above ge-

neralizations about the realization of a speech act pronoun are not valid. The clea-

rest example that comes to mind is the Finnish Bible:

(66) Ja minä käänsin sydämeni tutkimaan viisautta ja tietoa, mielettömyyttä ja tyhmyyttä, jaminä tulin tietämään, että sekin oli tuulen tavoittelemista. [Saarnaja 1:17]

And I gave my heart to know wisdom, and to know madness and folly: I perceived thatthis also is vexation of the spirit. [Ecclesiastes 1:17]

Page 361: systematic functional linguistic

353

Given the importance of the Bible in the development of literary Finnish (see 3.2,

p. 72), one cannot overlook its influence on the linguistic intuitions of speakers of

Finnish.

However, if we contrast standardized written Finnish with unselfconscious

spoken Finnish, then to use the examples given at the beginning of this section,

spoken informal (i) mä lähden ‘I’m leaving’ corresponds to standardized written

(ii) lähden ‘I’m leaving’. This variation poses problems for the way in which the

topical Theme can be identified in Finnish. If the topical Theme is the experien-

tial element in the position immediately preceding the verb, then we are forced to

say that the topical Theme in the spoken variant (i) is mä ‘I’, whereas in the writ-

ten variant, there is no Theme. The situation is even more complex, if there is

another experiential element at the beginning of the clause: spoken maanantaina

mä lähden Helsinkiin ‘On Monday I’m going to Helsinki’ corresponds to written

maanantaina lähden Helsinkiin ‘On Monday I’m going to Helsinki’. First and

second person pronouns, i.e. speech act pronouns, also need to be contrasted with

third person pronouns, which, as pointed out earlier, are typically realized unless

they are presupposed by ellipsis.

The problem can be illustrated by looking at some actual texts. The follo-

wing example is a transcription of an informal interview (quoted by Helasvuo

1987: 66). The text occurred in the middle of a long turn that was a response to

the question mitä sä harrastat? ‘what kind of pastimes do you have?’.

Page 362: systematic functional linguistic

354

1 The vowel in mä oon ‘I be-1SG’ was, in fact, marked short in several places in the text bythe transcriber. I do not know the reason for this.

(67) a. mun yks harrastus oli yhtee aikaaI+GEN one/a+NOM hobby+NOM was one+ILL time+ILL

villapukujen kutominenwool+dress+PL+GEN knitting-NOM

‘A hobby of mine was at one stage the knitting of woollen dresses’

b. et mä oon1 käsin saanut kaks villapukua aikaseksso/thus I-NOM be hand+INS get+PTC two-NOM wool+dress+PAR ready+TRA

‘and so I’ve completed two woollen dresses by hand’

c. toista mä tein puoltoista vuottaother+PAR I-NOM make+PAS+1SG one & a half year+PAR

‘one took me one and a half years’

d. toisen mä sain jo vuodessa valmiiksiother+GEN I-NOM get+PAS+1SG already year+INE ready+TRA

‘the other took me only a year to make’

Personal interviews such as this as well as much of our everyday conversation are

about people (you or me or someone else) and the people that figure in the world

of our experience often constitute the topic of our everyday talk. Personal inter-

views are about ‘me’, the person being interviewed, hence the presence of the

unstressed pronoun mä ‘I’ in the position preceding the verb. However, subsi-

diary experiential Themes are also woven into the thematic development of the

text. In the example above, the knitting of woollen dresses is introduced as the

Rheme of the first clause, whose topical Theme is mun yks harrastus ‘a hobby of

mine’. The speaker then mentions two dresses that she has knitted and they beco-

me the subsidiary Theme of the subsequent clauses. The subsidiary Themes in

67c and 67d, which are Goals in the experiential structure of the clause in Fin-

nish, also function cohesively: there is repetition of toinen ‘other’. (Repetition is

one type of cohesive tie recognized by Halliday & Hasan (1976) and Halliday

(1985a: Ch. 9).

The topical Theme in three of the clauses in the text above is the unstressed

pronoun mä ‘I’ (which is also cross-referenced on the verb) and there is also refe-

rence to the speaker in the first clause (mun ‘my’). The thematic development in

the text does not rest on this topical Theme, since it is simply repeated throughout

Page 363: systematic functional linguistic

355

the text. We could either say that it is taken for granted or that it is backgrounded.

The pronoun would not be realized in a standardized written variant of the text

above:

(68) a. olen käsin saanut kaksi villapukua aikaseksibe-1SG hand+INS get+PTC two-NOM wool+dress+PAR ready+TRA

‘I’ve completed two woollen dresses by hand’

b. toista tein puolitoista vuottaother+PAR make+PAS+1SG one & a half year+PAR

‘one took me one and a half years’

c. toisen sain jo vuodessa valmiiksiother+GEN get+PAS+1SG already year+INE ready+TRA

‘the other took me only a year to make’

This presents a problem for the recognition of topical Theme, since the element

preceding the verb in 68 above is the element that was the subsidiary Theme in

the spoken variant (67).

The following written text is taken from Vilkuna (1989: 93), who has taken

it from a paragraph in an interview published in a women’s magazine. According

to Vilkuna (1989: 93), who analyses the text in terms of a dialogue games appro-

ach (see Carlson 1983, 1984), the overall topic of the text is “What is my opinion

about different technical gadgets?”

(69) a. En omista videoita enkä kolvia,NEG+1SG own video+PL+PAR NEG+1SG+TIS soldering iron+PAR

‘I don’t own any videos or a soldering iron,’

b. olen käynyt kielilinjan jabe+1SG go+PTC language+line/strand+GEN &

‘I studied languages at high school and

c. nauttinut ainoastaan humanististen tieteiden koulutusta.receive+PTC only humanistic+GEN/PL science+PL+GEN education+PAR

only subjects in the humanities at university.’

d. Kodinkoneisiin suhtaudun erittäin varauksellisesti.house+GEN+appliance+PL+ILL relate+1SG extreme reservedly‘Household appliances are things I relate to with extreme reserve.’

Page 364: systematic functional linguistic

356

e. Kellari ja auma ovat minusta kymmennen kertaacellar+NOM & pit+NOM be+3PL me+ELA 10 time+PAR

parempia keksintöjä kuin paraskaan pakastin better+PL+PAR invention+PL+PAR than best+TIS deep freeze+NOM

‘The (cold storage) cellar and (cold storage) pit are in my opinion far superior inven-tions than even the best deep freeze.’

f. Tiskausta rakastan jadish-washing+PAR love+1SG &

‘Washing dishes is something I love and

g. pyykkikoneen ostoa vastustinwashing+machine+GEN purchase+PAR oppose+PAS+1SG

the purchase of a washing machine met with my opposition’

h. koska nyrkkipyykki on minusta suloista liikuntaabecause fist+washing+NOM be-3SG me+ELA sweet+PAR exercise+PAR

ja suunnatonta terapiaa.& boundless+PAR therapy+PAR

‘because in my opinion washing by hand is a pleasant form of exercise and of immensetherapeutic value.’

i. Autoonkin suhtaudun nyrpeästicar+ILL+TIS relate+1SG with disapproval‘A car meets with my disapproval’

j. koska se on kallis ja turvatonbecause it is expensive & unsafe

k. Televisio, tosin pieni ja mustavalkoinentelevision-NOM to be sure/albeit small+NOM & black+white+NOM

meillä on pelkästään vaimon painotuksesta.we+ADE be-3SG only wife+GEN pressure+ELA

‘A television set ) albeit small and black-and-white ) we’ve got only as the result ofmy wife’s insistence.’

In Vilkuna’s view, in those clauses in which there is a first person form of the

verb “the dropped pronoun continues to act as T” and the initial NP is K (Vilkuna

1989: 94). As indicated earlier (p. 323), Vilkuna’s T position is associated with

what would be referred to as the topical Theme in the approach adopted here. The

K position is associated with contrastive New, but it seems to me that this is not

true of all instances. (Vilkuna is clearly aware of this, and, as a consequence, does

Page 365: systematic functional linguistic

357

1 If there are experiential elements occurring in Vilkuna’s K and T positions, and if theelement in the T position also carries tonic prominence, then this would indicate that Theme-Rheme and Given-New vary independently, and the realization of a contrastive New elementcannot be tied to a particular linear position.

not unambiguously relate these “syntactic positions” to textual functions realized

in the clause.)

While it seems to me that Vilkuna is correct in maintaining that the topical

Theme of these clauses is ‘I’, there is a problem with Vilkuna’s formulation that

“the dropped pronoun continues to act as T”. This formulation treats T as though

it were a textual function. The problem is complex, and so Vilkuna’s ambivalence

is understandable. If Vilkuna’s T is regarded as a syntactic position, the only re-

course is to compare the written with the spoken and assume that they correlate.

(70) Kodinkoneisiin mä suhtaudun erittäin varauksellisesti.household appliances I relate+1SG extremely reservedly‘Household appliances are things I relate to with extreme reserve.’

(71) Kodinkoneisiin suhtaudun erittäin varauksellisesti.to household appliances relate+1SG extremely reservedly‘Household appliances are things I relate to with extreme reserve.’

The setting up of a correlation between spoken and written would have to assume

that the spoken is the more basic, which seems to me a valid assumption given the

number of times claims to this effect are repeated in introductory linguistics text-

books. It is also necessary to make this correlation because if a pronoun had been

realized in a written text comparable to 69 above, it seems to me that it would

have to be contrastive (cf. 61, p. 350), although this can only be ascertained by

looking at the text environments where the following type of structure occurs:1

(72) Kodinkoneisiin minä(kin) suhtaudun erittäin varauksellisesti.household appliances I + (TIS) relate+1SG extremely reservedly

Focus ‘Household appliances are things I relate to with extreme reserve.’

Page 366: systematic functional linguistic

358

The problem with the Theme-Rheme analysis of first and second person can

be overcome if we acknowledge the special status of the 1. and 2. person (speech

act pronouns), since in Finnish these can be marked on the verb, i.e. realized as

bound morphemes. With a third person form such as suhtautuu ‘relate-3SG’ or

suhtautuvat ‘relate-3PL’, we need to specify the Entity or Entities doing the ‘rela-

ting’, whereas with suhtaudun [relate+1SG] ‘I relate’ (and with other 1. and 2.

person forms) the Entity is already specified in the morphology of the finite verb.

A solution would be to say that the topical Theme is typically realized by an ex-

periential element occurring in the position immediately preceding the finite verb,

except in the case of the first and second person, where it can either 1) be realized

by this position and also cross-referenced on the verb or 2) it is simply realized by

a bound morpheme attached to the verb. It seems to me that this is also justified

since it sets up a correlation between clauses with a finite 1. and 2. person verb

and experiencer and possessive clauses, where the pronoun is typically realized in

written Finnish since the verb is in the third person singular: minua suututtaa ‘I

feel angry’, minulla on päänsärky ‘I’ve got a headache’.

If the possibility that a (backgrounded) topical Theme is realized in the

morphology of the verb is entertained, then Vilkuna’s text (“What is my opinion

about different technical gadgets?”) could be analysed as follows to show how the

two Themes ) ‘me’ and ‘technical gadgets’ are intertwined in the thematic deve-

lopment of the text.

(73) a. En omista videoita enkä kolvia,

not+I own videos not+I+and soldering iron

b. olen käynyt kielilinjan jahave+I studied language+strand &

c. nauttinut ainoastaan humanististen tieteiden koulutusta.received only humanistic sciences’ education

d. Kodinkoneisiin suhtaudun erittäin varauksellisesti.household appliances relate+I extreme reservedly

Page 367: systematic functional linguistic

359

e. Kellari ja auma ovat minusta kymmennen kertaa parempia keksintöjä kuincellar & pit are to me 10 times better inventions

paraskaan pakastin even the best deep freeze

f. Tiskausta rakastan jadish-washing love+I &

g. pyykkikoneen ostoa vastustinwashing machine’s purchase oppose+I

h. koska nyrkkipyykki on minusta suloista liikuntaabecause hand washing is to me pleasant exercise

ja suunnatonta terapiaa.& boundless therapy

i. Autoonkin suhtaudun nyrpeästicar relate+I with disapproval

j. koska se on kallis ja turvatonbecause it is expensive & unsafe

k. Televisio, tosin pieni ja mustavalkoinentelevision albeit small & black+white

meillä on pelkästään vaimon painotuksesta.we have only of wife’s insistence

The subsidiary Themes in this example have a cohesive function: they are in a

hyponymic relationship with “technical gadgets”. (See Halliday & Hasan (1976)

and Halliday (1985a: Ch. 9) for a discussion of hyponymy as a cohesive rela-

tionship.)

Page 368: systematic functional linguistic

360

7.5.3. Habitive and Manifestation Clauses

Two clause types that Vilkuna (1989) refers to as habitive and manifestation clau-

ses were not considered in the discussion of process types in Chapter 6. These

clause types are grammatically heterogeneous, and, in the approach adopted in

this study, they are not considered to be clause types, but constitute simultaneous

textual options that can be conflated with transitivity (experiential) options. Habi-

tive clauses are concerned with the Given-New structure in the clause. Manifesta-

tion clauses group together clauses in which it would appear that a Circumstance

is not realized. Manifestation clauses, like the clauses discussed in 7.5.1 (e.g.

Ärsyttääkö? ‘Irritated?’), appear to be topicless.

In a habitive clause, there is a pronoun or a proper noun in the adessive,

ablative or allative at the beginning of the clause. This constituent is referred to as

a habitive. Habitive clauses are characteristic of spoken Finnish:

(74) mulla lähtee juna ihan kohta [Vilkuna 1989:171]I+ADE leave-3SG train+NOM quite soon‘My train is about to leave’

(75) minulta jäi kotityöt tekemättä [Vilkuna 1989:174]I+ABL leave-3SG house+work+NOM/PL do+INF+ABE

‘I didn’t get round to doing the housework.’

(76) onks sulla sitte miten ystäväpiiri rakentunube-3SG+Q you-SG+ABL then how friend+circle-NOM build+PASS-REF+PTC

‘Then have you been able to build up a circle of friends?’

(77) Janikal(la) oli kouluvaihe aikaJanika+A(DE) be-3SG/PAS school+stage-NOM pretty

ikäväs vaihees [CA8:12]sad/depressing+(INE) stage+(INE)

‘It was a pretty depressing for Janika at that stage at school.’

As indicated by these examples, habitive clauses do not constitute a grammatical-

ly or semantically coherent clause type. Nevertheless, according to Vilkuna

(1989: 173) the primary semantic (i.e. experiential semantic) motivation for a

habitive clause is to present the person in question as being primarily affected by

Page 369: systematic functional linguistic

361

what happens. Wierzbicka (1988: 381 ff.) makes a similar claim concerning the

function of the Polish dative. While this is clearly the case in Finnish, it is true

whether or not the NP in the adessive, allative or ablative is at the beginning or

the end of the clause. In the following example, it comes at the end:

(78) tää on muute iha hirveethis+NOM be+3SG besides quite terrible+NOM

tilanne mulla [Tel2: 6]situation+NOM I+ADE

‘besides this is a quite a terrible situation for me’

While a “habitive” presents someone as being primarily affected by what

happens, what is significant from a textual perspective is that the habitive is typi-

cally Given: in the examples above, it is either a speech act person (I or you) or it

is information that was referred to earlier in the text. Example 77 is taken from

well over half way through a late night radio talk-back programme in which the

person in question (Janika) had been a topic of conversation from the start. It

would be highly unusual for a clause like this to be realized in a text if this were

the first reference to Janika.

The textual function of clauses like this can be illustrated by comparing

them to the following b-clauses:

(79) a. mulla lähtee juna ihan kohta [Vilkuna 1989:171]I+ADE leave-3SG train+NOM quite soon‘My train is about to leave’

b. mun juna lähtee ihan kohtaI+GEN train+NOM depart/leave-3SG quite soon‘My train is about to leave’

(80) a. onks sulla sitte miten ystäväpiiri rakentunube-3SG+Q you-SG+ABL then how friend+circle-NOM build+PASS-REF+PTC

‘Then have you been able to build up a circle of friends?’

Page 370: systematic functional linguistic

362

b. miten sun ystäväpiiri on rakentunuhow your friend+circle-NOM be-3SG+Q build+PASS-REF+PTC

‘Then have you been able to build up a circle of friends?’‘

OR:

b. entä ystäväpiiri? miten se on rakentunuhow about friend+circle-NOM how it be-3SG+Q build+PASS-REF+PTC

‘How about friends. Have friendships been established?’

Clauses like this are similar to what Quirk et. al. (1985: 1411 ff.) refer to as the

have-existential device in English: You’ve got a button undone, I’ve got a car

blocking my way. These clauses function in a similar way to clauses with there:

There’s a button undone on your shirt, There’s a car blocking my way. As dis-

cussed earlier (section 7.3.5), there are no “dummy elements” in Finnish, and so

what happens is that an element that is typically Given comes at the beginning of

a clause. The rest of the clause is presented as New: juna lähtee ihan kohta ‘train

is leaving soon’ , miten ystäväpiiri rakentunu ‘how circle of friends built up’. In

the b-clauses the pronouns are in the genitive in a NP with juna ‘train’ and ystä-

väpiiri ‘circle of friends’ as Head. Thus, the nominals juna ‘train’ and ystäväpiiri

‘circle of friends’ are part of an NP that is Given. The habitive clause allows these

NPs to come as New elements at the end of the clause.

The conflation of Theme with the experiential meaning that Vilkuna refers

to is, of course, important. As pointed out earlier, much of our everyday conversa-

tion is about “you” or “me” or someone else, and the people that figure in the

world of our experience often constitute the topic of our everyday talk. These

clauses with a habitive in initial position also allow us to get the unmarked theme

right. The clauses above are not about “trains” or “circles of friends”, but about

“me” and “you”. Significantly, examples 76 (80a) and 77 are from the same radio

talk-back show (Yöihmisiä ‘Night People’; currently called Yölinja ‘Night Line’),

where people phone in to talk about their problems. It is clear from these texts

that people generally phone in to talk about themselves and the host of the prog-

ramme directs the conversation along these lines.

Page 371: systematic functional linguistic

363

Vilkuna (1989: 165 ff.) also distinguishes a subset of “existential sentences”

in Finnish (see 6.8), which she refers to as “manifestation sentences”. From the

perspective taken in this study, these are all-new clauses in which the finite verb

is clause-initial and this is followed by the Domain (see 6.8) of the process, what

the process (of existing, coming into existence etc.) extends to. The following

examples are taken from Vilkuna (1989: 165-166):

(81) Oli ukko ja akka.be-3SG/PAS old man-NOM and old woman-NOM

‘There was an old man and an old woman.’

(82) a. Syttyi sota.light/break out-3SG/PAS war-NOM

‘There was a war/A war broke out.’

b. Syttyi kapinoita.light/break out-3SG/PAS revolt/riot+PL+PAR

‘There were riots/Riots broke out’

(83) Seurasi melkoinen kaaos.follow-3SG/PAS considerable/quite-NOM chaos-NOM

‘There was considerable chaos to follow/What followed was considerable chaos/

(84) Vallitsi syvä hiljaisuus.reign-3SG/PAS deep-NOM silence-NOM

‘There reigned a deep silence/A deep silence reigned.’

These clauses are considered problematic because they are verb-initial: “was old

man and old woman”, “broke out a war”, “followed considerable chaos”, “reig-

ned a deep silence”. If an experiential role is added in Finnish, it would not be a

“dummy element” like there or it, but a Circumstance of time or place: Silloin

syttyi sota “At that time broke out a war” ~ Maassa syttyi sota “In the country

broke out a war”.

According to Vilkuna (1989: 167), clauses 81 ) 84 are about “the temporal

domain of the universe of discourse”; in terms of dialogue game theory, these

examples are responses to questions like “What (about) then?” On the other hand,

it could be possible to extend on Vilkuna’s basic insight and argue that these

clausess are about either the temporal or the spatial domain of the universe of

discourse and, by doing this, we can also relate these clause to a far more per-

Page 372: systematic functional linguistic

364

vasive textual phenomenon in Finnish. As Vilkuna (1989: 168) points out, there

are indefinite clauses in Finnish that are similar in this respect. For example:

(85) From a children’s book: Tämä on Koiramäen talo. [KT: 2]

‘This is Dog Hill house.’

Eletään 1800-lukua.live+INDE 1800-decade/century+PAR

‘This is life in the nineteenth century/It’s the nineteenth century.’

(86) [From the beginning of an article on Finnish design:]Puhutaan suomalaisen muotoilun kriisistä. [SK 14.1.83: 25]speak+INDE Finnish+GEN design+GEN crisis+ELA

‘There’s talk about the crisis in Finnish design/People talk about the crisis in Finnish de-sign.’

These clauses are also about the temporal or spatial domain of the universe of

discourse. These examples are from written texts and their interpretation is tied up

with the concept of genre and the place of the text in the context of culture. The

Finnish Design example (example 86) is from a Finnish weekly magazine that

contains cultural and political articles. Thus the assumed Circumstance of Locati-

on is Finland and the Circumstance of Time is the time that the article was writ-

ten. With the Dog House example (85), on the other hand, the context is being

constructed by the writer, and the text is accompanied by elaborate illustrations.

These examples are similar to the ambient and meteorological processes

discussed in Chapter 6, where there is no circumstance of location or time:

(87) On kylmä.be-3SG cold-NOM

‘It’s cold’

(88) On kaunista.be-3SG cold+PAR

‘It’s beautiful’

The interpretative source of the Circumstance of time or location is in the ‘here

and now’ of the situation: for example, täällä ‘here’, ulkona ‘outside’ or nyt

‘now’ in the examples 87 and 88 above.

Page 373: systematic functional linguistic

365

If we regard text as always occurring in a context and if we see this context

as either being of two kinds: 1) words that precede or follow and 2) the situation

that surrounds something that has been said or written, then the examples discus-

sed here are no longer problematic. It is only when we isolate clauses and look at

them as a decontextualized bit of text that the “missing” situation is problematic.

In other words, there is a linguistic context (i.e. words) and an extra-linguistic

context (i.e. the situation, context of culture) and the analysis of language cannot

be separated from this context.

7.6. Summary

This chapter has discussed some aspects of textual organization in Finnish. It was

postulated that the Given and New structure in Finnish is realized in the pitch

movement in the tone group, and although no phonetic analysis was carried out,

this provided the basis for discussing Given-New and Theme-Rheme as distinct

structures that vary independently in the clause. It was argued that, in most cases,

the finite verb marks the movement from Theme (‘what the clause is about’) to

Rheme, and that we need to recognize two experiential Themes in Finnish. The

recognition criteria established for Given-New and Theme-Rheme were based on

the formal properties of the clause (or tone group). This fact, together with the

discussion of the non-realization of elements of the situation in the linguistic

structure, underscores the importance of the textual metafunction in the grammati-

cal organization of a language. Any analysis of grammatical structure necessarily

includes the analysis of textual structures in the clause: a text is also organized

textually at the rank of clause. The co-text and context is not something that can

be tacked on after we have completed a morpho-syntactic analysis of the clause.

Page 374: systematic functional linguistic

367

Chapter 8

Conclusion

8.1. Overview

In this study I have attempted to present some aspects of a systemic-functional

grammar of Finnish. One test for the efficacy of a functional grammar is its ability

to provide a revealing analysis of a text, and, thus, it seems appropriate to conclu-

de this study with the analysis of a text fragment. It is hoped that the analysis

presented in the next section (8.2) will provide an integrated view of how the

various kinds of structures presented in this study are conflated in the clause.

It needs to be emphasised here that the present study is conceived of as no

more than an introduction to a systemic-functional analysis of Finnish. Even the

grammatical description of the best described language of the world ) viz. Eng-

lish ) could hardly be said to be complete. It seems to me that the notion of a

complete grammatical description of any language is chimerical since languages

are neither clearly defined systems nor closed systems. While I am well aware

that the present work is far from a reasonably comprehensive description of

clause-rank structures in Finnish, the aim of this study has been to open up a new

perspective on the grammatical analysis of Finnish.

In section 8.3, I provide a very brief indication of what is “beside and bey-

ond” the analysis presented here: I discuss other patterns of linguistic organizati-

on that are beyond the scope of this study in order to show how the analysis pre-

sented here fits into a wider and more comprehensive framework. In the final

section of this conclusion (8.4), I shall attempt to bring together a number of re-

curring themes in this study.

Page 375: systematic functional linguistic

368

8.2. The Grammatical Analysis of a Text

The fragment of text that will be analysed is from spoken Finnish. The syllable in

bold type indicates what I perceived as the tonic syllable, however, I shall ignore

the prosody and also the pause in the first tone group. I shall also ignore the

fourth clause: since it was spoken extremely quickly and softly. I had to rewind

the tape many times before I actually heard what was said.

(1) viime, viime kesänä ((pause)) lähettii kesälomale [AA1: 3]last last summer leave+INDEF+PAS summer+holiday+ALL

‘last, last summer we went on our summer holiday and

(2) ja välttämättä kaveri halus& necessarily mate/partner want+PAS

‘and my partner insisted’

(3) et täytyy olla sormusthat must-3SG be ring-NOM

‘a ring is necessary ~ that one/I must have a ring’

(4) (no sitte se on)(‘well then it is’)

(5) mie olin sillon hiukan, laihempi,I-NOM be+PAS+1SG then slightly slim/lean+COMP/NOM

ehkä kymmenen killoo laihempiperhaps ten kilo+PAR lean/slim+COMP/NOM

‘I was slightly slimmer then, ten or so kilos lighter’

(6) se etu on ollut avioliitosta [[ että miethat advantage-NOM be+3SG be+PTC marriage+ELA that I-NOM

oon saanu kunnolla syyäkseni, ((laughingly)) lihoakseni ]]be+1SG get+PTC form+ADE eat+INF+TRA+POS1SG get fat+INF+TRA+POS1SG

/properly

‘The advantage of marriage is that I get to eat properly, and get to put on weight.’

A multi-layered, “polyphonic” analysis of this fragment is given on the next page

p. 370. (For the convenience of the reader, there is a loose copy of this page en-

closed in the pocket on the inside cover of this study.)

Clauses 1 ) 2 form a paratactic clause complex, and the continuing clause,

clause 2, projects the third clause. These clauses, thus, represent something that

Page 376: systematic functional linguistic

369

the speaker sees as being interrelated. Clause 5 is a simple clause, which contains

a phrase complex: hiukan laihempi, ehkä kymmenen killoo laihempi ‘slightly

leaner/slimmer, ten or so kilos lighter’. The final clause contains an embedded

clause, i.e. a clause that functions in the structure of the host clause. (The infiniti-

ve forms in the embedded clause can be regarded as non-finite clauses.)

Apart from clause 3 (täytyy olla sormus ‘(one/I) must have a ring), each

clause in this fragment is in the declarative and congruently realizes a STATE-

MENT, i.e. the clause is basically oriented to the exchange of language and the

role of the speaker in this orientation is that of giver. Clause 3 is a projected clau-

se introduced by the projecting clause välttämättä kaveri halus. While clause 3 is

in the declarative, it is not oriented to the exchange of language, but to a non-

linguistic response (or to an action). This is clear from the finite verb in the pro-

jecting clause: the verb haluta ‘to want’ realizes a mental process that can be

subclassified as a reactive process, i.e. it represents a mental reaction (of fear,

joy, desire etc.) to some phenomena. When a reactive verb is followed by a pro-

jecting clause beginning with a modal like täytyy, then täytyy can only be modu-

lating, i.e. action-oriented.

Page 377: systematic functional linguistic

370

(1) viime, viime kesänä lähettii kesälomalelast last summer went on summer+holiday

1 (initiating)Mood: declarative

* Theme: top * Rheme* Circ * Pro: mat + Ac (indef) * Circ

(2) ja välttämättä kaveri halus (3) et täytyy olla sormus& necessarily partner wanted that must be ring

2 (continuing [projected complex]) Projecting clause Projected clause

Mood: declarative Mood: declarative (metaphorical realization of a PROPOSAL)* Th: top * Rheme 5 Rheme

* Sen * Pro: men * * Pro: rel: circ * Poss:ed *

[Phen ` clause 3 (metaphenomenon)]

(5) mie olin sillon hiukan laihempi, ehkä kymmenen killoo laihempiI was then slightly slimmer perhaps ten kilos slimmer

Mood: declarative

* Th: top * Rheme* Car * Pro: rel: int * Circ * Att (paratactic phrase complex) *

(6) se etu on ollut avioliitosta [[ että mie oon saanu kunnolla syyäkseni, lihoakseni ]]that advantage has been from marriage that I have got properly fed, toget fatMood: declarative

* Th: top * Rheme[[ Th: top * Rheme ]]

* Id:ed * Pro: rel: int * Circ * Id:er ]]

Page 378: systematic functional linguistic

371

1 The ring that is introduced in this clause appears to be the topical Theme of clause 4, notanalysed here. The ring is not mentioned again in this fragment, but is picked up immediatelyafterwards.

The topical Theme of the first clause is viime kesänä ‘last summer’, the spea-

ker is concerned with saying something that relates to last summer. The finite

verb is indefinite, and a Finnish indefinite often implies that there is more than

one participant, particularly if it is preceded by a circumstantial element. The

other participant is the topical Theme of the second clause: kaveri ‘partner’. This

is followed by the projected clause discussed above. There is no topical Theme in

this clause, as it is oriented to action.1 The topical Theme of the next clause is

mie ‘I’ (dialect form of minä ‘I’). Clause 6 is an embedded complex. The topical

Theme is se etu ‘the/that advantage’ and the rest of the clause (including the em-

bedded clause) is the Rheme.

From an experiential perspective, the first clause realizes a material process.

It constructs a material event, something that happened. This is grammatically

reflected, for example, in the fact that the finite verb could have been modified by

an adverb of manner ending in -sti, e.g. viime kesänä lähettii kiireisesti kesälo-

male ‘Last summer we hurriedly left for our summer holiday’. The second clause

in the fragment above realizes a mental process, one that is concerned with inter-

nal human consciousness and external verbalized consciousness) and this is ref-

lected in the fact that the following clause is projected. Clauses 3, 5, and 6 realize

relational processes: the finite verb is olla ‘to be’. In clause 3, one of the inherent

roles is not realized in the linguistic structure: täytyy olla sormus [must be a

ring]. As it realizes a projected PROPOSAL, we can assume that the participant is

the person being addressed, who is the speaker of this text (sulla täytyy olla sor-

mus ‘you’ve got to have a ring’). Clauses 5 and 6 are intensive processes: 5 is

attributive and 6 is identifying. They can be distinguished from each other by the

fact that in an attributive process the inherent roles are unaffected by changes in

word order: the roles of Carrier and Attribute are the same in the following agnate

clause: hiukan laihempi mie olin sillon ‘slightly slimmer I was then’. Because

Page 379: systematic functional linguistic

372

clause 6 contains an embedded clause, it is difficult to vary the word order. Ho-

wever, the following is a possible variant:

(7) se [[ että mie oon saanu kunnolla syyäkseni ]]it/that that I-NOM be+1SG get+PTC properly eat+INF+TRA+POS1SG

on ollut etu avioliitossabe+3SG be+PTC advantage-NOM marriage+INE

‘The fact that I get to eat properly is an advantage of being married.’

I am not suggesting that these clauses could be changed in the original text, but as

argued throughout this study, grammar is an abstraction from text, it is based on

knowledge of what is possible in a language.

8.3. Alongside and Beyond this Analysis

The analysis above gives us some indication of the kinds of meaning that can

made at clause rank. This is, of course, only a thumbnail sketch. The morphologi-

cal and lexical meanings that can be realized in Finnish have barely been conside-

red at all.

I have also ignored the question of prosody in the analysis above. It can be

regarded as another kind of organization that moves hand in hand with grammati-

cal analysis: the tone group is a different order of patterning that moves “alongsi-

de” the clause. The analysis presented here needs to be extended by looking at the

way in which the tone group in Finnish realizes the system of information (what

the listener is being asked to attend to) and Key (the meanings associated with the

different melodic contours in the tone group). One would also need to consider

how the tone system expresses logical relations between successive information

units, and how it interacts with clause complexing and contributes to conjunctive

relations in a text.

Page 380: systematic functional linguistic

373

The clauses above are not a jumbled mess. If we are judging only on the

basis of this small fragment, then we can assume that the speaker is sane and not

aphasic, and can produce a text that coheres and makes sense. This is even more

apparent if we add the fragment presented below, which follows on immediately

after clauses 1 ) 6 discussed in the previous section.

(7) ((laughing)) niiyeah/so

(8) se sormus ei millään)that ring not in any ‘that ring won’t in any

(9) nyt ku sen panee mun sormeenow when it is put on my finger‘now when it’s put on my finger’

(10) ni se ku on hetken sielläso it when is a moment there‘and then when it’s there for a moment’

(11) ni nousee tämmönen pattiso rises this kind of swelling‘then a kind of swelling occurs

(12) et se ei lähe pois [AA1: 3].that it won’t come off‘so it won’t come off’

If we look at the resources that make a text into a text and not just a jumbled mess

of unconnected words, then we need to look at the resources in Finnish that cor-

respond to the kind of phenomena analysed by Halliday & Hasan (1976) and Hal-

liday (1985a: Chapter 9).

Cohesive devices include what Halliday & Hasan refer to as “reference” (i.e.

the presence of co-referential items e.g. sormus ‘ring’ ) se ‘it’) as well as ellipsis

and substitution. Conjunctive relations between clauses also contribute to the

textness of a text, for example nyt ‘now’ in 9 above might be considered to realize

a circumstance of time in the experiential structure, but textually it functions as a

continuative, opening up a new stage in the communication. Various types of

Page 381: systematic functional linguistic

374

lexical cohesion need to be taken into account: repetition, meronomy (part-who-

le), hyponymy, collocation and synonymy.

However, from the perspective of the meanings that are made and can be

made in Finnish, this is only a beginning. In the analysis above, I made no refe-

rence to the genre that the text was representative of. Neither did I make any refe-

rence to the gender of the speaker, and there is no lexico-grammatical reflex of

gender in the fragment above. To any Finn ) and indeed to any English-speaking

person ) it would come as no surprise that the speaker was male. One might also

hazard a guess that the fragment is from conversation, conversation that is unself-

conscious and its function is interpersonal. The fragment above is, in fact, a trans-

cribed turn from a conversation among a group of men that took place in a

restaurant-pub in Helsinki. The recording is old ) it was recorded in the late fif-

ties ) but the reference to the advantages of marriage places it in a certain dis-

course, which is still recognizable today.

Moving “upwards and outwards” from the grammar, as it were, takes us in

the realm of what Lemke (1989) refers to as discourse formations: what is re-

peatedly meant in a community. In Lemke’s approach, some of the kinds of mea-

ning that are repeatedly meant in a society can be grouped as follows:

- written and spoken genre structures, texts in their context that are recogniza-

ble as being a meaningful mode of action in a society, e.g. interpersonal con-

versation, an editorial in a newspaper, a service encounter (see, e.g., Bakhtin

1986, Hasan 1984b, Halliday & Hasan 1985, Kress 1985, Ventola 1987);

- thematic formations, experiential semantic patterns common to a large num-

ber of texts representing some particular view of the world, e.g. the conceptual

schemas of a science or a field of scholarship (see Lemke 1988, 1989, 1990);

Page 382: systematic functional linguistic

375

- the interactional (or dialogical) semantics of text: the intertextual and intra-

textual links in a text and with other texts and the heteroglossic relations whe-

reby the “voice” of a text establishes its stance towards other voices.

Analysing what is repeatedly meant in a community takes us into the con-

flicts and contradictions that exist in any society and the subcultures of that so-

ciety. As Halliday (1985a: 318.) puts it:

Discourse is a multidimensional process; a ‘text’ which is the product of that process not onlyembodies the same kind of polyphonic structuring as is found in the grammar ... but alsosince it is functioning at a higher level of the code, as the realization of semiotic orders ‘abo-ve’ the language, may contain in itself all the inconsistencies, contradictions and conflicts thatcan exist within and between such higher-order semiotic systems. Because it has this poten-tial, a text is not a mere reflection of what lies beyond; it is an active partner in the reality-making and reality-changing process.

If a text can contain “all the inconsistencies, contradictions and conflicts that can

exist within and between such higher-order semiotic systems”, then the analysis

of text that is not grounded on a theory-based grammar remains a question of

personal inclination and individual interpretation. In order to claim that a text is

racist or sexist, for example, one needs a principled way of demonstrating the

grammatical and semantic properties of the text that can be said to give it a racist

or sexist reading.

It is only by relating grammar and meaning that we can give an interpretati-

on of text that will be relevant to “the living of life” (Firth in Palmer ed. 1968:

169). A grammar that is relevant to the study of texts in living contexts of life )

in educational linguistics, translation studies, literary analysis, in studies of langu-

age and ideology etc. ) needs to be more than a simple formalism. It seems to me

that the kind of grammar outlined in this study would provide a viable basis for

the study of texts.

Page 383: systematic functional linguistic

376

8.4. Recurring Themes

As discussed in 2.2.5, in Firth’s view the study of language is the study of linguis-

tic meaning. Similarly, according to Robinson (1975: 139):

Meaning is not a component of language, but a characterization. When language is itself, itis meaningful, and the meaning is whatever makes language language. Hence grammar isindeed the study of meaning.

Meanings, however, are made in various ways and at various ranks and levels,

and they can be analysed in varying degrees of delicacy. This study has provided

the outline of how major clauses can be analysed in terms of a number of simul-

taneous grammatical structures. These structures are not simplistically conceived

frames into which words are slotted, but reflect the way in which clauses are

meaningfully organized. Meanings at the rank of clause have been dispersed )

to use once again Firth’s analogy of the prism. They have been dispersed into a

number of generalized types of meaning, using Halliday’s metafunctional appro-

ach. But like white light or a chord that is played by a musician, a clause in a

spoken or written text is an indivisible whole. The prism and the musical score

are the instruments of the observer. In this study, I have attempted to demonstrate

the advantages of dispersing the meanings that are made at the rank of clause by

taking a metafunctional approach to grammatical analysis.

I have also attempted to show the far-reaching repercussions of grounding

grammatical analysis in the description of language in use. Linguists who start

with decontextualized, intuited sentences, and then tack on another component

called “pragmatics”, which puts the sentence in a text and a context, are faced

with an inherent problem: what is the ontological status of a decontextualized,

intuited sentence? I would suggest that it is not a sentence, but a text, an au-

tonomous text that usually consists of one clause. This autonomous mini-text,

however, is unlike any other spoken or written text that we encounter in our eve-

ryday life. Linguists who start with decontextualized, intuited sentences thus

ground their analysis on data that is representative of a fictional genre. This ap-

Page 384: systematic functional linguistic

377

1 I am not suggesting that it is useless. There are many reasons why one might want to writea grammar, and not all grammarians want to write grammars that are applicable to languagein use.

proach is self-defeating if one is interested in a grammar that can be applicable to

actual texts in use.1 It also embodies a lot of assumptions about language that

affect that way in which language is analysed.

The implicit assumptions in the analysis of decontextualized sentences is

evident, for example, in the traditional analysis of pre- and postpositions discus-

sed in Chapter 4. In the traditional analysis the initial constituent in the clause

Takana oli jyrkkä kallio ‘Behind (there) was a steep cliff’ is an adverb but a post-

position in Veljesten takana oli jyrkkä kallio ‘Behind the brothers (there) was a

steep cliff’. The only formal criteria distinguishing these constituents is the pre-

sence of a nominal in the genitive. Apart from clauses that are intuited by phi-

losophers and linguists, clauses are always realized in a context of situation and in

a co-text of what is being said or written. The context can be an immediate one or

one that is created by a story-teller or novelist. If someone says Takana oli jyrkkä

kallio ‘Behind there was a steep cliff’ then the clause is understood in its cotext

and context. The NP in the genitive may be retrievable from the cotext or the

clause is interpreted with reference to some deictic centre. This kind of considera-

tion is (con)textual (i.e. based on the co-text and situational context), and a con-

cept of grammar in which the context is seen as an inherent part of the clause is

bound to differ from one in which the context is tacked on at a later stage.

Another example of how an approach that abstracts from language in use

differs from an approach that has decontextualized sentences as its starting point

is the analysis of clause types in Finnish. If clauses like Puhutaan suomalaisen

muotoilun kriisistä [speak-INDE about the crisis in Finnish design] ‘There’s talk

about the crisis in Finnish design’, Syttyi sota [was kindled war] ‘A war broke

out’, Sataa [rains] ‘it’s raining’ and oli myöhä [was late] ‘It was late’ are ana-

lysed in terms of the relations between their constituents as if they were au-

tonomous bits of text, then it is difficult to see what they have in common. If one

Page 385: systematic functional linguistic

378

1 A field linguist does not describe a system of sounds, she or he has the task of describinga meaning-making system, and in order to do so needs to gain an entry into the system.

band of the spectrum of grammatical organization in language looks at how a

clause is related to its context, then in spite of the fact that they realize different

processes with different kinds of inherent roles, they can be regarded as being

similar from a textual point of view.

Thus, what in other schools of linguistic thought might be regarded as a

move from morphology and syntax to semantics and pragmatics cannot be seen in

terms of a simple extension of the field of enquiry. A grammatical analysis based

on language in use necessarily entails a re-assessment of morphology and syntax.

From a metatheoretical perspective, what happens is that the recontextualization

of an object of enquiry irrevocably changes the way in which the object of en-

quiry is understood.

While, on the one hand, grammatical analysis ) if it is to be relevant to the

analysis of texts ) must be grounded in the description of language in use, this

should not be misunderstood in simplistic terms. Grammatical patterning is abst-

racted from language in use. Throughout this study I have stressed a second equ-

ally important consideration: any linguistic analysis is based on knowledge. To

begin with, we need to have some entry into the linguistic system; we need have

some degree of proficiency in the language. An English linguist with no knowled-

ge of Finnish or the Finnish way of life would have very little of significance to

say about a Finnish text, and similarly a Finnish linguist with no knowledge of

Gooniyandi or the Gooniyandi way of life would have very little of significance

to say about a Gooniyandi text.1 Moreover, even a thousand texts would constitu-

te only an infinitesimal fraction of the texts that might be considered to be Fin-

nish. The major data sources for this study ) the written and transcribed spoken

texts listed in Appendix 2 ) fit into two folders. Furthermore, as pointed out in

Chapter 2, an act of observation ) selecting a particular syntagm from a particular

place in a particular text (as opposed to selecting another syntagm from another

Page 386: systematic functional linguistic

379

place in another text) ) is no less subjective than producing a syntagm based on

one’s knowledge of the language being described.

Even if we know the language and we have a thousand texts in front of us,

in order to do some kind of linguistic analysis, we need to be able to systematical-

ly relate instance X with instance Y, and show how they are dissimilar from ins-

tance A, B, C .. and so on. This is not something that a text “tells us”. If we take

the clause (7) niin ‘yeah, so’ and clause (8) (et) se ei lähe pois ‘so that it won’t

come off’ from the example discussed on p. 373 above, then in order to determine

how they differ as clauses, we need to be able to relate niin to other sets of minor

clauses and (et) se ei lähe pois to other kinds of major clauses. In order to see

how the grammar of Finnish works, we need to be able to abstract clauses from

their particular context, and relate them in a principled and systematic way to

other clauses.

A third consideration that was briefly discussed in Chapter 1 and is implicit

in the approach taken in this study is the fact that observation presupposes theory.

If we want to relate instance X with instance Y, then what counts as an instance of

X? How do we know an X when we see one? If our X is a word, then in the final

analysis, as Firth (1957: 190) put it, “each word used in a new context is a new

word”. Moreover, if we say that X and Y are related, then we need to explicate the

nature of this relation, and we also need to explicate the nature of their relation to

A, B, C etc.

This focus on theory and on the precedence of knowledge in linguistic desc-

ription is not meant to undermine the importance of data. Many of the hypotheses

put forward in this study need to be tested with data. However, if we use a corpus,

we do not come to it with a tabula rasa, but with either an assortment of unarti-

culated assumptions and (theoretical) concepts or with assumptions that have

been articulated within a theoretical framework. The symbiotic relationship bet-

ween knowledge and data ) and between theory and description ) ensures what

Firth referred to as the “renewal of connection”. Armchair speculation about lan-

Page 387: systematic functional linguistic

380

guage in a vacuum and a grammar that is too theory-driven will have little of

relevance to say about actual texts in the context of living.

In this study, I have attempted to make explicit my assumptions about

language and the study of language within the framework of systemic-functional

theory. However, as pointed out in Chapter 1, the language of theory like langu-

age itself is inherently intertextual and heteroglossic, so the understanding of this

text, like any other text, will also depend on the reader of this particular text. A

reader who is unfamiliar with the systemic-functional texts referred to in this stu-

dy will understand this study in a different way to a reader who has read at least

some of the texts on systemic-functional theory that are referred to in the bibliog-

raphy, but, then by the same token no two people who are familiar with systemic-

functional linguistics will understand systemic-functional theory in exactly the

same way. Neither will they understand this study in the same way. Such is the

nature of language.

Page 388: systematic functional linguistic

381

Appendix 1: Analysis of the “Cat Text”

To the extent that it is possible, I have put the translation equivalents of Finnish constituentsin the same order in English as they are in Finnish. While this produces a lot of marked the-mes in English, it does not mean that the Theme is marked in Finnish, although it could be.

The topical Theme is marked by double underlining:koirat ovat purreetdogs have bitten.

A subsidiary experiential Theme has single underlining:Omakotitalossa kissa oppii pysymään omassa pihapiirissäin detached house cat learn to stay in its own yard

Interpersonal and textual points of departure are marked in cursive type:mutta myös koirat ovat purreet,but also dogs have bitten

A host clause that is the Theme in an embedded complex is marked as follows: On ihanaa,is wonderfulkun kissa kulkee viskellen villiä häntäänsäwhen cat moves about waving its wild tail

TITLE OF THE ARTICLE:

2 Lue tämä 2 ennen kuin otat kissan. 2read+IMP this+NOM before than take+2SG cat+GEN

‘Read this before you get yourself a cat.’

(1) 2 Tuskin on olemassa ihmistä,scarcely is+3SG being/existence+INE person+PAR

Û joka suhtautuu välinpitämättömästi kissaan. â 2who+NOM relate+PASS-REF+3SG indifferently cat+ILL

‘It’s rare for someone to be indifferent to cats.’

(2) 2 Yleensä ihmiset joko pitävät siitä 2generally person+NOM/PL either hold/like+3PL it+ELA

tai suhtautuvat siihen vähättelevästi, jopa vihamielisesti. 2or relate+PASS-REF+3PL it+ILL disparagingly even hostilely

‘Generally people either like cats or they relate to them disparagingly or even hostilely.’

Page 389: systematic functional linguistic

382

(3) 2 Monet, Û jotka vilpittömästi pitävätmany+NOM/PL who+NOM/PL sincerely hold/regard+3PL

itseään eläinystävinä â vieroksuvat kissaaself+PAR+POS/3 animal+friend+PL+ESS shy away from+3PL cat+PAR

sen oletetun petomaisuuden takia. 2it+GEN assume/suppose+INDEF/PTC+GEN predatory nature+GEN because

‘Many who sincerely regard themselves as animal lovers shy away from cats because of theirsupposed predatory nature.’

(4) 2 Kummallisen sitkeästi elävät käsitykset,strange+GEN toughly/perseveringly live+3PL idea/view+NOM/PL

Û että kissa saattaa purra poikki kaulavaltimon, 2that cat+NOM might+3SG bite+INF off/across carotid artery+GEN

repiä silmät päästä, 2 taiscratch+INF eyes+NOM/PL head+ELA or

että se yleensäkin on vain olemassathat it+NOM generally+TIS be+3SG only/just being/existence+INE

Ú saadakseen sopivana hetkenä raadellaget+INF+TRA+POS/3 appropriate+ESS moment+ESS maul+INF

jonkun elävän olennon. á â 2some+GEN living+GEN creature+GEN

‘How strangely persistent are the views (There is a strangely persistent view) that a cat mightbite through a neck artery, scratch a person’s eyes out, or that it generally only exists in orderto maul some living creature at the appropriate moment.’

(5) 2 Kaikenlaista ikävää onkin aikojenall+GEN+kind+PAR distressing/sad+PAR be+3SG+TIS time+PL+GEN

saatossa sattunut, 2procession+INE happen/occur+PTC

mutta myös koirat ovat purreet, 2but also/besides dog+NOM/PL be+3PL bite+PTC/PL

hevoset potkineet 2 ja pässit pökkineet. 2horse+NOM/PL kick+PTC/PL & ram+NOM/PL butt+PTC/PL

‘All kinds of distressing things have in fact occurred during the course of time, but it’salso true that dogs have bitten, horses have kicked and rams have butted people.’

Page 390: systematic functional linguistic

383

(6) 2 Silti koira on ihmisen paras ystävä, 2nevertheless dog+NOM be+3SG person+GEN best+NOM friend+NOM

hevonen jalo eläin 2 jahorse+NOM noble+NOM animal+NOM &

pässistäkin voi keksiä mukavaa sanottavaa. 2ram+ELA+TIS can+3SG think oft+INF nice+PAR say+INDE+PTC+PAR

‘Nevertheless a dog is a man’s best friend, a horse (is) a noble animal, and one can thinkof something nice to say about a ram too.’

(7) 2 Kissa vain on 2 ja pysyy petona, 2cat+NOM just/only be+3SG & remain+3SG beast of prey+ESS

eikä siksi ansaitse samaa huolenpitoaNEG+3SG+TIS it/that-TRA deserve same+PAR care+PAR

kuin muut kotieläimet. 2than/as other+NOM/PL domestic animal+NOM/PL

‘A cat simply is and will remain a beast of prey, and for this reason it does not deserve thesame care as other domestic animals.’

(8) 2 Mutta eipä silti, eivät kaikki suomalaisetbut for all that/nevertheless NEG+3PL all+NOM Finn+NOM/PL

ole aina ajatelleet, 2 että kissan paikkabe+PTC always think+PTC/PL that cat+GEN place+NOM

on vain ulkorakennuksissa hiirenpyynnissä. 2is+3SG only/just outbuilding+PL+INE mouse+GEN+catching+INE

‘But on the other hand (for all that), not all Finns have always thought that the only placefor a cat is in outbuildings catching mice.’

(9) 2 Olihan seitsemällä veljekselläkin oma temppujabe+PAS+3SG+TIS seven+ADE brothers+ADE+TIS own+NOM tricks+PL+PAR

tekevä Matti-kissa, Û jolla olimake/do+PTC/NOM Matti-cat+NOM who+ADE be+PAS+3SG

turvattu asema perhepiirissä. â 2secure+INDE/PTC position+NOM family+circle+INE

‘After all even the Seven Brothers had their own trick-performing cat called Matti, whohad a secure position in the family circle.’ (The reference is to “The Seven Brothers” byAleksis Kivi.)

Page 391: systematic functional linguistic

384

(10) 2 Koiran suosio lemmikkinä on helpostidog+GEN popularity+NOM pet+INE be+3SG easily

ymmärrettävä, 2 sillä koira on alistuvainen; 2understand+INDE/PTC as dog+NOM is+3SG submissive+NOM

sen perusluonteeseen kuuluu laumaeläimen tarveit+GEN basic nature+ILL belong+3SG pack+animal+GEN need+NOM

Ú saada totella johtajaansa.á 2get/be able/allowed+INF obey+INF leader+POS/3

‘The popularity of a dog as a pet can be easily understood: a dog is submissive; the needto be able to obey its leader is part of a dog’s basic nature.’

(11) 2 Tämä voi olla huonokin, 2 muttathis+NOM can+3SG be+INF bad+NOM+TIS but

koiran kiintymys on silti ehdotonta. 2dog+GEN devotion/attachment+NOM be+3SG nevertheless unconditional+PAR

‘Its leader may not be a good one, but a dog’s devotion is, nevertheless, unconditional.’

(12) 2 On olemassa suuri joukko ihmisiä,is+3SG being/existence+INE large+NOM group+NOM person+PL+PAR

Û joille näyttää olevan tärkeintäwho+PL+ADE seem+3SG be+PTC+GEN important+SUP+PAR

Ú saada olla eläimen palvonnan kohteena. á 2get+INF be+INF animal+GEN worship/adoration+GEN object+ESS

‘There is a large group of people for whom it is essential to get to be the object of an ani-mal’s adoration.’

(13) 2 Tällaiset henkilöt aliarvioivat kokonaan kissanthis kind+NOM/PL person+NOM/PL underestimate+3PL entirely cat+GEN

kyvyn Ú kokea myönteisiä tunteitacapacity/ability+GEN experience+INF positive+PL+PAR feeling+PL+PAR

ihmisiä kohtaan. á 2person+PL+PAR towards

‘People like this entirely underestimate a cat’s capacity to experience positive feelingstowards humans.’

Page 392: systematic functional linguistic

385

(14) 2 Kissaihmiset ovat tottuneet kuulemaancat+person+NOM/PL be accustomed/used to-PERF/3PL hear+ILL

hämmästyttäviä tokaisuja,surprising/astonishing+PL+PAR snapping out/jibe+PL+PAR

Û jotka saattavatepäilyksen alaisiksiwhich+NOM/PL lead+3PL suspicion+GEN subordinate/being under+TRA

sekä kissan että sen omistajan järjenjuoksun. â 2both cat+GEN and it+GEN owner+GEN reason+GEN+flow+GEN

‘Cat lovers are used to hearing astonishing jibes that put into question the intelligence of boththe cat and its owner.’

(15) 2 Joskus tekisi mieli lausua ääneensometimes make+CON+3SG mind+NOM express+INF sound+ILL

ajatuksensa: 2 « Jos eläisimme keskiajalla 2thought+GEN+POS/3 if live+CON+1PL middle+age+ADE

sinunlaisesi ihmiset varmaan polttaisivatyour+kind+POS/2SG person+NOM/PL for sure burn+CON+3PL

minut roviolla kissoineni... » 2I+ACC pyre+ADE cat+COM+POS/1SG

‘Sometimes you feel like voicing (have a mind to voice) out aloud the thought:”If we wereliving in the middle ages, people like you would surely burn me at the stake with mycat(s).”’

(16) 2 Ú Halu hallita toisia á on usein tyypillinendesire+NOM control+INF other+PL+PAR be+3SG often typical+NOM

ominaisuus koirien ihailijoille. 2characteristic+NOM dog+PL+GEN admirer+PL+ALL

‘The desire to control others is often typical of people who admire dogs.’

(17) 2 Kissan ystävät ovat suurpiirteisempiä,2cat+GEN friend+NOM/PL be+3PL generous/liberal/tolerant+COMP+PL+PAR

he hyväksyvät helpommin toisen yksilöllisyyden. 2they+NOM accept+3PL more easily other+GEN individuality+GEN

‘People who like cats are more tolerant, they more readily accept the individuality of ot-hers.’

Page 393: systematic functional linguistic

386

(18) 2 Lapsilla ja nuorilla, Û jotka kasvavatkissanchild+PL+ADE & young+PL+ADE who+NOM/PL grow+3PL cat+GEN

kanssa, â on mahdollisuus kehittyä toisia huomioonwith be+3SG chance+NOM develop+INF other+PL+PAR consideration+ILL

ottaviksi, antaviksi ihmisiksi, 2take+PTC+PL+TRA give+PTC+PL+TRA person+PL+TRA

kun lapsen on jatkuvasti muistettava, 2when/as child+GEN be+3SG continually remember+INDE/PTC

että kissalla on omakin tahto, 2that cat+ADE be+3SG own+NOM+TIS will+NOM

että ollaan ystäviä, 2that be+INDE friend+PL+PAR

mutta ei vain toisen ehdoilla. 2but NEG+3SG only other+GEN term+PL+ADE

‘Children and young people who grow up with a cat have the chance of developing intogiving adults who take others into consideration, since a child must continually rememberthat a cat has its own will too, that they can be friends, but not just on the child’s terms.’

(19) 2 Väite, Û että kissa kiintyy paikkaan,claim+NOM that cat+NOM attach+PASS+3SG place+ILL

ei ihmisiin, â pitää paikkansa, 2NEG+3SG person+PL/ILL keep+3SG place+POS/3

jos kissan annetaan vain Ú kuljeksia yksinään 1if cat+GEN give/let+INDE only wander about+INF by itself

ja yksin hakea elämäänsä kiintopisteitä. á 2& alone look for+INF life+PAR+POS/3 fixed+point+PL+PAR

‘The allegation that a cat gets attached to a place and not to people is true if a cat is left towander about by itself and look for fixed points in its life alone.’

Page 394: systematic functional linguistic

387

1 The relative pronouns are the topical Themes in the embedded clauses.

(20) 2 Mutta kissa, Û jolle1 osoitetaan kiinnostusta, 2but cat+NOM who+ALL indicate+INDE interest+PAR

jonka kanssa leikitään 2 ja jolle puhellaan ystävällisesti, âwho+GEN with play+INDE & who+ADE chat+INDE friendly/kindly

omaksuu hoitajan ystäväkseen. 2adopt/accept+3SG carer+GEN friend+TRA+POS/3

‘But a cat that you show an interest in, that you play with and chat with in a friendly man-ner, will accept you as its friend.’

(21) 2 Molemminpuolinen lämmin tunne on arvokas asia, 2mutual+NOM warm+NOM feeling+NOM be+3SG valuable+NOM thing+NOM

sillä kissan ystävyys on ansaittua, 2as cat+GEN friendship+NOM be+3SG earn/deserve+INDE/PTC+PAR

ei automaattisen vaiston ohjaamaa. 2NEG+3SG automatic+GEN instinct+GEN direct+PTC

‘A mutual and warm feeling is valuable, since the friendship of a cat is something that isearned, not something directed by automatic instinct.’

(22) 2 On ihanaa, 2is+3SG wonderful+PAR

kun kissa kulkee Ú viskellen villiä häntäänsä, á 2when cat+NOM move+3SG wave about+INS wild+PAR tail+PAR+POS/3

mutta katseessa tuike, Û joka kertoo onnellisestabut look+INE twinkle+NOM which+NOM tell+3SG happy+ELA

asuinkumppanuudesta. â 2living+partnership+ELA

‘It’s wonderful when a cat moves about waving its bushy tail about, but with a glint in its eyewhich reflects a happy living relationship.’

Page 395: systematic functional linguistic

388

(23) 2 Tämä jää näkemättä sellaiselta,this+NOM remain+3SG see+INF+ABE such kind+ABL

Û joka odottaa jatkuvaa hössöttämistäwho+NOM expect+3SG continue+PTC+PAR activity+PAR

ja hännän huisketta. â& tail+GEN whisking+PAR

‘This is missed by someone who expects continuous activity and tail-whisking.’

(24) 2 Kissasta ei ole seuraa kävelyretkille, 2cat+ELA NEG+3SG be+PTC company+PAR walk+trip+ALL

mutta toisaalta sitä ei myöskään tarvitsebut on the other hand it+PAR NEG+3SG also+TIS(NEG) need-PTC

kuljettaa ulos. âtransport+INF outside

‘A cat is no company on a walk but, on the other hand, it does not need to be taken out eit-her.’

(25) 2 Liikunnantarpeensa se saa tyydytetyksi jopaexercise+GEN+need+POS/3 it+NOM get+3SG satisfy+INDE/PTC+TRA even

kaupunkiasunnossa, 2 tarpeensa se suorittaa lehtisilpullacity+house+INE need-POS/3 it+NOM carry out-3SG paper+shred+ADE

tai muoviritilällä pohjustettuun vatiin. 2or plastic+grating+ADE ground/base+INDE/PTC+ILL basin+ILL

‘Its need for exercise gets to be satisfied even in an apartment house, its business can be donein a basin lined with paper shreds or with a piece of perforated plastic at the base.’

(26) 2 Kun hoito tapahtuu omassa kodissa, 2when care+NOM happen+3SG own+INE home+INE

ei lemmikkinsä takia joudu kiistaanNEG+3SG pet+POS/3 sake get into+PTC quarrel+ILL

muiden kanssa. 2other+PL+GEN with

‘When things are taken care of in your own home, then you’re unlikely to get into a quarrelwith others because of your pet.’

Page 396: systematic functional linguistic

389

(27) 2 Se onkin aikamoinen helpotus, 2it+NOM be+3SG+TIS quite+NOM relief+NOM

kun tietää 2 kuinka usein koirantaluttajatwhen know+3SG how often dog+GEN+walker+NOM/PL

saavat ulkopuolisten kiukut niskoilleen. 2get+3PL outsider+PL+GEN anger+NOM/PL neck+PL+ALL+POS/3

‘This is quite a relief when you know how often people walking dogs are subject to the abuseof others.’

(28) 2 Omakotitalossa kissa oppii pysymäändetached+house+INE cat+NOM learn+3SG stay+INF+ILL

omassa pihapiirissä,own+INE yard+circle+INDE

2 jos se on pienestä pitäen tottunutif it+NOM be+3SG little-ELA hold/keep+INS accustom+PTC

siihen, Û että turvallisinta ja hauskintait+ILL that secure+SUP+PAR & fun/interesting+SUP+PAR

on silloin Û kun oma perhe onbe+3SG it+by [time] when own+NOM family+NOM be+3SG

ainakin kuuloetäisyyden päässä. â â 2at least hearing+distance+GEN away-INE

‘In a detached house, on the other hand, a cat learns to stay within the confines of the yard, iffrom when it was small it has been used to the fact that it is safest and interesting when yourown family is at least within hearing distance.’

(29) 2 Jos tämä ei onnistu,if this+NOM NEG+3SG succeed+PTC

2 voi aluksi rakentaa ulkoilutarhan, 2can+3SG beginning/start+TRA build+INF outdoor+cage+GEN

niin vieraalta kuin ajatus voi tuntuakin. 2so/as strange-ABL as/then thought+NOM can+3SG feel/seem+INF+TIS

‘If this doesn’t succeed, (then) in the beginning you can build an outdoor cage as strange asthe idea may in fact seem.’

Page 397: systematic functional linguistic

390

(30) 2 Kissan voi myös pentuna opettaacat+GEN can+3SG also kitten+ESS teach+INF

ulkoilemaan valjaissa. 2be outside+INF+ILL leash+PL+INE

‘A cat can also be taught as a kitten to be outside on a leash.’

(31) 2 Erityisesti pikkulintujen lentoharjoitusten aikaan onparticulary small+bird+PL+GEN flying+practice+PL+GEN time+ILL be+3SG

kissan liikkumista rajoitettava.2cat+GEN moving+PAR restrict+INDE/PTC+NOM

‘When small birds are learning how to fly is one particular time when a cat’s movementsshould be restricted.’

(32) 2 Omakotialueilla syntyy paljondetached+house+area+PL+ADE create/is born+PASS+3SG a lot

epäviihtymystä siksi, Û että kissojen omistajatunenjoyment/unease+PAR it/that-TRA that cat+PL+GEN owner+PL+NOM

eivät huolehdi niiden leikkauttamisesta.â 2NEG+3PL take care of they+GEN desex/neutering+PAR

‘A lot of unpleasantness arises in an area with detached houses because owners do not takecare of neutering their cats.’

(33) 2 Kissan sukuvietti on voimakascat+GEN sexual instinct+NOM be+3SG strong+NOM

eivätkä niiden öiset serenadit oleNEG+3PL+and they+GEN nightly+NOM/PL serenade+NOM/PL be

kovin miellyttävää kuultavaa. 2very please+INDE/PTC+PAR hear+INDE/PTC+PAR

‘A cat’s sexual instinct is strong, and their nightly serenades are not the most pleasant thing tolisten to.’

Page 398: systematic functional linguistic

391

(34) 2 Leikkaamattomien kollikissojen hajumerkit eivätuncastrated+PL+GEN tomcat+PL+GEN scent+mark+NOM/PL NEG+3PL

nekään edistä naapurisopua. 2they+NOM+TIS(neg) promote neighbour+harmony+PAR

‘Neither do the scent marks of uncastrated tomcats do much to promote neighbourly love.’

(35) 2 On käsittämätöntä,be+3SG incomprehensible+PAR

Û että joidenkin mielestä ainoa todellinen kissa onthat some+PL+GEN only real+NOM cat+NOM be+3SG

tappeluissa parkkiintunut, repalekorvainen kolli 2fight+PL+INE toughen+PASS+PTC/NOM shred+eared+NOM tom+NOM

ja siksi he itsepintaisesti kieltäytyvät& it/that-TRA they+NOM stubbornly refuse+PASS-REF+3SG

leikkauttamasta sitä. â 2have neutered+INF+ELA it+PAR

‘It’s incredible that some people think that the only real cat is a tomcat who has beentoughened in fights and whose ears are in shreds, and, thus, they stubbornly refuse to haveit neutered.’

(36) 2 Kuitenkin tällainen kissa on ympäristönnevertheless this kind+NOM cat+NOM be+3SG environment+GEN

vaivana ja lisäksi hengenvaarassa; 2nuisance+ESS & furthermore life+GEN+danger/risk+INE

ennen pitkää se tekee retken,before long it+NOM do/make+3SG trip+GEN

Û jolta ei enää palaa. â 2which+ABL NEG+3SG longer return

‘Nevertheless this kind of cat is a nuisance to its environment and furthermore it is at risk oflosing its life; before long it will go on a trip from which it will no longer return.’

Page 399: systematic functional linguistic

392

1 The idiom in Finnish for “halcyon days” is “cat’s days”.

(37) 2 Paljon eläinystävällisempi vaihtoehto olisi a lot animal+friendly+COMP+NOM alternative+NOM be+CON+3SG

mukavat kissanpäivät kodin lähituntumassa. 2pleasant+NOM/PL cat+GEN+day+NOM/PL home+GEN close+touch+INE

‘A more animal-friendly alternative would be to spend halcyon days1 in the close vicinity ofthe home.’

(38) 2 Leikattu kissa ei liho, 2neutered+NOM cat+NOM NEG+3SG put on weight

eikä menetä vireyttään, 2NEG+3SG+TIS lose energy/vigour+PAR+POS/3

jos sitä ei syötetä liikaa. 2if it+PAR NEG+3SG feed+INDEF too much

‘A neutered cat does not put on weight nor lose its energy if it is not fed too much.’

(39) 2 Naaraan leikkaus on toimenpiteenä isompi,female+GEN spaying+NOM be+3SG procedure+ESS big+COMP+NOM

mutta yhtä välttämätön. 2but one+PAR (= as) unavoidable/necessary+NOM

‘The spaying of a female cat is a lengthier procedure, but just as necessary.’

(40) 2 Kissanpentujen jakaminen tuntemattomille sisältääcat+GEN+offspring+PL+GEN giving away+NOM unknown+PL+ALL include+3SG

aina vaaran, Û että ne joutuvatalways risk/danger+GEN that they+NOM fall/get into+3PL

huolimattomiin käsiin. â 2careless/uncaring+PL+ILL hand+PL+ILL

‘Giving kittens away to strangers always includes the risk that they will fall into uncaringhands.’

Page 400: systematic functional linguistic

393

(41) 2 Kissasta otettu maksu on edescat+ELA take+INDE/PTC+NOM payment+NOM be+3SG at least

jonkinlainen tae siitä, Û että sen vuoksisome kind+NOM guarantee it+ELA that it+GEN sake

ollaan halukkaita jotakin uhraamaankin. â 2be+INDE willing+PL+PAR something+PAR sacrifice+ILL+TIS

‘Taking payment for a cat is at least some sort of guarantee that people are willing to sacrificesomething for its sake.’

(42) 2 Mitä muuta nykyään lahjoitetaan ilmaiseksiwhich+PAR else+PAR nowadays give (as a present)+INDE free+TRA

pois kuin kissanpentuja? 2away than cat+GENoffspring+PL+PAR

‘What else nowadays can you get for free besides a kitten?’

(43) 2 Mikä muu on yhtä vähäarvoista? 2which+NOM else+NOM be+3SG one+PAR (= as) little+value+PAR

‘What else is of as little value?’

(44) 2 Juuri tästä syystä syntyy,precisely this-ELA reason-ELA emerge/create/is born+PASS+3SG

jopa keskelle kaupunkialueita, villikissalaumoja,even middle+ALL city+area+PL+PAR feral+cat+pack+PL+PAR

Û joiden elämä pitkinä talvikuukausina onwho+PL+GEN life+NOM long+PL+ELA winter+month+PL+ELA be+3SG

pelkkää kärsimistä. â 2mere/pure+PAR suffering+PAR

‘It is precisely for this reason that packs of feral cats emerge even in the middle of town areas,and their life during the long winter months is nothing but suffering.’

Page 401: systematic functional linguistic

394

(45) 2 Kuitenkin on aina myös niitä,nevertheless be+3SG always also they+PL+PAR

Û jotka henkeen ja vereen puolustavat tällaistenwho+NOM/PL spirit/soul+ILL & blood+ILL defend+3PL this kind+PL+GEN

kissaparkojen oikeutta Ú elää ja synnyttääcat+wretched/dear+PL+GEN right+PAR live+INF & give birth+INF

poikasiaan lumen ja jään keskelle.á â 2young+PAR+POS/3 snow+GEN & ice+GEN middle+ALL

‘Nevertheless there are also always people who will defend with their body and soul, the rightof these poor cats to live and give birth to their young in the middle of snow and ice.’

(46) 2 Tunteiden ristiaallokkoa täydentävät toiset,feeling+PL+GEN cross+waves/swell+PAR complete+3PL other+NOM/PL

Û jotka puolestaan vihaavat näitä syyttömästiwho+NOM/PL part+ELA+POS/3 hate+3PL these+PAR innocently

kärsimään joutuvia eläimiä luonnonsuffer+INF+ILL fall/get into+PTC+PL+PAR animal+PL+PAR nature+GEN

tuhoojina 2 ja ovat valmiit, vaikka kuinka tökeröindestroy+PL+ELA & be+3PL ready+PL+NOM no matter how clumsy+PL+INS

keinoin, päästämään ne päiviltä. â 2mean/manner+PL+INS let go/release+INF+ILL they+NOM day+PL+ABL

‘This conflict of emotions is complemented by others, who, for their part, despise these crea-tures, who have innocently had to suffer, as destroyers of nature and are ready to shorten theirlives by even the most ham-fisted means.’

(47) 2 Kissa on erittäin siisti eläin, 2cat+NOM be+3SG extremely tidy/clean+NOM animal+NOM

kokihan se Pasteurin bakteerikeksintöjenexperience+3SG+TIS it+NOM Pasteur+GEN bacteria+discovery+PL+GEN

aikaan aivan uudenlaisen arvonnousun. 2time+ILL completely new+GEN+kind+GEN value/esteem+GEN+rise+GEN

‘A cat is an extremely clean animal: as is evident from the fact that the time of Pasteur’s bac-teria discoveries, it experienced a completely new rise in esteem .’

Page 402: systematic functional linguistic

395

(48) 2 Kaiken haisevan ja likaisen se kiertää, 2all+GEN smelly+GEN & dirty+GEN it+NOM avoid+3SG

turkkinsa se pitää silkkisenä ja puhtaana 2coat+POS/3 it+NOM keep+3SG silky+ESS & clean+ESS

eikä se likaa edes liinavaatekaappia,2NEG+3SG+and it+NOM make dirty+3SG even linen+closet+PAR

jos se joskus päättää torkahtaa siellä hetken. 2if it+NOM sometimes decide+3SG doze+INF there moment+GEN

‘Anything smelly and dirty it avoids, its coat is kept silky and clean, and it doesn’t soil thelinen closet if it sometimes decides to take a nap there for a moment.’

(49) 2 Poikkeuksia on,exceptions+PL+PAR be+3SG

mutta yleensä kissa oppii erittäin helposti käymäänbut generally cat+NOM learn+3SG extremely easily go/visit+INF+ILL

omalla astiallaan. 2own+ADE dish/basin+ADE+POS/3

‘There are exceptions but generally a cat very easily learns to go to its own toilet basin.’

(50) 2 Astia on pidettävä puhtaana ja hajuttomana. 2basin+NOM be+3SG keep+INDE/PTC+NOM clean+ESS & scentless+ESS

‘The basin must be kept clean and should not smell.’

(51) 2 Opetuksen on oltava lempeää, muttainstruction+GEN be+3SG be+INDE/PTC+NOM tender+PAR but

päättäväistä, kuritus vain pahentaa asiaa. 2firm+PAR punishment+NOM only make worse+3SG matter+PAR

‘Discipline should be tender but firm, punishment only makes things worse.’

Page 403: systematic functional linguistic

396

(52) 2 Lukemattomat kissat ovat olleet koko elämänsäcountless+NOM/PL cat+NOM/PL be+3PL be+PTC/PL whole life+POS/3

täysin siistejä ilman, Û että ne edes ovatcompletely tidy+PL+PAR without that they+NOM even be+3PL

tarvinneet tässä asiassa opetusta. â 2need+PTC/PL this+INE matter+INE instruction+PAR

‘Countless cats have been completely tidy their whole life without needing to be disciplinedin this matter.’

(53) 2 Muutenkin kissan tavoille opettaminen on aikaotherwise+TIS cat+GEN way+PL+ALL teaching+NOM be+3SG quite

yksinkertaista, 2 parempien huonekalujensimple+PAR better+COMP+PL+GEN furniture+COMP+PL+GEN

raapiminen kielletään Ú ropauttamalla vettäscratching+NOM forbid+INDE squirt+INF+ADE water+PAR

suihkepullosta kuonon päälle, á 2spray+bottle+ELA nose+GEN head/on+ALL

eikä sitten paljon muuta tarvitakaan. 2NEG+3SG+TIS then much else+PAR need+TIS

‘In other respects teaching a cat how to behave is rather simple, you can prevent it from scrat-ching your best furniture by squirting water from a spray bottle on its nose, and then there’snot much else that you really need to do.’

(54) 2 Kissa katsoo, 2cat+NOM look/observe/watch+3SG

kuinka muut elävät 2how-Q other+NOM/PL live+3PL

ja asettuu sitten taloksi. 2& place+PASS-REF+3SG then house+TRA

‘A cat observes how others live and then settles down into a house.’

Page 404: systematic functional linguistic

397

(55) 2 Tärkeintä on,important+SUP+PAR be+3SG

Û että kiellettyjä asioita ei ole niin monta,that forbidden+PL+PAR matter+PL+PAR NEG+3SG be so many+PAR

Û että kissaa täytyy jatkuvasti komennella. â â 2that cat+PAR must/have to+3SG continually order about+INF

‘It’s important that (/what is important is that) there are not so many no-nos that you have tobe continually ordering a cat about.’

(56) 2 Kissa vaistoaa erittäin herkästi, 2cat+NOM feel intuitively+3SG extremely sensitively

onko se kodissa todella hyväksytty. 2be+3SG+Q it+NOM home+INE real+ADE accept+INDE/PTC

‘A cat has an extremely sensitive instinct concerning whether or not it is really accepted in ahome.’

(57) 2 Turvalliseksi olonsa tunteva kissa onsecure+TRA existence+POS/3 feel+PTC cat+NOM be+3SG

« kiltti », mutta loukkaantunut tai epävarmuudessa eläväwell-behaved+NOM but hurt/offend+PTC or uncertainty+INE live+INE

keksii kyllä tapoja, Û joilla härnätä jainvent+3SG surely way+PL+PAR which+PL+ADE pester/irritate+INF &

hyppyyttää perhettänsä. â 2make to jump+INF family+PAR+POS/3

‘A cat that feels secure is well-behaved, but one that is offended or lives in uncertainty is sureto think of ways to irritate its family and make them jump.’

(58) 2 Omaa kotikissaamme nimitetään kissapiireissäown+PAR home+cat+PAR+POS/1PL name+INDE cat+circle+PL+INE

eurooppalaiseksi lyhytkarvakissaksi. 2European+TRA short+(body)hair+cat+TRA

‘Our own house(hold) cat [i.e. in Finland] is referred to in cat-breeding circles as a Europeanshort-haired cat.’

Page 405: systematic functional linguistic

398

(59) 2 Parhaimmillaan siitä voidaan sanoamontabest+SUP+PL+ADE+POS/3 it+ELA can+INDE say+INF many+PAR

hyvää asiaa, 2 mutta miinuspuolelle kuuluugood+PAR matter+PAR but minus+side+ALL belong+3SG

se, Û että se ei ole ainait/that+NOM that it+NOM NEG+3SG be always

hyvä kaupunkikissa.â 2good+NOM town+cat+NOM

‘At its best, a lot of good things can be said about it, but on the minus side is the fact that it isnot always a good town cat.’

(60) 2 Paljon riippuu perintötekijöistä ja siitä a lot depend+3SG heredity+factor+PL+ELA & it/that+ELA

Û kuinka sitä on pienestä pitäen kohdeltu. â 2how it+PAR is+3SG small-ELA hold/keep+INS treat+INDE

‘A lot depends on heredity factors and how it has been treated from the time when it wassmall.’

(61) 2 Koskaan kissaa ei pitäisi ottaa(n)ever cat+PAR NEG+3SG have to/must+CON take+INE

noin vain mistä vain, 2just like that which+ELA just/only

sillä silloin onnistuminen on ainaas at that time/event succeeding+NOM be+3SG always

epävarmaa. 2uncertain+PAR

‘You should never accept a cat/A cat should never be accepted just like that from anywhere atall, as then success is bound to be uncertain.’

(62) 2 Ulkomaisista roduista on tällä hetkelläoverseas/foreign+PL+ELA breed+PL+ELA be+3SG this+ADE moment+ADE

suosituin persialainen, Û jonka pörröinen olemusfavour+INDE/PTC+SUP Persian+NOM which+GEN shaggy+NOM being+NOM

miellyttää monia. â 2please+3SG many+PAR

‘Of foreign breeds, at the moment, the most popular is a Persian, whose shaggy appearanceappeals to many.’

Page 406: systematic functional linguistic

399

(63) 2 Tästä kissasta haaveilevien on kuitenkinthis+ELA cat+ELA fancy/dream about+PTC+PL+GEN be+3SG nevertheless

otettava huomioon jokapäiväinen turkinhoito jatake+INDE+PTC notice+ILL (every+)daily+NOM fur+GENcare+NOM &

karvanlähtö, Û mikä järjestää runsaastikinhair+GENloss+NOM which+NOM organize+3SG plentifully+TIS

lisätöitä. â 2extra+work+PL+PAR

‘Those who would like a cat like this should nevertheless take into account the daily care ofits coat and its loss of hair, which can mean a lot of extra work.’

(64) 2 Luonteeltaan persialainen on tasaisen rauhallinennature+ABL+POS/3 Persian+NOM be+3SG evenly+NOM calm+NOM

2 mutta eri värien kesken onbut different colours+PL+GEN among be+3SG

temperamenttieroja olemassa. 2temperament+difference+PL+PAR be+INF+INE

‘By nature a Persian is even-tempered, but amongst the various colours there are differencesin temperament.’

(65) 2 Persialainen sopii myös perheisiin,Persian+NOM suit+3SG also family+PL+ILL

Û joissa se joudutaan jättämään päiväksi yksin. 2which+PL+INE it+NOM have to+INDE leave+INF+ILL day+TRA one+INS

‘A Persian is also suitable for a family in which it has to be left alone for the day.’

(66) 2 Lyhytkarvarotujen kauneimpiin kissoihinshort+hair+breed+PL+GEN beautiful+COMP+PL+ILL cat+PL+ILL

kuuluvat abessinialaiset. 2belong+3PL Abyssinian+NOM/PL

‘Among the more beautiful of short-haired breeds are Abyssinians.’

Page 407: systematic functional linguistic

400

(67) 2 Joidenkin yksilöiden sammumaton vaellushalusome+PL+GEN individual+PL+GEN insatiable+NOM wanderlust+NOM

saattaa tehdä abessinialaisen pitämisen hankalaksi. 2might+3SG make+INF Abyssinian+GEN keeping+GEN difficult+TRA

‘The insatiable wanderlust of some individuals might make the keeping of an Abyssiniandifficult.’

(68) 2 Nämä kissat osoittavat hellyyttä puuskittain 2these+NOM cat+NOM/PL show+3PL affection+PAR by gusts

ja viihtyvät sitten pitkät tovit& enjoy/are content to be+3PL then long+NOM/PL spell+NOM

omissa maailmoissaan, 2 siksi ne ovatown+PL+INE world+PL+INE+POS/3 it/that-TRA they+NOM be+3PL

sopivia myös työssä käyville. 2suitable+PL+PAR also work+INE go/visit+PTC+PL+ALL

‘These cats show affection at intervals and are thus content to be in their own world for longspells; for this reason they are also suitable for those who go to work.’

(69) 2 Siamilainen sinisine silmineen omaa vankanSiamese+NOM blue+COM eye+COM possess+3SG solid+GEN

kannattajajoukon.2supporter+group+GEN

‘The Siamese with its blue eyes commands a solid group of supporters.’

(70) 2 Kaikista kissoista se on eniten riippuvainenall+PL+ELA cat+PL+ELA it+NOM be+3SG most dependent+NOM

omistajansa seurasta. 2owner+GEN company+ELA

‘Of all cats it is the most dependent on it’s owner’s company.’

Page 408: systematic functional linguistic

401

(71) 2 Se on hauskasti puhelias, 2 muttait+NOM be+3SG agreeably/merrily talkative+NOM but

juoksuaikoina sen mahtavat äänivaratrut+time+PL+ESS it+GEN powerful+NOM/PL sound+reserve+NOM/PL

näännyttävät kaikki vähemmänmake to be exhausted+3PL all+NOM little+COMP+GEN

tositarkoituksella leikkiin ryhtyneet. 2real+intention/purpose+ADE game/play+ILL undertake/take on+PTC/PL

‘It is agreeably talkative, but when it is on heat its powerful sound reserves exhaust all ofthose who got into the game with less than real resolve.’

(72) 2 Tämä haitta saadaan poistetuksithis+NOM drawback+NOM get+INDE eliminate+INDE/PTC+TRA

leikkauksen avulla. 2operation+GEN help+ADE

‘This drawback can be eliminated by the help of an operation.’

(73) 2 Sama koskee myös pientä ja siroasame+NOM touch/affect+3SG also small+PAR & elegant+PAR

burmaa, siamilaisen lähisukulaista. 2Burmese+PAR Siamese+GEN close+relative+PAR

‘The same goes for the small and elegant Burmese, a close relative of the Siamese.’

(74) 2 Kiharaturkkisesta rex-kissasta on omistajillawavy+coated+ELA Rex-cat+ELA be+3SG owner+PL+ADE

paljon hyvää kerrottavana. 2a lot good+PAR relate/tell+INDE/PTC+ESS

‘As for Rex cats with their wavy coats, owners have a lot of good things to say about them.’

Page 409: systematic functional linguistic

402

(75) 2 Sen turkista irtoava karva on niin näkymätöntä,it+GEN coat+ELA fall off+PTC hair+NOM be+3SG so invisible+PAR

Û että ylimääräistä siivoustyötä se eithat extra+PAR cleaning+work+PAR it+NOM NEG+3SG

juurikaan aiheuta. 2hardly/scarcely+TIS cause

‘The hair that falls off its coat is so invisible that it scarcely causes any extra cleaning work atall.’

(76) 2 Rodun heikkona kohtana on, etenkin aikaisemmin,breed+GEN weak+ESS point+ESS be+3SG particularly earlier

ollut joidenkin yksilöiden liian ohut turkki,be+PTC some+PL+GEN individuals+PL+GEN too thin+NOM coat+NOM

vika, Û joka on enemmän tuottanut huoltafault+NOM which+NOM be+3SG more produce+PTC worry/care+PAR

ulkopuolisille kuin omalle isäntäväelle! â 2outsider+PL+ALL than own+ALL master+folk+ALL

‘A weak point in the breed is, or has been (particularly earlier) that some individuals have acoat that is too thin, a defect that has been of more concern to outsiders than the cat’s masterand mistress.’

(77) 2 Hauskuus korvaa karvojenkin puutteen. 2pleasure/enjoyment+NOM replace+3SG hair+PL+GEN+TIS lack+GEN

‘The pleasure it brings makes up for (even) the lack of hair.’

(78) 2 Kissat tarjoavat mahdollisuuden laajempaankincat+NOM/PL offer+3PL chance+GEN extensive+COMP+ILL+TIS

harrastukseen sekä kissojen että niiden ystävien parissa. 2pastime+ILL both cat+PL+GEN and they+GEN friend+PL+GEN among+INE

‘Cats offer the chance to get involved in a more extensive pastime amongst both cats and catlovers.’

(79) 2 Suomessa toimii Suomen RotukissayhdistystenFinland+INE function+3SG Finland+GEN pedigree+cat+society+PL+GEN

liitto ry, Û johon kuuluu neljä kissayhdistystä. 2association/union which+ILL belong+3SG four cat+society+PAR

‘Finland has a registered association of Pedigree Cat Societies, to which are affiliated four catsocieties.’

Page 410: systematic functional linguistic

403

(80) 2 Kaksi niistä toimii Helsingissä,two+NOM they+ELA function+3SG Helsinki+INE

2 Turussa ja Tampereella ovat omat yhdistykset. 2Turku+INE & Tampere+ADE be+3PL own+NOM/PL society+NOM/PL

‘Two of them function in Helsinki, Turku and Tampere have their own societies.’

(81) 2 Yhdistykset järjestävät näyttelyitä, Û joista seuraavatsociety+NOM/PL organize+3PL show+PL+PAR which+PL+ELA next+NOM/PL

pidetään lokakuun ensimmäisenä viikonloppunahold+INDE October+GEN first+ESS week+GEN+end+ESS

Helsingissä, Pohjois-Haagan Yhteiskoululla. 2Helsinki+INDE North-Haaga+GEN Co-Educational Secondary School+ADE

‘The societies organize shows, and the next ones are during the first weekend in October inHelsinki at the North Haaga Co-Educational Secondary School.’

(82) 2 Lauantaina on mahdollisuus tutustuaSaturday+ESS be+3SG chance+NOM acquaint+PASS+INF

pitkäkarvarotuihin, 2 sunnuntaina esitelläänlong+hair+breed+PL+ILL Sunday+ESS present/introduce+INDE

lyhytkarvakissoja. 2short+hair+cat++PL+PAR

‘On Saturday it’s possible for you to get acquainted with long-haired breeds, on Sunday youcan be introduced to the short-haired cats.’

(83) 2 Näyttelyssä tavataan! 2show+INE meet+INDE

‘See you at the show!’

Page 411: systematic functional linguistic

405

Appendix 2: Data Sources

Note: Many of the names of people and places in spoken and unpublished data have been chan-ged in excerpts quoted in this study.

AA1 - AA2 - Spoken texts (conversation) originally recorded for a sociological study conductedby Kettil Bruun. Transcribed for the Conversation Analysis Project, Finnish Department, Uni-versity of Helsinki. Transcription & recording kept at Dept. of Finnish, University of Helsinki.(See A. Hakulinen (ed.) 1989.)

AA1 - Vapaa keskustelu 1.b. ryhmä 5. Transcribed by Eeva-Leena Seppänen.

AA2 - Vapaa keskustelu 1 ryhmä 7. Transcribed by Marja-Leena Sorjonen.

ÄA - Äidinkielen Avain. Finnish textbook for school children.

Äidinkielen avain 3-4. Leena Laulajainen, Mervi Miettinen, Hannele Ilves, & Kari Evinsalo,Weilin+Göös, Espoo, 1982.

Äidinkielen avain 7. Ritva Koskipää, Pirjo Nallikari, Kaija Parko, Aulikki Pönttinen, Wei-lin+Göös, Espoo, 1982.

Ad1 - Advertisement for the movie Kärpänen II (The Fly II)

Ad2 - 4 billboard advertisements for the Finnish commercial television channel (MTV). Thebillboards appeared at bus and tram stops in Helsinki, November, 1991.

Ad3 - Advertisement for a pep-up pill (Gerimax), which appeared in Helsinki City buses during1991.

AR - Akvaariorakkautta. A novel by Anna-Leena Härkönen (born 1965). Otava, Helsinki,1991.

CA1 - CA11. Spoken texts recorded and transcribed by students for the Conversation AnalysisSeminar (Keskusteluntutkimuksen praktikumi), Dept. of Finnish, University of Helsinki. Trans-cription & recording kept at Dept. of Finnish, University of Helsinki.

CA1 - Liisa Hakala, HOAS:n kohteen asukkaiden kokous (Informal meeting of flat-mates).

CA2 - Birgitta Lindblad, Opiskelutovereiden (3 naista) keskustelu (Conversation - femalestudents).

CA3 - Tuula Sulonen, Keskustelu autossa (Conversation recorded in a car).

CA4 - Markku Haakana, Haastattelu kahvilassa, toimittaja ja vammaisten olympialaisiinvalmistautuva (tuttuja) (Informal interview).

CA5 - Raija Kumpunen, Aikuinen ja lapsi (4v 2kk): piirtelyä pöydän ääressä (Adult & child).

CA6 - Kari Kivinen, Koululaisten pelikeskustelua (Conversation amongst school childrenplaying a game).

Page 412: systematic functional linguistic

406

CA7 - Päivi Keränen, Aikuinen ja lapsi (7v) pelaamassa (Adult & child playing a game).

CA8 - Jaana Sihvo, Yöihmisiä-ohjelman kaksi keskustelua (Late-night radio talk-back prog-ramme: callers talk about personal problems).

CA9 - 2 ELMU:n historiikan miestoimittajaa haastattelee yhtä Radio City:n naistoimittaja(interview).

CA10 - Ulla Jäntti.

CA12 - Päivi Kiiski.

FP Kalevi Wiik, Fonetiikan perusteet. Werner Söderström, Helsinki, 1981. An elementaryphonetics textbook.

Finnish Bible ) see Pyhä Raamattu.

Helsinki City Ordinance

HKV - Computerized corpus of written Finnish. Various genres of written Finnish, includingnewpaper and magazine articles and encyclopaedia entries. Compiled by Auli Hakulinen, FredKarlsson & Maria Vilkuna. (See bibliography.)

HS - Helsingin Sanomat, the leading national newpaper published in Helsinki.

HS kl - Helsingin Sanomien kuukausiliite, colour magazine (published once a month) for thenewpaper Helsingin Sanomat.

IL - Iltalehti, tabloid published in Helsinki.

IN - Ilkka Niiniluoto, Tiede, filosofia ja maailmankatsomus. Otava, Helsinki, 1984, p. 65.

JYK - Fred Karlsson, Johdatus yleiseen kielitieteeseen. Gaudeamus, Helsinki, 1976. An ele-mentary linguistics textbook.

KT - Koiramäen talossa, Mauri Kunnas. 1982. A children’s book (fiction).

NS - Nykysuomen sanakirja (Dictionary of Contemporary Finnish). 6 volumes. Werner Söder-ström, Helsinki, 1963.

OE - Otavan Suuri Ensyklopedia.

OH Overheard - This refers to something that I have overheard and have written down.

PL1 Personal letter - A letter written to me by someone I did not know.

Pyhä Raamattu (The Holy Bible). Suomen kirkon sisälähetysseura, Pieksämäki, 1953.

S - Suomi. A current affairs journal.

SE - Sinuhe egyptiläinen. A novel by Mika Waltari. WSOY, Helsinki, 1964.

Page 413: systematic functional linguistic

407

SIIIM2d - Interview done for a Labovian sociolinguistic analysis of the Finnish spoken in Hel-sinki. Transcription at Dept. of Finnish, University of Helsinki.

SIIIM3b - Interview done for a Labovian sociolinguistic analysis of the Finnish spoken in Hel-sinki. Transcription at Dept. of Finnish, University of Helsinki.

SIIN3c - Interview done for a Labovian sociolinguistic analysis of the Finnish spoken in Helsin-ki. Transcription kept at Dept. of Finnish, University of Helsinki.

SK - Suomen kuvalehti. A “quality” current affairs journal.

SS - Suomea suomeksi 1 by Olli Nuutinen. A textbook for foreigners learning Finnish. Suoma-lainen Kirjallisuuden Seura [The Finnish Literature Society], Helsinki 1977.

T - Tiede 2000. (Science 2000). A monthly magazine with articles on various scientific andacademic topics directed at a general audience.

TB - Tourist brochure for the city of Joensuu. Joensuun kaupungin matkailutoimisto, 1991.

Tel1 - Tel3. Telephone conversations transcribed for the Conversation Analysis Project, FinnishDepartment, University of Helsinki. Transcription & recording kept at Dept. of Finnish, Univer-sity of Helsinki.

Tel1 - “Seija-Liisa” - telephone conversation between female friends.

Tel2 - “Läskipuhelu” - telephone conversation between male friends.

Tel3 - “Muuraripuhelu”. Telephone conversation between female friends.

TH - Taikurin hattu by Tove Jansson. Werner Söderström, 1973. Translated from Swedish intoFinnish by Laila Järvinen.

TIIIN3c - Interview done for a Labovian sociolinguistic analysis of the Finnish spoken in Hel-sinki. Transcription kept at Dept. of Finnish, University of Helsinki.

TP1 - Turun puhekieli (Turku Spoken Finnish Project). Vapaa keskustelu (Casual Conversati-on): Tyttöystävien keskustelu. Transcription/recording kept at Dept. of Finnish and GeneralLinguistics, University of Turku, Turku.

TP2 - Turun puhekieli (Turku Spoken Finnish Project). Tilannenauhoitus: kahvikutsut (CasualConversation “Coffee Party”). Transcription/recording kept at Dept. of Finnish and GeneralLinguistics, University of Turku, Turku.

TSM - Alpo Ruuth, Tavallisen suomalainen mies. Tammi, Helsinki. 1982. A novel.

TT - Tieteessä tapahtuu 6/91 p. 1. A bulletin published by the Tieteellisten seurain valtuuskun-ta (Commission for Academic Societies).

WALTARI - The University of Helsinki Language Corpus Server. Department of GeneralLinguistics, University of Helsinki.

Y Yliopisto-lehti. A magazine published by the University of Helsinki.

Page 414: systematic functional linguistic

409

Appendix 3: Form Glosses

The phonological representations and explanations given below are meant only as a guide: theplural morpheme -i-, for example, engenders further phonological changes. (For more detail, seeChapter 3 and Karlsson 1983b.) Only back vowel variants are given. A capital V in the secondcolumn stands for the lengthening of the stem vowel.

Gloss

ABE . . . .

ABL . . . .

ACC . . . .

ADE . . . .

ADV . . . .

ALL . . . .

CAUS . . .

COM . . .

COMP . .

CON . . . .

ELA . . . .

ESS . . . . .

GEN . . . .

ILL . . . . .

IMP . . . .

INDE(F) .

INE . . . . .

INF . . . . .

INS . . . . .

NOM . . .

PAR . . . .

PAS . . . .

PAS/AUT . .

PER . . . .

PL . . . . . .

PLU . . . .

POS . . . .

POT . . . .

PP . . . . . .

PRO . . . .

PTC . . . .

Q . . . . . . .

REF/PAS

SG . . . . .

SUP . . . .

TRA . . . .

TIS . . . . .

UF . . . . .

-tta . . . . . . . . . . . . . . . . .

-lta . . . . . . . . . . . . . . . . .

-t . . . . . . . . . . . . . . . . . . .

-lla . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

-lle . . . . . . . . . . . . . . . . .

-tta- . . . . . . . . . . . . . . . .

-ine- . . . . . . . . . . . . . . . .

-mpi/-mp(a)- . . . . . . . . .

-isi- . . . . . . . . . . . . . . . . .

-sta . . . . . . . . . . . . . . . . .

-na . . . . . . . . . . . . . . . . .

-n . . . . . . . . . . . . . . . . . .

-Vn, -hVn, -seen . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

-da-/-t(t)a- (+ Vn) . . . . .

-ssa . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

-n . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

-(t)a . . . . . . . . . . . . . . . .

-i- . . . . . . . . . . . . . . . . . .

-u- . . . . . . . . . . . . . . . . . .

olla ‘be’ + ptc . . . . . . . .

-i-/-t . . . . . . . . . . . . . . . .

past of olla ‘be’ + ptc . .

. . . . . . . . . . . . . . . . . . . . .

-ne- . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

-itse . . . . . . . . . . . . . . . .

-va, -nut/neet . . . . . . . . .

-ko, mi-, ku- . . . . . . . . . .

-utu- . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

-in/-imp(a)-~

-ksi (-kse) . . . . . . . . . . . .

-han, -kin/-kaan, etc. . . .

. . . . . . . . . . . . . . . . . . . . .

Name

abessive . . . . . . . . . . . . . . . .

ablative . . . . . . . . . . . . . . . . .

accusative . . . . . . . . . . . . . . .

adessive . . . . . . . . . . . . . . . .

adverb

allative . . . . . . . . . . . . . . . . .

causative . . . . . . . . . . . . . . .

comitative . . . . . . . . . . . . . .

comparative

conditional

elative . . . . . . . . . . . . . . . . . .

essive . . . . . . . . . . . . . . . . . .

genitive . . . . . . . . . . . . . . . .

illative . . . . . . . . . . . . . . . . .

imperative

indefinite (“passive”)

inessive . . . . . . . . . . . . . . . .

infinitive

instructive . . . . . . . . . . . . . .

nominative (/absolutive) . . .

partitive . . . . . . . . . . . . . . . .

past tense

passive/automative . . . . . . .

perfect tense

plural . . . . . . . . . . . . . . . . . . .

pluperfect tense

possessive suffix . . . . . . . . . .

potential

pre- or postposition

prolative . . . . . . . . . . . . . . . .

participle

interrogative element . . . . . .

reflexive-passive . . . . . . . . .

singular

superlative

translative . . . . . . . . . . . . . . .

textual/interpersonal suffix .

uninflected form

Explanation/Comment

‘without’

‘from (by/on/near)’

human pronouns only

‘by/on/near’, instrument, means

‘to (by/one/near)’

derivational morpheme

‘with’

‘from (in)side, intimate contact’

‘as’

marks a dependent element

‘to (inside)/intimate contact’

see Figure 3-4

‘in(side)/intimate contact’

see Figure 3-9

‘by’

unmarked form, bounded entity

non-bounded entity

derivational affix

-t in nominative, -i- in other cases

see below

‘by way of’

suffix, particle or pronoun

derivational affix

‘into’

“pragmatic particle”

Finite Endings (see Chapter 3):

1SG -n 1PL -mme2SG -t 2PL -tte3SG -V 3PL -vat

Possessive Suffixes:

1SG -ni 1PL -mme2SG -si 2PL -nne

3SG/PL -nsa/Vn

Page 415: systematic functional linguistic

410

Appendix 4: Function Glosses

Ac . . . . . . . . . . .amb . . . . . . . . . .Att . . . . . . . . . .beh . . . . . . . . . .Car . . . . . . . . . .Circ . . . . . . . . . .cir . . . . . . . . . . .Dom . . . . . . . . .exp . . . . . . . . . .Ext . . . . . . . . . .Foc . . . . . . . . . .G . . . . . . . . . . . .Go . . . . . . . . . . .Id . . . . . . . . . . .int . . . . . . . . . . .Ir . . . . . . . . . . . .Loc:ed . . . . . . . .mat . . . . . . . . . .Med . . . . . . . . .men . . . . . . . . . .met . . . . . . . . . .Mod . . . . . . . . .N . . . . . . . . . . . .(NR) . . . . . . . . .Phen . . . . . . . . .Posit:ed . . . . . . .poss . . . . . . . . .Poss:ed . . . . . . .Poss:er . . . . . . .Pro . . . . . . . . . .Ra . . . . . . . . . . .rel . . . . . . . . . . .res . . . . . . . . . . .Say . . . . . . . . . .Sen . . . . . . . . . .sen . . . . . . . . . Th: int . . . . . . . .Th: top . . . . . . .Th: sub . . . . . . .temp . . . . . . . . .Tk . . . . . . . . . . .Vl . . . . . . . . . . .Ver . . . . . . . . . .

Actorambient (relational process)Attributebehavioural (material process)CarrierCircumstancecircumstantial (relational process)Domain (macro-role)experiencer process (subtype of material process)Extent (spatial/material Extent of a process)Focus (in information structure)GivenGoalIdentifiedIntensive (relational process)IdentifierLocatedmaterial processMedium (macro-role) mental process (internal & external consciousness)meteorological (subtype of material process)ModalityNewnon-inherent rolePhenomenonPositionedpossessivePossessedPossessorProcessRange (in a material process)relationalresultativeSayerSensersensory (relational intensive) processTheme: interpersonalExperiential Theme: Topical ThemeSubsidiary experiential themetemporary (relational intensive process)TokenValueVerbiage

Page 416: systematic functional linguistic

411

Appendix 5: Notational Conventions

+ . . . . . . .

, . . . . . . . . .

1 . . . . . . . . .

*** . . . . . . .

[ ] . . . . . . .

[[ ]] . . . . .

: . . . . . . . . .

/ . . . . . . . .

ú . . . . . . . .

! . . . . . . . .

v . . . . . . . .

morpheme boundary

phrase (or group) boundary

clause boundary

clause complex boundary

embedded phrase

embedded clause

delicacy (subclassification) . . .

conflation . . . . . . . . . . . . . . . . .

realization

unsequenced structure . . . . . . .

sequenced structure . . . . . . . . .

e.g. Process: relational: intensive

e.g. Ra/Th - conflation (Range & Theme)

e.g. Actor ! Process ! Circ

e.g. Actor v Process v Circ

<A> Speaker A

<B> Speaker B

Phonological Conventions:

// tone group boundary

/ foot boundary

v silent beat

kissa tonic syllable

kissa word containing a tonic syllable in a handwritten manuscript~~~~~

Page 417: systematic functional linguistic

412

Appendix 6: System Network Conventions

Page 418: systematic functional linguistic

413

Bibliography:

Alhoniemi, Alho (1975) “Eräistä suomen kielen paikallissijojen keskeisistä käyttötavoista”.

Sananjalka 17: 5-24.

Allwood, J., L-G Anderson & Ö. Dahl (1980) Logic in Linguistics. Cambridge University

Press: Cambridge.

Anderson, John (1971) The Grammar of Case. Cambridge University Press: Cambridge.

Anderson, John (1977) On Case Grammar: Prolegomena to a Theory of Grammatical

Relations. Humanities Press: London.

Anttila, Raimo (1977) Analogy. Mouton: The Hague.

Austerlitz, Robert (1987) “Uralic Languages”. In Bernard Comrie (Ed.), The World’s Ma-

jor Languages. Croom Helm: London. Pp. 567-576.

Austin, J.L. (1962) How To Do Things With Words. Second Edition. Oxford University

Press: Oxford.

Bakhtin, M.M. (1981) “Discourse in the Novel”. Translated by Caryl Emerson & Michael

Holqvist. In Michael Holqvist (Ed.), The Dialogic Imagination: Four Essays by

M.M.Bakhtin. University of Texas Press: Austin. Pp. 259-422.

Bakhtin, M.M. (1986 [1952-53]) “The Problem of Speech Genres”. Translated by Vern W.

McGee. In Caryl Emerson & Michael Holqvist (Eds.), M.M. Bakhtin: Speech Genres and

Other Late Essays. University of Texas Press: Austin. Pp. 60-102.

Benson, James D., Michael J. Cummings, & William S. Greaves (Eds.) (1988) Linguis-

tics in a Systemic Perspective. John Benjamins: Amsterdam.

Bernstein, Basil (1987) “Social Class, Codes and Communication”. In Ulrich Ammon, Nor-

bert Dittmar, & Klaus J. Mattheier (Eds.), Sociolinguistics/Soziolinguistik: An Interna-

tional Handbook for the Science of Language and Society/Ein internationales Hand-

buch zur Wissenschaft von Sprache und Gessellschaft, Volume 1. Walter de Gruyter:

Berlin. Pp. 563-579.

Berry, Margaret (1975) An Introduction to Systemic Linguistics: 1 Structures and Sys-

tems. St. Martin’s Press: New York.

Berry, Margaret (1980) They’re All Out of Step Except Our Johnny: A Discussion of Moti-

vation (or the Lack of it) in Systemic Linguistics. (Department of English Studies, Universi-

ty of Nottingham (mimeo).)

Berry, Margaret (1987) “Is Teacher an Unanalysed Concept?”. In M.A.K. Halliday & Robin

P. Fawcett (Eds.), New Developments in Systemic Linguistics: Volume 1 Theory and

Description. Frances Pinter: London. Pp. 41-63.

Bloomfield, Leonard (1933) Language. Henry Holt: New York.

Bohm, D. (1980) Wholeness and the Implicate Order. Routledge and Kegan Paul: London.

Page 419: systematic functional linguistic

414

Branch, Michael (1987) “Finnish”. In Bernard Comrie (Ed.), The World’s Major Langu-

ages. Croom Helm: London. Pp. 593-617.

Bridgen, Nigel (1984) “Towards a Functional Grammar of Aspect in Finnish”. In Casper de

Groot & Hannu Tommola (Eds.) Aspect Bound: A Voyage into the Realm of Germanic,

Slavonic and Finno-Ugrian Aspectology. Foris Publications: Dordrecht. Pp. 179-198.

Brown, Gillian and George Yule (1983) Discourse Analysis. Cambridge University Press:

Cambridge.

Butler, Christopher S. (1985) Systemic Linguistics: Theory and Applications. Batsford:

London.

Butt, David (1988) “Randomness, Order and the Latent Patterning of Text”. In David Birch

& Michael O’Toole (Eds.), Functions of Style. Pinter: London. Pp. 74-97.

Carlson, Lauri (1981) “Aspect and Quantification”. In Philip J. Teseschi & Annie Zaenan

(Eds.), Syntax and Semantics Volume 14: Tense and Aspect. Academic Press: New York.

Pp. 31-64.

Carlson, Lauri (1983) Dialogue Games. D. Reidel: Dordrecht.

Carlson, Lauri (1984) “Well” in Dialogue Games: A Discourse Analysis of the Interjec-

tion “Well” in Idealized Conversation. Pragmatics & Beyond 5. John Benjamins: Amster-

dam.

Carroll, J.B. (Ed.) (1956) Language, Thought and Reality: Selected Writings of Benjamin

Lee Whorf. M.I.T. Press: Cambridge, Mass.

Chafe, Wallace L. (1976) “Givenness, Contrastiveness, Definiteness, Subjects, Topics and

Points of View”. In Charles N. Li (Ed.) Subject and Topic pp. 25-55. Academic Press: New

York.

Chafe, Wallace (1987) “Cognitive Constraints on Information Flow”. In Russell S. Tomlin

(Ed.) Coherence and Grounding in Discourse pp. 21-51. Typological Studies in Language

II. John Benjamins, Amsterdam.

Chafe, Wallace (Ed.) (1980) The Pear Stories: Cognitive, Cultural, and Linguistic As-

pects of Narrative Production. Ablex: Norwood, N.J.

Chalmers, A.F (1982) What Is This Thing Called Science? 2nd edition. University of

Queensland Press: St. Lucia/London.

Chesterman, Andrew (1977) “Definiteness in Finnish”. Papers and Studies in Contrastive

Linguistics 7: 111-120. UAM: Pozna½.

Chesterman, Andrew (1991) On Definiteness: A Study with Special Reference to English

and Finnish. Cambridge Studies in Linguistics 56. Cambridge University Press: Cambrid-

ge.

Chomsky, Noam (1957) Syntactic Structures. Mouton: The Hague.

Chomsky, Noam (1965) Aspects of the Theory of Syntax. MIT Press: Cambridge, Mass.

Chomsky, Noam (1968) Language and Mind. Harcourt Brace Jovanich: New York.

Page 420: systematic functional linguistic

415

Chomsky, Noam (1976) Reflections on Language. Fontana: Glasgow.

Chomsky, Noam (1986) Knowledge of Language: Its Nature, Origin and Use. Convergen-

ce. Praeger: New York.

Collins Cobuild English Language Dictionary. Collins: London. 1987.

Comrie, Bernard (1975) “The Anti-Ergative: Finland’s Answer to Basque”. In R. Grossman

et al (Eds.), Papers from the Regional Meeting of the Chicago Linguistic Society 11.

Dept. of Linguistics, University of Chicago: Chicago. Pp. 112-121.

Comrie, Bernard (1976) Aspect. Cambridge University Press: Cambridge.

Comrie, Bernard (1977) “In Defense of Spontaneous Demotion: the Impersonal Passive”. In

Peter Cole & Jerrold M. Saddock (Eds.), Syntax and Semantics Volume 8: Grammatical

Relations. Academic Press: New York. Pp. 47-58.

Dahl, Östen (1980) “Some Arguments for Higher Nodes in Syntax: A Reply to Hudson’s

‘Constituency and Dependency’”. Linguistics 18: 485-488.

Dahl, Östen (1981) “On the Definition of the Telic-Atelic (Bounded-Nonbounded) Distincti-

on”. In Philip J. Teseschi & Annie Zaenan (Eds.), Syntax and Semantics Volume 14: Tense

and Aspect. Academic Press: New York. Pp. 79-90.

Daneš, František (Ed.) (1974) Papers on Functional Sentence Perspective. Czechoslovak

Academy of Sciences. Academia: Prague.

Daneš, František (1987) “On Prague School Functionalism in Linguistics”. In René Dirven

& Vilém Fried (Eds.), Functionalism in Linguistics. John Benjamins: Amsterdam. Pp. 3-

37.

Décsy, Gyula (1990) The Uralic Protolanguage: A Comprehensive Reconstruction. Bib-

liotheca Nostratica. Eurolingua: Bloomington.

Denison, Norman (1957) The Partitive in Finnish. Suomalaisen Tiedeakatemia: Helsinki.

Dik, Simon (1978) Functional Grammar. North-Holland Publishing Company: Amsterdam.

Dik, Simon (1980) Studies in Functional Grammar. Academic Press: London.

Dik, Simon (1987) Some Principles of Functional Grammar. In René Dirven & Vilém Fried

(Eds.) Functionalism in Linguistics. John Benjamins: Amsterdam. Pp. 81-100.

Dinneen, Francis (1967) An Introduction to General Linguistics. Holt, Rinehart & Wins-

ton: New York.

Dirven, René & Fried, Vilém (Eds.) (1987) Functionalism in Linguistics. John Benja-

mins: Amsterdam.

Downes, William (1984) Language and Society. Fontana Paperbacks: London.

Downing, Angela (1991) “An Alternative Approach to Theme: A Systemic-Functional Pers-

pective”. WORD 42: 119-143.

Dowty, David R. (1979) Word Meaning and Montague Grammar. Reidel: Dordrecht.

Eades, Diana (1982) “You Gotta Know How to Talk ...: Information Seeking in South-East

Queensland Aboriginal Society”. Australian Journal of Linguistics 2: 61-82.

Page 421: systematic functional linguistic

416

Eggins, Suzanne, J.R. Martin & Peter Wignell (1987) Writing Project Report. Working

Papers in Linguistics No. 5. Linguistics Department, University of Sydney: Sydney.

Ellis, Jeffrey (1987) “The Logical and Textual Functions”. In M.A.K. Halliday & Robin P.

Fawcett (Eds.), New Developments in Systemic Linguistics Volume 1: Theory and Desc-

ription. Frances Pinter: London. Pp. 107-129.

Fawcett, Robin P. (1980) Cognitive Linguistics and Social Interaction: Towards an In-

tegrated Model of a Systemic Functional Grammar and Other Components of the Com-

municating Mind. Julius Groos Verlag and Exeter University: Heidelberg.

Fawcett, Robin P. (1987) “The Semantics of the Clause and Verb for Relational Processes in

English”. In M.A.K. Halliday & Robin P. Fawcett (Eds.), New Developments in Systemic

Linguistics: Volume 1 Theory and Description. Frances Pinter: London. Pp. 130-183.

Fawcett, Robin P. (1988) “What Makes a ‘Good’ System Network Good? ) Four Pairs of

Concepts for Such Evaluations”. In James Benson & William Greaves (Eds.), Systemic

Functional Approaches to Discourse: Selected Papers from the 12th International Sys-

temic Workshop. Advances in Discourse Processes: Volume XXVI. Ablex: Norwood, New

Jersey. Pp. 1-28.

Fillmore, Charles J. (1968) “The Case for Case”. In Emmon Bach & Robert T. Harms

(Eds.), Universals in Linguistic Theory. Holt, Rinehart and Winston: New York. Pp. 1-88.

Fillmore, Charles J. (1981) “Pragmatics and the Description of Discourse”. In Peter Cole

(Ed.), Radical Pragmatics. Academic Press: New York. Pp. 143-66.

Firth, J.R. (1930) Speech. Benn’s Sixpenny Library: London. [Republished as Firth 1964,

see below.]

Firth, J.R. (1937) The Tongues of Men. Publisher not indicated: London. [Republished as

Firth 1964, see below.]

Firth, J.R. (1957) Papers in Linguistics 1934-1951. Oxford University Press: London.

Firth, J.R. (1964) The Tongues of Men and Speech. Oxford University Press, London.

Fishman, Joshua A. (1971) “The Sociology of Language: An Interdisciplinary Social Scien-

ce Approach to Language in Society”. In Joshua A. Fishman, Advances in the Sociology of

Language 1. Mouton: The Hague.

Flint, Aili (1980) Semantic Structure in the Finnish Lexicon: Verbs of Possibility and

Sufficiency. Suomalaisen Kirjallisuuden Seura [Finnish Literature Society]: Helsinki.

Foley, William & Van Valin Jr., Robert D. (1984) Functional Syntax and Universal

Grammar. Cambridge University Press: Cambridge.

Givón, Talmy (1979) On Understanding Grammar. Academic Press: New York.

Givón, Talmy (1982) “Logic vs. Pragmatics, with Human Language as Referee: Towards an

Empirically Viable Epistemology”. Journal of Pragmatics 6:81-133.

Page 422: systematic functional linguistic

417

Givón, Talmy (1986) “Prototypes: Between Plato and Wittgenstein”. In Colette Craig (Ed.),

Noun Classes and Categorization. Typological Studies in Language, Volume 7. John

Benjamins: Amsterdam. Pp. 77-102.

Grace, George (1987) The Linguistic Construction of Reality. Croom Helm: London.

Gregory, Michael (1967) “Aspects of Varieties Differentiation”. Journal of Linguistics 3:

177-198.

Gregory, Michael (1987) “Meta-Functions: Aspects of Their Development, Status and Use

in Systemic Linguistics”. In M.A.K. Halliday & Robin P. Fawcett (Eds.), New Develop-

ments in Systemic Linguistics: Volume 1 Theory and Description. Frances Pinter: Lon-

don. Pp. 94-106.

Gregory, Michael (1988) “Generic Situation and Register: A Functional View of Com-

munication. In James D. Benson, Michael J. Cummings, & William S. Greaves (Eds.)

(1988) Linguistics in a Systemic Perspective. John Benjamins: Amsterdam. Pp. 301-329.

Gregory, Michael & Carroll, Susanne (1978) Language and Situation: Language Va-

rieties in their Social Contexts. Routledge & Kegan Paul: London.

Hackman, D. (1978) “Rhythm in Finnish and English”. In E. Gårding, G. Bruce & R. Ban-

nert (Eds.), Nordic Prosody: Papers from a Symposium. Travaux de l’Institut de Linguis-

tique de Lund 13. Pp. 263-269.

Hakanen, Aimo (1972) “Normaalilause ja eksistentiaalilause”. Sananjalka 14: 36-76. [Sum-

mary: Normalsatz und Existentialsatz (p. 74 ) 76).]

Hakanen, Aimo (1973) Kontrastiivista lauseanalyysia: eksistentiaalilauseet. Turun yli-

opiston suomen kielen laitoksen julkaisuja 1: Turku.

Hakulinen, Auli (1976) “Liitepartikkelin -han/-hän syntaksia”. Reports on Text Linguis-

tics: Suomen kielen generatiivista lauseoppia 2. Meddelanden från Stiftelsens för Åbo

Akademi forskningsinstitut, Nr 7. Åbo Akademi: Åbo/Turku. Pp. 1-60. [Summary: On the

Syntax and Pragmatics of the Clitic -han/-hän (5 pages).]

Hakulinen, Auli (1983) “Subjektikategoria vai nominaalijäsenten subjektimaisuus”. In Auli

Hakulinen & Pentti Leino (Eds.) Nykysuomen rakenne ja kehitys 1: Näkökulmia kielen

rakenteisiin. Tietolipas 93. Suomalaisen Kirjallisuuden Seura [Finnish Literature Society]:

Helsinki.

Hakulinen, Auli (1989) “Some Notes on Thematics, Topic, and Typology”. In Maria-Elisa-

beth Conte, János S. Petöfi and Emil Sözer (Eds.), Text and Discourse Connectedness pp.

53 ) 63. Studies in Language Companion Series, Volume 16. John Benjamins, Amsterdam.

Hakulinen, Auli (1991) “Erään haastattelun anatomia”. In Tapani Lehtinen & Susanna Shore

(Eds.), Kieli, valta, eriarvoisuus. Kieli 6. Helsingin yliopiston suomen kielen laitos [Dept.

of Finnish at the University of Helsinki]: Helsinki. Pp. 166-186.

Hakulinen, Auli (Ed.) (1989) Suomalaisen keskustelun keinoja I. Kieli 4. Dept. of Finnish,

University of Helsinki: Helsinki.

Page 423: systematic functional linguistic

418

Hakulinen, Auli & Karlsson, Fred (1979) Nykysuomen lauseoppia. Suomalaisen Kirjalli-

suuden Seura [Finnish Literature Society]: Helsinki.

Hakulinen, Auli, Karlsson, Fred & Vilkuna, Maria (1980) Suomen tekstilauseiden

piirteitä: kvantitatiivinen tutkimus. Department of General Linguistics, University of Hel-

sinki: Helsinki.

Hakulinen, Auli & Sorjonen, Marja-Leena (1986) “Keskustelun kuvaamisesta”. In Pentti

Leino & Jyrki Kalliokoski (Eds.), Kieli 1. Helsingin yliopiston suomen kielen laitos: Hel-

sinki. Pp. 5-38.

Hakulinen, Lauri (1961) The Structure and Development of the Finnish Language.

Translated by John Atkinson. Translation of an earlier edition of L. Hakulinen (1979). Ura-

lic and Altaic Series, Volume 3. Indiana University, Bloomington. Mouton: The Hague.

Hakulinen, Lauri (1979) Suomen kielen rakenne ja kehitys. 4th Amended and Revised

Edition. Otava: Helsinki.

Halle, Morris & Vergnaud, Jean Roger (1982) “On the Framework of Autosegmental

Phonology”. In Harry van der Hulst & Norval Smith (Eds.), The Structure of Phonological

Representations (Part I). Foris Publications: Dordrecht. Pp. 65-82.

Halliday, M.A.K. (1956) “Grammatical Categories in Modern Chinese”. Transactions of

the Philological Society 177-224.

Halliday, M.A.K. (1961) “Categories of the Theory of Grammar”. WORD 17: 241-292.

Halliday, M.A.K. (1963) “Class in Relation to the Axes of Chain and Choice in Language”.

Linguistics 2: 5-15.

Halliday, M.A.K. (1966a) “Some Notes on “Deep” Grammar”. Journal of Linguistics 2:

57-67.

Halliday, M.A.K. (1966b) “The Concept of Rank: A Reply”. Journal of Linguistics 2: 110-

118.

Halliday, M.A.K. (1967a) “Notes on Transitivity and Theme in English: Part 1". Journal of

Linguistics 3: 37-81.

Halliday, M.A.K. (1967b) “Notes on Transitivity and Theme in English: Part 2". Journal of

Linguistics 3: 199-244.

Halliday, M.A.K. (1967c) Intonation and Grammar in British English. Mouton: The Ha-

gue.

Halliday, M.A.K. (1968) “Notes on Transitivity and Theme in English: Part 3". Journal of

Linguistics 4: 179-215.

Halliday, M.A.K. (1973) Explorations in the Functions of Language. Edward Arnold:

London.

Halliday, M.A.K. (1975) “Language as Social Semiotic: Towards a General Sociolinguistic

Theory”. In Adam Makkai & Valerie Becker Makkai (Eds.) The First Lacus Forum 1974.

Hornbeam Press: Columbia, Southern Carolina. Pp. 17-46.

Page 424: systematic functional linguistic

419

Halliday, M.A.K. (1977) Aims and Perspectives in Linguistics. ALAA Occasional Papers

Number 1. Applied Linguistics Association of Australia: Mt. Gravatt, Brisbane.

Halliday, M.A.K. (1978) Language as Social Semiotic: The Social Interpretation of Lan-

guage and Meaning. Edward Arnold: London.

Halliday, M.A.K. (1981a) “Text Semantics and Clause Grammar: Some Patterns of Realizati-

on”. In J.E. Copeland & P.W. Davis (Eds.), The Seventh Lacus Forum 1980. Hornbeam

Press: Columbia, S.C. Pp. 31-59.

Halliday, M.A.K. (1981b [1965]) “Types of Structure”. In M.A.K. Halliday & J.R. Martin

(Eds.), Readings in Systemic Linguistics. Batsford Academic: London. Pp. 29-41.

Halliday, M.A.K. (1984) “Language as Code and Language as Behaviour: A Systemic-Func-

tional Interpretation of the Nature and Ontogenesis of Dialogue”. In Robin P. Fawcett,

M.A.K. Halliday, Sydney Lamb & Adam Makkai (Eds.), The Semiotics of Culture and

Language, Volume 1: Language as Social Semiotic. Frances Pinter: London. Pp. 3-35.

Halliday, M.A.K. (1985a) An Introduction to Functional Grammar. Edward Arnold: Lon-

don.

Halliday, M.A.K. (1985b) Spoken and Written Language. Deakin University Press: Deakin

University, Victoria.

Halliday, M.A.K. (1987) “Language and the Order of Nature”. In Nigel Fabb, Derek Attrid-

ge, Alan Durant, & Colin MacCabe (Eds.), The Linguistics of Writing: Arguments Bet-

ween Language and Literature. Manchester University Press: Manchester. Pp. 135-154.

Halliday, M.A.K. (1988a) “On the Ineffability of Grammatical Categories”. In James D.

Benson, Michael J. Cummings, & William S. Greaves (Eds.), Linguistics in a Systemic

Perspective. John Benjamins: Amsterdam. Pp. 27-51.

Halliday, M.A.K. (1988b) “On the Language of Physical Science”. Registers of Written

English: Situational Factors and Linguistic Features. Frances Pinter: London. Mohsen

Ghadessy. Pp. 162-178.

Halliday, M.A.K. (1991) “The Notion of ‘Context’ in Language Education”. In Thao Lê and

Mike McCausland (Eds.) Language Education: Interaction and Development (Procee-

dings of the International Conference held in Ho Chi Minh City, Vietnam, 30 March ) 1

April, 1991.) University of Tasmania: Launceston. Pp. 1-26.

Halliday, M.A.K. (in press) “How Do You Mean?”. In Martin Davies & Louise Ravelli

(Eds.) Recent Advances in Systemic Linguistics: Theory and Practice. (Papers from the

XVIIth International Systemic Congress, Stirling, Scotland, July 1990.) Frances Pinter:

London.

Halliday, M.A.K. (forthcoming) “The Act of Meaning”. To appear in Proceedings of the

Georgetown University Roundtable Meeting on Linguistics and Language Study,

Washington, D.C., 1992. Georgetown University Press.

Page 425: systematic functional linguistic

420

Halliday, M.A.K. & Fawcett, Robin P. (Eds.) (1987) New Developments in Systemic

Linguistics, Volume 1: Theory and Description. Frances Pinter: London.

Halliday, M.A.K. & Hasan, Ruqaiya (1976) Cohesion in English. Longman: London.

Halliday, M.A.K. & Hasan, Ruqaiya (1985) Language, Context, and Text: Aspects of

Language in a Social-Semiotic Perspective. Deakin University Press: Deakin University,

Victoria.

Halliday, M.A.K. & Martin, J.R. (Eds.) (1981) Readings in Systemic Linguistics. Bats-

ford Academic: London.

Halliday, M.A.K., McIntosh, A., & Strevens, P.D. (1964) The Linguistic Sciences and

Language Teaching. Longmans: London.

Harris, Roy (1980) The Language-Makers. Duckworth: London.

Harris, Roy (1981) The Language Myth. Duckworth: London.

Harris, Roy (1987a) “Language as Social Interaction: Integrationalism Versus Segregationa-

lism”. Language Sciences 9: 131-143.

Harris, Roy (1987b) The Language Machine. Duckworth: London.

Hartmann, R.R.K. & F.C. Stork (1972) Dictionary of Language and Linguistics. Ap-

plied Science Publishers Ltd.: London.

Hasan, Ruqaiya (1971) “Syntax and Semantics”. In John Morton (Ed.), Biological and

Social Factors in Psycholinguistics. Logos Press: London. Pp. 131-157.

Hasan, Ruqaiya (1984a) “Ways of Saying: Ways of Meaning”. In Robin P. Fawcett, M.A.K.

Halliday, Sydney Lamb & Adam Makkai (Eds.), The Semiotics of Culture and Language,

Volume 1: Language as Social Semiotic. Frances Pinter: London. Pp. 105-162.

Hasan, Ruqaiya (1984b) “The Nursery Tale as a Genre”. Nottingham Linguistic Circular

13. M. Berry, M. Stubbs, & R. Carter (Eds.): Special Issue on Linguistics. Nottingham Uni-

versity Press.

Hasan, Ruqaiya (1985a) “Meaning, Context and Text: Fifty Years After Malinowski”. In

James D. Benson & William S. Greaves (Ed.), Systemic Perspectives on Discourse, Volu-

me 1. Ablex: Norwood, New Jersey. Pp. 16-49.

Hasan, Ruqaiya (1985b) Linguistics, Language, and Verbal Art. Deakin University Press:

Deakin University, Victoria.

Hasan, Ruqaiya (1987a) “Directions from Structuralism”. In Nigel Fabb, Derek Attridge,

Alan Durant, & Colin MacCabe (Eds.), The Linguistics of Writing: Arguments Between

Language and Literature. Manchester University Press: Manchester. Pp. 103-122.

Hasan, Ruqaiya (1987b) “The Grammarian’s Dream: Lexis as Most Delicate Grammar”. In

M.A.K. Halliday & Robin P. Fawcett (Eds.), New Developments in Systemic Linguistics:

Volume 1 Theory and Description. Frances Pinter: London. Pp. 184-211.

Hasan, Ruqaiya (1988) “Language in the Processes of Socialisation: Home and School”. In

Linda Gerot, Jane Oldenburg, Theo Van Leeuwen (Eds.), Proceedings from the Working

Page 426: systematic functional linguistic

421

Conference on Language in Education Macquarie University 17-21 November 1986.

Macquarie University: Macquarie University, Sydney. Pp. 36-95.

Hasan, Ruqaiya (1989) “Semantic Variation and Sociolinguistics”. Australian Journal of

Linguistics 9: 221-275.

Hasan, Ruqaiya (in press) “The Conception of Context in Text”. In Peter Fries & Michael

Gregory (Eds.) Discourse in Society: A Systemic-Functional Perspective. Vol. 3 in the

series Meaning and Choice in Language: Studies for Michael Halliday. Ablex: Norwood,

New Jersey.

Hasan, Ruqaiya (MS) Offers in the Making: A Systemic-Functional Approach. Macquarie

University, Sydney.

Hasan, Ruqaiya & Cloran, Carmel (1990) “A Sociolinguistic Study of Everyday Talk

Between Mothers and Children”. In M.A.K. Halliday, John Gibbons, & Howard Nicholas

(Eds.) Learning Keeping and Using Language: Volume I. (Selected Papers from the 8th

World Congress of Applied Linguistics, Sydney, 16-21 August 1987. John Benjamins:

Amsterdam. Pp. 67-99.

Heinämäki, Orvokki (1981) “On the meaning of olla + 3rd infinitive inessive in Finnish”. In

Congressus Quintus Internationalis Finno-Ugristium, Turku 1980: Pars VI. Suomen

kielen seura: Turku. Pp. 383-388.

Heinämäki, Orvokki (1984) “Aspect in Finnish”. In Casper de Groot & Hannu Tommola

(Eds.) Aspect Bound: A Voyage into the Realm of Germanic, Slavonic and Finno-Ugri-

an Aspectology. Foris Publications: Dordrecht. Pp. 153-177.

Helasvuo, Marja-Liisa (1987) Subjekteina ja objekteina toimivat nominilausekkeet

puhutuissa teksteissä. (Unpublished Master’s Thesis, Dept. of Finnish, University of Hel-

sinki.)

Helasvuo, Marja-Liisa (1988) Subjekteina ja objekteina toimivat nominilausekkeet

puhutussa suomessa. (Unpublished Licentiate Thesis, Dept. of Finnish, University of Hel-

sinki.)

Helasvuo, Marja-Liisa (in press) “Relative Clauses in the Finnish Pear Stories”. In Urpo

Nikanne & Anders Holmberg (Eds.), Case and Other Topics in Finnish Syntax. Mouton:

The Hague.

Henderson, Eugénie J.A. (1987) “J.R. Firth in Retrospect: A View from the Eighties”. In

Ross Steele & Terry Threadgold (Eds.), Language Topics: Essays in Honour of Michael

Halliday. Volume I. John Benjamins: Amsterdam. Pp. 57-68.

Henrici, A. (1981 [1966]) “Some Noteson the Systemic Generation of a Paradigm of the

English Clause”. In M.A.K. Halliday & J.R. Martin (Eds.), Readings in Systemic Linguis-

tics. Batsford Academic: London. Pp. 74-98.

Hirvonen, Pekka (1970) Finnish and English Communicative Intonation. Publications of

the Phonetics Department of the University of Turku, No. 8: Turku.

Page 427: systematic functional linguistic

422

Hjelmslev, Louis (1953) Prolegomena to a Theory of Language. Translated by Francis J.

Whitfield. Waverly Press: Baltimore.

Hjelmslev, Louis (1959 [1938]) “Essai d’une theorie des morphemes”. In Louis Hjelmslev

(Ed.), Essais linguistiques. Nordisk Sprog- og Kulturforlag: Copenhagen.

Hockett, Charles (1942) “A System of Descriptive Phonology”. Language 18: 3-21.

Hockett, Charles F. (1968) The State of the Art. Janua Linguarum Series Minor Nr.

LXXIII. Mouton: The Hague.

Huddleston, Rodney (1965) “Rank and Depth”. Language 41: 574-586.

Huddleston, Rodney (1984) Introduction to the Grammar of English. Cambridge Univer-

sity Press: Cambridge.

Huddleston, Rodney (1988) “Constituency, Multi-Functionality and Grammaticalization in

Halliday’s Functional Grammar”. Journal of Linguistics 24: 137-74.

Hudson, R.A. (1974) “Systemic Generative Grammar”. Linguistics 139: 5-42.

Hudson, Richard A. (1976) Arguments for a Non-transformational Grammar. The Uni-

versity of Chicago Press: Chicago.

Hudson, Richard (1980a) “Constituency and Dependency”. Linguistics 18: 179-98.

Hudson, Richard (1980b) “A Second Attack on Constituency: A Reply to Dahl”. Linguis-

tics 18: 489-504.

Hudson, Richard (1984) Word Grammar. Basil Blackwell: Oxford.

Hudson, Richard (1986) “Systemic Grammar”. Linguistics 24: 791-815.

Hudson, R.A. (1987) “Daughter Dependency Theory and Systemic Grammar”. In M.A.K.

Halliday & Robin P. Fawcett (Eds.), New Developments in Systemic Linguistics: Volume

1 Theory and Description. Frances Pinter: London. Pp. 246-257.

Hyman, Larry M. (1975) Phonology: Theory and Analysis. Holt, Rinehart & Winston:

New York.

Hymes, Dell H. (1967) “Models of Interaction of Language and Social Setting”. Journal of

Social Issues 23.

Hymes, Dell (1986) “Models of the Interaction of Language and Social Life”. In John J.

Gumperz & Dell Hymes (Eds.), Directions in Sociolinguistics: The Ethnography of Com-

munication. Basil Blackwell: Oxford. Pp. 35-71.

Iivonen, Antti (1976) Fonetiikan terminologisia selityksiä. Unpublished teaching mimeo-

graph. Dept. of Phonetics, University of Helsinki.

Iivonen, Antti, Terttu Nevalinen, Reijo Aulanko, Hannu Kaskinen (1987) Puheen

intonaatio. Gaudeamus: Helsinki.

Ikola, Osmo (1954) “Suomen lauseopin ongelmia I ) III”. Virittäjä 58:209-245.

Ikola, Osmo (1977) Nykysuomen käsikirja. Weilin+Göös: Espoo.

Itkonen, Esa (1978) Grammatical Theory and Metascience. Benjamins: Amsterdam.

Itkonen, Esa (1983a) Causality in Linguistic Theory. Croom Helm: London.

Page 428: systematic functional linguistic

423

Itkonen, Esa (1983b) “The Use of a Corpus in the Study of Syntax: Historical and Method-

ological Considerations”. In Kaisa Häkkinen (Ed.), The Status of Corpus-Based Study in

Syntax. Publications of the Linguistic Association of Finland: Turku.

Itkonen, Esa (1990) “In principio erat verbum”. (Review of Anneli Pajunen, Verbien

leksikaalinen kuvaus). Virittäjä 94: 352-358.

Itkonen, Esa (1991) “Two Notions of Universal Grammar”. In Maria Vilkuna & Arto Anttila

(Eds.) SKY 1991. (1991 Yearbook of the Linguistic Association of Finland). Suomen

kielitieteellinen yhdistys/Linguistic Association of Finland, Helsinki. Pp. 53-90.

Itkonen, Terho (1974a) Erään sijamuodon ongelmia. Opuscula Instituti Linguae Fennicae,

Universitas Helsingiensis 53. Helsingin yliopiston suomen kielen laitos: Helsinki.

Itkonen, Terho (1974b) “Ergatiivisuutta suomessa I”. Virittäjä 78: 379-398.

Itkonen, Terho (1975) “Ergatiivisuutta suomessa II”. Virittäjä 79: 31-65.

Itkonen, Terho (1979) “Subject and Object Marking in Finnish: An Inverted Ergative Sys-

tem and an “Ideal” Ergative System”. In Frans Plank (Ed.), Ergativity. Academic Press:

London. Pp. 79-102.

Itkonen, Terho (1980) “Spesies suomessa ja germaanisissa kielissä”. Virittäjä 84: 27-38.

[Summary: Definiteness in Finnish and the Germanic Languages (pp. 37 ) 38).]

Jackendoff, Ray (1983) Semantics and Cognition. MIT Press: Cambridge, Mass.

Jespersen, Otto (1924) The Philosophy of Grammar. George Allen & Unwin: London.

Joseph, John E. & Taylor, Talbot J. (Eds.) (1990) Ideologies of Language. Routledge:

New York.

Kalliokoski, Jyrki (1989) Ja: rinnastus ja rinnastuskonjunktion käyttö. Suomalaisen

Kirjallisuuden Seura [Finnish Literature Society]: Helsinki. (English abstract (1 page): Ja:

Coordination and the use of the coordinating conjunction in Finnish.)

Kangasmaa-Minn, Eeva (1991) “Genetiivin tehtävistä “. Kielikello 2: 3-9.

Kangasmaa-Minn, Eeva (1971) “Verbaalilauseen perusrakenteesta”. Virittäjä 75: 255-262.

[Zur Grundsstruktur des Verbalsatzes (p. 261 -262).]

Karlsson, Fred (1978a) Finsk grammatik. Suomalaisen Kirjallisuuden Seura [Finnish Liter-

ature Society]: Helsinki.

Karlsson, Fred (1978b) “Nominaalilausekkeen tematiikka ja eksistentiaalilauseiden ongel-

ma”. Rakenteita. Publications of the Department of Finnish and General Linguistics, Uni-

versity of Turku: Turku. Pp. 293-305.

Karlsson, Fred (1983a) “Prototypes as Models for Linguistic Structure”. In Fred Karlsson

(Ed.), Papers from the Seventh Scandinavian Conference of Linguistics II. Publications

of the Department of General Linguistics, University of Helsinki No. 10. Department of

General Linguistics, University of Helsinki: Helsinki. Pp. 583-604.

Karlsson, Fred (1983b) Finnish Grammar. Werner Söderström: Helsinki.

Page 429: systematic functional linguistic

424

Karlsson, Fred (1983c) Suomen kielen äänne- ja muotorakenne. Werner Söderström: Hel-

sinki.

Karvonen, Pirjo (1991) “Kieliopillinen metafora”. In Tapani Lehtinen & Susanna Shore

(Eds.), Kieli, valta, eriarvoisuus. Helsingin yliopiston suomen kielen laitos [Dept. of Finn-

ish at the University of Helsinki]: Helsinki. Pp. 149-165.

Karvonen, Pirjo (1992) Oppikirjojen kieli ja tiedonkäsitys: Lukion biologian ja maantiedon

kirjojen leksikaaliskieliopillinen ja tekstuaalinen analyysi. (Unpublished Licentiate Thesis,

Dept. of Finnish, University of Helsinki.)

Keenan, Edward L. (1976) “Towards a Universal Definition of ‘Subject’”. In Charles N. Li

(Ed.), Subject and Topic. Academic Press: New York. Pp. 000-000.

Keenan, Edward L. (1984) “Semantic Correlates of the Ergative/Absolutive Distinction”.

Linguistics 22: 197-223.

Keller, Evelyn Fox (1985) Reflections on Gender and Science. Yale University Press: New

Haven.

Kelomäki, Tapani (1988) Identiteetti lauserakenteessa. Unpublished Licentiate Thesis, Dept.

of General Linguistics, University of Helsinki, Helsinki.

Kenny, Anthony (1984) The Legacy of Wittgenstein. Basil Blackwell: Oxford.

Koivisto, Vesa (1991) Suomen verbikantaisten -UtU-verbijohdosten semantiikkaa. Suomi

161. Suomalaisen Kirjallisuuden Seura [Finnish Literature Society]: Helsinki.

Korhonen, Mikko (1986) Finno-Ugrian Language Studies in Finland 1828-1918. Trans-

lated by Nigel Brigden. Societas Scientiarum Fennica: Helsinki.

Korhonen, Mikko (1981) Johdatus lapin kielen historiaan. Suomalaisen Kirjallisuuden

Seura: Helsinki.

Kress, Gunther (1985) Linguistic Processes in Sociocultural Practice. Deakin University

Press: Deakin University, Victoria.

Kress, Gunther (Ed.) (1976) Halliday: System and Function in Language. Oxford Uni-

versity Press: Oxford.

Kuhn, Thomas (1970) The Structure of Scientific Revolutions. The University of Chicago

Press: London.

Kuiri, Kaija (1984) Referointi Kainuun ja Pohjois-Karjalan murteissa. Suomalaisen

Kirjallisuuden Seura [Finnish Literature Society]: Helsinki.

Kulonen, Ulla-Maija (1989) The Passive in Ob-Ugrian. Suomalais-ugrilaisen Seuran

toimituksia/Mémoires de la Société Finno-Ougrienne Vol. 203. Suomalais-ugrilainen

Seura/Finno-Ugrian Society: Helsinki.

Kulonen-Korhonen, Ulla (1985) “Deverbaalisten U-johdosten semantiikkaa”. Virittäjä 89:

290-309. [Summary: Über die Semantik der finnischen deverbalen Verbalableitungen mit -

U- (p. 309).]

Page 430: systematic functional linguistic

425

Kytömäki, Leena (1978) “Kuratiivikausatiivit”. Rakenteita. (Juhlakirja Osmo Ikolan 60-

vuotispäiväksi.) Publications of the Department of Finnish and General Linguistics, Univer-

sity of Turku: Turku. Pp. 129 ) 150.

Kytömäki, Leena (1985) “Kielioppia kieliopin vuoksi: sillä-konjunktio äidinkielen

oppikirjoissa”. Sananjalka 27: 45-61. Suomen kielen seuran vuosikirja. [Summary:

Grammatik um der Grammatik willen: dies Darstellung der Konjunktion sillä ‘denn’ in den

muttersprachlichen Grammatiken.]

Kytömäki, Leena (1989) “Teettoverbit: johdon ja taivutuksen välimaastoa”. Sananjalka 31:

61 ) 79. Suomen kielen seuran vuosikirja. [Summary (2 pages): Factitive verbs: the grey

area between derivation and inflection.]

Laitinen, Lea (1988) “Naiset ja roolit. Koti- ja ansioäitien keskustelu 1987". In Lea Laitinen

(Ed.) Isosuinen nainen: tutkielmia naisesta ja kielestä. Yliopistopaino [University Press]:

Helsinki. Pp. 159-184.

Lakoff, George (1973) “Hedges: A Study in the Meaning Criteria and the Logic of Fuzzy

Concepts”. Journal of Philosophical Logic 2.

Lakoff, George (1977) “Linguistic Gestalt”. In W.A. Beach et al. (Eds.), Papers from the

Thirteenth Regional Meeting of the Chicago Linguistic Society. Chicago Linguistic Soci-

ety: Chicago. Pp. 236-287.

Lakoff, George (1982) “Categories and Cognitive Models”. Berkeley Cognitive Science

Report No. 2. Institute for Human Learning, University of California: Berkeley.

Lakoff, George (1987) Women, Fire, and Dangerous Things. The University of Chicago

Press: Chicago.

Lakoff, George & Johnson, Mark (1980) Metaphors We Live By. The University of Chi-

cago Press: Chicago.

Langacker, Ronald W. (1982) “Space Grammar, Analysability, and the English Passive”.

Language 58: 22-80.

Langacker, Ronald W. (1987) “Nouns and Verbs”. Language 63: 53-94.

Larjavaara, Matti (1990) Suomen deiksis. Suomalaisen Kirjallisuuden Seura [Finnish Lit-

erature Society]: Helsinki.

Larjavaara, Matti (1991) Aspektuaalisen objektin synty. Virittäjä 95: 372- 408. [Summary

(2 pages): The Origin of the Aspectual Object.]

Laury, Ritva (1991) “On the Development of the Definite Article se in Spoken Finnish. In

Maria Vilkuna & Arto Anttila (Eds.) SKY 1991. (1991 Yearbook of the Linguistic Associa-

tion of Finland). Suomen kielitieteellinen yhdistys/Linguistic Association of Finland, Hel-

sinki. Pp. 93-121.

Leech, Geoffrey (1974) Semantics: the Study of Meaning. Penguin: Harmondsworth.

Leech, Geoffrey N. (1983) Principles of Pragmatics. Longmans: London.

Page 431: systematic functional linguistic

426

Lehikoinen, Laila & Kiuru, Silva (1991) Kirjasuomen kehitys. Helsingin yliopiston suo-

men kielen laitos: Helsinki.

Lehtinen, Tapani (1984) Itämerensuomen passiivin alkuperästä. Suomi 129. Suomalaisen

Kirjallisuuden Seura [Finnish Literature Society]: Helsinki.

Leino, Pentti (1982) Suomen kielen lohkolause. Suomalaisen Kirjallisuuden Seura [Finnish

Literature Society]: Helsinki.

Leino, Pentti (1986) Language and Metre: Metrics and the Metrical System of Finnish.

Studia Fennica 31. Finnish Literature Society/Suomalaisen Kirjallisuuden Seura: Helsinki.

Leino, Pentti (1989) Paikallissijat ja suhdesääntö: kognitiivisen kieliopin näkökulma.

Virittäjä 93: 161-219. [Summary (4 pages): Local cases and the relation rule: an approach

from cognitive grammar.]

Leino, Pentti (1990) “Spatial Relations in Finnish: a Cognitive Perspective”. In Ingrid Alm-

qvist, Per-Erik Cederholm, & Jarmo Lainio (Eds.), Från Pohjolas pörten till kognitiv

kontakt. Stockholm Studies in Finnish Language and Literature No. 6 (Festschrift for Erling

Wande). Department of Finnish, University of Stockholm: Stockholm. Pp. 117-152.

Leino, Pentti (1991) Lauseet ja tilanteet: Suomen objektin ongelmia. Suomalaisen

Kirjallisuuden Seura [Finnish Literature Society]: Helsinki.

Leino, Pentti, Helasvuo, Marja-Liisa, Lauerma, Petri, Nikanne, Urpo & Onikki, Tiina

(1990) Suomen kielen paikallissijat konseptuaalisessa semantiikassa. Kieli 5. Helsingin

yliopiston suomen kielen laitos [Dept. of Finnish at the University of Helsinki]: Helsinki.

Lemke, Jay (1988) “Discourses in Conflict: Heteroglossia and Text Semantics”. In James

Benson & William Greaves (Eds.), Systemic Functional Approaches to Discourse: Se-

lected Papers from the 12th International Systemic Workshop. Advances in Discourse

Processes: Volume XXVI. Ablex: Norwood, New Jersey. Pp. 29-50.

Lemke, Jay (1989) “Semantics and Social Values”. WORD 40: 37-50.

Lemke, Jay L. (1990) “Technical Discourse and Technocratic Society”. In M.A.K. Halliday,

John Gibbons, & Howard Nicholas (Eds.) Learning Keeping and Using Language: Vol-

ume II. (Selected Papers from the 8th World Congress of Applied Linguistics, Sydney, 16-

21 August 1987. John Benjamins: Amsterdam. Pp. 435-460.

Leskinen, Heikki (1970) Imperatiivin muodostus itämerensuomalaisissa kielissä. Suomi

115. Suomalaisen Kirjallisuuden Seura [Finnish Literature Society]: Helsinki.

Levinson, Stephen C. (1983) Pragmatics. Cambridge University Press: Cambridge.

Lyne, A.A. (1988) “Systemic Syntax from a Lexical Point of View”. In James D. Benson,

Michael J. Cummings, & William S. Greaves (Eds.), Linguistics in a Systemic Perspec-

tive. Current Issues in Linguistic Theory, Volume 39. John Benjamins: Amsterdam. Pp. 53-

71.

Lyons, John (1966) “Firth’s Theory of Meaning”. In C.E. Bazell, J.C. Catford, M.A.K. Halli-

day, & R.H. Robin (Eds.), In Memory of J.R. Firth. Longmans: London. Pp. 288-302.

Page 432: systematic functional linguistic

427

Lyons, John (1977a) Semantics: Volume 1. Cambridge University Press: Cambridge.

Lyons, John (1977b) Semantics: Volume 2. Cambridge University Press: Cambridge.

Maamies, Sari (1990) Tulemaisillaan, lähtemäisillään ) suomen propinkvatiivin muoto- ja

merkityshistoriaa. (Unpublished Master’s Thesis, Dept. of Finnish, University of Helsinki.)

Malinowski, Bronislaw (1923) “The Problem of Meaning in Primitive Languages”. The

Meaning of Meaning. Pp. 296-336.

Martin, J.R. (1981) “How Many Speech Acts?”. University of East Anglia Papers in Lin-

guistics 14-15: 52-77.

Martin, James R. (1987) “The Meaning of Features in Systemic Linguistics”. In M.A.K.

Halliday & Robin P. Fawcett (Eds.), New Developments in Systemic Linguistics: Volume

1 Theory and Description. Frances Pinter: London. Pp. 14-40.

Martin, J.R. (1988) “Hypotactic Recursive Systems in English: Towards a Functional Inter-

pretation “. In James Benson & William Greaves (Eds.), Systemic Functional Approaches

to Discourse: Selected Papers from the 12th International Systemic Workshop. Ad-

vances in Discourse Processes: Volume XXVI. Ablex: Norwood, New Jersey. Pp. 240-270.

Martin, James R. (1991) “Nominalization in Science and Humanities: Distilling Knowledge

and Scaffolding Text”. In Eija Ventola (Ed.), Functional and Systemic Linguistics:

Approaches and Uses. Mouton de Gruyter: Berlin. Pp. 307-337.

Matthews, P.H. (1966) “The Concept of Rank in ‘Neo-Firthian’ Grammar”. Journal of

Linguistics 2: 101-110.

Matthews, P.H. (1981) Syntax. Cambridge University Press: Cambridge.

Matthiessen, Christian (1988a) “Representational Issues in Systemic Functional Grammar”.

In James Benson & William Greaves (Eds.), Systemic Functional Approaches to Dis-

course: Selected Papers from the 12th International Systemic Workshop. Ablex: Nor-

wood, New Jersey. Pp. 136-175.

Matthiessen, Christian (1988b) “Semantics for a Systemic Grammar: The Chooser and

Inquiry Framework”. In James D. Benson, Michael J. Cummings, & William S. Greaves

(Eds.) (1988) Linguistics in a Systemic Perspective. John Benjamins: Amsterdam. Pp.

221-242.

Matthiessen, Christian (1989) Lexicogrammatical Cartography: English Systems. (Dept. of

Linguistics, University of Sydney.)

Matthiessen, Christian & Halliday, M.A.K. (forthcoming) “Systemic-Functional Gram-

mar”. To appear in F. Peng & J. Ney (Eds.) Current Approaches to Syntax. Benjamins &

Whurr: London.

McGregor, William (1988) Mood and Subordination in Kuniyanti. In Peter Austin (Ed.),

Complex Sentence Constructions in Australian Languages. John Benjamins: Amsterdam.

Pp. 37-67.

Page 433: systematic functional linguistic

428

McGregor, William (1990a) A Functional Grammar of Gooniyandi. John Benjamins:

Amsterdam.

McGregor, William (1990b) “The Metafunctional Hypothesis and Syntagmatic Relations”.

Occasional Papers in Systemic Linguistics 4: 5-50.

McGregor, William (1991) “The Concept of Rank in Systemic Linguistics”. In Eija Ventola

(Ed.), Functional and Systemic Linguistics: Approaches and Uses. Mouton de Gruyter:

Berlin. Pp. 121-138.

McGregor (MS) Quoted Speech and Thought in Gooniyandi ) Perspectives from Functional

Grammar. Melbourne University, Melbourne.

McIntosh, Angus & Halliday, M.A.K. (1966) Patterns of Language: Papers in General,

Descriptive and Applied Linguistics. Longmans: London.

Milroy, Lesley (1987) Language and Social Networks. Second Edition. Basil Blackwell:

Oxford.

Mitchell, T.F. (1975) Principles of Firthian Linguistics. Longman: London.

Mühlhäusler, Peter (1987) “The Politics of Small Languages in Australia and the Pacific”.

Language & Communication 7: 1-24.

Nevalainen, Terttu (1990) “Some Aspects of Convergence in Finnish and English Prosody”.

In Kalevi Wiik & Ilkka Raimo (Eds.), Nordic Prosody V: Papers from a Symposium.

Phonetics, University of Turku: Turku. Pp. 236-255.

Newmeyer, Frederick J. (1980) Linguistic Theory in America. Academic Press: New

York.

Nikanne, Urpo (1990) Zones and Tiers: A Study of Thematic Structure. Studia Fennica 35.

Suomalaisen Kirjallisuuden Seura [Finnish Literature Society]: Helsinki.

Nuolijärvi, Pirkko (1991) “Monikielinen Suomi”. In Tapani Lehtinen & Susanna Shore

(Eds.), Kieli, valta, eriarvoisuus. Kieli 6. Helsingin yliopiston suomen kielen laitos [Dept.

of Finnish at the University of Helsinki]: Helsinki. Pp. 11-32.

Nykysuomen sanakirja (Dictionary of Contemporary Finnish) Volumes 1 ) 6. Suomalaisen

Kirjallisuuden Seura [Finnish Literature Society)/Werner Söderström: Helsinki. 1962 )

1964.

Pajunen, Anneli & Palomäki, Ulla (1982) Tilastotietoja suomen kielen muoto- ja

lauseopillisista yksiköistä. Lauseopin arkisto: Turku.

Palmer, F.R. (Ed.) (1968) Selected Papers of J.R. Firth 1952-59. Longmans: London.

Patten, Terry (1988) Systemic Text Generation and Problem Solving. Cambridge Univer-

sity Press: Cambridge.

Penttilä, Aarni (1963) Suomen kielioppi. Werner Söderström: Helsinki.

Page 434: systematic functional linguistic

429

Phillips, M.K. (1988) “Text, Terms and Meanings: Some Principles of Analysis”. In James

D. Benson, Michael J. Cummings, & William S. Greaves (Eds.), Linguistics in a Systemic

Perspective. Current Issues in Linguistic Theory, Volume 39. John Benjamins: Amsterdam.

Pp. 99-118.

Popper, Karl (1972) Objective Knowledge. Clarendon Press: Oxford.

Prince, E. F. (1981) “Towards a Taxonomy of Given ) New Information”. In P. Cole (Ed.)

Radical Pragmatics. Academic Press: New York.

Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, & Jan Svartvik (1985) A Com-

prehensive Grammar of English. Longman: London.

Räisänen, Alpo (1988) Suomen kielen u-johtimiset verbit. Suomi 141. Suomalaisen

Kirjallisuuden Seura [Finnish Literature Society]: Helsinki.

Ravelli, Louise (1988) “Grammatical Metaphor: An Initial Analysis”. In Erich H. Steiner &

Robert Veltman (Eds.), Pragmatics, Discourse and Text: Some Systemically-Inspired

Approaches. Pp. 133 ) 147.

Reddy, Michael J. (1979) “The Conduit Metaphor ) A Case of Frame Conflict in Our Lan-

guage about Language”. In Andrew Ortony (Ed.), Metaphor and Thought. Cambridge

University Press: Cambridge. Pp. 284-324.

Robins, R.H. (1961) John Rupert Firth (Obituary Article). Language 37: 191-200.

Robins, R.H. (1964) General Linguistics: An Introductory Survey. Longmans: London.

Robins, R.H. (1967) A Short History of Linguistics. Longman: London.

Robinson, Ian (1975) The New Grammarian’s Funeral: A Critique of Noam Chomsky’s

Linguistics. Cambridge University Press: Cambridge..

Ross, J.R. (1972) “The Category Squish: Endstation Hauptwort”. CLS No. 8. Chicago Lin-

guistics Society: Chicago.

Ross, J.R. (1973) “Nouniness”. In D. Fujimura (Ed.) Three Dimensions of Linguistics. TEC

Corp: Tokyo.

Ross, J.R. (1974) “Clausematiness”. In E.L.Keenan (Ed.) Semantics for Natural Languages.

Cambridge University Press: Cambridge.

Sadeniemi, Matti (1950) “Partisiippien ajanmerkityksestä”. Virittäjä 54: 358- 359.

Sadeniemi, Matti (1950) Kielenkäytön kysymyksiä. Tietolipas 18. Suomalaisen Kirjallisuu-

den Seura [Finnish Literature Society]: Helsinki.

Salmi, Tiina (1990) P-position ja adverbin yli merkitykset kognitiivisen kieliopin

näkökulmasta. Unpublished M.A. Thesis, Dept. of Finnish, University of Helsinki.

Sammallahti, Pekka (1991) “No Passing, no Xing: Traffic Regulations for Word Order”. In

Lea Laitinen, Pirkko Nuolijärvi & Mirja Saari (Eds.), Leikkauspiste: Kirjoituksia kielestä

ja ihmisestä. Suomalaisen Kirjallisuuden Seura [Finnish Literature Society]: Helsinki. Pp.

225-233.

Sampson, Geoffrey (1980) Schools of Linguistics. Stanford University Press: Stanford.

Page 435: systematic functional linguistic

430

Saussure de, F. (1983 [1916]) Course in General Linguistics. Translated by Roy Harris.

Duckworth: London.

Searle, John R. (1969) Speech Acts: An Essay in the Philosophy of Language. Cambridge

University Press: Cambridge.

Searle, J.R. (1979) Expression and Meaning. Cambridge University Press: Cambridge.

Setälä, E.N. (1891) Suomen kielen lauseoppi. 3rd Edition. K.E. Holm: Helsinki.

Shore, Susanna (1986) Onko suomessa passiivia. Suomi 133. Suomalaisen Kirjallisuuden

Seura [Finnish Literature Society]: Helsinki.

Shore, Susanna (1988) “On the So-Called Finnish Passive”. WORD 39: 151-176.

Shore, Susanna (1990) Review of Maria Vilkuna, Free Word Order in Finnish: Its Syntax

and Discourse Functions. Nordic Linguistic Journal 13: 89-101.

Shore, Susanna (1991a) “Kielitiede ja maailmankatsomus”. In Tapani Lehtinen & Susanna

Shore (Eds.), Kieli, valta ja eriarvoisuus. Helsingin yliopiston suomen kielen laitos: Hel-

sinki. Pp. 189-207.

Shore, Susanna (1991b) A Systemic-Functional Perspective on Mood. In Jussi Niemi (Ed.)

Papers from the Eighteenth Finnish Conference on Linguistics. Kielitieteellisiä

tutkimuksia/Studies in Language. Faculty of Arts, University of Joensuu: Joensuu. Pp. 235-

255.

Shore, Susanna (1991c) “Tarvitaanko kartta”. Virittäjä 95: 221-227.

Simon, H.F. (1953) “Two Substantival Complexes in Standard Chinese”. Bulletin of the

School of Oriental and African Studies, University of London XV: 327-355.

Siro, Paavo (1964) Suomen kielen lauseoppia. Sanoma: Helsinki.

Sorjonen, Marja-Leena & John Heritage (1991) “And-Prefacing as a feature of Question

Design”. In Lea Laitinen, Pirkko Nuolijärvi & Mirja Saari (Eds.), Leikkauspiste: Kirjoituk-

sia kielestä ja ihmisestä. Suomalaisen Kirjallisuuden Seura [Finnish Literature Society]:

Helsinki. Pp. 59-74.

Suomen murteiden sanakirja [Dictionary of Finnish Dialects] (1985) Volume 1.

Kotimaisten kielten tutkimuskeskuksen julkaisuja 36. Kotimaisten kielten

tutkimuskeskus/Research Centre for Domestic Languages: Helsinki.

Suomen murteiden sanakirja [Dictionary of Finnish Dialects] (1988) Volume 2.

Kotimaisten kielten tutkimuskeskuksen julkaisuja 36. Kotimaisten kielten tutkimuskes-

kus/Research Centre for Domestic Languages: Helsinki.

Tainio, Liisa, Eeva-Leena Seppänen, Liisa Raevaara, & Pia Purra (1991) “”Joku oli

sanonut jotain lapsista” ) Arkikertomukset heijastamassa ja rakentamassa naisten kokemuk-

sia”. In Lea Laitinen, Pirkko Nuolijärvi & Mirja Saari (Eds.), Leikkauspiste: Kirjoituksia

kielestä ja ihmisestä. Suomalaisen Kirjallisuuden Seura [Finnish Literature Society]: Hel-

sinki. Pp. 17-42.

Tarvainen, Kalevi (1977) Dependenssikielioppi. Gaudeamus: Helsinki.

Page 436: systematic functional linguistic

431

Tench, Paul (1991) The Roles of Intonation in English Discourse. Forum Linguisticum. P.

Lang: Frankfurt am Main.

Thibault, Paul J. (1987) “An Interview with Michael Halliday”. In Ross Steele & Terry

Threadgold (Eds.), Language Topics: Essays in Honour of Michael Halliday, Volume II.

John Benjamins: Amsterdam. Pp. 601-627.

Toivainen, Jorma (1980) Inflectional Affixes Used by Finnish-Speaking Children Aged 1

) 3 Years. Suomalaisen Kirjallisuuden Seura [Finnish Literature Society] : Helsinki.

Toivainen, Jorma (1986) “Easy for the Child, Hard for the Grammarian: Observations on the

Functions of the Partitive in Finnish”. In I.Kurcz, G.W.Shugar, & J.H.Danks (Eds.), Knowl-

edge and Language. Elsevier Science Publishers: Amsterdam. Pp. 431-461.

Traugott, Elizabeth Closs (1982) “From Propositional to Textual and Expressive Meanings:

Some Semantic-Pragmatic Aspects of Grammaticalization”. In Winfred P. Lehmann &

Yakov Malkiel (Eds.), Perspectives on Historical Linguistics. Amsterdam Series in the

Theory and History of Linguistic Science. Series IV ) Current Issues in Linguistic Theory,

Volume 24.. Pp. 245-271.

Trubetzkoy, N.S. (1969 [1935]) Principles of Phonology. Translated by Christiane A.M.

Baltaxe. University of California Press: Berkeley.

Vähämäki, K. Börge (1980) Spesieksen asema suomen kielen kuvauksessa. Fennistica 2.

Åbo Akademi: Åbo/Turku. [Summary: The Status of Determinedness in Finnish.]

Vähämäki, K. Börge (1984) Existence and Identity: A Study of the Semantics and Syntax

of Existential Sentences in Finnish. Meddelanden från Stiftelsens för Åbo Akademi

forskningsinstitut, Nr 99. Åbo Akademi: Åbo/Turku.

Vendler, Zeno (1967) Philosophy in Linguistics. Cornell University Press: Ithaca.

Ventola, Eija (1987) The Structure of Social Interaction. Pinter: London.

Ventola, Eija (1988) “Text Analysis in Operation”. In Robin P. Fawcett & David J. Young

(Eds.), New Developments in Systemic Linguistics: Volume 2 Theory and Application.

Pinter: London. Pp. 52-77.

Verschueren, Jef (1980) On Speech Act Verbs. Pragmatics & Beyond 4. John Benjamins:

Amsterdam.

Vilkuna, Maria (1980) Referenssi ja spesies suomenkielisten tekstien tulkinnassa. Unpub-

lished Licentiate Thesis. Dept. of Finnish, University of Helsinki.

Vilkuna, Maria (1984) Voiko kin-partikkelia ymmärtää? Virittäjä 88: 393-408. [Summary

(p. 408): Understanding Conventional Implicature: the clitic -kin/kAAn in Finnish].

Vilkuna, Maria (1989) Free Word Order in Finnish: Its Syntax and Discourse Functions.

Suomalaisen Kirjallisuuden Seura: Helsinki.

Vilppula, Matti (1984) “Kurkikin jo lähti”. Virittäjä 88: 47-60. [Summary (p. 59-60): On

the Meaning of the Clitic -kin.]

Page 437: systematic functional linguistic

432

Vološinov, V.N. (1973 [1930]) Marxism and the Philosophy of Language. Seminar Press:

New York.

Walsh, Michael (1987) “The Impersonal Verb Construction in Australian Languages”. In

Ross Steele & Terry Threadgold (Eds.), Language Topics: Essays in Honour of Michael

Halliday. Volume I. John Benjamins: Amsterdam. Pp. 425-438.

Wertsch, James D. (1983) “The Role of Semiosis in L.S.Vygotsky’s Theory of Human Cog-

nition”. In B. Bain (Ed.), The Sociogenesis of Language and Human Conduct. Plenum

Press: New York. Pp. 17-31.

Wertsch, James D. & Stone, C. Addison (1985) “The Concept of Internalization in Vygot-

sky’s Account of the Genesis of Higher Mental Functions”. In James V. Wertsch (Ed.),

Culture, Communication, and Cognition: Vygotskian Perspectives. Cambridge Univer-

sity Press: Cambridge. Pp. 162-179.

Wierzbicka, Anna (1980) The Case for Surface Case. Karoma: Ann Arbor.

Wierzbicka, Anna (1987) English Speech Act Verbs. Academic Press: London.

Wierzbicka, Anna (1988) The Semantics of Grammar. John Benjamins: Amsterdam.

Winograd, Terry (1972) Understanding Natural Language. Edinburgh University Press:

Edinburgh.

Wittgenstein, Ludwig (1953) Philosophical Investigations. Blackwell: Oxford.

Yule, George (1981) “New, Current and Displaced Entity Reference”. Lingua 55: 41-52.

Yli-Vakkuri, Valma (1969) “Miksi pojat joskus ovat iloiset?” In Alho Alhoniemi, Sirkka-

Liisa Hahmo, Osmo Nikkilä & Valma Yli-Vakkuri (Eds.) Juhlakirja Paavo Siron täyttäes-

sä 60 vuotta. Suomen kielen laitoksen julkaisuja 1. Acta Universitatis Tamperensis Ser. A

Vol. 26. Tampereen yliopisto: Tampere.

Page 438: systematic functional linguistic

435

Index:

aboutness 320-322

accusative 69, 82-84, 87-94, 96, 250, 260, 277,

291, 293, 294, 297

action processes 256 fn

active clause (in English) 66, 302-306

Actor 63, 64, 111, 113, 116, 206, 209, 211, 257

ff., 266, 288, 302-304, 331

actual (vs. potential) (see also potential) 23, 46,

115, 378

adjectival phrase 119, 121

adverb 133, 241, 263-266, 273, 295, 321-322,

343, 371

criticism of traditional notion 126, 127,

129, 130, 136, 377

adverbial of quantity case-marked like an object

261-262

adverbial phrase (AdvP) 133

Hakulinen & Karlsson’s adverbial phrase &

adverb phrase 115

AdvP: see adverbial phrase

agentless processes (cf. metaphor, experiential)

101, 265 ff., 301 ff.

agglutinative 74

agnateness 66, 100, 103, 217, 228, 229, 232,

285, 298, 371

Agricola 72, 73, 75

ambient process 214, 240-243, 256, 267, 364

analogy 67, 247, 379

ancillary (cf. constitutive) 29, 38, 187

arbitrary 40 ff., 45, 56 fn

association (vs. inclusion in circumstantial pro-

cesses) 253

assumptions (unexplicated assumptions about

language) 9 ff., 30, 376 ff.

Attribute (cf. genitive attribute) 5, 98, 156,

198, 214-219, 231, 232, 234-236, 238-242,

256, 371

attribution 214, 226

attributive intensive process: see intensive

Austin 32, 36, 168

automative (automative-passive) 302 ff.

autonomous linguistics 27, 28

autosegmental phonology 2, 3

Bakhtin 15, 33, 35, 170, 183, 374

Baltic-Finnic 70

baseline (in grammatical description) (see also

metaphor) 55, 108, 178, 284, 315

basic unit (vs. complex unit) 46-48, 105-106

behaviour potential 22, 23

behavioural process 265, 272-273 290

Bible, Finnish 149, 218, 352, 353

English Bible 215

bound morpheme 77, 132, 212, 217, 224, 289,

290, 301, 346-348, 351, 358

boundedness 69, 82, 88, 89, 93, 94-99, 103,

106, 216, 227, 250, 260, 273, 277, 293, 300,

301

Brown and Yule’s approach to Given and New

55, 308 ff.

Carrier 5, 156, 215-219, 232, 234, 236, 238,

239, 241, 288, 371

case, case-marking (Finnish case forms) 82 ff.,

260-263, 270-272, 276-277, 296-330

case-role analysis (Fillmore) 206

cataphoric demonstrative 155

category 18, 43, 46, 48-50, 60, 63-65, 209

causation 301, 304

causative affix 268, 269, 303, 304

censorious passive 207

central notions in SF theory Ch.2 (42 ff.)

chaos theory 276

characterizing demonstrative 155

check interrogative 194, 196

child language 184

metaphor & child language 181

Circumstance 115-116, 120, 133, 135, 141,

193, 240-248, 253, 256, 327, 329, 338, 360,

363-365

circumstantial process 214, 244-256

circumstantial role 211

class (social class) 17-18, 22, 24, 27, 73

class (word class, class of units) 48, 114, 127,

131

Classifier 114-116

clause 2-5, 46-47, 105-108, 137, 205

clause complex: see separate entry

clause vs. sentence (orthographical unit) 46,

106-107

clause vs. tone group (phonological unit)

315

elliptical (vs. full) clause 107, 108

major clause (clause) 108, 174

minimal clause 211

minor (vs. major) clause 108, 174, 180

clause as Theme 343 ff.

clause complex 46-48, 137-166, 280, 308,

341-346, 369

function of clause complex 137, 146-148

clause-initial interpersonal element 339-340

clause-initial textual element 339-340

Page 439: systematic functional linguistic

436

cohesion 7, 40, 57, 143, 144, 159, 308,

327-329, 331-333, 354, 359, 373

cohesive ties 373-374

collocation 374

command 170, 172-190, 204, 346 ff.

communicative competence 26

comparative affix 119, 130

competence and performance 17

complement adverbial (traditional Finnish

grammar) 233

complement (predikatiivi, traditional Finnish

grammar) 97-99, 150, 238, 241, 298

complement in English 198

complexes (basic vs. complex units) (see also

clause complex) 46-48, 105-106, 369

computer applications 43

computer corpora 8, 66, 101, 379

conflation 2, 49, 56, 198, 267, 315, 320, 325,

362

congruent, congruency: see metaphor

conjunction 48, 139, 141-143, 148, 155,

160-162, 195

conjunction vs. conjunctive (discourse ad-

junct) 141-143

conjunctive adjunct 141, 332

consciousness (see also: human consciousness,

mental process) 33, 34

consonant gradation 75, 76

constituency 3, 4, 47-50, 92, Ch.4 (esp.

105-112)

constituency & dependency 92, 109, 111

clause as constituent of another clause (see

also embedding, rank-shifting) 153

discontinuous constituent 111, 216

ordering of constituents (see word order)

ranked constituency 47-50, 105 ff.

constitutive (vs. ancillary) 29, 38, 177

content interrogative 193, 221

context of situation 14, 19-21, 34-39, 42, 55,

59, 169, 175, 176, 377

contextual semantics 46

contextual meaning (Firth) 14

continuative 208, 373

continuing clause (vs. initiating) 139, 140, 143,

145, 166, 369

contrastive New: see New

co-ordination: see parataxis

corpus 7, 8, 27-29, 101, 102, 283, 379

corpus vs. knowledge 27-29, 101, 379

co-text 7, 129, 131, 169, 242, 243, 250, 251,

365, 377

creation of meaning: see semogenesis

data (see also: corpus, computer corpora) 6-9,

12, 15, 43, 54, 66, 170, 226, 376-379

data on which this study is based 6-7, 29

daughter-dependency 92

declarative 50, 52, 78, 81, 108, 109, 135, 167,

169-196, 197, 243, 313, 323, 334, 369

decontextualized view of language 8, 29, 35,

365, 376, 377

default Theme (see also: Theme) 337-338

defining (identifying) process: see intensive

defining relative: see relative

definition (vs. recognition) criteria 320

deixis 82, 98, 114, 115, 121, 128, 135, 136,

164, 165, 200, 218, 315, 317, 377

deictic centre 135, 136, 377

deictic centre that is never explicated 135

delicacy 3, 6, 43, 53, 54-55, 59, 60, 61, 63,

119, 167, 196, 210, 244, 245, 265, 274, 284,

288, 306, 376

deontic modality 167, 189

dependency (see also: taxis) 91, 92, 105, 106,

109-111, 116, 120, 125, 139, 144, 165-166,

211

dependency & constituency 92, 109, 111

dependency vs. embedding 149

dependent clause (vs. dominant) 139-146, 148,

149, 164, 167, 344-346

derivation 74, 112, 265, 301-306

designed (vs. evolved) systems 64

dialectic between system and text 21, 207

dialectic between theory and description 12

dialectic (vs. causal) relation 55

direct representation (vs. projection) 159, 162,

278

direct speech: see projection

discontinuous constituent 111, 216

discourse formations (Lemke) 374

Domain 42, 155, 213, 242, 243, 288, 290, 291,

296-298, 301, 306, 363, 364

dominant clause (vs. dependent) 139-146, 148,

149, 164, 167, 344-346

dummy element 363

dynamic (vs. synoptic) 21, 67-68, 146, 148,

157, 287-288

E. Itkonen 8, 26, 33, 67, 212

educational linguistics 375

elaboration (in expansion) 141

ellipsis (cf. non-realization) 79, 81, 107, 108,

112, 117, 121, 144, 174, 243, 308-310, 345,

346, 349, 353, 373

elliptical clause 107, 108

Page 440: systematic functional linguistic

437

embedding (rankshifting) (see also clause, em-

bedded) 47-48, 91, 106, 120, 139, 146, 149,

150, 153, 155-159, 162, 166, 170, 218, 223,

224, 279-281, 285, 287, 326, 338, 339, 341-344,

369, 371, 372

enhancement (in expansion) 141

Entity (function in an NP) 114-118, 121-122,

127-129

Entity vs. Head in an NP 121-122, 127-129

Entities as topical Theme 320

entry condition 113

epistemic modality 167, 189

Epithet (function in an NP) 114, 115, 121

Estonian 69

ethnomethodological conversation analysis 9,

10, 176, 182

evolved (vs. designed) systems 64

evolution of SF theory 42

exchange (clause as exchange) 172-190

existential clause 96, 102, 103, 248-249, 250,

291-301, 324, 362, 363

existential subject (in traditional grammar)

102, 103, 291-301, 324

expansion (in complexes) 47, 48, 141, 166, 342

Experiencer 267-272, 290-291, 347

experiencer process 257, 267-272, 290, 303,

304, 358

experiential meaning 37, 49-51, 57, 74, 210,

284, 329, 362

experiential meaning & topical Theme 320

ff.

experiential structure 50-51, 68, 108, 111, 113,

205 ff.

experiential Theme: see Theme

exponence (see also realization) 43, 54

expression (vs. realization) 55, 293 fn

extended verb phrase 106

extension (in expansion) 141

Extent (spatial or temporal Extent) 133,

260-263, 266

external causation 301-305

face-to-face conversation ~ interaction 176-

177, 183

face-to-face language-in-action 177

Facet (in English NP) 128

cf. Orientation 128-131

fact (projected clause) 161-162, 281

field (variable in context of situation) 37, 38,

59-60, 374

first-order field 38

Fillmore 176, 206

finite verb forms in Finnish 77-81

Finite (as an interactional function) 197 ff.

Finnish dialects 71, 72

Finnish Sign Language 70

Finno-Ugric 69, 70, 89

first-order field 38

first-order reality 307

Firth 2, 8, 12-21, 23, 43, 46, 51, 60, 63, 65, 68,

122, 251, 300, 375, 376, 379

flux 68

focus (see: New, contrastive)

focus particles 127, 136, 316, 317

form (vs. substance) 45-46, 54

form (importance of forms and structures) 63,

206-208

form (grammar organized as meanings rather

than forms) 56-57

forward channel 195, 352

free indirect speech 160, 163

frequency (see: statistical analysis)

function of a clause complex 137, 146-148

functional component (cf. metafunction) 56

fuzziness 50, 142, 166, 190, 230, 253, 262, 265

general circumstantial process 244-249, 253

general linguistic theory 18

generic 35

genitive 82-84, 87-94, 96, 101, 103, 104, 114,

118, 123-128, 130, 134-136, 255, 260,

269-272, 276, 277, 291, 293, 298, 362, 377

genitive attribute (traditional grammar) 125

genitive-like accusative 88-90, 92, 93

genre 7, 29, 35, 325, 364, 374, 376

Given (see also: New) 5, 51, 55, 63, 220-225,

292-301, 307-318, 332, 333, 335, 357,

360-362, 365, 368

typically Given 315

Given-New (information structure) 5, 51, 55,

220-221, 292-296, 300, 307-318, 332-333,

334-336, 357-339, 360, 365

glosses 9, 93, 212, 321, 409, 410

Goal 63, 113, 116, 206, 209, 211, 247,

257-265, 266, 272-273, 276-278, 290, 291,

296, 331, 332

goods-&-services 172, 173, 177, 178, 180, 183,

185-190

grammatical case 84, 103

grammatical category 18, 64, 209

grammatical metaphor: see metaphor

grammatical proportionality 66-67

Page 441: systematic functional linguistic

438

grammatical subject

traditional definition 99-104, 292-301 (tra-

ditional notion of “existential subject”

reinterpreted as Domain)

as defined in this study 78, 99-100, 192,

201-204, 217, 241, 268, 271, 289 fn,

290, 308, 348-353

grammatical subject in English 185-187,

346

graphology 44, 46

Greek letters 116, 139

group 46-48, 50, 54, 82, 105-106, 119, 153,

285

habitive clause 360-362

Hakulinen & Karlsson (1979) 105, 115-116,

121, 125-126, 133, 170, 176, 205-206, 215,

227, 233, 235, 237-239, 241-243, 253, 257,

261, 263, 269-272, 276, 292, 323

Head 54, 82, 91-93, 110, 114-128, 133, 216,

233, 236, 253, 281, 321, 326, 338, 339, 362

Head in a Finnish PP 125

hedge 182, 183

heteroglossia 7, 170, 183, 380

historical linguistics 89-90, 93 fn, 123

historical overview of SF theory 43

Hjelmslev 14, 21, 44

Hockett 64

host clause 149-156, 162, 166, 279, 338,

341-343, 369

human consciousness (in mental processes)

161-162, 257, 274, 278-279, 306, 371

Hungarian 70

Hymes 26, 59

hypostatization (see also: reification) 18

hypotaxis (see also: taxis) 90, 91, 116, 123,

139-146, 149-151, 166-167, 344-345

idea (projected clause) 161-162, 164-167

ideational meaning 36-37

Identified 150, 214, 215, 220, 221, 223-227,

234, 251, 288, 353

Identifier 214, 220, 221, 224-227, 234

identifying relational process: see intensive

ideology 1, 375

illocutionary force 197

imperative 52, 79-81, 89, 167, 169-196, 198,

204

impersonal verb 101

inclination (modality) 4, 189, 271

inclusion (vs. association in circumstantial pro-

cesses) 253

inclusive processes 237-239, 256

indefinite form (“Finnish passive”) 77-81, 87,

89, 102, 112-113, 208, 304-306, 364, 371

indefinite article in English 309-310

indirect speech (see projection) 159-163, 278

individual (as defined by Firth) 16-17

ineffability 18-19, 63-64, 311

infinitives in Finnish 77, 85-87, 132, 167

number of infinitive & participial forms in

Finnish 132

1st infinitive 85-86

2nd infinitive 85-86

3rd infinitive 85-86, 246, 247, 264

information structure: see Given-New

inherent participant 88, 112, 132, 133, 202-203,

210-211

inherent role 211, 216, 218, 218, 220, 233-235,

240, 244, 250, 257, 269, 274, 276, 284, 302,

308, 346-348, 371, 376-378

initiating clause (in paratactic complex) 139-

140, 142, 143, 145, 166, 312

Instigator 268-269, 304

integrated lexicogrammar 60-63, 112, 206, 265

intensive relational process 54, 97, 210,

214-236, 237, 240-241, 251, 256, 258, 264,

288, 305, 371

attributive 156, 214-218, 223, 226, 229,

231-236, 239, 283, 371

identifying 54, 214-215, 219-230, 232,

234, 251, 289, 371

defining, naming, exemplifying 223,

227

other intensive processes 231, 233, 234, 236

resultative 203, 231-235, 264, 305

sensory 236

temporary 231, 233, 234

interactional function 186, 197-201

interactional structure in the clause Ch.5 (167

ff.)

interactive event (clause as interactive event)

4, 172-173, 186, 188, 198, 199, 202

interdependency (see also: taxis) 37, 138, 139,

161, 166

interface (semantics as an interface) 37, 59,

173, 175

internal human consciousness (see: human con-

sciousness)

inter-organism perspective 25-26

interpersonal meaning 36, 51, 57, 167, 200,

307

interpersonal theme: see clause-initial interper-

sonal element

interpretative source 129, 243, 365

interrogative 167-170, 172-174, 176, 187

English 52, 174, 313, 346

Page 442: systematic functional linguistic

439

Finnish 81, 135, 145-146, 154, 170-171,

181, 187, 192-198, 221, 313, 334, 343,

346

interrogative and Key 313

content interrogative 193-194

content check interrogative 194-195

polar interrogative 81, 160, 192, 313, 334

intertextuality 11, 375, 380

intonation (see also: Key, Given-New) 51,

147, 148, 169, 171, 194, 313, 318

intra-organism perspective 25-26

intratextuality 208, 375

intuition (see also: knowledge) 8, 26, 28, 29,

376, 377

inverted ergative system 297

Jespersen 292

Keller 286

kernel sentences 205

Key 30, 156, 191, 289, 290, 301, 334, 372

kin-particle in Finnish 74, 96, 151, 296, 301,

315-318, 324-325, 328, 332, 333, 357

kirjakieli: see standardized written

knowable (definite article) 311

knowability of reality 31

knowledge of language (cf. intuition) 6-8, 11,

13, 25-29, 31, 34, 67, 144, 212, 259, 319,

319, 372, 378, 379

knowledge (its accumulation & grammatical

metaphor) 287-288

knowledge, shared 309-311

kokijalause 269

ko/kö-interrogative 194-196

Kuhn 11, 28, 30

labels, labelling (of grammatical categories)

64, 66, 114, 128, 185, 209-210, 222, 234,

239, 249, 251, 274, 282, 298, 299

language and mind 33-34

language-in-action 177, 184

language myth 28

language-specificness (of categories) 18, 209

langue 17, 20, 23, 24

Lappish: see Saame

Latin-based categories 100, 139, 161, 262

lausepaino (see stress) 314

levels in language (planes, strata) 45-46

lexical metaphor: see metaphor

lexicogrammar 44-46, 60, 112, 181, 265, 315

integrated lexicogrammar 60-63, 112, 206,

265

lexicogrammar & metafunctions 56, 59

lexicogrammar & the tone group 315

lexicogrammatical resources 7, 21

theory-based & metafunctionally organized

205

lexis as most delicate grammar 60-63, 112,

206, 265

linguistic behaviour potential 22

location vs. movement in Finnish PPs 124, 130

locative case 82, 84, 240, 244, 247, 276, 278

locative cases & PPs 123-124, 130

locution (projected clause) 160, 161-162, 164,

166, 167

logical approach to language 32, 100-101, 210,

215, 223, 227

logical relations construed by language (see

also logical meaning) 116, 197-198

logical meaning 36-37, 48-49, 50, 58

logical operator 197

London School 13

macro-role 201, 290, 298

macrophenomenon 279-283

major clause (see also clause) 113

major process types 213

Malinowski 19, 36

manifestation clause 346, 360, 363

mapping view (vs. reality construction) 30, 31

marked vs. unmarked 52, 79, 215, 315-318,

325, 332, 334-336, 349-358

marked information focus 316-318, 325, 332,

334-336, 349-358

material process 52, 54, 61, 111, 213, 222, 233,

241, 246, 253, 256-273, 275-278, 287, 288,

290, 294, 296, 297, 306, 314, 316, 371

matrix clause (see also: clause) 91, 282

meaning (see also: semantics, semogenesis,

metafunctions) 7, 19-21, 34-39, 44-46,

50-51, 56-59, 60-62, 372-375

act of meaning 2

clause meaning:

as exchange 167

as model of reality 205-207

textual meaning 307

clause complex, meaning of 137, 146-147

lexical meaning 60-63, 206, 265

meaning as a mental process 19, 34

meaning as a mode of behaviour 13, 19

meaning potential 23

meaning & grammatical labels 64

Measure 121, 122

Medium 104, 201, 204, 213, 288-291, 298,

301, 304, 306, 348, 352

medium (spoken vs. written) 38, 68, 79, 352

mental process 50, 101, 162, 165, 260, 263,

264, 269, 274-283, 289, 290, 369, 371

Page 443: systematic functional linguistic

440

mental process, meaning as a mental process

19, 34

metafunction 56-58, 60, 242, 307, 365

metaphenomenon 166, 279, 280

metaphor 55, 67, 179-183, 284-288, 300, 306

congruent, congruent realization 55, 56,

177-183, 187, 284, 285, 287

congruent and metaphorical at the same

time 55-56, 180-181

grammatical metaphor 55-56, 179-183,

284-288, 306

interpersonal (interactional) 179-183

experiential 284-288

lexical metaphor 179, 284, 287

meteorological process 241, 242, 257,

265-267, 290, 303, 364

minimal clause 211

minor clause 108, 180

mnemonic device 64, 209-210

modality 4, 32, 36, 58, 132, 167, 182, 189, 191,

199, 200, 223, 271

semantic space between positive and nega-

tive 189 fn, 199

modalization (probaility, usuality) 167, 189

modulation (obligation, inclination) 4, 167,

189, 191, 199

mode (variable in context of situation) 37-38,

59-60

mode of behaviour 19, 34

mode of meaning 20, 56

Modifier 54, 117-119, 121-123, 125-130, 134,

136, 247, 255, 261, 262

mood 4, 36, 52, 61, 145, Ch.5 (167 ff.)

mood marker 197, 198

mood option 179

morphology 44, 73, 74, 77, 112, 305, 347, 358,

378

morphophonological variation 78

multi-structural and polysystemic 14, 15

multivariate 51, 68

negative (auxiliary) verb 80-81, 132, 201

network 35, 52, 61, 62, 167, 174, 196

New (see also: Given, Given-New) 5, 51, 55,

220-225, 292-301, 307-318, 323, 325,

331-333, 336, 349-357, 360, 362, 363, 365,

373, 379

contrastive New (focus, contrastive focus)

312, 315-318, 323, 325, 331-333, 336,

349-357

nominal phrase (NP, nominal group) 46, 82,

106-107, 114-122

nominalization 287, 288

nominative-like accusative 88

non-finite 85, 86, 115, 119, 120, 150, 160, 183,

190, 278, 279, 281-283, 287, 369

non-inherent role 111, 211, 212, 216, 284

non-productive 85

non-realization (vs. ellipsis) 121 fn, 308,

345-348, 365

norm 75, 177, 181

not previously mentioned 312

NP: see nominal phrase

Numerative 121, 122, 128

object 87-88, 94, 271, 278, 288-300

objectivist approaches 31, 310

obligation (modality) 4, 36, 167, 189, 271

offer 173, 175, 177, 183, 185-190, 195, 204,

346

Ogden and Richards 19, 39

option 22, 52, 54, 61, 108, 152, 178, 179, 188,

196, 201, 301, 306

Orientation (in p-positional phrase) 128, 131,

134, 135

orientation to action 80, 184-190, 204, 369

OSMA (objektinsijainen määrän adverbiaali)

261-262

paradigm 11, 35, 80, 176

paradigmatic relations 51-53

parataxis (co-ordination) 48, 139-146, 149,

155, 166, 167, 369

parole 23

part-to-part relationship 109

part-to-whole relationship 109

passive: see indefinite or automative

English passive 63, 66, 301-302

personality 16, 17

Phenomenon 274-281, 290

types of Phenomena 279-280

phonological prominence 309, 310, 312

phonology 2, 3, 14, 20, 45, 46, 54

phrase 46-48, 54, 57, 82, 106-108, 113-115,

118, 119, 121-123, 125, 131, 133, 150, 153,

155, 309, 326, 369

planes in language (levels, strata) 44-46

polar interrogative 81, 160, 192, 313, 334

polarity (negative vs. positive) 32, 89, 95-97,

100, 102-104, 132, 192, 199-210, 233, 250,

261, 285, 294

Polish dative 361

polyphonic 3, 122, 325, 368, 375

polysystemic & multistructural 14, 15

Popper 28

Positioned 244, 245, 249, 253-255, 286

Possessed 249, 250, 252, 294

possessive deictic 98

Page 444: systematic functional linguistic

441

possessive process 214, 244, 249-256, 269,

272, 294, 358

possessive suffix 85, 90, 98, 102, 114, 123,

125, 203, 331

Possessor 64, 249-252

postmodification 118

postposition: see p-position

potential 3, 22-24, 43, 47, 48, 53, 79, 113, 115,

146, 174, 191, 198, 284, 375

PP: see p-positional phrase

p-position (pre- or postposition) 123, 125-129,

130, 135-136, 322, 377

p-positional phrase (pre- or postpositional

phrase, PP) 123, 125-129, 130, 135-136,

322, 377

pragmatics 39, 376, 378

pre- and postpositional phrase (PP) 76, 105,

106, 114, 115, 122-125, 125, 133, 156, 173,

212, 217, 244, 260, 261, 290, 377

preposition: see p-position

previously mentioned 312, 315

primary clause (vs. secondary) 139

principle of grammaticalization 63

prism 20, 122, 300, 376

probability 4, 36, 167, 189, 271

Process 210-211

process type 49, 65, 213

projection 138, 139, 146, 155, 159-166, 189,

258, 275, 278-283, 285, 338

direct & indirect speech 159, 160, 161, 163

direct vs. indirect as meaningful distinction

164

dramatic flavour of quoting (direct speech)

164

fact 161

free indirect 162

human consciousness 161-162, 257, 274,

278-279, 306, 371

idea 161

locution 161

projection in mental processes 278-283

quote 162

report 162

person (pronouns, personal endings) 78-81

1. and 2. person 78-81, 113-114, 201 ff,

212, 290, 358-359

proposal 186, 189, 199, 204, 347, 371

proposition 100, 190, 199, 204

prosody 10, 14, 51, 149, 197, 311, 313, 314,

317, 368, 372

prototype 65, 66, 166, 276

prototypical subject 101

Proto-Uralic 89, 90

purpose of this study 1-3

question 168-189

Range 272-273, 284, 285, 289, 290

rank 43, 44, 46-49, 54, 58, 105-110, 112-114,

137, 139, 149, 151, 153, 155, 179, 209, 210,

279, 284, 306, 344, 365, 367, 372, 376

rank hierarchy hypothesis 47, 49, 105, 108-110,

112-112, 137

rank scale 46-48, 106-109, 113, 137

rank-shifting (see also: embedding) 106, 109

ranked constituency 49, 105, 108

reality construction:

Grace: 30, 31, 39

Halliday: 33, 64, 375

realization 43-45, 49, 54-55, 121 fn, 175,

177-179, 181, 284, 293 fn, 311

received view in traditional grammar 69

accusative 87-94

direct & indirect speech 159, 160-161

existential subject 102-103, 292 fn

genitive subject 103-104

subject 99-104

received view of language (cf. assumptions) 8,

11, 28-29

received view in historical linguistics 88-89, 93

fn

recognition (vs. definition) criteria 320

recontextualization 42, 49, 378

referatiivinen lauseenvastike (see also: projec-

tion) 282

reference 39-42

referent 32, 39, 40, 215

reification 18, 20, 322

relational process (see also: intensive, general

circumstantial, possessive, ambient, inclu-

sive) 5, 214-256, 258, 263, 264, 269, 283,

297, 306, 371

relative clause, relativization 138-139,

149-159, 341-343

relevance of defining vs. non-defining to

Finnish 149-153, 155

intonational differences & punctuation in

English 152, 152 fn

renewal of connection 12, 21, 379

repeatedly meant 374, 375

repetition (as cohesive tie) 354, 375

report: see projection

reported speech: see projection

reporting clause equivalent (see also: projec-

tion) 281

Research Centre for Domestic Languages 73

residue 197, 198

restricted language 15, 16

Page 445: systematic functional linguistic

442

restrictive relative clause: see relative clause

resultative (aspect) 94

resultative intensive process: see intensive

rhetorical function 167-170, 173, 175-176, 191

rich and poor (analogy of rich vs. poor) 167

role 1, 4, 28, 32, 33, 37, 210-212

Romany 70

Russia 70, 89

Saame 70, 72, 110

salient

contextually salient 309

experientially salient 122, 128

phonological salience: see tonic promi-

nence, phonological prominence, stress

Sayer 274, 280, 289

scale, scale of abstraction 43

scale and category grammar 13, 43, 46, 52, 60

scope of this study 1-4, 367

Searle 35, 168, 197

second-order field 38

second-order reality 307

secondary clause (vs. primary) 139

self-engendering 101, 301-304, 306

semantic continuum 60, 90, 112

semantic variant 179-180

semantics (see also: meaning, semogenesis,

text semantics) 19, 34-39, 44-46, 56-60,

175, 372-375

semi-transitive verb 276

semogenesis (meaning-making, creation of

meaning) 16, 23, 31, 32, 38, 53, 67, 181,

182, 207, 208, 284, 378

semiosis 208

semiotic potential (see also: potential, mean-

ing) 23, 284, 375

Senser 50, 274, 281, 282, 289

sensory process: see intensive

sentence (orthographical unit cf. clause) 46,

106-107

SF theory 2-4, Ch.2 (13 ff.)

shared information 312

signifiant 40, 293

simple dichotomy 79, 352

simultaneous paradigmatic options 52

singulary branching 47, 106, 109, 113

situational meaning 14

spatial Extent 260-263

speech act theory 168-169 fn, 182

speech act pronoun (see also person, 1. and 2.)

347, 352-353, 358, 361

speech fellowship 16, 17

speech function (cf. rhetorical function) 174,

176, 177, 179, 183, 186, 189, 346

standardized written Finnish 72-73, 78-81

statistics, statistical frequency 60, 101-102, 226

stratum (see also plane) 14, 44, 54

stress 313-314

structuralism 20

structure

as non-random organization 3, 49-51, 276

Firth 14-15

in scale and category grammar 43

in SF grammar 3, 49-51, 52, 276

multivariate structure 50-51

univariate structure 50-51

subject, English 185-187

subject, Finnish:

dispersion/deconstruction of “exis-

tential subject” 292

formally defined subject realizes Medium

100, 201-204, 288-291

traditionally defined in Finnish 99-104

subjectless clause 101, 102, 201, 204, 242, 290,

324

submodification 92, 119

subordination (see also: hypotaxis) 48, 90,

117, 139, 149, 151, 166

subscript (to indicate cross-referencing) 212

subsidiary topical Theme: see Theme

substance (form vs. substance) 45-46, 54

substitution (cohesive tie) 326, 373

Swedish 70, 72, 74, 75, 107, 192, 292

symmetry 187, 188, 229, 230

synoptic (vs. dynamic) 21, 67, 68, 109-111,

148, 149, 287

syntagm 108, 278, 378, 379

syntagmatic relation 14, 20, 49, 51, 113

system (i.e. systematicity, overall organization,

cf. langue)

as semogenic resource: see semogenesis

designed vs. evolved 63-64

Firth 17

Halliday 23-24

Hjelmslev’s system & process 21

system and text 21, 378

system (i.e. paradigmatic options in a particular

context or environment)

Firth 14-15

scale and category grammar 43

SF grammar 49-53

system and structure 14, 21, 52, 53

Firth 14-15

scale and category grammar 43

SF grammar 49-53

taxis (interdependency, see also: hypotaxis,

parataxis) 149-166

Page 446: systematic functional linguistic

443

taxonomy of information status (Prince) 309

temporary intensive process: see intensive

tenor (variable in context of situation) 37, 38,

59, 60

textual meaning 37, 51, 144, 218, Ch.7 (307

ff.)

textual structure 57, Ch.7 (307 ff.)

thematic formation (Lemke) 374

Theme 308, 319-345

clause as Theme 338-339, 343-344

default Theme 337

entities as topical Theme 320 ff.

experiential Theme: see topical Theme

predicated Theme 336

subsidiary topical Theme 327-333, 343,

354, 355, 359

subsidiary topical Theme in relative clause

343

topic: see topical Theme

topical Theme 319-326

topical Theme & aboutness 320-322, 336-

339

topical Theme: basic definition (cf. follow-

ing entries) 322-323

topical Theme & 1. & 2. person 348-359

(esp. 357-358)

topical Theme in post-verbal position

334-336

clause as Theme 338-339, 343-344

theme in clause complex 341-345, 371

Theme-Rheme 319-345

topic-worthiness 322, 336, 337

Theme-Rheme 319

theory (per se) 9-12, 375, 378-380

thought dependent on semiotic systems 31

Token 26, 225, 226, 229, 251, 255, 380

tone group 5, 191, 218, 300, 307, 311-315, 317,

318, 365, 368, 372

tonic prominence 311, 313-318, 350, 357

transitive-intransitive dichotomy 256

tree diagram 47, 110, 111, 137

triplanar organization of language 44

truth (falsity, morality) 31, 32, 36, 200-201

truth-conditional semantics 31, 32, 56, 189,

200, 201, 215

tulosrakenne 231

types of structure 27, 36, 37, 49, 55, 59, 61, 84,

88, 105, 106, 108, 115, 119, 120, 141, 161,

167, 210, 211, 244, 266, 279, 373, 376

typicality 24, 53, 59, 124, 177, 178

unit 4, 10, 46-48, 51, 56, 68, 91-93, 105-107,

110, 125, 137, 180, 182, 195, 307, 313, 314,

315

univariate 50, 51, 68

unknowability of reality 31

unmarked 52, 104, 191, 242, 243, 284, 303,

311, 313-316, 318, 325, 362

unmodified p-position 135

unspecified human participant (see also: indefi-

nite) 112, 304

Uralic 69, 70, 89, 90

usuality (modality) 4, 189, 271

vai-interrogative: see interrogative, content

check 195

Value 14, 20, 41, 212, 225, 226, 229, 325, 340,

356

variables:

in interactive event 173

in context of situation 37-38

variation (dialectic & diatypic) 18, 21, 23, 24,

27, 34, 79, 166, 192, 193, 349, 353

morphonological variation 78

verb phrase (VP) 106, 112-113, 131-132

extended verb phrase 106

Verbiage 274, 278, 281, 282, 290

virke 10, 106

vowel harmony 76, 77

VP: see verb phrase

Vygotsky 33, 34

Watch this! function 312, 314

Whorf 28, 50, 63-64

word order 74, 110, 192, 215-216, 218-219,

232, 307, 315, 318, 325, 327, 332, 334, 335,

350

word order as indicator of Key 334-335

word order & unmarked information struc-

ture 315

word order & marked information structure

318, 332, 334, 350

ordering of clauses 149, 344-345