Systemic Functional Language Description: Making Meaning Matter [1 ed.] 9780815395089, 9781351184533, 9781032176956

Citation preview

Systemic Functional Language Description

This volume showcases previously unpublished research on theoretical, descriptive and methodological innovations for understanding language patterns grounded in a Systemic Functional perspective. Featuring contributions from an international range of scholars, the book demonstrates how advances in SFL have developed to reflect the breadth of variation in language and how descriptive methodologies for language have evolved in turn. Using examples from a wide variety of languages, the first part of the book looks at new methods for understanding grammatical variation in SFL. The second part of the book highlights a top-down approach by extending SFL methods to understanding grammar in context and examining text patterns in discourse semantics, register and genre, with a particular focus on languages and semiotic resources that have been understudied in SFL. Taken together, the volume offers a comprehensive account of Systemic Functional Language description, providing a foundation for practice and further research for students and scholars in descriptive linguistics, Systemic Functional Linguistics and theoretical linguistics. J.R. Martin is Professor in the Department of Linguistics at the University of Sydney, Australia, where he is also Deputy Director of the LCT Centre for Knowledge-Building. Y.J. Doran is Lecturer in the Department of Linguistics and Research Fellow at the LCT Centre for Knowledge-Building at the University of Sydney, Australia. Giacomo Figueredo is a Researcher and Senior Adjunct Professor of Linguistics at the Federal University of Ouro Preto, Brazil.

Routledge Studies in Linguistics

Metonymy and Language A New Theory of Linguistic Processing Charles Denroche A Forensic Linguistic Approach to Legal Disclosures ERISA Cash Balance Conversion Cases and the Contextual Dynamics of Deception James F. Stratman Conceptual Conflicts in Metaphors and Figurative Language Michele Prandi The Language of Pop Culture Edited by Valentin Werner Perspectives from Systemic Functional Linguistics Edited by Akila Sellami-Baklouti and Lise Fontaine Time Series Analysis of Discourse Method and Case Studies Dennis Tay Heart- and Soul-Like Constructs across Languages, Cultures, and Epochs Edited by Bert Peeters Systemic Functional Political Discourse Analysis A Text-based Study Eden Sum-hung Li, Percy Luen-tim Lui and Andy Ka-chun Fung Systemic Functional Language Description: Making Meaning Matter Edited by J.R. Martin, Y.J. Doran and Giacomo Figueredo For more information about this series, please visit: www.routledge.com/ Routledge-Studies-in-Linguistics/book-series/SE0719

Systemic Functional Language Description Making Meaning Matter Edited by J.R. Martin The University of Sydney, Australia Shanghai Jiao Tong University, China Pontificia Universidad Católica de Chile, Chile

Y.J. Doran The University of Sydney, Australia

Giacomo Figueredo Federal University of Ouro Preto, Brazil

First published 2020 by Routledge 52 Vanderbilt Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2020 Taylor & Francis The right of J.R. Martin, Y.J. Doran and Giacomo Figueredo to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data A catalog record for this book has been requested ISBN: 978-0-8153-9508-9 (hbk) ISBN: 978-1-351-18453-3 (ebk) Typeset in Sabon by Apex CoVantage, LLC

Contents

List of Illustrations List of Contributors 1 Describing Languages, Understanding Language: Systemic Functional Theory and Description

vii x

1

Y.J. DORAN AND J.R. MARTIN

PART I

Understanding Grammar 2 Axial Argumentation Below the Clause: The Verbal Group in Khorchin Mongolian

33

35

DONGBING ZHANG

3 Axial Argumentation and Cryptogrammar in Interpersonal Grammar: A Case Study of Classical Tibetan MOOD

73

PIN WANG

4 Experiential Cryptotypes: Reasoning About PROCESS TYPE

102

BEATRIZ QUIROZ

5 Axial Argumentation and Cryptogrammar in Textual Grammar: THEME in Brazilian Portuguese

129

GIACOMO FIGUEREDO

6 Arguments for Seeing Theme-Rheme and Topic-Comment as Separate Functional Structures RANDY J. LAPOLLA

162

vi

Contents

PART II

Contextualizing Grammar 7 Functional Language Typology: A Discourse Semantic Perspective

187

189

J.R. MARTIN AND BEATRIZ QUIROZ

8 Construing Relations Between Scientific Activities Through Mandarin Chinese

238

JING HAO

9 The Baboon and the Bee: Exploring Register Patterns Across Languages

273

DAVID ROSE

10 Language Shift: Bilingual Exchange Structure in Classroom Interactions

307

HARNI KARTIKA-NINGSIH

11 Academic Formalisms: Toward a Semiotic Typology

331

Y.J. DORAN

Index

359

Illustrations

Tables 2.1 3.1 3.2 3.3 4.1 4.2 5.1 5.2 5.3 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 7.10 7.11 8.1 11.1

Imperative Clauses and the Realizations of the Predicator Parallel of Terminology Between SFL and Traditional Language Typology Particles and Their Metafunctions Verbal Stems of Classical Tibetan Main Reactances for Facts and Ideas in Mental Clauses Patterns Associated With the Nature of Participants in Mental Processes Periodicity in Text 1 Periodicity and mood orientation in Text 2 Occurrences and Relative Frequency of theme, mood and transitivity in BP Spanish Demonstratives Discourse Semantic Systems (Organized by Metafunction) Entity Tracking in Spanish (la mujer ‘the woman’) Entity Tracking in English (the woman) Entity Tracking in Spanish (el anciano ‘the old man’) Entity Tracking in English (the old man) Spanish Verb Morphology Paradigm (Indicative Mood, Past Tense) Theme in the ‘Valladares to the rescue’ Phase of La Ola Maldita Theme in the English Translation of the ‘Valladares to the rescue’ Phase of La Ola Maldita Personal Pronouns (Clause Rank) in Chilean Spanish Pronominal Clitics (Group Rank) in Chilean Spanish Summary of Markings in Parataxis and Hypotaxis Field Meanings Realized by Academic Formalisms

63 75 79 85 123 123 137 148 154 194 198 208 209 210 211 214 223 225 236 236 262 339

viii 11.2 11.3 11.4 11.5

Illustrations Grammatical Organization and Field Relations in System Networks Grammatical Organization and Field Relations in Tree Diagrams Grammatical Organization and Field Relations in Mathematical Symbolism Grammatical Organization and Field Relations in Nuclear Equations

344 346 348 352

Figures 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 2.1 2.2 2.3 2.4 3.1 3.2 3.3 3.4 4.1 4.2 4.3

4.4

Syntagms and Structures (Visualization) for Example (2) Simplified Network of Clause Types (English transitivity) Nominal Group System and Structure (for Examples 1–18) Simplified Network of Clause Types (English mood) Cross-classification of Clauses (mood and transitivity Options) Discourse Semantic Systems, organized by Metafunction Language Strata Language, Register and Genre Mood Preselecting vg deixis in Khorchin Mongolian Transitivity Preselecting vg type in Khorchin Mongolian vg type, vg polarity, vg tense and aspect in the Elaborated Verbal Group The Verbal Group in Khorchin Mongolian System Network of the Imperative Mood System Network of the Interrogative Mood System Network of the Declarative Mood Synopsis of mood System in Classical Tibetan The Interplay of Agnation and Enation Across Sentences Potential Relations Between Sentences in Examples (6) and (7), Based on Gleason One Proportionality Relevant to the Distinction Between Declaratives and (Polar) Interrogatives in English Interlocking Agnation Relations Underlying Main Distinctions in process type

4 6 7 8 9 13 13 14 38 41 53 68 88 91 96 97 107 108

110 112

Illustrations 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9

A Model of Discourse Flow Discourse Flow in Text 1 The System of contextualization in BP The System of textual theme in BP The System of interpersonal theme in BP [absolute] in BP mood orientation in BP transitive orientation in BP Co-selection of mood and transitive orientation in BP 5.10 theme in Brazilian Portuguese 7.1 Extrinsic Functionality (Context) and Intrinsic Functionality (Language) 7.2 Language Strata 7.3 Basic ideation Systems 7.4 Discourse Semantic Systems 8.1 CONNEXION System 8.2 Network of Consequential Relations in English 9.1 Connotative and Denotative Semiotics 9.2 Cline of Instantiation 9.3 Story Genres Realized Intrastratally by Staging 9.4 Fields of Stories 9.5 Story Plots Realized Structurally by Story Phases 9.6 Expectancy and Sourcing in Story Plots 10.1 The Cycle of Teaching/Learning in R2L 10.2 Orbital Structure of R2L Learning Exchange 10.3 Options for Evaluate 10.4 Options for Language Shift in Exchanges 11.1 Tree Diagram 11.2 Minimal System Network 11.3 System With Three Choices 11.4 System Network With Two Levels of Delicacy 11.5 Simultaneous Systems 11.6 System Network Realizing Multiple Interdependent Classification Taxonomies 11.7 Basic Tree Diagram 11.8 Tree Diagram With Increased Depth 11.9 Tree Diagram With Increased Breadth 11.10 Lexical Substitution in Tree Diagrams

ix 131 133 136 141 142 144 145 151 153 155 196 197 199 206 248 267 275 276 278 279 280 301 309 311 311 318 334 340 341 341 343 343 344 345 346 353

Contributors

Y.J. Doran is Research Fellow in the LCT Centre for Knowledge-Building and Lecturer in the Department of Linguistics at the University of Sydney. His research focuses on language, semiosis, knowledge and education from the perspectives of Systemic Functional Linguistics and Legitimation Code Theory, spanning the interdisciplinary fields of educational linguistics, multimodality and language and identity. His research has developed grammatical, discourse semantic, register and genre-based descriptions of mathematics, language (focusing on English and Sundanese), images and academic formalisms, as well as fine-grained analytical tools for understanding variations in knowledge practices. In addition, he considers the organization of values systems that position communities of practice, how these are invoked through highly evaluative language and the theoretical issues that arise from these for our understanding of language, semiosis, knowledge and pedagogy. Giacomo Figueredo completed his Ph.D. on the description of Brazilian Portuguese Grammar. Currently, he is Senior Lecturer of Linguistics at the Federal University of Ouro Preto, Brazil, where he leads the Multilingu@Research Group, housing B.A., M.A. and Ph.D. projects on language description, modeling and generation. His research interests include systemic theory, language description, multilingual studies, language typology and translation. His research has been on language description and multilingual text production, focusing on exact descriptions of systemic probability and dynamics across text types in different languages. Jing Hao is a researcher who has held positions at the University of Sydney, Australia and The Hong Kong Polytechnic University, following her doctorate in linguistics at the University of Sydney. Her research explores disciplinary literacy, discourse analysis and knowledge building through English and Mandarin Chinese and their interaction with other semiotic modes, including images and body language.

Contributors

xi

Harni Kartika-Ningsih is Researcher in the Linguistics Department of Universitas Indonesia. She received her Ph.D. in linguistics from the University of Sydney and did her postdoctoral fellowship in the Department of Curriculum & Instruction, the Chinese University of Hong Kong. Her interests include Sydney School genre pedagogy, bilingual education and classroom discourse. Randy J. LaPolla (Ph.D. in linguistics from the University of California, Berkeley; Fellow of the Australian Academy for the Humanities) is Professor of Linguistics with a courtesy appointment in Chinese at Nanyang Technological University, Singapore. His research focuses on the history and typology of Sino-Tibetan and Austronesian languages and issues related to the nature of communicative behavior and functional explanations for the patterns found in languages and human behavior generally. J.R. Martin is Professor of Linguistics at the University of Sydney, where he is also Deputy Director of the LCT Centre for Knowledge-Building. His research interests include systemic theory, functional grammar, discourse semantics, register, genre, multimodality and critical discourse analysis, focusing on English, Korean, Spanish and Tagalog—with special reference to the transdisciplinary fields of educational linguistics, forensic linguistics and social semiotics. Recent publications include a collection of interviews with Wang Zhenhua (Interviews with J R Martin, Foreign Language and Research Press, Beijing 2015); a book on teaching academic discourse on-line (Genre Pedagogy in Higher Education, Palgrave Macmillan 2016), with Shoshana Dreyfus, Sally Humphrey and Ahmar Mahboob; and a book on Youth Justice Conferencing (Discourse and Diversionary Justice, Palgrave Macmillan 2018), with Michele Zappavigna. Eight volumes of his collected papers (edited by Wang Zhenhua, Shanghai Jiao Tong University Press) have been published in China (2010, 2012). Professor Martin was elected a fellow of the Australian Academy of the Humanities in 1998 and was Head of its Linguistics Section from 2010–2012; he was awarded a Centenary Medal for his services to Linguistics and Philology in 2003. In April 2014, Shanghai Jiao Tong University opened its Martin Centre for Appliable Linguistics, appointing Professor Martin as Director. Professor Martin is also affiliated with Pontificia Universidad Católica de Chile. Beatriz Quiroz is Assistant Professor in Language Sciences at the Faculty of Letters, Pontificia Universidad Católica de Chile (PUC-Chile), where she teaches and supervises undergraduate and postgraduate students in linguistics. She holds a Ph.D. in linguistics from the University of Sydney, Australia. Her current research, informed by Systemic Functional Linguistics (SFL), focuses on a metafunctionally integrated description

xii

Contributors of clause systems in Chilean Spanish, with special emphasis on the system-structure principle embodied by the theoretical dimension of axis. Other research interests include the interaction between lexicogrammar and discourse-semantics, and systemic functional language typology. Relevant publications include the forthcoming book Experiential Grammar in Systemic Functional Linguistics: Assumptions and Challenges (Equinox).

David Rose is Director of Reading to Learn, an international literacy program that trains teachers across school and university sectors in Australia, Africa, Asia and western Europe (www.readingtolearn.com.au). He is an Honorary Associate of the University of Sydney. His research includes analysis and design of classroom discourse, literacy teaching practices and teacher professional learning, together with language typology, language evolution and social semiotic theory. His work has been particularly concerned with Indigenous Australian communities, languages and education programs, with whom he has worked for over 30 years. Pin Wang is a lecturer and researcher at the Martin Centre for Appliable Linguistics of the School of Foreign Languages, Shanghai Jiao Tong University, China. His chief research interests are: systemic theory, functional grammar, and functional language typology, with particular focus on classical languages, Mandarin and minority languages of China. Dongbing Zhang is a Ph.D. candidate in linguistics at The University of Sydney whose research focuses on developing a systemic functional account of Khorchin Mongolian. His recent work includes an exploration of interpersonal systems in Khorchin Mongolian grammar, their orientation to discourse and their role in negotiating bonding and consensus.

1

Describing Languages, Understanding Language Systemic Functional Theory and Description Y.J. Doran and J.R. Martin

1. Systemic Functional Linguistics (SFL) For most readers SFL is probably best known for its descriptions of English grammar (e.g. Halliday 1985 and subsequent editions), cohesion and discourse structure (e.g. Halliday and Hasan 1976; Martin 1992), genre (e.g. Martin and Rose 2008), appraisal (e.g. Martin and White 2005) and multimodality (e.g. Kress & van Leeuwen 1990 and subsequent editions; Bateman 2008). Less well known are its descriptions of a wide range of languages and language families, as exemplified in Caffarel et al. (2004), Martin and Doran (2015a), Martin (2018), Martin et al. (in press a)— and surveyed in Mwinlaaru and Xuan (2016). In this section we introduce the basic theoretical architecture informing all this work. In order to maximize the accessibility of this introduction we will take English grammar as a point of departure, leaving it to subsequent chapters to deploy SFL’s basic principles for other languages and modalities. Examples are taken or adapted from the legal discourse text presented at the end of this chapter.1 The key references for this introduction are Matthiessen and Halliday (2009), Martin, Wang and Zhu (2013), and Martin et al. (in press b); for foundational papers see Halliday and Martin (1981) and Martin and Doran (2015b); for SFL handbooks see Bartlett and O’Grady (2017), Thompson et al. (2019); for surveys of the range of work carried out in SFL see Hasan et al. (2005, 2007), Halliday and Webster (2009), Webster (2015). Like all linguists, SFL linguists are concerned with non-random (i.e. informationally redundant) combinations of linguistic entities—be they phonological, grammatical, discoursal or beyond. Consider for example the words making up There will be a seat for him. In this example we can begin by naming the word classes involved (leaving aside the word there for now): 1. word class

will be auxiliary verb verb

a determiner

seat for noun preposition

him pronoun

2 Y.J. Doran and J.R. Martin We can take a further step by combining these words into three significant combinations (technically, syntagms): will be, a seat and for him. We can name these syntagms of words as classes in their own right—as verbal group, nominal group and prepositional phrase in this example. This means that if we encounter a syntagm like a seat elsewhere (e.g. it might give him a seat), we can refer to it as a whole (i.e. as a nominal group) and not simply as a non-random sequence of determiner and noun. 1’.

will

be

a

seat

for

him

word class

auxiliary verb verb determiner noun preposition pronoun

group/phrase class

verbal group

nominal group

prepositional phrase

In this grammar, we can push up by naming non-random sequences of groups and phrases as a class of clause (e.g. declarative); and we can push down to allow for non-random sequences of morphemes constituting words (e.g. do-n’t and seat-s). This gives us a scale of units in our grammar, with clause classes consisting of one or more group/phrase classes, group/phrase classes consisting of one or more word classes, and word classes consisting of one or more morpheme classes. This SFL constituency hierarchy is referred to as rank. 2. ranks

you don’t have vacant seats you do -n’t

have vacant

seat

stem stem

stem pl

morpheme class

2sg

prs

word class

pronoun

auxiliary verb verb

neg

group/phrase class

nominal group verbal group

clause class

relational clause

-s

adjective noun nominal group

For a given grammatical unit, syntagmatic relations of this kind are often represented as an image in a phrase structure tree—popularized in the formal syntax inspired by Chomsky’s postgraduate research 1955– 1956, 1957.2 Phrase structure trees are typically binary branching, drawing on what American structuralists referred to as “immediate constituent (IC) analysis” (cf. Wells 1947). Various abbreviations, regularly drawing on Harris (1946), are used in these visualizations for the names of classes. In SFL, recognizing classes is just one step as far as grammatical description is concerned. We also need to understand what each class is doing in any instance—i.e. we need to determine the function of each class. One reason for this is that the same function can be performed by more than one class and the same class can perform more than one function.3 A verb for example can be used as the Event function in a verbal group: (verb as Event) 3. So you always overbook

Describing Languages, Understanding Language

3

or it can be used as a Classifier function in a nominal group: (verb as Classifier) 4. the overbooking policy Similarly, a Classifier can also be realized by a noun or an adjective. (noun as Classifier) 5. a plane booking (noun as Classifier) (adjective as Classifier) 6. a standard booking When functioning as Classifiers, potentially gradable classes (e.g. adjectives and verbs) cannot be graded because they are making categorical distinctions. The overbooking policy is here a type of policy, and similarly the standard booking here is opposed to a non-standard booking (rather than being relatively more or less standard); 7. *the heavily overbooking policy 8. *a rather standard policy4 We can accordingly expand 2 as 2’, including an analysis of alternating syntagms (configurations of classes) and structures (configurations of functions). By convention in SFL, class labels are written in lower case and function labels being with an initial upper case letter (with the exception of the morpheme classes where we have followed the standard nonSFL Leipzig glossing conventions for capitalization): 2’. rank

you don’t have vacant seats you do -n’t

have

vacant

seat

-s

morpheme class

2sg

stem

stem

stem

pl

prs

neg

morpheme function

Head

Head Suffix

word class

pronoun

auxiliary verb verb

word function

Thing

Tense

group/phrase class

nominal group

verbal group

group/phrase function Carrier Process clause class relational clause

Head

Head

Head Suffix

adjective

noun

Event Epithet

Thing

nominal group Attribute

An analysis of this kind can be alternatively presented as a tree, with nodes labeled for function and class (i.e. for both what they do and for what they are; see Figure 1.1).

4 Y.J. Doran and J.R. Martin

Figure 1.1 Syntagms and Structures (Visualization) for Example (2)

The distinction between class and function is of course a familiar one in linguistics, although many theories make little or no use of function labels in grammatical analysis per se. In SFL this distinction between syntagms (as sequences of classes) and structures (as configurations of functions) is crucial because it is relations among structures that are the focus of the theory.5 This enables SFL to produce richer grammatical descriptions than if it focused on syntagms alone. In our focus text, for example, the lawyer Taylor is introduced as representing his client Edmonds, which means that he was acting on Edmond’s behalf in a legal disputation. The clause syntagm in (9) is the same as that in (2)—nominal group followed by verbal group followed by nominal group; 9. He

was representing a guy called Edwards

nominal group verbal group 2. You

don’t have

nominal group verbal group

nominal group vacant seats nominal group

Despite the similarity in the syntagm, these sequences of classes realize different structures. We can see this from the fact that they enter into distinctive relationships with other clauses that distinguish one structure

Describing Languages, Understanding Language

5

from the other. For example, (9) alternates with passive clauses, but (2) does not. 10. Edmonds was being represented by Taylor. 11. *Vacant seats weren’t had by the airlines. And for activity concurrent with the moment of speaking (9) uses [present in present] tense, whereas (2) uses [present]. 12. He is representing Edmonds. 13. *You aren’t having vacant seats. Had [present] tense been deployed in (2) it would have shifted the temporality from concurrence with the moment of speaking to habitual behavior, as in (14). 14. Taylor represents Edmonds (whenever he needs legal aid). Note that we are focusing on relations between clauses here, not relations among verbs. The verb represent can be used in clauses which pattern like (2) as far as tense is concerned—but when it does, as in (15), the structure does not construe forensic activity but rather a representational relationship between an entity and its symbolization. 15. The logo represents British Airways. Unlike (2) this clause does have a passive. 16. British Airways is represented by that logo. But unlike (10), it doesn’t have an ‘agentless’ one; both nominal groups have to be present. 17. Edwards is being represented (by Taylor). 18. British Airways is represented *(by that logo). We will not pursue further this discussion of the different structures that can be realized through a nominal group followed by verbal group followed by nominal group syntagm (for one comprehensive SFL description of these in English see Halliday 1985 and subsequent editions). Rather, what we are foregrounding here is the way in which SFL reasons about relationships among structures rather than focusing simply on classes and their sequence (i.e. syntagms). In doing so, SFL draws on Gleason’s (1965) notions of agnation and enation and Whorf’s (1945)

6 Y.J. Doran and J.R. Martin discussion of overt and covert categories (his phenotypes and cryptotypes respectively), explored in detail in Quiroz (this volume). The distinction between 1, 2, 15 and 16 on the one hand and 3, 9, 10, 12, 14 and 17 on the other for example is a distinction between relational vs. material clauses (Halliday 1985)—in Whorf’s terms a cryptogrammatical distinction. Cryptogrammars are based on what Whorf called reactances—the distinctive syntagms that emerge or not as agnate units are explored (such as those involving the alternations of tense and voice touched on earlier). 1.1 Cryptogrammatical Reasoning Cryptogrammatical reasoning of this kind is the basis for the development of systems in SFL (such as the simplified network of clause types in Figure 1.2). Each type of clause has distinguishing agnation patterns that motivate the distinctions (just some of which we noted earlier). As far as structure is concerned, distinctive configurations of functions are typically proposed for more general systems. Halliday 1985 suggests Actor, Process and Goal for material clauses (e.g. He [Actor] was representing [Process] a guy called Edwards [Goal]), Senser, Process and Phenomenon for mental clauses (e.g. Edmonds [Senser] liked [Process] that sunshine [Phenomenon]) and Token, Process and Value for relational clauses (the logo [Token] represents [Process] British Airways [Value]). For further discussion of axial reasoning in relation to networks of this kind see Quiroz (this volume).

Figure 1.2 Simplified Network of Clause Types (English transitivity)

At this point we can return to the rank scale outlined previously and reconsider it from a paradigmatic perspective (i.e. from the perspective of system). What ranks actually represent are generalized points of origin for system networks. That is, they are bundles of relations among structures realized as clauses, among structures realized as groups and phrases, among structures realized as words and among structures realized as morphemes. For example, the nominal groups realizing the participant functions Actor, Goal, Senser Phenomenon, Token and Value in material, mental and relational clauses can be generalized in simplified terms as [designating] or [specifying]; if [designating], as [naming] involving a name, or [pronaming] involving a pronoun; and if [specifying], as optionally [classified], by inserting a Classifier, and/or [determined], by inserting a Deictic.

Describing Languages, Understanding Language

7

The network for these systems is presented as Figure 1.3, along with realization statements relating paradigmatic choices to syntagmatic structure. The square bracket ‘[’ represents ‘or’ relations and the curly bracket (the brace ‘{’) represents ‘and’. Realization statements are prefaced with a downward slanting arrow. In these statements, ‘+’ means insert a function, ‘^’ sequences functions, ‘#’ represents the first or last position of a unit (so ‘#^’ means insert first and ‘^#’ means insert last) and ‘:’ indicates the class realizing a function—typically at the rank below (e.g. the group function Thing realized by the word class pronoun).

Figure 1.3 Nominal Group System and Structure (for Examples 1–18)

As we can see, the realization statements allow us to transition between ranks. If realization statements had been provided for the clause systems in Figure 1.2, they would have selected options from the nominal group network in Figure 1.3 (since clause functions such as Actor or Senser can be realized as nominal groups); and as we have seen, the realization statements in Figure 1.3 select from word class networks that we have not specified (but would include classes such as nominals, adjectives, verbals etc.). For details of this notation and the axial reasoning involved, see Martin et al. (2013, in press a); and see Zhang (this volume) and Wang (this volume) for detailed explorations of the distribution of systems and structures across clause, group and word ranks in Mongolian and Classical Tibetan. 1.2 Metafunction At this point let’s return to our original example and focus on an additional syntagm there—namely there will. We have purposely left this

8 Y.J. Doran and J.R. Martin aside to this point, as it illustrates a second factor in organizing grammars having to do with metafunction. 1. There will be a seat for him. Related to this clause, small adjustments in the sequence of this syntagm can involve tags, interrogatives (both polar and elemental) and elliptical responses: 19. There will be a seat for him, will there? 20. Will there be a seat for him? 21. When will there be a seat for him? 22. Will there? And the absence of this syntagm from a clause marks a shift from indicative to imperative mood: 23. ø Be an advocate for him. Drawing on terms from Halliday (1985) we can generalize the function of the syntagms here as a Subject•Finite structure.6 The position and sequence of both the Subject and Finite in any given clause plays a critical role in positioning the clause as an interact in dialogue (referred to in SFL as the system of mood), as well as establishing the nub and terms of its negotiability (Martin 2018). A network for some of the key systems and their realization in structure is in presented in Figure 1.4. These systems reveal a different organization of the clause from what we have seen earlier. In fact, they cross-classify the system in Figure 1.3, since the different transitivity choices (distinguishing between, for example, material, mental and relational clauses) combine freely with the different mood choices in Figure 1.4.

Figure 1.4 Simplified Network of Clause Types (English mood)

Describing Languages, Understanding Language

9

In the mood system, the realization statements specify the presence of a Predicator (+Pred) for all clauses and the presence of a Subject and Finite function (+Subj, +Fin) for indicative clauses. The realization statements also indicate the realization of the Predicator as a non-finite verbal group in imperative clauses (Pred: non-finite) and a finite verbal group in indicative ones (Pred: finite), the sequence of Subject followed by Finite (Subject^Finite) for declarative clauses and Finite followed by Subject (Finite^Subject) for interrogative clauses, and the presence of a Wh function (+Wh), in initial position (#^Wh), followed by Finite (Wh^Finite) in elemental interrogatives.

Figure 1.5 Cross-Classification of Clauses (mood and transitivity Options)

The way in which clauses are cross-classified by the transitivity and mood options introduced here is formalized in the system network in Figure 1.5. The network offers a glimpse of the way in which systems subclassify and cross-classify one another as clause relations are formalized. Based on his analyses of English grammar (Halliday 1967a, b, 1968, 1969, 1970, 1973), Halliday noted that clause rank systems tended to bundle into three groups based on their relative interdependence on one another (adding theme systems to the crossclassification in Figure 1.5; see Figueredo in this volume for theme in Brazilian Portuguese). These bundles, Halliday suggested (with reference to the work of the Prague School; see LaPolla this volume) reflect a general metafunctional organization of language—with ideational meaning construing a model of the world outside and inside us, interpersonal meaning enacting our social relations and textual meaning

10 Y.J. Doran and J.R. Martin composing information flow (as reflected in the clause rank systems of transitivity, mood and theme respectively). In SFL this metafunctional perspective on system dependencies complements the rankbased dependencies introduced earlier. The key point here is that in a paradigmatically organized grammar the global architecture (rank and metafunction to this point in our discussion) is determined by axis (i.e. system/structure relations).

1.3 Strata The third piece of SFL architecture we will introduce here has to do with levels of language known as strata. So far we’ve been focusing on grammar.7 Looking from ‘below,’ we also have to account for the phonology (Tench 1992; Halliday and Greaves 2008), graphology (Mountford 1998) and ‘signology’ (Rudge 2017) that encodes language in ways we materialize as physical ‘information’ we can hear and see. Looking from ‘above’ we face the challenge of meaning beyond the clause (Martin and Rose 2007). This means taking into account (1) co-textual relations of indefinite extent realized within or between clauses (Martin 2015) and (2) generalizations of meanings realized across a range of grammatical systems (including what Halliday 1985 introduces as congruent and metaphorical realizations). We can exemplify this by drawing on examples dealing with reference and modalization (assessments of probability) in our source text. Co-textual relations are illustrated in (24), where text reference (highlighted in bold) is used within and between clause complexes (i.e. orthographic sentences) to track the idea (in italics) that in extreme circumstances, because of overbooking, a customer may not get a seat. 24. They did not reveal this idea that in very extreme circumstances there may not be a seat. Didn’t reveal that. So British Airways’ argument was “Well this is so remote. It is so unlikely to occur that we don’t worry about it and you really can’t class it as a misrepresentation, a deceit.” Text references of this kind can be used retrospectively or prospectively to identify indefinitely long phases of discourse—viz this referencing the whole of Lord Wilberforce’s judgement. 25. Well the courts tend to take a dim view of anyone who lies irrespective of the reasons, at least in commerce anyway. And Lord Wilberforce said this.

Describing Languages, Understanding Language

11

“The letter, taken with the ticket, would be taken as a statement that Mr. Edmonds had a certain booking, which statement, in view of the overbooking policy, was untrue, since his booking, though very likely to be a firm one, was exposed to risk, small, but as events proved real, and it might give him a seat on the aircraft, it might not. This was a statement of fact rather than a statement of mere intention.” Compare the use of that to identify implied entities (i.e. sunshine in the Caribbean and fog and dampness in London). 26. Edmonds had lived on the island of Bermuda for a long time, down there in the Caribbean. Liked that sunshine. Decided after many years of this he’d like to take a holiday in London. Experience some of that fog and dampness. Turning to meanings realized across a range of grammatical systems, consider the more subjective (27 and 29) and more objective (28 and 30) realizations of probability (Halliday 1982). 27. there may not be a seat 28. It is so unlikely to occur that 29. it might give him a seat on the aircraft, it might not 30. though very likely to be a firm one We can expand the paradigm by taking verbal and clausal realizations into account. 31. It’ll perhaps give him a seat. 32. I reckon it’ll give him a seat. Halliday scales these alternatives from most subjective to most objective as follows: 32. I reckon it’ll give him a seat. 29. it might give him a seat on the aircraft, it might not 31. It’ll perhaps give him a seat. 33. It is likely to give him a seat. And he further suggests that the ends of the scale involve grammatical metaphor, with the grammar symbolizing the semantics rather than directly realizing it. The explicitly subjective realization I reckon for example is literally a mental clause with a Senser; (32) has not been modalized with

12 Y.J. Doran and J.R. Martin a modal verb or adverb. But we do not negotiate an example of this kind with reference to its projecting clause. Rather we negotiate the projection (i.e. the idea it reports), because the mental clause is symbolizing subjective modulization; it is not realizing a cognitive figure. 34. *I reckon it’ll give him a seat, don’t I? 35. I reckon it’ll give him a seat, won’t it? – Yes, it will. The explicitly objective realization it is likely literally realizes the modulization as an Attribute in a relational clause—as a description of how the seat will be made available. If graded (e.g. very likely) and nominalized (e.g. considerable likelihood) this objectifying lexicalization of probability starts to merge with digital construals of probability (ultimately formalized mathematically as statistics). This shift from interpersonal assessment to ideational factuality is an important dimension of argumentation in the legal disputation we are using to contextualize examples here (cf. 36 and 37). 36. Only two times, the facts indicated, in every 10,000 bookings does someone not make it. 37. since his booking, though very likely to be a firm one, was exposed to risk, small, but as events proved real, Co-textualization and generalization of this kind for ideational, interpersonal and textual meaning encouraged the development of a stratified content plane in SFL (Martin 2014, 2016)—comprising a discourse semantics realized through lexicogrammar. The major discourse semantic systems organized by metafunction are outlined in Figure 1.6 (Martin and Rose 2007) and introduced in more detail in Martin and Quiroz (this volume). In short, negotiation comprises resources for organizing moves in exchanges, appraisal comprises resources for sharing feelings, ideation comprises resources for construing experience as figures, connexion comprises resources for connecting figures, identification comprises resources for introducing and tracking entities and periodicity comprises resources for phasing discourse as waves of information. The interpersonal assessment systems introduced earlier for modalization realize appraisal—specifically the sub-systems of engagement positioning voices in discourse (Martin and White 2005). This third dimension of SFL architecture is outlined in Figure 1.7. The co-tangential circles represent the idea that discourse semantic patterns need to be interpreted as patterns of lexicogrammatical patterns, and

Describing Languages, Understanding Language

13

Figure 1.6 Discourse Semantic Systems, Organized by Metafunction

Figure 1.7 Language Strata

lexicogrammatical patterns as patterns of phonological, graphological or signed ones (a relationship referred to technically as metaredundancy). Their increasing size reflects the fact that our focal unit of analysis tends to get bigger as we move from phonology (e.g. syllable system and

14 Y.J. Doran and J.R. Martin

Figure 1.8 Language, Register and Genre

structure) through lexicogrammar (e.g. clause system and structure) to discourse semantics (e.g. text system and structure). A further step into SFL’s architecture arises from the fact that to understand why a particular set of linguistic choices are made in any instance, models of language need a model of context. This involves understanding the social purposes of the language in focus, and the broader semiotic variables that underpin why a particular text is how it is. In the model of SFL generally assumed in this volume, context is modeled as two higher level strata: genre and register. Genre offers a means for understanding the overall social purpose of a text and the particular stages it will go through. For our law lecture text, Hood and Maggiora (2016) describe it from the perspective of genre as an exemplum (the more specific genre analysis below is based on Martin in press). An exemplum is a story genre where there is an unresolved disruption of activity that is then interpreted on moral or ethical grounds. In broad terms, it is a genre oriented to evaluation of behavior in terms of particular social standards (Martin and Rose 2008). This two-fold purpose of detailing a disruptive event and judging it ethically or morally is reflected in the typical exemplum staging of an Incident, which records the problematic events, followed by an Interpretation, which evaluates the behavior in the Incident. In this instance, the genre is relatively self-contextualizing. So in addition to the Incident^Interpretation stages, it also includes an Abstract and

Describing Languages, Understanding Language

15

an Orientation. These are opening stages available for all story genres that predict what is to come (Abstract), and situate the story in its time and/or place and introduce the people involved in it (Orientation). In addition the Interpretation is technicalized in terms of legal principles. In particular, it is concerned with whether British Airways’ promise of a seat for Edwards can be classed as a statement of fact, in which case it was a misrepresentation by virtue of there not being a seat for Edwards, or whether it was a statement of intention, in which case it was neither deceitful nor a misrepresentation. The final judgement involves three perspectives. One is that of British Airways, who are positioned by the lecturer as arguing that “Well this is so remote. It is so unlikely to occur that we don’t worry about it and you can’t really class it as a misrepresentation, a deceit.” In contrast, Edwards is said to argue that he has “Been waiting for years for my holiday. Wanted to get up to London. Now I can’t get there. Everything’s mucked up.” But as the text arises in a pedagogic setting focused on teaching the law, the crucial component of the Interpretation here is the legal judgement by Lord Wilberforce. In this judgement, Lord Wilberforce classes British Airways’ promise of a seat as “a statement of fact.” The three perspectives in the Interpretation reflect the complex exigencies of this pedagogic context. On the one hand, it simply retells an unfortunate event experienced by Edwards which reflects poorly on British Airways. But its relevance to the legal system arises from the fact that this story figures as a legal precedent regarding what constitutes a statement of fact rather than a statement of intention. Perhaps most importantly, this is all underpinned by the pedagogic purpose of the law lecture— namely to apprentice law students into the nuances of the legal system, not just tell a story with legal implications. So . . . British Airways and Taylor.

Abstract

Taylor was a guy who worked for a consumer agency, a govern- Orientation ment agency. He was representing a guy called Edmonds. Edmonds had lived on the island of Bermuda for a long time, down there in the Caribbean. Liked that sunshine. Decided after many years of this he’d like to take a holiday in Incident London. Experience some of that fog and dampness. Get up there. Take in some shows. Go to Wimbledon. Hadn’t had a holiday for a long, long time and so this is a big deal for him. Cheap airfare was on offer if you were the early bird. Get in a couple of months early, you book your name, much cheaper. So he did this. Books his flight for London. Plans his holidays. Becomes a little bit concerned that maybe his place will get taken if there’s a crowd, and someone who paid full fare would be in his seat. So he rings them up, gets some reassurance that that’s not the case. There will be a seat for him. And in fact gets a letter confirming the seat that’s there. Obviously, he turns up at the airport in Bermuda, no seat.

16 Y.J. Doran and J.R. Martin It turns out that the policy of British Airways, along with many Interpretation of its competitors, was to overbook because it was realized that some people don’t turn up, they get sick, there’s misadventure. So you always overbook and that way you don’t have vacant seats and you are able to maximize the income of the flight. Didn’t work this time. Only two times, the facts indicated, in every 10,000 bookings does someone not make it. But the airline had said to him “There is a seat for you.” They did not reveal this idea that in very extreme circumstances there may not be a seat. Didn’t reveal that. So British Airways’ argument was “Well this is so remote. It is so unlikely to occur that we don’t worry about it and you really can’t class it as a misrepresentation, a deceit.” Of course Edmonds had the other view. “Been waiting for years for my holiday. Wanted to get up to London. Now I can’t get there. Everything’s mucked up.” So you see the logic of both arguments. Well the courts tend to take a dim view of anyone who lies irrespective of the reasons, at least in commerce anyway. And Lord Wilberforce said this. “The letter, taken with the ticket, would be taken as a statement that Mr. Edmonds had a certain booking, which statement, in view of the overbooking policy, was untrue, since his booking, though very likely to be a firm one, was exposed to risk, small, but as events proved real, and it might give him a seat on the aircraft, it might not. This was a statement of fact rather than a statement of mere intention.” Hence, this was considered then a statement of fact. Why? Because you told me today there was a seat for me in the future. It was not really, as you’ve masked it, a statement today that you intend to give me a seat in the future. Do you see the subtle difference? Not sure? Too subtle? Not getting a lot of feedback here. Is anyone unsure of the subtlety, the difference between someone? Remember the example we used the other day when we were talking about this sort of stuff. If someone says to you, “Will you meet me for coffee tomorrow, at ten o’clock?” and you say “Yes.” And at the time you say that you have no intention of turning up, even though it is in the future, it is still a lie today. Correct? This is the same thing in this case.

The different stages in this genre are realized through distinct register patterns. In the first instance, they are different in terms of their orientation to the field of the text—i.e. the type of phenomena they construe (Doran and Martin 2020). In the Incident stage, for example, there is a phase of momented activity, where events are listed in temporal sequence, one after the other (‘^’ indicates sequence—only Edwards’ activities are listed here): Decided after many years of this he’d like to take a holiday in London. ^

Describing Languages, Understanding Language

17

So he did this. Books his flight for London. ^ Plans his holidays. ^ Becomes a little bit concerned that maybe his place will get taken if there’s a crowd, and someone who paid full fare would be in his seat. ^ So he rings them up, ^ gets some reassurance that that’s not the case. There will be a seat for him. ^ And in fact gets a letter confirming the seat that’s there. ^ Obviously, he turns up at the airport in Bermuda, ^ no seat. In the Interpretation stage, in contrast, the meanings are not oriented to simply stepping through events, but rather with determining whether the sale of a ticket and letter of confirmation are legally classified as a statement of fact or a statement of interpretation (and thus, whether the ticket and letter of confirmation involve a misrepresentation or not). Lord Wilberforce sums this up by stating unequivocally: This was a statement of fact rather than a statement of mere intention. These two stages thus construe two types of phenomena. The Incident construes the field as a dynamic unfolding of events, while the Interpretation construes the field as a static taxonomizing of (abstract) items (i.e. the statements). At the same time, the text displays shifts in the two other variables of register, tenor (involving the interpersonal relations of power and solidarity) and mode (involving shifts in the contextualizing role language is playing). As noted earlier, the lecturer regularly moves back and forth between language that is careful to make explicit the relations among its meanings and language that is more implicit about these meanings (and so more open to interpretation). This reflects shifts between language constituting a field (language as reflection) and language more ancillary to a field (language as action). Lord Wilberforce’s judgement involves more reflective language, leaving very little to the listener to ‘read between the lines’: The letter, taken with the ticket, would be taken as a statement that Mr. Edmonds had a certain booking, which statement, in view of the

18 Y.J. Doran and J.R. Martin overbooking policy, was untrue, since his booking, though very likely to be a firm one, was exposed to risk, small, but as events proved real, and it might give him a seat on the aircraft, it might not. This was a statement of fact rather than a statement of mere intention. In contrast, when the lecturer is attempting to check what the students understood, the language shifts to being more reliant on the physical and social setting it is spoken in, as well as becoming considerably more dialogic: Do you see the subtle difference? Not sure? Too subtle? Not getting a lot of feedback here. Is anyone unsure of the subtlety, the difference between someone? In common-sense terms, Lord Wilberforce’s judgement is more ‘written’ in the way it construes meaning, while the lecturer’s attempt at engaging students is more ‘spoken.’ Engaging students at this point in the lecture of course involves more than spoken language. It also involves the ‘paralinguistic’ resources of body language (Hood and Maggiora 2016; Martinec 1998, 2000, 2001; Martin and Zappavigna 2019), the spatial configuration of the room (positioning the lecturer below the students who sit in tiered rows of seats above) (e.g. Ravelli and McMurtrie 2016), as well as the slides projected onto the screen and the range of various image or other resources projected there at any time (e.g. symbols, O’Halloran 2005; Doran 2018; music, van Leeuwen 1999; or film, Bateman and Schmidt 2012). In general terms then, for a comprehensive analysis of discourse, we need a multimodal model of semiosis. SFL theory has been foundational to the development of a social semiotic model of discourse, inspired by seminal descriptions of images by Kress and van Leeuwen (1990) and O’Toole (1994).

2. Contributions to This Volume This volume is organized into two parts. Part I, Understanding Grammar, explores lexicogrammatical systems and structures across languages. It focuses in particular on how meaning-making resources are distributed across ranks and diversified across metafunctions. Zhang (Chapter 2) carefully illustrates this through a description of the verbal group in Khorchin Mongolian. Beginning with clause-based distinctions, he shows that in order to understand both interpersonal and ideational clause distinctions in Mongolian, the description must focus on the highly elaborated systems of the verbal group that realize them. This in turns implicates a number of word-class distinctions and ultimately morphological markings. Working along these lines, Zhang shows clearly how

Describing Languages, Understanding Language

19

a language can distribute metafunctionally diversified meanings across ranks. Throughout the description the chapter carefully illustrates axial argumentation—the motivation of a description in both systemic and structural terms—in ways that underpin each of the chapters in the rest of the volume. The three following chapters focus more specifically on clause grammar, foregrounding different metafunctions. Wang (Chapter 3) explores the enactment of interpersonal meaning in Classical Tibetan. Taking the discourse semantic system of negotiation as its point of departure (the system oriented to dialogic exchanges), the description shows that Classical Tibetan does not make a generalizable distinction between clause types that give or demand information (i.e. between interrogative and declarative clauses within mood). Rather, Classical Tibetan enacts dialogue through an intricate interplay of three main clausal systems— mood, polarity and a third system termed finalisation which is realized through Predicator-final interpersonal particles. This linguo-centric description underscores the need for careful axial argumentation— argumentation that does not take the familiar distinction between interrogative and declarative moods for granted. Quiroz (Chapter 4) turns her attention to cryptogrammatical argumentation in experiential clause grammar, focusing on transitivity in Chilean Spanish. Following an in-depth discussion of what a cryptogrammar involves, Quiroz explores ways of reasoning about process types in Systemic Functional Linguistics. She exemplifies this by progressively building up a range of reactances for distinguishing types of mental clauses in Spanish. Throughout the chapter Quiroz shows that sets of reactances cluster together in ways that enable distinctions that cannot be recognized if, on the one hand, there was a focus on verb types (rather than clause types) and if, on the other, there was a reliance on explicit markings (i.e. on phenotypes rather than cryptotypes). Figueredo (Chapter 5) extends this discussion of clause grammar into the textual grammar of Brazilian Portuguese. Here Figueredo explores the crucial role textual resources play in the logogenesis of discourse. This involves first establishing clear distinctions in the discourse semantic system of periodicity (concerned with waves of informational prominence in discourse) and in the ways texts shift or maintain their gaze on a particular field. From this, Figueredo is able to establish grammatical distinctions in the clausal system of theme and their realization in the group/phrase rank below—realizations that compose these discourse semantic variations. This chapter shows the importance of looking at grammar in relation to co-text, and also illustrates how descriptions sensitive to discourse semantic patterns can be developed. LaPolla (Chapter 6) rounds out the focus on clause grammar in this volume by exploring in detail the distinctions between Theme-Rheme, Topic-Comment and Given-New as they are used in linguistic typology,

20 Y.J. Doran and J.R. Martin Systemic Functional Linguistics, the Prague School and other approaches. He does this through an exploration of information organization in the Tagalog clause (with reference to both Mandarin Chinese and English)— arguing that a clear distinction between Theme-Rheme and Topic-Comment structures is needed in order to understand typological variation across languages. Aside from this important typological proposal, this chapter also presents a useful positioning of Systemic Functional grammatical description in relation to a range of alternative approaches. Part II comprises studies that extend the Systemic Functional descriptive methods to areas contextualizing grammar—including larger-scale discourse patterns, social functionality and multimodality. Martin and Quiroz (Chapter 7) frame this concern by considering what it means to develop functional language typology within Systemic Functional Linguistics. They suggest that Systemic Functional language typology will necessarily be a defeasible typology; every generalization or comparison will always be subject to reconsideration from a different vantage point in the architecture of the theory overall. This means that to compare languages we must be prepared to ascend from the syntagmatic axis to the paradigmatic axis (i.e. from system to structure), from lower ranks to higher ranks (i.e. from morphemes up to clauses), from phonology to lexicogrammar to discourse semantics (i.e. across strata), from language to register and genre (i.e. from text to context)—all the while being prepared to descend again, shifting our gaze across metafunctions and multiple semiotic resources. This contribution to Systemic Functional language description and typology exemplifies its perspective through a comparison of Chilean Spanish and English and the resources they use to realize a range of discourse semantic meanings. Hao (Chapter 8) extends this approach to language description by focusing methodologically on ‘top-down’ language description. This chapter focuses on how scientific causality is construed in Mandarin Chinese, taking meaning at the contextual levels of register and genre as its point of departure. It then considers how contextual choices at these levels affect language choices, in terms of discourse semantic connexion and then in turn in terms of lexicogrammatical choices—including clause complexing, verbal aspect and modality. Hao’s top-down approach complements the familiar ‘bottom-up’ perspective that so often moves from phonology to grammar to semantics and pragmatics, and/or within the grammar from morphology up to syntax. In addition it offers an insightful means for developing descriptions along appliable lines, as it illustrates how a description can be formulated with respect to whatever aspect of language is of special relevance for the applied goals in mind. Rose (Chapter 9) takes contextualization one step further by developing a typology of registerial patterns in their own right. Focusing on phases within story genres, this chapter illustrates the relative similarity of such patterns across languages (specifically Pitjantjatjara, Oromo

Describing Languages, Understanding Language

21

and Mandarin Chinese), taking into account the significant differences in the way phases are realized in discourse semantics and lexicogrammar. In the first instance, this chapter offers a clear illustration of the approaches explored in both Martin and Quiroz (Chapter 7) and Hao (Chapter 8), where the similarity of patterns ‘from above’ (in this case register patterns) differ considerably when looking at their realization at lower levels. In the second instance, this chapter pushes our understanding of typology further than the typical focus on phonological, grammatical or even semantic systems, by developing a contextual typology with explicit reference to register. Kartika (Chapter 10) explores language description from the most explicitly appliable perspective in this volume. She is concerned with how we can design language use in multilingual classroom settings. She illustrates her approach through dialogic exchanges involving Indonesian and English (with some reference to the Sundanese also used in her data). This allows Kartika to contextualize code-switching from the perspective of the discourse semantic system of negotiation, with the aim of improving learning outcomes. Complementing Hao’s chapter, Kartika shows how one can come at language description from an appliable underpinning, in this case the adaptation of Reading to Learn methodology (Rose and Martin 2012) for bilingual education. In the final chapter, Doran (Chapter 11) extends the functional description and typology perspective developed in this book toward functional semiotic typology. In particular he explores four key academic formalisms used in academic discourse—system networks and tree diagrams in linguistics, and algebraic mathematical symbolism and nuclear equations used in physics. Echoing the approach of Martin and Quiroz (Chapter 7), he illustrates that depending on the perspective, the formalisms look either significantly different or remarkably the same. From the perspective of register (in a particular field), they each construe significantly different meanings oriented to their particular disciplinary environments. But from the perspective of their structural (syntagmatic) organization in the grammar, they display a remarkable similarity by virtue of their functionality in relation to language. The chapter thus synthesizes Systemic Functional approaches to both multimodality and language typology to help build a Systemic Functional semiotic typology.

3. Contributions of This Volume The chapters in this volume each orient to their own languages, their own problems and their own solutions. But throughout there is significant coherence in the approaches they take. This has largely arisen from the close interactions between most of the authors in this book and the shared commitment to extending both Systemic Functional Linguistics to new languages and to developing a responsible, appliable linguistics.

22 Y.J. Doran and J.R. Martin This shared commitment and close interaction means that the volume presents a number of key contributions to the description of languages, as well as our understanding of language in general and our ways of doing linguistics. These cut across methodologies for description, perspectives on functional language typology, the role of multilinguality and multimodality in typology and the appliability of linguistics. 3.1 Methodology Throughout the volume, our focus has been on methodologies for describing language. At stake here is the challenge of comparing like with like in a discipline where different theories of language are explicitly adopted (or implicitly assumed), different kinds of argumentation for categorization are privileged and different kinds of data underpin grammatical descriptions. This book explicitly addresses this challenge from the perspective of SFL by considering the functionality of language from four perspectives: i. ii. iii. iv.

axial functionality metafunctionality co-textual functionality contextual functionality

Axial functionality underpins the discussions and grounds each of the perspectives in this volume, including those from rank, strata, metafunction and context. This explicit focus on system/structure relations (introduced in Section 1 earlier, and explored in Quiroz in this volume and Martin et al. 2013) ensures that each chapter approaches the variation in language by reasoning from above, around and below (either in terms of rank or strata) through paradigms of enation and agnation. In this way, each description motivates their distinctions in terms of both the particular systems under study and their realization in appropriate function structures and class syntagms. This enables descriptions to move away from developing categories based on their notional ‘meanings’ and lessens the possibility for distinctions to be unproblematically transferred from one language to another (often drawing on Halliday’s 1985 description English). Similarly it puts forward a principled understanding of how the other aspects of language—metafunction, rank and strata—arise from axially motivated linguistic patterns and how they can be integrated into a wholistic description. Metafunctionality ensures that descriptions in this volume focus carefully on both the interdependency of systems and the multifunctionality of any particular instance. The chapters of Quiroz (experiential), Wang (interpersonal), Figueredo and LaPolla (both textual) explicitly address one metafunction in the clause in a way that brings out the particular ways of arguing needed to grasp one particular dimension of meaning,

Describing Languages, Understanding Language

23

and Zhang shows how the realization of metafunctionally diversified meanings can be distributed across ranks. Metafunctional description means understanding the same instances from different angles; it involves seeing that there are general functions all languages perform in terms of organizing our experience, enacting our social relations and composing texts and privileging any one of these functions (in theory, in analysis or in practice) gives a distorted picture of the way language has evolved for people to use it to live. Co-textual functionality ensures that discussions in this volume are oriented to patterns in discourse. The approaches in each chapter are not based on isolated instances cut off from their wider co-textual background. Rather, they explore in detail the relations between linguistic patterns and their orientation to larger text patterns. This is most explicitly seen in the chapters of Martin and Quiroz (exploring functional typology from discourse semantics), Figueredo (detailing the interaction of textual grammar with discourse semantic periodicity) and Kartika (exploring multilingual interaction through the system of negotiation). These considerations mean that co-textual relations (i.e. meanings beyond the clause) are carefully considered alongside both lexicogrammatical relations and contextual relations. Finally, contextual functionality ensures that the descriptions in this volume engage responsibly with meaning beyond language and on to the registers and genres they are situated in. Such a focus enables explorations of what aspects of language are used for what purposes and in what situations. One key manifestation of this is explicitness in terms of the registers and genres of the data that underpin the descriptions in most chapters. Rose, for example, makes clear his comparison between languages and orients to phases within story genres, while Hao steps through her description of Mandarin Chinese from a ‘top-down’ perspective in terms of the resources used in explanations of scientific causality. Doran takes this one step further by exploring the organization of semiotic systems other than language in terms of the registerial meanings they realize. This opens up the possibility of investigating how semiotic systems serve particular contextual functionalities in relation to those realized through language. On top of this, an approach exploring contextual functionality enables contextual meanings to be described not as a static ‘extra-semiotic’ reality, but as parameters of semiotic variation in themselves. This is fundamental to a linguistics and semiotics that engages with an indefinitely wide range of instances and data, and so is a crucial focus of this volume. 3.2 Functional Language Typology The second key contribution of this volume is to the development of a functional language typology. As noted earlier, a number of chapters in this volume consider explicitly how Systemic Functional Linguistics can

24 Y.J. Doran and J.R. Martin deal with language comparison and the generalization of patterns across languages—more specifically how it can develop a functional typology, as opposed to a typology of syntagms. We have already stepped through what we mean by functional here; but methodologically, the key chapter in this regard is that of Martin and Quiroz. They explain that the key manoeuvre when considering typology is ascent from ‘lower’ levels to ‘higher’ ones. What may look different in terms of syntagmatic organization (structure) may look similar paradigmatically (system); what differences there are in the morphological organization of a grammar may be neutralized when considered from a higher rank such as the clause; what may be distinct in terms of the phonological or grammatical organization of language may realize similar meanings at the levels of discourse semantics, register or genre. And conversely, what may look similar at any of these lower levels may realize quite distinct meanings at the higher levels. Developing a functional language typology means being able to compare and contrast languages from as many perspectives as possible. In fact this volume does more than outline an approach to Systemic Functional typology; it also includes two chapters that explicitly extend the boundaries of such research: Rose shows how SFL can develop registerial typology (i.e. a typology of contextual meaning) and Doran takes a step toward building semiotic typology (i.e. a typology of meaning-making in general). In this sense, this volume pushes Systemic Functional typology further than it has gone before. 3.3 Multilinguality and Multimodality Approaches that deal with text and context implicate two further variables: multilinguality and multimodality. As SFL descriptions and typologies expand their scope, understanding how multiple languages and multiple semiotic resources are used to form coherent texts will become an increasingly pervasive issue. Indeed, as decades of research into ‘code-switching’ and multimodality have shown, such multilingual/ multimodal texts are the norm rather than the exception. In this volume, Kartika’s chapter offers one way of managing multilinguality by considering its functionality in terms of the discourse semantic meanings it organizes. In the bilingual classrooms under study, Kartika shows that the shifts between language can be understood from the perspective of the dialogic resources of negotiation. Such a perspective is vital if we are to understand the multilingual texts holistically, rather than as an un-theorized switching from one language to another. Similarly, Doran’s approach to understanding the diversity of semiotic resources relies on considering their functionality from a range of perspectives. This enables typology to move beyond relatively common-sense distinctions such as ‘image’ vs. ‘symbolism’ to show that there is an underlying similarity between the resources being studied despite their apparent differences at face value.

Describing Languages, Understanding Language

25

3.4 Appliability Perhaps the most important aspect of the approaches in this volume is their commitment to an appliable linguistics (Halliday 2005). SFL is known for its work across a range of applications—including educational linguistics, clinical linguistics, forensic linguistics, translation studies, computational linguistics and beyond. The crucial component in all of these applications is that they are founded upon rich descriptions of the language (and semiosis) in play. As applications expand across the world, the need for descriptions that take seriously the specific functionality of the language being used is becoming increasingly pressing. In particular, descriptions that can target the relevant areas of language (be they at particular ranks, strata or metafunctions) and typologies that can see similarity and difference depending on the perspective taken are becoming vital. Descriptions and typologies that take into account the functionalities described earlier are much more powerful than those that do not. Similarly, approaches to language that do not create separate silos for ‘applied linguistics’ and ‘linguistics’ are of much greater use to the world than those that do. The dialectic of theory and practice, encompassed by the term appliable linguistics, is what underpins the work in this volume. SFL’s long-standing commitment to social responsibility is in many ways what has driven the development of its extravagant theory and descriptive apparatus, as new situations, new contexts and new challenges arise that necessitate an expansion of its approach into new areas of language and semiosis and a reconsideration of already ‘understood’ components of language (Martin 2016; Halliday 2013). In this book we have leaned toward the descriptive end of appliable linguistics, but this should not obscure SFL’s commitment to application. The papers here should be read in conjunction with the increasing scope of applications of SFL that arise each year (see Caldwell et al. in prep. for a survey of numerous SFL scholars’ shifts between applied and theoretical/descriptive concerns and how this has expanded SFL’s applicable framework). Similarly, everything in this book should be reconsidered in this sense as new challenges arise, new problems are tackled and new situations are encountered. We can summarize this in terms of Halliday’s dream of a “socially accountable linguistics, and this in two distinct though related senses: that it put language in its social context, and at the same time it put linguistics in its social context, as a mode of intervention in critical social practices” (1993: 73). We hope this volume takes steps toward this goal and in turn encourages others to develop their work in socially responsible ways.

Notes 1. The text is taken from data analyzed in Hood and Maggiora (2016) and used with their permission. 2. A representative sample can be accessed at the following URL: www.google. com/search?q=phrase+structure+trees+images&tbm=isch&source=univ&clie

26 Y.J. Doran and J.R. Martin

3. 4.

5. 6. 7.

nt=firefox-b-d&sa=X&ved=2ahUKEwjvpZfDucngAhWTTX0KHdtYCQ0Qs AR6BAgFEAE&biw=2145&bih=1301&dpr=2 Outside SFL classes are often referred to as categories, and functions as relations. In this instance, grading with rather indicates a shift in function from Classifier— a non-gradable discrete classification as standard or not-standard—to an Epithet, with varying degrees of standardness (but this is not the intended meaning in this context). For another example from linguistics, one can use standard as a Classifier to distinguish a ‘standard variety’ of a language from a ‘non-standard’ variety. Here there are no degrees of standardness; it is a categorical distinction. Relations among syntagms are accounted for indirectly, since they realize structures (cf. the function and class relations in example 2’ and Figures 1.1, 1.3 and 1.4). The dot ‘•’ indicates that the Subject and Finite are not in the same sequence across all clause types (i.e. various syntagms realizing the same structure). More specifically, we can say we have been focusing on lexicogrammar. The term lexicogrammar signals that in SFL grammar and lexis are not treated as separate modules. Rather, lexis is treated as the realization of increasingly delicate grammatical systems (Hasan 1987), as we move from say the systems triggering the insertion of a Modal Adjunct realized by modal adverbs to specific choices for modalization (realized by lexical items such as perhaps, probably, certainly etc.).

References Bartlett, T. and O’Grady, G. (eds.). 2017. Routledge Handbook of Systemic Functional Linguistics. London: Routledge. Bateman, J.A. 2008. Multimodality and Genre: A Foundation for the Systematic Analysis of Multimodal Documents. London: Palgrave Macmillan. Bateman, J.A. and Schmidt, K. 2012. Multimodal Film Analysis: How Films Mean. London: Routledge. Caffarel, A., Martin, J.R. and Matthiessen, C.M.I.M. (eds.). 2004. Language Typology: A Functional Perspective. Amsterdam: Benjamins (Current Issues in Linguistic Theory). Caldwell, D., Knox, J. and Martin, J.R. (eds.). (in prep.) Developing Theory: A Handbook in Appliable Linguistics and Semiotics. London: Bloomsbury. Chomsky, N. 1955–1956. The Logical Structure of Linguistic Theory. Cambridge, MA: Harvard University MA Thesis. Chomsky, N. 1957. Syntactic Structures. The Hague: Mouton. Doran, Y.J. 2018. The Discourse of Physics: Building Knowledge Through Language, Mathematics and Image. London: Routledge. Doran, Y.J. and Martin, J.R. 2020. Field relations: Understanding scientific explanations. In K. Maton, J.R. Martin and Y.J. Doran (eds.), Studying Science: Knowledge, Language, Pedagogy. London: Routledge. Gleason, H.A. Jr. 1965. Linguistics and English Grammar. New York: Holt, Rinehart & Winston. Halliday, M.A.K. 1967a. Notes on transitivity and theme in English: Part 1. Journal of Linguistics 3.1. 37–81. Halliday, M.A.K. 1967b. Notes on transitivity and theme in English: Part 2. Journal of Linguistics 3.2. 199–244.

Describing Languages, Understanding Language

27

Halliday, M.A.K. 1968. Notes on transitivity and theme in English: Part 3. Journal of Linguistics 4.2. 179–215. Halliday, M.A.K. 1969. Options and functions in the English clause. Brno Studies in English 8. 81–88. Halliday, M.A.K. 1970. Functional diversity in language, as seen from a consideration of modality and mood in English. Foundations of Language 6.3. 322–361. Halliday, M.A.K. 1973. The functional basis of language. In B. Bernstein (ed.), Class, codes and control, Volume 2: Applied studies towards a sociology of language. London: Routledge & Kegan Paul. 343–366. Halliday, M.A.K. 1982. The de-automatization of grammar: From Priestley’s An Inspector Calls. In J.M. Anderson (ed.), Language Form and Language Variation: Papers Dedicated to Angus McIntosh (Current Issues in Linguistic Theory 15). Amsterdam: John Benjamins. 129–159. Halliday, M.A.K. 1985. An Introduction to Functional Grammar. London: Arnold (2nd edition 1994, 3rd edition revised by C.M.I.M. Matthiessen 2004, 4th edition revised by C.M.I.M. Matthiessen 2014). Halliday, M.A.K. 1993. Language in a changing world. Occasional Papers of the Applied Linguistics Association of Australia. 13. Halliday, M.A.K. 2005. Interview with Anne Burns. In J.R. Martin (ed.) 2013. Interviews With M.A.K. Halliday: Language Turned Back on Himself. London: Bloomsbury. Halliday, M.A.K. 2013. Interviews With M.A.K. Halliday: Language Turned Back on Himself. J.R. Martin (ed.). London: Bloomsbury. Halliday, M.A.K. and Greaves, W.S. 2008. Intonation in the Grammar of English. London: Equinox. Halliday, M.A.K. and Hasan, R. 1976. Cohesion in English. London: Longman (English Language Series 9). Halliday, M.A.K. and Martin, J.R. (eds.). 1981. Readings in Systemic Linguistics. London: Batsford. Halliday, M.A.K. and Webster, J.J. (eds.). 2009. Continuum Companion to Systemic Functional Linguistics. London: Continuum. Harris, Z.S. 1946. From morpheme to utterance. Language 22. 161–183. Hasan, R. 1987. The grammarian’s dream: Lexis as most delicate grammar. In M.A.K. Halliday and R.P. Fawcett (eds.), New developments in systemic linguistics Volume 1: Theory and description. London: Pinter. 184–211. Hasan, R., Matthiessen, C.M.I.M. and Webster, J.J. (eds.). 2005. Continuing Discourse on Language: A Functional Perspective. Vol. 1. London: Equinox. Hasan, R., Matthiessen, C.M.I.M. and Webster, J.J. (eds.). 2007. Continuing Discourse on Language: A Functional Perspective. Vol. 2. London: Equinox. Hood, S. and Maggiora, P. 2016. The lecturer at work: Language, the body and space in the structuring of disciplinary knowledge in law. In H. de Silva Joyce (ed.), Language at Work in Social Contexts: Analysing Language Use in Work, Educational, Medical and Museum Contexts. Newcastle upon Tyne, UK: Cambridge Scholars. 108–128. Kress, G. and van Leeuwen, T. 1990. Reading Images. Geelong: Deakin University Press. Martin, J.R. 1992. English Text: System and Structure. Amsterdam: Benjamins. Martin, J.R. 2014. Evolving systemic functional linguistics: Beyond the clause. Functional Linguistics 1.3.

28 Y.J. Doran and J.R. Martin Martin, J.R. 2015. Meaning beyond the clause: Co-textual relations. Linguistics and the Human Sciences 11.1–2. 203–235. Martin, J.R. 2016. Meaning matters: A short history of systemic functional linguistics. Word 61.2. 35–58. Martin, J.R. (ed.). 2018. Interpersonal meaning: Systemic Functional Linguistics perspectives. Special Issue of Functions of Language 25.1. 2018. Martin, J.R. in press. Genre and activity: A potential site for dialogue between Systemic Functional Linguistics (SFL) and Cultural Historical Activity Theory (CHAT). Mind, Culture, Activity. (Special Issue on SFL and CHAT edited by Michael Cole). Martin, J.R. and Doran, Y.J. (eds.). 2015a. Grammatical Descriptions. London: Routledge (Critical Concepts in Linguistics: Systemic Functional Linguistics, Vol. 2). Martin, J.R. and Doran, Y.J. (eds.). 2015b. Grammatics. London: Routledge (Critical Concepts in Linguistics: Systemic Functional Linguistics, Vol. 1). Martin, J.R., Quiroz, B. and Figueredo, G. (eds.). in press a. Interpersonal Grammar: Systemic Functional Linguistic Theory and Description. Cambridge: Cambridge University Press. Martin, J.R., Quiroz, B., Wang, P. and Zhu, Y. in press b. Systemic Functional Grammar: Another Step Into the Theory—Grammatical Description. Beijing: Higher Education Press. Martin, J.R. and Rose, D. 2007. Working With Discourse: Meaning Beyond the Clause (Second Edition). London: Continuum. Martin, J.R. and Rose, D. 2008. Genre Relations: Mapping Culture. London: Equinox. Martin, J.R., Wang, P. and Zhu, Y. 2013. Systemic Functional Grammar: A Next Step Into the Theory—Axial Relations. Beijing: Higher Education Press. Martin, J.R. and White, P.R.R. 2005. The Language of Evaluation: Appraisal in English. London: Palgrave. Martin, J.R. and Zappavigna, M. 2019. Embodied meaning: A systemic functional perspective on body language. Functional Linguistics 6.1. Martinec, R. 1998. Cohesion in action. Semiotica. 120.1/2. 161–180. Martinec, R. 2000. Types of process in action. Semiotica. 130.3/4. 243–268. Martinec, R. 2001. Interpersonal resources in action. Semiotica. 135.1/4. 117–145. Matthiessen, C.M.I.M. and Halliday, M.A.K. 2009. Systemic Functional Grammar: A First Step Into the Theory. Beijing: Higher Education Press. Mountford, J. 1998. An Insight Into English Spelling. London: Hodder & Stoughton. Mwinlaaru, I.N. and Xuan, W.W. 2016. A survey of studies in systemic functional language description and typology. Functional Linguistics 3.8. O’Halloran, K.L. 2005. Mathematical Discourse: Language, Symbolism and Visual Images. London: Continuum. O’Toole, M. 1994. The Language of Displayed Visual Art. Cranbury: Associated University Presses. Ravelli, L. and McMurtrie, R. 2016. Multimodality in the Built Environment: Spatial Discourse Analysis. London: Routledge. Rose, D. and Martin, J.R. 2012. Learning to Write/Reading to Learn: Genre, Knowledge and Pedagogy in the Sydney School. London: Equinox.

Describing Languages, Understanding Language

29

Rudge, L. 2017. Analysing British Sign Language Through the Lens of Systemic Functional Linguistics. University of Western England PhD thesis, Bristol. Tench, P. (ed.). 1992. Studies in Systemic Phonology. London: Pinter Publishers. Thompson, G., Bowcher, W.L., Fontaine, L. and Schönthal, D. (eds.). 2019. The Cambridge Handbook of Systemic Functional Linguistics. Cambridge: Cambridge University Press. van Leeuwen, T. 1999. Speech, Music, Sound. London: Macmillan. Webster, J.J. (ed.). 2015. The Bloomsbury Companion to M A K Halliday. London: Bloomsbury. Wells, R. 1947. Immediate constituents. Language 23.212–226. Whorf, B.L. 1945. Grammatical categories. Language 21.1–11.

Appendix (source text contextualizing examples in this chapter)

So . . . British Airways and Taylor. Taylor was a guy who worked for a consumer agency, a government agency. He was representing a guy called Edmonds. Edmonds had lived on the island of Bermuda for a long time, down there in the Caribbean. Liked that sunshine. Decided after many years of this he’d like to take a holiday in London. Experience some of that fog and dampness. Get up there. Take in some shows. Go to Wimbledon. Hadn’t had a holiday for a long, long time and so this is a big deal for him. Cheap airfare was on offer if you were the early bird. Get in a couple of months early, you book your name, much cheaper. So he did this. Books his flight for London. Plans his holidays. Becomes a little bit concerned that maybe his place will get taken if there’s a crowd, and someone who paid full fare would be in his seat. So he rings them up, gets some reassurance that that’s not the case. There will be a seat for him. And in fact gets a letter confirming the seat that’s there. Obviously, he turns up at the airport in Bermuda, no seat. It turns out that the policy of British Airways, along with many of its competitors, was to overbook because it was realised that some people don’t turn up, they get sick, there’s misadventure. So you always overbook and that way you don’t have vacant seats and you are able to maximise the income of the flight. Didn’t work this time. Only two times, the facts indicated, in every 10,000 bookings does someone not make it. But the airline had said to him “There is a seat for you.” They did not reveal this idea that in very extreme circumstances there may not be a seat. Didn’t reveal that. So British Airways’ argument was “Well this is so remote. It is so unlikely to occur that we don’t worry about it and you really can’t class it as a misrepresentation, a deceit.” Of course Edmonds had the other view. “Been waiting for years for my holiday. Wanted to get up to London. Now I can’t get there. Everything’s mucked up.” So you see the logic of both arguments. Well the courts tend to take a dim view of anyone who lies irrespective of the reasons, at least in commerce anyway. And Lord Wilberforce said this.

Describing Languages, Understanding Language

31

“The letter, taken with the ticket, would be taken as a statement that Mr. Edmonds had a certain booking, which statement, in view of the overbooking policy, was untrue, since his booking, though very likely to be a firm one, was exposed to risk, small, but as events proved real, and it might give him a seat on the aircraft, it might not. This was a statement of fact rather than a statement of mere intention.” Hence, this was considered then a statement of fact. Why? Because you told me today there was a seat for me in the future. It was not really, as you’ve masked it, a statement today that you intend to give me a seat in the future. Do you see the subtle difference? Not sure? Too subtle? Not getting a lot of feedback here. Is anyone unsure of the subtlety, the difference between someone? Remember the example we used the other day when we were talking about this sort of stuff. If someone says to you, “Will you meet me for coffee tomorrow, at ten o’clock?” and you say “Yes.” And at the time you say that you have no intention of turning up, even though it is in the future, it is still a lie today. Correct? This is the same thing in this case.

Part I

Understanding Grammar

2

Axial Argumentation Below the Clause The Verbal Group in Khorchin Mongolian Dongbing Zhang

1. Introduction This chapter shows how axial argumentation (as outlined in Martin 2013) can inform systemic functional language description at ranks below the clause. For a synopsis of this type of argumentation, see Martin, Doran and Figueredo (this volume). At the same time, the chapter provides a systemic functional account of the verbal group in Khorchin Mongolian.1 The description takes as its starting point the role of the verbal group in relation to how a clause in which it functions enacts interpersonal meaning and construes experiential meaning in discourse. In terms of the SFL descriptive principle of the “trinocular perspective” (Halliday 2009), this chapter begins by establishing the basic paradigmatic organization of the Khorchin Mongolian verbal group ‘from above.’ The role of the verbal group in the interpersonal organization of a clause is examined first, followed by the role of the verbal group in the experiential organization of a clause. The interpersonal organization of a Khorchin Mongolian clause is shaped by the negotiation of information and goods and services. A verbal group system, vg deixis,2 is established in relation to the basic clause distinctions in mood. The features in the system of vg deixis are [elaborated] and [restricted] (see Section 2). The experiential organization of a Khorchin Mongolian clause is shaped by the construal of different types of experience. A verbal group system, vg type, is established in relation to the basic clause distinctions in transitivity. The features in the system of vg type are [eventive] and [noneventive] (see Section 3). The two systems, vg deixis and vg type, are co-selecting simultaneous systems. The co-selections from the two systems result in the choices [elaborated eventive], [elaborated non-eventive], [restricted eventive] and [restricted non-eventive]. Section 4 establishes the further paradigmatic organization of the Khorchin Mongolian verbal group ‘from roundabout’ in terms of the interdependencies between systems and ‘from below’ in terms of the structural patterns motivating the systems. The systems dependent on the choices from

36

Dongbing Zhang

vg deixis and vg type are discussed in relation to the structural configurations that realize them. In Section 4.1 the major resources involved in the elaborated verbal groups, i.e. vg polarity, vg tense, aspect, vg modality and relative tense are discussed. An elaborated eventive verbal group selects from all the five systems; an elaborated non-eventive verbal group cannot select from vg modality and relative tense. In Section 4.2, attention shifts to the resources involved in the [restricted] verbal groups, i.e. vg person and aspect. Both eventive and non-eventive restricted verbal groups select from these two systems. In Section 5, conclusions are drawn based on the axial argumentation throughout Section 2 to Section 4. The meaning-making resources of the Khorchin Mongolian verbal group are summarized with a system network in this section. The description in this chapter shows how axial argumentation with a discourse semantic orientation (Martin and Quiroz in this volume) provides a useful way of making explicit the valeur of a particular grammatical category. This is a special focus in the discussions of aspect and relative tense in Section 4.1.3, Section 4.1.5 and Section 4.2.2.

2. The Verbal Group in Service of the Interpersonal Organization of a Clause If we understand communication as exchanges, we can exchange either knowledge or action. In Khorchin Mongolian, an exchange of knowledge and an exchange of action are typically realized by different grammatical categories (see also Wang in this volume for Classical Tibetan). An exchange of knowledge is typically realized by indicative clauses; an exchange of action is typically realized by imperative ones. There are a number of motivations for the distinction between [indicative] and [imperative] in the Khorchin Mongolian interpersonal clause system of mood. The most important motivation is the meaning-making potential of the verbal group3 in the two clause types. The verbal group which functions in an indicative clause is more elaborate in its potential than the one that functions in an imperative clause. The verbal group in an indicative clause can choose from the systems vg polarity, vg tense, aspect, vg modality and relative tense; the verbal group in an imperative clause, on the other hand, can only choose from the systems vg person and aspect. The first set of systems is introduced in Section 4.1; the second set is introduced in Section 4.2. I will use the terms [elaborated] and [restricted] to refer to the classes of the verbal group which function in an indicative and an imperative clause respectively. The name of this system is referred to as vg deixis, following Quiroz (2013).4 If we use the function label Predicator to capture the function of the verbal group in the interpersonal organization of a clause, Predicators preselect different classes of the verbal group from the system of vg

Axial Argumentation Below the Clause

37

deixis. The Predicator in an indicative clause is realized by an elaborated verbal group; the Predicator in an imperative clause is realized by a restricted verbal group. The exchange of knowledge in (1)5 exemplifies indicative clauses. Predicator in each clause is realized by an elaborated verbal group. In the exchange, a government official (O) demands a piece of information from a peasant (P); the peasant then gives that piece of information. In terms of SFL work on exchange structure (Berry 1981; Martin 1992, 2018; Martin and Rose 2007), we have a knowledge exchange where a K2 move demands information and a K1 move gives information. (1) O:

K2

tʰɐn

ɐs6

ɐsu-sɘn

mɛ

2pl

abl

ask-pst

qp

Predicator elaborated verbal group “Did (they) ask you?” P:

K1

ɐsu-sɘn

kuɛ

ask-pst

neg

Predicator elaborated verbal group “(They) didn’t ask (us).”

The verbal groups in (1) selects from both vg tense and vg polarity. The verbal group in K2 selects [past/positive]; the verbal group in K1 selects [past/negative], indicated by the negation adverb kuɛ. The choices in the system of vg tense and vg polarity are introduced in Section 4.1. The exchange of action in (2) on the other hand exemplifies imperative clauses. Predicator in the imperative clauses in (2.1) and (2.3) is realised by restricted verbal group. In the exchange, a grandmother (G) asks her granddaughter, Hairhan (H), to eat some more oranges. The request is first rejected by Hairhan and then accepted when her grandmother insists. In terms of exchange structure, we have an action exchange where an A2 move demands an action and a compliant A1 move (realized nonverbally) performs the action; but these two moves are first interrupted by a challenging move (ch) and a response to the challenge move (rch). Here we will focus only on the imperative clauses—(2.1) and (2.3): (2) G: 1. A2 xɛːrxɐn

itə

Hairhan eat.imp.2 Predicator restricted verbal group “Hairhan, eat.”

38

Dongbing Zhang H: 2. ch ɔː ukuɛ nɛi it-x oh neg grandma eat-npst “Oh, no, grandma, I won’t eat.” G: 3. rch tʃ ʰɐmɐ t 2sg

ʃiɔʃiɔ

nɛ

uɛ neg

lɛ res

uk-jə

dat small gen give-imp.1 Predicator restricted verbal group

“Let (me) give you the smaller ones.” H: 4. A1 [Accepts and eats the orange.]

The types of clause the verbal groups function in condition the resources available to them. The verbal groups in (2.1) and (2.3) show distinct choices from those in (1). Whereas those functioning in the indicative clauses in (1) select from vg tense and vg polarity, those functioning in imperative clauses in (2) select from vg person. In this case, the verbal group in (2.1) selects [second person] and the verbal group in (2.3) selects [first person]. The choices in the system of vg person are explored in detail in Section 4.2. The relationship between clause types and verbal group types can be represented as a relationship of preselection between system networks. Here, choices in the clause system of mood preselect choices in the verbal group system of vg deixis as shown in Figure 2.1.7

Figure 2.1 Mood Preselecting vg deixis in Khorchin Mongolian

Axial Argumentation Below the Clause

39

3. The Verbal Group in Service of the Experiential Organization of a Clause Complementing the view of conversations as enacting exchanges of knowledge or action as in Section 2 (i.e. an interpersonal perspective), we can also consider them in terms of the kinds of experience construed (i.e. an experiential perspective). From this perspective, examples (1) and (2) earlier, are concerned with experiences that are relatively dynamic (saying and doing) while example (3) is concerned with an experience that is relatively static (being). In terms of the SFL work on ideation in discourse (Hao 2019, this volume; Martin 1992; Martin and Quiroz this volume; Martin and Rose 2007), we are concerned with the construal of occurrence figures in (1) and (2), and the construal of a state figure in (3). In (3), the husband (H) queries if there is some more flour left, to which the wife (W) gives a positive answer. (3) H:

pɐ

kʊjir

still

flour cop-npst

pɛː-ø

mɛ qp

verbal group “(Is there) still some flour left?” W: pɛː-nɐ cop-npst verbal group “(There) is.”

In Khorchin Mongolian, state figures and occurrence figures are realized by different grammatical categories. A state figure is typically realized by a relational clause as in (3); an occurrence figure is typically realized by a non-relational clause as in (1) and (2) earlier in Section 2. There are a number of motivations for the distinction between [relational] and [non-relational] in the experiential clause system of transitivity. The most important motivation is once again the meaning-making potential of the verbal group. The verbal group that functions in a nonrelational clause is more elaborate in its potential than the one that functions in a relational clause. Verbal groups in non-relational clauses can choose from the systems of vg polarity, vg tense, aspect, vg modality, relative tense and vg person, whereas verbal groups in relational clauses can choose from the systems vg polarity, vg tense, aspect and vg person, but not from vg modality or relative tense. The choices in these systems are introduced in Section 4. I will use the terms [eventive] and [non-eventive] to refer to the classes of the verbal group which function in non-relational and relational clauses respectively. This system is referred to as vg type.

40

Dongbing Zhang

If we use the function label Process to capture the function of the verbal group in the experiential organization of a clause, then we can say that Process is realized by different classes of the verbal group from the system vg type. The Process in a non-relational clause is realized by an eventive verbal group; the Process in a relational clause is realized by a non-eventive verbal group. The exchange in (3) exemplifies relational clauses. The first move is repeated as (4). The Process in the relational clause in (4) is realized by a non-eventive verbal group. (4) pɐ

kʊjir

pɛː-ø

mɛ

still flour cop-npst

qp

Process non-eventive verbal group “(Is there) still some flour left?”

The verbal group in (4) selects [non-past/positive] from vg tense and vg polarity. However, it cannot select from vg modality or relative tense. The exchanges in (1) and (2) above on the other hand exemplify nonrelational clauses. Move (2.2) is adjusted as (5) here. The Process in the non-relational clause in (5) is realized by an eventive verbal group. (5)

ɔː

ukuɛ

nɛi

it-x

uɛ

oh

neg

grandma

eat-npst

neg

Process eventive verbal group “Oh, no, grandma, I won’t eat.”

The verbal group in (5) selects [non-past/negative] from vg tense and vg polarity. However, in contrast to the non-eventive verbal group in (4) it also has the potential to select from vg modality and relative tense (see Section 4.1.4 and 4.1.5). Like for the interpersonal perspective given earlier, this establishes a relationship of preselection between the clause system of transitivity and the verbal group system of vg type, as shown in Figure 2.2. In this figure, the interpersonal system of vg deixis is also included in the system network. So far, two verbal group systems, vg deixis and vg type, have been established in relation to the role the verbal group plays in a clause in Khorchin Mongolian. The vg deixis system is related to the interpersonal organization of a clause; the vg type system is related to the experiential organization of a clause. In terms of ‘trinocular perspective,’ this

Axial Argumentation Below the Clause

41

Figure 2.2 Transitivity Preselecting vg type in Khorchin Mongolian

section has established the primary paradigmatic organization of the Khorchin Mongolian verbal group ‘from above.’

4. Verbal Group System and Structure This section examines the Khorchin Mongolian verbal group from two further perspectives to provide a holistic description. It considers the verbal group ‘from roundabout’ in terms of the interdependencies between systems (i.e. paradigmatic environment) and ‘from below’ in terms of the structural patterns motivating the systemic choices (i.e. syntagmatic organization). The systemic choices in the verbal group are realized by configurations of verbal group functions, which in turn are realized by classes of verb. To see how these configurations work, we first need to introduce the set of verb classes, exemplified here with a lexical verb xiː ‘do’ and a copular verb pɛː ‘be’ respectively. Their possible variations with respect to either the vowel of the verb stem or a following particle are also included.

42

Dongbing Zhang xiː ‘do’ pɛː ‘be’ variations

1.

v

xiː

pɛː

-

2.

v-j

xiː-jə

pɛː-jɐ

-jɐ, -jə, -jɔ

3.

v-k

xiː-k

pɛː-k

-k, -kɐ, -kə, -kɔ, -ɘk

4.

v-n

xiː-n

pɛː-n

-n, -nɐ, -nə, -nɔ, -ɘn

5.

v-tʃɛ

xiː-tʃɛ

pɛː-tʃɛ

-tʃ

6.

v-x

xiː-x

pɛː-x

-x, -ø

7.

v-sɘn xiː-sɘn

pɛː-sɘn

-sɘn, -sɘ

8.

v-ntʃ xiː-ntʃ

-

-

9.

v-tʃ

pɛː-tʃ

-

xiː-tʃ

Verb classes 1 to 3 are related to the restricted verbal group system vg person. Verb classes 4 to 8 on the other hand are related to the elaborated verbal group systems of vg tense and relative tense. Verb class 9 functions either in relation to vg modality, a system available for elaborated verbal groups, or aspect, a system available to both restricted verbal groups and elaborated verbal groups. Section 4.1 and Section 4.2 are organized in terms of the choices from the vg deixis system discussed in Section 2: [elaborated] and [restricted]. Section 4.1 describes the resources available to the elaborated verbal group. The relevant systems are vg polarity, vg tense, aspect, vg modality and relative tense. Section 4.2 describes the resources available to the restricted verbal group. The relevant systems are vg person and aspect. The choices from the vg type system discussed in Section 3, [eventive] and [non-eventive], are discussed in turn in each subsection. 4.1 The Elaborated Verbal Group In this section, I explore the systems available for an elaborated verbal group. Elaborated verbal groups function in indicative clauses. Three systems are selected by both the eventive and the non-eventive verbal group. They are vg polarity, vg tense, and aspect. They are described in Section 4.1.1, 4.1.2, and 4.1.3 respectively. In each section, the eventive verbal group is discussed and exemplified first; this is followed by discussion and exemplification of the non-eventive verbal group. In Section 4.1.4 and 4.1.5, the focus is on the additional meaningmaking potential of an eventive verbal group. An eventive verbal group has the potential to choose from the systems vg modality and relative tense, which are not available for a non-eventive verbal group. 4.1.1 VG POLARITY One of the basic functions of a language is to allow the speakers to query or confirm the polarity of a piece of information. The verbal group in

Axial Argumentation Below the Clause

43

Khorchin Mongolian plays a key role in this regard. In terms of the function played by a verbal group in the interpersonal organization of a clause, when an indicative clause is positive, the Predicator is realized by a positive verbal group; when it is negative, the Predicator is realized by a negative verbal group. The exchange in (6) (repeating (1)) exemplifies the vg polarity of an eventive verbal group. The verbal group in (6.1) selects [positive]; the verbal group in (6.2) selects [negative]. (6) O:

1. tʰɐn ɐs 2pl

ɐsu-sɘn

mɛ

abl ask-pst

qp

Predicator positive verbal group “Did (they) ask you?” P:

2. ɐsu-sɘn

kuɛ8

ask-pst

neg

Predicator negative verbal group “(They) didn’t ask (us).”

Note that a positive verbal group as in (6.1), when there is no selection from the other systems discussed later (i.e. aspect, vg modality and relative tense), comprises only one verb. When [negative] is selected (or choices from any other verbal group systems are made) a combination of words constitutes the verbal group (as in example (6.2)). The [positive] and the [negative] features are thus realized by different structures. The verbal groups in (6) are analyzed in (7) and (8). The function of the lexical verb in an eventive verbal group is called Event; the function of the negation adverb is called Negator. When an eventive verbal group is negative, a Negator function is inserted and is realized at the final position as in (7); when it is positive, there is no function which explicitly realizes the feature as in (8). (7) ɐsu-sɘn

kuɛ

ask-pst

neg

Event

Negator

‘didn’t ask’ (8) ɐsu-tʃɛ ask-pst Event ‘asked’

44

Dongbing Zhang

In contrast, the exchange in (9) exemplifies the vg polarity of a noneventive verbal group. The first verbal group in (9.1) selects [positive]; the second one selects [negative]. All the other verbal groups in this exchange select [positive]. T1 asks if a student’s sentence is grammatically correct in (9.1). T2, who is senior in Mongolian language teaching, affirms that the sentence is grammatically correct in (9.2). T3 double checks this answer in (9.3) and is reaffirmed by T1 in (9.4). A question mark is used to show the rising intonation in (9.3). Note that the English translation here is misleading; the word tʰɐːr translated as ‘correct’ in Khorchin Mongolian is in fact a verb. (9) T1:

1.

ɘn

uːlpɘr

tʰɐːr-ø9

mu

tʰɐːr-x

uɛ

mu

prox

sentence

correct-npst

qp

correct-npst

neg

qp

verbal group

verbal group

“Does this sentence correct or not correct? (= Is or isn’t this sentence correct?)” T2:

2. tʰɐːr-nɐ correct-npst verbal group “Correct.”

T3

3. tʰɐːr-nɐ? correct-npst verbal group “Correct?”

T1

4. tʰɐːr-nɐ correct-npst verbal group “Correct.”

The verbal groups in (9) are analyzed in (10) and (11). The function of the lexical verb is here referred to as State. When the non-eventive verbal group is negative, a Negator is inserted and is realized at the final position as in (10); when it is positive, there is no function that explicitly realizes the feature as in (11). (10) tʰɐːr-x

uɛ

correct-npst

neg

State

Negator

‘not correct’

Axial Argumentation Below the Clause

45

(11) tʰɐːr-nɐ correct-npst State ‘correct’

To summarize, an elaborated verbal group selects either [positive] or [negative]; when it selects [positive], there is no overt realization. When it selects [negative], a Negator function is inserted and is realized at the end of the verbal group.

4.1.2 VG TENSE Vg tense in Khorchin Mongolian grammaticalizes the sequencing of figures in relation to the speech time. The choices in this system are [past] and [non-past]. The feature [non-past] construes a figure as either concurrent with or following the speech time (‘present’ or ‘future’); they are not structurally distinct. The exchange in (12) exemplifies the negotiation of a proposition in terms of vg tense. The two teachers in (12), T1 and T2, are discussing the mother of one of their students. This student’s sister used to go to the same school. The proposition they are arguing about has to do with the modes of transport the mother uses to take her children to school: whether it is always by motorbike or also sometimes by motorized trike in the past. The translations of any modal particles (mp) are shown in square brackets ([ ]). The tense choices in the verbal group are highlighted in bold. (12) T1

1. sɘtʃ ʰin niːtɘm Secin

mʊtʰʊ

ɛː-n

kər jɐp-tʃ

xʊi

always motorbike ins commute-prog cop-NPST mp verbal group

“[I saw that] Secin is always commuting by motorbike.” 2. tʰɘr dist

ixin

ɘn

ʊrtʰɛ

sɐnlʊ kər jɐp-tʃ

ɛː-tʃ

daughter poss before trike ins commute-prog cop-PST mp verbal group

“[You know that] her daughter was commuting (to school) by trike before.” T2

3. ukuɛ neg

ʃɛ

sɘtʃʰin

niːtɘm

Secin

always motorbike ins

mʊtʰʊ

kər

jɐp-tʃ

ɛː-nɐ

commute-prog cop-NPST verbal group

“No, Secin is always commuting by motorbike.”

46

Dongbing Zhang T1

4. ukuɛ neg

sɐnlʊ

kər

pɐs

trike

ins

also

jɐp-tʃ

ɛː-tʃ

ʃʊ

commute-prog

cop-PST

mp

verbal group “No, [I know that] (she was) also commuting by trike.” T2

5. ɔː Oh “Oh.”

The verbal groups in (12) exemplify the selections from vg tense in the eventive verbal group. They are contrasted in (13). The verbal groups in (12.1) and (12.3) are the same; they select [non-past]. The verbal groups in (12.2) and (12.4) are the same; they select [past]. Here we will use the lengthened form of the past tense suffix -tʃɛ, rather than -tʃ given in (12), to avoid confusion with the progressive aspect marker -tʃ discussed in Section 4.1.3.10 (13) from (12.1) and (12.3)

ɛː-n

jɐp-tʃ

commute-prog cop-npst ‘is commuting’ from (12.2) and (12.4)

jɐp-tʃ

ɛː-tʃɛ

commute-prog cop-pst ‘was commuting’

In contrast to the verbal groups in Section 4.1.1, where the constituents of the verbal group are a verb and an adverb, the verbal groups in (13) comprise two verbs: a lexical main verb followed by an auxiliary copular verb. This arises when there is a selection from aspect. When this is the case, the aspect choice is shown as a suffix on the lexical verb, and the vg tense choice is shown as a suffix on the following auxiliary copula. However, to simplify our discussion of vg tense here, we will focus on instances without aspect as illustrated in (14) for [nonpast] and (15) for [past]. We will come back to how this interplays with aspect in Section 4.1.3. (14) jɐp-ɘn commute-npst ‘commutes’ (15) jɐp-tʃɛ commute-pst ‘commuted’

Axial Argumentation Below the Clause

47

As these instances indicate, the distinction in vg tense is shown by suffixes that arise from distinct verb classes. We will call the verb class that realizes [non-past] in instances such as (14) ‘v-n’ and the verb class that realizes [past] in instances such as (15) ‘v-tʃɛ’. As far as the structure of the verbal group is concerned, I use the term Tense to refer to the function generated by a selection from the vg tense system. The importance of a distinct function of Tense arises from the fact that, as shown by the aspectual examples included, vg tense does not always occur on Event or State; this is only the case for certain types of verbal groups (e.g. aspectless). Tense can also occur on auxiliaries and negation adverbs (discussed later). The analyses in (16) and (17) illustrate the structures that realize the co-selection of [positive] from vg polarity and vg tense in an eventive verbal group. In this case, the Event and Tense functions are conflated. (16) jɐp-ɘn cop-npst Event/Tense ‘commutes’ (17) jɐp-tʃɛ commute-pst Event/Tense ‘commuted’

As far as vg polarity is concerned, both (16) and (17) select [positive]. As we’ve seen, when an eventive verbal group selects [negative], the Negator is realized separately by a negation adverb. However, negation also changes the verb class realizing vg tense on the lexical verb. In the following negative verbal groups, Tense in [non-past] is realized by v-x in (18); Tense in [past] is realized by v-sɘn in (19). (18) jɐp-x

uɛ

commute-npst

neg

Event/Tense

Negator

‘doesn’t commute’ (19) jɐp-sɘn

kuɛ

commute-pst

neg

Event/Tense

Negator

‘didn’t commute’

The distinctive realizations of vg tense in relation to the co-selection from vg polarity are summarized here.11

48

Dongbing Zhang positive negative non-past v-n

v-x

past

v-sɘn

v-tʃɛ

The previous examples also show that the realizations of Negator vary with the choice of vg tense. When [non-past], the Negator is lexicalized as uɛ as in (18); when [past] it is lexicalized as kuɛ, as in (19).12 Non-eventive verbal groups also co-select from vg polarity and vg tense. As the following examples show, the patterns for vg tense choices in non-eventive verbal groups are the same with that in eventive verbal groups. In a non-eventive verbal group, the co-selection of [non-past] and [positive] conflates the State and Tense functions as in (20); the co-selection of [non-past] and [negative] conflates the State and Tense functions, which are followed by Negator as in (21). (20) tʰɐːr-ɘn correct-npst State/Tense ‘correct’ (21) tʰɐːr-x

uɛ

correct-npst

neg

State/Tense

Negator

‘is not correct’

The structural configurations for [past] is the same. The co-selection of [past] and [positive] conflates State and Tense as in (22); the co-selection of [past] and [negative] conflates State and Tense, which is followed by Negator as in (23). (22) tʰɐːr-tʃɛ correct-pst State/Tense ‘was correct’ (23) tʰɐːr-sɘn

kuɛ

correct-pst

neg

State/Tense

Negator

‘was not correct’

So far, the co-selection from vg type, vg polarity and vg tense in the elaborated verbal group has been discussed and exemplified. The features in vg type [eventive] and [non-eventive] are realized

Axial Argumentation Below the Clause

49

by insertion of functions Event and State. The feature [negative] in vg  polarity is realized by insertion of a Negator function. The features in vg tense are realized by insertion of a Tense function. When the co-selections are [non-past/positive] and [past/positive], Tense is realized by v-n and v-tʃɛ respectively. When the co-selections are [nonpast/negative] and [past/negative], Tense is realized by v-x and v-sɘn respectively; at the same time, the corresponding Negator is lexicalized as uɛ and kuɛ. When the features from [progressive] in aspect, [modality] in vg modality and [future] in relative tense as they are described in the remainder of Section 4.1 are not selected, Event and State are conflated with Tense. 4.1.3 ASPECT Apart from vg polarity and vg tense, another resource shared between eventive and non-eventive verbal groups is aspect. Unlike vg polarity and vg tense, in aspect the verbal group optionally selects the feature [progressive].13 The choice of [progressive] aspect construes an occurrence as ongoing (i.e. the middle part of the ‘internal constituency’ of an occurrence). We have seen aspect at work in the eventive verbal group in (12), repeated as (24). The verbal groups in (24.1) and (24.3) co-select [progressive], [non-past], and [positive]; those in (24.2) and (24.4) co-select [progressive], [past], and [positive]. The suffixes that mark aspect and vg tense are highlighted in bold.

(24) T1 1. sɘtʃʰin niːtɘm Secin

mʊtʰʊ

ɛː-n

kər jɐp-tʃ

xʊi

always motorbike ins commute-PROG cop-PST

mp

verbal group “[I saw that] Secin is always commuting by motorbike.” 2. tʰɘr dist

ixin

ɘn

ʊrtʰɛ

sɐnlʊ kər

daughter poss before trike

ins

jɐp-tʃ

ɛː-tʃ

ʃɛ

commute-PROG cop-PST mp verbal group

“[You know that] her daughter was commuting (to school) by trike before.”

T2 3. ukuɛ sɘtʃʰin niːtɘm neg

Secin

always

mʊtʰʊ

kər

motorbike ins

jɐp-tʃ

ɛː-nɐ

commute-PROG cop-NPST verbal group

“No, Secin is always commuting by motorbike.”

50

Dongbing Zhang T1 4. ukuɛ neg

sɐnlʊ kər pɐs trike

ɛː-tʃ

jɐp-tʃ

ʃʊ

ins also commute-PROG cop-PST mp verbal group

“No, [I know that] (she was) also commuting by trike.” T2 5. ɔː Oh “Oh.”

The verbal groups in (24) together with the choices they instantiate are shown in (25). The feature [progressive] is realized through verb class v-tʃ and an auxiliary copula ɛː.14 (25) [progressive/non-past/positive]

jɐp-tʃ

ɛː-n

commute-prog cop-npst ‘is commuting’ [progressive/past/positive]

jɐp-tʃ

ɛː-tʃɛ

commute-prog cop-pst ‘was commuting’

Structurally, I use the terms Aspect and Auxiliary to refer to the functions of v-tʃ and ɛː in the verbal group. The selection of [progressive/positive] is realized by inserting Aspect and Auxiliary as shown in examples from (26) to (30). The selection of [progressive/negative] is realized differently in relation to the selection from vg tense. When [past] is selected, both Aspect and Auxiliary are required in the structure as in (31) and (32); on the other hand, when [non-past] is selected, only an Aspect function is required as in (33) and (34). The structural configuration of the co-selection [progressive/positive] and the features from vg tense in an eventive verbal group is shown in (26) and (27). The features selected from vg tense in (26) and (27) are [non-past] and [past] respectively. In both cases, Aspect is conflated with Event and Auxiliary is conflated with Tense. There is no explicit realization of [positive]. (26) jɐp-tʃ

ɛː-n

commute-prog

cop-npst

Event/Aspect

Auxiliary/Tense

‘is commuting’ (27) jɐp-tʃ

ɛː-tʃɛ

commute-prog

cop-pst

Event/Aspect

Auxiliary/Tense

‘was commuting’

Axial Argumentation Below the Clause

51

In the same way, a non-eventive verbal group can also select from aspect. This is exemplified in (28). The mother (M) asks about her daughter’s (D) physiological state because the weather is warming up. The non-eventive verbal group in (28.1) co-selects [progressive], [nonpast] and [positive]. (28) M: 1. tʃʰi

xɐlut-tʃ

ɛː-ø15

mɛ

2sg feel.hot-prog cop-npst qp verbal group “Are you feeling hot?” D:

2. ukuɛ neg “No.”

As far as the structure of the verbal group in (28.1) is concerned, State in the non-eventive verbal group is conflated with Aspect and Tense is conflated with Auxiliary. The structural configurations of the co-selection [progressive/positive] and the features from vg tense in a non-eventive verbal group are shown in (29) and (30). The features selected from vg tense in (29) and (30) are [non-past] and [past] respectively. (29) xɐlut-tʃ

ɛː-n

feel.hot-prog

cop-npst

State/Aspect

Auxiliary/Tense

‘is feeling hot’ (30) xɐlut-tʃ

ɛː-tʃɛ

feel.hot-prog

cop-pst

State/Aspect

Auxiliary/Tense

‘was feeling hot’

In contrast to [progressive/positive], the co-selection [progressive/ negative] from aspect and vg polarity generates different structures. When the co-selection is [negative/progressive/past], Aspect is conflated with Event or State and Auxiliary is conflated with Tense; the Negator generated by [negative] is lexicalized as kuɛ. This is illustrated for the eventive verbal group in (31) and the non-eventive verbal group in (32). ɛː-sɘn

kuɛ

commute-prog

cop-pst

neg

Event/Aspect

Auxiliary/Tense Negator

(31) jɐp-tʃ

‘was not commuting’

52

Dongbing Zhang ɛː-sɘn

kuɛ

feel.hot-prog

cop-pst

neg

State/Aspect

Auxiliary/Tense Negator

(32) xɐlut-tʃ

‘was not feeling hot’

In contrast, when the features selected are [negative/progressive/nonpast], no Auxiliary function is required. Aspect is conflated with Event or State. Tense on the other hand does not have an identifiable realization; it is conflated with Negator, which is lexicalized as ukuɛ. This is illustrated for the eventive verbal group in (33) and the non-eventive verbal group in (34). (33) jɐp-tʃ

ukuɛ

commute-prog

neg

Event/Aspect

Tense/Negator

‘is not commuting’ (34) xɐlut-tʃ

ukuɛ

feel.hot-prog

neg

State/Aspect

Tense/Negator

‘is not feeling hot’

Figure 2.3 formalizes the features selected from aspect, vg tense and vg polarity, along with their structural realizations. The realization statements on the right only account for selections which concern [progressive]. For the realizations of vg tense and vg polarity without the selection of [progressive], see Section 4.1.2. Note that the conflation of Aspect with Event or State is not included in the diagram. The sequencing of Event or State also varies in relation to selections from vg modality and relative tense as they are discussed in Section 4.1.4 and 4.1.5. A double colon (::) is used to show lexicalization (e.g. Negator:: ukuɛ means “Negator is lexicalized as ukuɛ”). In the remainder of Section 4.1, two further systems, vg modality and relative tense, are discussed. The two systems are only available for the eventive verbal group, and relative tense can only occur when neither aspect nor vg modality occurs. We will begin with vg modality. 4.1.4 VG MODALITY One defining feature of an eventive verbal group is that it selects from vg modality. This is the case when what is being negotiated is a proposal (i.e. goods and services). A proposal in Halliday’s sense (1994: 89) can be modulated. In Khorchin Mongolian, one of the ways to modulate a proposal is through the verbal group.16

Axial Argumentation Below the Clause

53

Figure 2.3 vg type, vg polarity, vg tense and aspect in the Elaborated Verbal Group

Two kinds of modulation are possible: [permission] and [ability]. The verbal group in (35.1) exemplifies the modulation of a proposal in terms of [permission]. The aunt (A) reminds her nephew that the way his dumpling has been made is not permitted (because the fillings will come out). Modulation in Khorchin Mongolian selects from vg tense. (35) A: 1. ŋət

pɛr-tʃ

pɔl-x

uɛ

kɔlɔi

this way make-cvb allow-npst neg dear verbal group “Dear, (you) are not allowed to make (it) this way.” 2. ɘn

pɔl

xəkə-x

uɛ

prox theme meet.the.standard-npst neg verbal group “This does not meet the standard.”

54

Dongbing Zhang

The modulation of a proposal in terms of [permission] is achieved through the modal verb pɔl ‘be allowed to’ as in (35.1). I use the term Modality to refer to the function of the modal verb in the verbal group. Structurally, the Modality function is realized after Event; Event is realized by v-tʃ.17 Modality is conflated with Tense when there is no selection from aspect. The structure of the verbal group in (35.1) is analyzed in (36). (36) pɛr-tʃ

pɔl-x

uɛ

make-cvb

allow-npst

neg

Event

Modality/Tense Negator

‘is not allowed to make’

In (36), although Modality and Tense are conflated at group rank, the realization of each function is identifiable. Modality is lexicalized as pɔl; Tense is realized by v-x. When the verbal group selects [positive], on the other hand, there is no explicit function which realizes the feature; Event is followed by the conflation of Modality and Tense as in (37). Modality is lexicalized as pɔl; Tense is realized by v-n. (37) pɛr-tʃ

pɔl-ɘn

make-cvb

allow-npst

Event

Modality/Tense

‘is allowed to make’

Similarly, an eventive verbal group can realize the modulation of a proposal in terms of [ability] through the modal verb jɔl ‘be able to.’ The exchange in (38) is an excerpt from the negotiation of shifts between teachers. The teacher states her preference for the evening shift. (38) 1.

pi

pɔl

urlə

ir-tʃ

jɔl-x

uɛ

1sg theme morning come-cvb be.able.to-npst neg verbal group “I am not able to come in the morning.” 2.

œrœn

tʃitʃur pɔl

evening shift

jɔl-ɘn

cond be.able.to-npst verbal group

“If (it is) evening shift, (I) am able to (come).”

The verbal groups in (38.1) and (38.2) illustrate the co-selections of [ability/non-past/negative] and [ability/non-past/positive] respectively. Note that in (38.2) the verb that realizes Event, ir-tʃ ‘come-cvb’, is elided.

Axial Argumentation Below the Clause

55

Like modulation of [permission], Modality in (38) is realized after the Event. It is conflated with Tense when [progressive] in the aspect system is not selected. The structure of the verbal group in (38.1) and (38.2) are analyzed in (39) and (40). (39) ir-tʃ

jɔl-x

uɛ

come-cvb

be.able.to-npst neg

Event

Modality/Tense Negator

‘am not able to come’ (40) ir-tʃ

jɔl-ɘn

come-cvb

be.able.to-npst

Event

Modality/Tense

‘am able to come’

In (39) and (40), although Modality and Tense are conflated, the realization of each function is identifiable. Modality is lexicalized as the modal verb jɔl; Tense is realized by v-x when the verbal group selects [negative] as in (39) and by v-n when the verbal group selects [positive] as in (40). Modality also co-selects with aspect. Though it is not attested in my data, instances such as (41) are not rare. For example, it could be something uttered when the wellbeing of a patient is asked. (41) tʰɘr 3sg

pɐtɐ it-ɘtʃ jɔl-tʃ meal eat-cvb be.able.to-prog

ɛː-n cop-npst

verbal group Event

Modality/Aspect Auxiliary/Tense

“He is being able to eat meals.”

In (41), Event is lexicalized as it ‘eat.’ Modality is conflated with Aspect, and Auxiliary is conflated with Tense. The interaction between vg modality, aspect and vg polarity follows the pattern of interaction between aspect and vg polarity as it is discussed in Section 4.1.3. Example (42) illustrates [ability/progressive/ non-past/negative]. (42) tʰɘr 3sg

pɐtɐ itɘ-tʃ jɔl-tʃ meal eat-cvb be.able.to-prog

ukuɛ neg

verbal group Event

Modality/Aspect Tense/Negator

“He is not being able to eat meals.”

56

Dongbing Zhang

In (42), Modality is conflated with Aspect; Tense is conflated with Negator; the conflation of Tense and Negator follows the conflation of Modality and Aspect. Modality is lexicalized as the modal verb jɔl; Aspect is realized by v-tʃ and Negator by ukuɛ. Tense does not have an identifiable realization; it is conflated with Negator. The pattern is different when the verbal group selects [past]. In such cases, Modality is conflated with Aspect; Auxiliary is conflated with Tense; Negator is realized at the final position. The instance in (43) exemplifies the selection [ability/progressive/past/negative]. (43) tʰɘr 3sg

pɐtɐ itɘ-tʃ jɔl-tʃ meal eat-cvb be.able.to-prog

ɛː-sɘn cop-pst

kuɛ neg

verbal group Event

Modality/Aspect Auxiliary/Tense Negator

“He was not being able to eat meals.”

In (43), Modality is lexicalized as the modal verb jɔl; Aspect is realized by v-tʃ, and Tense by v-sɘn; Negator is lexicalized as kuɛ. To summarize, an eventive verbal group optionally selects [modulated]; when the feature is selected, a Modality function is inserted. Modality is either lexicalized as pɔl: ‘be allowed to’ or jɔl: ‘be able to,’ which respectively realize features [permission] and [ability]. When [progressive] from aspect is not selected, Modality is realized after Event and is conflated with Tense. When [progressive] is selected, on the other hand, Modality is realized after Event and is conflated with Aspect.

4.1.5 RELATIVE TENSE Another defining feature of an eventive verbal group is that it selects from relative tense. Relative tense can only occur when there is no selection of [progressive] from aspect or [modulated] from vg modality. It allows the construal of a relative [future] in relation to the ‘absolute time’ construed by vg tense, which deals with temporality in relation to the time of a speech event (i.e. [past] or [non-past]). The move (44.3) exemplifies the co-selection of [future/non-past]. The exchange is part of a conversation between a wife (W) and her husband (H) on their phone through WeChat. The wife is waiting for her husband to come back for dinner. She asks what time he is coming home in (44.1). The husband states that he is going to be awhile since the carwash is washing other people’s cars in (44.2). He then reassures his wife by telling her that their car is going to be washed in (44.3) and asks her to keep waiting in (44.4).

Axial Argumentation Below the Clause (44) W: 1. pɐs nɘktʃʰi also awhile

ʊt-ø last-npst

57

mɛ tʃʰi qp 2sg

verbal group “Are you going to be awhile?” H. 2. uər xun nɛ tʰɘrɘk iː ʊɐ-tʃ ɛː-nɐ other people gen car acc wash-prog cop -nspt verbal group “(They) are washing other people’s cars.” 3. ɔtɔ mɐn nɛ ki ʊɐ-ntʃ ɛː-nɐ now 1pl gen acc wash-fut cop-npst verbal group “(They) are going to wash ours now.” 4. xulɛ-tʃ

ɛː

wait-prog cop.imp.2 verbal group “Stay waiting.”

The verbal group in (44.2) ʊɐ-tʃ ɛː-nɐ ‘wash-prog cop-npst’ co-selects [progressive], [non-past] and [positive] as discussed in Section 4.1.3; the verbal group in (44.4) xuliɛ-tʃ ɛː ‘wait-prog cop.imp.2’ functions in an imperative clause, which is described in Section 4.2. In (44.3), the verbal group construes an occurrence that is going to happen after the speech time. In terms of ideation in discourse (Hao 2015, 2019; Martin 1992; Martin and Quiroz in this volume; Martin and Rose 2007) the co-selection from relative tense (i.e. [future]) and from vg tense (i.e. [non-past]) sequences a figure at a ‘future’ time in relation to the ‘present’ established in the discourse. Therefore, the features selected from the two systems are named as [‘relative tense’ in ‘absolute tense’]—[future in past] and [future in non-past] (following Halliday 1976, 1985 [1994]). As illustrated in (45), the feature [future] in relative tense is realized by the verb class v-ntʃ and an auxiliary copula ɛː; the selection from vg tense ([non-past] in this instance) is realized on the auxiliary copula. (45) mɐn 1pl

nɛ ki ʊɐ-ntʃ ɛː-nɐ gen acc wash-fut cop-npst verbal group

“(They) are going to wash ours.”

58

Dongbing Zhang

The reference time can also be [past] (i.e. [future in past]). The example (45) is adapted as (46) to exemplify the positioning of a figure at a ‘future’ time in reference to the ‘past.’ (46) mɐn nɛ 1pl

ʊɐ-ntʃ

ki

ɛː-tʃɛ

gen acc wash-fut cop-pst verbal group

“(They) were going to wash ours.”

Brosig (2014) treats the syntagm v-ntʃ^ɛː as a marker of ‘prospective’ aspect. His position is not adopted on the following two grounds: first, the temporal meaning construed is not concerned with the “internal temporal constituency” of a situation (Comrie 1976); rather the meaning is concerned with the positioning of a figure as a whole in relation to another; second, the meaning construed by v-ntʃ^ɛː has rather different valeur than the choice [progressive] in the system of aspect as it is discussed in Section 4.1.3. The choice of [progressive] is not restricted to the vg type of verbal group ([eventive] or [non-eventive]); nor is it restricted to the vg deixis of the verbal group ([elaborated] or [restricted]).18 Furthermore, [progressive] can also co-select [modulated]. In contrast, relative [future] is a choice possible only for an eventive elaborated verbal group, and is mutually exclusive with vg modality. The structures that realize [future in past] and [future in non-past] are exemplified in (47) and (48). The functions of v-ntʃ and ɛː are referred to as Relative Tense (R.Tense) and Auxiliary respectively. In this way, when an eventive verbal group selects [future] from relative tense and [positive] from vg polarity, two functions are inserted: R.Tense and Auxiliary. R.Tense is realized by v-ntʃ and is conflated with Event; Auxiliary is lexicalized as ɛː and is conflated with Tense. (47) ʊɐ-ntʃ wash-fut

ɛː-tʃɛ cop-pst

Event/R.Tense Auxiliary/Tense ‘was going to wash’ (48) ʊɐ-ntʃ wash-fut

ɛː-nɐ cop-npst

Event/R.Tense Auxiliary/Tense ‘is going to wash’

Similar to the interaction between aspect and vg polarity, the coselection of [future in non-past] and [negative] deserves special attention.

Axial Argumentation Below the Clause

59

When the choice is [future in non-past/negative], only R.Tense is required for the realization of [future]. There is no identifiable realization of Tense; it is conflated with Negator, which is lexicalized as ukuɛ. The choice [future in non-past/negative] is exemplified in (49). (49) ʊɐ-ntʃ wash-fut

ukuɛ neg

Event/R.Tense Tense/Negator ‘is not going to wash’

On the other hand, when the co-selection is between [future in past] and [negative], an Auxiliary function is required and is conflated with Tense. This co-selection is exemplified in (50). (50) ʊɐ-ntʃ wash-fut

ɛː-sɘn

kuɛ

cop-pst

neg

Event/R.Tense Auxiliary/Tense Negator ‘was not going to wash’

In (50), Event is conflated with R.Tense, realised by v-ntʃ; Auxiliary is conflated with Tense, realized by v-sɘn; Negator is lexicalized as kuɛ, following Tense. The structures which realize the interaction between relative tense, vg tense, and vg polarity are summarized here. The feature [future] in relative tense in an eventive verbal group is an optional choice. When [future] is selected we find the following possibilities: [future] ↘ + R.Tense; R.Tense: v-ntʃ [positive] ↘ + Auxiliary; R.Tense^Auxiliary; Auxiliary:: ɛː; Auxiliary/Tense [negative] ↘ + Negator; Negator^# [future in past/negative] ↘ + Auxiliary; R.Tense^Auxiliary; Auxiliary:: ɛː; Auxiliary/Tense; Negator:: kuɛ [future in non-past/negative] ↘ R.Tense^Tense; Tense/Negator; Negator:: ukuɛ

4.2 The Restricted Verbal Group Restricted verbal groups function in imperative clauses. A restricted verbal group obligatorily selects from the system vg person and optionally selects [progressive] from the system aspect. Both eventive and noneventive verbal groups select from these two systems. Section 4.2.1 and 4.2.2 describes vg person and aspect respectively. In each section, eventive verbal groups are discussed first, followed by discussions of noneventive verbal groups.

60

Dongbing Zhang

4.2.1 VG PERSON The system vg person has three features: [first person], [second person] and [third person]. A restricted verbal group realizes the Predicator in an imperative clause (see Section 2) and obligatorily chooses from one of the three features. In terms of discourse, the realization of Predicator has to do with the modally responsible participant (i.e. the participant that is responsible for the provision of the goods and services). When the modally responsible participant includes the speaker (i.e. speaker inclusive), the Predicator of the imperative clause is realized by a verbal group that selects [first person]. When the modally responsible participant does not include the speaker (i.e. speaker exclusive), the Predicator of the imperative clause is realized by a verbal group which selects [second person]. In contrast, when the verbal group realization of the Predicator selects [third person], the modally responsible participant may either include or exclude the speaker. The exchange in (51) (repeating (2) earlier) exemplifies the role of the eventive verbal group in both speaker exclusive and speaker inclusive imperative clauses. The imperative clause in (51.1) positions the addressee as the modally responsible participant, i.e. speaker exclusive; the imperative clause in (51.3) positions the speaker as the modally responsible participant, i.e. speaker inclusive (G = grandmother; H = Hairhan, the granddaughter). The clause in (51.2) is not an imperative clause, and thus is not included in the analysis. Note that the name xɛːrxɐn in (51.1) is a Vocative realized on a separate tone group. (51) G

1. xɛːrxɐn itə Hairhan eat.imp.2 Predicator verbal group “Hairhan, eat.”

H 2. ɔː

ukuɛ nɛi

oh neg

it-x

uɛ

lɛ

grandma eat-npst neg res

“Oh, no, grandma, I won’t eat.” G

3. tʃʰɐmɐ t 2sg

ʃiɔʃiɔ

nɛ

uk-jə

dat small gen give-imp.1 Predicator verbal group

“Let (me) give you the smaller ones.” H 4. [Accepts and eats the orange.]

Axial Argumentation Below the Clause

61

The clause in (51.1) exemplifies one type of speaker exclusive imperative clause. It positions the addressee as modally responsible without any explicit affixation in the verbal group. In contrast, the speaker inclusive imperative clause in (51.3) uses the verb class v-j to realize the modally responsible participant in the Predicator. The verbal group realizations of Predicator in imperative clauses agree with the pronominal realization of the ‘actor.’ Examples (52) and (53) illustrate speaker inclusive imperative clauses with explicit pronominal realization of the participant performing the action of giving (i.e. ‘speaker only’ in (52) and ‘speaker and non-interlocutor’ in (53)). (52) pi 1sg

tʃʰɐmɐ t

ʃiɔʃiɔ

2sg

small gen give-imp.1

dat

nɛ

uk-jə Predicator verbal group

“Let me give you the smaller ones.” (53) pɛtɘn tʃʰɐmɐ t 1pl

2sg

ʃiɔʃiɔ

nɛ

uk-jə

dat small gen give-imp.1 Predicator verbal group

“Let us give you the smaller ones.”

The suffix -jə in the verbal groups ‘agrees with’ the first person pronouns: first person singular in (52) and first person plural in (53). This is why the verbal group option is called [first person]. In the same way, the second person pronoun in the speaker exclusive imperative clause exemplified in (51.1) can be made explicit. This is illustrated in (54) and (55). (54) tʃʰi itə 2sg eat.imp.2 Predicator verbal group “You eat.” (55) tʃʰɛtɘn itə 2pl eat.imp.2 Predicator verbal group “You eat.”

62

Dongbing Zhang

The non-inflected verbs in the verbal groups agree with the second person pronouns: second person singular in (54) and second person plural in (55). This is why the verbal group option is called [second person]. The third type of imperative clause in Khorchin Mongolian is ambiguous as to whether the modally responsible participant includes or excludes the speaker. Unlike the speaker inclusive and the speaker exclusive imperative clauses, this type is used to permit the occurrence of the action realized in the verbal group. For a detailed account of this type of imperative clause in relation to the other two, see Zhang (2020). In (56) the sister informs her brother what she is going to do. The move in (56.1) positions her brother as modally responsible for permitting the boiling of the pot.19 ɛː-kɐ cop-imp.3

(56) 1. tʰɔkɔ pœʃɘl-tʃ pot boil-prog Predicator

verbal group “Let the pot boil,” 2. pi kɐxɐ ki tʰitʃə-kət 1sg pig acc feed-pfv

ir-jə come-imp.1

Predicator verbal group verbal group “Let me feed the pigs and come back.”

In (56.1), Predicator is realized by a verbal group with the final component realized by the verb class v-k. The ‘actor’ tʰɔkɔ cannot be replaced by either the first person or the second person pronouns. It can be replaced by the third person pronouns as in (57) and (58). (57) tʰɘr

pœʃɘl-tʃ

3sg boil-prog

ɛː-kɐ cop-imp.3

Predicator verbal group “Let it boil.” (58) tʰɘtɘn pœʃɘl-tʃ 3pl

boil-prog

ɛː-kɐ cop-imp.3

Predicator verbal group “Let them boil.”

The suffix -kɐ in the verbal groups ‘agrees with’ the third person pronouns: third person singular in (57) and third person plural in (58). This is why the verbal group option is called [third person].

Axial Argumentation Below the Clause

63

Structurally speaking, we will use the term Person to refer to the function realizing choices from vg person. In eventive verbal groups as exemplified so far, Event is conflated with Person. The selections of [first person], [second person] and [third person] in an eventive verbal group are illustrated in (59), (60) and (61) respectively via the verb uk ‘give.’ (59) uk-jə give-imp.1 Event/Person ‘(Let me) give’ (60) uk give.imp.2 Event/Person ‘(you) give’ (61) uk-ɘk give-imp.3 Event/Person ‘(Let her) give’

Table 2.1 shows the correspondence between the different types of imperative clause and the verb classes involved in the verbal group which realizes the Predicator in each type. The verb it ‘eat’ is used for exemplification. Note that when there is no selection of [progressive] from the aspect system discussed in Section 4.2.2, the verbal group comprises only one verb. When [progressive] is selected a combination of verbs constitute the verbal group. In other words, these ‘verbs’ are described as verbal groups because of their potential selection from different verbal group systems. Table 2.1 Imperative Clauses and the Realizations of the Predicator clause rank feature in MOOD

group rank feature in VG PERSON

word rank verb class

examples

v-j v v-k

it-jə it it-ɘk

realizing the Predicator speaker inclusive speaker exclusive speaker inclusive or exclusive

first person second person third person

The non-eventive verbal group interaction with vg person is similar to that of the eventive verbal group. Although there is no instance of a non-eventive verbal group selecting vg person in my corpus, instances such as (62) are acceptable. It is commonly used from parents to children

64

Dongbing Zhang

when the children’s behavior needs to be regulated. In a non-eventive verbal group, State conflates with Person. (62) tʃukər quiet

pɛː cop.imp.2 verbal group State/Person

“(You) be quiet.”

In (62), the verbal group selects [second person]. The addressee, e.g. the children, is positioned as modally responsible for being quiet. The adjective tʃukər in (62) can in fact be used in different senses in relation to the different choices speakers make from the verbal group system vg person. When the verbal group selects [first person] as in (63), the adjective tʃukər means ‘stay out of.’ (63) tʃukər

pɛː-jɐ

stay.out.of

cop-imp.1 verbal group State/Person

“Let me stay out of (this).”

In (63), the clause positions the speaker as modally responsible for staying out of the situation. When the verbal group selects [third person] as in (64), the adjective tʃukər is best interpreted as ‘undisturbed’. (64) tʰɘr 3sg

tʃukər

pɛː-kɐ

undisturbed cop-imp.3 verbal group State/Person

“Let him be undisturbed. (= Leave him alone.)”

In (64), the clause positions the addressee as modally responsible for permitting a non-interlocutor to stay undisturbed. To summarize, the structures generated by the co-selections from vg type and vg person are listed here: [eventive] ↘ + Event [non-eventive] ↘ + State [restricted] ↘ + Person [first person] ↘ Person: v-j [second person] ↘ Person: v

Axial Argumentation Below the Clause

65

[third person] ↘ Person: v-k [eventive/restricted] ↘ Event/Person [non-eventive/restricted] ↘ State/Person

4.2.2 ASPECT As with elaborated verbal groups, restricted verbal groups optionally select [progressive] from the aspect system. We have seen the co-selection of aspect and vg person in an eventive verbal group in (56), repeated as (65). The verbal group in (65.1) co-selects [progressive] and [third person]. The Event and Aspect functions are conflated and the Auxiliary and Person functions are conflated. (65) 1.

tʰɔkɔ pœʃɘl-tʃ

ɛː-kɐ

pot

cop-imp.3

boil-prog verbal group

Event/Aspect Auxiliary/Person “Let the pot boil,” 2.

pi

kɐxɐ ki

1sg pig

tʰitʃə-kət

acc feed-pfv

ir-jə come-imp.1

verbal group verbal group “Let me feed the pig and come back.”

In (65.1) Event is lexicalized as pœʃɘl; Aspect is realized by v-tʃ; Auxiliary is lexicalized as ɛː; and Person is realized by v-k. In (65.2), we have a verbal group complex. It is adjusted as (66) to exemplify the co-selection of [progressive] and [first person] in a single verbal group.20 (66) pi 1sg

kɐxɐ ki pig

tʰitʃə-tʃ

acc feed-prog

ɛː-jɐ cop-imp.1

verbal group Event/Aspect Auxiliary/Person “Let me feed the pigs.”

Similarly, with the co-selection of [progressive] and [third person], the choices from each system have distinct realizations. Event is lexicalized as tʰitʃə; Aspect is realized by v-tʃ; Auxiliary is lexicalized as ɛː; and Person is realized by v-j. The clause in (66) is adjusted as (67) to exemplify the co-selection of [progressive] and [second person]. Event is conflated with Aspect and Auxiliary is conflated with Person.

66

Dongbing Zhang

(67) tʃʰi

kɐxɐ ki

2sg pig

ɛː

tʰitʃə-tʃ

acc feed-prog

cop.imp.2

verbal group Event/Aspect Auxiliary/Person “You feed the pigs.”

In (67), Event is lexicalized as tʰitʃə; Aspect is realized by v-tʃ; Auxiliary is lexicalized as ɛː; and Person is realized by v. As far as non-eventive verbal groups are concerned, their selections from aspect and vg person resemble that of eventive verbal groups. The verbal group in (68) exemplifies the co-selection of [progressive] and [second person] in a non-eventive verbal group. The speaker requests the addressee to stay wherever she is. Structurally, State is conflated with Aspect and Auxiliary is conflated with Person. (68) tʃʰi

tʰɘntər pɛː-tʃ

2sg there

cop-prog

ɛː cop.imp.2

verbal group State/Aspect Auxiliary/Person “You be there.”

The clause in (68) positions the addressee as modally responsible for staying at a place for a prolonged period of time. In the verbal group, State is lexicalized as the copular verb pɛː; Aspect is realized by v-tʃ; Auxiliary is lexicalized as ɛː; and Person is realized by v. The clauses in (69) and (70) adjust that in (68) to exemplify the coselection of [progressive] and [first person], and [progressive] and [third person] in a non-eventive verbal group. The modal responsibility of staying at a location is assigned to the speaker and the addressee in (69); the assignment of the modal responsibility of permitting a non-interlocutor’s staying at a location is ambiguous in (70). (69) pɛtɘn tʰɘntər pɛː-tʃ 1pl

there

cop-prog

ɛː-jɐ cop-imp.1

verbal group State/Aspect Auxiliary/Person “Let’s be there.” (70) tʰɘr 3sg

tʰɘntər pɛː-tʃ

ɛː-kɐ

there

cop-imp.3

cop-prog verbal group

State/Aspect Auxiliary/Person “Let her be there.”

Axial Argumentation Below the Clause

67

In the verbal groups in (69) and (70), States are lexicalized as the copular verb pɛː; Aspects are realized by v-tʃ; Auxiliaries are lexicalized as ɛː; and Person is realized by v-j in (69) and v-k in (70). The structures that realize the co-selection from vg person and [progressive] in aspect in the restricted verbal group are summarized here. [restricted] ↘ + Person [first person] ↘ Person: v-j [second person] ↘ Person: v [third person] ↘ Person: v-k [eventive] ↘ + Event [non-eventive] ↘ + State [progressive] ↘ + Aspect; + Auxiliary; Aspect: v-tʃ; Auxiliary:: ɛː [progressive/eventive/restricted] ↘ Event/Aspect; Auxiliary/Person [progressive/non-eventive/restricted] ↘ State/Aspect; Auxiliary/Person

5. Conclusions This chapter provides a systemic functional account of the verbal group in Khorchin Mongolian. Two basic verbal group systems, vg deixis and vg type, are established based on the role the verbal group plays in a clause. The vg deixis system includes the features [elaborated] and [restricted]. Elaborated verbal groups function in indicative clauses, whereas restricted verbal groups function in imperative clauses. The system vg type includes the features [eventive] and [non-eventive]. Eventive verbal groups function in non-relational clauses, whereas non-eventive verbal groups function in relational clauses. The systems vg deixis and vg type are simultaneous co-selecting systems. An eventive elaborated verbal group selects from the systems vg polarity, vg tense, aspect, vg modality and relative tense; a non-eventive elaborated verbal group selects from the systems vg polarity, vg tense and aspect. Both eventive and non-eventive restricted verbal groups select from the systems vg person and aspect. These systemic relations are summarized as a system network in Figure 2.4. The structures that realize the co-selections from these systems are excluded to avoid repetition. They can be found in their respective sections. The interactions between the systems in Figure 2.4 are based on axial argumentation at ranks below the clause. In Section 2 and Section 3, the paradigmatic relations in the systems vg deixis and vg type are established from above in relation to the role the verbal group plays in a Khorchin Mongolian clause. In Section 4, the systems of vg polarity, vg tense, aspect, vg modality, relative tense and vg person are motivated from below with respect to the structural patterns in the syntagmatic organization of the verbal group and from roundabout with respect to the interdependencies between systems.

68

Dongbing Zhang

Figure 2.4 The Verbal Group in Khorchin Mongolian

In terms of the syntagmatic organization of the verbal group, conflations of function structures are frequently observed in Section 4 (e.g. Modality/Aspect, Auxiliary/Tense). One of the explanations is that for the features realized by these structures the “division of the grammatical labour” is at group rank, but the “location of the grammatical labour” is at word rank (Matthiessen 2015). For example, systems such as vg tense and vg person interact with the other systems at group rank. However, the distinctions in these systems are established at word rank through suffixes (-n and -x for [non-past], and -tʃɛ and -sɘn for [past] in the vg tense system; -j for [first person] and -k for [third person] in the vg person system). Similar conflation of function structures due to the distribution of the division and the location of the grammatical labor across rank scale is also observed between clause, group, and word rank in other languages (e.g. Spanish realization of the participant roles through pronominal clitics at group rank and inflectional morphology at word rank, see Martin et al. in prep; Quiroz 2017, this volume).

Axial Argumentation Below the Clause

69

In terms of the paradigmatic organization of the verbal group, the system network usefully shows the valeur of a specific grammatical category. For example, the categories that have previously been described as ‘aspect’ markers (e.g. Brosig 2014) are described here as [progressive] and [future] in two separate systems, aspect and relative tense. The system network in Figure 2.4 shows that [progressive] and [future] interact with the other systems in different ways. The choice [progressive] is not restricted to the vg type of verbal group ([eventive] or [non-eventive]); nor is it restricted to the vg deixis of the verbal group ([elaborated] or [restricted]); [progressive] also co-selects [modulation]. In contrast, relative [future] is a choice possible only for the elaborated eventive verbal group; and it is mutually exclusive with vg modality. This chapter also attests to the usefulness of starting the description of lower ranking units from above drawing on resources in the higher ranking units (e.g. mood and transitivity) or resources in the more abstract strata (e.g. exchange structure and ideation in discourse semantics). This approach sheds light on issues that await further exploration. One area of description in Khorchin Mongolian that needs further attention is verbal group complexes, which we encountered in (65), pi kɐxɐ-ki tʰitʃə-kət ir-jə ‘1sg pig-acc feed-pfv come-imp.1.’ This phenomenon needs to be accounted for in relation to the logical meaning the verbal group realizes in clause complexing. Following the line of reasoning adopted in this chapter, the system of connexion21 developed in SFL work on discourse semantics (Hao 2018; Martin 1992; Martin and Quiroz this volume; Martin and Rose 2007) seems like a promising starting point.

Acknowledgements I am grateful for the constructive and stimulating feedback from the editors on the earlier drafts of this chapter. I also would like to acknowledge the tremendous help on my writing from A/Prof. Susan Hood. Finally, this research would not have been possible without the generous consent from my informants for recording their conversations during my field trip. They are teachers from the Mongolian Sector at Hʉgjiltʉ Primary School and Baoxiang’s family from Guóyíng Mùchǎng (国营牧场), and Nara’s family from Aldartu Somu (ālādá’ěrtǔ 阿拉达尔吐) at Jalaid Banner, Hinggan League in Inner Mongolia.

Notes 1. Khorchin Mongolian is a variety of Mongolian spoken in the eastern part of Inner Mongolia, China. The variety under investigation is the language of the Mongols living in Jalaid Banner, Hinggan League. 2. vg = verbal group. This abbreviation is used in the front of some system names to distinguish them from (1) systems in a preselecting relationship

70

3. 4.

5.

6. 7.

8.

9. 10. 11.

12.

Dongbing Zhang from clause or word rank (e.g. polarity, tense); (2) systems available to other group rank classes (e.g. deixis and person for nominal groups). The group rank is assumed in Section 2 and Section 3. Its motivation is explained in more detail in Section 4.1 when more specific systems are discussed. In Quiroz (2013), the Spanish verbal group system deixis has two terms: [restricted] and [unrestricted]. The former functions in imperative clauses while the latter in indicative ones. The type of verbal group in Khorchin Mongolian that functions in an indicative clause is called ‘elaborated’ rather than ‘unrestricted’ because its potential is not unrestricted, i.e. it cannot choose from vg person (cf. Spanish unrestricted verbal group selects from person). The examples in this chapter consist of at least three lines: (1) phonemic transcription, (2) morpheme-by-morpheme glossing and (3) idiomatic translation. The abbreviations used in the glossing are: 1 = first person; 2 = second person; 3 = third person; abl = ablative; cond = conditional; cop = copula; cvb = converbal; dist = distal; fut = future; imp = imperative; ins = instrumental; mp = modal particle; neg = negation; npst = non-past; pfv = perfective; prox = proximal; pst = past; qp = question particle; res = resultative. ‘Case marking’ has been described as nominal suffixes in the literature (e.g. Bayancogtu 2002). They are treated as lexis from the perspective of systemic functional theory. For a discussion of this see Zhang (2020). Although at this primary delicacy the two systems are in a one-to-one relationship, they have different valeur, as they enable selections from different systems. For example, at the clause rank the further choices in [indicative] are [informative] and [interrogative], whereas at the group rank the further choices in [elaborated] are from the systems vg tense and vg polarity; and there is by no means a one-to-one relationship between these more delicate choices across ranks. For a detailed account of the mood system in Khorchin Mongolian, see Zhang (2020). The words kuɛ and uɛ are allomorphs of ukuɛ. Ukuɛ is described as a particle in Nasunbayar et al. (1982: 410–412). But the full form can function independently in a turn; so it is not considered a particle here. For example,—O: tʰɐn ɐs ɐsu-sɘn mɛ “Did (they) ask you?”—P: ukuɛ “No.” In contrast, kuɛ and uɛ are described as suffixes in Bayancogtu (2002: 290–296). But the counter-expectant particle pɐs ‘even’ can be realized between the verb and the negationː ɐsu-sɘn pɐs ukuɛ ‘ask-pst even neg’; so it is not considered a suffix here. It might be more accurate to consider it a clitic, a category between word and suffix. However, this category needs careful examination in relation to delicacy and rank scale. I will privilege its potential to function in a response move on its own and consider it a word. Note that this exponent of non-past is different from the ones in the next verbal group and the one in (9.2); for a discussion of these see Section 4.1.2. The suffixes -tʃɛ (~ -tʃ) for past tense and -tʃ for progressive aspect are not allomorphs; they have different valeur (also see Section 4.1.3 and 4.2.2). Due to the constraint of space, this chapter restricts its discussion of the realization of tense to the ‘non-assessed’ declarative clauses only. When the interrogative particles mɛ and iː, the exclamative particle iː, and a small amount of modal particles ( ʃɘmu, wanjaŋ, ʃidə) are used after the verbal group, the realization of Tense is v-x for [non-past] and v-sɘn for [past]. For a detailed discussion of the preselection of the mood features from verb features, see Zhang (2020). Phonologically speaking, the different lexicalizations of Negator are sensitive to the pattern of syllabification in Khorchin Mongolian. When the vg tense

Axial Argumentation Below the Clause

13.

14.

15. 16. 17. 18. 19. 20. 21.

71

is [non-past], the tense suffix -x forms a syllable with the negation adverb; when the vg tense is [past], the tense suffix -sɘn and the negation adverb forms separate syllables. This chapter does not deal with what Brosig (2014) terms ‘neutral’ and ‘continuative-resultative’ aspects. More research needs to be done with respect to rank scale and complexing of units to account for these two types systemically. Bayancogtu (2002: 335) treats this syntagm as an affix, -tʃɛː-. However, particles such as il can be realized between -tʃ and ɛː to construe a repetitive continuous occurrence, e.g. jab-tʃ il ɛː-na ‘keeps walking (back and forth),’ which means that -tʃ and ɛː are separate morphemes. The [non-past] is realized by -ø due to its interaction with the question particle mɛ. In contrast, propositions are modalized mainly through adverbial groups or clause final particles at clause rank. The suffix -tʃ is traditionally glossed cvb ‘converbal’; this will not be problematized here. The selection from aspect in the restricted verbal group is discussed in Section 4.2. The same structure can also be used to position both the speaker and the addressee as modally responsible for permitting the occurrence of an action. For exemplifications and discussions on this point, see Zhang (2020). To account for the perfective in (65.2), we need to introduce the way verbal group complex works in Khorchin Mongolian, which needs another occasion. This system is called conjunction in Martin (1992).

References Bayancogtu. 2002. horcin aman ayalgun u sodulul (A study of Khorchin dialect). Hohhot: Inner Mongolia University Press. Berry, Margaret. 1981. Systemic linguistics and discourse analysis: A multilayered approach to exchange structure. In Malcolm Coulthard and Martin Montgomery (eds.), Studies in discourse analysis, 120–145. London: Routledge & Kegan Paul. Brosig, Benjamin. 2014. The tense-aspect system of Khorchin Mongolian. In Pirkko Suihkonen and Lindsay J. Whaley (eds.), On diversity and complexity of languages spoken in Europe and North and Central Asia, 3–65. Amsterdam/ Philadelphia: John Benjamins. Comrie, Bernard. 1976. Aspect: An introduction to the study of verbal aspect and related problems. Cambridge: Cambridge University Press. Halliday, M.A.K. 1976. The English verbal group. In Gunther R. Kress (ed.), Halliday: System and function in language, 136–158. Oxford: Oxford University Press. Halliday, M.A.K. 1985. An introduction to functional grammar. London: Arnold. Halliday, M.A.K. 1994. An introduction to functional grammar (2nd edn.). London: Arnold. Halliday, M.A.K. 2009. Methods—techniques—problems. In M. A.K. Halliday and Jonathan Webster (eds.), Continuum companion to systemic functional linguistics, 59–86. London: Continuum. Hao, Jing. 2015. Construing biology: An ideational perspective. Sydney: The University of Sydney.

72

Dongbing Zhang

Hao, Jing. 2018. Reconsidering “cause inside the clause” in scientific discourse— From a discourse semantic perspective in systemic functional linguistics. Text & Talk 38(5). 525–550. doi:10.1515/text-2018-0013. Hao, Jing. 2019. Analysing scientific discourse from a systemic functional perspective: A framework for exploring knowledge building in biology. London: Routledge. Martin, J.R. 1992. English text: System and structure. Amsterdam/Philadelphia: John Benjamins. Martin, J.R. 2013. Systemic functional grammar: A next step into the theory— Axial relations. Beijing: Higher Education Press. Martin, J.R. 2018. Introduction. Interpersonal meaning: Systemic functional linguistics perspective. Functions of Language 25(1). 2–19. doi:10.1075/fol.17018. mar. Martin, J.R., Beatriz Quiroz, Pin Wang and Yongsheng Zhu. in prep. Systemic functional grammar: Another step into the theory—grammatical description. Beijing: Higher Education Press. Martin, J.R. and David Rose. 2007. Working with discourse (2nd edn.). London/ New York: Continuum. Matthiessen, C.M.I.M. 2015. Systemic functional morphology: The lexicogrammar of the word. In Edson Rosa Francisco de Souza (ed.), Estudos de descri̧ cão funcionalista: Objetos e abordagens, 150–199. Muenchen: Lincom GmbH. Nasunbayar, Haserdeni, Tʉrgen, Cogtu, Tawadagba, Naranbatu and Secen. 1982. orcin cag un monggul hele (Modern Mongolian). Hohhot: Inner Mongolia Education Press. Quiroz, Beatriz. 2013. The interpersonal and experiential grammar of Chilean Spanish: Towards a principled systemic-functional description based on axial argumentation. Sydney: The University of Sydney. Quiroz, Beatriz. 2017. The verbal group. In Tom Bartlett and O’Grady Gerard (eds.), The Routledge handbook of systemic functional linguistics, 301–318. New York: Routledge. Zhang, Dongbing. 2020. An interpersonal grammar of Khorchin Mongolian. Sydney: The University of Sydney.

3

Axial Argumentation and Cryptogrammar in Interpersonal Grammar A Case Study of Classical Tibetan mood Pin Wang

1. An Overview of Classical Tibetan The Tibetan language belongs to the Tibeto-Burman branch in the SinoTibetan family, spoken mainly by Tibetan peoples who live across a wide area of East Asia bordering the Indian subcontinent, including the Tibetan Plateau and northern Indian subcontinent in Nepal, Bhutan, Pakistan, India etc. Modern Tibetan includes a cluster of mutually unintelligible dialects,1 among which the most widely spoken is Standard Tibetan, which is based on the speech of Lhasa, a Central Tibetan dialect, also known as U-tsang dialect. Besides U-tsang dialect, there are two other prominent modern Tibetan dialects, i.e. Amdo dialect and Khams dialect. U-tsang and Khams dialects have tones, while Amdo dialect is not tonal (Janhunen and Norbu 1999: 250). Classical Tibetan, the study object of this chapter, refers to the middle phase of the three stages in the development of the Tibetan language, namely Old, Classical and Modern. Some scholars believe that there is an earlier form before Old Tibetan called Archaic Tibetan and a Medieval form of the Tibetan language between Classical and Modern; therefore, Classical Tibetan sits right in the middle of five historical stages: Archaic—Old—Classical—Medieval—Modern (Hodge 1993: vii). To be more specific, Classical Tibetan refers to the literary Tibetan language that was standardized around the 9th century during the reign of King Khri-lde Srong-btsan for translations of Buddhist canonical texts from Indic languages, especially Sanskrit (Hodge 1993: vii). It assumes historical and literary significance in that it has been used to record a large number of religious, literary and medical documents, and it forms the basis of vocabulary and grammar generally used by Tibetan speakers down to the present day. Classical Tibetan also sets the standard for the contemporary written form of the Tibetan language. Classical Tibetan vocabulary can be classified most generally into content words, including nouns, verbs, adjectives, adverbs etc., and particles that are postpositively dependent on content words and indicate

74

Pin Wang

topicality, various cases, moods or logical relationships with the subsequent clause. This dichotomy of vocabulary reflects the complementarity between lexis and grammar in Classical Tibetan, with content words toward the lexis end and particles toward the grammar end (see Halliday 2008). Tibetan dictionaries mainly provide entries of content words, while Tibetan grammar books largely comprise descriptions of the uses of various particles. From the perspective of more traditional linguistic description, Classical Tibetan adopts an SOV word order, with the verbal group placed at the final position of a clause, as shown in the following example, where the transliterated Classical Tibetan clause is written in italics, with a line of word-for-word glossing2 and a line of free translation here. Note that the abbreviations in uppercase letters are used in the traditional account of Classical Tibetan grammar for particles: TOP for topical particle, ERG for ergative particle etc. (1) da ltar now

ni

bdag gis

TOP I

khyim la

dur

tsam du

’dzin

no3

ERG house ALL grave only TER regard FIN

“Now I regard my house as nothing but a grave.”

Example (1) demonstrates the interaction between content words and particles within a clause. Apart from the zero-marked absolutive dur, all other content words are followed by a particle, which is like mortar holding bricks together. Particles are responsible for making grammatical meaning, which will be addressed in the following sections from the metafunctional perspective.

2. Metafunctional Profile Before addressing the mood systems and structures in Classical Tibetan, it would be helpful to provide a metafunctional profile of the language in question, especially how its experiential and textual meanings are made through its linguistic resources. 2.1 Experiential Metafunction Ideationally, if we explore the ‘transitivity’ system of Classical Tibetan, we find that Classical Tibetan is systemically an ergative language, both in the traditional Latinate sense and in the SFL tradition, following Halliday and Matthiessen (2014: 336–355) and Martin et al. (2010: 109–111). In traditional Latinate terms, the subject of a transitive verb is marked by an ergative particle (called byed sgra in Tibetan), whereas its object and the subjective of an intransitive verb takes the zero-marked absolutive case. The ergative particle in Classical Tibetan has five allomorphs: kyis, gis, gyis, -s and yis, depending on the sound at the final position of the

Axial Argumentation and Cryptogrammar

75

preceding word. For example in (1) earlier, bdag ‘I’ takes gis as the ergative marker. Following the principles formulated by Halliday in his grammar of English (Halliday and Matthiessen 2014: 336–355), the ergative model stands as a grammatical interpretation complementary to the transitive model. Take material process for example: the transitive model views the Process and the Actor as constituting the nucleus of a clause, and this central configuration can extend to another Participant, i.e. Goal, if there is one that is created, changed or influenced by the Process. The clause is transitive where there is a Goal, intransitive where there is not. By the ergative model, the Process is also in the nucleus of a clause, but the Participant closest to the Process and together with it in the core is one through which the process is actualized, called Medium. This Process  + Medium core can then extend to the external cause of the Process, i.e. the Agent, and other Participants such as Range and Beneficiary. The inner configurations of both transitive and ergative models can extend further to peripheral, circumstantial functions such as Manner, Cause, Location etc. So there is a conspicuous parallel between the transitive and ergative models of SFL on the one hand, and the nominative–accusative and ergative–absolutive types of traditional language typology on the other. If we adopt the symbolism in traditional language typology—‘A’ (from ‘agent’) stands for the subject of a transitive verb, ‘P’ (from ‘patient’) stands for the object of a transitive verb, and ‘S’ (from ‘subject’) stands for the subject of an intransitive verb, we will have such correspondence of terminologies as summarized in Table 3.1. Table 3.1 Parallel of Terminology Between SFL and Traditional Language Typology

A S P

nominative-accusative type

SFL transitive model

ergativeabsolutive type

SFL ergative model

nominative

Actor

ergative

Agent

accusative

Goal

absolutive

Medium

In English, the Agent has mixed features of both participants and circumstances, reflected in that the Agent may enter into a clause either as a nominal group or as a prepositional phrase headed by by (Halliday and Matthiessen 2014: 348). In Classical Tibetan, however, the Agent is always indicated by the ergative particle, and the same ergative marker can also pinpoint the functions of Cause or Manner. Using this as criterial, this suggests that the Agent resides closer to other particle-marked Circumstantial elements than to the zero-marked Medium. This can be seen in examples (2) and (3). Here a line of Functions is added to the glossing.

76

Pin Wang

(2) rgyal po de king

-s

dgra bo rnams bcom

that -ERG enemies

Agent

Medium

mo

conquered FIN Process

“The king conquered the enemies.” (3) bcom ldan ’das

kyis

de

epithet of Buddha ERG that Agent

snying rje pa -s compassion

Medium Manner

gsungs so

-ERG said

FIN

Process

“The Buddha said that with compassion.”

In example (2), bcom mo ‘conquered’ as the Process and dgra bo rnams ‘the enemies’ as the Medium constitute the clause nucleus, on which basis the clause extends outwards to the Agent rgyal po de-s ‘that king.’ A more literal translation into English would be like this: “By means of the king (as agent), a conquering-of-the-enemies took place” (Hahn 2005: 48). Example (3) demonstrates the linkage between the Agent and the Circumstantial element of Manner as they both use the ergative particle. The core of the clause is again composed of Process and Medium—‘a saying took place’—and then the ergative particles introduce two outer elements—the saying was performed by the Agent the Buddha and in the Manner of being compassionate. Classical Tibetan does not distinguish between active and passive voices. The voice of a clause is determined by the system of agency. The clause is effective if it has agency, middle if it does not. The middle voice further splits into actualized and non-actualized, depending on whether the clause involves a Medium (Wang 2012). Examples (4)–(6) provide demonstrations of each type. Note that for examples (4) and (5) two free translations are given, in order to capture the feature of non-distinction between active and passive voices in Classical Tibetan. (4) effective bdag

gis

I

ERG lama

Agent

bla ma

mthong ngo saw

FIN

Medium Process

“I saw a lama.”/“A lama was seen by me.” (5) actualized middle bla ma

mthong

ngo

lama

saw

FIN

Medium

Process

“(Someone) saw a lama.”/“A lama was seen.”

Axial Argumentation and Cryptogrammar

77

(6) non-actualized middle sems can

thams cad la

sentient being all

byin no

ALL gave FIN

Beneficiary

Process

“(Someone) gave (something) to all sentient beings.”

2.2 Textual Metafunction Textually, the group-rank structural element in the initial position of a clause serves as the Theme of that clause, representing the point of attention in the field, which gets maintained or shifted in the unfolding of text that is to follow. In examples (4)–(6), the unmarked Themes are bdag gis, bla ma and sems can thams cad la respectively. In contrast, marked Themes in Classical Tibetan clauses are realized through the ‘topical particle’ ni, which isolates the word group from the remainder of the clause to show its topicality, giving that group a particular prominence in the field, e.g. sang rgyas ni ‘as far as sang rgyas (Buddha) is concerned’ in example (7) as follows. Here a line of glossing for textual functions is provided below the word-for-word gloss. (7) sang rgyas

ni

Buddha

mi rnams kyi

TOP men

Theme (marked)

mchog yin no

GEN best

be

FIN

Rheme

“As concerns the Buddha, he is the most excellent of people.”

It is important to note that the marked Theme does not necessarily appear at the initial structural position of a clause. The markedness is realized through the topical particle ni rather than position. In this case the initial element of the clause is deprived of the thematic status as the “angle on the field” (Martin and Rose 2008: 35) is now shifted to the element explicitly marked by the topical particle ni. Besides, when the textual prominence is given to a clausal element that is already marked by a particle that contributes to the experiential meaning, the experiential particle is positioned prior to ni. See example (8). (8) rigs ngan

gyis

ral gri

-s

ni

bsad

do

executioner ERG sword -ERG TOP killed FIN Rhe . . .

Theme (marked)

. . . me

“It is with a sword that (someone) was killed by the executioner.”

A sample short text is hereby provided to demonstrate how the text unfolds and shifts its ‘gaze’ on the field and establish contrasts through deployment of unmarked and marked Themes (in boldface).

78

Pin Wang

(9a) ’dzam bu gling ’di na India

rgyal po shing rta chen po zhes bya ba zhig yod 4

INE king

Theme (unmarked)

Mahāratha

so called

a

de

there be SF

Rheme

“In India there was a king called Mahāratha.”

(9b) rgyal po de king

la

sras gsum mnga’

that ALL son

Theme (unmarked)

three

ste

there be SF

Rheme

“The king had three sons.” (9c) rab first

ni

sgra chen po

TOP

Mahāpraṇāda so called FIN

zhes bya

’o

Theme (marked) Rheme “The first son is called Mahāpraṇāda.” (9d) ’bring po ni middle

lha chen po zhes bya

TOP

Theme (marked)

Mahādeva

’o

so called FIN

Rheme

“The middle son is called Mahādeva.” (9e) tha chung ni youngest

sems can chen po zhes bya ste

TOP Mahāsattva

so called SF

Theme (marked) Rheme “The youngest son is called Mahāsattva.” (9f) sras tha chung

de

chung ngu nas

son youngest that little

byams ba

dang snying rjer

ELA benevolence and

ldan

te

compassion endowed SF with

Theme (unmarked) Rheme “The youngest son was endowed with benevolence and compassion since he was little.”

In (9a), the text sets its initial prominence of message at the land of India and then introduces the king as new information. In (9b), the king is picked up as the point of departure of the message in this clause, followed by an introduction of the three sons. From (9c) to (9e), marked Themes are deployed not to signal new phases in the discourse but to provide direct contrast amongst the three clauses, bringing prominence to each one of the sons in relation to the other two. In (9f), the perspective is shifted to the youngest of the sons by giving it the thematic status.

Axial Argumentation and Cryptogrammar

79

3. Types of Particles As is mentioned cursorily in Section 1, Classical Tibetan particles are postpositively dependent on content words to form word groups, indicating various nominal group cases, logical relationships between elements, moods of the clause and topicality of certain element of the clause. From the metafunctional perspective, the particles can thus be viewed as falling into the following types based on their functions: ideational (experiential and logical) particles, interpersonal particles and textual particles. In examples (1), (7) and (9c)–(9e), the particle ni indicates the topic of the clause. This means the nominal group ending with it is the (marked) Theme, which realizes a function textually; therefore, it shall be called a textual particle. Associated with the system of transitivity, within the experiential component of the ideational metafunction, the Agent is indicated by the ergative particle, as in examples (1)–(4) and (8), which, besides Agent, can also realize Means, Quality and Cause. The allative particle la, as in examples (1) and (6), is used for direction or recipient of the process, and the la nominal group usually functions as Location or Beneficiary. The terminative particle du in example (1) makes the nominal group attached to it function as Manner in the clause; on other occasions it can be used for Location or Range. What is traditionally called semi-final particle de, te and ste in examples (9a), (9b), (9e) and (9f) realize paratactic relation with the forthcoming clause, and so forth. These particles are ideational ones, as they fulfill experiential or logical metafunctions. The particles no, mo, so etc. at the end of the clauses in many of the previous examples, marked in the first gloss line with FIN (for ‘final particle’ in traditional Classical Tibetan grammar) play their part in the mood system, all indicating the declarative mood. It represents a facet of the interpersonal metafunction, hence their classification as interpersonal particles. Table 3.2 provides a classifications of particles that have thus far appeared. Table 3.2 Particles and Their Metafunctions Metafunction

Particle

Ideational Experiential

Logical

gis, -s, kyis, gyis (ERG) la (ALL) du (TER) kyi (GEN) na (INE) nas (ELA)

de, te, ste (SF)

Interpersonal

Textual

no, mo, so, ngo, do, ’o (FIN)

ni (TOP)

80

Pin Wang

As noted previously, although various particles of Classical Tibetan are appearing in a same syntagmatic structure, they may realize functions from different metafunctions. It goes beyond doubt that there are far more particles in the Classical Tibetan language than those given in the previous examples; however, any particle will fit into one of the previous categories in accordance with their metafunctions. This chapter explores the mood system of Classical Tibetan; thus it focuses on the interpersonal particles, specifically those particles that have a part to play in the realization of the mood structure. Those particles are, as a matter of fact, important elements in the manifestation of mood of a clause; function labels will be subsequently rendered as per SFL practice to the final particles realizing mood, at the group rank.

4. MOOD: A Look From Above In this chapter, the exploration of the mood system in Classical Tibetan is based on the theoretical and methodological principles of axial argumentation and cryptogrammar. Such principles deal with the complementarity between system and structure, foreground paradigmatic relations in the sense that system gives rise to structure and explore systemic contrasts for the interpretation of underlying grammatical patterns (see e.g. Halliday 1985; Halliday and Matthiessen 1999; Davidse 1992; Martin 1996a, 1996b, 2013). To address the interpersonal grammar of mood, this chapter approaches the description from above, since lexicogrammar in SFL has evolved to interact with the stratum of discourse semantics. Specifically, the mood system is responsible for the realization of the discourse semantic system of negotiation, which is concerned with the structure of interactive exchanges—how discourse roles are assigned and how moves are organized in relation to one another (Martin and Rose 2007: 219). For a full account of exchange in English, see Martin (2013: 101). Under the model of exchange presented in Martin (1992), utilized here, exchanges oriented toward the negotiation of goods and services are known as action exchanges, in which the provider of goods and services is called the primary actor. The core move of an action exchange involves the primary actor giving the good or doing the service (or at least promising to) and is labeled A1. In contrast, the initiator or demander of the goods and services is known as the secondary actor, and the move that does the demanding is known as A2. These two moves are exemplified by the following dialogue. Speaker 1 A2 “Do not let the horse escape!” Speaker 2 A1 “OK.” [stopping the horse]

Axial Argumentation and Cryptogrammar

81

In contrast to action exchanges, the negotiation of information is known as a knowledge exchange. Like in action exchanges, in knowledge exchanges, the one that gives information is called the primary knower and takes the K1 move (the move that controls the information), and the one who requests information is the secondary knower and takes the K2 move. These two moves are exemplified by the following dialogue. Speaker 1 K2 “Where are you going to?” Speaker 2 K1 “To the palace.”

Besides the four major moves, there are also moves that delay the delivery of goods and services and information, Da1 and Dk1; moves that follow it, A2f and K2f; tracking moves and response to track, Tr and Rtr; and challenging moves and response to challenge, Ch and Rch. However, these moves are not potentially germane to the current discussion; therefore, this chapter is not going to focus on them.

5. MOOD in Classical Tibetan The present study draws on text-based data for analysis. Specifically, the description in this chapter is grounded upon a Buddhist folktale in Tibet entitled Khyim bdag dbyig pa can gyi le’u “Householder Daṇḍin’s Chapter,” written in standard literary Classical Tibetan. Buddhist folktales are typical of written literary texts and are celebrated in Tibet, where Buddhism has exerted profound influence in Tibet’s culture and customs as the most extensively practiced religion of Tibetans. The following is an excerpt from the text, segmented clause by clause with quoted speech indented, on which the present description and analysis are mainly based. The type of move of each clause is indicated to the left side of the text. K1 khye’u

de

srog dang bral ba -r

young boy that life

from free

gyur

-TER became FIN

“That young boy lost his life.” K1 de

-s

kyang dbyig pa can de

she -ERG also

Daṇḍin

to

bzung ste

that seized

SF

“And she seized Daṇḍin,” K1 khyod kyis nga -’i bu bsad kyis you ERG I -GEN son killed ERG “‘Because you killed my son,’”

82

Pin Wang A2 nga -’i I

bu

byin cig

-GEN son give

FIN

“‘give my son (back)!’” K1 de -s smras pa he -ERG said NOM “He said,”" K1 nga -s I

ni

ma

bsad

na

-ERG TOP not killed INE

‘Since I did not kill (him),’ K1 khyod kyi you

bu

nga -s

GEN son I

ji ltar sbyin

-ERG how

shall give

“‘how shall I give your son (back)?’” A2 de

-s

tshur

shog

she -ERG hither come “She (said), ‘come here!’” A2 rgyal po -’i king

drung

’dong dang

du

-GEN vicinity TER go

SOC

“‘Let’s go to the king’s place!’” K1 de -s

’u bu cag

he -ERG we

gi

zhal che gcod

GEN decision will decide FIN

“‘He will make a decision for us.’” K1 zhes

smras nas

thus

said

ELA

“Thus said (she).” K1 rgyal po -s king

dbyig pa can

-ERG Daṇḍin

la

smras pa

ALL said

“The king said to Daṇḍin,” K2 khyod kyis you

glang brnyas

ERG ox

borrowed

so FIN

“‘(I) borrowed (it).’”

sam

borrowed FIN

“‘Did you borrow the ox?’” K1 brnyas

do

NOM

Axial Argumentation and Cryptogrammar

83

In our analysis, the narrator’s voice is seen as imparting knowledge to the readers in an author–reader dialogue, thus being all K1 moves. The dialogues proper in the story are set apart from the narrative by further indentation. Congruently, the moves A2, K2 and K1 are realized grammatically in the system of mood as imperative, interrogative and declarative respectively, whereas in our data there is also a metaphorical realization of the K1 move as the interrogative mood. The system and structure of each mood type is accounted for later. 5.1 Imperative Mood Looking from above at exchanges, the primary contrast in the discourse semantic system of negotiation is between an exchange of action and that of knowledge. This is reflected in the grammatical system of mood that first and foremost, there is a fundamental distinction between the imperative mood on the one hand, and the interrogative and declarative moods—combined as the indicative—on the other, on the grounds that the imperative mood is involved in the exchange of action whereas the indicative realizes moves in the exchange of knowledge. In Classical Tibetan, the imperative mood is typically responsible for the realization of A2 moves that tend to involve demanding goods and services. In the data under study there are three A2 moves in the negotiation structure, each realized by the imperative mood. These are reproduced as examples (10)–(12) with the key interpersonal function Predicator analyzed more fully. (10) A2 nga -’i I

bu

byin

-GEN son give

cig FIN

Predicator verbal group Event Imperator verb “Give my son (back)!” (11) A2 tshur

shog

hither come Predicator verbal group Event verb “Come here!”

particle

84

Pin Wang

(12) A2 rgyal po -’i king

drung

du

’dong dang

-GEN vicinity TER go

SOC

Predicator verbal group Event Linker verb

particle

“Let’s go to the king’s place!”

The metalanguage used in the three examples is expanded on the basis of previous ones. The gloss for lexical words and labels of particles are kept to provide lexical meaning of content words and a useful link with traditional Classical Tibetan grammar. The next line is a clause rank function in the interpersonal perspective, Predicator, which plays a key role in the mood system in Classical Tibetan. It is realized by a verbal group, indicated in the next line. The verbal group has an obligatory function labeled Event, specifying the nature of the process itself and realized by a verb; following the Event, optionally, there is an Imperator or a Linker realized by a particle, which will be discussed later. The verbal group, which is pivotally involved in the realization of the imperative mood, is displayed in boldface (as it will be for polar interrogative and declarative examples later). From the three examples, it is notable that the imperative mood in Classical Tibetan rests within the clause rank function Predicator, which is structurally realized through the verbal group. Within the verbal group, the Event verb is directly associated with the mood choice. To explore this, we need first to outline the organization of verbs in Classical Tibetan. For each verb, excluding auxiliary ones, modifications to vowels, and/or initial, radical or ending consonants may lead to the creation of four different verbal stems. Each of the four stems performs a different grammatical function, and they are called in traditional Classical Tibetan grammar the present stem, the perfective stem, the future stem and the imperative stem respectively. Some verbs possess four different forms for the four stems, while others have some overlapping among the stem forms, so that they may have one, two or three stem forms for the four stem functions. Therefore, the number of stem forms for each verb is between one at the least and four at the most. In dictionaries verbs are listed with the present verbal stem form. Table 3.3 is a synopsis of verbal stems of Classical Tibetan with examples. Among the four verbal stem forms, the present stem, e.g. byed ‘do’ in Table 3.3, can be regarded as the neutral form, which is deployed in a context where there is not any restriction in terms of tense, aspect, mood or modality (this form is also that which is itemized in dictionaries and vocabulary lists). The perfective stem, e.g. byas (roughly ‘did’) is used

Axial Argumentation and Cryptogrammar

85

Table 3.3 Verbal Stems of Classical Tibetan present perfective future stem stem stem

imperative stem

byed

byas

bya

byos

rnyed

brnyes

brnyed rnyed

bsnar

bsnar

snor

’khol

bkol

’khol

khol

sbyin

byin

sbyin

byin

‘to meet’ ’phrad

phrad

phrad

phrad

‘to come’ gshegs

gshegs

gshegs

shog

‘to go’

dong

’dong

’dong

4 stem ‘to do’ forms 3 stem ‘to forms obtain’

‘to snar lengthen’ ‘to employ’ 2 stem ‘to give’ forms

1 stem ‘to see’ form

’dong

mthong mthong

mthong mthong

stem forms all distinct identical present and imperative stem forms identical perfective and future stem forms identical present and future stem forms identical present and future stem forms; identical perfective and imperative stem forms identical perfective, future and imperative stem forms identical present, perfective and future stem forms identical present, future and imperative stem forms all stem forms identical

for the perfective aspect, which describes a completed whole action. The future stem, e.g. bya ‘shall do,’ so called in traditional Tibetan grammar, is not actually an indicator of the future tense, but of the modality of necessity or volition. It implies the wish to carry out an action that has not been started yet (Hahn 2005: 55). The imperative stem, e.g. byos ‘do!’— as the name suggests—is the verbal stem used for a positive imperative. Thus in examples (10)–(12), the verbs byin, shog and ’dong are all in the imperative stem form, in opposition to other stem forms, as can be seen from Table 3.3. One optional structural element in an imperative clause is the imperative type of ‘final particle,’ so-called in traditional Tibetan grammar. This particle, as has appeared in many of the previous examples, indicates the end of a complete sentence. There are three types of final particles, all interpersonal: imperative, declarative and interrogative. The imperative particle is part of the verbal group and realizes the function Imperator at the group rank.

86

Pin Wang

The imperative particle has three allomorphs, cig, zhig and shig. The exact form is determined by the final sound (represented with letters and symbols in transliteration) of the preceding verb: 1. cig is used after a final -g, -b or -d, e.g. sod → sod cig ‘kill!’; 2. zhig is used after a final -ng, -n, -m, -r, -l, or -’, e.g. nyon → nyon zhig ‘listen!’; 3. shig is used after a final -s, e.g. byos → byos shig ‘do!’5 Therefore, to look from below, the verb is primarily responsible for the imperative mood in a Classical Tibetan clause by adopting the imperative stem. It can be followed by an optional imperative particle realizing the Imperator, as shown in example (10). But the Imperator can be dispensed with as in example (11). Alternatively, the position of the imperative particle can be occupied by a linking particle, functionally labeled as Linker, as in example (12), where the sociative particle dang after the verb in its imperative stem form ’dong does not directly have to do with the realization of the imperative mood, but it is a logical marker linking the clause with the following one. We can term the imperative clauses with the Imperator or Linker as ‘finalized,’ those without it ‘non-finalized,’ on the basis that finalized imperatives have a Predicator realized by a particled verbal group. So far we have only addressed the positive imperative (the jussive) in Classical Tibetan. Opposing this, the negative imperative (the prohibitive) does not require the imperative stem of the verb, but is realized by placing the negation adverb ma before the present or perfective stem of the verb. The function this negation adverb realizes will be labeled Negator, which is also a function at the verbal group rank. This is shown in examples (13)–(16) that compare the jussive with the prohibitive. Both jussive and prohibitive can be either non-finalized, as in examples (13)– (14), or finalized, as in examples (15) and (16). (13) imperative: jussive, non-finalized ltos look Predicator verbal group Event verb “Look!” (14) imperative: prohibitive, non-finalized ma

lta6

NEG

look

Axial Argumentation and Cryptogrammar

87

Predicator verbal group Negator

Event

adverb

verb

“Do not look!” (15) imperative: jussive, finalized rta

thong

horse

set free FIN

zhig

Predicator verbal group Event

Imperator

verb

particle

“Set the horse free!” (16) imperative: prohibitive, finalized rta

ma

horse NEG

btang7

zhig

set free FIN

Predicator verbal group Negator Event

Imperator

adverb

particle

verb

“Do not let the horse escape!”

Figure 3.1 summarizes choices involved in imperative clauses in Classical Tibetan with a system network. For all major clauses there must be a clause rank function Predicator responsible for the mood system. For the imperative mood to be realized there must also be the group rank function Event, and so this is inserted at this entry condition stage in the realization statement.8 To distinguish between realizations at the clause and group ranks, realizations within the verbal group are put in boxes. The imperative mood chooses between jussive and prohibitive, with the distinction being realized in the verbal group by the choice of verb stem and the insertion of a Negator in the prohibitive. The Negator is lexicalized as ma and placed before the Event in sequence. As both can have an optional particle at the final position of the verbal group, a simultaneous system is drawn up to capture this choice, indicated by the right-facing brace. For the non-finalized option, the Event itself constitutes the verbal group, which appears at the end of the clause. For the choice finalized, an Imperator or a Linker is inserted in the verbal group, and is placed after the Event in sequence. We can pursue the difference in finalized imperatives in further delicacy by dividing it into ending and linking, depending

88

Pin Wang jussive Event: verb in imperative stem prohibitive Event: verb in present or perfective stem; +Negator; Negator:: ma; Negator^Event non-finalised

imperative

Predicator: non-particled verbal group

+Predicator +Event

ending +Imperator; Imperator: imperative

finalised Predicator: particled verbal group

particle; Event^Imperator^# linking +Linker; Linker:: dang; Event^Linker^#

Figure 3.1 System Network of the Imperative Mood

on what kind of particle is attached in the verbal group: the Imperator is realized by the imperative particle; the Linker is lexicalized as dang. 5.2 Interrogative Mood The K2 move in the exchange of knowledge is congruently realized by the interrogative mood at the stratum of lexicogrammar. In the dialogues in our excerpt of data there is one such move, presented in example (17), together with the response move K1 in example (18). (17) K2 khyod kyis you

glang brnyas

ERG ox

sam

borrowed FIN

“Did you borrow the ox?” (18) K1 brnyas borrowed

so FIN

“(I) borrowed (it).”

Such a request of knowledge in the K2 move demands that the response be either an explicit acknowledgement or denial of the proposition that is put forward in the request. In our case the response is in the affirmative brnyas so “(I) borrowed (it).” From this round of exchange of knowledge it is readily observable that the Predicator again plays an essential part in posing a polar question and making a response in Classical Tibetan. Example (17) is represented with function and class glossing.

Axial Argumentation and Cryptogrammar

89

(17) interrogative: polar khyod kyis you

glang brnyas

ERG ox

sam

borrowed FIN Predicator verbal group Event

Interrogator

verb

particle

“Did you borrow the ox?”

Here the verbal group has a controlling stake in the realization of the polar interrogative, with the Event appearing in the non-imperative stem of the verb. We can also observe that the verb realizing the Event remains unchanged in the interrogative and the response in the declarative, and it is the final particle that differentiates one from the other. The final particle, traditionally known as the interrogative particle, is a required element in the polar interrogative. Functionally it can be termed Interrogator, which means polar interrogatives are always finalized. The basic realization form of the Interrogator is the interrogative particle ’am. The interrogative particle also has allomorphs conditioned by the ending sound of the preceding word. Here are the rules: 1. If the preceding word ends with a vowel, ’am is directly attached to the preceding word, e.g. bya ‘shall do’ → bya’am “Shall (it) be done?”;9 2. If the preceding word ends with ’, the ’ is dropped out to avert duplication of it, and then ’am is appended to the remaining form, e.g. bda’ ‘drive out’ → bda’am “Is (it) driven out?”; 3. If the preceding word ends with a consonant, the consonant needs to be repeated as the initial of the interrogative particle, e.g. yin ‘be’ → mi de dgra yin nam ‘Is that man an enemy?’; ’ongs ‘arrived’ → pha ’ongs sam “Has father arrived?” Like the imperative particle, the d-dropping verb forms (see Note 4) require an interrogative particle not conditioned by the final consonant at the Classical Tibetan stage of orthography, but by the historical form with a final -d. However, the formation of interrogative particle for d-dropping verb form does not follow rule 3, giving do, but has a special form to, e.g. stor (from stord) ‘lost’ → stor tam “Was (it) lost?” As mentioned previously, there is a K1 move in our data incongruently realized as an interrogative clause, represented in example (19). Although this move in the exchange is meant by the primary knower to impart knowledge, that is, to express to the character in the story the proposition “I can’t give your son back (since I didn’t kill him),” grammatically

90

Pin Wang

this is an elemental interrogative clause, ostensibly seeking some missing information—‘in what manner’ shall I give your son back to you? (19) interrogative: elemental khyod kyi you

bu

nga -s

GEN son I

ji ltar

-ERG how Inquirer

sbyin shall give Predicator

adverbial group verbal group Event verb “How shall I give your son (back)?”

In this elemental kind of interrogative clause in Classical Tibetan, the equivalent of Wh element in English that seeks the missing information is retained at the original position, where the piece of information would be placed in the declarative. Let’s apply to that element the functional label Inquirer. In example (19) the Inquirer lexically realized as ji ltar ‘how’ appears before the Predicator, which is a position where, ideationally, the Circumstance of Manner would normally hold. In addition to this, the elemental interrogative differs from the other mood types in that it cannot take any kind of particle at the end of the verbal group, i.e. it is non-finalized. An exchange outside our data is used in examples (20) and (21) to further illustrate the structural configuration of elemental interrogative mood. Here the Inquirer su ‘who’ demands the information about the Attribute in the relational process and is placed structurally where that Attribute is. (20) K2 mi

de

su

man that who Inquirer Carrier

yin be Predicator

Attribute Process

“Who is that man?” (21) K1 mi

de

dgra

man that enemy

yin be Predicator

Carrier

Attribute Process

“That man is an enemy.”

Thus far a system network of the interrogative mood in Classical Tibetan can be displayed in Figure 3.2. At the clause level, the insertion of a Predicator is required as the realization statement for the entry feature

Axial Argumentation and Cryptogrammar

91

polar +Interrogator; Interrogator: interrogative particle;

interrogative

Event^Interrogator^#

+Predicator elemental

+Event;

+Inquirer; Inquirer/

Event: verb in non-imperative stem

Figure 3.2 System Network of the Interrogative Mood

interrogative; this Predicator is realized at the group rank through a verb in a non-imperative verbal stem as the Event. For polar interrogative, an Interrogator realized by the interrogative particle is positioned at the end of the verbal group and after the Event; for elemental, in addition to the insertion of the Inquirer at the clause rank, it should be conflated with the function served by the sought information, retained at the original position as in the declarative. 5.3 Declarative Mood The K1 move in the exchange of knowledge is congruently realized by the declarative mood, typically involved in the giving of information. To address the grammatical system and structure of the declarative mood, we need first to divide the K1 moves in our data into three groups. The first group features the use of final interpersonal particles at the end of the clause, as in examples (18) (reproduced here) and (22)–(23). (18) K1 brnyas

so

borrowed

FIN

“(I) borrowed (it).” (22) K1 khye’u

de

srog dang bral ba -r

young boy that life

from free

gyur

to

TER became FIN

“That young boy lost his life.” (23) K1 de -s

’u bu cag gi

he ERG we

zhal che gcod

GEN decision will decide

do FIN

“He will make a decision for us.”

Like in the imperative and interrogative moods, the final particles here also serve to indicate the end of complete sentences. But the particles here are of the declarative type, morphologically featuring the ’o form. Like the

92

Pin Wang

interrogative particle, the declarative particle also has allomorphs under different circumstances. The rules are as follows: 1. If the preceding word ends with a vowel, ’o is directly attached to the preceding word, e.g. ’dre ‘mix’ → ’dre’o;. 2. If the preceding word ends with ’, the ’ is dropped out to avert duplication, and ’o is then attached to the remaining form, e.g. bda’ ‘drive out’ → bda’o;. 3. If the preceding word ends with a consonant, the consonant is repeated as the initial of the declarative particle, e.g. ’tshong ‘sell’ → ’tshong ngo; smras ‘say’ → smras so. Also, like in the imperative and interrogative, the d-dropping words do not conform to rule 3, but uses to as the interrogative particle, conditioned by the historical phonological and graphological form, e.g. stor (from stord) ‘lost’ → stor to. Example (22) is represented here with glossing for functions and classes. Again, the verbal group is the unit responsible for mood. Within the verbal group, the Event is realized by a verb that can take any of the three stems other than the imperative. This resembles the interrogative, but the final particle is of the declarative type, functionally labeled as Declarer here. (22) declarative: finalized khye’u

de

srog dang bral ba -r

young boy that life

from free

gyur

to

-TER became FIN Predicator verbal group Event

Declarer

verb

particle

“That young boy lost his life.”

Similar to the Imperator in the imperative mood, the Declarer in the declarative is optional. This brings us to the second group of declarative clauses in our data. They forsake the use of the Declarer realized by the final declarative particle, in whose place other kinds of particles are put to establish connection with the subsequent clause, as in examples (24)–(26). Among them, the semi-final particle in example (24), so-called in traditional Tibetan grammar, is a paratactic linker forming parallelism between this clause and the ensuing one. Examples (25) and (26) are β clauses in hypotactic clause complexes. In example (25), the ergative particle kyis appended to the verb in its perfective stem bsad ‘killed’ bears causal relation; the inessive particle na in example (26) establishes

Axial Argumentation and Cryptogrammar

93

a temporal or conditional relation, hence its usual English translation ‘if,’ ‘when’ or ‘since.’ (24) K1 de

-s

kyang dbyig pa can

she -ERG also

de

Daṇḍin

bzung ste

that seized

SF

“And she seized Daṇḍin,” (25) K1 khyod kyis you

nga -’i

ERG I

bu

bsad

kyis

-GEN son killed ERG

“Because you killed my son,” (26) K1 nga -s I

ni

ma

bsad

na

-ERG TOP NEG killed INE

“Since I didn’t kill (him),”

Declarative clauses in the previous examples are also finalized in that the clause has a Predicator that is realized by a particled verbal group. The particles connect the clause with a following clause in a logical relationship, forming clause complexes. Those logical particles are given the function label Linker, operating at the group rank. Example (27) is a functional representation of a clause complex conjoining example (25) with the ensuing α clause (as has appeared in example (10)): (27) declarative: finalized: linking khyod kyis you

nga -’i

ERG I

bu

imperative bsad

kyis

-GEN son killed ERG

nga -’i I

bu

byin

-GEN son give

cig FIN

α

×β Predicator

Predicator

verbal group

verbal group

Event Linker

Event Imperator

verb

verb

particle

“Because you killed my son,”

particle

“give my son (back)!”

The third group of realizations of K1 moves as the declarative do not insert the declarative particle represented by ’o realizing the Declarer function after the Event verb. Nor are there Linkers after the Event to logically connect such declarative clauses with subsequent ones. They employ at the position after the Event verb what is in traditional Classical Tibetan description known as the ‘nominal particle’ pa,10 as in examples (28) and (29). (28) K1

de

-s

smras

pa

she

-ERG

said

NOM

“She said,”

94

Pin Wang

(29) K1 rgyal po -s king

dbyig pa can la

-ERG Daṇḍin

smras pa

ALL said

NOM

“The king said to Daṇḍin,”

In some traditional Classical Tibetan grammatical descriptions it is suggested that the ‘nominal particle’ transforms verbal stems into ‘verbal substantives (gerunds)’ (Hahn 2005: 27), or ‘verbal nouns’ or ‘participles’ (Hodge 1993: 13). And it is not uncommon in Classical Tibetan texts for this particle to be attached to verbs in non-imperative stems in the declarative mood. However, to observe this annexation of ‘nominal particle’ with a systemic functional eye by looking axially, we find that the so-called ‘nominal particle’ does not nominalize the verbal groups here; nor does it downrank the clauses into a nominal group. Examples (28) and (29) justifiably stand as ranking clauses, only that the verbal groups in the clauses end with a different particle. A functional representation of example (29) is presented here, in contrast to a declarative clause that bears the same ideational meaning but has the declarative particle in the verbal group, numbered as (29’). We can recognize the particle pa as another type of Declarer functionally, which also makes the declarative clause finalized but does not logically link it to the next clause. This type of finalized declarative is termed ‘ending’, as opposed to the ‘linking’ type prior. The different realizations of the ending declarative, in further delicacy, lie in modal assessment of the proposition. The Declarer realized by the particle pa puts forward the proposition as reliable and rules out alternative viewpoints, hence a dialogic contracting resource in the discourse semantic system of engagement (Martin and White 2005: 102). This explains why in the story smras pa ‘said’ is always followed by the projected clauses—the readers are oriented toward concerning themselves with the message in what the characters say, rather than assessing the plausibility of the saying process itself. This type of ending declarative is termed ‘guaranteed.’ In contrast, the Declarer realized by the particle so gives the proposition a status of bare assertion. Bare assertions do not reference other voices or recognize alternative positions, but are claimed categorically and accord a monoglossic communicative context to what is being exchanged (Martin and White 2005: 99). This type of ending declarative is termed ‘asserted.’ The English translation for examples (29) and (29’) uses ‘said’ and ‘spoke’ respectively, in an effort to capture the nuance in modal assessment. (29) declarative: finalized: ending: guaranteed rgyal po -s king

dbyig pa can la

-ERG Daṇḍin

smras pa

ALL said NOM Predicator verbal group

Axial Argumentation and Cryptogrammar Event

Declarer

verb

particle

95

“The king said to Daṇḍin,” (29’) declarative: finalized: ending: asserted rgyal po -s king

dbyig pa can

-ERG Daṇḍin

la

smras so

ALL said

FIN

Predicator verbal group Event

Declarer

verb

particle

“The king spoke to Daṇḍin.”

Besides the three types of finalized declarative clauses in our data, there is a fourth kind of declarative clause in Classical Tibetan, which is non-finalized, in that the Predicator is realized by a non-particled verbal group. Sometimes where there is a succession of declarative clauses in paratactic relations, the final declarative particle can only occur in the last verbal group. This renders the clause(s) before the last one nonfinalized declaratives, as shown in example (30). Also, if the verb is yin ‘to be’ or yod ‘to have,’ the final declarative particle may be omitted, as shown in example (21) reproduced here. Both examples are not from the story in the present study. (30) declarative: non-finalized

declarative: finalized

bud med rgan mo shi

bu mo gzhon pa nyams

woman

girl

old

died

1

young

+2 Predicator

Predicator

verbal group

verbal group

Event

Event

Declarer

verb

verb

particle

“The old woman died, and the young girl suffered harm.” (21) declarative: non-finalized mi

so

harmed FIN

de

dgra

yin

man that enemy be Predicator verbal group Event verb “That man is an enemy.”

96

Pin Wang

The declarative mood in Classical Tibetan is thus summarized with a system network in Figure 3.3. The only necessary realization rule for the declarative is also the insertion of a Predicator at the clause rank, realized by an Event at the group rank, which in turn is realized by a verb that takes the present, perfective, or future, i.e. non-imperative verbal stem. Then there is the choice between finalized and non-finalized, the former, realized by a particled verbal group as Predicator, further splitting into ending and linking. The ending types have a Declarer inserted and placed after the Event in sequence, at the end of the verbal group. For asserted, the Declarer is realized by the declarative final particle; for guaranteed, the Declarer is realized by the particle pa. The linking type has a Linker at the final position of the verbal group, after the Event. For non-finalized, the Event is simply left alone in the verbal group. asserted ending finalised Predicator: declarative +Predicator +Event; Event: verb in

particled verbal group

+Declarer; Event^Declarer^#

Declarer: declarative particle guaranteed Declarer:: pa

linking +Linker; Event^Linker^#

non-finalised Predicator: non-particled verbal group

non-imperative stem

Figure 3.3 System Network of the Declarative Mood

5.4 Synopsis of MOOD: Another Perspective The account of these imperative, interrogative and declarative moods focuses primarily on units within the verbal group as they are responsible for most mood choices. If we approach the mood system with clause as the entry condition, we find that at the most general level, there are three simultaneous systems. Besides the distinction between indicative and imperative as the general mood types, there are the system of finalisation, which essentially specifies whether the verbal group has a particle attached, across all mood types, and that of polarity, since all moods can select in this system between positive and negative. The negative polarity for the indicatives is also realized by inserting a Negator before the Event, as in the imperatives. The system network is represented in Figure 3.4. The network suggests that motivated by realizations, the indicatives are grouped together on the basis that the Event in the Predicator has a verb in a non-imperative system, whereas the imperatives take the imperative stem. This is in parallel with the argument in the discourse

POLARITY

MOOD

Negator^Event

+Negator;

negative

positive

imperative

Event: verb in non-imperative stem

indicative

Negator:: ma

declarative particle

Declarer:

asserted declarative

Declarer:: pa

non-finalised declarative

+Inquirer; Inquirer/

elemental interrogative

linking declarative

+Interrogator

Event: verb in present/perfective stem;

prohibitive

Event: verb in imperative stem

jussive

linking-particled verbal group

Predicator:

linking

Predicator: non-particled verbal group

non-finalised

particled verbal group

Predicator:

finalised

Figure 3.4 Synopsis of mood System in Classical Tibetan

+Event

+Predicator

clause

FINALISATION

ending

polar interrogative

+Declarer

ending declarative

guaranteed declarative

98

Pin Wang

semantics—indicatives are responsible for exchange of knowledge, as opposed to exchange of action realized by imperatives. However, the realizations suggest that interrogatives and declaratives are actually split across two systems, according to whether the Predicator is realized by a verbal group that takes a particle after the Event verb. This goes against the argument from the perspective of discourse semantics, where interrogatives congruently realize K2 moves and declaratives congruently realize K1 moves; rather, the congruent grammatical realizations of K2 and K1 are distributed across multiple systems. This indicates that sometimes there is tension between motivation from realizations and argument from above when it comes to grammatical systems and structures. They provide different perspectives toward grammatical description.

6. Conclusion This chapter attempts to probe into the interpersonal grammar in Classical Tibetan with a central focus on the mood system, discussing how the exchange of knowledge and action in the discourse semantic system of negotiation are realized by the lexicogrammatical resources in the Classical Tibetan language. On the basis of text-based data and axial reasoning, it is revealed that, in Classical Tibetan, mood manifests itself mainly through the group rank function Predicator, the key element that establishes the clause as a negotiable move, differentiating mood types through use of different verbal stems and co-occurrence with relevant particles. Thus Classical Tibetan is a ‘Predicator-based’ language in the typology of mood system and structure (see Teruya et al. 2007: 913). The present study is an application of Systemic Functional theory to the understanding and description of a language genetically unrelated to English, a challenge taken up by a good number of SFL typologists. An important issue of such work is to sort out how the general descriptive mechanism proposed by SFL typology, such as axial argumentation, approaching grammar from above, trinocular perspective etc. can be localized in particular languages, so as to avoid foisting the English code on other languages and “making all other languages look like imperfect copies of English” (Halliday 1994: F56–57). Through exploration of the cryptogrammar of mood for example, in particular languages, we do not ask what elements in a clause are equivalent to the English Subject and Finite in a language, but ask what contrast allows us to understand a linguistic unit as an interactive event, or negotiable move (see Halliday 1984, 1992). The general theoretical framework and particular languages logically constitute a complementarity—SFL typologists use SFL theory to account for the way a language makes meaning on the one hand, and need to conduct research on various languages to attest to the appliability of the theory on the other.

Axial Argumentation and Cryptogrammar

99

Abbreviations ALL ELA ERG FIN GEN INE NOM SF SOC TER TOP

allative elative ergative final genetive inessive nominal semi-final sociative terminative topical

Notes 1. Although mutual intelligibility is often used as a yardstick by which to tell apart linguistic varieties as languages or dialects—dialects if mutually intelligible, languages if not—this chapter respects the tradition of viewing varieties of the Tibetan language as dialects, e.g. Jäschke (1883). 2. Separate glossing is provided for morphemes s (ergative), ’i (genitive) r (terminative) from Example (2) onward because they are morphological variations to word level ergative, genititive and terminative particles. 3. The transliteration of Tibetan alphabet in this chapter uses the Wylie System, a widely adopted transliteration scheme in Tibetan studies devised by Turrell V. Wylie. 4. The translation of names of people in this short text may look confusing, as the English glossing bears no apparent phonological similarity to the original Classical Tibetan transliteration. This is because names in Buddhist texts are translated from the original Sanskrit into Classical Tibetan based not on the sound but the meaning, whereas in English the Sanskrit transliteration is preserved. Here in the name Mahāratha, mahā means ‘big,’ equivalent to chen po in Classical Tibetan; ratha means ‘chariot,’ equivalent to shing rta in Classical Tibetan. The other names are translated likewise. 5. Note that there is an exception to these rules. After the reform of the Tibetan script, a final consonant -d after some verb forms (known as ‘d-dropping’ verb forms) ending in -n, -r or -l, called da drag, was lost. Although the da drag does not appear in the new orthography, it leaves a trace in the selection of final particles, and imperative stems that used to have this da drag select cig as its imperative particle, e.g. gyur (from gyurd) → gyur cig, not gyur zhig. 6. The verb ‘to look’ has the following stems: present lta, perfective bltas, future blta and imperative ltos. 7. The verb ‘to set free’ has the following stems: present gtong, perfective btang, future gtang and imperative thong. 8. For the meaning of symbols in the realization statements, see Martin (2013: 41). The basic statements that are important to this chapter are: +F: insert Function; F^F: sequence Functions; F^#: position Function last; F:c: realize Function by class; F::word: realize Function by a specific lexical item; F/F: conflate Functions. 9. Note that in an ergative language like Classical Tibetan, the participant that is semantically closest to the Process is the Medium, hence the English translation in the passive voice.

100

Pin Wang

10. The nominal particle pa has an allomorph ba used after final vowel sounds and after final consonants ng, ’, r and l. Where r and l were originally followed by da drag (see Note 4), the particle pa remains unchanged.

References Caffarel, A., J. R. Martin and C. M. I. M. Matthiessen. 2004. Language Typology: A Functional Perspective. Amsterdam/Philadelphia: John Benjamins. Davidse, K. 1992. Transitivity/ergativity: The Janus-headed grammar of actions and events. In Davies, M. and L. Ravelli (Eds.) 2002, Advances in Systemic Linguistics: Recent Theory and Practice. London/New York: Pinter Publishers, 105–135. Hahn, M. 2005. A Textbook of Classical Literary Tibetan, translated from German into English by Ulrich Pagel. London: SOAS. Halliday, M. A. K. 1984. On the ineffability of grammatical categories. In Webster, J. J. (Ed.) 2002. On Grammar. London: Continuum, 291–322. Halliday, M. A. K. 1985. Systemic background: Systemic perspectives on discourse. In J. J. Webster (Ed.), The Collected Works of M. A. K. Halliday (Vol. 3). London: Continuum. Halliday, M. A. K. 1992. Systemic grammar and the concept of a ‘science of language’. In Webster, J. J. (Ed.) 2003, On Language and Linguistics. London: Continuum, 199–212. Halliday, M. A. K. 1994. An Introduction to Functional Grammar (2nd ed.). London: Arnold. Halliday, M. A. K. 2008. Complementarities in Language. Beijing: Commercial Press. Halliday, M. A. K. and C. M. I. M. Matthiessen. 1999. Construing Experience Through Meaning: A Language-based Approach to Cognition. London: Continuum. Halliday, M. A. K. and C. M. I. M. Matthiessen. 2014. Halliday’s Introduction to Functional Grammar (4th ed.). London: Routledge. Hodge, S. 1993. An Introduction to Classical Tibetan. Warminster, England: Aris & Phillips. Janhunen, J. and K. Norbu. 1999. Aspects of Amdo Tibetan segmental phonology. Studia Orientalia Electronica, 85: 249–276. Jäschke, H. A. 1883. Tibetan Grammar (2nd ed.). London: Trübner & Co. Martin, J. R. 1992. English Text: System and Structure. Philadelphia/Amsterdam: John Benjamins. Martin, J. R. 1996a. Metalinguistic diversity: The case from case. In Hasan, R., Cloran, C. and Butt, D. (Eds.), Functional Descriptions: Theory in Practice. Amsterdam/Philadelphia: John Benjamins, 325–374. Martin, J. R. 1996b. Transitivity in Tagalog: A functional interpretation of case. In Berry, M., Butler, C., Fawcett, R. and Huang, G. (Eds.), Meaning and Form: Systemic Functional Interpretations. Norwood: Ablex Publishing Corporation, 229–296. Martin, J. R. 2013. Systemic Functional Grammar: A Next Step into the Theory— Axial Relations (Chinese translation and expansion by Wang Pin and Zhu Yongsheng). Beijing: Higher Education Press.

Axial Argumentation and Cryptogrammar

101

Martin, J. R. and P. R. R. White. 2005. The Language of Evaluation: Appraisal in English. New York: Palgrave Macmillan. Martin, J. R. and D. Rose. 2007. Working With Discourse: Meaning Beyond the Clause (2nd ed.). London/New York: Continuum. Martin, J. R. and D. Rose. 2008. Genre Relations: Mapping Culture. London/ Oakville: Equinox. Martin, J. R., C. M. I. M. Matthiessen and C. Painter. 2010. Deploying Functional Grammar. Beijing: Commercial Press. Teruya, K., E. Akerejola, T. H. Andersen, A. Caffarel, J. Lavid, C. M. I. M. Matthiessen and F. Smedegaard. 2007. Typology of mood: A text-based and systembased functional view. In Hasan, R., Matthiessen, C. M. I. M. and Webster, J. (Eds.), Continuing Discourse on Language: Functional Perspective (Vol. 2). London: Equinox, 859–920. Wang, P. 2012. A systemic functional interpretation of ergativity in Classical Tibetan. In Knox, J. (Ed.), To Boldly Proceed: Papers From the 39th International Systemic Functional Congress, 105–110.

4

Experiential Cryptotypes Reasoning About process type1 Beatriz Quiroz

1. Experiential Clause Types in SFL In Systemic Functional Linguistics (hereafter, SFL), experiential analysis has been extensively associated with the identification of clause types or process types, including material, mental, relational, behavioral, verbal and existential clauses (e.g. Halliday, 1969/1976; Halliday and Matthiessen, 1999). This analysis, also known as ‘transitivity analysis’ (e.g. Halliday, 1994; Halliday and Matthiessen, 2014; Matthiessen, 1995), involves the identification of specific participant roles as well as attending circumstances in the structure of the clause. Over the years, these experiential clause types along with their associated structural functions have been productively used in the analysis of naturally occurring texts, especially in research focusing on specific registers and genres, as well as in research more generally concerned with relations between texts and various social practices, including the study of ideology (e.g. in (critical) discourse studies). This area of experiential lexicogrammar poses a number of challenges for both research in English and across languages. In English, a productive discussion has been initiated by O’Donnell, Zappavigna, and Whitelaw (2008) on the nature of criteria used by researchers in the SFL community world-wide, and particularly the lack of agreement on criteria. This discussion, followed-up by Tucker (2014) and Gwilliams and Fontaine (2015), has pointed to the need for explicit reasoning that deploys grammatical evidence to ground the functional, meaning-making orientation to linguistic description and analysis. SFL research concerned with the cross-linguistic account of process types faces difficulties that are, to a certain extent, comparable to those of practitioners working on English. In available work (e.g., Caffarel, Martin, and Matthiessen, 2004) the criteria on which the establishment of experiential types in different languages is based are not always clear. A valid question that can be raised is the extent to which some of the descriptions of process types across languages are based on conceptual analyses that largely involve ad hoc interpretations of the lexical meaning of relevant word classes (i.e., verbs), and/or the re-interpretation of

Experiential Cryptotypes

103

‘probes’ that are adapted from English accounts in ways that are not necessarily self-evident. In the Hallidayan tradition, these difficulties are often explained in terms of the specific complexity of experiential lexicogrammar and the ‘covert’ nature of the clause patterns involved2 (e.g. Halliday and Matthiessen, 1999, pp. 16, 26; Matthiessen and Halliday, 1997/2009, p. 61 ff.). Thus far, however, neither widely known work on English lexicogrammar (e.g. all four editions of An introduction to functional grammar) nor SFL cross-linguistic work (e.g. Caffarel et al., 2004) have fully addressed the nature of this complexity nor the underlying reasoning at stake.3 The aim of this chapter is to explore the reasoning underpinning the inherent complexity of experiential lexicogrammar in SFL, with particular attention to its implications for the description of experiential clause types across languages. Section 2 focuses on the paradigmatic reasoning underlying SFL descriptive work in general, drawing first on the contributions of Whorf and Gleason, then approaching the specific ways in which experiential clause types have been modeled in SFL, and homing into the specific complexities involved, particularly in relation to the English clause. Section 3 then turns to an illustration of the paradigmatic reasoning under focus, by reviewing key patterns that are relevant to the description of mental processes in Spanish. The implications of the approach adopted for the description of systemic and structural categories will be discussed in the last Section of this chapter.

2. Cryptotypes, Agnation and Proportionalities in SFL Lexicogrammatical Analysis The specific ways in which paradigmatic reasoning has been deployed in SFL descriptive work owes much to Whorf (1945)’s emphasis on the configurative nature of meaningful grammatical categories, as well as to Gleason (1965)’s exploration of paradigmatic relations between sentences. In this section, the notions of ‘covert category,’ ‘cryptotype’ and ‘reactance,’ as first developed by Whorf, are reviewed along with Gleason’s original notion of ‘agnation.’ These contributions will then take us back to the use of proportionalities in SFL as means to access (and model) paradigmatic relations in the description of lexicogrammatical systems in English. This will lay the groundwork for a consideration of the discussion put forward by Davidse (1998) in relation to the specific complexity of experiential clause types in English. 2.1 Whorf’s Contribution In his posthumously published paper “Grammatical categories” (1945), Whorf discusses the need for a systematic method to address language

104

Beatriz Quiroz

description that allows both meaningful generalizations about human language and revealing accounts of language-specific patterns. The first concern aims at an understanding of linguistic phenomena and their fundamental relation to ‘meaning.’ The second refers to Whorf’s general interest in the study of the complex interrelations between language and culture (see specifically Whorf, 1956). Whorf (1945) proposes that grammatical categories must be explored in relation to patterns that are configurative in nature. In order to explain this idea, he introduces the distinction between overt and covert categories in what he establishes as ‘specific descriptive categories’ (p. 10). An overt category is one “having a formal mark which is present (with only infrequent exceptions) in every sentence containing a member of the category” (1945, p. 2). From Whorf’s perspective, overt categories include the ‘classical’ morphological categories, like those recognized in case, number and gender distinctions in most Indo-European languages (1956, p. 105). Nonetheless, he notes that overt categories are not restricted to morphological categories, nor to morphological marking within a given word class. Overt categories may be manifested in a wider structural environment. For example, when a typically overt category such as ‘plural’ in English is not marked morphologically on a noun, i.e. by adding -s or -es to a lexical root, its meaning is likely to be openly manifest elsewhere within the same sentence, e.g. through the absence/presence of determiners (among other patterns).4 Whorf calls those categories that are directly recognizable within the same sentence phenotypes. However, Whorf stresses that there is another kind of category that is very important in linguistic description. These are covert categories that cannot be immediately recognized by a ‘mark’ or by any self-evident patterning within the same sentence. He suggests in fact that these categories only emerge when seen in light of patterns across sentences. Whorf exemplifies a covert category as follows: In English, intransitive verbs form a covert category marked by lack of the passive participle and the passive and causative voices; we cannot substitute a verb of this class (e.g. go, lie, sit, rise, gleam, sleep, arrive, appear, rejoice) into such sentences as It was cooked, It was being cooked, I had it cooked to order. An intransitive thus configuratively defined . . . is a true grammatical class marked by these and other constant grammatical features, such as non-occurrence of nouns or pronouns after the verb; one does not say I gleamed it, I appeared the table. (Whorf, 1945, p. 2) As Whorf later points out, the fact that the category ‘intransitive verb’ can be shown to be covert does not mean that ‘the same verb’ cannot be used both intransitively and transitively. The category itself is not defined

Experiential Cryptotypes

105

by the specific verb selected, but rather by a recurrent patterning found across associated sentences. Whorf calls grammatical categories established in this way cryptotypes. As already noted, both ‘overt’ phenotypical and ‘covert’ cryptotypical categories are ‘configurationally’ defined. What distinguishes cryptotypes from phenotypes is that the former can only be accessed through a ‘distinctive treatment’: A covert category is marked, whether morphemically or by sentencepattern, only in certain types of sentence and not in every sentence in which a word or element belonging to the category occurs. The classmembership of the word is not apparent until there is a question of using it or referring to it in one of these special types of sentence, and then we find that this word belongs to a class requiring some sort of distinctive treatment, which may even be the negative treatment of excluding that type of sentence. This distinctive treatment we may call the reactance of the category. (1945, p. 2, emphasis in the original) The reactance of a grammatical category is thus a characteristic recurrent pattern that emerges when contrasting configurations across sentences. Put another way, reactances ‘reveal’ a covert category by consistently showing an opposition between sentences. In Whorf’s view, phenotypes and cryptotypes are equally important for language description. For him, they may reveal important commonalities and differences in the ways in which different languages make sense of the world: a given comparable pattern may be phenotypic in one language and cryptotypic in another. Among other things, Whorf raises questions on comparable grammatical patterns that are phenotypic in non-Indo-European languages but cryptotypic in languages like English (or vice versa), suggesting an interesting area of cross-linguistic research. Importantly for the discussion of this chapter, he stresses that both phenotypic and cryptotypic categories need to be configuratively established; that is, both imply looking at broader grammatical patterning rather than isolated elements in structure—a tendency he criticizes not only in ‘traditional’ approaches to language description but also in descriptivist (e.g. Bloomfieldian) approaches being developed at the time. Whorf’s ideas, including the notion of reactance, are certainly crucial for an understanding of covert grammatical categories as semantically and experientially relevant in the broader environment of the sentence. However, his examples still mostly concern morphological classes and word classes that may be comparable (or not) across languages. In order to fully understand the reasoning underpinning descriptive work on experiential clause types in SFL, the important notion of agnation, as modeled by Gleason, needs to be considered.

106

Beatriz Quiroz

2.2 Gleason’s Contribution At a time when transformational grammar was emerging with a strong emphasis on syntagmatic relations at the expense of the paradigmatic, Gleason (1965) suggested that the descriptive focus on patterns within sentences covers only one aspect of linguistic relations in structure. A full description of sentence structure, in his view, necessarily needs to account for contrastive relations between sentences. Gleason begins his reasoning with the observation that some sentences share a special kind of ‘constructional identity,’ which he refers to as enation. This kind of identity can be established based on (1) the classes of units recognizable within sentences ‘at equivalent places,’ and (2) the kind of construction, within the sentence, where those units are (syntactically) related. He offers two examples of sentences sharing their constructional make-up—examples (1) and (2): (1) The dog bit the man. (2) The cat ate the canary. Sentences (1) and (2) can be considered enate because they offer the same ‘parsing’ result—that is, they involve the same classes (‘parts of speech’), and the same relations between those classes in the sentences in which they appear (Gleason, 1965, p. 197). Gleason’s enate sentences are thus related through structural identity: their internal constituents have comparable relations among one another within each sentence. Gleason however goes on to problematize this kind of structural identity per se in an adequate account of grammatical relations by considering a third sentence in relation to examples 1 and 2: (3) The man was bitten by the dog. From the viewpoint of structural identity, (3) is clearly different from both (1) and (2). But Gleason notes that there is a sense in which (1) and (3) do seem to be related—as opposed to (2). This kind of relation, he suggests, can be stated in terms of the ‘meaning’ of the sentences concerned, but cannot be derived simply from the fact that they share the same basic lexical items—since an alternative sentence such as The man bit the dog, having the same lexical items as (3), would not be said to be related to (1) and (3) in the same way in which they are interrelated with one another. Moreover, regardless of the criteria or combination of criteria that an analyst could produce in trying to account for the relatedness of sentences (1) and (3), Gleason argues that the relation between both ultimately rests “in just one feature of the two sentences” (1965, p. 198)—in other words, the possibility of active/passive alternation. For this kind of contrastive pattern relating two sentences consistently, Gleason coins the term agnation (1965, p. 199).

Experiential Cryptotypes

107

Two sentences are thus agnate when they reveal a recurrent contrastive relation—in the previous discussion, examples (1) and (3) are agnate sentences. As a bidirectional relation, the representation of agnation is independent from the sequential ordering of the contrasted sentences. For this reason, Gleason represents agnation by means of a double-headed arrow, as in example (4): (4) The dog bit the man. ↔ The man was bitten by the dog. In example (5), the order of agnate sentences is reversed, but the nature of the paradigmatic relatedness between sentences remains the same: (5) The man was bitten by the dog. ↔ The dog bit the man. According to Gleason, agnation relations between sentences are as grammatically significant as enation relations, and they are equally pervasive in the linguistic system. In other words, two agnate sentences are as grammatically related as two enate sentences are. Now, it is not the case that any set of contrasts found by the analyst would count as agnation. In fact, the only way to ‘test’ for agnation relations between two sentences is in relation to enation relations across sentences. In Figure 4.1, vertically arranged sets represent sentences that are in principle grouped by enation (structural similarity). However, these sentences can be considered enate only to the extent they can be equally related horizontally to sentences across vertically arranged sets, that is, if they all share the same basic contrast. Thus sentences in Figure 4.1 can be classified as ‘transitive’ if there are both (syntagmatically) enate and (paradigmatically) agnate. For Gleason agnation and enation are thus mutually defining aspects of grammatical relations, and they in fact presuppose each other (1965, p. 199, footnote 2). Structural similarity is not enough, in its own right, to establish a systematic relation between enate sentences; rather, identity

Figure 4.1 The Interplay of Agnation and Enation Across Sentences

108

Beatriz Quiroz

depends on agnation-based ‘paradigmatic identity’ and vice versa. Gleason shows this through two additional examples, (6) and (7), which look similar from the point of view of their structural patterning: (6) The man saw a stranger. (7) The man seemed a stranger. Sentences (6) and (7) can be shown to be both structurally and paradigmatically unrelated when their rather diverging agnation/enation patterning is systematically explored, as shown in Figure 4.2 (1965, p. 203).

Figure 4.2 Potential Relations Between Sentences in Examples (6) and (7), Based on Gleason (1965)

The un-relatedness of examples (6) and (7) is evidenced in contrasts between kinds of sentences rather than between words classes within them (e.g. the ‘type of verb’). The analysis in Figure 4.2 also shows that grammatical identity may need to be established through more than one enation/agnation pattern—Figure 4.2 shows at least two different agnation sets are at stake. Ultimately, a full exploration of agnation/enation relations associated with groups of sentences can reveal more specific or ‘delicate’ distinctions within a given category—between sentences that are agnate in some respects but not in all respects (1965, p. 205). Given this crucial interdependency, Gleason concludes that the term agnation alone is enough to account for this mutually defining interaction between structural and contrastive patterns. Gleason’s agnation relations are not only descriptively significant, but they are also crucial from a methodological point of view. Thanks to the systematic interactions involved, any contrasts that are not truly significant between pairs of sentences can be ruled out. As Davidse (2011) later suggests, Gleason’s notion of agnation allows for a principled approach to so-called ‘alternations’ in grammatical analysis, particularly when these are used to account for the ‘meaning’ of the grammatical categories at stake.5 Drawing on Gleason’s work, we can now reconsider Whorf’s cryptotypic categories in terms of agnation relations. We saw previously that Whorf’s cryptotypes need to be accessed through reactances emerging in contrasts between sentences. In Gleason’s terms, generalizations across

Experiential Cryptotypes

109

sentences and their specific recurring structuring can be systematically established through agnation relations that are contrastive in nature. But Gleason (1965) takes a step further, by moving beyond the account of the meaning of morphological and word classes toward the meaning of sentences. This allows him to show that paradigmatic relations between sentences are not only pervasive in the linguistic system, but that they are also of grammatical significance. Importantly, agnation-based reasoning helps him to put forward a descriptive methodology that can systematically show whether sentences are related to one another, and if they are, in what respects. 2.3 Toward Experiential Clause Types in SFL We can now turn to the paradigmatic reasoning underlying the ways in which (meta)functionally motivated clause systems have been argued for in SFL. Accounts of clause systems are best known in relation to English lexicogrammar. It makes sense then to begin our discussion with the crucial role of Halliday’s (1985, 1994) proportionalities in SFL grammatical analysis, before addressing the implications of this perspective for crosslinguistic descriptive work. In the preface to the second edition of his grammar of English, Halliday commits very openly to the idea that “there are no meanings waiting around to be encoded; the meaning is created in language” (Halliday, 1994). In Halliday’s view, however, determining ‘semantically significant’ grammatical categories is far from a straightforward enterprise. Proportionalities are the key to revealing the distinctions explicitly embodied in lexicogrammar, as they show recurrent patterns that would be otherwise inaccessible. The role of proportionalities in grammatical reasoning can be first illustrated with respect to interpersonal grammatical categories such as clause moods in English. In SFL accounts, lexicogrammatical mood types are ‘semantically relevant’ distinctions that can be related to the specific ways in which speakers interactively negotiate roles and semiotic commodities in dialogue. For example, declarative and interrogative clauses in the mood system are congruently related to the negotiation of information, as opposed to the negotiation of goods and services in imperative clauses (e.g. Halliday, 1984). Simultaneously, each of these clause types congruently embody the kind of interactive orientation that can be adopted by the speaker: giving (information) vs. demanding (information). The difference between these clause types is manifest in—technically, is realised by—specific configurations of the English clause structure. Figure 4.3 illustrates this through one proportionality: recurrent similarity among declarative clauses is shown vertically within the same column, and their recurrent contrast with polar interrogative clauses is shown horizontally across columns.

110

Beatriz Quiroz

Figure 4.3 One Proportionality Relevant to the Distinction Between Declaratives and (Polar) Interrogatives in English

From top to bottom, the clausal proportionality in Figure 4.3 can be read as follows:6 John has seen the play is to Has John seen the play? as They will build the house is to Will they build the house? (etc.) In this proportionality, the English declarative clauses share a key structural pattern, even if they differ from one another in terms of tense and modality selections, as well as in the kind of lexical items involved. This shared structural pattern involves a nominal element (in bold) followed by a verbal element (underlined), and both elements, in specific environments, may also enter into agreement relations. The significance of this pattern for the specific interpersonal distinction at stake can only be appreciated with reference to the recurrent contrast across clause types—since it turns out that (polar) interrogatives in English reverse the sequence of the prior elements.7 This interpersonally significant contrast in the English mood system holds true regardless of other similarities and differences that the same clause pairings distributed in rows from top to bottom in Figure 4.3 may display with respect to other systems (e.g. experiential and/or textual clause systems). The proportionality-based reasoning underpinning interpersonal clause types in English does not seem to be the same deployed for the account of experiential clause types. Functionally relevant distinctions seem to involve grammatical interactions of a different kind. These interactions are often addressed in SFL literature in terms of the specifically complex, ‘covert’ and ‘cryptotypical’ nature of experiential clause types.

Experiential Cryptotypes

111

Davidse (1998) addresses this problem by re-interpreting Whorf’s distinction between phenotypic and cryptotypic categories in SFL terms. She notes that relevant phenotypic categories can be better seen as ‘overt’ if considered in terms of a ‘realizational’ definition, that is, in terms of how they can be recognized in clause structure. In English, for example, the Subject function may be considered an ‘overt’ category to the extent that its identification in interpersonal clause structure is relatively straightforward: i.e. by looking at the nominal group establishing agreement relations with the finite verb realizing the Finite, its relative positioning with respect to the Finite in different clause moods and its recoverability in interrogative tags. However, Davidse agrees with Halliday (1984/2002) in that such overt categories might still be very elusive in terms of a ‘value’ definition, i.e. in relation to their meaning in a system (Davidse, 1998, p. 284). With respect to categories such as ideational clause types (e.g. ‘goaloriented’ transitives vs. ‘descriptive’ transitives, in her discussion), Davidse claims that they are elusive both from the viewpoint of their realization (how the category can be recognized in structure) and their value (the meaning or systemic contrast the category can be related to) (1998, p. 285). In other words, these categories are difficult to establish in terms of their ‘outward’ manifestation in structure as well as in terms of the contrast they represent in an underlying lexicogrammatical system. One key reason for this difficulty is that ideational clause types, as modeled in SFL, are established based on sets of configurational patterns organized into what Davidse calls agnation paradigms. In terms of our discussion this means, first, that the recognition and the meaning of process types involves a cluster of agnation patterns (‘probes’) that are distinctive for each clause type. For example, the distinction between material and mental clause types in English involves the possibility of projection in the latter not the former, but it also involves the different nature of participants in each, and the likelihood of passive alternation in each, and the potential for unmarked present tense selections (simple present in the former, present continuous in the latter) and so forth (e.g. Halliday and Matthiessen, 2014). Overall, each contrast in the system of process type represents a generalization about whole sets of agnation patterns. These sets, or agnation paradigms, “are presumed to share crucial semantic features pertaining to different [ideational] types” (Davidse, 1998, p. 293, emphasis in the original). Second, ‘single’ agnation patterns within such sets may overlap across process types, thus defining intermediate clause types (e.g. behavioral processes, which in English share some agnation patterns with material processes and some others with mental processes). Interlocking agnation paradigms shaping experiential clause distinctions can be graphically illustrated by the diagram in Figure 4.4, which has been slightly adapted from Martin (1996b).8

112

Beatriz Quiroz

Figure 4.4 Interlocking Agnation Relations Underlying Main Distinctions in process type (adapted from Martin, 1996a, p. 365)

There are at least two key implications of the complex interaction of agnation relations outlined in Figure 4.4. First, as Davidse (1998, 2011) argues, understanding the specific kind of complexity underpinning ideational clause types is crucial for a systematic analysis of what is conceptualized in other relevant (functional) frameworks as semantically relevant ‘schemas’ or ‘constructions.’ If configurational patterns organized in agnation paradigms are used as the basis of the classification into clause types, the analyst does not need to resort to informal (i.e. notional) analyses of ‘verb types,’ for which construction types are then brought in as a second step. Second, and probably most importantly, the kind of paradigmatic reasoning underlying these categories in SFL, as Davidse (1998) notes, does more than provide criteria to identify clause types in text analysis; it also represents a particular kind of heuristic for interpreting the experiential meaning of clausal configurations in a given language (p. 293). This heuristic is one of the most important contributions of Halliday’s work to the study of meaning, as introduced in the 1960s (Halliday, 1967a, 1967b, 1968). 2.4 PROCESS TYPES as Cryptotypes As explored thus far the sense in which process types are ‘covert’ or cryptotypic in SFL can be more clearly explained in terms of the complex patterning of paradigmatic relations underlying the experiential

Experiential Cryptotypes

113

distinctions they represent. In fact, when looked at more closely, the ‘probes’ used in the description and recognition of process types in English more often than not combine agnation relations across metafunctions (groupings of systems motivated by contextual functions), across strata (groupings of systems at different levels of semiotic abstraction) and across ranks (groupings of systems for different units within lexicogrammar); for discussion of the systemic organisation of different dimensions in SFL theory, see Martin, Wang, and Zhu (2013) and Matthiessen (2015). Exploring agnation relations thus presupposes exploring the specific ways in which systems of relations are ‘internally’ organized in a language, and specific ways in which such systems are interconnected within the overall linguistic system. The agnation-based reasoning underpinning current SFL accounts of process type distinctions in English is relevant both to cross-linguistic work as well as to the application of such work to text analysis. Descriptively, establishing meaningful experiential clause types and associated structural functions in a given language necessarily involves the exploration of sets of clause-wide configurational patterns that can be related to features in systems. In text analysis, such configurational patterns can be productively used (and shared) as criteria for the recognition of clause types in naturally occurring texts. In the next section, the implications of the reasoning reviewed thus far for SFL cross-linguistic work will be illustrated in more concrete terms through the exploration of relevant interrelations in Spanish experiential lexicogrammar.

3. An Illustration in Another Language: Mental Clauses in Spanish As we have seen, both systemic and structural categories associated with experiential lexicogrammar are motivated by complex interrelations across clause structures and/or across clause systems. As pointed out by Halliday, interrelations in the general linguistic system are language-specific and should be described while taking this specificity into account (Halliday, 1992/2003, 1996/2002). The explicitly paradigmatic and agnationbased reasoning required for the account of these interrelations will now be illustrated through key structural and systemic patterns associated with mental clauses in Spanish. SFL descriptive work centered on cryptogrammatical patterns across languages has suggested that the specific nature of elements associated with the Process in the structure of mental clauses is critical for distinguishing them from other process types (i.e. Davidse, 1991/1999 on English; Martin, 1996b on Tagalog; Quiroz, 2013 on Chilean Spanish). Mental clauses necessarily involve a Participant endowed with consciousness, the Senser; and they often include a second element that is

114

Beatriz Quiroz

‘processed’ by sensorial perception, cognition or emotive reaction (Halliday, 1969/1976, p. 166). In examples (8) to (11), this second element fulfils the structural function of Phenomenon:9 (8) A Alan

lo = v-eía

sm Alan

cotidianamente en el campus San Joaquín

acc.3sg = see- daily impf.1sg

in the campus San Joaquín

Phenomenon Process/Senser Circumstance

Circumstance

(ad) nom gr

prepositional phrase

verbal group

(FP)

adv group

“I used to see Alan every day at the San Joaquin campus.” (9) Esta chica [[que se = mov-ía con mucho desplante]]

me = asust-ó (FP)

This girl [[rel refl = move-impf.3sg with much self-confidence]]

dat.1sg = scare-pst.3sg

Phenomenon

Senser/Process

nominal group

verbal group

“This girl who would go around with so much self-confidence scared me.” (10) Desde los doce años

comprend-ió

las diferencias sociales

(FP)

Since the twelve years understand-pst.3sg the differences social Circumstance

Process/Senser

Phenomenon

prep phrase

verbal group

nominal group

“Since she was twelve years old she understood social differences.” (11) Pato

escuch-aba

música clásica

Pato

listen-impf.3sg music classical

Senser

Process

nom gr verbal group

(FP)

Phenomenon nom group

“Pato used to listen to classical music.”

As discussed in Martin and Quiroz (this volume), Participants in Spanish clause structure can be realized alongside and/or inside verbal groups. As the Senser in examples (8), (9) and (10), a Participant may be realized inside the verbal group alone, either through pronominal clitics or the inflectional morphology of the finite verb. Alternatively, Participants may be realized only by nominal groups, as the Phenomenon of (10) and (11).10 Very frequently, however, Participants are simultaneously realized both by nominal groups and co-referential verbal group resources, as is the case with the Phenomenon in (8) and (9), and the Senser in (11).11 For the sake of simplicity, in cases where the Participant is realized both ways, only the nominal group realization has been

Experiential Cryptotypes

115

assigned a functional label in the examples. In those cases where the Participant is exclusively realized by verbal group resources, its corresponding label is shown as conflated with the Process (marked by ‘/’). As can be seen, the sequence in which experiential functions can appear in the Spanish clause is not as restricted as in English due to the fact that the ordering of constituents is not constricted by mood choices, as it is in English (Quiroz, 2018). The reactance that can be used in Spanish to check the ‘conscious’ or ‘conscious-like’ nature of a Participant involves agnation with a clause where this element can be realized by the so-called ‘salient’ personal pronouns within the nominal group,12 as shown in examples (8a) to (11a) for the Senser (in bold): (8) a. A Alan

yo lo = v-eía

sm Alan I

cotidianamente en el campus San Joaquín.

acc.3sg.m = daily see-impf.1sg

in the campus San Joaquin

“I used to see Alan every day at the San Joaquin campus.” (9) a. Esta chica [[que se mov-ía con mucho desplante]] This girl [[rel move-impf.3sg with much self-confidence]]

me = asust-ó

a mí

dat.1sg = scare-pst.3sg

sm 1sg

“This girl [[who would go around with so much self-confidence]] scared me.” (10) a. Desde los doce años Since the twelve years

ella comprend-ió

las diferencias sociales

she

the differences social

understandpst.3sg

“Since she was twelve years old she understood social differences.” (11) a. Él escuch-aba

música clásica

he listen- impf.3sg music classical ‘He used to listen to classical music’

In agnate examples (8a) to (11a) above, the Senser is realized by (salient) pronouns yo, mí, ella and él. This kind of realization is extremely rare in the case of non-conscious Participants, which would be largely realized by demonstratives (or pronominal clitics) instead.13 This reactance is important for the distinction of mental (and behavioral) clauses, since only these clauses require at least one Participant construed as ‘conscious’ or ‘conscious-like.’ The element that is processed by the Senser is far less restricted than the nominal group realizing the Senser itself. As seen in the previous examples, the Phenomenon characteristically involves the construal of

116

Beatriz Quiroz

a wide range of phenomena of differing natures. It may be construed as a Phenomenon in a very concrete physical realm, including entities that are conscious or non-conscious, animate or inanimate, concrete or abstract—such as a Alan (‘Alan’), las diferencias sociales (‘the social differences’) y música clásica (‘classical music’) in the previous examples. But this element can be also more complex in nature, as in examples (12)–(15): (12) [. . .] sent-ían

el agua [[subiendo despacio || hasta inundarlo todo]]

feel-impf.3pl

the water [[rise.grnd slowly || until flood.inf everything]]

Process/Senser

Phenomenon

verbal group

[[non-finite clause complex]

(FP)

“They felt the water [[rising slowly || until everything was flooded]].” (13) Me = carg-a

[[hacer transbordo]]

|| no me = gust-a

dat.1sg = displease-prs.3sg

[[do.inf transfer]]

|| not dat.3sg = please-prs.3sg

Senser/Process

Phenomenon

|| Senser/Process/ Phenomenon

verbal group

[[non-finite clause]]

|| verbal group

“I hate [[doing transfers (e.g. from bus to subway)]].” (14) Me = asust-ó

(FP)

|| “I don’t like it.”

[[que est-uviera enojado]] (CE)

acc.1sg = scare-pst.3sg that be-pst.sbj.3sg angry Senser/Process

Phenomenon

verbal group

[[finite clause]]

“(It) scared me that he was angry.” (15) Sig-an sabiendo

|| que más temprano que tarde se abrirán las grandes alamedas. (FP)

keep-prs.3pl know.grnd

|| that more early than late open-fut.3pl the grand avenues

Process/Senser

|| => β (finite projected clause)

verbal group (complex) “Keep on knowing || => that sooner than later the grand avenues will open.”

The elements construing what is brought about by the Senser’s consciousness in examples (12) to (15) are clearly of a different nature, as

Experiential Cryptotypes

117

they are realized by clauses rather than by nominal groups. These clauses may be embedded and non-finite, as in (12) and (13), and so involving a gerund or an infinitive verb within the verbal group respectively. Or they can be embedded and finite, as in (14), where the verbal group shows distinctions in tense, verb mood, modality and person. In example (15), the finite clause construing what is processed by the Senser is not treated as part of the experiential structure of the main mental clause, but is rather (hypotactically) projected as a separate clause in a clause complex (a recurrent pattern associated with cognition mental processes, as we will discuss later).14 For Halliday and Matthiessen (1999), these possibilities indicate that lexicogrammatical resources can construe experiences of different kinds. From a discourse-semantic perspective (Halliday and Matthiessen, 1999; Hao, 2015, forthcoming), nominal groups tend to (congruently) construe entities—‘micro’ phenomena in the lexicogrammar. Embedded and projected clauses, on the other hand, construe discourse-semantic figures—macrophenomena or metaphenomena in the lexicogrammar.15 These clauses have been systematically associated in English accounts with the realization of acts (embedded macro-phenomena), facts (embedded meta-phenomena) and ideas (projected meta-phenomena) (e.g. Halliday, 1985). One key configurative pattern in mental clauses is precisely their recurrent association with macro- and metaphenomenal clauses of the kind exemplified in (12)–(15). In other words, mental clauses involve much wider possibilities for the construal of ‘phenomenality,’ as opposed to other clause types (as seen in Figure 4.4). In SFL cross-linguistic work, acts, facts and ideas have been found to recur typically with distinct mental subtypes: macrophenomenal acts tend to pattern with perception mental processes, metaphenomenal facts with reaction mental processes, and metaphenomenal ideas with cognition mental processes (Halliday, 1985; Martin, 1996b; Quiroz, 2013). This does not mean they are exclusively and necessarily associated with one subtype. In English descriptions, for example, fact-clauses have been also found in cognition mental processes (and they appear recurrently realizing Participants in relational clauses as well). Since macrophenomenal clauses in Spanish perception clauses show very distinctive patterns that have been fully explored elsewhere (Martin et al., forthcoming; Quiroz, 2013), the discussion will specifically address here metaphenomenal clauses patterning with cognition and reaction mental clauses. Seen in general terms all metaphenomenal clauses in Spanish look very similar, since they are all realized by que-clauses (comparable to thatclauses in English). Being metaphenomenal, idea-clauses and fact-clauses additionally share one key reactance: they are agnate with neuter substitute forms, either the accusative pronominal clitic lo (‘it’) or the neuter

118

Beatriz Quiroz

demonstratives eso (‘that’) or esto (‘this’).16 Agnate clauses (15) to (16) show these substitute forms in bold in (b) through (c): (15) a. Sig-an sabiendo keep-prs.ind.3pl know.grnd

|| => que más temprano que tarde

se abrir-án

that more early than late

open-fut.ind.3pl

‘Keep on knowing || => that sooner than later las grandes alamedas

(FP)

the grand avenues the grand avenues will open’ b. Sigan sabiéndo = lo keep-prs.ind.3pl know.grnd = acc.n “Keep on knowing it.” c. Sigan sabiendo

esto

keep-prs.ind.3pl know.grnd this.n “Keep on knowing this.” (16) a. El director [. . .] The director

tem-ía

[[que se fuera a malinterpretar políticamente]] (CE)

fear-impf.ind.3sg [[that is go-pst.sbj.3sg misinterpret.inf politically]]

“The director feared [[that it would be going to be misconstrued politically]].” b. El director

lo = temía

The director acc.n = fear-impf.ind.3sg “The director feared it.” c. El director

temía

eso

The director fear-impf.ind.3sg that.n “The director feared it.”

While the substitute-form agnation suggests that the metaphenomenal clauses are comparable clause constituents, several other patterns suggest that the behavior of idea-clauses is different from that of fact-clauses. The key point here is that ideas are propositions directly projected by the main cognition clause involved, unlike embedded facts (Halliday and Matthiessen, 2014, p. 516, 537). As projected propositions, ideas allow for interpersonal distinctions in the interpersonal system of speech function (i.e. despite their ‘bound’ status). The implication of this is that idea-clauses display a wide range of possibilities for interpersonal

Experiential Cryptotypes

119

deixis (distinctions in person, tense, modality) when compared to factclauses. Examples (17) to (19) show idea-clauses projected by cognition mental processes in Spanish (in bold): (17) Uno p-uede pensar

=> que dismin-uye

One may-prs.ind.3sg think.inf

el riesgo

that decrease-pst. the risk ind.3sg

“You may think => the risk decreases || pero es

al revés (FP)

|| but

other.way.around

be-pst.3sg

|| but it is the other way around.” (18) ¿Sab-ías

=> que hoy en Chile

know-prs. ind.2sg

se = p-uede determinar

that today in Chile se = can-prs.ind.3sg determine.inf

“Did you know => that today in Chile the magnitude of an earthquake can la magnitud de un temblor

en seis minutos? (FP)

the magnitude of an earthquake

in six minutes?

be established in six minutes?” (19) [. . .] rec-uerda => que p-uedes denunciar informar = te remember-imp

o

that can-prs.ind.2sg report.inf or inform.inf = refl.2sg

en: 800104 00 (FP) in: 80010400

“Remember => that you can report or get informed at 80010400.”

Examples (17) to (19) show that in Spanish the verbal groups of projected ideas favor the selection of the so-called ‘indicative verb mood’ at word rank (glossed as ‘ind’ in examples throughout). The implications of this are two-fold: verbal groups selecting for ‘indicative’ verb mood admit the full range of selections in tense, person and modality; as a result, projected ideas can select for all these interpersonal systems, including polarity, independently.17 Embedded fact-clauses on the other hand typically involve the selection of ‘subjunctive’ verb mood within the verbal group (glossed as sbj in examples (14) and (16)). Accordingly their potential for tense and modality distinctions at group rank are quite restricted. Subjunctive inflectional morphology does allow the basic distinction between ‘present’ and ‘past’; but this distinction is not strictly comparable to primary tense selections made in projected clauses. As can be seen in example (13), if the verbal group realizing the Process in the reaction clause selects for primary present (indicative), then the selection in the fact-clause is ‘present subjunctive.’

120

Beatriz Quiroz

When the Process selects for primary past (indicative verb mood), the fact-clause selects for ‘past subjunctive,’ as in example (14) and (16). However, if the Process of the reaction clause selects for ‘future,’ the fact-clause still needs to select for ‘present subjunctive,’ as in the agnate clause (16d): (16) d. El director

tem-erá

The director fear-fut. ind.3sg

[[que se = vaya a malinterpretar [[that se = go-prs.sbj.3sg lnk misinterpret.inf

políticamente]] politically]]

“The director will fear [[that it will be going to be misconstrued politically]].”

Since projected ideas associated with cognition clauses allow for a greater range of interpersonal distinctions, they are, unlike fact-clauses, open to debate—in other words, they are arguable. This means that in dialogic exchange projected ideas may replay the polarity and, in some cases, the modality of a projected proposition, as shown in bold in the agnates of examples (15), (17) and (19) displayed in (b): (15) a. Sig-an sabiendo

=> que [. . .] se abrirán las grandes alamedas (FP)

keep-prs.sbj.3pl that the grand avenues will open know.grnd b. Sig-an sabiendo

=> que sí

keep-prs.sbj.3pl know.grnd that yes’ “Keep on knowing => that yes.” (17) a. Uno p-uede pensar One

b. Uno p-uede pensar One

=> que disminuye el riesgo

may-prs.ind.3sg think.inf => that the risk decreases => que no

may-prs.ind.3sg think.inf

that no

“You may think => that no.” (19) a. Rec-uerda => que puedes denunciar o informarte en 800104 00 remember-imp => that you can report or get informed at 80010400 b. Rec-uerda => que tal vez that maybe remember-imp “Remember => that maybe”

The possibility of replaying the embedded proposition in the factclause of reaction processes is much less likely:18 (14) a. *Me = asust-ó [[que sí ]] acc.1sg = scare-pst.3sg [[that yes]] *“(It) scared me [[that yes]].”

Experiential Cryptotypes

121

(16) e. *El director tem-ía [[que no]] The director fear-impf.ind.3sg [[that no]] *“The director feared [[that yes]].”

The restricted interpersonal deixis shown by fact-clauses (i.e. within the limited range of subjunctive verb mood), locate them in between temporal deixis proper and what has been referred to sometimes as the ‘realis/irrealis mood’ contrast (e.g. Palmer, 2001, p. 1ff). In other words, in finite fact-clauses the distinction between reality phase and temporality is ‘blurred’ by the subjunctive verb mood. Also, subjunctive fact-clauses do not allow the selection of ‘conditional verb mood,’ which in indicative clauses admits further modality distinctions through verb morphology. These more restricted patterns allow a better understanding of why facts are interpreted as ‘pre-projected’ propositions. Being metaphenomenal in nature, they are not propositions directly brought about by the conscious processing construed by the reaction configuration; their existence is independent from the construal of the reaction itself. In other words, they are modeled as pre-existing metaphenomena impinging upon the Senser’s consciousness, rather than brought into existence by a cognitive mental processing (Halliday and Matthiessen, 2014, p. 543ff). Based on the lexicogrammatical patterns explored thus far in Spanish, it makes good sense to treat them as part of the multivariate structure of the clause. The possible insertion of el hecho de (‘the fact of’) brings out their down-ranked status more clearly. Naturally occurring examples of reaction clauses including el hecho de are shown in (20) and(21): (20) P-uede divertir = nos

el hecho [[de [[que el reconocimiento por distinción

can-prs.ind.3sg amuse-inf = the fact [[of [[that the recognition by dat.1pl distinction “[[The fact that the recognition of distinction is not but recognition of conformity]] no sea [. . .]

sino un reconocimiento por conformidad]] ]]

not be-prs.sbj.3sg

but a recognition by conformity.

(CE)

may amuse us.” (21) Me = sorprend-ió

el hecho [[de [[que el secretario del Tesoro afirm-ase [. . .]

dat.1sg = surprise-pst.ind.3sg the fact [[of [[that the secretary of Treasure affirm-prs.sbj.3sg “[[The fact that the Secretary of the Treasury asserted ‘there is something ‘Hay algo extraño en el debate actual’]] ]]’ have-prs.ind.3sg something strange in the debate current’ strange in the current debate’]] surprised me.”

(CE)

122

Beatriz Quiroz

Restricted interpersonal deixis in fact-clauses is also evidenced when the modal responsibility is presupposed by the main reaction clause. When the modally responsible person in both the main clause and the metaphenomenal fact coincides, the fact-clause is necessarily a non-finite clause (with an infinitive verb in the verbal group). The embedded nonfinite clauses in example (13), repeated below, and examples (22)–(23) can be preceded by el hecho de (‘the fact of’). (13) Me = carg-a dat.1sg = displease-prs.3sg

[[hacer transbordo]]

|| no me = gust-a

[[do.inf transfer]]

|| not dat.3sg = please-prs.3sg

(FP)

“I hate [[doing transfers (e.g. from bus to subway)]], || I don’t like it.” (22) Me = alegr-ó

[[sentir => que [[lo que hac-emos]] [. . .] sí tiene eco]] (FP)

dat.1sg = make.happy-prs.3sg [[feel.inf that [[what do-prs.1pl]] yes have-prs.3sg echo]] “[[Feeling that what we do has indeed an impact]] made me glad.” (23) No nos = asust-a not dat.1pl = scare-prs. ind.3sg

[[limpiar baños

tomar una brocha,

cortar pasto

[[clean.inf bathrooms take.inf a cut.inf paint.brush grass

“[[Cleaning bathrooms, taking a paint brush, cutting the lawn or taking the garbage out]] o sacar basura [. . .]]]

|| cuando el pago es

digno

or take.out.inf garbage]]

|| when

be-prs.3sg decent

the pay

(FP)

does not scare us || when the pay is decent.”

As pointed out by Davidse (1991, p. 349), what all of these patterns reveal is that the (interpersonal) arguability potential of the metaphenomenal fact-clauses that frequently appear in reaction processes is much more restricted than that for idea-clauses appearing in cognition mental clauses. As seen already in the examples (14) and (16), the proposition embodied by the embedded fact is ‘shielded off’ from debate, unlike projected ideas. The distinctive configurational patterns associated with metaphenomenal ideas and metaphenomenal facts are summarized in Table 4.1. Table 4.2 summarizes the patterns reviewed in this section, which specifically deal with the nature of the elements associated with the Process in the structure of Spanish mental clauses. These patterns, along with many others not addressed here (see Quiroz, 2013 for a full account) are part of a bundle of reactances underpinning the agnation paradigm that is relevant to mental clauses in Spanish.

Experiential Cryptotypes

123

Table 4.1 Main Reactances for Facts and Ideas in Mental Clauses HYPER-PHENOMENALITY

FACT-CLAUSE [[META-PHENOMENON]]

PROJECTED IDEA

typical mental subtype

reaction

cognition

finiteness: interpersonal deixis

restricted

unrestricted

finiteness: verbal group

non-finite (infinitive) or finite (subjunctive)

finite (indicative)

finiteness: presupposed modal responsibility

if co-referential with main reaction process

—

arguability

—

que sí, que no, etc.

tracking reference

group rank: neuter accusative clitic clause rank: neuter demonstrative

META-PHENOMENON

Table 4.2 Patterns Associated With the Nature of Participants in Mental Processes PARTICIPANT-HOOD

REACTION

COGNITION

Senser: inherent

conscious or conscious-like

conscious or conscious-like

Phenomenon: inherent

unspecified phenomenal hyperphenomenal: [[embedded fact]]

unspecified phenomenal hyperphenomenal: → projected idea

Agnation relations illustrated here have been shown to be useful not only for the identification of elements of structure in text analysis but also as criteria for the distinction of clause types and subtypes in the Spanish system of process type.

4. Conclusion Description of experiential clause types is a challenging area for SFL research. As reviewed in O’Donnell et al. (2008), even when comprehensive (English) descriptions are available, there are no unified criteria in the analysis of process types across the community of researchers. In cross-linguistic work, descriptive and methodological divergence in process type analysis makes it difficult for individual researchers to develop new descriptions or build on the ones previously developed for a given language or a language family. As a result, there is not yet a principled

124

Beatriz Quiroz

approach available to the systematic comparison and contrast of experiential systems across languages. The point of departure for the discussion in this chapter was the idea that at the core of these challenges lies the complex interrelations underpinning experiential grammatical categories in SFL. It has been proposed that the key to address this complexity in the account of process types across languages is the deployment of an explicit paradigmatic reasoning. In the account of meaningful experiential distinctions, it has been shown that this reasoning involves the simultaneous consideration of a number of agnation-based patterns (a series of ‘probes’ in agnation paradigms), which may overlap across systemic distinctions (i.e., process types and subtypes) and/or across associated functions in structure. Careful consideration of the language-specific interaction of clause distinctions and structural patterns has been illustrated in this chapter through a review of the relevant patterns associated with Spanish mental processes. In particular, the nature of participants as a key criterion to access experiential cryptotypes has been reviewed through an agnation-based reasoning that involves the exploration of a series of Spanish-specific contrastive patterns. This review has shown that generalized criteria, such as the nature of participants and the potential for projection, are not necessarily grounded on the same agnation relations (or ‘proportionalities’) in Spanish and in English (see e.g. Halliday, 1967a, 1967b, 1976; Davidse, 1991; Halliday and Matthiessen, 2014). A question that can also be raised is the extent to which such ‘generalized’ criteria are necessarily valid across languages (cf. Matthiessen, 2004 ff.). Put another way, other useful patterns may emerge as more agnation-based descriptions of experiential clause types are developed across languages. Some other key questions that have not been addressed here include the nature and weight of ‘semantic’ criteria in the description (see Hao, this volume; Martin and Quiroz, this volume) and/or an account of experiential types across languages. As is well known, SFL theory includes competing models of the semantic stratum (e.g. Hao, 2015, forthcoming, vs. Halliday and Matthiessen, 1999; see also Taverniers, forthcoming, for a discussion of the theoretical nature of the semantic stratum in SFL). Different models of semantics are bound to influence the semantic criteria used to motivate experiential clause types ‘from above.’ The study of the relevant interactions between process types and higher level patterns would thus be crucial to address the question of the relevant criteria and agnation-based patterns that should be privileged in cross-linguistic work.

Notes 1. The theoretical and descriptive discussion presented in this chapter was originally part of doctoral studies conducted between 2009–2012 at the University

Experiential Cryptotypes

2. 3.

4.

5. 6. 7.

8.

9.

10.

11.

12.

125

of Sydney; the theoretical and empirical research extending this work has been supported by CONICYT-FONDECYT through grant N°11170674. For an approach to this discussion from the perspective of the Cardiff Grammar, see Fawcett (2009, 2011). Notable exceptions include early work by Halliday on English experiential grammar in the 1960’s (Halliday, 1967a, 1967b) and Davidse’s key contribution to the expansion of such work (e.g. Davidse, 1991/1999). See later passages for her specific contributions to the discussion put forward in this chapter. “In fish appeared the absence of any article denotes plural, in the fish will be plentiful a pluralizing adjective denotes it, in the Chinese arrived and the Kwakiutl arrived, the definite article coupled with lack of a singular marker like person, Chinaman, or Indian denotes plural. In all these cases plural is overtly marked, and so with few exceptions are all noun plurals in English, so that noun-plural is an overt category in English” (Whorf, 1945, p. 2) Compare with the discussion put forward by Lamb (1964, p. 106), who argues that ‘alternations’ in their own sake do not have necessarily any ‘deep’ grammatical significance. See also the discussion in Davidse (2011). See Nesbitt and Plum (1988, pp. 13–14) for a similar way of framing grammatically relevant proportions in SFL. These key elements at stake have been re-interpreted by Halliday as Subject and Finite. But these interpersonal structural functions in English are not only motivated by the systemic distinction between declarative and interrogative. Other syntagmatic patterns—e.g. tagging and potential for passive voice— are also relevant throughout the English mood system (see Martin, Wang, and Zhu, 2013 for a step-by-step exploration of the reasoning involved). For an example of interpersonal functions motivated by language-specific syntagmatic and paradigmatic clause relations throughout the mood system, see the account proposed Quiroz (2017, 2018) for Spanish. The diagram is part of the discussion put forward by Martin (1996a) stressing the key role of configurative clause patterns in SFL descriptive work—as opposed to work focusing on isolated classes (e.g. Fillmore’s work on cases made manifest by nominal groups, or lexicase work focusing on verb types). Naturally occurring examples used in this chapter, aligned on a group-bygroup basis, come from two sources: (1) texts from Chilean Spanish collected as part of the data used by the FONDECYT-11170674 research project (FP), and (2)) Chilean Spanish examples selected from the Web/Dialects online database in Corpus del Español (CE) (www.corpusdelespanol.org). These examples are modified as needed to show various agnation relations. As in these specific examples, Participants in Spanish may be realized by adpositional nominal groups preceded by the structural marker a—here glossed as ‘sm’ to avoid confusion with prepositions proper. See Quiroz (2013) for the full reasoning underlying the distinction between adpositional nominal groups and prepositional phrases in Spanish (pp. 201–203, 315). This simultaneous realization along the rank scale, including so-called ‘agreement’ and ‘clitic doubling,’ is a key reactance distinguishing Participants from Circumstances in Spanish. For an in-depth review of the specific ways in which Participants are realized in Spanish clause structure (with specific reference to Chilean Spanish), see Quiroz (2013), and Martin, Quiroz, Wang and Zhu (forthcoming). In Spanish descriptive work, the label ‘tonic’ is traditionally used to distinguish these pronouns from non-salient pronominal clitics, whose domain is the verbal group (and which are still treated by many reference grammars as ‘atonic’ pronouns).

126

Beatriz Quiroz

13. If we try agnate clauses using a salient personal pronoun for the Phenomenon in (10) and (11), we get the ungrammatical examples *Desde los doce años las comprendió a ellas (*“Since he was twelve years old she understood them”), and *Pato la escuchaba a ella (*“Pato used to listen to her”). These agnates would only be acceptable if the Phenomenon was a human or conscious-like Participant. No neuter personal pronoun comparable to ‘it’ is available in Spanish (the pronoun ello ‘it’ is only rarely used, and mostly for text reference). 14. For a full discussion on the distinction between projection and embedding in SFL, see Matthiessen and Martin (1991). 15. The cover term ‘hyper-phenomenal’ has been sometimes used for both macroand metaphenomenal clauses in SFL literature (e.g. Matthiessen, 1995, p. 258ff). 16. With the caveat that this pattern applies specifically to metaphenomenal clauses in like-type (or emanating) structures, as in examples (15) and (16). However, in please-type (impinging) clauses illustrated in (13) and (14), the fact-clause is the modally responsible Participant, thus co-referential with the inflectional morphology of the verbal group realizing the Process. The please-type pattern is the most typical one in Spanish reaction clauses, being another piece of evidence contributing to their distinction from perception and cognition mentals, which largely prefer the like-type pattern (see Quiroz, 2013). 17. In this respect, projected ideas in Spanish are very close to simple (unbound) indicative clauses in the system of mood (Quiroz, 2018). 18. In the Web/Dialects database of Corpus del Español, elliptical forms such as que sí-que no are very clearly associated in Spanish with cognition mental processes—with verbs such as pensar (‘think’), creer (‘believe’), parecer (‘seem’), estimar (‘consider’)—or verbal processes projecting metaphenomenal locutions—and co-occurring with verbs such as decir (‘say’), responder (‘reply’), criticar (‘criticise’), asegurar (‘assure’) and informar (‘inform’), among others. For the distinction between (hypotactically) projected ideas and locutions, see Halliday (1985, 1994) and Halliday and Matthiessen (2014).

References Caffarel, A., Martin, J. R., and Matthiessen, C. M. I. M. (Eds.). (2004). Language typology: a functional perspective. Amsterdam: John Benjamins. Davidse, K. (1991/1999). Categories of experiential grammar (vol. 11, Monographs in systemic linguistics). Nottingham: Department of English and Media Studies, Nottingham Trent University. Davidse, K. (1998). Agnates, verb classes and the meaning of construals: The case of ditransitivity in English. Leuvense Bijdragen, 87(3–4), 281–313. Davidse, K. (2011). Alternations as a heuristic to verb meaning and the semantics of constructions. In P. Guerrero Medina (Ed.), Morphosyntactic alternations in English: functional and cognitive perspectives (pp. 11–37). London: Equinox. Fawcett, R. (2009). Seven problems to beware of when analyzing processes and participant roles in texts. In D. Slembrouck, M. Taverniers and M. van Herreweghe (Eds.), From will to well: studies in linguistics offered to Anne-Marie Simon-Vandenbergen (pp. 209–224). Gent: Academia Press. Fawcett, R. (2011). Problems and solutions in identifying processes and participant roles in discourse analysis (Part I: Introduction to a systematic procedure for identifying processes and participant roles). Annual Review of Functional Linguistics, 3, 33–87.

Experiential Cryptotypes

127

Gleason, H. A. (1965). Linguistics and English grammar. New York: Holt, Rinehart and Winston. Gwilliams, L., and Fontaine, L. (2015). Indeterminacy in process type classification. Functional Linguistics, 2(8), 2–8. Halliday, M. A. K. (1967a). Notes on transitivity and theme in English: part 1. Journal of Linguistics, 3(1), 37–81. Halliday, M. A. K. (1967b). Notes on transitivity and theme in English: part 2. Journal of Linguistics, 3(2), 199–244. Halliday, M. A. K. (1968). Notes on transitivity and theme in English: part 3. Journal of Linguistics, 4(2), 179–215. Halliday, M. A. K. (1969/1976). Types of process. In G. Kress (Ed.), Halliday: system and function in language (pp. 159–173). Oxford: Oxford University Press. (Reprinted from: Unpublished paper written as a Nuffield Program Work paper, 1969). Halliday, M. A. K. (1984). Language as code and language as behaviour: a systemic functional interpretation of the nature and ontogenesis of dialogue. In R. Fawcett, M. A. K. Halliday, S. Lamb, and A. Makkai (Eds.), The semiotics of culture and language (vol. 1, Language as social semiotic, pp. 3–35). London: Frances Pinter. Halliday, M. A. K. (1984/2002). On the ineffability of grammatical categories. In J. Webster (Ed.), On grammar (vol. 1, Collected Works of M.A.K. Halliday, pp. 291–322). London: Continuum. (Reprinted from: A. Manning, P. Martin and K. McCalla (Eds.). (1984). The tenth LACUS forum, pp. 3–18). Halliday, M. A. K. (1985). An introduction to functional grammar (1st ed.). London: Edward Arnold. Halliday, M. A. K. (1992/2003). Systemic grammar and the concept of a ‘science of language’. In J. Webster (Ed.), On language and linguistics (vol. 3, Collected Works of M.A.K. Halliday, pp. 199–212). London: Continuum. Halliday, M. A. K. (1994). An introduction to functional grammar (2nd ed.). London: Edward Arnold. Halliday, M. A. K. (1996/2002). On grammar and grammatics. In J. Webster (Ed.), On grammar (vol. 1, Collected Works of M.A.K. Halliday, pp. 384–417). London: Continuum. Halliday, M. A. K., and Matthiessen, C. M. I. M. (1999). Construing experience through meaning: a language-based approach to cognition. London: Continuum. Halliday, M. A. K., and Matthiessen, C. M. I. M. (2014). Halliday’s introduction to functional grammar (4th ed.). London: Routledge. Hao, J. (2015). Construing biology: an ideational perspective. Unpublished PhD Dissertation, University of Sydney, Sydney, Australia. Hao, J. (forthcoming). Analysing scientific discourse from a systemic functional linguistic perspective: a framework for exploring knowledge building in Biology. London: Routledge. Lamb, S. (1964). On alternation, transformation, realization and stratification. Monograph Series on Languages and Linguistics, 17, 105–122. Martin, J. R. (1996a). Metalinguistic diversity: the case from case. In R. Hasan, C. Cloran and D. Butt (Eds.), Functional descriptions: theory in practice (pp. 325–374). Amsterdam: John Benjamins. Martin, J. R. (1996b). Transitivity in Tagalog: a functional interpretation of case. In M. Berry, C. Butler, R. Fawcett, and G. Huang (Eds.), Meaning and form: systemic functional interpretations (pp. 229–296). Norwood: Ablex Publishing Corporation.

128

Beatriz Quiroz

Martin, J. R. and Matthiessen, C. M. I. M. (1991). Systemic typology and topology. In F. Christie (Ed.), Literacy in Social Processes: papers from the inaugural Australian Systemic Linguistics Conference, held at Deakin University, January 1990 (pp. 345–383). Darwin: Centre for Studies in Language in Education, Northern Territory University. Martin, J. R., Quiroz, B., Wang, P., and Zhu, Y. (forthcoming). Systemic functional grammar: another step into the theory – grammatical description. Beijing: Higher Education Press. Martin, J. R., Wang, P., and Zhu, Y. (2013). Systemic functional grammar: a next step into the theory—axial relations. Beijing: Higher Education Press. Matthiessen, C. M. I. M. (1995). Lexicogrammatical cartography: English systems. Tokyo: International Language Sciences Publishers. Matthiessen, C. M. I. M. (2004). Descriptive motifs and generalizations. In A. Caffarel, J. R. Martin, and C. M. I. M. Matthiessen (Eds.), Language typology: a functional perspective (pp. 537–664). Amsterdam: John Benjamins. Matthiessen, C. M. I. M. (2015). Halliday on language. In J. Webster (Ed.), The Bloomsbury companion to M. A. K. Halliday (pp. 137–202). London: Bloomsbury. Matthiessen, C. M. I. M., and Halliday, M. A. K. (1997/2009). Systemic functional grammar: a first step into the theory. Beijing: Higher Education Press. Matthiessen, C. M. I. M., and Martin, J. R. (1991). A response to Huddleston’s review of Halliday’s introduction to functional grammar. Occasional Papers in Systemic Linguistics, 5, 5–74. Nesbitt, C., and Plum, G. (1988). Probabilities in a systemic grammar: the clause complex in English. In R. Fawcett and D. Young (Eds.), New developments in systemic linguistics (vol. II, Theory and application, pp. 6–38). London: Pinter. O’Donnell, M., Zappavigna, M., and Whitelaw, C. (2008). A survey of process type classification over difficult cases. In C. Jones and E. Ventola (Eds.), From language to multimodality: new developments in the study of ideational meaning (pp. 47–64). London: Equinox. Palmer, F. R. (2001). Mood and modality (2nd ed.). Cambridge: Cambridge University Press. Quiroz, B. (2013). The interpersonal and experiential grammar of Chilean Spanish: towards a principled systemic-functional description based on axial argumentation. PhD Dissertation, University of Sydney, Sydney, Australia. Quiroz, B. (2017). Gramática interpersonal básica del español: una caracterización sistémico-funcional del sistema de MODO. Lenguas Modernas, 49, 157–182. Quiroz, B. (2018). Negotiating interpersonal meanings: reasoning about MOOD. Functions of Language, 25(1), 135–163. Taverniers, M. (2019). Semantics. In G. Thompson, W. Bowcher, L. Fontaine and D. Schönthal (Eds.), The Cambridge handbook of systemic functional linguistics (pp. 55–91). Cambridge: Cambridge University Press. Tucker, G. (2014). Process types and their classification. In K. Kunz, E. Teich, S. Hansen-Schirra, S. Neumann, and P. Daut (Eds.), Caught in the middle—language use and translation (pp. 401–416). Saarbrücken: Saarland University Press. Whorf, B. L. (1945). Grammatical categories. Language, 21(1), 1–11. Whorf, B. L. (1956). Discussion of hopi linguistics. In J. B. Carrol (Ed.), Language, thought and reality: selected writings of Benjamin Lee Whorf (pp. 102–111). Cambridge, MA: The MIT Press.

5

Axial Argumentation and Cryptogrammar in Textual Grammar theme in Brazilian Portuguese Giacomo Figueredo

1. Theme Semiotic Address Structural resources realizing theme build the clause as a wave of information (Pike, 1982; Halliday, 2002) in the grammar, leading up to comparable wave patterns in whole texts (Martin, 1992). When presenting information, such structures give prominence to certain elements, which in turn determine how discourse flows. For this reason, this chapter takes into account both interpretations of theme (1) from the vantage point of the clause; and (2) from the vantage point of discourse. From the vantage point of the clause, Theme is characterized as “the point of departure for the message. It is the element the speaker selects for ‘grounding’ what he is going on to say” (Halliday, 1994, p. 34). It is fundamental to note that ‘point of departure’ is not a positional notion of ‘what comes first’ in the clause. So, Theme is not what comes first (analogous to a Process is not a verb; or a Subject is not a noun). Rather, the notion of ‘point of departure’ leads to an organization such that Theme is the basis from which the clause is to be interpreted. Theme is “the resource for manipulating the contextualization of the clause; it is the resource for setting up a local context for each clause in a text” (Matthiessen, 1995, p. 531). Consequently, textual organization has a different nature from that of ideational and interpersonal organization. Theme conflates with ideational and interpersonal functions to give them thematic prominence. Reasoning along these lines, a ‘message’ is a semantic unit; in broad terms it can be explained as a ‘contextualized clause in the text.’ Thus the question we need to ask to identify the Theme is “what is the point of departure of this clause?” Theme is realized by position in Brazilian Portuguese (henceforth BP), and so its identification is related to structure. The different kinds of selections for theme features reveals their valeur; thematic choice is thus related to system. In order to identify the Theme we need to ask “where do waves in the flow of discourse begin, change and end?” and from a discourse perspective we ask “what causes shifts in the flow to happen?” When vantage points from both the clause and discourse are taken into account, the possibilities for identifying Theme structures and systems open up to a trinocular account of theme.

130

Giacomo Figueredo

We treat this description of theme trinocularly. We begin in the next section by viewing theme from above in order to contextualize its patterns in discourse. In Section 2 theme will be viewed from below so Theme functions will be related to their realizing structures in lower ranks. Finally, in Section 3 it will be viewed from roundabout, and subsystems of theme will be described in more detail.

2. Approaching From Above: Discourse Semantics In this section, we will draw upon Martin (1992) and Martin and Rose (2007) with respect to the textual organization of discourse (especially periodicity). Despite being a clausal system in the grammar, theme contributes fundamentally to text production (Halliday, 2002, p. 369). Thus we start our approach from above by relating the system of theme—a grammatical system of the clause in BP—to discourse semantics—the stratum above grammar. Because language strata operate together to make meaning, approaching theme from above—i.e., as preselections of discourse semantics functions—increases the description in its comprehensiveness. 2.1 Discourse Flow and Periodicity Looking at discourse from above—i.e., from context—the enabling of ideational meanings (sequence of activities in the field) and interpersonal meanings (negotiations in the tenor) is realized in discourse semantics as text. This enabling defines text as a configuration of meanings determined by field and tenor. Textually, for every new orientation (Martin and Rose, 2007, p. 191) to the field or tenor there is generally a new wave in the discourse flow, causing discourse to either follow the orientation or, conversely, to cause a shift in the flow. Every new wave in the flow— either following or shifting orientation—is introduced by a message with the specific function of realizing this follow or shift orientation. In turn, message functions are realized in the grammar (i.e., the stratum below) by theme. As discourse unfolds, messages accumulate, extending over greater lengths in the flow (Martin, 1992, p. 436). This process of accumulation from one message to many and the development from one quantum to many is defined textually as ‘wave behavior’ (Pike, 1982; Halliday, 1978; Martin and Rose, 2007). Its thematic organization is enabling in nature and is oriented toward enabling the development of field and tenor. Accumulation means extending the principle of waves to ever growing groups of messages. “Little waves” (realized by the function of Theme) accumulate together into “bigger waves” (realized by the function of hyperTheme), which in turn accumulate into “tidal waves” (realized by the function of macroTheme) (cf. Martin and Rose, 2007). These form hierarchies in periodicity. Martin and Rose (2007, p. 188) explain that

Axial Argumentation and Cryptogrammar

131

“periodicity is concerned with information flow: with the way in which meanings are packaged.” The process of accumulation often reaches a culminating point, in which the information from the waves is distilled into the core meaning of the wave (realized by a New, a hyperNew and a macroNew respectively) (Figure 5.1). Here we show an example of textual organization and the accumulation of meaning in hierarchies of periodicity. Text 1 is a procedure—an excerpt from the SEBRAE (an institution to support small businesses) handbook on making flour, extracted from our reference corpus CALIBRA (cf. Figueredo, 2014). Text 1’s purpose is to give prominence to the steps of bagging flour. Major clauses are numbered and Theme is marked in bold. A translation of Text 1 into English is also provided. Because Theme is realized by position in structure, we will keep the Brazilian clause structure (and group structure when needed) in the translation and mark the structures that would be Theme in BP in bold. Being a procedure, Text 1 is organized around sequencing activities (field) and recommending the readers to control their behavior (tenor). It starts with message 1—the title of the text, ‘Bagging.’ Message 1 is the initial message of the whole text and it performs the function of predicting the accumulation of all information in the text as it sets off the flow of discourse. Thus message 1 is the macroTheme of the whole text.

Figure 5.1 A Model of Discourse Flow

132

Giacomo Figueredo

Text 1 [XPL_EM_16] (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)

Ensacamento Antes do ensacamento a farinha deve ser classificada. A classificação será conforme a demanda do mercado consumidor. Para isto existem conjuntos de peneiras de diferentes tamanhos de malha. Durante a classificação o movimento da peneira gera grande quantidade de pó. Nesta etapa é indispensável o uso de equipamento de proteção respiratória. A farinha poderá ser acondicionada em sacos de ráfia com capacidade para 50Kg ou em pacotes de 1Kg, esta última preferencialmente deve ser feita através de ensacadeiras automáticas que evitam o manuseio do produto acabado. Para os sacos com 50Kg existem dois tipos de procedimento: enchimento mecânico e fechamento com costuradeira elétrica e enchimento e costura manual. Nesta etapa a geração de resíduos sólidos é proveniente das embalagens danificadas, que devem ser acondicionadas em fardos para uma posterior reutilização ou venda. No enchimento dos sacos existe a emissão de particulados. Nesta etapa deve-se garantir uma boa ventilação no local e, também, recomenda-se o uso de equipamento de proteção respiratória. Como nas etapas anteriores deve ser observado o uso correto da energia elétrica.

Text 1 [English] (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)

Bagging Before bagging the flour must be classified. Classification depends on consumer market demands. For this there are sets of sieves with different mesh sizes. During classification sieving produces a great amount of dust. In this step it is fundamental to use respirator masks. The flour can be put in raffia bags of 50 or 1Kg. The latter must preferably be filled using automatic baggers, which avoid touching the finished product. For the 50Kg bags there are two procedures: mechanical bagging and machine sewing or manual bagging and sewing. In this step solid waste generation results from damaged bags, which latter can be recycled or sold. During bagging there is emission of particles. For this step a good ventilation must be provided to the bagging area and also we recommend the use of respirator masks. As in all previous steps the correct use of electricity must be observed.

Next, textual organization gives prominence to the sequences of activities by separating them into two waves starting at messages 2 and 7. Messages 2 and 7 begin the contextualization of sequences of activity and culminate with the same behavior control from the tenor. The periodicity of Text 1 and its accumulation of information can be divided as: ‘little waves’ (messages from 1 through 14) accumulating into ‘big waves’ (messages 1; 2–4; 5–6; 7–8; 9–10; 11–13; 14), accumulating into ‘bigger

Axial Argumentation and Cryptogrammar

133

waves’ (messages 2–6; 7–13), accumulating into a ‘tidal wave’ (messages 1–14). Figure 5.2 models the discourse flow in Text 1. In Figure 5.2, common terms ‘little, big, bigger and tidal’ are replaced by their technical counterparts ‘wave, hyperwave, macrowave 2nd level and macrowave 1st level.’ Accordingly, the messages performing the function of initiating each wave are labeled as Theme, hyperTheme and macroTheme. Further explanation will be given later (for the original and comprehensive source, cf. Martin and Rose, 2007). In Figure 5.2 we can see how textual organization accumulates information in Text 1 through periodicity. It is important to remark that in this chapter—differently from Martin and Rose (2007)—we use the term ‘macrowave’ for ‘tidal waves’ whose initial message performs the function of a macroTheme. A ‘macrowave,’ consequently, consists of a (macroTheme) + (following messages composing the body of the wave). There are three macrowaves in Text 1. Macrowave 1 includes the whole text, from messages 1 through 14. Macrowave 2, which is a second level macrowave, includes messages from 2 through 6. Macrowave 3, again a second level macrowave, includes messages from 7 to 13.

Figure 5.2 Discourse Flow in Text 1

134

Giacomo Figueredo

The function of macroThemes of messages 2 and 7 demarcate second level macrowaves of periodicity in the discourse. Message 2 “Before bagging the flour must be classified”1 contextualizes the activities of ‘preparing for bagging.’ Message 7 “The flour can be put in raffia bags of 50 or 1Kg” contextualizes the activities of ‘bagging itself.’ Interpersonally, the prominence is given to recommendations to the reader, contextualized in the end of each wave. Message 6 “In this step it is fundamental to use respirator masks” is the last message of macrowave 2. It is not part of the activity sequence of ‘preparing for bagging,’ but a recommendation to control the reader’s behavior. Messages 12 and 13 operate similarly in the end of macrowave 3. Finally, message 14 “As in all previous steps the use of electricity must be observed” culminates the whole discourse flow, distilling the information from macrowave 1. Each of these macrowaves are themselves the accumulation of ‘big waves,’ for which we use the term ‘hyperwave.’ A hyperwave textualizes the ideational meanings of an activity from the sequence. Again we are introducing a different terminology from Martin and Rose (2007) to define a ‘hyperwave’ as the group of accumulating messages of a (hyperTheme) + (following messages composing the body of the wave). The activity sequence of macrowave 2 ‘preparing for bagging,’ involves ‘classification’—initiated by message 2, functioning as hyperTheme 1— and a specific aspect ‘during classification’—initiated by message 5 “During classification sieving produces a great amount of dust,” functioning as hyperTheme 2. Hyperwaves do not necessarily culminate and not always include a hyperNew, as is the case here. An important development from this to Text 1 is related to the textualization of interpersonal meanings. Because Text 1 shifts orientation from ‘commanding’ to ‘safety’ in accumulating messages, hyperwaves 1 and 2 have shifts in the discourse flow caused by ideational meanings related to the sequence of activities in the field—but not to different negotiations in the tenor. A similar pattern can be seen within macrowave 3. There are 3 hyperwaves related to the activities in the sequence. For ‘bagging itself,’ there is ‘bagging’—initiated by message 7, functioning as hyperTheme 3; ‘different types of bagging’—initiated by message 9 “For the 50Kg bags there are two procedures,” functioning as hyperTheme 4; and ‘during bagging’—initiated by message 11 “During bagging there is emission of particles,” functioning as hyperTheme 5. 2.2 The System of CONTEXTUALIZATION The model of periodicity exemplified earlier enables us to understand the textual discourse organization of Text 1, which can be summarized as follows:

Axial Argumentation and Cryptogrammar

135

1. The smallest unit of textual discourse semantics is the message. 2. In the flow of discourse, the meanings of messages accumulate to form bigger units of meaning—i.e., waves. This accumulation is the result of the textual property of ‘wave behavior’ assigned to messages. 3. Waves accumulate to form hyperwaves, and hyperwaves accumulate to form macrowaves. 4. Accumulation of meaning is organized by the textual metafunction— more specifically by its enabling property. Enabling means bringing together ideational meanings and interpersonal meanings into a seamless flow of discourse. In Text 1, for example, the meanings of ‘bagging flour’ and ‘controlling the reader’ were brought together. 5. Because the textual metafunction enables ideational and interpersonal meanings, the flow of discourse accommodates shifts in activity sequences from the field and different negotiations from the tenor. In Text 1, for example, there are two sequences of activities divided by a shift in the discourse flow (starting at message 7). There are also two negotiations culminating in each macrowave. Messages can have different discourse functions. These functions, in turn, form the structure of a wave, shaping the flow of discourse. Firstly, a message can set off the discourse flow by generating a wave. As a result, there is a mandatory message function in every wave of discourse, so (1) messages can start the flow of discourse—or a new wave within the flow— and function as macroTheme. Also, a message can operate to distill the core meanings accumulated in the flow, so (2) messages can culminate the wave and function as macroNew (and less frequently as hyperNew, depending on the text). Instead of initiating or culminating, messages can, alternatively, be part of the body of the wave. In between the initial and the culminative message there can be a given number of messages varying proportionally to sequences of activities and negotiations. So (3) messages can simply follow the preceding messages in the wave and keep the flow functioning as continuing messages. Finally, messages can be part of a wave but change the course of discourse flow by either enabling ideational changes between activities in a sequence and/or enabling interpersonal changes in types of negotiation. So (4) messages can introduce shifts of ideational and interpersonal perspective within the wave—due to gaze shifts to the field and tenor (cf. Martin and Rose, 2007) and function as hyperTheme. Message functions—like any other functions—realize features of systems. As a result, by establishing the agnations among these functions, we can posit a system for messages. We will call the system contextualization, since it includes messages in the discourse flow (cf. Matthiessen et al., 2010, p. 106). Figure 5.3 presents contextualization with examples extracted from macrowave 2 of Text 1.

136

Giacomo Figueredo

Figure 5.3 The System of contextualization in BP

2.2

THEME

and CONTEXTUALIZATION

As mentioned earlier, message functions preselect the clause system of theme. Message functions are realized by preselection in the stratum below—the grammar. Most frequently, message functions are realized by the clause. Thus we can examine how these functions map onto Theme function-structures. Going through the features of contextualization and their correlating clause textual function-structure, we can correlate choices in contextualization and theme.2 First, examining textual clause structure motivation from above shows evidence that Theme is realized structurally in BP—more specifically by the element in first position. This is due to the fact that, for every selection in contextualization, there is a change in clause structure—more specifically related to the element in first position in the clause. In Text 1, for example, every contextualization function of messages selected in the flow of discourse causes a change in the type of element in the first position of clauses realizing messages. This is mainly due to the behavior of discourse flow called continuity. Martin and Rose (2007, p. 215) explain that: “interaction of discourse systems in a text tells us a lot about what’s continuing and what’s changing, from one phase to the next. . . . We’ll refer to meanings that continue as ‘continuity’, and meanings that change as ‘discontinuity.’” The discourse semantics of message, cohesion and lexical chains interact in the text and, as a result, different types of Theme work complementarily to manage continuity (Table 5.1). The approach from above is important to show how the point of departure is connected to discourse, and how Theme choices are correlated to initiating, continuing, discontinuing or culminating discourse. As Table 5.1 shows, Theme needs to perform different functions, both to

Table 5.1 Periodicity in Text 1 message

1 Ensacamento Bagging 2 Antes do ensacamento a farinha . . . Before bagging the flour . . . 3 A classificação será . . . Classification depends . . . 4 Para isto existem . . . For this there are . . . 5 Durante a classificação o movimento . . . During classification sieving . . . 6 Nesta etapa é fundamental . . . In this step it is fundamental . . . 7 A farinha poderá ser . . . The flour can be put . . . 8 Esta última preferencialmente . . . The latter must preferably . . . 9 Para os sacos com 50Kg existem . . . For the 50kg bags there are . . . 10 Nesta etapa a geração . . . In this step solid waste . . . 11 No enchimento dos sacos existe . . . During bagging there is . . . 12 Nesta etapa deve-se . . . For this step a good . . . 13 e, também, recomenda-se o uso . . . and also we recommend the use . . . 14 Como nas etapas anteriores deve . . . As in all previous steps the correct . . .

CONTEXTUALIZATION clause function ideational

interpersonal

initial (macro 1) initial (macro 2 & hyper 1)

nominal group Circ. Time: p.phr

—

continuing

Token:ng

Subject

continuing

Circ. Cause: p.phr Circ. Time: p.phr

—

discontinuing (hyper 2)

—

—

culminative

Circ. Place: p.phr

—

initial (macro 3 & hyper 3) continuing

Goal:ng

Subject

Goal:ng

Subject

discontinuing (hyper 4)

Client:p.phr.

—

continuing

Circ. Place: p.phr.

discontinuing (hyper 5)

Circ. Time: p.phr.

—

culminative

Client:p.phr

—

continuing

Verbiage:ng

—

culminative

Circ. Manner:p. phr.

—

138

Giacomo Figueredo

realize continuity in the flow of discourse and to enable ideational and interpersonal meanings (by conflating with ideational and interpersonal clause functions). As we can see in the analysis of Text 1, a great number of clause functions may be enabled by theme to perform the job of handling discourse flow. Thus a single text example cannot show a clear pattern of one-toone correspondence between discourse and textual grammar functions. Instead, what we can show in this section is that theme is directly connected to discourse in terms of managing the discourse flow locally by offering different realizational options for contextualization. What we could not show here—due to the scope of the chapter—but can be found elsewhere (Figueredo, 2011) is patterned configurations of discourse and textual grammar.

3. Approaching From Below: Group Rank Realizing Theme In the previous section, Table 5.1 shows that discourse flow is realized by the first element in the clause structure, which tends to be realized by elements in lower ranks, such as groups. For example, the nominal group ‘bagging’ realizes the first macroTheme. Consequently, approaching theme from contextualization needs to be complemented by an approach from below in terms of the groups operating as clause elements that realize Theme. The group rank items operating in the clause are also connected to contextualization features. The most frequent patterns—as exemplified by Text 1—are explained as follows. These are supposed to expound on the structures presented in Table 5.1. 1. initial ↓ macroTheme are correlated to different clause functionstructures. Predominantly, these clause structures have a nominal group as the element in first position. If it is the first macroTheme in the text, then [initial]3 tends to occur with systems at group rank. This is commonly seen in titles of texts and sections, e.g., ‘Bagging.’ Example 1 [XPL_EM_16] Ensacamento nominal group Bagging ‘Bagging’

If it is not the first macroTheme in the text, [initial] is realized by clauses with a nominal group in first position. But in this case the nominal group realizes a different Participant ideationally, a different Mood

Axial Argumentation and Cryptogrammar

139

interpersonally and co-instantiates with different lexical items (from different lexical chains). This indicates shifts in the field or tenor. Example 2 [XPL_EM_16] A farinha

poderá ser acondicionada em sacos de ráfia

Themenom.gp

Rheme

The flour

will.can be put

in bags of raffia

“The flour can be put in raffia bags.”

An [initial] can also be realized by clauses with a prepositional phrases as the element in first position. Example 3 [XPL_EM_16] Antes do ensacamento

a farinha

Themeprep.phr

Rheme

Before of.the bagging

the flour

deve ser classficada must be classified

“Before the bagging the flour must be classified.”

2. culminative ↓ macroNew—In this situation, grammatical Themes occur within clause structures having a nominal group or a prepositional phrase as elements in first position. However, in contrast to [initial] these elements may present a cohesive function to refer to a previous message or to the whole wave. In example 4, ‘nesta (in this)’ refers back to the meanings of the whole wave. Example 4 [XPL_EM_16] Nesta etapa

deve-se garantir

uma boa ventilação

Themeprep.phr Rheme In.this step

must provide

a good ventilation

“For this step good ventilation must be provided.”

3. continuing—Within continuing messages, Themes occur in clauses having a nominal group as the elements in first position, in particular for declarative and material or relational clauses, as is the case of Text 1. The lexical items belong to the same lexical chains as those in previous messages as well. Example 5 [XPL_EM_16] A classificação

será

conforme a demanda

Themenom.group Rheme The classification

will.be

depending the demand

“Classification depends on demand.”

140

Giacomo Figueredo

4. discontinuing ↓ hyperTheme—here, Theme occurs in clauses having prepositional phrases, nominal groups, adverbial groups and verbal groups as the elements in first position. The element realizes clause functions that cause shifts in the flow of discourse, such as a temporal perspective (message 5) or a subclassification (message 9). (As we will see in Section 4, these can also cause other types of shift, such as place, manner, angle, mood and modality, among others). Example 6 [XPL_EM_16] Durante a classificação

o movimento da peneira

Themeprep.phr

Rheme

During the classification the movement of the sieve

gera grande quantidade de pó generates big quantity of dust

“During classification sieving produces a great amount of dust.” Example 7 [XPL_EM_16] Para os sacos com 50kg existem dois tipos de procedimento Themeprep.phr

Rheme

For the bags with 50kg

exist

two types of procedure

“For the 50kg bags there are two procedures”

Examining theme composition from below shows the motivation for selection of elements in first position of the clause. This can be summarized as: [initial]: nominal group alone, different lexis or less frequently prepositional phrases when overlapping with hyperwaves. [culminating]: nominal groups (or less frequently prepositional phrases) that have a cohesive function. [continuing]: nominal groups with lexis continuing a cohesive chain; or less frequently a prepositional phrase with a cohesive function. [discontinuing]: prepositional phrase. In the next section, we will examine theme from roundabout. We will use the evidence from above and below to sort out agnations among thematic choices to gain an understanding in the system organization.

4. THEME From Roundabout In this section we will examine features from the grammatical system of THEME in more detail. We will use examples extracted from other texts of CALIBRA. Since clauses will be plucked from their original texts, they will be decontextualized, so their purpose is only to exemplify the

Axial Argumentation and Cryptogrammar

141

thematic organization. Whenever possible, bigger excerpts of text will be given.4 theme has ‘major clause’ as its entry condition. At the first level of delicacy three simultaneous systems occur: textual theme, interpersonal theme and theme selection. 4.1 The System of TEXTUAL THEME Textual Themes (see Figure 5.4) structure messages in discourse through textual resources. The system has three features: [conjunctive], [continuative] and [relative]. As we can see in the examples, the choice textual:conjunctive involves conjunction groups realizing the Theme. These are mostly deployed within a single hyperwave and conjunct clauses. Example 8—Conjunctive [EXP_FM_09] Mas

vocês

já podem fazer seu cadastro

Textual Th.conj.g Theme But

you

already can make your application

“But you can already make your application.”

Textual:continuative Themes are realized by continuatives and tend to relate a clause to some bigger unit, generally the previous hyperwave. Example 9—Continuative [RMD_EM_14] Também

ø

defendemos soluções socialmente justas

Textual Th.continuative (Theme) Also

(we)

defend

“Also, we defend socially just solutions.”

Figure 5.4 The System of textual theme in BP

solutions socially just

142

Giacomo Figueredo

Finally, textual:relative Themes are realized by relative pronouns and make reference to a Theme from a previous clause. Example 10—Relative [CAP_EM_12] A comissão

apresentará relatório que

Theme

Rheme

The commission will.present a report

será publicado no Diário do Senado

Theme: Rheme Relative which

will.be published

in Journal of Senate

“The commission will present a report, which will be published by the Senate Journal.”

4.2 The System of INTERPERSONAL THEME interpersonal theme (see Figure 5.5) involves thematic prominence for modal assessments of propositions and the discursive roles they enact in the text. Modal assessment is realized interpersonally through Adjuncts of Polarity, Mood and Commentary. The different types of assessment conflate with interpersonal Themes to realize [evaluation] as the interpersonal point of departure in the clause. From a discourse perspective, this is how theme manages the enabling domains of evaluation, including meanings associated with attitude within appraisal (cf. Martin and White, 2005 for a general overview), which are conflated with the Interpersonal Theme when thematized.

Figure 5.5 The System of interpersonal theme in BP

Axial Argumentation and Cryptogrammar

143

Example 10—Polarity [RMD_FM_16] Não

tem

amostra grátis

Interpersonal Theme Theme Not

exist

sample free

“There aren’t free samples.” Example 11—Modality [EXP_FD_06] Provavelmente

isso

não tem valor jurídico

Interpersonal Theme Theme Probably

this

not

have value legal

“Probably this does not have legal value.” Example 12—Comment [COM_ED_22] Infelizmente

eu

não estava bem

Interpersonal Theme Theme Unfortunately

I

not was well

“Unfortunately, I wasn’t well.”

When the Theme organizes interpersonal meanings associated with enacting discursive roles it selects as a point of departure different forms of interaction between speaker and listener. Interpersonal Themes enable meanings associated with the orientation of evaluation, or the validation of propositions and proposals, as they are speaker oriented. In this case, Theme: Interpersonal conflates with Modal Particles. Example 13—Assessment [REL_EM_26] Ué

agora

você vai lá

Interpersonal Th./Modal Particle Theme Obvious

now

you

go

there

‘It’s obvious that now you must go there’.

These Theme: Interpersonal types co-select with a subsystem of vocative. This system can specify the listener by conflating Theme: Interpersonal with Vocative. Example 14—Vocative [REL_EM_26] Ana Maria,

ø

está novamente com a palavra

Interpersonal Theme (Theme) Ana Maria

(you)

are

“Ana Maria, you have the floor again.”

again

with the word

144

Giacomo Figueredo

4.3 The System of THEME SELECTION The system of selection manages the primary relationship between textual grammar, discourse and context. It is the main resource theme possesses to enable ideational and interpersonal functions. As seen previously (Text 1), texts are ideationally and interpersonally oriented. This means theme must have resources to orient textualization toward ideational and interpersonal meanings. Orientation is the main job performed by theme selection. 4.4

THEME SELECTION:

Absolute

Orientation is often very specific in selecting ideational and interpersonal meanings to conflate with Theme. In general, it is functions from transitivity and mood that are mostly selected. Sometimes, however, orientation can be generalized and not bound by any other clause grammar system. This is the function of Theme called Absolute. This function enables selecting a Theme in BP with only a vague or generic orientation. Accordingly, the first feature in the system of theme selection is the introduction of a lexical item/items realizing some aspect of the context— which in turn is selected to be Theme. This feature is [absolute]—which in turn is realized by the function of Absolute (see Figure 5.6). In this situation, there is no conflation between ideational or interpersonal functions of the clause with the Theme: Absolute. From discourse, [absolute] generally realizes a change in the discourse flow, particularly between waves—which is followed by a shift in activities in the field or in negotiations in the tenor. Example 16—Absolute [REL_FD_34] A apresentação de power point nós Theme: Absolute

vamos ter que ler esse livro

Rheme Actor

Process

Goal

Subject Finite Predicator The power point presentation

we

will have to read

this book

“The power point presentation, we will have to read this book.”

Figure 5.6 [absolute] in BP

Axial Argumentation and Cryptogrammar

145

In example 16, the Theme is ‘apresentação de power point (power point presentation)’. The presentation does not have any ideational or interpersonal functions in the clause; it is not a Participant, Subject etc. Its only function is Theme. It is there only to serve as point of departure and to realize the beginning of a new wave in discourse. 4.5 MOOD ORIENTATION theme in BP has a system to manage the contextualization of interpersonal meanings. This system is responsible for enabling the clause as exchange. It is called mood orientation. The first job done by this system is to decide whether a clause will be interpersonally oriented or not—in other words, whether crucial functions realizing mood types will become the point of departure of the clause by conflating with Theme. If interpersonal orientation is selected, then through mood orientation Theme selection can place interpersonal meanings to shape the flow of discourse (see Figure 5.7). In the grammar of BP, the crucial interpersonal clause functions are: (1) Predicator for imperative clauses; (2) Subject and Finite for indicative clauses; and (3) ‘Qu-’element for interrogative: elemental clauses. (See examples in the next section). mood orientation enables different functions to be the point of departure depending on mood choices. The Predicator is chosen in imperatives. The Subject—responsible for undertaking the proposition— is chosen in declaratives and polar interrogatives (there is no valeur between these two options in relation to mood orientation). The ‘Qu-’ element is chosen in elemental questions. As shown, modal and assessment interpersonal functions are managed by the system of interpersonal theme and are different from mood orientation. The meanings realized by modality, comment, polarity and assessment contribute to state the speakers positioning in a wave of the discourse flow because they are associated with attitude and engagement. However imperative, indicative and interrogative are directly related to

Figure 5.7 mood orientation in BP

146

Giacomo Figueredo

negotiation and speech functions because they realize exchanges and turns between speakers. Thus their crucial functions are selected by the orientation system to conflate with Theme in order to textualize interpersonal shifts and follow in the discourse flow of BP texts. Text 2 illustrates the job of mood orientation—enabling interpersonal functions as point of departure. Text 2 is a gossip text (Eggins and Slade, 1997)—an excerpt from a casual conversation among the members of a family during a dinner table conversation. Text 2 involves four members of the family (speakers S1, S2, S3 and S4). They are talking about their cousin’s daughter and her partner, a teenage couple who recently learned they are pregnant. The four speakers are talking specifically about the teenagers quitting school and getting married. As with Text 1 earlier, Text 2 is also extracted from CALIBRA. A translation of Text 2 into English is provided and Theme is marked in bold. Text 2 [COM_FD_15] (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)

S1 S2 S2 S3 S3 S3 S3 S1 S1 S2 S3 S1 S1 S1 S1 S4

Deixa ela correr atrás. Claro que ela não vai ficar longe da filha dela. Ela vai morar junto com a criança, ué. Ela faz de tudo pela internet pra ficar com o neném no colo, ué. Ela se vira. Ø Faz faculdade pela internet, ué. Não, o neném não vai atrapalhar de estudar não. Fala com ela que ela pode ficar até os 18 anos solteira para ela pensar. Eles tão pensando em casar, S1, os dois? Ele tá querendo. Ele quer, né? Ela não quer nada, né? Ela não sabe o que ela quer, né? Ela não sabe lavar um copo! Ela não faz nada!

Text 2 [English] (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)

S1 S2 S2 S3 S3 S3 S3 S1 S1 S2 S3 S1 S1 S1 S1 S4

Let her figure this out on her own. Of course she will not be away from her baby. She must live with her child, obviously. She can study online so she can take care of her baby, obviously. She must find a way. Ø (she) can take online courses. No, the baby will not cause her to quit school. Tell her that she can stay single at least until she’s 18, so she can figure this out. Are they considering getting married, S1, the both of them? He wants to. He does, right? She doesn’t want anything, does she? She doesn’t know what she wants. She can’t properly wash a single cup. She can’t do anything.

Axial Argumentation and Cryptogrammar

147

Text 2 is a typical gossip in that it fits the most probable generic organization for the genre and accompanying discourse and grammar configurations (cf. Eggins and Slade, 1997 for a comprehensive explanation and Santiago, 2012 for a generic description of gossip in BP). In broad terms, we can see the speakers sharing values and opinions while passing judgment on a third party. The discourse flow in Text 2 has a single macrowave. It has message 1 “Let her figure this out on her own” as an [initiating] macroTheme and message 16 “She can’t do anything” as a [culminative] macroNew. In addition, macrowave 1 is the accumulation of three hyperwaves. The first hypewave starts with message 2 “Of course she will not be away from her baby” and culminates with message 7 “No, the baby will not cause her to quit school.” Hyperwave 2 has message 8 “Tell her” as an [initial] hyperTheme and message 9 as a [continuing] message. Finally, the third hyperwave starts with message 10 “Are they considering getting married . . . ?” up until message 16. In Text 2, shifts in tenor and interpersonal meanings shape changes in discourse flow. In the grammar, different types of interpersonal clauses are selected to realize these changes. The flow starts with an imperative clause (macroTheme 1), then moves to a declarative clause (hyperTheme 1), then to another imperative clause (hyperTheme 2) and finally to an interrogative clause (hyperTheme 3). Textually, theme needs to manage these different negotiations, which is done as follows. The system of mood orientation enables different interpersonal clause types to place their crucial realizing function in the first position of clause structure. The interpersonal meanings of gossip related to speech functions are frequently realized interpersonally by placing: (1) someone with power over the teenagers as the recipient of demand: goods and services in imperative clauses (Let her; tell her); and (2) the third party members—the teenagers and the baby—as Subject (i.e., as modally responsible) of most propositions in the text (she; they; he; the baby). Consequently, imperative clauses have the Predicator conflating with Theme. And most importantly the declaratives have the Subject conflating with Theme, which consequently means the teenagers and their baby are both modally responsible and the point of departure for most of the text. For example the mom-to-be is Subject/Theme 11 times out of 16 clauses. Table 5.2 shows how contextualization is realized by mood orientation clause functions. When compared to Table 5.1—which shows a similar analysis for Text 1, we can see the selection for the features [mood non-oriented] Text 1 vs. [mood oriented] Text 2 is relevant for different types of text. 4.6

TRANSITIVE ORIENTATION

theme in BP also has a system to manage the contextualization of ideational meanings. This system is responsible for enabling the clause as representation. The system of transitive orientation selects an

148

Giacomo Figueredo

Table 5.2 Periodicity and mood orientation in Text 2 message

CONTEXUALIZATION clause function ideational interpersonal

1 Deixa ela correr . . . Let her figure . . . 2 Claro que ela não vai . . . Of course she will not be 3 Ela vai morar . . . She must live 4 Ela faz . . . She can study . . . 5 Ela se vira. She must . . . 6 ∅ Faz faculdade . . . ∅ (she) can take 7 Não, o neném não vai . . . No, the baby will not cause 8 Fala com ela Tell her 9 que ela pode ficar . . . that she can stay . . . 10 Eles tão pensando . . . Are they considering . . . 11 Ele tá querendo He wants to 12 Ele quer . . . He does 13 Ela não quer . . . She doesn’t 14 Ela não sabe . . . She doesn’t know 15 Ela não sabe . . . She can’t properly 16 Ela não faz . . . She can’t do

initial (macroTheme 1) initial (hyperTheme 1)

Process: Material Actor

Predicator

continuing

Actor

Subject

continuing

Actor

Subject

continuing

Actor

Subject

continuing

Actor

Subject

culminative

Actor

Subject

discontinuing (hyperTheme 2) continuing

Process: Verbal Carrier

Predicator

discontinuing (hyperTheme 3)

Senser

Subject

continuing

Senser

Subject

continuing

Senser

Subject

continuing

Senser

Subject

continuing

Senser

Subject

continuing

Actor

Subject

culminative (macroNew 1)

Actor

Subject

Subject

Subject

ideational function of the clause (Process, Participant or Circumstance) as point of departure by conflating it with Theme (see Figure 5.8). In Brazilian Portuguese, this selection follows the principle of directionality together with the degree of involvement in the process (cf. Halliday and Matthiessen, 1999, p. 172 for a general explanation). The concept of directionality describes how the arrangement (positioning in structure) of ideational clause functions realize the quantum of change of a figure. Accordingly, directionality is primarily identified through clause structures. In BP, transitivity functions can be classified as follows:

Axial Argumentation and Cryptogrammar

149

Process (Prs), or the function responsible for realizing a figure’s quantum of change; Processes are key functions realizing clause types such as material, relational, mental, verbal and existential (Pagano et al., 2015). Participant 1 (P1) is the function realizing entities directly involved in the Process—i.e., representing a cause or source for the Process to happen. More specifically in BP, P1 can be an Actor, Carrier, Senser (emanating), Phenomenon (impinging), Sayer (Figueredo, 2011; Alves, 2017; Paula, 2017). Participant 2 (P2) is the function realizing entities impacted or created by the Process, or those which are not the cause or source for the Process to occur. The types of P2 are Goal, Attribute, Existent, Senser (impinging), Phenomenon (emanating), Target, Verbiage. For example: Example 17 [COM_FD_15] Ela

não faz

P1 She

nada

Process P2 not

do

nothing

“She can’t do anything.” Example 18 [REL_FD_34] nós vamos ter que ler esse livro P1

Process

P2

we

will have to read

this book

“We will have to read this book.”

In addition, there can be the function of Participant 3 (P3), which realizes marginal involvement in the Process mainly as beneficiary or receptor. P3s in BP are Recipient, Client, Scope, Attribute (material). Circumstances (C1) are functions even more marginally involved in the Process that can project, extend and elaborate the Process. Functions of C1 are Role, Accompaniment, Matter. Finally, there is yet another type of Circumstance (C2) that is even more marginal to the Process, and their job is to enhance the Process. C2 includes Extension, Location, Manner, Cause and Contingency. For example: Example 19 [XPL_EM_16] Para os sacos com 50kg existem dois tipos de procedimento P3

Process

P2

For the bags with 50kg

exist

two types of procedure

“For the 50kg bags there are two procedures.” Example 20 [XPL_EM_16] Antes do ensacamento a farinha deve ser classficada C2

P2

Process

Before of.the bagging

the flour

must be classified

“Before the bagging the flour must be classified.”

150

Giacomo Figueredo

The ideational structure of the clause in BP allows different types of directionality and realizes the quantum of change of a figure in many different structures. Basically, all functions can be arranged in any order. So, different arrangements of function structures of ideational clause in BP are possible. Below are some possibilities as illustration (though these are some of the possible permutations, but many others are also possible). Ideational functions conflated with Theme are marked in bold:5 P1^Prs^P2^P3^C2 → Ela não faz nada pra mim hoje (“She can’t do anything for me today”). P2^P3^C2^P1^Prs → Nada pra mim hoje ela não faz (“Anything for me today she can’t do”). P3^P1^C2^Prs^P2 → Pra mim ela hoje não faz nada (“For me she today can’t do anything”). C2^Prs^P1^P2^P3 → Hoje não faz ela nada pra mim (“Today can’t do she anything for me”). Prs^P1^P2^P3^C2 → Não faz ela nada pra mim hoje (“Can’t do she anything for me today”). As the examples show, all ideational functions can be enabled by transitive orientation to conflate with Theme and become point of departure. However, there are some preferred ideational structures for the clause in BP that impact on its textual organization. Ideationally speaking (see Table 5.3, Section 4.7) the most probable ideational clause structure in BP is: P1^Process^(P2)^(P3)^(C) (as in Ela não faz nada pra mim hoje. “She can’t do anything for me today”). For this reason, this structure arrangement is called default directionality. The only exception to the arrangement of default directionality that is ideationally motivated happens in existential clauses. This is due to the fact that there is no P1 in existential clauses and that existence is realized by Prs^P2. The reason why the structure P1 ^Process ^ (P2) ^ (P3) ^ (C)—and Process ^ P2 ^ (P3) ^ (C) for existentials—is the default directionality is the fact that it is used in BP as the most frequent clause structure that realizes, ideationally speaking, initial and continuity messages. Accordingly, other ideational clause structures are used to realize discontinuity. This can be done by shifting the discourse flow to a specific aspect of the field. In this case, the preferred structures are those having a C2 (Enhancing Circumstance) conflating with Theme. This is due to the fact that these Circumstances focus on some aspect of the Process, such as time, place, manner and cause. This, in turn, realizes the kind of shift that discourse is making in the field. As a result, these tend to be more global and realize hyperThemes of longer hyperwaves. Alternatively, discontinuity can be carried out by other ideational clause functions different from P1 and C2. These also promote shifts in the discourse flow, but not as a focus on some specific part—but rather

Axial Argumentation and Cryptogrammar

151

as a change or contrast. Consequently these tend to be more localized and realize hyperThemes of localized shorter hyperwaves.

Figure 5.8 transitive orientation in BP

Text 3 is an example of the work done by transitive orientation. The conflation of Theme and ideational functions is annotated. The text is an excerpt of a paper by Berber Sardinha. More specifically, Text 3 explains the history of corpus linguistics. The meaning of ‘time through history’ is used in order to develop the field and, consequently, to shift the discourse flow. Again, the English translation is given and BP clause structure is kept as part of the example. Text 3 [EXP_EM_17] (1) A Linguística de Corpus P1/Theme (2) Como tal, Ø Textual Th. P1/Theme (3) (4) (5) (6) (7) (8) (9) (10)

ocupa-se da coleta e exploração de corpora. dedica-se à exploração da linguagem através de evidências empíricas, extraídas por meio de computador. já que o sentido original da é ‘corpo’, ‘conjunto de documentos’. Textual Th. palavra ‘corpus’ P1/Theme Na Grécia Antiga Alexandre, o Grande definiu o C2/Theme Corpus Helenístico. Na Antiguidade e na Idade Média produziam-se corpora de citações da C2/Theme Bíblia. Durante boa parte do século XX houve muitos pesquisadores que C2/Theme se dedicaram à descrição da linguagem por meio de corpora. Há duas diferenças fundamentais entre Prs/Theme esta época e a atual. A primeira obviamente, é que os corpora não eram P1/Theme eletrônicos, ou seja, eram coletados, mantidos e analisados manualmente A segunda é que a ênfase destes trabalhos era em P1/Theme geral o ensino de línguas. Atualmente o que prepondera na literatura é a C2/Theme descrição de linguagem e não a pedagogia.

152

Giacomo Figueredo

Text 3 [English] (1) Corpus Linguistics P1/Theme (2) As such, Textual Th.

Ø (It) P1/Theme

(3)

Since Textual Th.

(4)

In Ancient Greece C2/Theme In Classical Antiquity and the Middle Ages C2/Theme During a great part of the 20th Century C2/Theme

(5) (6)

(7)

There are Prs/Theme

(8)

The first P1/Theme

(9)

The second P1/Theme

the original sense of the word ‘corpus’ P1/Theme

(10) Today C2/Theme

works on the collection and exploration of corpora. focuses on the exploration of language through empirical evidence, extracted by the computer. is ‘body,’ ‘group of documents.’ Alexander, the Great defined the Helenistic Corpus. Bible quotation corpora were produced. there were many researchers dedicated to the description of language through corpora. two fundamental differences between those times and today. obviously is that corpora were not electronic, that is, collected, kept and analyzed manually. is that the emphasis of those works was language teaching. what dominates the literature is language description and not pedagogy.

As we can see in Text 3, P1s are used to realize messages initiating (1 and 7) and following the flow (2, 3, 8, 9). Circumstances of Location: Time (C2) are used to shift discourse (4, 5, 6, 10). 4.7

THEME SELECTION:

Subtypes of Orientation

After presenting the feature [orientation] of theme in BP, we are now in the position to describe how features are co-selected to form more delicate subsystems. The selection of a Theme is involved with the messages it realizes and the orientation to the field and tenor. This has implications for mood orientation and transitive orientation in terms of: (1) experiential meanings co-selecting with mood features and (2) interpersonal meanings co-selecting with transitivity features (see Figure 5.9). The systems shown in Figures 5.7 and 5.8 describe individually each type of orientation in their own right, not considering their co-selection. However, when co-selection is added to the description we see that

Axial Argumentation and Cryptogrammar

153

Figure 5.9 Co-Selection of mood and transitive orientation in BP

Theme may select or not select a crucial mood function and select or not select default directionality. As mentioned earlier default directionality is important to determine the job performed by transitive orientation. Accordingly, it works together with mood orientation in textual organization. So, the core functions of mood are as important as directionality for enabling. The probabilities of particular co-selections in BP indicate the part they play in the overall textual organization of texts. To explore this, we analyzed 8,394 major clauses. We looked for evidence of textual enabling by identifying all existing types of directionality in the corpus together with mood functions and how they conflate with Theme. The findings are shown in Table 5.3. The columns show mood functions and the rows show transitivity functions. The cells in the table show how many times these functions conflated with Theme. For example, the first cell tells that a Predicator of a declarative clause co-selects with a Process, and they conflate as: Predicator/Process/Theme. This happened 484 times, or 6.55% of all declarative clauses. Table 5.3 also shows that the crucial mood functions Subject is regularly conflated with Theme in declaratives (67.68%) and polars (65.21%). The ‘Qu-’ element tends strongly to conflate with Theme in elementals (98.52%), while Predicators overwhelmingly conflate with Theme on imperative clauses (85.88%). Textual enabling selects default directionality and the degree of involvement functions for ideational meanings in the following proportions: P1 (74.14%); P2 (3.21%); P3 (0.4%); C1 (0.95%); C2 (21.32%). The remainder of occurrences, those which are not part of the default textual organization, become more salient in the sense they are rarer and

154

Giacomo Figueredo

Table 5.3 Occurrences and Relative Frequency of theme, mood and transitivity in BP Indicative Declarative (Subject; Predicator in Existential or Passive)

Imperative (Predicator)

total

572 85.88% 18 2.7% 4 1.2% — — — — — — 68 10.21% — — — — 662 100%

1088 5136 270 30 28 8 1790 2 42 8394

Interrogative Polar (Subject)

Process 484 6.55% 28 20.28% Participant 1 4996 67.68% 90 65.21% Participant 2 142 1.92% 6 4.34% Participant 3 30 0.4% — — Extension 28 0.37% — — Elaboration 8 0.11% — — Enhancement 1592 21.56% 14 10.14% Matter 2 0.02% — — Angle 42 0.56% — — total 7324 100% 138 100%

Elemental (Qu-) 4 1.48% 32 11.85% 118 43.7% — — — — — — 116 42.96% — — — — 270 100%

less expected, both in terms of feature selection and clause structural organization. These occurrences, then, are deployed to emphasize their prominence. Ultimately, the interplay between default and prominent textual organization is the resource developed by theme to handle discourse flow in realizing contextualization (initial, phase: continuity and phase: discontinuity). In Brazilian Portuguese, a default theme selection (i.e., default ideational and interpersonal) following another default theme selection is mostly responsible for realizing continuity (see Figure 5.10). Prominent theme selections (i.e., prominent ideational and interpersonal) following defaults are mostly responsible for realizing phases of discontinuity (Figueredo, 2014). The more delicate features of discontinuity and the co-selections with scaffolding are then handled by more delicate prominence selections. The co-selections in mood and transitive orientation in BP, generate four possibilities for organization. These, in turn, are conjunctive entry conditions for four subsystems of theme. The more delicate subsystems will be further explained. 1.

THEME DEFAULT:

Oriented and Directional

The system of theme default is realized by a function of P1 (or Process: Existential) conflated with the Subject of indicative clauses, or the Process conflated with Predicator for imperative clauses. This is the default feature of theme because it (1) from above is frequently selected to realize initial and continuity messages, thus following the discourse flow and consequently the orientation to field and tenor; (2) from below, it is

Axial Argumentation and Cryptogrammar

155

Figure 5.10 theme in Brazilian Portuguese—Basic Systems

realized by the most frequent structure for clauses (both in terms of mood and directionality elements); and (3) from roundabout, it enables the greatest variation in interpersonal and ideational meanings. For example: Example 21 [COM_FD_15] Ela

não

P1

faz

nada

Process

P2

do

nothing

Subject Theme She

not

“She can’t do anything.”

156 2.

Giacomo Figueredo

THEME ELEMENTAL:

Oriented and Non-Directional

This system exclusively manages elemental interrogative functions. It is generated by the co-selection of [mood oriented], because it contextualizes the critical interpersonal feature, ‘qu-elemental’, as well as [nondirectional], because it does not contextualize any element directly involved in the transitive directional structure. For example: Example 22 [XPL_FD_02] Quanto que é o salário mínimo na Argentina, na Bolívia, ou no Paraguai? P2

Process P1

C2

is

in Argentina in Bolivia or in Paraguay

Qu-element Theme How much

the wage minimum

“How much is the minimum wage in Argentina, in Bolivia or in Paraguay?”

3.

THEME ANGLE:

Non-Oriented and Directional

theme angle gives textual prominence to ideational functions for projection of thought and saying in the clause. This system is generated by the feature [non-oriented] because there is no contextualization of any Mood Element function, and [directional] because Circumstance: Angle indicates the source of thought or saying, thus being agnate function to that of Senser or Sayer and directionally an agnate structure of that of P1. Corroborating evidence for including Angle in directionality is that Circumstance: Angle occurred in the corpus 44 times in total. Out of these, 42 times it was conflated with Theme (the other two times were found in news reports, which have the particular style of treating quotes as a single element placed in first position). This means that Angle conflates with Theme in BP 95.45% of the time if we extrapolate this sample to the language. Example 23 [EXP_FM_12] Segundo Paulo Freire

a democracia

é

uma conquista de todos

C2: Angle

P1

Process P2

Theme According Paulo Freire the democracy is

an achievement of all

“According to Paulo Freire, ‘democracy is an achievement of all.’”

4.

THEME PROMINENT:

Non-Oriented and Non-Directional

The system of theme prominent is generated by the co-selections of [non-oriented] interpersonally and [non-directional] ideationally. This

Axial Argumentation and Cryptogrammar

157

means that the least expected functions are conflated with Theme. Discursively, [prominent] is a feature that contrasts with [default] to realize discontinuity and change in discourse flow. Accordingly, [prominent] options signal prominence to meanings in waves and, accordingly, are a frequent co-variation with hyperThemes. Examples of this are in messages 2, 9 and 11 in Text 1 and messages 4, 5, 6 and 10 in Text 3.

5. A Note on Theme Markedness Textual markedness is mostly associated with the grammatical function of Marked Theme (Halliday, 1994; Rose, 2001). In many languages (Caffarel et al., 2004), a Marked Theme operates as an unexpected point of departure. From the point of view of discourse Marked Themes signpost shifts in the discourse flow (Martin and Rose, 2007; Matthiessen, 2004). In English the description of theme establishes the functions of Unmarked and Marked Theme. These functions are both grammatically (Halliday and Matthiessen, 2014) and discursively (Martin and Rose, 2007) motivated. Under the interpretation of BP in this chapter, in contrast, this distinction does not apply because markedness is not seen as a distinct category but rather as a probability notion. In linguistics the notion of markedness is often given by the opposition of unmarked vs. marked categories. An unmarked category is recognized as “the most expected, common and unremarkable case” (Butt et al., 1994, p. 139). However, determining markedness may vary according to different criteria and perspectives. Textual markedness can be understood when considering the dimensions of stratification and instantiation. Markedness viewed from stratification is related to the most expected preselections. Assigning textual markedness to Theme from above means markedness signals certain types of shift in the discourse flow. This is related to shifts in field orientation or tenor orientation. In BP the most frequent conflation to realize [initial] and [continuing] is through the [default] thematic choice: Theme/P1/Subject (67.68%). In this case, [default] can be considered the unmarked option. Conversely, the most frequent conflation to realize [discontinuity] is through the thematic choice of [enhancing] realized by Theme/C2 (21.56%). In this case, [enhancing] is the unmarked option. Thus each preselection has its own unmarked/marked options that cannot be generalized. Markedness viewed from instantiation is related to the relative frequency of occurrences for a given Theme. At first glance, frequencies tend to be generalized for the whole language irrespective of the environment in which Theme occurs. So, [default] would be considered unmarked because it occurs more times than other options. However, there are three different subtypes of [default], which vary according to the type of clause (default: imperative, default: declarative, default: interrogative). Thus

158

Giacomo Figueredo

there are three types of unmarked Theme in BP. In addition, relative frequency constrained to text types perturbs these generalized frequencies, which makes the notion of markedness unhelpful for description. For example, in Text 1 earlier, the relative frequency of [default] is 21.43%, or less than a third of what would be expected. A more complicated problem, though, is to determine which type of Theme is the Marked. Even if we assume that generalized frequency is a valid criterion to determine Unmarked Theme, then seven features are considered marked. However, their frequency is discrepant: [source 0.83%], [point-of-view 1.11%], [circumstantial 5.40%], [complement 5.49%], [intensive 9.87%] and [enhancing 77.35%]. Despite the fact that all these features would be considered Marked, for every occurrence of, say, [source], there are 1.33 occurrences of [pointof-view], 6.44 [circumstantial], 6.55 [complement], 11.66 [intensive] and 92.22 [enhancing]. In turn, the ratio of enhancing to default is 3.09. So, in terms of probabilities, [enhancing] is significantly closer to ‘unmarked’ then it is to ‘marked.’ Instead of using markedness as a category, the approach we deployed in this description of BP offers us a set of possibilities, each having their own functional configuration motivated by discourse flow and managed by textual enabling.

6. Conclusion We began this chapter approaching theme from above. By examining discourse flow structure, we identified shifts in the flow and contrasted different structures. Our focus was the smallest unit of textual discourse semantics, or ‘message.’ The reasoning for establishing contrasts enabled us to determine textual discourse semantic functions—macroTheme, hyperTheme and continuity—responsible for structuring discourse. We could then apply axial reasoning to develop the systems responsible for organizing messages—the system of contextualization. Then we moved on to approaching theme from below, in terms of its realization by groups that constitute the clause. Different classes of group are deployed depending on the needs of Theme organization. Nominal groups operating as Theme are related to the beginning and continuing of discourse. Prepositional phrases are related to shifts in the flow. Examining the clause from roundabout shows that the textual structure of the clause is arranged in relation to discourse flow. More specifically, the function of Theme manages ideational and interpersonal meanings by enabling transitivity and mood functions to conflate and become the point of departure. Accordingly, the conflation between Theme and ideational and interpersonal functions offers different possibilities of

Axial Argumentation and Cryptogrammar

159

functional combinations that may be used to realize structures of discourse flow. In the system of theme, this is the feature [orientation]. The co-selections between ideational orientation/non-orientation and interpersonal orientation/non-orientation generate more delicate features of theme. At the third level of delicacy (Figure 5.10), theme in BP has nine features: [absolute], [imperative], [indicative], [complement], [circumstantial], [source], [point-of-view], [enhancing] and [intensive]. Except for more delicate subsystems of [indicative] and [intensive], this description includes the whole of theme in BP. However, at this point, we felt the need to address an important typological (or descriptive) difference between BP and other languages—the notion of markedness. Throughout this description, we tried to show that in BP there is no discursive or grammatical motivation to establish a clear cut separation between unmarked and marked Themes. Rather, this difference could mostly be established in terms of general frequency of occurrences, which is not practical since texts obey genre configurations—and not language as a whole. Moreover, there is no objective linguistic criterion to separate frequencies, ratios and proportionalities. We hope that this chapter may contribute to future descriptions of textual grammar systems in other languages, especially with respect to including axial reasoning, trinocular view and frequency of occurrences as part of the methodology.

Notes 1. For ease of reading, all examples in this section looking from discourse semantics use an English translation. However this does not affect the analysis. 2. Due to our limited space in this chapter, we will continue to use Text 1 as our source of reasoning in this section. It is worth mentioning, though, that this mapping of message onto Theme in BP has been extensively done in our research corpus CALIBRA and is documented in Figueredo (2011, 2014) and Figueredo and Araújo (2013). 3. All systemic features presented in the body of the text will be written in square brackets, as in [initial] or [culminative]. 4. The examples will be glossed according to the Systemic Functional Glossing Conventions, which can be found at https://systemiclanguagemodelling.word press.com/glossing/. However, we will only gloss functions and classes that are being illustrated in the examples. 5. Constructed clauses based on example 17.

References Alves, L. (2017). Uma proposta de descrição sistêmico-funcional das orações materiais do português brasileiro orientada para os estudos multilíngues [A systemic functional description of the material clause in Brazilian Portuguese for multilingual studies]. Mariana, Brazil: Universidade Federal de Ouro Preto—UFOP.

160

Giacomo Figueredo

Butt, D., Fahey, R., Feez, S., Spinks, S. and Yallop, C. (1994). Using functional grammar: An explorer’s guide. Sydney: Macquarie University. Caffarel, A., Martin, J. R. and Matthiessen, C. M. I. M. (2004). Language typology: A functional perspective. Amsterdam/Philadelphia: John Benjamins. Eggins, S. and Slade, D. (1997). Analysing casual conversation. London: Cassell. Figueredo, G. (2011). Introdução ao Perfil Metafuncional do Português Brasileiro: Contribuições para os Estudos Multilíngues [Introduction to the metafunctional profile of Brazilian Portuguese: Contributions to multilingual studies]. Belo Horizonte, Brazil: Universidade Federal de Minas Gerais—UFMG. Figueredo, G. (2014). Os sistemas textuais de focalização na organização funcional da gramática do Português Brasileiro [Textual systems of focus in the functional grammar of Brazilian Portuguese]. São Paulo: DELTA. Documentação de Estudos em Linguística Teórica e Aplicada, v.30. Figueredo, G. and Araújo, C. (2013). Uma breve história da divulgação científica: as metáforas ideacionais e o sistema de mensagem na construção do discurso da ciência [A brief history of popular science: Ideational metaphors and the system of message in scientific discourse]. Entrepalavras, v.3. Halliday, M. A. K. (1978). Language as social semiotic: The social interpretation of language and meaning. London: Edward Arnold. Halliday, M. A. K. (1994). An introduction to functional grammar (2nd ed.). London: Arnold. Halliday, M. A. K. (2002). On grammar. London and New York: Continuum. Halliday, M. A. K. and Matthiessen, C. M. I. M. (1999). Construing experience through meaning: A language-based approach to cognition. London: Continuum. Halliday, M. A. K. and Matthiessen, C. M. I. M. (2014). Halliday’s introduction to functional grammar (4th ed.). London: Routledge. Martin, J. R. (1992). English text: System and structure. Philadelphia/Amsterdam: John Benjamins. Martin, J. R. and Rose, D. (2007). Working with discourse: Meaning beyond the clause (2nd ed.). London/New York: Continuum. Martin, J. R. and White, P. R. R. (2005). The language of evaluation: Appraisal in English. New York: Palgrave Macmillan. Matthiessen, C. M. I. M. (1995). Lexicogrammatical cartography: English systems. Tokyo: International Language Sciences Publishers. Matthiessen, C. M. I. M. (2004). Descriptive motifs and generalisations. In A. Caffarel, J. R. Martin and C. M. I. M. Matthiessen (Eds.), Language typology: A functional perspective. Amsterdam/Philadelphia: John Benjamins, pp. 537–673. Matthiessen, C. M. I. M., Teruya, K. and Lam, M. (2010). Key terms in systemic functional linguistics. London and New York: Continuum. Pagano, A., Figueredo, G. and Ferreguetti, K. (2012). Equivalência em foco: padrões na tradução para a língua inglesa de significados existenciais do português brasileiro [Equivalence in translation of existentials of English into Brazilian Portuguese]. In Viana, V. and Tagnin, S. (Org.). Corpora na Tradução (1st ed.). São Paulo: Hub, 2015, v. 1, pp. 211–241. Paula, A. (2017). Orações verbais—uma descrição sistêmico funcional dos processos de representação do dizer do português brasileiro [Verbal clauses—a systemic functional description of saying in Brazilian Portuguese]. Mariana, Brazil: Universidade Federal de Ouro Preto—UFOP.

Axial Argumentation and Cryptogrammar

161

Pike, K. L. (1982). Linguistic concepts: An introduction to tagmemics. Lincoln: University of Nebraska Press. Rose, D. (2001). Some variations in theme across languages. Functions of Language, 8.1: 109–145. Santiago, L. (2012). Modelagem da conversa cotidiana em português brasileiro [Modeling casual conversation in Brazilian Portuguese]. Mariana, Brazil: Federal University of Ouro Preto—UFOP.

6

Arguments for Seeing ThemeRheme and Topic-Comment as Separate Functional Structures* Randy J. LaPolla

Theme-Rheme and Given-New The early Prague School of linguistics is known for the concepts of Functional Sentence Perspective, Theme-Rheme, and Communicative Dynamism (CD). Functional Sentence Perspective is essentially what we talk about now as information structure, but the Prague School focused on Theme-Rheme and communicative dynamism. In discussing Theme, Firbas (1964[2009: 260]), citing Mathesius (1939), says, “Mathesius defines the Theme as ‘that which is known or at least obvious in the Given situation, and from which the speaker proceeds’ in his discourse (234),” though later (Firbas 1987) argues that the Theme is not necessarily ‘known,’ and it is not necessarily the initial element of a clause, but it is the item with the lowest degree of communicative dynamism in the clause. That is, “the information conveyed by the theme contributes least to the further development of the communication within the sentence” (Firbas 1987: 138). But it provides the foundation (using Mathesius’s základ ‘foundation’) for the information provided in the rest of the clause.1 He mentions that Daneš (1964) pointed out that Mathesius used three different terms, východiště ‘point of departure,’ téma ‘Theme’ and základ ‘basis, foundation,’ and that in 1939 Mathesius explicitly stated that the point of departure was not necessarily always identical with the theme (cf. Mathesius 1939: 171, 1947: 235, 1982: 174). Mathesius, however, did not explain the difference between the two. Later he dropped the term východiště [‘point of departure’] altogether and used the terms téma [‘theme’] and základ [‘foundation’] synonymously, in fact returning to a practice he already chose in 1929. (cf. Mathesius 1929, 1982: 29–38, 1983: 121–142). (Firbas 1987: 140) Firbas goes on say that the feature ‘aboutness’ is always part of the Theme (that is, it is the Topic of the clause), but the Theme is not necessarily ‘context-dependent,’ so the two are separate features. The important

Arguments for Seeing Theme-Rheme

163

thing for us is that Mathesius felt the two parts of the semantic structuring of the sentence were “what is being spoken about” (the basis, foundation) and “what is being said about it” (the ‘core’; Mathesius 1982: 120, cited in Firbas 1987: 144), that is, Topic and Comment, and Mathesius for some time saw this as separate from the point of departure. Firbas (1964[2009: 265]) argues against another Prague School linguist, František Trávníček, who had argued (1962) that the Theme was the initial element of the clause. Firbas says, “Identifying the theme with the beginning of the sentence, Trávníček disregards both the criterion of the degree of CD and the criterion of known or unknown information.” Firbas also mentions that the linguist K. Boost had a similar understanding of ‘Thema,’ and that a third linguist, Eduard Beneš (1959), suggested a distinction between ‘basis’ and ‘theme’: By ‘basis’ he understands the phenomenon that ‘as the opening element of the sentence links up the utterance with the context and the situation, selecting from several possible connections one that becomes the starting point, from which the entire further utterance unfolds and in regard to which it is orientated’ (216). The term ‘theme’ would be applied by him to the phenomenon defined by us here as the element(s) carrying the lowest degree of CD within the sentence. We believe this differentiation to be sound. It prevents the student of language structure from mixing up aspects that should be kept separate. (Firbas 1964[2009: 267]) So here we see at least some Prague School linguists accepting a three-way distinction between what is ‘known’ or not; what is the starting point of the utterance; and what the clause is about and also has the lowest degree of communicative dynamism in the clause. Michael Halliday was one of the first linguists outside the Prague School to talk about information structure (e.g. 1967b, 1970), and the typology that he outlined influenced Knud Lambrecht (1986, 1994, 2000), whose typology of information structure has become very influential and adopted in, for example, Role and Reference Theory (e.g. Van Valin and LaPolla 1997: Ch. 5). Halliday distinguishes the status of referents as identifiable or not in the mind of the addressee from the structure of focus and pragmatic presupposition, using ‘New’ and ‘Given’ for the latter concepts, respectively. He states (1967b: 204), What is focal is ‘new’ information; not in the sense that it cannot have been previously mentioned, although it is often the case that it has not been, but in the sense that the speaker presents it as not being recoverable from the preceding discourse.  .  . . If we use

164

Randy J. LaPolla the—admittedly rather inappropriate—term ‘given’ to label what is not ‘new’, we can say that the system of information focus assigns to the information unit a structure in terms of the two functions ‘given’ and ‘new.’

This is quite similar to Lambrecht’s definitions of focus, pragmatic presupposition and focus structure (1994: Ch. 5; where he cites Halliday 1967b at length). Halliday (1967b) identified ‘unmarked focus,’ where all but the Topic or the entire clause is in focus, and ‘marked focus,’ where the domain of focus is limited to a single constituent of the clause. Lambrecht expanded this typology, not only distinguishing broad focus (Halliday’s unmarked focus) from narrow focus (Halliday’s marked focus), but within broad focus distinguishing between ‘predicate focus,’ the most common type of focus, where there is a Topic, and the rest of the clause is a Comment about that Topic, and ‘sentence focus,’ where the whole clause is in focus and there is no Topic. In narrow focus, although all but the single focused element is within the pragmatic presupposition (which is an open proposition), there is also no Topic that the clause is about. Halliday explicitly recognized that there are utterances where the entire clause is in focus (i.e. all ‘New’), yet unlike Lambrecht seems to assume that there is a topical Theme in all clauses (see footnote 23), and this I think has given rise to problems in the analysis of texts.2 I’ll return to this later. Halliday adopts the Prague School conception of Theme-Rheme, though he distinguishes it as a separate functional structure from GivenNew (communicative dynamism). He argues that Topic is just one type of Theme (1994: 38). He identifies English Theme3 as the initial element of the clause but states clearly that is not how it is defined; it is defined functionally, in construction with the Rheme, as a message is made up of a Theme and a Rheme, and within that “the Theme is the starting-point for the message; it is the ground from which the clause is taking off” (Halliday 1994: 38). Matthiessen and Halliday (2009: 65) state, “The system of THEME sets up a local environment, providing a point of departure by reference to which the listener interprets the message.” From the examples used where the topical Theme is not a nominal group (e.g. Halliday 1994: 39), such as [with sobs and tears]Theme [he sorted out those of the largest size]Rheme, we can see that the Theme is not always what the clause is about, and Downing (1991) on this basis argues that “what is being talked about” and “the point of departure for the clause as a message” should not be conflated, as “[t]he point of departure of the message is not necessarily what the message is about” (p. 122). McGregor (1992) also argues for a position similar to Downing’s. Distinct from the system of THEME is the “degree of newsworthiness” (Matthiessen and Halliday 2009: 66), what was called communicative dynamism by the Prague School, “represented as a configuration of Given + New,” the system of information focus (ibid).

Arguments for Seeing Theme-Rheme

165

So we can distinguish three concepts, Topic-(Comment),4 Given-(New), and Theme-(Rheme), though Topic and Theme seem to be collapsed in Mathesius’ later view, and in Halliday’s view there seems to be the assumption that Theme always includes Topic. What I would like to argue is that there are very good reasons for the Theme being the initial element of the utterance, and that the collapsing of Topic and Theme into one concept is problematic, and so we should separate out three functional structures: 1. Theme-Rheme (where Theme is defined as the beginning of the speaker’s utterance, so the structure necessarily always has Theme before Rheme, and does not necessarily include Topic or Given); 2. Given-New (where New is defined as that which is not recoverable from the preceding discourse, and Given is that which is not New (Halliday 1967b, cited earlier); the structure can be New-Given in some languages as a marked pattern and in some languages as the unmarked pattern); and 3. Topic-Comment (where Topic is defined as what the clause is about and Comment as what is said about the Topic; this structure can be Comment-Topic in some languages as the unmarked pattern). Before I can make my point, though, I need to digress a bit and talk about how communication happens, and why the initial position in an utterance is important independent of being the Topic or not.

The Creation of Meaning Meaning doesn’t exist externally in the world; it is created subjectively in our minds.5 That is, even in linguistic communication, there is no meaning in words or sounds; we create meaning from the actions of other people by putting together certain assumptions in which the actions ‘make sense.’ This includes linguistic actions (speech/writing), as language is not a thing, but a behavior, and many of the principles and cognitive abilities are the same as for other aspects of human behavior.6 The inference used to ‘make sense’ of some phenomenon is called abductive inference (Peirce 1940; called ‘inference to the best explanation’ in the philosophy of science), and it is how we understand the natural world and how we understand the motivations of other people when they do something. We have a natural instinct to ‘make sense’ of things, as it is crucial to our survival, and that means trying to hypothesize a reason for some phenomenon, whatever it is. When we observe a natural phenomenon, we take whatever information is available to us and create a context of interpretation in which that phenomenon makes sense to us. When we see another person doing something, we instinctively hypothesize why they are doing it, using abductive inference. In communication we apply this ability to inferring someone’s intentions when they want us to infer their intention in performing some action (see LaPolla 2015 for details).

166

Randy J. LaPolla

So communication is not coding and decoding, it is ostension7 and inference, that is, the communicator doing an action ostensibly to show the desire to communicate, and then the addressee using abductive inference to infer the reason for the person’s action, whether it was linguistic, non-linguistic, or both.8 Communication can happen with or without language, but language serves to constrain the creation of the context of interpretation in particular ways relevant to the culture and cognition of the speakers. Each language is unique, as it emerges out of the communicative behavior of a unique society of speakers, and so each language will differ in terms of what semantic domains the speakers have constrained often enough for the particular pattern to become conventionalized on the societal level and habitualized on the individual level, and they will differ in terms of how much they constrain the interpretation of a semantic domain if they do so, and they will differ in the particular linguistic mechanisms used to constrain it if they do so. As perception of a speech act is linear, interpretation is also linear, and this is why the initial segment of the utterance is so important: we don’t wait for the whole utterance to be completed before we start creating the context of interpretation; we start building the context of interpretation with the very first word, regardless of the structure of the utterance, and then that context of interpretation influences the creation of the context for interpreting the rest of the utterance.9 We anticipate what is to come, and languages can give us clues as to what to expect.10 Signaling to the hearer what is about to come in the interaction on the part of the speaker and inferring what is about to come in the interaction on the part of the hearer is what is talked about in Interactional Linguistics as ‘projection’ (e.g. Auer 2005; Thompson and Couper-Kuhlen 2005; Hopper and Thompson 2008; Hopper 2012, 2013). This use of the word ‘projection’ is quite different from the use of the word in Halliday’s framework (for projecting quotes and ideas) and in generative syntax (for consistency in categorization between the lexical item and the larger structures based on it). Sandra Thompson, a key person in Interactional Linguistics since its inception, in an early paper in the development of the field (1985) explicitly builds on work by Halliday (e.g. 1967a, b, 1968) and Fries (1983) on the importance of initial position in the clause and the function of Theme. She explains the difference between the uses of initial and final purpose clauses in terms of the three metafunctions recognized by Halliday: the initial purpose clause helps to guide the attention of the reader, by signalling, within the portion of the text in which it occurs, how the reader is expected to associate the material following the purpose clause with the material preceding it. The final purpose clause does not play this role. In Hallidayan terms, the difference could be stated

Arguments for Seeing Theme-Rheme

167

by referring to the three functions of language: ideational (content), textual (text-organizing), and interpersonal (pragmatic). While final purpose clauses serve at the ideational level, initial purpose clauses operate simultaneously at the ideational and at the textual levels (Halliday 1973). (1985: 61, italics original) Projection in Interactional Linguistics includes both the ability to guess what is coming up in the interaction (anticipating the speaker’s intentions), and also the grammatical mechanisms for helping the hearer to make such guesses (“telegraphing” one’s intentions, to use the boxing metaphor).11 Typologically different languages allow for different types of projections (e.g. see Tanaka 2000, 2001; Ono and Thompson 2017 on projection in Japanese), and languages differ typologically also in terms of what is obligatorily thematic and what is not. The clause is seen in Interactional Linguistics as the locus of interaction, as “the clause is precisely that unit which permits significant projectability” (Thompson and Couper-Kuhlen 2005: 485). The same authors (p. 487) point out that “Schegloff (1987, 1996) has proposed that the beginning of the turn in English is the key locus for projectability.” Hopper (2012: 308) argues that [p]rojection is what makes verbal communication an open and collaborative affair; as participants develop a sense of where the discourse is going, they tacitly mold it, allow it to continue, harmonize with the speaker’s goals, interrupt it with their own contribution, offer supportive tokens of various kinds, or predict when their turn will come.12 In the next two sections we will look at the different uses of Theme to enhance that interaction (assist in projection) that English speakers and Tagalog speakers have conventionalized. In English the initial segment influences the creation of the context of interpretation in terms of constraining the interpretation of the mood, such that the addressee can project what sort of interaction the speaker intends to accomplish with the hearer. Like English, Tagalog also makes good use of the clauseinitial position to aid the hearer’s projection of the speaker’s intention, but uses it for somewhat different purposes than English does, and does not generally put the Topic in initial position. Other languages put different elements in initial position to help the hearer project aspects of the interaction salient to the speakers.

The Use of Theme in English Halliday (1967b, 1994: Ch. 3) showed that speakers of English have conventionalized a particular use of the speaker’s starting point (Theme) to

168

Randy J. LaPolla

mark the grammatical mood of the clause in the unmarked cases by what appears in the Theme:13 Indicative: declarative: The unmarked Theme (the most usual Theme) in English declarative clauses is the Subject, as in (1)–(2): Theme

Rheme

topical 1) The boy lost his notebook. 2) The cup was smashed to pieces.

Indicative: interrogative: yes/no question: The unmarked Theme in English yes/no interrogatives includes the Finite Verbal Operator (is, isn’t, does, doesn’t etc.—that which embodies the expression of tense and polarity or modality) and the Subject, in that order: Theme interpersonal 3) Did 4) Will 5) Can

Rheme topical you eat yet? he eat the ice cream? I do it?

Indicative: interrogative: WH-word question: The Theme in English question-word type interrogatives is the question word, i.e. that which requests the missing piece of information: Theme interpersonal/topical 6) Who 7) What 8) Where

Rheme left the cat out? are we having for dinner? did he say we are going?

Imperative: The unmarked Theme in English non-negative second person imperatives is the Predicator (the function of be in (9)). The unmarked Theme in negative second person imperatives (prohibitives) is Don’t plus the Predicator (as in (10)). The unmarked Theme in first person imperatives is Let’s, as in (11). Theme topical 9) Be 10) Don’t be 11) Let’s

Rheme quiet! (second person Imperative) so talkative! (second person Prohibitive) have lunch together! (first person Imperative)

This system is important and functionally useful because it allows the hearer to project right at the beginning of the clause what sort of interaction is being carried out (assuming congruency between mood and speech

Arguments for Seeing Theme-Rheme

169

act), and if it is declarative, what the clause is about, and if it is interrogative, what sort of information is being asked for.

The Use of Theme in Tagalog Tagalog, a Malayo-Polynesian language spoken in the Philippines, is a consistently focus-initial language and so in general the predicate appears in initial position and the information structure is Comment-Topic and New-Given rather than Topic-Comment and Given-New (using Halliday’s (1967b: 204) definition of ‘New’ as what “the speaker presents . . . as not being recoverable from the preceding discourse”; cf. Naylor (1975: 48) on the predicate as New and the Topic as Given in Tagalog). The predicate in most cases marks aspect, realis/irrealis, and often the semantic role of the Topic of the clause, and so when it appears in initial (Theme) position, it allows the hearer to project the situation predicated, its reality status and aspect and the semantic role of the main participant in the situation, the Topic. Topic here is actually a Tagalog-specific grammatical status, as well as a pragmatic status, as it is an argument singled out for special morphological treatment, and is also what the clause is about.14 Generally any argument, whether participant or circumstance, can be the Topic of the clause, though usually (but not always) it is one that is identifiable to the hearer, and in many cases the predicate takes a form to reflect the semantic role of that argument.15 The representation of the Topic argument (if it appears in the clause and is not a pronoun) generally takes either a marker of specificity (ang, or si with proper names) or a Topic form of demonstrative pronoun plus linker (most commonly ’yung) to mark it as the Topic. The set of pronouns also distinguishes topical from non-topical referents, with the latter appearing as possessive forms or dative forms, and in the second person singular there is also a special form used when the 2sg referent is focal, and as such it always appears in initial (focus) position. Let’s look at some examples of the use of the different Topic-marking affixes (natural examples from my own fieldwork, the conversation ‘Making Salsa’; see also Schachter 2008: 337–338, for sets of constructed parallel examples with the same arguments but with different choices of Topic). 12) 1. Jirehel: Madali lang kasi’ng gumawa ng salsa eh. ma-dali stat-easy

lang kasi ang gawa ng salsa eh just because spec make poss sauce emphatic

“Because making sauce is just easy.” (Lit.: “Because one’s making of salsa is easy.”)

2. Wendy: Oo, madali lang. oo ma-dali lang yes stat-easy just “Yes, it’s really easy.”

170

Randy J. LaPolla

3. Gawin mo lang ketchup, gawa-in mo lang ketchup do-irr:pt 2sgposs just catsup “Just make it with ketchup,

4. tapos lagyan mo ng tomatoes, tapos finish

lagay-an mo ng tomatoes put-lt 2sgposs poss tomatoes

then you add tomatoes to it, 5.

lagyan mo ng salt and pepper to taste, tapos na. lagay-an put-lt

mo ng salt and pepper to taste tapos na 2sgposs poss salt and pepper to taste finish cs

add salt and pepper to taste to it; then (it’s) done.”

In this example, the first speaker, Jirehel, refers to the making of salsa using a form (gumawa) that includes the Actor-Topic infix -um-,16 but then the second speaker, Wendy, uses the same root in line 3 of the example, but with Irrealis Patient-Topic suffix -in, to profile the event from the point of view with the salsa as the Topic. She then follows this in lines 4 and 5 with two tokens of the root lagay ‘put, add’ with the LocationTopic suffix -an, to keep the salsa as the Topic, but with it now having the semantic role of the location where the tomatoes and salt and pepper are to be added. Notice how there is no overt reference to the Topic in any of Wendy’s utterances (e.g. ’yung salsa ‘the salsa’ could have been added to the end of each of Wendy’s utterances in lines 3–5, but it wasn’t), yet we can tell what is being talked about (what is the pragmatic and grammatical Topic) because of the marking on the predicate, and this ‘switch function’ referent-tracking system allows us to track a single Topic even while the function of the Topic changes. In a series of insightful papers, James Martin (1981, 1988, 1990, 1995b, 1996, 2004, Martin and Cruz 2018) discusses several aspects of Tagalog grammar from a Hallidayan point of view. He points out that Tagalog has a “predisposition for loading interpersonal meaning onto the front of the clause” (1990: 36). Interrogative elements, for example, obligatorily appear in initial position (as in (13); with some interrogative words this is achieved by the use of clefting to isolate the interrogative word as the utterance-initial predicate), the ‘attentiondirecting’ deictic pronouns (heto, hayan) and interpersonal Comments such as mabuti pa ‘it would be best if .  .  .’ always appear in initial position, and the marking of negation and modality is also generally utterance-initial, as in (14) and (15). Aside from the immediately initial position, there is a large number of clitics that mark textual or interpersonal meaning that occur in second position in the clause (those in

Arguments for Seeing Theme-Rheme

171

bold type in (13)–(15)), and the personal pronouns (e.g. mo in Wendy’s utterances in (12) and in (15), and siya, which is the Topic in (13)) are also second position clitics. (Examples (13)–(16) adapted from Martin 1990: 19, 22, 12.) 13) Kailan kaya siya tatakbo. kailan when

kaya siya ta-takbo speculation 3sgT redup-run

“When do you suppose she’ll run?”

14) Hindi pa rin ho ba lumabas si Cory. hindi neg

pa rin ho ncs also respect

ba labas si Q outside spec

Cory pn

“Didn’t Cory leave anyway, sir/ma’am?”

15) Baka naman gusto mong magkape. baka maybe

naman gusto mo=ng contrast want 2sgposs=lnk

mag-kape iat-coffee

“But maybe you’d like to have coffee.”

Martin not only identifies these elements as thematic but also considers Topic phrases at the end of the clause, such as si Cory in (14), thematic. This view is also followed in Matthiessen and Halliday (2009). It is also possible for the Topic or circumstantial phrases and clauses to appear in focus position preceding the predicate, generally linked to the predicate with the particle ay, as in (16),17 which Martin (1990: 20) classifies as having a marked Theme. This construction is usually used for scene-setting or contrastive Topics (cf. Fox 1985), and for narrow focus questions. (See also Naylor 1975: 54ff. on the predicate as the unmarked Theme and anything other than the predicate appearing in the initial position as marked Theme.) 16) Si Aquino pala ay pinatay ni Marcos. si spec

Aquino pala ay patay pn surprise ppt dead

ni Marcos poss pn

“Surprisingly Aquino was killed by Marcos.”

In the major theoretical discussions of Theme, from Mathesius (1929) up through to the present, the languages under discussion have been Topic-Comment languages where the Topic appears in initial position, and so there has been an assumption that the function of the initial part of the utterance or clause (the speaker’s starting point) must include presenting the Topic.18 A language like Tagalog allows

172

Randy J. LaPolla

us to see that the starting point of the utterance does not have to include the Topic, yet it is still important as the hearer’s starting point for creating the context of interpretation, and for understanding how the speaker facilitates the interaction by structuring the clause in a way that makes it easier for the hearer to infer certain aspects of the speaker’s intention. What I’d like to argue is for defining Theme as the beginning of the utterance. The elements that appear there influence the hearer’s projection of what is to follow. A Topic that comes at the end of the clause would not be considered part of the Theme, though the clause-initial Topics would be. There has been much discussion of the complementarity of the themeRheme and Given-New functional structures in English, giving the clause peaks of speaker (Theme) vs. hearer (New) prominence and troughs of non-prominence in a sort of periodic wave-like structure (e.g. Halliday 1979, 1994; Matthiessen 1988, 1992; Martin 1992), and in fact Matthiessen (1992: 42) states that “the textual metafunction is concerned with creating contrasts between prominence and non-prominence in meaning as an aid in the processing of text.” As Tagalog consistently has the ‘New’ utterance-initially (when it is not the entire utterance), conflating Theme and Topic would mean that there is no such wave-like structure in Tagalog, but if we separate Theme and Topic we can say that there is still such a complementary periodicity, but it is Topic vs. New rather than Theme vs. New, which is actually what Halliday meant in talking about Theme vs. New in English, as it is the topical Theme that he felt is the most important in this regard. This then leaves us with the question of where to draw the line between what I am now defining as Theme and the rest of the clause (Rheme) in Tagalog. In Halliday’s analysis of English (1994), everything up to the topical Theme, defined as the first participant, process or circumstance in the clause, is considered part of the Theme, but if in a language like Tagalog the Topic is not part of the Theme, how do we draw the line between Theme and Rheme? In an example like (14), if we follow the rule given for English, then the entire predicate (including lumabas, which could be considered an ideational Theme rather than topical Theme) is within the Theme, leaving only the Topic (si Cory) outside the Theme. Evidence for this sort of analysis is the fact that in some cases the initial element and the predicate have to be linked into an overt phrase using the na/-ng linker. In such cases, if there are second position clitics, they occur after the first word of the phrase and before the linker, as in (17), where the phrase pwedeng kunin ‘can take’ is interrupted by the 1sg non-Topic pronoun and the question marker, and so the linker appears at the end of the question marker. So if we take the first phrase as the Theme, then we should include the other elements as well, again leaving only the Topic (yun leaves) as the Rheme.

Arguments for Seeing Theme-Rheme

173

17) Pwede ko bang kunin yun leaves . . . ? [pwede can

ko ba=ng 1sgposs q=lnk

kunin] ’yung take:ut that+lnk

leaves leaves

“Can I take the leaves . . . ?” https://delishably.com/beverages/How-to-Make-Malunggay-Tea-Home-MadeMoringa-Tea

In a clause like (16), where the Topic is in initial position, the Theme would just be the initial Topic alone. It then would not include the particle of surprise pala, which might seem counterintuitive given that si Aquino pala is one information unit in Halliday’s view (e.g. 1967b). But it might be seen as similar to unfortunately in John unfortunately left before Melanie came,19 where the Comment Adjunct would not be seen as part of the Theme. In a case like line 4 of (12), we could consider the first word, tapos, as a textual Theme and the predicator, lagyan, as the ideational Theme. But as the other two elements actually form a phrase together with the predicator (mo and ng tomatoes are both possessive modifiers)20 set off from the Topic (which is left unexpressed by Wendy), we could consider the entire utterance to be thematic, with only the Topic again not part of the Theme. We also commonly find clauses such as (18): 18) Kaibigan ito ni Waki . . . kaibigan ito ni Waki friend this:topic poss pn “This is Waki’s friend . . .” http://cheriepaanashaven-collab.blogspot.com/2013/12/calle-pogi-series-3lian.html?m=1

In this clause the Topic comes between the two parts of the possessive phrase ‘Waki’s friend,’ and so if we see Theme as linear, then the whole clause is within the Theme. On rereading the Prague School writings, it seems we may not need to draw a clear line at all. Their view was that there was a continuous progression, and not a clear break between the two, though there are some cases, such as (14), (16) and (12.4), where we can draw a line. As Geoff Thompson pointed out in his plenary talk at the conference, Theme ‘tails off’ (see also Matthiessen 1992: 50–51). If we take Theme as those elements which aid projection, then we also do not need to draw a clear line, as throughout the utterance the hearer is using what has been said already to project what is to come or the end of the turn.

Separating Theme-Rheme and Topic-Comment Now, this chapter is not really about Tagalog; I am just using Tagalog as an example. I am arguing for a methodological principle. I want to argue that

174

Randy J. LaPolla

we should distinguish between the Theme, defined as the speaker’s starting point, and the Topic, what the clause is about. The Theme forms a structure with the Rheme, and the Topic forms a structure with the Comment. In Tagalog the Topic is explicitly marked, and does not depend on position within the clause.21 In other languages, such as Chinese, Topic is defined by its position in the clause, and only reference phrases that are clause-initial (in the case of primary Topics) or at least preverbal (in the case of secondary Topics) will be seen as a Topic, and so word order is used to distinguish referents that are Topics (and distinguish primary and secondary Topics) from referents that are not Topics. This means that in Chinese it is being made thematic that allows the reference to a referent to be understood as referring to the Topic of the clause. See the following example (LaPolla and Poa 2006: 278, from Rúlín Wàishıˇ, an 18th-century vernacular novel): 19)

(a) Yuán pn chū-le emerge-pfv

cháo dynasty yī-ge one-cl

mò end qiānshílěiluò honest.and.upright

nián, year de assoc

yě céng also exp rén. person

“At the end of the Yuan dynasty, there appeared an honest and upright person.” (b.1) Rén xìng Wáng, (b.2) míng Miǎn, person surnamed pn given.named pn “(This) person was surnamed Wang, and had the given name Mian.” (b.3) zài Zhūjì-xiàn xiāngcūn jūzhù; loc pn-county countryside live “(he) lived in in the countryside of Zhuji county,” (b.4) qī suì shí sǐ-le fùqīn, seven years.old time die-pfv father “when he was seven his father died,” (c.1)

tā mǔqīn 3sg mother

zuò xiē zhēnzhǐ, do some sewing

“his mother did some sewing,” (c.2)

gōngjǐ supply

tā dào cūn xuétáng-lǐ qù dúshū. 3sg all village school-inside go study

“to give him money to go to the village school to study.”

This is a very typical stretch of Chinese narrative text. The first clause introduces a new referent in post-verbal (non-Topic) position, then this referent becomes the Topic of the following four clauses.22 A related referent is then the Topic of the next two clauses. The structure of all the clauses except the first is Topic-Comment. The first clause is presentative, a ‘sentence focus’ construction (Lambrecht 1994; LaPolla 1995), and so

Arguments for Seeing Theme-Rheme

175

does not have a Topic (the temporal expression locates the event in time, but is not the Topic of the predication—see LaPolla 1995 for discussion). The clause in (19b.4) involves two parts, a Topic and a Comment, but the Topic is not mentioned and the Comment (the entire line) has the reference to the person who died in post-verbal position, just as in the presentative construction in the first line of this text. In this case the reference to the father appears in post-verbal position not because the father is being introduced for further development as a Topic, but simply to mark it as not a Topic, and so his dying is seen as an event that happened to the Topic of the Clause (Wang Mian) when he was seven years old. What is important is that the father not be interpreted as the Topic of the Clause. This allows for consistency in the choice of Topic, what is called a ‘Topic chain’ in Chinese linguistics. Had reference to the father been before the verb, then the father would have been understood as the Topic, and the statement would have been about him, not about Wang Mian. That would be saying something very different from what the author intends in this stretch of text. Notice that distinguishing Theme from Topic allows us to still talk about Theme in these clauses, even though the Topic is not overtly mentioned, resolving an issue that frequently comes up when scholars try to apply Halliday’s analysis of English Theme to typologically different languages. Even in English separating Theme and Topic can help us resolve some problems that come up when we try to apply the view that necessarily conflates initial position and what the clause is about. For example, Matthiessen and Martin (1991: 45) discuss the thematic nature of English initial there in existential clauses, arguing that “The Theme is one of the elements that realises the feature ‘existential’; it sets up as the point of departure that an Existent will be presented. The new information comes within the Rheme as the Existent.” Martin (1992: 165), says “As unmarked Theme, there is anticipatory; it signals that something is coming—namely the new participant at the end of the clause.” Martin (1995a: 306) also discusses this construction in similar terms: for Halliday the fact that placing there first is a systemic choice precisely parallel to beginning a clause with its Subject in any other process type means that there is a perfectly ordinary candidate for Theme. The fact that it is not assigned a participant function by Halliday . . . is beside the point; there does realise ideational meaning, helping distinguish existential from other relational clauses (cf. existential there was a record player in the corner, attributive the record player was in the corner, wasn’t it? and identifying that’s the record player in the corner. The function of clause-initial there in this construction then is the function of projection that I have been talking about, which I see as different

176

Randy J. LaPolla

from Topic. The controversy here is about whether clause-initial there can be considered ‘what the clause is about,’ i.e. can be considered the Topic of the clause. If instead of insisting that Theme and Topic must coincide, and also recognize that not all clauses have Topics (see Lambrecht’s notion of “sentence focus” clauses, where there is no Topic, a prime example being existential clauses), then we can just say clauseinitial there is Theme, that it is helping the addressee project the mood of the clause and the nature of the clause as existential. The clause is not “about” there, which is non-referential, though, so there is not a Topic, even if it is Subject, and in fact there is no Topic in the clause.23 In discussing the fact that WH-elements are obligatorily clause-initial in English, Halliday says (1970: 358): There is no intrinsic reason why the WH-element should be in first position. In many languages, it occupies whatever position in the clause is appropriate to its status in transitivity. Where the WHelement is assigned first position this must be because first position has some independent structural significance as the expression of another function with which the WH-function is typically associated, a function that will automatically be carried by the WH-element if the clause is of the WH-type. This seems to be recognition that initial position in the clause has a function independent from what the clause is about. Martin (1995a: 304–305) says, “In defining Theme as ‘the point of departure for the message . . . that with which the clause is concerned’, Halliday is attempting to gloss the rationale for placing information first (Theme) rather than last (as unmarked New).” This makes perfect sense, for English, as English generally has New at the end of the clause, and so the information about the Topic relative to which the New is to be understood should ideally come before it. But in a language where New is at the beginning of the clause, this motivation does not hold. There is still motivation for helping the addressee infer (project) the direction in which the interaction is most likely to proceed, but that is not the same as Topic, what the clause is about. As Halliday (1967b) and Knud Lambrecht (1994) have argued, the pragmatic status of referents in the minds of the speaker and hearer (what Lambrecht calls identifiable vs. unidentifiable) is a different type of information from that involved in what Halliday calls ‘information focus’ and Lambrecht calls ‘focus structure’ (e.g. Topic-Comment is ‘predicate focus structure’), and so these two concepts also need to be distinguished, giving us three different semantic domains: the identification of referents;24 the identification of what the clause is about and what is being said about it; and the projection of what is to follow in the interaction.

Arguments for Seeing Theme-Rheme

177

Of course the use of Topics as Themes is highly motivated, as it tells the hearer right away what the clause is about and helps in projecting the content to come (see Fries 1983; Martin 1992), including Martin’s (1992) macro-Theme, hyper-Theme and Theme (i.e. macro-Topic, hyper-Topic, Topic), but I am arguing that we need to expand the typology to allow for languages that do not conflate Theme and Topic, like Tagalog, or that allow identifiable Topics to remain unexpressed, like Chinese. The fact that distinguishing between Given and New also helps the hearer correctly project the speaker’s intention (the reason for marked focus constructions like clefts), and the fact that Topics are generally identifiable referents, and the fact that speakers often use the Topic as the starting point to allow the hearer to better project what the speaker’s intention in the interaction will be, has led to the concepts being collapsed in many theories, but the Tagalog examples where this correlation doesn’t happen show that we need to distinguish different functional structures.

Abbreviations Used 1plinclt

LT

Locative Topic suffix

NCS

No Change of State marker negative

ALL ASSOC

First person inclusive Topic pronoun First person singular possessive pronoun Second person singular possessive pronoun Third person singular pronoun Third person singular Topic pronoun Allative verb Associative marker

AT CL CS

Actor Topic infix Classifier Change of State marker

PT Q REDUP

EXP

Experiential aspect marker Irrealis Actor Topic prefix Linker Locative marker (Tagalog); Locative verb (Chinese)

RUT

1sgposs 2sgposs 3sg 3sgt

IAT LNK LOC

NEG PFV PN POSS PPT

Perfective aspect marker Proper Name

SPEC

Possessive linker Pre-Predicate Topic marker (creates a construction with the Topic preceding the predicate) Patient Topic suffix Question particle Reduplication of initial syllable for marking imperfective and planned actions Realis Undergoer Topic infix Specific referent

STAT

Stative predicate

178

Randy J. LaPolla

Notes * A draft of this chapter was presented at the 40th International Systemic Functional Congress (ISFC40), Sun Yat-sen University, Guangzhou, 15–19 July 2013. I’d like to thank Michael Halliday, Ruqaiya Hasan, David Butt and several others for helpful comments on the ideas presented then, and to thank Jim Martin, David Butt, David Rose, Le Tuan Anh, Jura Matela, Lukáš Zádrapa, Sergey Zinin, Jesse Gates, Gael Fonken and Siva Kalyan for helpful discussions since then. Prof. Halliday passed away on 15 April 2018, and so I would like to dedicate this chapter to him to show my respect, admiration and affection. 1. The earliest mention of the beginning of the utterance as the ground on which the interaction depends is probably Weil 1844[1887: 29]: “it was necessary to lean on something present and known, in order to reach out to something less present, nearer, or unknown. There is then a point of departure, an initial notion which is equally present to him who speaks and to him who hears, which forms, as it were, the ground upon which the two intelligences meet; and another part of discourse which forms the statement (l’énonciation), properly so called. This division is found in almost all we say.” 2. As Butler (2005) points out, a major difference between Lambrecht’s approach and Halliday’s approach is that Lambrecht pays much attention to the different ways that focus structure (the morphosyntactic and/or prosodic differentiation of focus and presupposition to aid the hearer’s interpretation) is manifested in different languages, while Halliday has only discussed English. This has unfortunately led some of his students to assume that his analysis of English, where focus is most often marked only by prosody, is applicable to all languages, but Halliday, as well as his teacher, J. R. Firth, did not approve of imposing the categories of one language on another; each language must be studied inductively. He once expressed to me his dismay that some had taken his analysis of English and applied it directly to other languages. 3. Following best practice in typology, for language-specific (descriptive) categories and constructions I will capitalize the initial letters of the name of the category or construction, but for comparative concepts I will not capitalize the first letter. 4. Halliday (1967b: 200) states that he does not use the terms ‘topic’ and ‘comment’ “because they have tended to be used in a way which conflates what are here regarded as distinct functions, with ‘topic’ meaning both ‘given’ and ‘theme,’” though he conflates topic and point of departure in saying, “The theme is what is being talked about, the point of departure for the clause as a message” (1967b: 212). I am using ‘topic’ as what the clause is about, which I argue is independent of, though often overlaps with, both ‘given’ and ‘point of departure’ (Theme). 5. In Wendy Bowcher’s plenary talk at the conference where this paper was presented, she used the example of a fist, and all the meanings one could ascribe to it; in his plenary talk Geoff Thompson showed a picture that could be understood in different ways. These examples highlight the subjective nature of meaning. 6. As David Butt said about the system networks in his talk at the conference, they are a representation of ‘the flow of behavior.’ I also want to stress, as Ruquaiya Hasan pointed out after my presentation, the assumptions related to language and other cultural aspects that we bring to the context of interpretation are socially constructed, that is, they are social conventions (as well as personal habits), and I do not intend to downplay the social aspect

Arguments for Seeing Theme-Rheme

7.

8.

9.

10.

11.

12. 13.

179

of language use, but the idea is that our personal understanding of language is simply memories of how we have seen or heard language being used, and how we have used it ourselves. The use of the term ‘ostension’ for a communicative act and some other insights that led to this view of communication and cognition are due to Sperber and Wilson (1986), though the theory presented here diverges in many important ways from Relevance Theory. See LaPolla (1997) for discussion. Halliday (2002: 8) states: “it should be possible in the next decade or two to crack the semiotic code, in the sense of coming fully to understand the relationship between observed instances of language behaviour and the underlying system of language—something that has eluded us up till now, so that we have even turned the two into different disciplines, calling only one of them ‘linguistics’ and labelling the other ‘pragmatics’.” My view is that this split occurred because the Structuralists divorced language from communication and made linguistics solely the study of linguistic structure, but this is based on a problematic notion of how language works (see Harris 1981—even Reddy 1979 did not go far enough in refuting what he called the ‘conduit metaphor’), and so the way to ‘crack the code’ is to realize that there is no code; communication isn’t based on coding and decoding, but on inferring the communicator’s intention in performing a communicative act (see LaPolla 2015, 2016). Cf. MacWhinney’s (1977: 152) view of ‘starting points’: “The speaker uses the first element in the English sentence as a starting point for the organization of the sentence as a whole. Similarly, the listener uses the first element in a sentence as a starting point in comprehension.” Gernsbacher and Hargreaves (1988, 1992) also argue for the importance of initial position, discussing many of the experimental results showing what they call ‘the privilege of primacy,’ how what appears in initial position influences the interpretation of the whole utterance. Kim and Kuroshima (2013: 269) cite a large number of studies on interaction in different languages, and argue that they “demonstrate how turn beginnings serve as a prime location in interaction regardless of the language’s typology.” Cf. Schegloff (1987: 71): “Turn-beginnings project a ‘shape’ for the turn, and they’re in that respect potentially critical elements for a speaker’s recipients, who, in having to analyze the turn as it develops, may need the turn beginning resources as part of the materials for their analysis.” See also Heritage (2013). The ability to use abductive inference to anticipate someone’s future actions, or anything for that matter, is a general cognitive mechanism and not limited to communication. There is considerable psycholinguistic evidence for projection, sometimes called ‘anticipation’ or ‘prediction’ in the psychology literature. See e.g. Altmann and Kamide (1999), DeLong et al. (2005), Bubic et al. (2010), Cohn and Paczynski (2013), Holler et al. (2015), Huettig (2015), Levinson and Torreira (2015), Barthel et al. (2016), Barthel et al. (2017), Sauppe (2016). See Hopper (2011), Obana and Haugh (2015), and Blöndal (2015) for examples of collaborative production of the structures used in communication. Within this system, declarative and polar interrogative clauses are distinguished by the order of Subject and Finite, so the Subject must appear in such clauses in order to mark the grammatical mood. This then gives us a very plausible functional explanation for why English is not a so-called ‘pro-drop’ language, and given the fact that it is a grammaticalized aspect of English, and English is the odd-man out in this regard typologically, it would actually make more sense to call English a ‘pro-retaining’ language.

180

Randy J. LaPolla

14. The notion ‘what the clause is about’ is not as straightforward as it might seem, as this can differ between languages. Tagalog has a much richer set of morphosyntactic possibilities for maintaining the reference to particular referents as Topic than English does, and so very often what is the Topic in Tagalog, what the clause clearly is about, will not be the topical Theme or even notional Topic in the English translation (e.g. (12.3–5), where the salsa is the Topic in the Tagalog clauses but cannot be in the English translations in any natural way). English can achieve some of the same results in some contexts using marked constructions like clefts but does not do so to the extent or as naturally as Tagalog does. That the grammatical Topic in Tagalog is what the clause is about can be seen from the fact that “[a]ny predication minus its topic can function as a nominalization understood to denote what would be the topic of that predication” (Adams and Manaster-Ramer 1988: 81). 15. This is not active-passive, but similar to the choice of A construction vs. O construction in Jawarawa, discussed by Dixon (2000, 2004), depending on what is considered to be the Topic of the clause, just with more choices for Topic than just A and O. 16. It happens that in this line the speaker has embedded the relevant clause with the Actor-Topic-marked predicate as the Topic of the main clause (the X in ‘X is easy’), but the phenomenon of Actor-Topic marking is the same whether it is a main clause or embedded. Cf. Gumawa ng salsa ’yung lalaki “That man made sauce.” 17. It is also possible to only have a pause before the predicate instead of using ay. Tagalog is rigidly focus-initial, and there are also other marked-focus constructions for allowing different elements to appear in initial focus position, e.g. circumstantial elements (which usually occur after the predicate), without using the ay construction, e.g. Sa Martes pa tayo bibili ng mesa [loc Tuesday ncs 1plinclt redup-buy poss table] “We won’t buy a table until Tuesday,” a narrow focus construction where the day the table will be bought is the focal information of the clause. 18. This is the problem with non-demonstrative inference, in this case, induction, first identified by Hume (1739): we make generalizations based on our experience, but there is no certainty that future experience will not contradict our generalizations. In this case the scholars only looked at one type of language, and so drew inductive generalizations based on that data, but those generalizations are only valid for that set of data. 19. Assuming this is said without intonation breaks before and after unfortunately, which would mark it as a separate information unit. 20. Compare for example, Kaibigan lang siya ng aking tatay [friend only 3sgt poss 1sgdat+lnk father] “He is only my father’s friend,” with Tinanong lang siya ng aking tatay “My father only asked him,” where kaibigan ng aking tatay “my father’s brother” and tinanong ng aking tatay “asked by my father” are phrases of the same type in Tagalog, though in English we translate them very differently. The fact that they form phrases with the predicate linked by ng is why such arguments cannot appear in initial position the way ang-marked or sa-marked arguments can. See LaPolla (2014) for more on the different types of phrases in Tagalog, and Naylor (2005) and references therein on the isomorphy of referential and predicative phrases. 21. The rare exceptions to this being where ang or demonstrative-marked arguments occur in the predicate and so are not Topics, and when unmarked topics occur in initial position in the construction with ay. In these cases position in the construction helps identify the element as the Topic. There is also the

Arguments for Seeing Theme-Rheme

181

exclamative construction, a referential use in which the pragmatic Topic, if it appears, is not treated as a grammatical Topic but as a possessive modifier, e.g. Ang ganda ng buhok mo! [spec beauty poss hair 2sgposs] “The beauty of your hair!”—cf. Maganda ang buhok mo [beautiful spec hair 2sgposs] “Your hair is beautiful.” This is also true when the predicate is a property concept modified by the intensifier napaka-, e.g. Napakaganda ng buhok mo “Your hair is very beautiful.” 22. This is true of Tagalog as well: existential clauses do not have Topics and can be used to introduce referents that will go on to become Topics in later clauses (see Adams and Manaster-Ramer 1988). 23. Halliday himself, in earlier work (e.g. 1970: 357) argued that all independent indicative clauses have a theme, “with the possible exception of those beginning with dummy it and there, which may be best regarded as having no thematic element in their structure.” In this he seems to have been influenced more by the sense of what the clause is about rather than the starting point of the utterance. Martin (1983) recognizes that not all clauses in Tagalog have Topics. 24. This is only talking about the status of referents in the minds of the speaker and hearer, and not the tracking of referents in discourse. The latter is known as ‘referent tracking’ or ‘reference tracking’ (see Heath 1975; Foley and Van Valin 1984: 321–374; Van Valin 1987; Comrie 1989, LaPolla, to appear, on the typology of grammatical relations as referent-tracking devices). Although Tagalog does manifest a referent tracking system in its marking of the semantic role of the argument that is the Topic, as we can see in (12), I am not specifically talking about referent tracking here except as one aspect of the use of Theme in Tagalog for helping with projection.

References Adams, Karen L. and Alexis Manaster-Ramer. 1988. Some questions of topic/ focus choice in Tagalog. Oceanic Linguistics 27.1/2: 79–101. Altmann, Gerry T.M. and Yuki Kamide. 1999. Incremental interpretation at verbs: Restricting the domain of subsequent reference. Cognition 73: 247–264. doi:10.1016/S0010-0277(99)00059-1 Auer, Peter. 2005. Projection in interaction and projection in grammar. Text 25.1: 7–36. Barthel, Mathias, Antje S. Meyer and Stephen C. Levinson. 2017. Next speakers plan their turn early and speak after turn-final “go-signals”. Frontiers in Psychology 8, Article 393. doi:10.3389/fpsyg.2017.00393 Barthel, Mathias, Sebastian Sauppe, Stephen C. Levinson and Antje S. Meyer. 2016. The timing of utterance planning in task-oriented dialogue: Evidence from a novel list completion paradigm. Frontiers in Psychology 7, Article 1858. doi:10.3389/fpsyg.2016.01858 Beneš, Eduard. 1959. Začátek německé věty z hlediska aktuálního členění větného [The beginning of the German sentence from the point of view of Functional Sentence Perspective]. Časopis pro moderní filologii [Journal of Modern Philology] 41: 205–217. Blöndal, þórunn. 2015. Where grammar meets interaction: Collaborative production of syntactic constructions in Icelandic conversation. University of Helsinki PhD dissertation. Published as vol. 42 of Nordica Helsingiensia. Helsinki:

182

Randy J. LaPolla

Department of Finnish, Finno-Ugrian and Scandinavian Studies, University of Helsinki. Bubic, Andreja, D. Yves von Cramon, and Ricarda I. Schubotz. 2010. Prediction, cognition and the brain. Frontiers in Human Neuroscience 4, Article 25. doi:10.3389/fnhum.2010.00025 Butler, Christopher S. 2005. Focusing on focus: A comparison of Functional Grammar, Role and Reference Grammar and Systemic Functional Grammar. Language Sciences 27: 585–618. Cohn, Neil and Martin Paczynski. 2013. Prediction, events, and the advantage of agents: The processing of semantic roles in visual narrative. Cognitive Psychology 67: 73–97. doi:10.1016/j.cogpsych.2013.07.002 Comrie, Bernard. 1989. Some general properties of reference-tracking systems. In Doug Arnold, Martin Atkinson, Jacques Durand, Clair Grover and Louisa Sadler (eds.), Essays on grammatical theory and universal grammar, 37–51. Oxford: Clarendon. Daneš, František. 1964. Téma/základ/východisko výpovédi [Theme/foundation/ point of departure of the utterance]. Slovo a Slovesnost 25: 148–149. DeLong, Katherine A., Thomas P. Urbach and Marta Kutas. 2005. Probabilistic word pre-activation during language comprehension inferred from electrical brain activity. Nature Neuroscience 8.8: 1117–1121. doi:10.1038/nn1504 Dixon, R. M. W. 2000. A-constructions and O-constructions in Jawawara. International Journal of American Linguistics 66: 22–56. Dixon, R. M. W. 2004. The Jarawara language of Southern Amazonia. Oxford: Oxford University Press. Downing, Angela. 1991. An alternative approach to theme: A systemic functional perspective. Word 42.2: 119–143. doi:10.1080/00437956.1991.11435835 Firbas, Jan. 1964[2009]. On defining the theme in functional sentence analysis. Travaux linguistiques de Prague 1: 267–280. Reprinted in Aleš Svoboda, Jana Chamonikolasov, and Ludmila Urbanov (eds.), Collected works of Jan Firbas, Volume One (1951–1967). Brno: Masaryk University Press. Firbas, Jan. 1987. On the delimitation of the theme in functional sentence perspective. In René Dirven and Vilém Fried (eds.), Functionalism in linguistics, 137–156. Amsterdam and Philadelphia: John Benjamins. Foley, William A. and Robert D. Van Valin, Jr. 1984. Functional syntax and universal grammar. Cambridge: Cambridge University Press. Fox, Barbara. 1985. Word-order inversion and discourse continuity in Tagalog. Text 5.1–2: 39–54. Fries, Peter. 1983. On the status of theme in English: Arguments from discourse. In J. S. Petöfi and Emel Sözer (eds.), Micro and macro connexity of texts, 116– 152. Hamburg: Helmut Buske Verlag. Gernsbacher, Morton Ann and David J. Hargreaves. 1988. Accessing sentence participants: The advantage of first mention. Journal of Memory and Language 21: 699–717. Gernsbacher, Morton Ann and David J. Hargreaves. 1992. The privilege of primacy: Experimental data and cognitive explanations. In Doris L. Payne (ed.), Pragmatics of word order flexibility, 83–116. Amsterdam and Philadelphia: John Benjamins. Halliday, M. A. K. 1967a. Notes on transitivity and theme in English: Part 1. Journal of Linguistics 3.1: 37–81.

Arguments for Seeing Theme-Rheme

183

Halliday, M. A. K. 1967b. Notes on transitivity and theme in English: Part 2. Journal of Linguistics 3.2: 199–244. Halliday, M. A. K. 1968. Notes on transitivity and theme in English: Part 3. Journal of Linguistics 4.2: 179–215. Halliday, M. A. K. 1970. Functional diversity in language as seen from a consideration of modality and mood in English. Foundations of Language 6.3: 322–361. Halliday, M. A. K. 1973. The functional basis of language. In M. A. K. Halliday (ed.), Explorations in the functions of language, 14–39. London: Edward Arnold. Halliday, M. A. K. 1979. Modes of meaning and modes of expression: Types of grammatical structure and their determination by different semantic functions. In D. J. Allerton, Edward Carney and David Holdcroft (eds.), Function and context in linguistic analysis: A festschrift for William Haas, 57–79. Cambridge: Cambridge University Press. Halliday, M. A. K. 1994. An introduction to Functional Grammar (2nd edition). London: Arnold. Halliday, M. A. K. 2002. Introduction: A personal perspective. In Jonathan Webster (ed.), Collected works of M. A. K. Halliday, Volume I: On grammar, 1–14. London and New York: Continuum. Harris, Roy. 1981. The language myth. London: Duckworth. Heath, Jeffrey. 1975. Functional relationships in grammar. Language 51.1: 89–104. Heritage, John. 2013. Turn-initial position and some of its occupants. Journal of Pragmatics 57: 331–337. Holler, Judith, Kobin H. Kendrick, Marisa Casillas and Stephen C. Levinson. 2015. Editorial: Turn-taking in human communicative interaction. Frontiers in Psychology 6, Article 1919. doi:10.3389/fpsyg.2015.01919 Hopper, Paul. 2011. Emergent grammar and temporality in interactional linguistics. In P. Auer and S. Pfänder (eds.), Constructions: Emerging and emergent, 22–44. Berlin: Walter de Gruyter. Hopper, Paul. 2012. Emergent grammar. In James Gee and Michael Handford (eds.), The Routledge handbook of discourse analysis, 301–314. London and New York: Routledge. Hopper, Paul. 2013. Usage and syntax. In Silvia Luraghi and Claudia Paro (eds.), The Bloomsbury companion to syntax, 405–420. London: Bloomsbury Academic. Hopper, Paul and Sandra A. Thompson. 2008. Projectability and clause combining in interaction. In Ritva Laury (ed.), Crosslinguistic studies of clause combining: The multifunctionality of conjunctions, 99–123. Amsterdam: John Benjamins. Huettig, Falk. 2015. Four central questions about prediction in language processing. Brain Research 1626: 118–135. doi:10.1016/j.brainres.2015.02.014 Hume, David. 1739. A treatise of human nature. Oxford: Oxford University Press. Kim, Hye Ri Stephanie and Satomi Kuroshima. 2013.Turn beginnings in interaction: An introduction. Journal of Pragmatics 57: 267–273. (Special issue on turn-beginnings in interaction). Lambrecht, Knud. 1986. Topic, focus, and the grammar of spoken French. PhD dissertation, University of California, Berkeley.

184

Randy J. LaPolla

Lambrecht, Knud. 1994. Information structure and sentence form. Cambridge: Cambridge University Press. Lambrecht, Knud. 2000. When subjects behave like objects: An analysis of the merging of S and O in sentence focus constructions across languages. Studies in Language 24.3: 611–682. LaPolla, Randy J. 1995. Pragmatic relations and word order in Chinese. In Pamela Downing and Michael Noonan (eds.), Word order in discourse, 297– 329. Amsterdam and Philadelphia: Benjamins. LaPolla, Randy J. 1997. Grammaticalization as the fossilization of constraints on interpretation: Towards a single theory of cognition, communication, and the development of language. City University of Hong Kong Seminar in Linguistics, November 6, 1997. http://randylapolla.net/Papers/LaPolla_1997_Gram maticalization_as_the_Fossilization_of_Constraints_on_Interpretation.pdf LaPolla, Randy J. 2014. Constituent structure in a Tagalog text. Language and Linguistics 15.6: 761–774. LaPolla, Randy J. 2015. On the logical necessity of a cultural connection for all aspects of linguistic structure. In Rik De Busser and Randy J. LaPolla (eds.), Language structure and environment: Social, cultural, and natural factors, 33–44. Amsterdam and Philadelphia: John Benjamins. LaPolla, Randy J. 2016. Review of The language myth, by Vyvyan Evans. Studies in Language 40.1: 235–252. doi:10.1075/sl.40.1.09lap LaPolla, Randy J. to appear. Grammatical relations. In The Cambridge handbook of Role and Reference Grammar. Cambridge: Cambridge University Press. LaPolla, Randy J. and Dory Poa. 2006. On describing word order. In Felix Ameka, Alan Dench, and Nicholas Evans (eds.), Catching language: The standing challenge of grammar writing, 269–295. Berlin: Mouton de Gruyter. Levinson, Stephen C. and Francisco Torreira. 2015. Timing in turn-taking and its implications for processing models of language. Frontiers in Psychology 6, Article 731. doi:10.3389/fpsyg.2015.00731 MacWhinney, Brian. 1977. Starting points. Language 53.1: 152–168. Martin, James R. 1981. Conjunction and continuity in Tagalog. In M. A. K. Halliday and James R. Martin (eds.), Readings in systemic linguistics, 310–336. London: Batsford. Martin, James R. 1983. Participant identification in English, Tagalog, and Kâte. Australian Journal of Linguistics 3: 45–74. Martin, James R. 1988. Grammatical conspiracies in Tagalog: family, face and fate—with reference to Benjamin Lee Whorf. In J. D. Benson, M. J. Cummings and W. S. Greaves (eds.), Linguistics in a systemic perspective, 243–300. Amsterdam and Philadelphia: John Benjamins. Martin, James R. 1990. Interpersonal grammatization: mood and modality in Tagalog. Philippine Journal of Linguistics 21.1: 2–51 (Special Issue on the Silver Anniversary of the Language Study Centre of Philippine Normal College 1964–1989—Part 2). Martin, James R. 1992. Theme, method of development and existentiality: The price of reply. Occasional Papers in Systemic Linguistics 6: 147–183. Martin, James R. 1995a. More than what the message is about: English theme. In M. Ghadessy (ed.), Thematic development in English texts, 223–258. London: Pinter (Open Linguistics Series).

Arguments for Seeing Theme-Rheme

185

Martin, James R. 1995b. Logical meaning, interdependency and the linking particle {-ng/na } in Tagalog. Functions of Language 2.2: 189–228. Martin, James R. 1996. Transitivity in Tagalog: A functional interpretation of case. In Margaret Berry, Christopher Butler, Robbin Fawcett and Guowen Huang (eds.), Meaning and form: Systemic functional interpretations, 229– 296. Norwood: Ablex Publishing Corporation. Martin, James R. 2004. Metafunctional profile: Tagalog. In Alice Caffarel, James R. Martin and Christian M. I. M. Matthiessen (eds.), Language typology: A functional perspective, 255–304. Amsterdam and Philadelphia: John Benjamins. Martin, James R. and Priscilla Cruz. 2018. Interpersonal grammar of Tagalog: A systemic functional linguistic perspective. Functions of Language 25.1: 53–91. Mathesius, Vilém. 1929. Zur Satzperspektive im modernen English. In Archiv f. das Studium der neueren Sprachen und Literaturen, Band 155: 203–210. Braunschweig, Berlin, Hamburg: G. Westermann. Mathesius, Vilém. 1939[1947]. O tak zvaném aktuálním členění větném [On the so-called Functional Sentence Perspective]. Slovo a Slovesnost 5: 171–174, reprinted in Čeština a obecný jazykozpyt [The Czech language and general linguistics], Prague 1947, 234–242. Mathesius, Vilém. 1982. Jazyk, kultura a slovesnost (Language, culture and literature). Edited by Joseph Vachek. Prague: Odeon. Mathesius, Vilém. 1983. Functional linguistics. In Josef Vachek and Libuše Dušková (eds.), Praguiana: Some basic and less known aspects of the Prague Linguistic School, 121–142. Amsterdam and New York: John Benjamins. Matthiessen, Christian. 1988. Representational issues in systemic functional grammar. In James D. Benson and William S. Greaves (eds.), Systemic functional approaches to discourse, 136–175. Norwood: Ablex Publishing Corporation. Matthiessen, Christian. 1992. Interpreting the textual metafunctions. In Martin Davies and Louise Ravelli (eds.), Advances in systemic linguistics: Recent theory and practice, 37–81. London and New York: Pinter Publishers. Matthiessen, Christian and M. A. K. Halliday. 2009. Systemic Functional Grammar: A first step into the theory. Beijing: Higher Education Press. Matthiessen, Christian and James R. Martin. 1991. A response to Huddleston’s review of Halliday’s Introduction to Functional Grammar. Occasional Papers in Systemic Linguistics 5: 5–74. McGregor, William. The place of circumstantials in Systemic Functional Grammar. In Martin Davies and Louise Ravelli (eds.), Advances in systemic linguistics: Recent theory and practice, 136–149. London and New York: Pinter Publishers. Naylor, Paz Buenaventura. 1975. Topic, focus, and emphasis in the Tagalog verbal clause. Oceanic Linguistics 14.1: 12–79. Naylor, Paz Buenaventura. 2005. On the stative predicate: Tagalog “existentials” revisited. In Hsiu-chuan Liao and Carl R. Galvez Rubino (eds.), Current issues in Philippine linguistics and anthropology parangal kay Lawrence A. Reid. Manila: The Linguistic Society of the Philippines and SIL Philippines. Obana, Yasuko and Michael Haugh. 2015. Co-authorship of joint utterances in Japanese. Dialogue and Discourse 6.1: 1–25. Ono, Tsuyoshi and Sandra A. Thompson. 2017. Negative scope, temporality, fixedness, and right- and left-branching: Implications for typology and cognitive processing. Studies in Language 41.3: 543–576. doi:10.1075/sl.41.3.01ono

186

Randy J. LaPolla

Peirce, Charles Sanders. 1940. Abduction and induction. In J. Buchler (ed.), Philosophical writings of Peirce, 150–156. New York, NY: Routledge. Reddy, Michael J. 1979. The conduit metaphor—a case of frame conflict in our language about language. In Andrew Ortony (ed.), Metaphor and thought, 284–324. Cambridge: Cambridge University Press. Sauppe, Sebastian. 2016. Verbal semantics drives early anticipatory eye movements during the comprehension of verb initial sentences. Frontiers in Psychology 7, Article 95. doi:10.3389/fpsyg.2016.00095 Schachter, Paul, revised by Lawrence A. Reid. 2008. Tagalog. In Bernard Comrie (ed.), The world’s major languages (2nd edition), 833–855. London: Routledge. Schegloff, Emanuel. 1987. Recycled turn beginnings: A precise repair mechanism in conversation’s turn-taking organization. In G. Button and J.R.E. Lee (eds.), Talk and social organization, 70–85. Clevedon, England: Multilingual Matters Ltd. Schegloff, Emanuel. 1996. Turn organization: One intersection of grammar and interaction. In Elinor Ochs, Emanuel A. Schegloff and Sandra A. Thompson (eds.), Interaction and grammar, 52–133. Cambridge: Cambridge University Press. Sperber, Dan and Deirdre Wilson. 1986. Relevance: Communication and cognition. Cambridge, MA: Harvard University Press. Tanaka, Hiroko. 2000. Turn projection in Japanese talk-in-interaction. Research on Language & Social Interaction 33.1: 1–38. Tanaka, Hiroko. 2001. Adverbials for turn projection in Japanese: Toward a demystification of the “telepathic” mode of communication. Language in Society 30.4: 559–587. Thompson. Sandra A. 1985. Grammar and written discourse: Initial vs. final purpose clauses in English. Text 5.1–2: 55–84. Thompson, Sandra A. and Elizabeth Couper-Kuhlen. 2005. The clause as a locus of grammar and interaction. Language and Linguistics 6.4: 807–837. Trávníček, František. 1962. O tak zvaném aktuálním členění větném [On the socalled functional sentence perspective]. Slovo a Slovesnost 22: 163–171. Van Valin, Robert D. Jr. 1987. Aspects of the interaction of syntax and pragmatics: Discourse coreference mechanisms and the typology of grammatical systems. In Jef Verschueren and Marcella Bertuccelli-Papi (eds.), The pragmatic perspective: selected papers from the 1985 International Pragmatics Conference (Pragmatics and Beyond Companion Series 5), 513–531. Amsterdam and Philadelphia: John Benjamins. Van Valin, Robert D. Jr. and Randy J. LaPolla. 1997. Syntax: Structure, meaning and function. Cambridge: Cambridge University Press. Weil, Henri. 1844[1887]. De l’ordre des mots dans les langues anciennes comparées aux langues modernes: Question de grammaire générale. Paris: Joubert. Translated, with notes and additions, by Charles W. Super, as The order of words in the ancient languages compared with that of the modern languages (Boston: Ginn & Company, 1887).

Part II

Contextualizing Grammar

7

Functional Language Typology A Discourse Semantic Perspective1 J.R. Martin and Beatriz Quiroz

1. Functional Language Typology In this chapter we will use the term ‘typology’ in a very broad sense, to embrace what Matthiessen (e.g. 2018; Matthiessen et al. 2008) refers to a ‘multilingual studies’—including consideration of similarities and differences between two or more languages anywhere along the cline of instantiation from individual texts to generalized systems. We are particularly concerned with exploring what the term ‘functional’ involves when we approach typology from the perspective of Systemic Functional Linguistics (hereafter SFL), concentrating here on discourse semantic systems (Martin 1992; Martin and Rose 2007). In general terms there are five dimensions of functionality to consider in SFL. Fundamental to all of these is what Matthiessen (2014a: Section 3) calls the ‘axial rethink’ whereby Halliday and his colleagues founded the theory (for seminal papers and discussion see Halliday and Martin 1981; Martin 2015a, 2016; Martin and Doran 2015a). As introduced step-by-step in Martin et al. (2013), this rethink involves a paradigmatic orientation to the organization of grammatical description—an orientation grounded syntagmatically in both grammatical relations (functions such as Actor or Epithet) and the grammatical categories (classes such as nominal group or adjective) realizing them. The first two editions of Halliday’s An Introduction to Functional Grammar (1985a, 1994) focus on the syntagmatic axis, introducing the function structures and class syntagms grounding paradigmatic relations; in the next two editions (2004, 2014), revised by Matthiessen, the paradigmatic axis based on these structures and syntagms is brought explicitly into play. The bridging role played by relational categories (i.e. function structures) in this axial rethink thus represents one key respect in which SFL inflected language typology is functional—SFL attends in other words to relations among function structures (e.g. Actor Process Goal or Deictic Epithet Thing) as well as syntagms (i.e. configurations of classes). This foregrounding of paradigmatic relations underpins the second dimension of functionality characterizing SFL’s orientation to language

190

J.R. Martin and Beatriz Quiroz

typology, namely metafunctions. The basic idea here is that systems of choice in language bundle in terms of their relative interdependency on one another—into three sets of relations referred to as ideational, interpersonal and textual.2 Broadly speaking, ideational resources construe the activities, taxonomies and associated properties distinguishing one walk of life from another; interpersonal resources enact social relations of status and solidarity through which we commune as strangers, acquaintances, peers and colleagues, kith and kin; and textual resources compose ideational and interpersonal meaning as waves of information flow adjusted to our channels of communication as texts unfold. Metafunctions represent the intrinsic functional organization of language, generalizing the work that languages have evolved to do (Matthiessen 2014b: Section 4.1). The textual metafunction draws our attention to the fact that language manifests as discourse, bringing a third dimension, co-textual functionality, into the picture. At stake here are the resources language has evolved to relate phases of text of indefinite extent to one another—ideationally in terms of taxonomic relations between entities and conjunctive relations between occurrences, interpersonally in terms of the negotiation of proposition and proposals and feelings in exchange structure and textually in terms of entity tracking and longer wave-lengths of information flow (Martin 2015b). Textual meaning also draws our attention to the fact that co-textual relations obtain in specific contexts of language use, bringing contextual functionality into the picture. The key variables here are register, including field (what is going on or being talked about), tenor (the status and affiliation of the interlocutors), mode (the affordances of the channel of communication as far as context dependency is concerned) and genre (the staged goal-oriented social process involved). Register and genre model the extrinsic functionality of social semiosis, the fourth dimension of functionality germane to SFL (Bartlett 2017; Bowcher 2017; Martin and Rose 2008; Tann 2017). Finally, there is the question of appliability—the use that is made of the model. SFL is a theory designed to be useful, in contexts ranging from educational linguistics to clinical linguistics through forensic linguistics and beyond (e.g. Rochester and Martin 1979; Rose and Martin 2012; Zappavigna and Martin 2018). The four dimensions of functionality described thus far (axially functional, metafunctional, co-textually functional and contextually functional) have all evolved as part of a theory/ practice dialectic inspired by Halliday’s conception of linguistics as an ideologically committed form of social action (Halliday 1985b). SFL involves deployable functionality in this respect. In this chapter we will concentrate on the third dimension of SFL functionality, co-textual functionality, and develop the role it might play with respect to similarities and differences among languages as we explore their instantiation in texts.

Functional Language Typology

191

2. Upward Bound Comparing languages is a problematic enterprise, especially so for linguists using SFL, precisely because of the privileging of paradigmatic relations by their theory. As linguists influenced by Saussure have long acknowledged, the meaning of any linguistic choice lies in its valeur. And if the valeur of a choice in one language is different from its valeur in another because it enters into different oppositions, then the meaning of that choice is not the same. Language learners are of course familiar with this problem as they engage with gender, tense, determiner and pronoun systems different from those in languages they already control. And languages may not even be consistent in themselves. A language like Spanish for example has a distinct set of oppositions for clause rank personal pronouns, verbal group pronominal clitics, possessive determiners and possessive pronouns, not to mention the variations from one dialect of Spanish to another3 and the possibility of realizing an entity through verb inflection alone. What does it mean to compare pronoun systems in Spanish, with one another, or to pronoun systems in English, or any other language? If the valeur is different, what is the common ground? Let’s explore this problem of common ground step-by-step to see where it takes us. We confidently wrote pronoun systems earlier, as if it was clear what we were talking about. Consider the following Spanish clauses (everyday invitations to the cinema). The Circumstance of Accompaniment in the first example is a phrase, with the pronoun nosotras as Complement—con nosotras ‘with us.fem’; in the second example the Circumstance is a word—conmigo ‘with.me,’ whose morphological analysis we will set aside here.4 The issue at stake here is whether conmigo is a pronoun. (1) ¿Quieres ir con nosotras al cine? qu-ieres

ir

con

want-prs.2sg go.inf with Process/Actor

nosotras al 1pl.f

cine

to.det.def.sg.m cinema

Accompaniment Location

“Would you like to go to the movies with us?”

(2) ¿Quieres ir conmigo al cine? qu-ieres

ir

conmigo

want-prs.2sg go.inf with.1sg Process/Actor

al

cine

to.det.def.sg.m cinema

Accompaniment Location

“Would you like to go to the movies with me?”

192

J.R. Martin and Beatriz Quiroz

The wrinkle here is that we have what appears to be the same meaning but divergent structural realizations. In SFL we can bring axis into play at this point and argue that as far as paradigmatic relations are concerned, we have the same systems in play—the same accompaniment system choice at clause rank and the same person system choices at group rank. The contrast between con nosotras (a prepositional phrase) and conmigo (a word) is thus not systemic but structural. In (1) the choices for accompaniment and person are realized as a phrase and in (2) as a word. This example shows how the dimension of axis, with its complementarity of paradigmatic and syntagmatic relations (the latter realizing the former) can be used to explore similarity and differences between languages. For example, as the glosses included reveal, English always realizes comparable choices as phrases (with us, with me), not single words. Consider now the following Spanish clauses, recounting a moment in a trip to the cinema. In both we have three Participants realized by nominal groups at clause rank.5 In the first example all three nominal groups have nouns realizing the nuclear Thing function (niña ‘girl,’ entrada ‘ticket,’ amigo ‘friend’), in bold; in (4) the first and third of these nominal groups are realized by a pronoun (ella ‘she,’ él ‘him’). (3) La niña le compró una entrada a un amigo . la

niña

le compr-ó

una

entrada a

un

amigo

det.def.sg.f girl

dat.3sg = det. ticket buy-pst.3sg indf.sg.f

sm det.indf. friend sg.m

Actor

Process

Beneficiary

Goal

“The girl bought a ticket for a friend.”6

(4) Ella le compró una entrada a él . ella

le = compr-ó

una

entrada a

él

she.3sg.f dat.3sg = buy-pst.3sg det.indf.sg.f ticket

sm 3sg.m

Actor

Beneficiary

Process

Goal

“She bought him a ticket.”

Spanish clause Participants can be alternatively realized inside verbal groups by pronominal clitics and inflectional morphology, as exemplified here. (5) Se la compró. se = la= compr-ó dat.3 = acc.3sg.f = buy-pst.3sg Beneficiary/Goal/Process/Actor “She bought it for him.”

Functional Language Typology

193

And the phenomenon of ‘clitic doubling’ means that Participants can be realized both ways, both inside and alongside a verbal group.7 (6) La entrada la niña se la compró a un amigo. la

entrada la

niña se = la= compr-ó

a

un

amigo

det.def. ticket sg.f

det.def. girl sg.f

dat3 = acc3s.f = sm det.def. friend buy-pst.3sg sg.m

Goal

Actor

Process

Beneficiary

“The girl bought a ticket for a friend.”

(7) La entrada ella se la compró a él. La

entrada ella

se lacompr-ó

a

él

det.def.sg.f ticket

she.3sg.f dat3 = acc3sg.f sm 3sg.m = buy-pst.3sg

Goal

Actor

Process

Beneficiary

“The girl bought a ticket for him.”

The wrinkle here then is that in Spanish Participants can be realized by nominal groups at clause rank, or inside verbal groups, or both—whereas a language like English, lacking pronominal clitics, is restricted to realizing Participants as nominal groups at clause rank. In SFL the concept of rank can be used to generalize across these two realization systems, with transitivity systems at clause rank showing how the languages are similar, and realization statements potentially diversifying the realization of the Spanish system across clause and group rank structures (bringing verbal group systems into the picture where required). Consider now the following Spanish clause, which uses the demonstrative eso ‘that’ to refer to the ticket.8 (8) La niña le compró eso a un amigo. la

niña le = compr-ó eso

a

un

amigo

det.def.sg.f girl

dat.3sg = buy-pst.3sg

dem.n sm det.indf.sg.m friend

Actor

Process

Goal

Beneficiary

“The girl bought it for/from a friend.”

Spanish uses a demonstrative rather than a pronoun as Thing here because at clause rank personal pronouns are rarely used to realize Things referring to non-conscious entities in nominal groups. The ‘ticket’

194

J.R. Martin and Beatriz Quiroz

Participant can on the other hand be alternatively realized as a pronominal clitic (la ‘it’);9 consciousness is not an issue for pronominal clitics. (9) La niña se la compró a un amigo. la

niña se = la = compr-ó

a

un

amigo

det.def.sg.f girl

dat3 = acc3sg.f = buy-pst.3sg sm det.indf.sg.m friend

Actor

Goal/Process

Beneficiary

“The girl bought it for a friend.”

This brings another system into the picture, the system of demonstratives, which is used to realize either a Modifier or Head in nominal group structure (e.g. esa entrada ‘that ticket’ or esa ‘that’). This system involves a very different set of oppositions to those for Spanish possessive determiner, possessive pronoun, personal pronoun and pronominal clitic systems (since it involves near speaker, near listener and near neither options, outlined in Table 7.1).10

Table 7.1 Spanish Demonstratives number

singular

plural

gender

masculine

feminine

neuter

masculine

feminine

near speaker near listener near neither

este ese aquel

esta esa aquella

esto eso aquello

estos esos aquellos

estas esas aquellas

The wrinkle here is that in nominal groups referring to non-conscious entities, demonstratives play the same role as pronouns. We have two systems, but the same role. But what does saying ‘the same role’ mean? In SFL the concept of stratification can be used to make this explicit. The relevant discourse semantic system is identification (Martin 1992; Martin and Rose 2007) whose options of presuming track entities in discourse. These options can be used to generalize across a range of lexicogrammatical realizations, including Spanish personal pronoun, pronominal clitic and demonstrative systems just reviewed. They also provide us with a comparable vantage point from which to explore differences in participant tracking resources and their use across languages (as in Martin 1983). Unlike Spanish, for example, English personal pronouns can track non-conscious entities (e.g. the wave toppled the trees/it toppled them). And unlike English, Spanish is comfortable tracking entities with implicit unmarked Themes between

Functional Language Typology

195

clause complexes—there are no Spanish personal pronouns tracking the movie goers in the second and third clauses of the sequence that follows. (10) La niña le compró una entrada a un amigo. la

niña le = compr-ó una

entrada a

un

amigo

det.def. girl sg.f

dat3sg = buy-pst.3sg

det.indf. ticket sg.f

sm det.indf. friend sg.m

Actor

Process

Goal

Beneficiary

“the girl bought a ticket for a friend”

(11) Disfrutaron la película disfrut-aron

la

enjoy-pst.3pl

det.def.sg.f film

película

Process/Senser Phenomenon “they enjoyed the movie”

(12) y volvieron a casa. y

volv-ieron

a

casa

and return-pst.3pl to house Process/Actor

Location

“and returned home”

The general rule of thumb we are developing here is that when things looks different, move ‘up’ and look again (Matthiessen 2014b: 42). From the point of view of axis, different structures may be realizing comparable systems; from the point of view of constituency, structures at different ranks may be realizing comparable systems; and from the point of view of stratification, an array of lexicogrammatical resources may perform comparable discourse semantic functions. In SFL we can in principle push this two steps further on, asking how different discourse semantic systems realize comparable field, mode and tenor variables, and how different combinations of register variables cooperate to realize comparable genre systems. As far as functional language typology is concerned, what is important is to be explicit about the perspective we are comparing languages from—which stratum, which rank, which axis? We also need to keep in mind that any generalizations we make will have to be treated as defeasible, since

196

J.R. Martin and Beatriz Quiroz

what looks the same or different from one vantage point can look rather different from another. With this in mind we will now present on overview of the major discourse semantic systems that language typologists can bring into play when comparing diverging lexicogrammatical realizations of comparable discourse phenomena across languages. We will then focus on two of these systems, identification and periodicity, and compare English and Spanish in terms of how they track entities in discourse. For foundational work on discourse semantic language typology, see Gleason (1968), Longacre (1976, 1983) and Callow (1974).

3. Discourse Semantic Parameters In SFL the extrinsic functional organization of language (field, tenor and mode) is correlated with the intrinsic functional organization of language (ideational, interpersonal and textual metafunctions). Field is treated as by and large construed through ideational meaning, tenor as by and large enacted through interpersonal meaning and mode as by and large composed through textual meaning. For Martin and his colleagues these register variables are coordinated by genre, stage by stage as texts unfold (Martin and Rose 2008). As Figure 7.1 implies, genre and register are modeled as strata of meaning—embracing the social semiotic context functionality outlined in Section 1.

Figure 7.1 Extrinsic Functionality (Context) and Intrinsic Functionality (Language)

Functional Language Typology

197

Alongside being intrinsically organized by metafunction, in SFL language is organized by stratum—as choices in discourse semantic systems realized by choices in lexicogrammatical systems realized in turn by choices in phonological systems (Figure 7.2). At the level of discourse semantics we are concerned with meaning realized both inside and between clauses, whether these clauses have any grammatical relation to one another or not. The focus in other words is on co-text. As argued in Martin (2015b), the units between which discourse semantic relations obtain are of indefinite extent.

Figure 7.2 Language Strata

We will briefly introduce six discourse semantic systems here, organized by metafunction in Table 7.2. For a detailed presentation of these systems see Martin (1992), Martin and White (2005) and Martin and Rose (2007); useful introductory overviews include Butler (2003), Martin (2009, 2014, 2015c, 2019) and Tann (2017). For each system we will note the diversification of grammatical systems realizing discourse semantic ones, including what we will refer to as metaphorical realizations; we will also exemplify what we mean by co-textual relations between units of discourse that are not grammatically related to one another. By metaphorical realizations we mean grammatical choices that symbolize discourse semantic ones rather than realizing them directly (as discussed in Halliday 1985a/2014: Chapter 10; Simon-Vandenbergen et al. 2003). Each system will be exemplified for English, followed by an open-ended list of suggestions for considering the realization of comparable discourse functions in lexicogrammar across languages.

198

J.R. Martin and Beatriz Quiroz Table 7.2 Discourse Semantic Systems (Organized by Metafunction) metafunction

discourse semantic systems

ideational

ideation connexion negotiation appraisal identification periodicity

interpersonal textual

3.1 Ideational Discourse Semantics The key ideational systems are ideation and connexion.11 ideation comprises resources for construing experience as occurrences and relations. In terms of diversification it allows us, for example, to position figures lexicogrammatically through a range of clause types (Halliday and Matthiessen 1999; Hao 2015, forthcoming_a, forthcoming_b). (13) “They’ve won,” she mused. (behavioral clause realizing the positioning of an occurrence figure) (14) She knew they’d won. (mental clause realizing the positioning of an occurrence figure) (15) She was aware they’d won. (relational clause realizing the positioning of an occurrence figure) In terms of grammatical metaphor ideation allows us to realize an occurrence figure congruently as a clause or metaphorically as a nominal group: (16) She knew they’d win. (congruent realization of an occurrence figure)12 (17) Their win delighted her. (metaphorical realization of an occurrence figure) In terms of meaning beyond the clause ideation allows us for example to anticipate activity—using one clause to name what’s to come (taking stock) and others to spell it out:

Functional Language Typology

199

(18) and it was time to take stock: ⇓ 250 people had needed the lifesavers to pull them out, of whom 210 were OK once back on land. Thirty-five needed mouth to mouth to be restored to consciousness, while five people perished. (FitzSimons 2005) The basic framework for analyzing ideation adopted here is based on Halliday and Matthiessen (1999), Hao (2015, forthcoming a) and Doran and Martin (forthcoming). As flagged in Figure 7.3, it deals with sequences consisting of one or more figures and figures consisting of one or more elements; and the main types of element are occurrences, entities and qualities—which in various combinations constitute figures.

Figure 7.3 Basic ideation Systems

Drawing on this framework we might ask, across languages: With respect to which realms of practice are entities taxonomized (via classification and/or composition)—everyday, specialized, technical, metadiscursive etc.? What realms of quality are available to describe these entities? How are occurrences construed (as single or serial verb events)? What realms of manner are available to qualify these occurrences (and to what extent is manner lexicalized verbally)? How are occurrences, entities and settings organized in figures? How are figures positioned, evaluated and agentively instigated? Which realms of practice are sequenced, and through how many tiers? Connexion comprises resources for relating discourse semantic figures (both occurrence figures and state figures) to one another in sequences. In

200

J.R. Martin and Beatriz Quiroz

terms of lexicogrammatical diversification it allows us to connect various types of clause and clause complex to one another (via addition, comparison, time and cause): (19) Thanks to the lifesavers pulling them out, 210 were OK once back on land. (20) Because the lifesavers pulled them out, 210 were OK once back on land. (21) The lifesavers pulled them out so that 210 were OK once back on land. (22) The lifesavers pulled them out. So 210 were OK once back on land. In terms of grammatical metaphor connexion allows us to realize relations between figures metaphorically as single clauses: (23) (congruent clause complex construing cause) They were restored to consciousness because the lifesavers gave them mouth-to-mouth. (24) (metaphorical cause in the clause) Mouth-to-mouth resuscitation ensured their restoration to consciousness. In terms of meaning beyond the clause connexion allows us to connect indefinitely long phases of discourse to one another (viz the long justification of the preceding sentence introduced by for here): (25) In their long and glorious history, this still stands as the finest hour of the Australian lifesaving movement. ⇑ For, ignoring their own possible peril, the Bondi boys now charged into the surf, some attached to one of the seven reels available, some relying only on their own strength. As one, they began pulling the people out. On the shore, many survivors were resuscitated, as the Bondi clubhouse was turned into a kind of emergency clearing house, and ambulances from all over Sydney town descended and carried the victims away. Finally, just half an hour after the waves hit, the water was cleared of bobbing heads and waving arms, and it was time to take stock: 250 people had needed the lifesavers to pull them out, of whom 210 were OK once back on land. Thirty-five needed mouth to mouth to be restored to consciousness, while five people perished. (FitzSimons 2005)

Functional Language Typology

201

Looking across languages as far as comparable sequencing resources are concerned, we might ask: What range of conjunctive resources (e.g. additive, comparative, temporal and causal)? How do these resources interact with other resources such as ‘continuity’ (i.e. engagement adverbs/particles such as still, yet, already), tense/ aspect, temporal setting and experience reiteration to connect figures? To what extent are cohesive (inter-sentential), paratactic and hypotactic realizations available? Do these resources interact with participant identification (‘switchreference’ systems) and/or thematic or informational systems? Is there a set of ‘internal’ conjunctive resources specialized for ‘public’ (≈ written) discourse? 3.2 Interpersonal Discourse Semantic Resources The key interpersonal systems are negotiation and appraisal. negotiation comprises resources for enacting social relations in dialogue. In terms of diversification it allows us, for example, to realize greeting moves through a range of more and less lexicalized structures. (26) Hey. (27) Good morning. (28) How’s it going? (29) What a surprise! (30) Lovely to see you. (31) Didn’t know you were back in town. In terms of grammatical metaphor, negotiation allows us to realize moves directly or metaphorically through so-called ‘indirect speech acts.’ (32) What’s your name? (congruent interrogative clause requesting information) —Lionel. (33) Tell me your name. (metaphorical imperative requesting information) —Lionel. (34) Your name is? (metaphorical declarative requesting information) —Lionel.

202

J.R. Martin and Beatriz Quiroz

In terms of meaning beyond the clause, negotiation allows us to relate moves in conversation, including moves comprising several clauses (as in the following request and compliance sequence): (35) What I want you to do now is, I need you to tell us in a really loud voice, OK, what happened on that particular day, alright. So I want you to tell me everything that happened on that day that led to you being stopped by the police with the telephone, OK. Can you do that for us? Thanks. Off you go. ⇓⇑ —When I was walking to my mate’s house, this guy just came up to me and he goes ‘Do you want to buy a phone?, and I go ‘Nah’, and I go ‘Do you want to swap?’ He wanted to swap, with my phone. And he looked at my phone and he goes, ‘Yeah, swap’. And we swapped, and I went and stayed at my mate’s house and then, when it came to night time and I was going back home, I was walking and he was walking me up the road and the police just came and brought us. (Zappavigna and Martin 2018) Looking across languages as far as comparable dialogic resources are concerned, we might ask: How are congruent moves realized, in initiations and responses (first vs. second pair parts)? What range of resources establish the arguability of an exchange (via tense/aspect, modality, polarity, evidentiality etc.)? Are the entities most at risk in an exchange made explicit—necessarily or as needed—via position, adposition, cliticization or inflection, and with or without agreement as a modally responsible ‘Subject?’ Which resources position the speaker in relation to the addressee (e.g. deference, vocation)—necessarily or as needed? Are they mainly grammatical or lexical? Are there resources positioning the speaker in relation to non-addressees (e.g. honorification)? Are they mainly grammatical or lexical? What patterns of ellipsis characterize compliant negotiations of initiating moves (second pair parts)? What patterns of minor and major clause characterize non-compliant negotiation (i.e. tracking and challenging moves, aka insertion sequences)?

Functional Language Typology

203

Appraisal comprises resources for enacting social relations by sharing attitudes (Martin and White 2005). In terms of diversification it allows us, for example, to realize effects across a range of grammatical structures: (36) Sadly they lost. (37) They sadly made their way home. (38) They were sad because they lost. (39) The sad fans drowned their sorrows. (40) The score saddened them. In terms of grammatical metaphor appraisal allows us to realize feelings as if they were things and deploy them accordingly. (41) They were so palpably sad that it darkened proceedings. (congruent adjectival feeling) (42) Their palpable sadness darkened proceedings. (metaphorical nominalized feeling) In terms of meaning beyond the clause, appraisal allows us for example to evaluate indefinitely long phases of discourse. The extended connection example we used does more than justify a proposition; it also glorifies the lifesavers’ rescue operations: (43) In their long and glorious history, this still stands as the finest hour of the Australian lifesaving movement. ⇓ For, ignoring their own possible peril, the Bondi boys now charged into the surf, some attached to one of the seven reels available, some relying only on their own strength. As one, they began pulling the people out. On the shore, many survivors were resuscitated, as the Bondi clubhouse was turned into a kind of emergency clearing house, and ambulances from all over Sydney town descended and carried the victims away. Finally, just half an hour after the waves hit, the water was cleared of bobbing heads and waving arms, and it was time to take stock: 250 people had needed the lifesavers to pull them out, of whom 210 were OK once back on land. Thirty-five needed mouth to mouth to be restored to consciousness, while five people perished. (FitzSimons 2005) Looking across languages as far as comparable resources for negotiating feeling are concerned, we might ask: How are attitudes lexicalized? How suitable is their classification as affect, judgement or appreciation?

204

J.R. Martin and Beatriz Quiroz What resources are deployed to grade attitude (e.g. sub-modification, repetition, graded lexis, prosodic phonology)? How is attitude propagated prosodically (e.g. via saturation, dominance or intensity)? What range of resources lexicogrammaticalize attitude (comment adverbs, manner adverbs, mental processes, mental state relational processes, epithets)? How is attitude focalized (i.e. sourced via projection, circumstance of angle)? How are relevant voices acknowledged, sourced, affirmed or shut down (via projection, modality, evidentiality, negation, concession or comment particles or adverbials)?

The key textual systems are identification and periodicity. Identification comprises resources for composing discourse in terms of introducing entities and tracking them once there.13 In terms of diversification it allows us for example to track entities through a range of nominal resources: (44) Pope Francis arrived. (45) He was dressed in white. (46) The Argentinian thanked Chileans for their hospitality during his Jesuitical training there. (47) This pope publicly acknowledged the child abuse scandals. In terms of meaning beyond the clause identification allows us to identify and track indefinitely long phases of discourse; the pronoun it is used in this way to reference the activity that subsequently unfolds.14 (48) At three o’clock there was still not the slightest clue that this afternoon would forever be known as “Black Sunday” in the annals of Sydney. Then it happened. ⇓ With a roar like a Bondi tram running amok, an enormous wave suddenly rolled over the thousands in the surf, including those many standing on the large sandbank just out from the shore—knocking them all over as it went. And then another wave hit, and then another. The huge waves, just like that, piggy-backed their way further and further up the beach and grabbed everything they could along the way—from babies to toddlers to adolescents to beach umbrellas, to old blokes and young sheilas alike, and then made a mad dash for the open sea again, carrying all before it and sweeping everyone off the sandbank and into the deep channel next to it in the process. (FitzSimons 2005)

Functional Language Typology

205

Looking across languages as far as comparable resources for introducing and tracking entities are concerned, we might ask: How are entities introduced in a text? Is the ‘prominence’ of entities in the discourse flagged in some way when they are introduced? How are entities tracked in a text, via what range of entity coding (deixis), Theme selection (including ‘switch-reference’) and ellipsis? What resources are deployed to track discourse (text reference)? Does introducing and tracking entities interact with transitivity (e.g. introducing entities in existential clauses or restrictions on identifiability for participant or circumstantial roles)? Periodicity comprises resources for composing text as waves of information (phases of a field, we might say). The basic idea here is that there is a hierarchy of periodicity, extending from the small wave-lengths of tone group and clause to an indefinite number of indefinitely long phases of discourse. In the example that follows we have a Kicker foreshadowing an entire feature article about a beachside tragedy, a topic sentence introducing what happened when the waves rolled up the beach further down and a retrospective summary of the effect of the inundations. A wide range of resources, including text reference and ideational grammatical metaphor (in bold), cooperate with one another to scaffold information flow along these lines. (49) Big waves and Bondi Beach have always gone together, writes Peter FitzSimons, but no one had ever seen the ocean rise up with a strength such as this. ⇒ At three o’clock there was still not the slightest clue that this afternoon would forever be known as “Black Sunday” in the annals of Sydney. Then it happened. ⇒ With a roar like a Bondi tram running amok, an enormous wave suddenly rolled over the thousands in the surf, including those many standing on the large sandbank just out from the shore—knocking them all over as it went. And then another wave hit, and then another. The huge waves, just like that, piggy-backed their way further and further up the beach and grabbed everything they could along the way—from babies to toddlers to adolescents to beach umbrellas, to old blokes and young sheilas alike, and then made a mad dash for the open sea again, carrying all before it and sweeping everyone off the sandbank and into the deep channel next to it in the process.

206

J.R. Martin and Beatriz Quiroz ⇐ In no more than 20 seconds, that peaceful scene had been tragically transformed into utter chaos. Now, the boiling surf, with yet more large waves continuing to roll over, was filled with distressed folk waving for help. (FitzSimons 2005)

Looking across languages as far as comparable resources for composing text as waves of information are concerned, we might ask: How is an orientation to the field established and sustained, via explicit or implicit unmarked Themes, encoded by position, adposition or inflection (± agreement)? How are shifts in orientation to the field managed, e.g. via marked or absolute Themes, encoded by position, adposition or inflection and/ or through intonation? How is new information highlighted, by position, adposition or inflection and/or through intonation? What is the unmarked dis/association of Theme and New (assuming such categories are useful)? Through what resources can this mapping be adjusted? How are longer wave-lengths of theme and information managed (e.g. via elaborating figures and sequences and/or rhythm and intonation and/or body language)? The discourse semantic resources briefly reviewed here are outlined by metafunction in Figure 7.4. Space precludes detailed consideration of the interaction of these systems in the realization of register and genre. In the

Figure 7.4 Discourse Semantic Systems

Functional Language Typology

207

following section we will however consider the interaction of identification and periodicity in two short phases of Spanish discourse.

4. Textual Discourse Semantics In Section 2 we toured a range of grammatical resources realizing the discourse semantic system of identification in Spanish, by way of showing the importance of reconsidering structural similarity and difference from the perspective of system, reconsidering lower rank similarity and difference from the perspective of higher ranks and reconsidering lower stratum similarity and difference from the perspective of higher strata. In this section we will explore these resources a little further, including consideration of their interaction with periodicity. We’ll deal with an excerpt from a text titled La ola maldita (Guzmán 2010), a focus text from Martin et al. (forthcoming). La ola maldita, ‘the hellish wave,’ is a feature article focusing on the impact of an earthquake and tidal wave disaster in the Maule and Biobío regions of Chile in 2010. The first excerpt deals with the rescue of an elderly man who was left behind in the city of Constitución when his relatives fled to the mountains. We will use this text to focus on the introduction and tracking of the elderly man and his carer (detailed glossing for this excerpt is provided)). (50) En su retirada, el sargento Valladares vio que una mujer pedía ayuda. Al entrar en la casa medio derrumbada, encontró a un anciano en silla de ruedas que los miraba con angustia. Sus familiares estaban en los cerros. La señora que pedía auxilio era la mujer que lo cuidaba. Ella vivía en otro lugar y corrió a ver al anciano y lo encontró solo. Estaba mojado y gemía de miedo. El sargento lo sacó de ahí y lo llevó a la casa de la mujer. El saqueo No es difícil imaginar la angustia de ese anciano abandonado. Tampoco es incomprensible el miedo que deben haber sentido sus parientes. (Guzmán 2010) ‘During his return, Sergeant Valladares saw that a woman was asking for help. On entering the half-demolished house, he met an old man in a wheel chair who was looking at them anxiously. His family were in the hills. The lady who was asking for help was the woman who looked after him. She was living in another place and ran towards the old man and found him alone. He was wet and was moaning with fear. The sergeant took him away from there and left him in the woman’s house. The Looting

208

J.R. Martin and Beatriz Quiroz It is not difficult to imagine the distress of that old abandoned man. Neither is the fear that his relatives must have felt incomprehensible.’15

The Spanish resources used to introduce (presenting reference) and track (presuming reference) the carer are presented in Table 7.3. The carer is introduced with non-specific deixis as una mujer, ‘a woman,’ and then tracked through a range of resources, including specific deixis (la, ‘the’), a personal pronoun (ella, ‘she’), a pronominal clitic including reference to both her and Sergeant Valladares (los, ‘them’) and two third person singular past tense suffixes (-ió and -ó) in clauses without any other reference to the carer. We repeat the relevant part of the excerpt here, with the presenting reference marked in italics and presuming reference in bold. (51) En su retirada, el sargento Valladares vio que una mujer pedía ayuda. Al entrar en la casa medio derrumbada, encontró a un anciano en silla de ruedas que los miraba con angustia. Sus familiares estaban en los cerros. La señora que pedía auxilio era la mujer que lo cuidaba. Ella vivía en otro lugar y corrió a ver al anciano y lo encontró solo. Estaba mojado y gemía de miedo. El sargento lo sacó de ahí y lo llevó a la casa de la mujer.

Table 7.3 Entity Tracking in Spanish (la mujer ‘the woman’) presuming

↓

nominal group Deictic specific determiners

nominal group Thing personal pronouns

↓

presenting

una mujer det.indf.sg.f ‘a woman’

verbal group: pronominal clitics

verbal group: inflectional morphology only

los acc. 3pl.m la(señora . . . ) det.def.sg.f ‘the lady’ la(mujer . . . ) det.def.sg.f ‘the woman’ ella 3sg.f -ió pst.3sg -ó pst.3sg la(mujer) det.def.sg.f ‘the woman’

Functional Language Typology

209

Table 7.4 Entity Tracking in English (the woman) presenting nominal group Thing personal pronouns

↓

↓

nominal group Deictic specific determiners a woman

them the(lady . . .) the(woman . . .) she the(woman’s)

The English resources used in the translation presented here are outlined in Table 7.4. Like Spanish, English uses non-specific deixis to introduce the carer (a woman); the two languages are similar in their insistence on making presenting vs. presuming reference explicit in every nominal group (unlike say Tagalog, Korean or Chinese, which can choose to make this explicit or not). Subsequently the carer is presumed through specific deixis (the) and personal pronouns (them, she). English has no pronominal clitics and so uses the personal pronoun them to translate the Spanish clitic los. And the English verbs deployed make no number or person distinctions in past tense, and so there are no explicit resources tracking entities corresponding to Spanish corr-ió ‘run-pst.3sg’ and encontr-ó ‘meet-pst.3sg’. We repeat the relevant part of the translation here, with the presenting reference marked in italics and presuming reference in bold. (52) During his return, Sergeant Valladares saw that a woman was asking for help. On entering the half-demolished house, he met an old man in a wheel chair who was looking at them anxiously. His relatives were in the hills. The lady who was asking for help was the woman who looked after him. She was living in another place and ran towards the old man and found him alone. He was wet and moaning with fear. The sergeant took him away from there and carried him to the woman’s house.\ Turning to the old man, as outlined in Table 7.5, he is introduced with non-specific deixis as un anciano, ‘an old man,’ and then tracked through a range of resources, including specific deixis (sus, ‘his,’; al, ‘to the,’;16 ese, ‘that’), pronominal clitics (lo, ‘him’) and two third person singular past tense suffixes (-aba and -ía) in clauses without any other marking of his role. We repeat the text here, with the presenting reference marked in italics and presuming reference in bold.

210

J.R. Martin and Beatriz Quiroz

Table 7.5 Entity Tracking in Spanish (el anciano ‘the old man’) presuming

↓

nominal group Deictic specific (possessive) determiners un anciano det.indf. sg.m ‘an old man’

nominal group Thing personal pronouns

↓

presenting

verbal group: pronominal clitics

verbal group: inflectional morphology only -aba impf.3sg

sus(familiares) det.poss.3pl ‘his relatives’ lo acc.3sg.m al(anciano) to.det.def.sg.m ‘the old man’ lo acc.3sg.m -aba impf.3sg -ía impf.3sg lo acc.3sg.m lo acc.3sg.m ese(anciano . . . ) det.dem.sg.m ‘that old man’ sus(parientes) det.poss.3pl ‘his relatives’

(53) En su retirada, el sargento Valladares vio que una mujer pedía ayuda. Al entrar en la casa medio derrumbada, encontró a un anciano en silla de ruedas que los miraba con angustia. Sus familiares estaban en los cerros. La señora que pedía auxilio era la mujer que lo cuidaba. Ella vivía en otro lugar y corrió a ver al anciano y lo encontró solo. Estaba mojado y gemía de miedo. El sargento lo sacó de ahí y lo llevó a la casa de la mujer. El saqueo No es difícil imaginar la angustia de ese anciano abandonado. Tampoco es incomprensible el miedo que deben haber sentido sus parientes.

Functional Language Typology

211

The English resources used in the translation presented are outlined in Table 7.6. Like Spanish, English uses non-specific deixis to introduce the old man. Subsequently the old man is presumed through specific deixis (his, the, that) and personal pronouns (him, he). English has no pronominal clitics and so uses the personal pronoun him to translate the Spanish clitic lo. Subject ellipsis is not possible in the first clause of a paratactic clause complex in English and so a pronoun Subject is deployed in the he was wet clause. The English verb be has a distinctive third person singular past tense realization was, which can be taken as presuming the old man in the continuing and was moaning with fear clause in that clause complex. We repeat the translation here, with the presenting reference marked in italics and presuming reference in bold. (54) During his return, Sergeant Valladares saw that a woman was asking for help. On entering the half-demolished house, he met an old man in a wheel chair who was looking at them anxiously. His family were in the hills. The lady who was asking for help was the woman who looked after him. She was living in another place and ran towards the old man and found him alone. He was wet and was moaning with fear. The sergeant took him away from there and left him in the woman’s house. The Looting It is not difficult to imagine the distress of that old abandoned man. Neither is the fear that his relatives must have felt incomprehensible. Table 7.6 Entity Tracking in English (the old man) presenting

an old man . . .

nominal group Thing personal pronouns

↓

↓

nominal group Deictic specific (possessive) determiners

verbal group: inflectional morphology only

his (family) him the (old man) him he was (moaning) him him that (old . . . man) his (relatives)

212

J.R. Martin and Beatriz Quiroz

In relation to our discussion of entity tracking so far, it might be objected that we have glossed over the contribution of periodicity (in particular what is sometimes referred to as ‘Subject ellipsis’ in English). Ellipsed Subjects are common in paratactic clause complexes in English, as illustrated in the translations repeated here (the ellipsed Subject has been inserted in parentheses in the English translations). A comparable pattern of uncommitted meaning is also found in Spanish, where however it is common both within and between clause complexes. So we find a clause like estaba mojado, literally ‘was wet’ in the following, far more commonly in fact than él estaba mojado, ‘he was wet’—as shown in the following example (55–60) (transitivity analysis based on the criteria for Spanish in Quiroz 2013). (55) Ella vivía en otro lugar ella

viv-ía

3sg.f

live-impf.3sg in

Carrier Process

en otro

lugar

other place

Location

“She lived in another place”

(56) y corrió y

corr-ió

and ran-pst.3sg Process/Actor “and (she) ran”

(57) a ver al anciano a

v-er

lk see-inf

al

anciano

sm.def.sg.m

old.man

Process Phenomenon “to check on the old.man”

(58) y lo encontró solo. y

lo = encontr-ó

and acc3sg.m = find-pst.3sg

solo alone

Phenomenon/Process/Senser Attribute “and (she) found him alone.”

Functional Language Typology

213

(59) Estaba mojado est-aba

mojado

be-impf.3sg

wet

Process/Carrier Attribute “(He) was wet”

(60) y gemía de miedo. y

gem-ía

and moan-impf.3sg

de

miedo

from fear

Process/Behaver Range “and (he) was moaning from fear.”

It is sometimes felt by speakers of languages like English that it is its rich verb morphology that facilitates ‘implicitness’ between clause complexes in Spanish. An exemplary paradigm is presented in Table 7.7, for indicative mood, past tense in Chilean Spanish. Note that the English translations (i.e. lost, did . . . look, did . . . look, didn’t allow, didn’t find, didn’t keep, gave) provide no comparable information about the person or number of an agreeing participant. But there is abundant evidence against this explanation of Spanish, as provided by the many languages that do not mark person or number in their verbal groups but favor the same patterns of uncommitted meaning between clause complexes as Spanish (for example Tagalog, Korean and Chinese; see Matthiessen (2004) for relevant discussion). For a more general explanation we have to bring the realization of the discourse semantic system of periodicity into the picture. In order to explore this system in Spanish we will draw on work by Moyano (2016) and Martin et al. (forthcoming). Their analysis suggests that by and large it is the entity potentially realized as a co-referential nominal group in a Spanish clause that functions as unmarked Theme. Further, by way of maintaining our orientation to the field in a phase of discourse, the unmarked Theme is usually left implicit (i.e. the potential to be realized by a nominal group is usually not taken up). Let’s explore this now in a longer excerpt from La ola maldita, an exemplum dealing with Sergeant Cristián Valladares’s experiences on the day (including the excerpt used to explore identification earlier). This phase of the feature article is presented ranking clause by ranking clause here, with their verbal groups underlined and agreeing entities, where present, in bold.

214

J.R. Martin and Beatriz Quiroz

Table 7.7 Spanish Verb Morphology Paradigm (Indicative Mood, Past Tense)17 indicative mood past tense

verbal group

example

speaker

First person singular

addressee (informal)

Second person singular

addressee (formal)

Third person singular

non-interlocutor

Third person singular

speaker plus

First person plural

addressee plus

Third person plural

non-interlocutors

Third person plural

yo me vin-e a Constitución el año 85 “I moved to Constitución in the year 1985.” ¿Cómo era cuando lleg-aste acá? “What was it like when you got here?” Lo que pasa es que usted t-uvo el servicio en enero “What happens is that you had the service in January.” Ricardo Fuentes oy-ó los ruegos de Nora “Ricardo Fuentes heard Nora begging.” Hay dos cuerpos que no encontr-amos. “There are two bodies we didn’t find.” Muy bien, ¿escuch-aron lo que dijo Pablo? “Very good, did you hear what Pablo said?” Los adultos empez-aron a pedir ayuda. “The adults started to ask for help.”

(61) Otros ancianos, en cambio, vieron [[cómo sus familias escapaban y los dejaban botados]]. otros ancianos

en cambio

v-ieron

[[clause]]

other old people on the other hand see-pst.3pl (see below) Senser

Process

“Other old people, however, saw . . .”

(62) [[cómo sus familias escapaban . . . ]] cómo

sus

how

their families escape-impf.3pl

Manner Actor

familias escap-aban Process

“. . . how their families escaped . . .”

Phenomenon

Functional Language Typology

215

(63) [[ . . . y los dejaban botados]] y

los = dej-aban

botados

and acc.3pl = leave-impf.3pl

abandoned

Attributor/Carrier/Process Attribute “. . . and left them abandoned.”

(64) El sargento de la Armada, Cristián Valladares, se encontró con uno de ellos el

sargento de la

Armada Cristián Valladares se = encontr-ó

the sergeant of the Navy

con

uno de ellos

Cristián Valladares refl = with one of 3pl.m met-pst.3sg

Actor

Process

Goal

“Navy Sergeant Cristián Valladares met one of them”

(65) cuando intentaba llegar a la Capitanía de Puerto cuando intent-aba when

lleg-ar

a

la

Capitanía de Puerto

try-impf.3sg arrive-inf at the command of port Process/Actor

Location

“when he tried to reach the Harbour Command”

(66) para ayudar. para ayud-ar to

help-inf Process

“to help.”

(67) Eran las seis de la mañana, er-an

las

seis de la

be-pst.3pl det.def.pl.f six Process

Ambience

“It was six in the morning,”

of

mañana

det.def.sg.f morning

216

J.R. Martin and Beatriz Quiroz

(68) ya estaba clareando. ya

est-aba

clare-ando

already

be-impf.3sg

clear-ger

Process “(and) was already growing light.”

(69) Mientras se acercaba a la costa mientras se = acerc-aba while

a

la

costa

refl = approach-impf.3sg to det.def.sg.f coast Process/Actor

Location

“As (he) was approaching the coast”

(70) oyó los gritos de la isla Orrego oy-ó

los

gritos de

hear-pst.3sg

det.def.pl.m cries

la

isla

from det.def.sg.f island Orrego

Process/Senser Phenomenon “(he) heard the cries from Orrego island”

(71) y se preguntó y

Se = pregunt-ó

and refl = ask-pst.3sg Receiver/Process/Sayer “and (he) asked himself”

(72) si habría un bote [[con qué ir a buscar a la gente]]. si habr-ía

un

bote

[[clause]]

if have-cond.3sg det.indf.sg.m boat (see below) Process

Existent

“if there were a boat . . .”

(73) [[con qué ir a buscar a la gente]] con

qué

ir

a

busc-ar

a

la

gente

with which go.inf lk pick.up-inf sm det.def.sg.f people Manner

Process

Orrego

Goal

“with which to go to pick up the people.”

Functional Language Typology

217

(74) Entonces vio [[cómo el río se recogía y se formaba otra ola gigantesca]]. entonces vi-o

[[clause]]

then

(see below)

see-pst.3s

Process/Senser Phenomenon “Then (he) saw . . .”

(75) [[cómo el río se recogía . . . cómo

el

río

se = recog-ía

how

det.def.sg.m river refl = gather-pst.3sg

Manner Actor

Process

“. . . how the river receded . . .”

(76) . . . y se formaba otra ola gigantesca]] y

se = form-aba

otra

ola

gigantesca

and refl = form-impf.3sg other wave gigantic Process

Actor

“. . . and another gigantic wave was forming.”

(77) Mientras huía mientras hu-ía while flee-impf.3sg Process/Actor “While (he) was fleeing”

(78) vio [[que esa ola tapaba los árboles]], vi-o

[[clause]]

see-pst.3sg

(see below)

Process/Senser “(he) saw that wave cover the trees”

(79) [[que esa ola tapaba los árboles]], que

esa

ola

tap-aba

los

árboles

that det.dem.sg.f wave cover-impf.3sg det.def.pl.m trees Actor “that wave cover the trees”

Process

Goal

218

J.R. Martin and Beatriz Quiroz

(80) por lo que calcula por

lo

que

calcul-a

because of 3sg.n that estimate-prs.3sg Cause

Process/Senser

“because of which (he) estimates”

(81) que tendría unos 10 metros. que

tendr-ía

unos

10 metros

that have-cond.3sg det.indf.pl.m 10 meters Process/Carrier Attribute “that (it) would be some 10 meters high.”

(82) Debe haber sido la tercera o cuarta gran ola [[que azotó Constitución]]. deb-e

hab-er

s-ido

la

musthave-inf bedet. prs.3sg prtcp def. sg.f Process/Token

tercera o 3rd

cuarta gran ola

or 4th

big

[[clause]]

wave (see below)

Value

“(It) must have been the third or fourth biggest wave that struck Constitución.”

(83) [[que azotó Constitución]] que

azot-ó

Constitución

rel

batter-pst.3sg Constitución

Actor Process

Goal

“that struck Constitución.”

(84) En su retirada, el sargento Valladares vio [[que una mujer pedía ayuda]]. en su in

retirada el

det. retreat poss. 3sg

Location

sargento Valladares vi-o

det. sergeant def. sg.m Senser

[[clause]]

Valladares see(see below) pst.3sg Process

Phenomenon

“In his retreat Sergeant Valladares saw a woman asking for help.”

Functional Language Typology

219

(85) [[que una mujer pedía ayuda]] que

una

mujer

ped-ía

ayuda

that det.indf.sg.f woman ask-impf.3sg help Sayer

Process

Verbiage

“a woman asking for help”

(86) Al entrar en la casa medio derrumbada, al

entr-ar

en

la

casa

medio derrumb-ada

to.det. enter-inf into det.def.sg.f house half def.sg.m Process

torn.down-prtcp.sg.f

Location

“On entering the half-disintegrating house”

(87) encontró a un anciano en silla de ruedas [[que los miraba con angustia]]. encontr-ó

a

find-pst.3sg

sm det. old.man in indf.sg.m

un

anciano

en silla

de

ruedas [[clause]]

chair lk wheels (see below)

Process/Senser Phenomenon “(he) found an old man in a wheel chair that was watching them in anguish.”

(88) [[que los miraba con angustia]] que

los = mir-aba

rel

acc.3pl.m = watch-impf.3sg with anguish

con

Behaver Scope/Process

angustia

Manner

“that was watching them in anguish.”

(89) Sus familiares estaban en los cerros. sus

familiares est-aban

en los

det.poss.3pl family

be-impf.3pl in

Carrier

Process

“His family was in the hills.”

cerros

det.def.pl.m hills

Location/Attribute

220

J.R. Martin and Beatriz Quiroz

(90) La señora [[que pedía auxilio]] era la mujer [[que lo cuidaba]]. la

señora [[clause]]

det.def. lady sg.f

er-a

la

mujer

[[clause]]

(see below) be-impf.3sg det.def. woman (see below) sg.f

Token

Process

Value

“The lady who was asking for help was the woman who looked after him.”

(91) [[que pedía auxilio]] que

ped-ía

rel

ask-impf.3sg help

Sayer Process

auxilio Verbiage

“who was asking for help”

(92) [[que lo cuidaba]] que

lo = cuid-aba

rel

acc.3sg.m = look.after-impf.3sg

Actor Goal/Process “who looked after him.”

(93) Ella vivía en otro lugar ella

viv-ía

3sg.f

live-impf.3sg in

Carrier Process

en otro

(94) y corrió y

corr-ió

and run-pst.3sg Process/Actor “and ran”

other place

Location

“She lived in another place”

lugar

Functional Language Typology (95) a ver al anciano a

v-er

lk see-inf

al

anciano

sm.det.sg.m old.man

Process Phenomenon “to check on the old man”

(96) y lo encontró solo. y

lo = encontr-ó

solo

and acc.3sg.m = find-pst.3sg

alone

Phenomenon/Process/Senser Depictive Attribute “and found him alone.”

(97) Estaba mojado est-aba

mojado

be-impf.3sg

wet

Process/Carrier Attribute “(He) was wet”

(98) y gemía de miedo. y

gem-ía

de

and moan-impf.3sg

miedo

from fear

Process/Behaver Range “and was moaning from fear.”

(99) El sargento lo sacó de ahí el

sargento lo = sac-ó

de

ahí

det.def.sg.m sergeant

acc.3sg.m = take.out-pst.3sg from there

Actor

Goal/Process

“The sergeant took him out of there”

Location

221

222

J.R. Martin and Beatriz Quiroz

(100) y lo llevó a la casa de la mujer. y

lo = llev-ó

and acc.3sg.m = carry-pst.3sg

a

la

casa

de la

to det. house of def.sg.f

mujer

det. woman def.sg.f

Goal/Process/Actor Location “and carried him to the woman’s house.”

The pattern of Themes in this text is presented in Table 7.8. Column 1 (linkers) presents realizations of connexion. Column 2 presents nonagreeing ideation realized before the Predicator (i.e. a non-agreeing Participant or a Circumstance). Column 3 presents nominal group Themes (i.e. agreeing participant roles)—in parenthesis where not explicitly realized. Column 4 presents the Predicator. Column 5 allows for the possibility of nominal group Themes following the Predicator (a marked choice in Spanish that is not taken up in this text).18 The entity realized by pronouns is spelled out in square brackets (e.g. ella [la señora]; implicit Themes are made explicit in parentheses—e.g. (Valladares)); and the hyphen indicates the absence of a Theme. As the Table proposes, there are seven explicit Themes (otros ancianos [‘other old people’], El sargento de la Armada, Cristián Valladares [‘Navy Sergeant Cristián Valladares’], el sargento Valladares [‘Sergeant Valladares’], sus familiares [‘his family’], La señora que pedía auxilio [‘the woman asking for help’], ella [‘she’] and el sargento [‘the sergeant’]). Sergeant Valladares is reintroduced by name twice in the story—once to return our gaze to him following consideration of the size of the wave, and once to return our gaze following consideration of the old man he rescued and his carer. Elsewhere our gaze is fixed on the field through implicit Themes, both within and between clause complexes (i.e. within and between sentences). As noted earlier there are no explicit Themes following the Predicator in this story; and there is just one piece of non-agreeing ideation in pre-Predicator position (en su retirada, ‘in his retreat’). For three of the clauses we have not recognized a Theme. The first is a time clause, with agreement between the Predicator and a co-referential nominal group in bold (cf. singular era la una, ‘it was one o’clock) (101) Eran las seis de la mañana, “It was six in the morning,” The second is a weather clause, with a Predicator realizing modal responsibility (estaba clareando, ‘was growing light’); but from an ideational perspective it is assigning responsibility to no entity in particular.

Functional Language Typology

223

Table 7.8 Theme in the ‘Valladares to the rescue’ Phase of La Ola Maldita linkers

pre-Predicator ideation

cuando

nominal group Theme

Predicator

Otros ancianos El sargento . . . Cristián Valladares (Valladares)

vieron se encontró

para

(Valladares) — —

Mientras

(Valladares) (Valladares) (Valladares) — (Valladares) (Valladares) (Valladares) (Valladares) (esa ola) (esa ola)

y si Entonces Mientras por lo que que En su retirada Al

y y y

el sargento Valladares (Valladares) (Valladares) Sus familiares La señora que pedía auxilio Ella [la señora] (la señora) (la señora) (el anciano) El sargento (Valladares)

nominal group Theme

intentaba llegar ayudar Eran estaba clareando se acercaba oyó se preguntó habría vio huía vio calcula tendría Debe haber sido vio entrar encontró estaban era vivía corrió lo encontró Estaba lo sacó lo llevó

(102) ya estaba clareando. “(and) was already growing light.” The third is an existential clause, with agreement assigning modal responsibility19 to the Existent. (103) si habría un bote [[con qué ir a buscar a la gente]]. “if there were a boat with which to go to pick up the people.”

224

J.R. Martin and Beatriz Quiroz

Our reasoning for not recognizing a Theme in these clauses is partly grammatical. The co-referential nominal group in these clause types is either necessarily missing or fairly fixed in final position. And it has to either be there or not; in clauses like the ones exemplified there is no explicit Theme vs. implicit Theme choice readily available. Our reasoning is also in part discourse semantic. These clauses are not involved in sustaining our gaze on the field. Existential clauses typically initiate an orientation to the field by introducing a new entity, which is thereafter sustained by other clause types. And both time and weather clauses add occasional information about the setting of a phase of discourse, but they do not contribute to its overall method of development. Turning to the English translation, the pattern of Themes in this text is presented in Table 7.9. Column 1 (linkers) presents realizations of connexion. Column 2 presents non-agreeing ideation realized before the Predicator. Column 3 presents nominal group Themes (i.e. agreeing entities)—in parenthesis where not explicitly realized. Column 4 presents the Predicator. Word initial uppercase letters indicate follow standard English orthography (indicating either a proper name or the beginning of a sentence). The entity realized by pronouns is spelled out in square brackets (e.g. he [= Valladares]; implicit Themes are made explicit in parentheses—e.g. (the woman). As a comparison of Tables 7.8 and 7.9 reveals, the main difference between Spanish and English is the number of explicit Themes—seven in the Spanish source text and 20 in the English translation. Nonetheless, both Tables reflect a general trend across languages to sustain a focus on the field through relatively repetitive implicit resources. For Spanish, this is via implicit Themes, while for English, this is via pronouns. So the major issue to be explained has to do with why a language might prefer more commitment of meaning than another. English, as we have seen, uses an explicit Subject pronoun when moving from one clause complex to another where Spanish does not—in (108) for example (in bold); but like Spanish, English prefers ellipsis in paratactic extending clauses (in 105, 107 and 109 as follows). (104) Ella vivía en otro lugar “She lived in another place” (105) y corrió “and ran” (106) a ver al anciano “to check on the old man” (107) y lo encontró solo. “and found him alone.”

Functional Language Typology

225

Table 7.9 Theme in the English Translation of the ‘Valladares to the rescue’ Phase of La Ola Maldita linkers

pre-Predicator ideation

when to While and if Then While so that In his retreat On

and and

and

nominal group Theme

Predicator

Other old people Navy Sergeant Cristián Valladares he [Valladares] (Valladares) It [ambience] (ambience) he [Valladares] he [Valladares] (Valladares) there he [Valladares] he [Valladares] he [Valladares] he [Valladares] it [the wave] It [the wave] Sergeant Valladares (Valladares) he [Valladares] His family [old man] The woman asking for help She [the woman] (the woman) (the woman) He [= old man] (the old man) The sergeant (Valladares)

saw met tried to reach help was was growing was approaching heard asked himself was saw was escaping saw estimates was must have been saw entering met was was lived ran found him was was moaning took him left him

(108) Estaba mojado “He was wet” (109) y gemía de miedo. “and was moaning from fear.” We see a comparable use of an explicit Subject pronoun in (112). But like Spanish, English prefers implicitness in dependent non-finite clauses (111). (110) En su retirada, el sargento Valladares vio [[que una mujer pedía ayuda]]. “In his retreat Sergeant Valladares saw a woman [[asking for help]].”

226

J.R. Martin and Beatriz Quiroz (111) Al entrar en la casa medio derrumbada, “On entering the half-disintegrating house” (112) encontró a un anciano en silla de ruedas [[que los miraba con angustia]]. “he met an old man in a wheel chair [[that was watching them in anguish]].”

And English deploys a pronoun where there is no entity to presume—for example the non-referential it pronoun in the first of the following ambient clauses. (113) Eran las seis de la mañana, “It was six in the morning,” (114) ya estaba clareando. “(and) was already growing light.” What is it that ties these differences between Spanish and English together? To explore this we need to shift our attention from textual meaning to interpersonal meaning and ask questions about how the mood resources of Spanish and English position clauses as moves in exchange structure. As outlined in Quiroz (2013, 2015, 2017a, 2018) for Spanish and Halliday (1985a) for English, the crucial difference is that Spanish positions moves through its verbal group morphology and intonation, while English relies on the presence or absence of Subject and Finite functions and their sequence. The contrasts for declarative, interrogative and imperative mood are illustrated here. For Spanish, declarative and interrogative are distinguished from one another by intonation (falling vs. rising, indicated by the arrows in 115 and 116), and from imperatives by verb morphology (and clitic positioning); for English, declarative and interrogative are distinguished form one another in terms of Subject before Finite vs. Finite before Subject sequences (usually with a difference in intonation comparable to Spanish), and from imperatives by the absence of the Subject and Finite functions. (115) Vio [[cómo el río se recogía y se formaba otra ola gigantesca]].↓ vi-o

[[clause]]

see-pst.3sg Process/Senser

Phenomenon

“He saw how the river receded and another giant wave was forming.”

Functional Language Typology

227

(116) ¿Vio [[cómo el río se recogía y se formaba otra ola gigantesca]]? vi-o

[[clause]]

see-pst.3sg Process/Senser

Phenomenon

“Did he see how the river receded and another giant wave was forming.”

(117) Ve [[cómo el río se recoge y se forma otra ola gigantesca]]. v-e

[[clause]]

see-imp.2sg Process/Senser

Phenomenon

“See how the river recedes and another giant wave forms.”

This shows that English relies on the presence of an explicit Subject to show mood distinctions (whereas Spanish does not). The explicit English Subject is there for interpersonal reasons, not textual ones. This accounts for the explicit Subject in the initial clause of paratactic clause complexes; the Subject function can be ellipsed as long as the mood is held constant. (118) He was wet and was moaning from fear. To shift the mood, the Subject would have to be made explicit. (119) He was wet but was he moaning from fear? Non-finite clauses are not negotiable moves—this is signaled by the missing Finite function (so there are no terms, i.e. tense or modality, to argue with). This means that an explicit Subject is only necessary to clarify a transitivity role that might otherwise be ambiguous; it is not required for mood. (120) On entering the half-disintegrating house he met an old man in a wheel chair that was watching them in anguish. Ambient clauses dealing with time and weather on the other hand are negotiable, and require a Subject and Finite function even though there is no entity available to fill a relevant transitivity role.

228

J.R. Martin and Beatriz Quiroz (121) It was six in the morning, and was already growing light.

English existential clauses need a Subject for the same reasons, as reflected in the translation from the Spanish here (cf. the tagged existential, confirming There as Subject: There was a boat, wasn’t there?). (122) si habría un bote [[con qué ir a buscar a la gente]] “if there were a boat with which to go to pick up the people.” In short in order to compare and contrast identification in Spanish and English we need to bring periodicity into the picture—since implicit Themes play an important role in tracking entities in both languages. Then, to explain why Spanish makes use of implicit Themes more often than English, we need to turn from textual resources to interpersonal ones and take into account the constraints on Subject ellipsis in English deriving from the realization of the discourse semantic system of negotiation in its grammar of mood. A discourse semantic perspective encourages us to reason around in a way that is sensitive to the common need to introduce and track entities while attending to the distinctive interplay of the systems involved.

5. Defeasible Language Typology In Section 1 of this chapter we explored what it means to position SFL as a functional theory of language—as axially functional, metafunctional, co-textually functional, contextually functional and ‘appliably’ functional. In Section 2 we took what are commonly referred to as ‘pronouns’ in Spanish, and argued that any analyses we propose are fundamentally defeasible—since analyses of structural similarity and difference have to be reconsidered from the perspective of system, that analyses of similarity and difference at a lower rank have be reconsidered from the perspective of a higher one and that analyses of similarity and difference at a lower stratum have to be reconsidered form the perspective of a higher one. Section 3 introduced the discourse semantic systems that we used to afford this perspective on lexicogrammar. With this repositioning in mind we then brought SFL’s multifunctional vision to bear on a comparison of Spanish and English resources for introducing entities in discourse and keeping track of them once there. This was explored in one context of language use, a feature article comprising several story genres (Martin and Rose 2008), the most relevant appliability for the fields of translation studies (Kim et al. in preparation) and second or foreign language teaching. Adopting a ‘top-down’

Functional Language Typology

229

perspective we began by exploring the lexicogrammatical realization of identification in Spanish and English. This led on to a consideration of periodicity in order to account for the contribution made by implicit Themes. And this led on in turn to a consideration of the realization of the interpersonal discourse semantic system of negotiation in the lexicogrammatical system of mood in the two languages—since the involvement of English Subject in the realization of mood constrains Subject ellipsis and therefore the possibility of implicit Themes. This exercise brought axis, metafunction and co-text into play. For a richer account, these three dimensions of SFL functionality need of course to be supplemented by consideration of a wider range of contexts (i.e. additional fields, modes, tenors and genres) and a wider range of applications (e.g. in clinical, educational or forensic linguistics). The following text, for example,20 from a law lecture, introduces the case of British Airways and Taylor and does so using a mode and tenor in which English Subjects are regularly elided between clause complexes (marked  Ø). Everything we said prior, in other words, about identification and periodicity in Spanish and English, has to be reconsidered from the perspective of the spoken mode and relaxed tenor21 of this legal field (namely ‘misrepresentation’) and the case presentation genre. (123) So . . . British Airways and Taylor. Taylor was a guy who worked for a consumer agency, a government agency. He was representing a guy called Edmonds. Edmonds had lived on the island of Bermuda for a long time, down there in the Caribbean. Ø Liked that sunshine. Ø Decided after many years that he’d like to take a holiday in London. Ø Experience some of that fog and dampness. Ø Get up there. Ø Take in some shows. Ø Go to Wimbledon. Ø Hadn’t had a holiday for a long, long time and so this is a big deal for him. Cheap airfare was on offer if you were the early bird. Ø Get in a couple of months early you book your name, Ø much cheaper. So he did this. Ø Books his flight for London. Ø Plans his holidays. Ø Becomes a little bit concerned that maybe his place will get taken if there’s a crowd, and someone who’d paid full fare would be in his seat. So he rings them up gets some reassurance that that’s not the case. There will be a seat for him. And in fact gets a letter confirming the seat that’s there. Obviously, he turns up at the airport in Bermuda, Ø no seat. Our goal here, for reasons of space, was however necessarily constrained. We have only been able to offer a glimpse of how functional language typology might evolve in a multifunctional model of language such as SFL (inspired by pioneering work such as Martin 1983; Rose 2001;

230

J.R. Martin and Beatriz Quiroz

Caffarel et al. 2004; Teruya et al. 2007; Matthiessen et al. 2008; Teruya and Matthiessen 2015; Martin and Doran 2015b; Mwinlaaru and Xuan 2016; Martin 2018; Mwinlaaru et al. 2018). The metaphor of shunting22 has long been iconized in SFL, deriving as it does from Halliday’s (1961) seminal articulation of the model. The main message of this paper is simply that of rearticulating the need to shunt—to compare and contrast languages from multiple perspectives—and not privilege categories over relations (class over function in SFL terms), syntagmatic relations over paradigmatic ones (structure over system in SFL terms), morphemes over clauses (lower ranks over higher ones in SFL terms), syntax over semantics (lexicogrammar over discourse semantics in SFL terms), semantics over pragmatics (ideational meaning over interpersonal and textual meaning in SFL terms) or language over context (text over field, tenor, mode and genre in SFL terms). All of these complementarities need to be respected as such, with our gaze shifting systematically ‘upwards’ and ‘downwards’ to recontextualize what we find.

Notes 1. This chapter draws on discussions that are elaborated in Martin et al. (forthcoming), a project based in large part on Quiroz’s foundational work on Spanish grammar (especially, for this chapter, Quiroz 2013, 2015, 2017a, 2017b, 2018). The work on Spanish presented in this chapter has been partially funded by grant FRCAI1720 and by CONICYT-FONDECYT grant 11170674. 2. Distinctive types of structure have been associated with these metafunctions (e.g. Halliday 1979; Martin 1996)—particulate structures with ideational meaning, prosodic structure with interpersonal meaning and periodic structures with textual meaning. 3. The four types of ‘pronoun’ referred to here are illustrated in the following sentence (in sequence, personal pronoun ella [‘she’], pronominal clitic le [‘for her’], non-salient possessive determiner su [‘her’], salient possessive determiner mía [‘mine’]): la entrada que ella le compró a su amiga no es mía [‘the ticket she bought her friend isn’t mine’]. The main dialectal variation has to do with second person (the use or not of os, vos, vosotros, vosotras, vuestro, vuestra) and the blurring of the distinction between accusative and dative pronominal clitics when they refer to animate entities (e.g. la/le quiero conocer, ‘I want to meet her’). Paradigms for the oppositions in these systems are presented in the Appendix. 4. Note that the preposition con followed by the oblique form of the first person pronoun, mi, is ungrammatical: *con mi (cf. contigo [‘with you’], which patterns like conmigo vs. con usted [‘with you singular formal’], con él [‘with him’], con ella [‘with her’] etc., which pattern like con nosotras). 5. We are treating a un amigo, ‘for a friend,’ as an adpositional nominal group here, on the grounds that it can be realized inside the verbal group as a pronominal clitic. See full argumentation distinguishing them from prepositional phrases in Quiroz (2013, pp. 201–203; 315–316). 6. This example can alternatively be translated as “The girl bought a ticket from a friend.”

Functional Language Typology

231

7. The specific conditions favoring ‘doubling’ vary across Spanish varieties. See, for example, Belloro (2007), Silva-Corvalán (1981), Vázquez Rozas and García Salido (2012) and Pericchi Paga et al. (2017), among others, for more detail. 8. The neuter form eso is used here in preference to the feminine esa (which would sound contrastive). The neuter demonstrative is used for text reference and reference to very general entities. 9. The ticket participant can also be ‘doubled’ in a clause such as esa la niña se la compró a un amigo or esa se la compró la niña a un amigo, ‘The girl bought it for a friend.’ 10. Here, person has been generalized in terms of ‘distance,’ in the proportions first person : near speaker :: second person : near listener :: third person : near neither. 11. The system name connexion is taken from Hao (2015, 2018, forthcoming a, this volume), replacing Martin’s earlier term conjunction in order to more clearly differentiate discourse semantic and lexicogrammatical terminology (reacting in particular to confusion arising from the use of grammatical terminology for semantic description in Halliday and Matthiessen 1999). 12. The term congruent is used in SFL for direct non-metaphorical realizations. 13. For a complementary ‘west-coast functionalist’ perspective, see Du Bois (1980). 14. Martin (1992) treats text references of this kind as textual grammatical metaphors, since they involve a nominal group referring to a phase of discourse rather than an entity (a person, place or thing). It is however difficult to see how reference of this kind is symbolizing a phase of discourse, and there is arguably no stratal tension (i.e. there are not two meanings in play, one of which could be interpreted literally). In retrospect the concept of grammatical metaphor was probably being pushed too far. 15. The English translation was undertaken by the authors before consideration of the identification and periodicity analyses discussed in this chapter. 16. The contracted word al is used in Spanish in place of the ungrammatical syntagm *a el. 17. Examples for ‘addressee informal,’ ‘addressee formal’ and ‘addressee plus’ in Table 7 have not been taken from Guzmán (2010) but from other texts making up the data of project CONICYT-FONDECYT 11170674—including the sociolinguistic corpus of Spanish from Santiago, Chile (PRESEEA-SA 2014–2019). 18. In order to discuss the realization of Theme following the Predicator, we would have to bring ideational clause grammar (transitivity systems) and information systems realized through intonation into the picture, two undertakings beyond the scope of this study; for discussion, see Martin et al. (forthcoming). 19. Traditionally, Spanish grammarians have argued that in existential clauses with ‘impersonal’ haber there should be no agreement between the Process and the Existent. Their ‘subjectless’ nature is backed up by the fact that the Existent in these clauses may be tracked with the neuter accusative clitic lo. There is however a clear example of agreement in our text: ya no habían embarcaciones, ‘there weren’t any boats any longer.’ As in English (e.g. here’s an example, here’s some examples/here are some examples), there is tension in the grammar as far as agreement is concerned. 20. Our thanks to Susan Hood for sharing this text with us. 21. In adopting this informal style the law lecturer is in effect impersonating Edmonds, the plaintiff. 22. In British railway terminology ‘shunting’ involves pushing or pulling a train or one of its cars from one line of rails to another in order to drop off, pick up or rearrange cars (comparable to ‘switching’ in American railway

232

J.R. Martin and Beatriz Quiroz

terminology). Halliday uses the metaphor specifically to insist on the need to move up and down among ranks in grammar description; we are arguing for the need for movement among levels of all kinds here.

References Bartlett, T. 2017. Context in SFL. In Bartlett and O’Grady (eds.). 375–390. Bartlett, T. and G. O’Grady (eds.). 2017. The Routledge handbook of systemic functional linguistics. London: Routledge. Belloro, V. 2007. Spanish clitic doubling: A study of the syntax-pragmatics interface. Ph.D. Dissertation, State University of New York at Buffalo. Bowcher, W. 2017. Field, tenor and mode. In Bartlett and O’Grady (eds.). 391–403. Butler, C.S. 2003. Structure and function: A guide to three major structuralfunctional theories—Part 2 From clause to discourse and beyond. Amsterdam: Benjamins (Studies in Language Companion Series 64). Butler, C.S. 2005. Functional approaches to language. In C.S. Butler, M. de los Angeles Gomez-Gonzalez and S.M. Doval-Suarez (eds.), The dynamics of language use. Amsterdam: Benjamins. Caffarel, A., J.R. Martin and C.M.I.M. Matthiessen (eds.). 2004. Language typology: A functional perspective. Amsterdam: Benjamins (Current Issues in Linguistic Theory). Callow, K. 1974. Discourse considerations in translating the word of god. Grand Rapids: Zondervan. Doran, Y.J. and J.R. Martin. forthcoming. Field relations: Understanding scientific explanations. In K. Maton, J.R. Martin and Y. Doran (Eds.), Studying science: Language, knowledge and pedagogy. London: Routledge. Du Bois, J.W. 1980. Beyond definiteness: the trace of identity in discourse. In W.L. Chafe (ed.), The pear stories: Cognitive, cultural and linguistic aspects of narrative production. Norwood: Ablex Publishing Corporation. 203–274. FitzSimons, P. 2005. Placeintime Bondi beach. The Sydney Magazine. February. Sydney: Sydney Morning Herald. 23. Gleason, H.A. Jr. 1968. Contrastive analysis in discourse structure. In Monograph series on languages and linguistics 21. Washington, DC: Georgetown University Institute of Languages and Linguistics. 39–63. Guzmán, J.A. 2010. La ola maldita. Revista Paula. Retrieved on December 2017 from www.paula.cl/reportajes-y-entrevistas/reportajes/la-ola-maldita/. Halliday, M.A.K. 1961. Categories of the theory of grammar. Word 17.3. 241–292. Halliday, M.A.K. 1979. Modes of meaning and modes of expression: Types of grammatical structure, and their determination by different semantic functions. In D.J. Allerton, E. Carney and D. Holcroft (eds.), Function and context in linguistics analysis: Essays offered to William Haas. Cambridge: Cambridge University Press. 57–79. Halliday, M.A.K. 1985a. An introduction to functional grammar. London: Arnold. (2nd edition 1994, 3rd edition revised by C.M.I.M. Matthiessen 2004, 4th edition revised by C.M.I.M. Matthiessen 2014). Halliday, M.A.K. 1985b. Systemic background. In J.D. Benson and W.S. Greaves (Eds.), Systemic perspectives on discourse, vol. 1: Selected theoretical papers from the 9th International Systemic Workshop. Norwood: Ablex Publishing Corporation. 1–15.

Functional Language Typology

233

Halliday, M.A.K. and J.R. Martin (eds.). 1981. Readings in systemic linguistics. London: Batsford. Halliday, M.A.K. and C.M.I.M. Matthiessen. 1999. Construing experience through meaning: A language-based approach to cognition. London: Continuum. Hao, J. 2015. Construing biology: An ideational perspective. Unpublished PhD thesis, University of Sydney. Hao, J. 2018. Reconsidering “cause inside the clause” in scientific discourse— from a discourse semantic perspective in systemic functional linguistics. Text & Talk 38.5. 525–550. Hao, J. forthcoming a. Analysing scientific discourse from a systemic functional linguistic perspective: A framework for exploring knowledge building in biology. London: Routledge. Hao, J. forthcoming b. Nominalisations in scientific English: A tristratal perspective. Functions of Language. Kim, M., J. Munday, P. Wang and Z. Wang (eds.). in press. Systemic functional linguistics in translation studies. London: Bloomsbury. Longacre, R.E. 1976. An anatomy of speech notions. Lisse: Peter de Ridder. Longacre, R.E. 1983. The grammar of discourse. New York: Plenum. Martin, J.R. 1983. Participant identification in English, Tagalog and Kâte. Australian Journal of Linguistics 3.1. 45–74. Martin, J.R. 1992. English text: System and structure. Amsterdam: Benjamins. Martin, J.R. 1996. Types of structure: Deconstructing notions of constituency in clause and text. In E.H. Hovy and D.R. Scott (ed.), Computational and conversational discourse: Burning issues—an interdisciplinary account. Heidelberg: Springer (NATO Advanced Science Institute Series F—Computer and Systems Sciences, Vol. 151). 39–66. Martin, J.R. 2009. Discourse studies. In J.J. Webster (ed.), Continuum companion to Systemic Functional Linguistics. London: Continuum. 154–165. Martin, J.R. 2014. Evolving systemic functional linguistics: Beyond the clause. Functional Linguistics 1.3. Martin, J.R. 2015a. Halliday the grammarian: Axial foundations. In J.J. Webster (ed.), The Bloomsbury companion to M A K Halliday. London: Bloomsbury. 257–290. Martin, J.R. 2015b. Meaning beyond the clause: Co-textual relations. Linguistics and the Human Sciences 11.1–2. 203–235. Martin, J.R. 2015c. Cohesion and texture. In D. Schiffrin, D. Tannen and H. Hamilton (eds.), Handbook of discourse analysis (2nd edition). Chichester, UK: John Wiley & Sons, Ltd. 61–81. Martin, J.R. 2016. Meaning matters: A short history of systemic functional linguistics. Word 61.2. 35–58. Martin, J.R. 2018. (ed.) Interpersonal meaning: Systemic functional linguistics perspectives. Special Issue of Functions of Language 25.1. Martin, J.R. 2019. Discourse semantics. In G. Thompson, W. Boucher, L. Fontaine and D. Schönthal (eds.), The Cambridge handbook of systemic functional linguistics. Cambridge: Cambridge University Press. 358–381. Martin, J.R. and Y.J. Doran (eds.). 2015a. Grammatics. London: Routledge (Critical Concepts in Linguistics: Systemic Functional Linguistics, Vol. 1). Martin, J.R. and Y.J. Doran (eds.). 2015b. Grammatical descriptions. London: Routledge (Critical Concepts in Linguistics: Systemic Functional Linguistics, Vol. 2).

234

J.R. Martin and Beatriz Quiroz

Martin, J.R. and D. Rose 2007. Working with discourse: Meaning beyond the clause (2nd edition). London: Continuum. Martin, J.R. and D. Rose. 2008. Genre relations: Mapping culture. London: Equinox. Martin, J.R., B. Quiroz, P. Wang and Y. Zhu. forthcoming. Systemic functional grammar: Another step into the theory—grammatical description. Beijing: Higher Education Press. Martin, J.R., P. Wang and Y. Zhu. 2013. Systemic functional grammar: A next step into the theory—axial relations. Beijing: Higher Education Press. Martin, J.R. and P.R.R. White. 2005. The language of evaluation: Appraisal in English. London: Palgrave. Matthiessen, C.M.I.M. 2004. Descriptive motifs and generalisations. In Caffarel et al. 537–673. Matthiessen, C.M.I.M. 2014a. Halliday on language. In J.J. Webster (ed.), The Bloomsbury companion to M A K Halliday. London: Bloomsbury. 137–202. Matthiessen, C.M.I.M. 2014b. The notion of a multi-lingual meaning potential: A systemic exploration. Draft, Department of English. PolySystemic Research Group, Faculty of Humanities, Polytechnic University Hong Kong. Matthiessen, C.M.I.M. 2015. Systemic functional morphology: The lexicogrammar of the word. In E. Rosa de Souza (ed.), Estudos de Descrição Funcionalista: objetos e abordagens. LINCOM Studies in Theoretical Linguistics 55. München: LINCOM. 150–199. Matthiessen, C.M.I.M. 2018. The notion of a multi-lingual meaning potential: A systemic exploration. In A. Sellemi-Baklouti and L. Fontaine (eds.), Perspectives from systemic functional linguistics. London: Routledge. 90–120. Matthiessen, C.M.I.M., K. Teruya and C. Wu. 2008. Multilingual studies in a multi-dimensional space of interconnected language studies. In J.J. Webster (ed.), Meaning in context: Implementing intelligent applications of language studies. London: Continuum. 146–220. Moyano, E. 2016. Theme in English and Spanish: Different means of realisation for the same textual function. English Text Construction 9.1. 190–220. Mwinlaaru, I.N., C.M.I.M. Matthiessen and E. Akerajola. 2018. A system-based typology of mood in Niger-Congo languages. In A. Agwuele and A. Bodomo (eds.), The handbook of African Linguistics. London: Routledge. 93–117. Mwinlaaru, I.N. and W.W. Xuan. 2016. A survey of studies in systemic functional language description and typology. Functional Linguistics 3.8. Pericchi Paga, N., B. Cornillie, F. Van de Velde and K. Davidse (2017). La duplicación de clíticos en español como estrategia de marcación inversa. Revue Romane, 52.2. 190–206. PRESEEA-SA. (2014–2019). Corpus de Santiago de Chile del Proyecto para el Estudio Sociolingüístico del Español de España y de América. Alcalá de Henares: Universidad de Alcalá [http://preseea.linguas.net]. Consultado: [20 de diciembre de 2019]. Quiroz, B. 2013. The interpersonal and experiential grammar of Chilean Spanish: Towards a principled systemic-functional description based on axial argumentation. Unpublished PhD Dissertation, University of Sydney, Sydney, Australia. Quiroz, B. 2015. La cláusula como movimiento interactivo: Una perspectiva semántico-discursiva de la gramática interpersonal del español. Documentação de Estudos em Linguística Teórica e Aplicada 31.1. 261–301.

Functional Language Typology

235

Quiroz, B. 2017a. Gramática interpersonal básica del español: una caracterización sistémico-funcional del sistema de modo. Lenguas Modernas 49. 157–182. Quiroz, B. 2017b. The verbal group. In T. Bartlett and G. O’Grady (eds.), The Routledge handbook of systemic functional linguistics. London: Routledge. 301–318. Quiroz, B. 2018. Negotiating interpersonal meanings: Reasoning about mood. Functions of Language 25.1. 135–163. Rochester, S. and J.R. Martin. 1979. Crazy talk: A study of the discourse of schizophrenic speakers. New York: Plenum (Cognition and language: A series in psycholinguistics). Rose, D. 2001. Some variations in theme across languages. Functions of Language 8.1. 109–145. Rose, D. and J. R. Martin. 2012. Learning to write, reading to learn: Genre, knowledge and pedagogy in the Sydney School. London: Equinox. Silva-Corvalán, C. 1981. The diffusion of object-verb agreement in Spanish. Papers in Romance, 3.2. 163–176. Simon-Vandenbergen, A.-M., M. Taverniers, and L.J. Ravelli (eds.). 2003. Metaphor: Systemic and functional perspectives. Amsterdam: Benjamins. Tann, K. 2017. Context and meaning in the Sydney architecture of SFL. In T. Bartlett and G. O’Grady (eds.), The Routledge handbook of systemic functional linguistics. London: Routledge. 438–456. Teruya, K., E. Akerejola, T.H. Anderson, A. Caffarel, J. Lavid, C.M.I.M. Matthiessen, U.H. Petersen, P. Patpong and F. Smedegaard. 2007. Typology of mood: A text-based and system-based functional view. In R. Hasan, C.M.I.M. Matthiessen and J. Webster (eds.), Continuing discourse on language: A functional perspective (Volume 2). London: Equinox. 859–920. Teruya, K. and C.M.I.M. Matthiessen. 2015. Halliday in relation to language comparison and typology. In J.J. Webster (ed.), The Bloomsbury companion to M A K Halliday. London: Bloomsbury. 427–452. Vázquez Rozas, V. and M. García Salido. 2012. A discourse-based analysis of object clitic doubling in Spanish. In K. Davidse, T. Breban, L. Brems, and T. Mortelmans (Eds.), Grammaticalization and language change: New reflections. Amsterdam: John Benjamins. 269–296. Zappavigna, M. and J.R. Martin. 2018. Discourse and diversionary justice: An analysis of youth justice conferencing. London: Palgrave.

Appendix

Paradigms for personal pronouns, pronominal clitics and non-salient and salient possessive determiners in Chilean Spanish (see Tables 7.10 and 7.11).

Table 7.10 Personal Pronouns (Clause Rank) in Chilean Spanish Personal pronouns (clause rank) in Chilean Spanish number

person

singular

interactant

plural

non-interactant interactant non-interactant

non-oblique 1st 2nd 2nd (formal) 3rd 1st 2nd 3rd

oblique

yo mí tú ti usted él (masc.), ella (fem.), ello (neut.) nosotros (masc.), nosotras (fem.) ustedes ellos (masc.), ellas (fem.)

Table 7.11 Pronominal Clitics (Group Rank) in Chilean Spanish Pronominal clitics (group rank) in Chilean Spanish number

person

singular

interactant

plural

accusative 1st 2nd 2nd (formal)

non-interactant

3rd

interactant

1st 2nd

non-interactant

3rd

me te lo (masc.), la (fem.) lo (masc.), la (fem.) nos los (masc.), las (fem.) los (masc.), las (fem.)

dative

le/se1 le/se les/se les/se

Functional Language Typology

237

Non-salient possessive determiners in Chilean Spanish: singular : plural :: mi : mis :: ‘my’ tu : tus :: ‘your’ su : sus :: ‘his/her’ nuestro : nuestros (nuestra : nuestras) ‘our’ Salient possessive determiners in Chilean Spanish: masculine (singular : plural) : feminine (singular : plural) :: (mío : míos) : (mía, mías) :: ‘mine’ (tuyo, tuyos) : (tuya, tuyas) :: ‘yours’ (suyo, suyos) : (suya, suyas) :: ‘his/hers’ (nuestro, nuestros) : (nuestra, nuestras) ‘ours’

Note 1. Second and third person se is the dative pronominal clitic used before an accusative pronominal clitic: e.g. se lo dio (‘s/he gave it to him/her’); this paradigm does not include the neuter singular clitic lo.

8

Construing Relations Between Scientific Activities Through Mandarin Chinese Jing Hao

1. Introduction Description of the language used in specialized and disciplinary contexts has been developed over several decades for English, but less so for other languages. Understanding the use of Mandarin Chinese (hereafter Chinese) for academic and specialized purposes has become increasingly important given the growing number of learners of Chinese as a second and foreign language for the purposes of business, tourism, law, medicine and diplomacy. Many researchers and language teachers have drawn attention to the teaching of Chinese for Specific Purposes (CSP) (e.g. Q. Li 2011; Zhou 2015), picking up on ideas developed in English for Specific Purposes (ESP) theory and practice. As for ESP, a primary concern in CSP is the development of pedagogical materials which can be used to effectively teach the language of specialized knowledge. In order to do so, we need to undertake foundational work in linguistics to describe the language of specialized knowledge (Q. Li 2011). This study contributes to this linguistic endeavor, focusing on describing the language used for scientific communication, specifically for explaining activities in biological science. Systemic Functional Linguistics (SFL) provides a useful linguistic theory for describing the language of specialized knowledge, as it offers a rigorous theoretical account of the inherent relationship between knowledge and language. In this model, knowledge is conceptualized largely with respect to what are known as register variables (Martin 2017). As far as scientific knowledge is concerned, of central relevance is the register variable field, which construed through ideational linguistic resources (Halliday and Martin 1993; Hao 2015).1 For Chinese, ideational meanings have been described both at the clause rank (Halliday and McDonald 2004) and group rank (E. Li 2017; Fang 2015). These grammatical overviews are based by and large on the language used in everyday and common-sense discourse. Regarding scientific language, discussion has been largely focused on nominal groups and grammatical metaphors (Halliday 2005 [1984], 2005 [1993]; Yang 2015). These studies have

Construing Relations

239

provided foundational work on Chinese grammatical patterns. Much work, however, is needed to make the connexion between grammatical choices and a diverse range of contextual purposes, including disciplinespecific contexts. In order to make this connexion, this study takes a ‘top-down’ approach—it takes choices made in genre and register (Martin and Rose 2008) as a point of departure and examines how contextual choices are realized by choices of discourse patterns, which are realized in turn by grammatical patterns. The chapter examines in explanation genres in particular, to see how logical relationships between scientific activities are developed through language. The logical relationship in focus here is referred to in SFL as implication relations (Doran and Martin 2019; c.f. Martin 1992). At the level of language, both discourse patterns in a text and grammatical patterns in a clause are considered. Data construing scientific activities is taken from a textbook used for teaching undergraduate biology at a leading University in Mainland China—细胞生物学 (Cell Biology) (Wang 2003). The chapter is organized as follows. Section 2 introduces the rationale for a ‘top-down’ approach underpinned by SFL theory; it then reviews a recent SFL description of scientific activities from the perspective of field. Section 3 focuses on the discourse semantic realizations of implication relations, including temporal and causal connexions. Section 4 identifies relevant grammatical realizations, including downranking and complexing of clauses, verbal aspect and modal verbs. Section 5 concludes the chapter by discussing the theoretical and methodological implications of a top-down approach in systemic functional language description.

2. Theoretical Foundation 2.1 A ‘Top-Down’ Approach SFL models the use of language in relation to its context. The most abstract level, context, considers the social purposes of our communication in a culture. Martin and Rose (2008) describes genre as a system of semiotic choices, realized as staged goal-oriented social processes. Genres are realized through register variables, including our experience of ‘what is going on’ (field), our social values and relationships (tenor) and our channels of communication (mode). As far as scientific knowledge is concerned, of particular relevance is the register variable field (Martin 2017). Typical genres associated with scientific field include report, explanation and procedural recount. Choices of genre and register are realized by choices in language (along with other semiotic resources such as images and body language). The most directly relevant language choices are those organized in text-wide

240

Jing Hao

patterns at the level of discourse semantics (Martin 1983, 1992). Discourse semantic choices are further realized through choices in lexicogrammar (Halliday and Matthiessen 2014; Martin 2013), realized in turn through phonology, graphology and sign. Alongside stratification, language is also organized metafunctionally, as kinds of meaning—ideational meaning for construing field, interpersonal meaning for enacting tenor and textual meaning for composing mode. Field activity, including scientific activity, accordingly, implicates ideational choices in language. Ideational meaning is further specified as experiential meaning construing goings-on and logical meaning connecting goings-on as they unfold in discourse. A further dimension to consider involves choices organized by rank. For instance, at the grammatical stratum clauses are constituted by groups and phrases, which are in turn constituted by words. Drawing on this multi-dimensional perspective on the organization of language, SFL informed language descriptions adopt a threefold ‘trinocular’ perspective. That is to say, choices at any given level (stratum or rank) are not considered in isolation but are based simultaneously on related choices at other levels (strata and ranks), including a higher level (a perspective from ‘above’), a lower level (a perspective from ‘below’) and the given level (a perspective from ‘around’) (Halliday and Matthiessen 2014: 35; Matthiessen and Halliday 2009). To describe Chinese in scientific texts, this study takes the perspective from ‘above’—specifically description of the register variable field as a first step to begin to explore choices at other levels. This ‘top-down’ perspective is adopted for two reasons. First, it has been suggested that across languages choices are more comparable and generalizable at higher level strata (including genre, register and discourse semantics) than at the grammatical level (Martin 1983, Martin and Quiroz, this volume; Rose 2001, this volume). A field perspective enables the study to draw on a well-developed description of scientific fields as a point of departure for describing Chinese systems and structures. Second, although many SFL studies have provided grammatical descriptions of Chinese, these descriptions are less suitable as a starting point. For one thing, the available descriptions are concerned largely with grammatical patterns in common-sense fields and in spoken discourse. Only limited register-specific language patterns have been revealed. A further important issue is that SFL grammatical descriptions of languages other than English, including Chinese, have often ‘borrowed’ descriptive categories from English (following one or another edition of Halliday’s Introduction to Functional Grammar, e.g. Halliday and Matthiessen 2014) without providing language-specific reasoning (see Quiroz 2013, 2018, this volume). Looking from ‘above’ allows us to critically review the appliability of the available grammatical descriptions and thus provides a productive complementary view to previous predominant ‘bottom-up’ considerations.

Construing Relations

241

I now introduce features of scientific field revealed through English texts (Doran and Martin 2019).

2.2 Field In its broadest terms, a field can be seen from two simultaneous perspectives— a static perspective and a dynamic perspective (Doran and Martin 2019; c.f. Martin 1992). Seeing field statically involves mapping taxonomic relations among items and potentially their properties (such as recognizing different types of cells based on size and structures). Seeing field dynamically involves identifying the unfolding of activities “orienting to some global everyday, professional or institutional purposes.” The dynamic perspective allows us to make a distinction between activities which are unmomented and activities that are momented. Unmomented activities allow the activity to be encapsulated as a single, undivided whole. In contrast, momented activities are an activity series that specifies what is going on moment by moment. Using an example from Barthes’ (1977), the unmomented activity ‘having a drink,’ can be momented into an activity series as order a drink, obtain it, drink it, pay for it. This study adopts a dynamic perspective to understand the unfolding of field activities through momenting. A second important distinction with respect to activities is made between expectancy activities, within which (if momented) one activity expects another, and implication activities, within which (if momented) one activity absolutely leads to another. Activities momented through expectancy relations afford the risk of being interrupted. For example, in the activity momented as order a drink, obtain it, drink it, pay for it in an everyday context, there can be interruptions—e.g. the bar tender may be distracted and forget to bring the drink (this of course creates the possibility of complications in story genres). In contrast, activities momented through implication relations are highly determined and run a minimal risk of being interrupted. Examples of implication activities from science are the water cycle and life cycle of a cell. Implication activities in biology are typically co-textualized as explanation genres that make explicit how and why a scientific activity takes place. The focus of this chapter is on the language realizing implication relations in Chinese.

3. Identifying Discourse Semantic Resources 3.1 Identifying Instances of ‘Implication’ The data used for exploring implication activities is selected from a chapter in the textbook Cell Biology (Wang 2003). The purpose of the chapter, “Endomembrane system, protein sorting and membrane traffic,” is to

242

Jing Hao

explain activities in a cell. I begin with a brief account of how instances of implication activities are identified in the textbook data. When identifying implication activities in English, Hao (2015, in press) found that a momented implication activity can be realized in discourse semantics through a series of occurrence figures, which are related through logical connexions in a sequence.2 An occurrence figure is configured by an occurrence and at least one entity that construes a scientific item in the field. An occurrence is typically realized in the lexicogrammar through a Process realized by a verbal group, and an entity is typically realized through a Participant realized by a nominal group. The following is an example of an occurrence figure in Chinese. Annotations are provided for meanings at three strata (field, discourse semantics and lexicogrammar).3 (1) field:

activity

discourse semantics: entity

lexicogrammar:

occurrence

entity

荚膜

保护 着

细胞

jiamo

baohu zhe

xibao

capsule

protect ASP*:durative cell

Participant Process

Participant

“capsules protect the cells” *Key: ASP is short for aspect, a grammatical realization of ‘time’. The aspect here is a durative aspect, construing the occurrence as ongoing (see Halliday and McDonald 2004: 353).

To construe activities in an activity series, an occurrence figure needs to be related to other occurrence figure(s) in a sequence through a connexion (Martin 1992). Example (2) presents three occurrence figures connected into a sequence. Notice that several elements (in bold) fall outside the configuration of figures, and indicate connexions between the figures. (2) figure a con . . .

entity

occurrence

当

神经冲动

刺激

肌肉细胞

时

dang

shenjing-chongdong ciji

jirouxibao

shi

at

nerve impulse

muscle cells

time

stimulates

entity

…nexion

“(at the time) when nerve impulse stimulates muscle cells” figure b entity

connexion

occurrence

entity

钙离子

就

穿过

内质网膜

gailizi

jiu

chuanguo

neizhiwangmo

calcium

then

goes through

endoplasmic reticulum membrane

“calcium goes through the endoplasmic reticulum membrane”

Construing Relations figure c

connexion

occurrence

entity

然后

回到

细胞溶胶 中

er

huidao

xibaorongjiao zhong

and then

come back to

in the cytosol

243

“and comes back in the cytosol”

In addition to explicit connexions as shown in (2), connexions may be realized implicitly in the discourse (these are grammatically ‘unmarked’ by conjunctions and will be discussed further in Section 4.2). When connexions are realized implicitly, it is necessary to make them explicit for the purpose of data analysis. In example (3), the first three connexions are implicit; one way in which they could have been made explicit is shown in parentheses. (3) figure a connexions

entity

occurrence

(首先)

蛋白质转运通道

打开，

shouxian

danbaizhidakai zhuanyuntongdao

first

protein transport channel

open

“first, protein transport channel is opened,” figure b connexions

entity

occurrence

(然后)

核糖体与通道

结合，

ranhou

hetangti yu tongdao

jiehe

and then

ribosome and the channel

unify

“and then, ribosome and the channel unity,” figure c

(connexions) entity

occurrence

entity

(接下来)

新生的肽

插进

通道。

(jiexialai)

xinsheng de tai

chajin

tongdao

(and then)

newly formed peptide

entre into

channel

“(and then) new peptide entre into the channel.” figure d connexion

entity

occurrence

entity

于是

释放的SRP

又回到

胞质溶胶中。

yushi

shifang de SRP

you huidao baozhirongjiao zhong

so

released SRP

return to

“So the released SRP return to cytosol.”

inside the cytosol

244

Jing Hao

As the examples show, connexions can be realized between clause complexes (i.e. orthographic sentences) or within them (between interdependent clauses). This study focuses on congruent mapping of discourse semantic sequences in lexicogrammar. That is to say, grammatical metaphor, i.e. the re-mapping of a sequence onto a clause (Martin 2008), is set aside in this study (see Hao 2015, 2018 for discussion). An example of mapping sequence metaphorically onto a clause, within which both figures are mapped as Participants realized by nominal groups is shown in (4). Its congruent realization is provided for comparison. (4) figure a metaphorical Bip的功能丧失 realization Bip de gongneng sangshi

connexion

figure b

导致

蛋白质在ER中的聚集

daozhi

danbaizhi zai ER zhong de juji

loss of the function of Bip lead to

aggregation of proteins in ER

Participant

Participant

Process

“The loss of the function of Bip leads to the aggregation of proteins in ER.” congruent realization

Bip 丧失了功能

于是

蛋白质在ER中聚集

Bip sangshi le gongneng

yushi

danbaizhi zai ER zhong juji

Bip loses its function

so that

proteins aggregate in ER

clause

conjunction clause

“Bip loses its function, so that proteins aggregate in ER.”

This paper addresses the 189 instances of congruently realized momented implication activities which were identified in the textbook chapter.

3.2 Types of Connexion In her study of implication activities in English, Hao (in press) notes that both temporal and causal connexions can be used to realize implication relations in scientific fields. Similar features were also found in Chinese. With temporal sequencing, the figure realizing the activity happening first is annotated as the Anterior figure, and the figure realizing the activity that follows is annotated as Posterior. As the activity series develops,

Construing Relations

245

each Posterior can in turn becomes an Anterior. In (5), two figures are connected with a temporal connexion ranhou (and then). (5) Anterior 要进入

内质网的腔,

danbaizhi shouxian

首先

yao jinru

neizhiwang de qiang

protein

should enter into lumen of endoplasmic reticulum

蛋白质

first

“the proteins should first enter into endoplasmic reticulum lumen,” Posterior 然后

通过选择性的分泌过程

输出到

细胞外

ranhou

tongguo xuanzexing de fenmiguocheng

shuchu dao

xibaowai

and then

through selective secretion export to

outside the cells

“and then the proteins are transported out from the cells through selective secretion.”

The same activity series can also be realized through a causal sequence. If so, the connexion pre-selects a different set of grammatical realizations. Example (6) exemplifies a causal sequence comparable to the temporal one in (5). In the causal sequence, a different set of conjunctions ruguo  .  .  ., jiu .  .  . (‘if .  .  ., then .  .  .’) is used, and a modal verb yao (‘should’) used initially in the Anterior figure (5) needs to be used again in the Posterior figure (6) (i.e. hui (will) as underlined). In the causal sequence, the Anterior and Posterior function at the same time as a Cause and an Effect. (6) Anterior/Cause 如果

进入

内质网的腔,

danbaizhi

ruguo

jinru

neizhowang de qiang

protein

if

enter into

lumen of endoplasmic reticulum

蛋白质

“if the proteins enter into endoplasmic reticulum lumen,” Posterior/Effect 就

会

通过选择性的分泌过程

输出到

细胞外

jiu

hui

tongguo xuanzexing de fenmiguocheng

shuchu dao

xibaowai

then

will through selective secretion export to

outside the cells

“then the proteins will be transported out from the cells through selective secretion.”

More delicate choices of temporal and causal connexions can be made. In terms of temporal connexions, a further distinction can be made between

246

Jing Hao

successive and overlapping ones. When figures are connected successively, one figure follows another temporally in a sense that the Anterior figure realizes an activity that “does not continue beyond the beginning of the following activity” (Martin 1992: 185). This is exemplified by (7) which uses a temporal connexion shouxian . . . ranhou (‘first . . . and then’). (7) Anterior 蛋白质

首先

要进入

内质网的腔,

danbaizhi

shouxian

yao jinru

neizhowang de qiang

protein

first

must enter into

lumen of endoplasmic reticulum

“the proteins must first enter into endoplasmic reticulum lumen,” Posterior 然后

通过选择性的分泌过程

输出到

细胞外

ranhou

tongguo xuanzexing de fenmiguocheng

shuchu dao

xibaowai

and then

through selective secretion

export to

outside the cells

“and then the proteins are transported out from the cells through selective secretion.”

In contrast to successive connexions, Anteriors and Posteriors can be connected as overlapping—that is, the activity realized by Posterior starts before the completion of that realized by Anterior. This is exemplified in (8), where the connexion is indicated by dang . . . shi, . . . jiu . . . (‘at the time (when) . . . then . . .’). (8) Anterior 肌肉细胞

时，

dang shenjingzhongdong ciji

当

神经冲动

刺激

jirouxibao

shi,

at

muscle cells time,

nerve impulse

stimulates

“(at the time) when nerve impulse stimulates muscle cells,” Posterior 钙离子

就

穿过

内质网膜

gailizi

jiu

chuanguo

neizhiwangmo

calcium then

goes through endoplasmic reticulum membrane

“calcium goes through the endoplasmic reticulum membrane”

Turning to causal connexions, three delicate types were identified in the data, consequence, condition and purpose. Consequential

Construing Relations

247

connexions establish a realis relation between a Cause and a definite Effect, whereas conditional and purposive connexions construe an irrealis relation involving contingencies and hypotheticalities (Martin 1992: 193). The sequence in (9) exemplifies a realis consequential relationship between a Cause and an Effect. The connexion is indicated by yushi (‘so that’). (9) Anterior/Cause 当

这种信号

到达

肌肉细胞

时，

dang

zhezhong xinhao

daoda

jirouxibao

shi

at

this kind of signal

reach

muscle cells

time

“(at the time) when this kind of signal reaches the muscle cells,” 为

该细胞质膜中的受体蛋白

所识别,

wei

gai xibaozhimo zhong de shoutidanbai

suo shibie

by

the receptor proteins in the cytoplasm membrane’ is recognized

“the signal is recognized by the receptor proteins in the cytoplasm membrane” Posterior/Effect 于是

引起了

细胞中的信号转导途径

发生作用。

yushi

yinqile

xibaozhong de xinhaozhuandaotujing

fasheng zuoyong

the signal transfer path inside the cells

take effect

so that cause

“So the signal causes signal transfer path inside the cells to take effect.”

In contrast, when connected conditionally, both Cause and Effect are irrealis, as exemplified in (10). (10) Anterior/Cause (irrealis) 如果

ER 滞留蛋白质

被

错误地

ruguo

ER zhiliudanbaizhi

bei cuowude

if

ER retention protein by

mistakenly

包进

分泌泡

baojin

fenmipao

is embraced secretory vacuole

“if ER retention proteins are mistakenly embraced into the secretory vacuole,” 而

离开了

ER

er

likai le

ER

and

leave

ER

“and leave the ER” Posterior/Effect (irrealis)

248

Jing Hao

高尔基复合体膜上的这种 信号受体蛋白 gaoerjifuhetimo shang de zhezhong xinhaoshoutidanbai this kind of signal receptor proteins on Golgi complex membrane

就

会

与ER蛋白

结合

jiu

hui

yu ER danbai

jiehe

then

will

with ER protein

unite

“this kind of signal receptor proteins on Golgi complex membrane will then unite with ER.”

When figures are connected through a purposive connexion, a realis Cause is connected to an irrealis Effect such as in (11). (11) Posterior/Effect (irrealis) 为了

能

准确地

weile

neng

zhunque de

in order to

be able to accurately

运送

蛋白质

yunsong

danbaizhi

transport

protein

“in order to be able to accurately transport protein” Anterior/Cause (realis) 在进化过程中

每种蛋白质

形成了

zai jinghua guocheng zhong

mei zhong danbaizhi

xingcheng le yige mingque de dizhiqian

一个明确的地址签

in the course of evolution

each type of protein

form

a clear address

“in the course of evolution, each type of protein forms a clear address.”

The system network in Figure 8.1 outlines the different sequencing options in the data.

Figure 8.1 CONNEXION System

Construing Relations

249

In the next section I take a further step in order to examine how connexions are realized grammatically.

4. Grammatical Realizations As noted earlier, connexions in the discourse can be congruently realized either within clause complexes (e.g. (2)) or across clause complex boundaries (e.g. (3)). Realizations of connexion between clause complexes have been treated as involving ‘non-structural’ cohesive conjunctions (Halliday and Hasan 1976; Halliday and Matthiessen 2014). This study limits its scope to a consideration of clauses in clause complexes, and so the connexions realized across clause complexes will be set aside. In identifying the relationship between clauses, it is necessary to first label all the clauses in the analysis according to their sequence in a text. In the following discussion, ‘initial’ is used to label the clause that comes first, and ‘continuing’ labels the clauses which follow.4 In previous SFL informed descriptions of Chinese (e.g. Halliday and McDonald 2004; E. Li 2007; Hsu 2017; Zhang 2012), the term taxis is used to describe the ‘degree of interdependency’ between clauses (Halliday and Matthiessen 2014: 440).5 In the system of taxis, a distinction is made between parataxis and hypotaxis. With parataxis, clauses are in ‘equal’ relationship with one another—annotated with the Arabic numbers based on their sequence in the text (1, 2, 3 etc.). With hypotaxis, clauses are in ‘unequal’ relationship with one another—annotated by the Greek alphabet (α, β, γ, etc.). The clauses are unequal in a sense that β clause is dependent on α, and γ is dependent on β, etc. In addition to taxis, however, descriptions of English and other languages (Anderson and Holsting 2018; Huddleston, Hudson, Winter and Henric1968; Martin 1988) have had to deal with criteria for distinguishing a dependent β clause from an embedded clause. The principal difference between a β clause and an embedded clause is that a β clause is a ranking clause that does not form part of a larger unit, whereas an embedded clause is downranked (to the rank of group/phrase), part of a ranking clause. Two different structures are in play here—i.e. a univariate interdependency structure realizing logical meaning and a multivariate constituency structure realizing experiential meaning. In what follows I begin with a discussion of the borderline instances of embedding and complexing. In the data the problematic instances are mainly realizations of temporal connexions. 4.1 Embedding vs. Complexing The embedded clauses in focus on here are typically embedded in a Circumstance as far as experiential meaning is concerned (Halliday and Matthiessen 2014). A typical realization of a Circumstance is coverbal

250

Jing Hao

phrase. A Chinese coverbal phrase is a combination of a coverb and a nominal group (Halliday and McDonald 2004: 317).6 Embedded clauses take the place of the nominal group in a coverbal phrase. I will first illustrate the typical structure of a coverbal phrase, and then coverbal phrases with embedded clauses realizing connexion. The embedded clauses will then be distinguished from hypotactic clauses. After establishing their differences, I will introduce borderline examples having both characteristics of embedding and complexing. A typical coverbal phrase (not involving an embedded clause) is illustrated in (12). The coverb zai (analogous to English ‘(being) at’ or ‘during’) precedes a nominal group indicating time—ci qijian, ‘this time period’. (12) 在

[此

期间],

活细胞

会摄取

具有放射性的氨基酸

dang

[ci

qijian],

huoxibao

hui shequ

juyou fangshexing de anjisuan

at/ during

[this

time],

living cells

will absorb radioactive amino acids

Circumstance

Participant Process

Participant

coverbal phrase

nominal group

nominal group

verbal group

Minor Minor Range Process coverb

[nominal group] Deictic Thing determiner

noun

“During this time, living cells will take up radioactive amino acids.”

As we can see in (12) a coverbal phrase is situated one rank below the clause and at the same rank as a nominal or verbal group. However, while a group can be seen logically as a univariate structure (i.e. an interdependency structure in which one word is dependent on another), a coverbal phrase is constituted experientially as a multivariate structure— like the structure of a clause (a ‘shrunken clause’ for Halliday and Matthiessen 2014: 329). The coverb can thus be seen as a Minor Process and the nominal group as a Minor Range (following the terminology used for the structure of English prepositional phrases). Given that realizations of a group/phrase are at the word level (e.g. coverb), but the nominal group realizing the Minor Range is itself a unit at the group rank, we treat the nominal group as a downranked group—that is, an embedded group in the coverbal phrase. Embedded groups are annotated with a single-squared bracket. The downranked nominal group in example (12) is constituted by a Deictic and a Thing.

Construing Relations

251

A coverbal phrase involving an embedded clause can be used to realize a temporal sequence. In doing so, one figure is realized by the ranking clause, and the other figure is realized inside the coverbal phrase, specifically through a clause that is embedded inside the embedded nominal group (the Minor Range), as exemplified in (13) and (14). The embedded clauses are annotated with double-squared brackets. (13) 某

一

适当

时机], 蛋白质转 运通道

向两侧

打开

[ [[zhuanyun]] de

mou

yi

shidang

shiji],

danbaizhi zhuanyun tongdao

xiang liangce

dakai

[ [[transport] ]

certain one

suitable

time

protein transport channel

to both sides

open

在

[ [[转运]]

zai

at

的

Circumstance

Participant Circum- Process stance

coverbal phrase

nominal group

coverbal verbal phrase group

Minor Minor Range Process coverb

[nominal group] Qualifier

Linker Deictic Measure Epithet

[[embedded clause]]

link

determiner

numerative

Thing

adjective noun

“At [a suitable time for [[transporting (protein)]] ], the protein transport channel opens to both sides.”

(14) 当

[ [[蛋白质由核糖 体合成]]

之

后]

必须

准确无误地

运送到

细胞的 各个部位

dang

[ [[danbaizhi you hetangti hecheng]]

zhi

hou]

bixu

zhunquewuwu de

yunsong dao

xibao de gege buwei

[at/ during

[the time after [[that proteins are synthesized by ribosomes]] ]

must

accurately

transport to

all parts of the cell

Circumstance

Pro . . .

Circumstance

. . . cess

Participant

coverbal phrase

verbal . . .

adverbial group

. . . group

nominal group

Minor Process

Minor Range

coverb

[nominal group]

252

Jing Hao Qualifier

Linker

Thing

[[embedded clause]]

link

noun

“After [the time [[that proteins are synthesized by ribosomes]] ], the proteins must be transported accurately to all parts of the cell.”’

As realizations of temporal sequence, the Things of the embedded nominal groups refer to time—shiji, ‘timing/time’ in (13) and hou, ‘ensuing time’ in (14). In (13) the Thing is modified by an Epithet shidang (‘appropriate’), a Measurer yi (‘one’), a Deictic mou (‘some/certain’), and a Qualifier zhuanyun (‘transport’). In (14) the Thing is modified only by a Qualifier danbaizhi you hetangti hecheng (‘proteins are synthesized by ribosomes’). Both Qualifiers are realized by embedded clauses, which are downranked from the clause rank to the word rank (Fang 2015; E. Li 2017). One distinguishing feature of examples (12), (13) and (14) is that the coverbal phrases at the beginning of the clause can all be relocated after the first Participant of the ranking clause, as exemplified in (13b).7 (13b) 蛋白质转 运通道

在

[ [[转运]]

danbaizhi zhuanyun tongdao

zai

[ [[zhuanyun]] de

protein transport channel

at

[a suitable time for [[transporting]] ],

的

某

一

适当

时机], 向两侧

打开

mou

yi

shidang

shiji], xiang liangce

dakai

to both sides

open

Participant Circumstance

Circum- Process stance

nominal group

coverbal verbal phrase group

coverbal phrase Minor Minor Range Process coverb

[nominal group] Qualifier

Linker Deictic Measure Epithet

[[embedded clause]]

link

determiner

numerative

Thing

adjective noun

“At [a suitable time for [[transporting (protein)]] ], the protein transport channel opens to both sides.”’

This provides one criterion for distinguishing a coverbal phrase (with an embedded clause as part of the coverbal phrase) from α β clause in a hypotactic clause complex. As shown in the pair of examples that follow, the hypotactic β clause in (15a) is less comfortably relocated after the first Participant of the α clause (i.e. xinshengtai de hecheng, ‘the synthesis of

Construing Relations

253

nascent peptide’) as in (15b). It could create ambiguity as to whether it is the ‘synthesis of the nascent peptide’ or ‘SRP’ that terminates ‘translation of proteins,’ and even whether it is the ‘translation of proteins’ or the ‘blocking’ that is being terminated. No structures like (15b) were found in the data. (15a) β

因为

SRP

终止了

yinwei

SRP

zhongzhi le danbaizhi de fanxi

because

SRP

terminate

Participant Process

蛋白质的翻译, translation of proteins Participant

“Since SRP has terminated the translation of the protein,” α

新生肽的合成

被阻断

xinshengtai de hecheng

bei zuduan

the synthesis of nascent peptide is blocked Participant

Process

“the synthesis of the nascent peptide is blocked.”

*(15b) 新生肽的合成

因为

SRP

xinshengtai de hecheng

yinwei

SRP zhongzhi le danbaizhi de fanxi

终止了

the synthesis of the because SRP terminate nascent peptide

蛋白质的翻译 被阻断

translation of proteins

bei zuduan is blocked

“Since SRP has terminated the translation of the protein, the synthesis of the nascent peptide is blocked.”

Embedded clauses inside a coverbal phrase can also be distinguished from a hypotactic clause by their limited interpersonal potential (Anderson and Holsting 2018; Huddleston et al. 1968). This is reflected in science data in that if an activity series, such as in (13) and (14), does not take place as expected, the negation is realized in the ranking clause—by choosing negative polarity in the verbal group (in bold), as shown in (13c) and (14b). The choices of polarity in the embedded clause are not affected. (13c) [在

[[转运]] 的某一 适当时机],

蛋白质转运通道

[zai

[[zhuanyun]] de mou yi shidang shiji],

danbaizhi meiyou zhuanyuntongdao

没有

向两侧

打开

xiang liangce dakai

254

Jing Hao

[at

a suitable time for protein transport [[transporting]] ], channel

not

to both sides

open up

Circumstance

. . . cess

Circumstance

Participant

Pro . . .

coverbal phrase

nominal group

verbal . . . coverbal phrase

. . . group

neg coverb nominal group “*[At a suitable time for [[transporting (protein)]], the protein transport channel did not open to both sides.”

(14b) 当

[ [[蛋白质由核糖 之 体合成]]

后]

没有

准确无误地

运送到

细胞的 各个部位

dang

[ [[danbaizhi you zhi hetangti hecheng]]

hou]

meiyou

zhunquewuwu de yunsong dao

xibao de gege buwei

[at/ during

[the time after [[that proteins are synthesized by ribosomes]] ]

not

accurately

transport to

all parts of the cell

Circumstance

Pro . . .

Circumstance

. . . cess

Participant

coverbal phrase

verbal . . . adverbial group

. . . group nominal group

neg Minor Minor Range Process coverb

[nominal group] Qualifier

Linker Thing

[[embedded clause]]

link

noun

“After [the time [[that proteins were synthesized by ribosomes]] ], the proteins did not transported accurately to all parts of the cell.”

In contrast, in a hypotactic clause complex, both α and β clauses have more of the interpersonal meaning potential of a ranking clause, and both can select polarity—as shown in (15c). (15c) β 因为 yinwei

SRP

没有

终止

蛋白质的翻译,

SRP

meiyou

zhongzhi

danbaizhi de fanxi

because SRP does not terminate translation of proteins Participant Process Participant nominal group

verbal group

nominal group

neg “Since SRP did not terminate the translation of the protein,”

Construing Relations α 新生肽的合成

没有

被阻断

xinshengtai de hecheng

meiyou

bei zuduan

the synthesis of the nascent peptide

is not

blocked

Participant

Process

nominal group

verbal group

255

neg “the synthesis of the nascent peptide was not blocked.”

The two prior criteria were used in the analysis to distinguish realizations of figures by an embedded clause from those by a β clause. However, there are borderline examples having both characteristics of embedding and complexing. Example (16) realizes a temporal sequence, relating an Anterior figure—shenjingchongdong ciji jirouxibao (“nerve impulse stimulates muscle cells”) and a Posterior figure—gailizi chuanguo neizhiwangmo (“calcium goes through the endoplasmic reticulum membrane”). We can identify the realization of the Anterior figure as an embedded clause forming part of a coverbal phrase, since there is a coverb dang (‘at/during’) and nominal group with a Thing shi, ‘time’ (shihou in its full lexical realization). The embedded clause functions as a Qualifier modifying the Thing. (16) 当

[ [[神经冲动 刺激肌肉细胞]]

dang at/ during

时]，

钙离子

穿过

内质网膜

[ [[shenjingchongdong shi] ciji jirouxibao]]

gailizi

chuanguo neizhiwangmo

[the time [[when nerve impulse stimulates muscle cells]] ],

calcium

goes through

endoplasmic reticulum membrane

Circumstance

Participant Process

Participant

coverbal phrase

nominal group

nominal group

verbal group

Minor Minor Range Process coverb

[nominal group] Qualifier

Thing

[[embedded clause]]

noun

“At [the time [[when nerve impulse stimulates muscle cells]] ], calcium goes through the endoplasmic reticulum membrane.”

Similar to (12), (13) and (14), the coverbal phrase in (16) can be relocated after the first Participant of the ranking clause, as in (16b).

256

Jing Hao

(16b) 钙离子

当

[ [[神经冲动 刺激肌肉细胞]]

时]，

穿过

内质网膜

gailizi

dang

[ [[shenjingchongdong ciji jirouxibao]]

shi]

chuanguo

neizhiwangmo

calcium

at/ during

[the time [[when nerve impulse goes through stimulates muscle cells]] ],

Participant

Circumstance

Process

Participant

nominal group

coverbal phrase

verbal group

nominal group

Minor Process

Minor Range

coverb

[nominal group] Qualifier

Thing

[[embedded clause]]

noun

endoplasmic reticulum membrane

“At [the time [[when nerve impulse stimulates muscle cells]] ], calcium goes through the endoplasmic reticulum membrane.”

At the same time the coverbal phrase in (16) differs from a typical coverbal phrase in several ways. First, the noun realizing the Thing shi (‘time’) commits very little meaning in the nominal group and can in fact be omitted—as shown in (16c). There the realization of the Minor Range is realized directly through an embedded clause. The noun shi is an example of what is referred to as a ‘light noun’ in the literature—a kind of ‘nominalizer’ for Yap and Grunow-Hårsta (2010); see also Chao (1968: 119). (16c) 当

[ [[神经冲动 刺激肌肉细胞]] ]

钙离子

穿过

内质网膜

dang

[ [[shenjingchongdong ciji jirouxibao]] ]

gailizi

chuanguo

neizhiwangmo

at/ during

[(the time when) [[nerve calcium goes through impulse stimulates muscle cells]] ], Circumstance Participant Process

endoplasmic reticulum membrane Participant

coverbal phrase

nominal group

Minor Process

Minor Range

coverb

[[embedded clause]]

nominal group

verbal group

“At [the time [[when nerve impulse stimulates muscle cells]] ], calcium goes through the endoplasmic reticulum membrane.”

Construing Relations

257

A second untypical feature of (16) is that the coverb dang (‘at/during’) in (16) can also be elided, as shown in (16d). The Circumstance is now realized directly through a nominal group, rather than a coverbal phrase. (16d) [ [[神经冲动 刺激肌肉细胞]] 时]，

钙离子

穿过

内质网膜

[ [[shenjingchongdong ciji jirouxibao]]

gailizi

chuanguo

neizhiwangmo

calcium

goes through endoplasmic reticulum membrane

Circumstance

Participant

Process

nominal group

nominal group verbal group nominal group

shi]

[the time [[when nerve impulse stimulates muscle cells]] ],

Qualifier

Thing

[[embedded clause]]

noun

Participant

“At [the time [[when nerve impulse stimulates muscle cells]] ], calcium goes through the endoplasmic reticulum membrane.”

The lightness of the ‘nominalizer’ and the potential omission of the coverb make the embedded clause less ‘bound’ as an embedded clause. In addition, like a ranking clause, it can also realize polarity, as shown in (16e). (16e) 神经冲动不刺激肌肉细 钙离子 不穿过 内质网膜 胞 时， dang shenjingchongdong bu gailizi bu chuanguo neizhiwangmo ciji jirouxibao shi at/ the time when nerve calcium (does) not go through endoplasmic during impulse does not reticulum membrane stimulate muscle cells, “[At the time [[when nerve impulse does not stimulate muscle cells]] ], calcium does not go through the endoplasmic reticulum membrane.” 当

Example (16) accordingly has characteristics of both a ranking clause involving an embedded clause and a hypotactic clause complex. Its structure can be seen from two complementary perspectives—i.e. experientially in terms of its constituency and logically in terms of interdependency (Martin and Matthiessen 1991: 26; see also Martin 1995: 199–204 for both perspectives on Tagalog clause projection). The analyses from both perspectives are illustrated in (16f). From a logical perspective, optional marking through a conjunction (typically jiu (‘then’)) can be deployed in the α clause.

258

Jing Hao

(16f) 当

神经冲动

刺激

dang shenjing ciji chongdong at

experiential

nerve impulse

肌肉细胞

时， 钙离子 (就)

穿过

jirouxibao

shi,

chuanguo neizhiwangmo

gailizi

(jiu)

stimu- muscle cells time, calcium (then) goes lates through

Circumstance

内质网膜

endoplasmic reticulum membrane

Participant

Process

Participant

n.g.

v. g

n.g.

coverbal phrase Minor Process

Minor Range

coverb

[nominal group] Qualifier

Thing

[[embedded clause]]

noun

logical

(conj) β

α

“(at the time) when nerve impulse stimulates muscle cells, calcium goes through the endoplasmic reticulum membrane”

To summarize, a temporal sequence can be realized through a clause with a coverbal phrase, when a figure is realized through an embedded clause inside the coverbal phrase. And a temporal sequence can be realized in a way that is best characterized by both an embedding and a complexing perspective. We now take a further step in our consideration of clause complexing, to distinguish between parataxis and hypotaxis. 4.2 Clause Complexing: Parataxis vs. Hypotaxis In Chinese, parataxis and hypotaxis can be distinguished with respect to the way they are marked by conjunctions. Conjunctions typically appear either at the beginning of a clause, or before its first Participant. Three patterns of marking are available, which I refer to as zero marking, relative marking and co-relative marking. Zero marking means that no conjunctions are involved; relative marking means that there is a conjunction in one of the clauses; and co-relative marking means that there is a conjunction in both related clauses. Parataxis and hypotaxis differ on whether they need to be marked with a conjunction or not. Parataxis allows for zero marking, as can be seen in (17). In this instance, there is a temporal connexion in the discourse semantics, but grammatically neither clause is dependent on the other.

Construing Relations

259

(17) 打开，

1 蛋白质转运通道 danbaizhi zhuanyun tongdao

dakai

protein transport channels

open

“protein transport channels are opened up” 2 核糖体与通道

结合，

hetangti yu tongdao

jiehe

ribosomes and channels

combined

“ribosomes combine with the channels” 3 新生的肽

插进

通道。

xinsheng de tai

chajin

tongdao

new peptides

insert into channels

“new peptides are inserted into the channels.”

By contrast, hypotactic relations must be marked by conjunctions, and marking on the dependent β clause is necessary. In (18), yinwei (‘since/ because’) in β marks its dependent status. (18) β 因为

SRP

终止了

蛋白质的翻译,

yinwei

SRP

zhongzhi le danbaizhi de fanxi

because

SRP

terminate

translation of proteins

“Since SRP has terminated the translation of the protein,” α 新生肽的合成

被阻断

xinshengtai de hecheng bei zuduan the synthesis of the nascent peptide

is blocked

“the synthesis of the nascent peptide is blocked.”

Another distinguishing feature has to do with relative marking, in terms of whether to mark the initial or continuing clause(s). As shown, in hypotaxis marking has to occur on the β clause. So its relative marking is on β clause. With parataxis on the other hand, although marking is optional, if it does occur, the conjunction appears in continuing clauses. For example in (19) continuing clauses 2 and 5 are marked. (19) 1 光面内质网 guangmian neizhiwang

能够

合成

nenggou hecheng

胆固醇， danguchun

260

Jing Hao

smooth endoplasmic reticulum

can

synthesize

cholesterol

“smooth endoplasmic reticulum first can synthesize cholesterol” 2 然后

将

胆固醇

氧化、

ranhou

jiang

danguchun yanghua

and then

make

cholesterol

oxidize

“and then make cholesterol become oxidized” 3 还原、 huanyuan deoxidize “(make cholesterol become) deoxidized,” 4 水解, shuijie hydrolyse “(make cholesterol become) hydrolysed,” 5 进一步

转变成

各种类固醇激素。

jinyibu

zhuanbiancheng gezhong leiguchun jisu

further

turn into

kinds of steroid hormones

“and further turned into kinds of steroid hormones.”

If with parataxis there is more than one continuing clause and only one conjunction is deployed, then it appears in the final clause—as in (20). (20) 1

2

在进化 过程中，

细菌的质膜

有可能

内陷，

zai jinhua guochengzhong,

xijun de zhimo

youkeneng

neixian,

during the evolution process

bacterial plasma membrane

probably

cave in

“In the course of evolution, bacterial plasma membrane probably caves inward.” 内陷， 附着在质膜上的DNA 也 随之 fuzhuozai zhimo shang de DNA

ye

DNA attached on the plasma membrane also

suizhi

neixian

with them

cave in

“the DNA attached on the plasma membrane are also caved in” 3

然后

逐渐

形成了

ranhou

zhujian

xingcheng le shuangceng de wanquan jiang DNA baobei qilai de hemo

双层的完全将DNA包被起来的核膜。

and then

gradually

form

double-layer SUB completely make DNA encapsulate nuclear membrane

“and then gradually formed a double layered nuclear membrane that completely encapsulates DNA.”

Construing Relations

261

A third way of marking taxis is co-relative marking, where both initial and continuing clauses in a clause complex are marked with conjunctions. In the hypotactic example (20) provided earlier, the necessary marking appears in β clause. The same connexion can also be marked on both clauses in (20b). (20b) β 因为

SRP

终止了

yinwei

SRP

zhongzhi le danbaizhi de fanxi

蛋白质的翻译,

because

SRP

terminate

translation of proteins

“Since SRP has terminated the translation of the protein,” α

所以/于是

新生肽的合成

被阻断

suoyi/yushi xinshengtai de hecheng so

bei zuduan

the synthesis of the nascent peptide is blocked

“the synthesis of the nascent peptide is blocked.”

Similarly, in parataxis both initial and continuing clauses can be marked, as in (21). (21) 1

2

要进入

内质网的腔,

danbaizhi shouxian

蛋白质

首先

yao jinru

neizhowang de qiang

protein

must enter into lumen of endoplasmic reticulum

first

“the proteins must first enter into endoplasmic reticulum lumen,” 然后 通过选择性的分泌过程 输出到 细胞外 ranhou

tongguo xuanzexing de fenmiguocheng

shuchu dao xibaowai

and then

through selective secretion

export to

outside the cells

“and then the proteins are transported out from the cells through selective secretion.”

Table 8.1 summarizes the patterns of marking in parataxis and hypotaxis. The differences in zero marking and relative marking provide important criteria for making a distinction between the relationship between clauses. In the data considered here, with hypotaxis, the dependent clause β tends to precede the independent clause α. However, it is important to note here that these structural patterns were identified in a specific set of instances of implication activities, and we need to be careful not to immediately generalize them across registers.

262

Jing Hao

Table 8.1 Summary of Markings in Parataxis and Hypotaxis

parataxis

zero marking

relative marking

correlative marking

✓

marking only for continuing clause(s) (for all or only culminative link) marking clause β

✓

hypotaxis ✗ (clause β must be marked)

✓

In addition to relationships between clauses, two additional grammatical features—verbal aspect and modal verbs—also demonstrate significant patterns in realizing implication relations.

4.3 Verbal Aspect In Chinese, verbal aspect is a grammatical feature that functions to construe the experience of time within a verbal group (which in turn functions as a Process at the clause rank). Through aspect, an occurrence can be construed as either “unfolding (‘imperfective’, realized with zhe and (zheng)zai), culminating (‘perfective’, realized with le and guo), or neither (‘neutral’)” (Halliday and McDonald 2004: 380). Of particular relevance to the realization of implication activities is the perfective aspect le marking the completion of the occurrence. The aspect le has a particular way of interacting with temporal and causal connexions in that it occurs in the realizations of Posterior figures in temporal sequences and realis Cause/Anterior figures in purposive sequences. In temporal sequencing the perfective aspect le in the Posterior is used to mark the completion of the implication activity. Grammatically, it appears either in a ranking clause related to a borderline embedded/β clause, as in (22), or in a continuing clause in parataxis, as in (23). The aspect markers are underlined. (22) Anterior

Posterior

当

[[神经递质与肌细 胞表面受体

结合]]， 触发

了

信号的 级联反应

dang

[[shenjingdizhi yu jixibaobiaomian shouti

jiehe]]

le

xinhao de jilianfanying

chufa

Construing Relations at

[[neurotransmitters bind]] and muscle cell surface receptors

experiential Circumstance

263

trigger

cascade response of the signal

Process

Participant

Minor Minor Range Process coverb logical

[[embedded clause]]

verb

β

ASP: pf n.g.

α

“(at the time) when neurotransmitters combine with the surface receptors of muscle cells, cascade response of the signal is triggered”

(23) Anterior 1

ER滞留蛋白质

被

错误地

ER zhiliudanbaizhi

bei cuowude

ER retention protein by

包进

分泌泡

baojin

fenmipao

mistakenly is embraced secretory vacuole

“ER retention proteins are mistakenly embraced into the secretory vacuole,” Posterior 2

然后

离开

了

ER

ranhou and then

likai le leave

ER ER

verb

ASP: pf

“and then leave the ER.”

In parataxis, if there are more than one continuing clauses, the aspect typically appears in the final clause. When causal connexions are realized, aspect le appears in realis figures, particularly the Cause/Anterior in the purposive sequence. The realis figure is realized typically through an α clause, as in (24). The use of aspect not only marks the completion of the Anterior/ Cause but also indicates the high possibility the irrealis Effect will take place. (24) Effect: irrealis/Posterior β

为了

能

准确地

weile

neng

zhunque de yunsong

运送

in order to

would

accurately

蛋白质 danbaizhi

transport protein

“in order to accurately transport protein”

264

Jing Hao

Cause: realis/Anterior α

在进化过程中

每种蛋白质

形成

了

zai jinghua guocheng zhong

mei zhong danbaizhi

xingcheng le

in the course of evolution

each type of form protein

一个明确的地址签 yige mingque de dizhiqian a clear address

ASP: pf

verb

“in the course of evolution, each type of protein forms a clear address”

To sum up, perfective aspect marks the completion of an activity in temporal sequencing and the ‘reality’ of figures in causal sequencing. It is used to emphasize the ‘definiteness’ of the unfolding of implication activities. Note that the perfective aspect does not appear in β clauses or borderline embedded/β clauses. 4.4 Modal Verbs A further resource that is involved in realizing implication activities is modal verbs, which are treated as a realization in the interpersonal system of modality in Halliday and McDonald (2004: 339 ff). and Zhu (1996).8 Modal verbs are part of the verbal group, and they cannot precede the first Participant in a clause—as exemplified in (25) (underlined). (25) 破裂的膜碎片

能够

自己 融合形成

polie de mosuipian

nenggou

ziji

球形的小囊泡

ronghe qiuxing de xiaonangpao xingcheng

ruptured membrane would/will itself fuse fragments

spherical vesicles

“The ruptured membrane fragments can fuse themselves to form spherical vesicles.”

Identification of modal verbs involves making a distinction between their expression of modalization and modulation. This distinction is closely related to the opposition between propositions and proposals (Halliday and McDonald 2004; Zhu 1996). In explanations of implication activities, the language is oriented to modalization of propositions rather than modulation of proposals.9

Construing Relations

265

From a grammatical perspective, modal verbs do not appear in embedded clauses or borderline embedded/β clauses. Otherwise there is lack of consistent pattern with respect to how modal verbs interact with clause complexing. Modal verbs were found in both initial clauses and continuing clauses, and in both parataxis and hypotaxis. There is a significant discourse patterns, however, in terms of how modal verbs interact with sequencing. In temporal sequence, both Anterior and Posterior figures in temporal sequencing can be marked by modal verbs. When the Anterior figure is marked, it is typically realized as the first clause in a paratactic complex. In example (26) the modal verb yao (‘would/should’) is used in the Anterior figure realized by clause 1. (26) Anterior 1

蛋白质

首先

要进入

内质网的腔,

proteins

shouxian

yao jinru

neizhiwang de qiang

protein

first

would/should enter into endoplasmic reticulum

“The proteins would first entre into endoplasmic reticulum,” Posterior 2

然后

通过选择性的分泌过程

输出到

细胞外

ranhou

tongguo xuanzexing de fenmiguocheng

shuchu dao

xibaowai

and then through selective secretion export to

outside the cells

“and then the proteins are transported to the outside of the cell through a selective secretion process.”

A posterior figure can also be marked with modal verbs. In (27), the modal verb nenggou in clause 2 expresses medium modality would/will. (27) Anterior 1

Posterior/Anterior

当

[[细胞被匀浆破碎]]

dang

[[xibao bei yunjiang shi posui]]

xibaozhimo polie cheng mosuipian

at

[[cells are broken by time homogenization]]

plasma membrane

β

时， 细胞质膜

破裂成

rupture

膜碎片

membrane fragments

α

“When the cells are broken by homogenization, the plasma membrane of the cell ruptures into membrane fragments.”

266

Jing Hao Posterior

2

破裂的膜碎片

能够

自己

融合形成

polie de mosuipian

nenggou

ziji

ronghe qiuxing de xingcheng xiaonangpao

itself

fuse

ruptured membrane would/will fragments

球形的小囊泡

spherical vesicles

“The ruptured membrane fragments will fuse themselves to form spherical vesicles.”

By marking the high modality of a Posterior figure, the modal verb emphasizes the ‘determining’ nature in implication relations—i.e. an activity will result from the previous activity without interruption. In addition to the interactions with temporal connexions, instances of modal verbs also interact with irrealis Effect figures connected by purposive and conditional connexions. As exemplified in a purposive sequence in (28), the modal verb neng (i.e. nenggou) is used in the irrealis Effect figure; like in the temporal example prior, the modal verb can be interpreted as median probability. (28) Effect: irrealis/Posterior β

为了

能

准确 地

运送

蛋白质

weile

neng

zhunque de

yunsong

danbaizhi

in order to

would

accurately

transport protein

‘in order to accurately transport protein’ Cause: realis/Anterior α

在进化过程中

每种蛋白质

形成

了

一个明确的地址签

zai jinghua guocheng zhong

mei zhong danbaizhi

xingcheng

le

yige mingque de dizhiqian

in the course of evolution

each type of protein

form verb

a clear address ASP: pf

‘in the course of evolution, each type of protein forms a clear address’

In (29), exemplifying a conditional sequence, the modal verb is used to mark the irrealis Effect realized by an α clause. (29) Cause: irrealis/Anterior β ER滞留蛋白质 ER zhiliudanbaizhi

如果

被

错误地

ruguo bei cuowude

包进

分泌泡

baojin

fenmipao

Construing Relations ER retention protein

if

by

mistakenly

267

is embraced secretory vacuole

“if ER retention proteins are mistakenly embraced into the secretory vacuole,” Effect: irrealis/Posterior α 高尔基复合体膜上的这种信号受体蛋白 (就) gaoerjifuhetimo shang de zhezhong xinhaoshoutidanbai

(jiu)

this kind of signal receptor proteins on (then) Golgi complex membrane

会

与ER蛋白 结合

hui

yu ER danbai

jiehe

will

with ER protein

unite

“this kind of signal receptor proteins on Golgi complex membrane will then unite with ER”

The significant co-patterning of modal verbs with sequencing, rather than with different clause constructions, suggests that they play an important role in marking discourse semantic meaning. The co-patterning of modal verbs with sequencing can be related to Martin’s (1992) analysis of connexions in English. He recognizes distinctive ‘modal’ features of connexions, particularly the causal ones, as shown in Figure 8.2 (reproduced from Martin (1992: 194)). His proposal suggests that modality is inherent in causal connexions. Complementing this proposal, this study reveals that modality also interacts with temporal connexions; and modality interacts with causal when figures are irrealis.

Figure 8.2 Network of Consequential Relations in English (Martin 1992: 194)

5. Conclusion This chapter has explored the language features involved in construing relationships between activities in biological science. The study has provided a description of discourse semantic connexions and their realization in grammar, and it has clarified the distinction between embedded clauses and clause complexes, and within clause complexes the differences between parataxis and hypotaxis. It has also shed new light on the

268

Jing Hao

use of aspect and modal verbs in Chinese in relation to meanings made at different strata. In particular, it has revealed a discourse semantic interaction between modality (usually seen as an interpersonal resource) and connexion (logical meaning). Methodologically, the study has demonstrated a way of describing academic language by looking from ‘above’—taking context (the register variable field in particular) as a point of departure. This approach allows grammatical description to be closely related to meaning in discourse semantics that is further related to contextual meanings in register and genre. This approach to language description can be productive for exploring typology of languages, as it allows both the similarities at the higher level strata and differences at the lower level strata to be revealed. Pedagogically, since academic language described in this way is directly relevant to the language patterns required in the specialized register, the description can be appliable to teaching Chinese for specific/academic purposes and helping language learners develop effective communication of academic knowledge in a specialized field.

Acknowledgement I would like to acknowledge the numerous constructive comments made by the three editors on the earlier drafts of this article. Special thanks must also go to Professor Randy LaPolla, Dr. Wang Pin, Dr. Beatriz Quiroz and other colleagues who offered their valuable comments when I presented earlier findings of this work at the Systemic Language Typology Forum in 2017 and 2018. Needless to say, any remaining errors and mistakes are mine.

Notes 1. For a recent three-metafunctional discussion of disciplinary knowledge incorporating field, tenor and mode, see Martin (2017). 2. Hao (2015) uses the labels ‘figure’ and ‘sequence’ presented in Halliday and Matthiessen (1999) because of their familiarity, while the ways in which figure and sequence are identified in the discourse is influenced strongly by Martin’s (1992) discourse semantic description of ‘messages,’ ‘conjunctions’ and ‘activity sequences.’ Note also that following Hao (2015, 2018), the term ‘connexion’ is used here to substitute ‘conjunction’ as a discourse semantic system described in Martin (1992). ‘Conjunction’ is only used to refer to grammatical choices as described in Halliday and Matthiessen (2014). Conjunctions are a grammatical resource that realize discourse semantic connexions. 3. Occurrence figures need to be distinguished from state figures that are excluded from the data selection. State figures construe relationships and do

Construing Relations

4.

5. 6.

7. 8.

9.

269

not involve occurrences—e.g. 短的突出物是菌毛, duan de tuchuwu shi junmao (‘short protuberances are pili’). They are used to describe and taxonomize items in the scientific field. Note that in previous studies the terms initial clause and continuing clause are only used with reference to paratactic clause complexes. In this chapter I refer to the clauses in paratactic complexes only with Arabic numbers—1, 2, 3 etc. The data set in this study does not include projecting clause complexes. A coverbal phrase is comparable to an English prepositional phrase. However, a coverb differs from a typical English preposition in that it can function as a Process in a clause. Coverbs are treated in Chinese as a class of verb (Halliday and McDonald 2004: 317). Note however that if the first Participant in the embedded clause is coreferential with that in the ranking clause (e.g. danbaizhi (‘protein’) in (14)), this test is less revealing. Studies have shown that both auxiliary verbs and adverbs can be used to express modality (Halliday and McDonald, 2004; Zhu, 1996). Modal adverbs typically follow the first Participant but may also precede it in becoming thematic (analogous to English probably, possibly, etc.). However, modal adverbs are not a salient feature in the data set concerned in this study. Some instances revealed in the data, such as hui and yao (analogous to English would, should and will), can be used to express both modalization and modulation. These instances can be treated as modalization in the analysis. The modality type of some other instances, which are typically used in Chinese as modulation (Zhu, 1996), appear to be more ambiguous in the science textbook. These instances include bixu, which were recognized previously as expressions of high obligation in modulation; and ke(yi) and nenggou, both of which were recognized as instances of positive ability (‘can/be able to’), although it is possible to paraphrase bixu, keyi and nenggou with their counterparts, which are more clearly oriented to modalization. If bixu (‘must’) is used as modalization, it is analogous to yiding hui (‘definitely will’), which expresses high probability. If keyi and nenggou (‘can/be able to’) are used to express modalization, they are analogous to hui (‘would/will’), which expresses medium probability.

References Anderson, T. H. and Holsting, A. E. M. (2018). Clause Complexing in Systemic Functional Linguistics—Towards an Alternative Description. Functional Linguistics, 5, 10. Barthes, R. (1977). Introduction to the Structural Analysis of Narratives. In Image, Music, Text. New York: Hill and Wang. Chao, Y. R. (1968). A Grammar of Spoken Chinese. Berkeley and Los Angeles: University of California Press. Doran Y. J. and Martin, J. R. (2019). Field Relations: Understanding Scientific Explanations. In K. Maton, J. R. Martin and Y. J. Doran (Eds.), Studying Science: Language, Knowledge and Pedagogy. London: Routledge. Fang, J. (2015). The Systemic Functional Grammar of Chinese Nominal Groups: A Text-Based Approach. PhD thesis. Sydney: Macquarie University.

270

Jing Hao

Halliday, M. A. K. (2005 [1984]). Grammatical Metaphor in English and Chinese. In J. Webster (Ed.), Studies in Chinese Language (Vol. 8 in the Collected Works of M. A. K. Halliday) (pp. 325–333). London and New York: Continuum. Halliday, M. A. K. (2005 [1993]). Analysis of Scientific Texts in English and Chinese. In J. Webster (Ed.), Studies in Chinese Language (Vol. 8 in the Collected Works of M. A. K. Halliday) (pp. 325–333). London and New York: Continuum. Halliday, M. A. K. and Hasan, R. (1976). Cohesion in English. London: Longman. Halliday, M. A. K. and Martin, J. R. (1993). Writing Science: Literacy and Discursive Power. London: Falmer Press. Halliday, M. A. K. and Matthiessen, C. M. I. M. (1999). Construing Experience Through Meaning: A Language-Based Approach to Cognition. London: Cassell. Halliday, M. A. K. and Matthiessen, C. M. I. M. (2014). Halliday’s Introduction to Functional Grammar. London and New York: Routledge. Halliday, M. A. K. and McDonald, E. (2004). Metafunctional Profile of the Grammar of Chinese. In A. Caffarel, J. R. Martin and C. M. I. M. Matthiessen (Eds.), Language Typology: A Functional Perspective (pp. 305–396). Amsterdam and Philadelphia: John Benjamins. Hao, J. (2015). Construing Biology: An Ideational Perspective. Unpublished PhD thesis, University of Sydney. Hao, J. (2018). Reconsidering “Cause Inside the Clause” in Scientific Discourse— From a Discourse Semantic Perspective in Systemic Functional Linguistics. Text & Talk, 38(5), 520–550. https://doi.org/10.1515/text-2018-0013 Hao, J. (in press). Nominalisation in Scientific English: A Tristratal Perspective. Functions of Language. Hsu, F.-m. (2017). Chinese Characteristics of Clause Complex: The SF Perspective of Achievements From Former Accounts. In J. Webster and X. Peng (Eds.), Applying Systemic Functional Linguistics: The State of the Art in China Today. London: Bloomsbury. Huddleston, R. D., Hudson, R. A., Winter, E. O., and Henrici, A. (1968). Sentence and Clause in Scientific English. Report on the linguistic properties of scientific English. Communication Research Centre, Department of General Linguistics, University College London. Li, E. S.-H. (2007). A Systemic Functional Grammar of Chinese: A Text-Based Analysis. London: Continuum International Publishing Group. Li, E. S.-H. (2017). The Nominal Groups in Chinese. In T. Bartlett and G. O’Grady (Eds.), The Routledge Handbook of Systemic Functional Linguistics. London: Taylor & Francis Group. Li, Q. (2011). 论专门用途汉语教学 (Teaching Chinese for Specific Purposes and Its Textbook Compilation). 语言文字应用 (Applied Linguistics). August, 3, 110–117. Martin, J. R. (1983). Participant Identification in English, Tagalog and Kâte. Australian Journal of Linguistics, 3(1), 45–74. Martin, J. R. (1988). Hypotactic Recursive Systems in English: Toward a Functional Interpretation. In J. D. Benson and W. S. Greaves (Eds.), Systemic Functional Approaches to Discourse (pp. 240–270). Norwood, NJ: Ablex.

Construing Relations

271

Martin, J. R. (1992). English Text: System and Structure. Amsterdam: Benjamins. Martin, J. R. (1995). Logical Meaning, Interdependency and the Linking Particle {-ng/na} in Tagalog. Functions of Language, 2(2), 189–228. https://doi. org/10.1075/fol.2.2.04mar Martin, J. R. (2008). Incongruent and Proud: De-Vilifying “Nominalization”. Discourse & Society, 19(6), 801–810. Martin, J. R. (2013). Systemic Functional Grammar: A Next Step Into the Theory: Axial Relations (Y. Zhu and P. Wang, Trans.). Beijing: Higher Education Press. Martin, J. R. (2017). Revisiting Field: Specialized Knowledge in Ancient History and Biology Secondary School Discourse. In J. R. Martin, K. Maton and B. Quiroz (Eds.), Onomázein (special issue on knowledge and education) (pp. 111–148). Martin, J. R. and Matthiessen, C. M. I. M. (1991). Systemic Typology and Topology. In F. Christie (Ed.), Literacy in Social Processes: Papers From the Inaugural Australian Systemic Functional Linguistics Conference, Deakin University, January 1990 (pp. 345–383). Darwin: Centre for Studies of Language in Education, Northern Territory University. Reprinted in J. R. Martin (2010), SFL theory, Volume 1 in The collected works of J. R. Martin, edited by W. Zhenhua. Shanghai: Shanghai Jiao Tong University Press, 167–215. Martin, J. R. and Rose, D. (2008). Genre Relations: Mapping Culture. London: Equinox. Matthiessen, C. M. I. M. and Halliday, M. A. K. (2009). Systemic Functional Grammar: A First Step Into the Theory. Beijing: Higher Education Press. Matthiessen, C. M. I. M. and Martin, J. R. (1991). A Response to Huddleston’s Review of Halliday’s Introduction to Functional Grammar. Occasional Papers in Systemic Linguistics, 5, 55–84. Quiroz, B. (2013). The Interpersonal and Experiential Grammar of Chilean Spanish: Towards a Principled Systemic-functional Description Based on Axial Argumentation. Unpublished PhD thesis. Department of Linguistics, The University of Sydney. Quiroz, B. (2018). Negotiating Interpersonal Meanings: Reasoning About mood. Functions of Language (Special issue Interpersonal Meaning: Systemic Functional Linguistics Perspectives), 25(1), 130–158. Rose, D. (2001). Some Variations in Theme Across Languages. Functions of Language, 8(1), 109–145. Wang, J. (2003). 细胞生物学 (Cell Biology). 科学出版社. Yang, Y. (2015). Grammatical Metaphor in Chinese. Sheffield: Equinox. Yap, F. H. and Grunow-Hårsta, K. (2010). Non-Referential Uses of Nominalization Constructions: Asian Perspectives. Language and Linguistics Compass, 4, 12, 1154–1175. Zhang, C. (2012). Hànyǔ Fújù Fènlèi Xìtǒng de Gōngnéng Jiǎodù Gòujiàn (汉语复句分类系统的功能角度构建, Classification System of Clause Complex in Chinese: A Functional Perspective), In G. Huang (Ed.), Studies in Functional Linguistics and Discourse Analysis, vol. IV. Beijing: Higher Education Press.

272

Jing Hao

Zhou, Yan-Song (2015). 中医汉语课程与教学的目标定位 (The Aim of the Course and Teaching of Chinese Medicine Chinese). 海外华文教育 (Overseas Chinese Education). No. 76, 415–420. Zhu, Y. (1996). Modality and Modulation in Chinese. In M. Berry, C. Bulter and R. Fawcett (Eds.), Meaning and Form: Systemic Functional Interpretations. Norwood, NJ: Ablex Publishing Corporation.

9

The Baboon and the Bee Exploring Register Patterns Across Languages David Rose

1. Introduction The Saussurean trinity is completed by the concept he attaches to the French word langage—language in general; le langage comprises the linguistic tendencies of the general human faculty. Language in general is a power, a part of human nature, social, individual, heterogeneous and multiform . . . if we take away all the overt individual acts of sujets parlants of any given community, we have the all-important residue, a silent highly-organised system of signs existing apart from and over and above the individual as sujet parlant. Langage minus parole gives you langue, and now we come to the main conclusion: that it is the study of this langue which is the real purpose and object of linguistics. (Firth 1950: 41)

Firth did not entirely agree with this conclusion of Saussure’s, for In the most general terms we study language as part of the social process, and what we may call the systematics of phonetics and phonology, of grammatical categories or of semantics, are ordered schematic constructs, frames of reference, a sort of scaffolding for the handling of events. The study of the social process and of single human beings are simultaneous and of equal validity. (ibid) His vision anticipated major dimensions of systemic functional theory, particularly stratification, from phonetics up to social context, instantiation, from langue to parole, and individuation from langage to ‘single human beings,’ Saussure’s sujets parlants. And equally significant is his view of the social function of linguistic theory, its appliability as ‘scaffolding for the handling of events.’ We can use these criteria to interrogate the purpose and object of typological linguistics, a century after Saussure. In terms of instantiation, typological studies assume certain general potentials that are shared

274

David Rose

by languages, ‘multilingual meaning potentials’ in Matthiessen’s (2015) terms, of which each language is an instance. In terms of stratification and rank, a traditional concern of language typology has been with variations in general potentials at lower ranks of phonology and morphology. Functional language typology assumes some general common potentials in clause rank systems, such as transitivity, mood and theme, in order to describe more delicate variations (Caffarel, Martin and Matthiessen 2004), along with group rank and discourse semantic categories to some extent (Hao 2018; Hao and Wang 2018; Zhang 2018). This chapter takes another step up the stratal hierarchy, to explore some general potentials at the levels of genre and register, and down the clines of instantiation and individuation, to interpret how people might read texts in diverse cultures. The study it reports on began with typological studies of stories in Rose 2001a and 2005. Its data set has since grown as a corpus of traditional stories in languages around the world (accessible in Rose 2018b). Patterns of genre and register are illustrated here with three stories from China (Chinese), Australia (Pitjantjatjara) and Africa (Oromo). The findings suggest that systemic potentials at the level of grammar, such as process types, have co-evolved with register potentials, and hence mirror their organization, and that instantiation in language cannot be divorced from instantiation in register and genre. Indeed, linguistic study cannot avoid displaying instances of register in order to exemplify linguistic features. But the register of examples is usually backgrounded in order to focus attention on patterns of language. This study reverses the focus, backgrounding grammar to attend to patterns of register. 1.1 Theory 1.1.1 Stratification and Instantiation The study is grounded in the model of language in context proposed by Martin (1992). Martin builds on Hjelmslev’s proposal to treat context as a connotative semiotic realized by language as a denotative semiotic (Taverniers 2008). Malinowski’s ‘context of culture’ is modeled as constellations of genres that configure fields of social activity, tenor of social relations and modes of meaning-making. Genre and register (field, tenor, mode) are organized as systems of meaning whose features are realized in structures of social semiosis. Each genre is a configuration of general selections in register systems that is predictable to members of a culture. Register features are realized in language by selections from discourse semantic systems serving ideational, interpersonal and textual metafunctions. Discourse systems are realized in turn by selections from metafunctional systems in grammar, expressed by selections in sound or writing systems. The model is schematized in Figure 9.1.

The Baboon and the Bee

275

Figure 9.1 Connotative and Denotative Semiotics

The term realization is used in SFL for relations of abstraction between (1) strata in semiotic systems, (2) features in systems and structures that expound them and (3) classes in rank hierarchies and functions that compose them. A primary concern in this study is with feature/structure relations, or axis. The axial relation is one-to-one correspondence; each systemic contrast is realized by a unique structural contrast. However, stratal realization is more variable and probabilistic. A familiar example is the relation between functions of tone in phonology and mood in grammar. There are typical ‘unmarked’ relations between tone contours and mood of clauses, such as falling & declarative or rising & yes/no interrogative, but variations are possible, which diversify meaning potential at the stratum of discourse semantics, in the valeur of speech functions (Halliday and Greaves 2008; Halliday and Matthiessen 2014; Rose 2018a). I will use the term co-instantiation for such simultaneous selections from systems at different strata, as texts unfold. Each stratum contributes its own layer of unfolding patterns to a text. Genre contributes the global organization of texts oriented to social goals, structurally realized as text staging. Field contributes series of activities involving people and things, which tenor negotiates and evaluates, and mode organizes as dialogue and monologue, that accompanies or constitutes fields of activity. Stories are usually monologic texts that constitute their own fields; they are ‘self-contextualising’ (Martin and Matruglio 2013). Their fields are specific activities, people and places, while their

276

David Rose

Figure 9.2 Cline of Instantiation

tenor is variable: they negotiate expectancy with listeners/readers, and they may share feelings, judge people and/or appreciate activities, things and places (Martin 1992; Martin and Rose 2007, 2008, 2012). Out of the total reservoir of possible meanings available to speakers, certain selections are more or less likely to co-instantiate, or ‘couple’. From this perspective, an unfolding pattern of coupling is what we call a text. Predictable patterns of coupling can be classified as text types, described by Firth as ‘generalised actuals,’ where the actual is each text. Generalizing further gives us genres on one hand and registers on the other. A text type couples genre with certain selections in register. For example, a fable is a story type that could be a narrative (resolving a complication) or exemplum (unresolved). Its field is imaginary activities involving personified animals, its mode is field constituting spoken monologue1 and its tenor involves the listener with the characters in order to present a moral message. Out of such delimitations of meanings at risk, particular selections as a text unfolds further delimit probabilities for the meanings to follow. The meanings that different speakers apprehend in each instance have been termed ‘readings.’ While this term appears to privilege writing, both spoken and written texts afford variable specific readings. Whether its mode is dialogue or monologue, a text is always an exchange, in which interlocutors attempt to steer each other toward particular readings. These narrowing delimitations of meanings at risk have been modeled as a ‘cline of instantiation’ schematized in Figure 9.2 (Martin 2006; after Halliday and Matthiessen 1999). 1.2 Realizations of Fields in Register and Discourse A field is composed of recurrent sequences of activities. Because they are recurrent, any sequence is to some extent predictable within a field, so that variations from such sequences are counterexpectant. (Martin and Rose 2007: 101)

Recurrence, of course, refers to iterated instantiations, and expectancy depends on speakers’ exposure to recurrences. Following Martin’s (1992)

The Baboon and the Bee

277

model of field agnation, Hao (2015, forthcoming) analyzes the field of undergraduate biology into sub-fields that are structurally realized as series of activities, such as observation, operation and review. In the terms developed by Martin and Rose (2007, 2008, 2012), such series of activities are termed phases. In other words, phases are units of structure at the level of register that realize options in field systems (together with tenor and mode). Hao describes the discourse semantic systems and structures that realize field types interstratally. In particular, phases are realized in discourse by sequences of figures that involve entities in occurrences, or relate entities and their dimensions. In actual texts, occurrences and entities are instantiated by lexical items that specify fields, or presumed by reference items. Within each phase of a text, the sequence of lexical and reference items instantiating figures tends to be expectant, but between phases, lexical relations may be counterexpectant. The term activity is reserved by Hao (2015, forthcoming) and Doran and Martin (2020) for the level of register. The terms entity, occurrence, figure and figure sequence are reserved for the level of discourse semantics. This terminology clarifies and brings together Martin’s (1992) terms and Halliday and Matthiessen’s (1999) use of figure, also used in Rose’s (2001b) description of Pitjantjatjara. The term momented activity now refers to register (in place of ‘activity sequence’) and the term sequence is reserved for discourse semantics (Doran and Martin 2020). The terms process and participant generalize transitivity functions in grammar (Halliday and Matthiessen 2014). Crucially, any activity at the level of register may have variable realizations in discourse, as figure, occurrence or entity, which may also have variable realizations in grammar, as clause, group/phrase or word. Register cannot be interpreted directly from the grammar in which it is realized, except perhaps in the simplest of texts. 1.3 Story Genres Realized Intrastratally by Staging For the purposes of this study, we can apply this type of modeling to relations within and between genre and register, focusing on story genres and their fields. At the level of genre, relations between stories are structurally realized by their staging. Five story genres are described in Martin and Rose (2008) (following Martin and Plum 1997), including recount, narrative, anecdote, exemplum and observation. Each story genre may include an orienting and an optional evaluating stage. Their obligatory stages are shown in Figure 9.3 as realization statements. The first choice in the system is disruption. Recounts differ from other story genres, as their core stage is a Record of activities that follow expectantly. The others are all organized around a core stage that disrupts expectant activity. The second choice is resolution. The disrupting stage of a narrative, the Complication, is resolved through a Resolution stage. The disruption in other genres is not resolved, but it is responded

278

David Rose

Figure 9.3 Story Genres Realized Intrastratally by Staging

to, explicitly or implicitly. Anecdotes, exemplums and observations are distinguished by how the disruption is evaluated, either reacted to, interpreted or commented on. 1.4 Fields of Stories Figure 9.4 offers a tentative modeling of fields in stories, borrowing a few traditional terms from literature. The aim is to position the plot of stories and their overall event sequence in relation to other dimensions of their fields. Plot refers to the overall structuring of activities in in a story that either follow expectantly or counter expectancy. This structuring is outlined in the following section and exemplified throughout the paper. Characters may be people, or personified animals or things. Themes are the abstractions construed by the story, its interpretations of social significance, that may or may not be made explicit. (It may be possible to classify types of story themes, but this is not attempted here.) Setting distinguishes contemporary and historical stories (temporality) that narrate actual experiences or imagined events (actuality). The term actuality is taken from Malinowski, who compares Trobriand myths with historical and imagined stories and contemporary experience: the really important thing about the myth is its character of a retrospective, ever-present, live actuality. It is to a native neither a fictitious story, nor an account of a dead past; it is a statement of a bigger reality still partially alive. It is alive in that its precedent, its law, its moral, still rule the social life of the natives. (1926: 183) Contemporary stories may narrate events experienced by the teller, or a previous teller, or they may be imagined (fictional). Historical stories may

The Baboon and the Bee

279

Figure 9.4 Fields of Stories (Tentative)

also narrate events experienced by actual people, or by imagined people or animals, as in legends, historical fiction, folk tales and fables. But as Malinowski explains, classification of myths depends on perspective. To the bearers of a culture, its myths may narrate actual historical events, while outsiders may see them as imagined. Finally, setting in time and place co-select with these options in temporality and actuality. For example, locations in myths are likely to be actual places known to speakers, whereas locations in folk tales are more often imaginary. The time period in historical stories may be specific, whereas time period is indeterminate in both myths and folk tales. These options in story fields are sub-potentials in a culture’s overall field potential. That is, they are not equally available to other genres, but they are afforded by story genres in particular. On the cline of instantiation, they are a step down from the overall semiotic systems of a culture, co-instantiated with particular selections in the culture’s genre systems. Traditional stories instantiate two fields at once—the particular mundane field of the story plot and the transcendent generalizing field of its theme (Malinowski’s ‘bigger reality’). This generalizing function may

280

David Rose

also be a characteristic of stories in general (Halliday 1978: 58; Rothery and Stenglin 1997). Although this point is also assumed by mythologists such as Propp, Jung, Levi-Strauss or Joseph Campbell, this study is not concerned with reducing story plots and themes to these theorists’ pancultural ‘monomyths’;2 rather, its aim is to compare their organization. 1.5 Story Plots Realized Intrastratally by Story Phases Story phases in English are described in Martin and Rose (2007) and for stories in a range of language families in Rose (2005). They have been described as the basic building blocks of plot structure across story genres. Interstratally, each stage of a genre is realized by one or more phases. Intrastratally, phases realize types of field, so story phases realize types of plot elements. Plot elements include events that either follow expectantly or create or release tension, as well as phases that contextualize the events and interpretations by characters or by narrators. Figure 9.5 describes story phases relationally, in terms of their functions to either contextualize or sequence the story’s plot. That is, story phases can only be described in relation to the sequence of preceding and following phases. As phases are function structures, they are denoted with initial capitals. Contextualization may be internal to the plot or an exegetic commentary by a narrator or character. Internal contextualizations differ in

Figure 9.5 Story Plots Realized Structurally by Story Phases

The Baboon and the Bee

281

position in the story. Settings precede other phases, presenting the time, place or characters’ activities. Descriptions pause the action to describe qualities or behaviors of a character, thing or place. Exegetic contextualizations differ in their source, or voice. Comments stop the action for the narrator to explain or evaluate the activity. Reflections are characters’ thoughts about the activities that may pause or continue the action. Comments interpret activities from outside the story. Reflections interpret from within. Sequencing phases select for expectancy; they either follow preceding phases expectantly, or they counter expectancy. Counterexpectant phases create or release tension. Problems create tension through activities that disrupt expectancy. Solutions release this tension through activities that restore equilibrium. Problems need not have negative effects, but Solutions are positive in some respect. Expectant phases may be modulated as obligatory, where one phase is construed as a consequence of another. Results are material consequences, that may be an activity or quality. Reactions are characters’ feelings, expressed as behaviors or qualities. Expectant phases that are not modulated as consequences are simply termed Events.3 1.6 Story Genres Realized Interstratally by Story Phases Genre systems and structures are realized interstratally by register systems and structures. In particular, the types and stages of story genres are realized by types and sequences of story phases. The counterexpectant stages of a narrative, anecdote, exemplum or observation are necessarily realized by at least one Problem, and narrative Resolutions must involve a Solution, but beyond this the potential for variation is considerable, although not random. Rather there are predictable variations in phasal patterns within each story genre. For example, a Complication stage may be foreshadowed by minor Problem phases within the Orientation, or tension may be built within a Complication by a series of worsening Problems and intensifying Reactions. A recount may consist simply of a series of expectant Events or a series of expectant episodes that include other phases (for examples, see Martin and Rose 2012). The term episode will be used for a series of story phases. Episodes are demarcated by a significant shift in field, and may be signaled by textual markers such as marked Themes (Martin 1992: 447–450). Any story genre may include a series of episodes in its core stage. From the perspective of instantiation, story genres and story phases co-instantiate. This co-instantiation is a step down the instantiation cline from the sub-potentials of story genres, toward the generalized patterns of story types. Story types may include recognizable types of plots, characters, settings and themes, such as the legends, historical fiction, folk tales and fables noted earlier.

282

David Rose

Research to date suggests that systems of story genres and story phases outlined previously are common potentials across language families (Rose 2001a, b, 2005). Questions for this study are how these common potentials are instantiated in particular cultures, and particular story types, and how these instantiations are realized in language.

2. The Stories Three stories are analyzed to illustrate the instantiation of these ‘multilingual meaning potentials’ across cultures and languages. The first is a traditional exemplum from Chinese, now spoken by around 1 billion people, traditionally peasant grain farmers. The second is a mythic narrative from Pitjantjatjara, a dialect of the Australian Western Desert language, spoken by around 7,000 people, who were traditionally arid lands huntergatherers. The third story is a mythic exemplum from the Borana dialect of Oromo, an Afroasiatic language in the Cushitic family, spoken by about 70,000 people in southern Ethiopia and northern Kenya, traditionally nomadic cattle herders. These examples cover a cross-section of pre-modern economic types and language families, but they exemplify a larger data set of stories from various language families and cultures (Rose 2018b). 2.1 Chinese Story “The Farmer and The Hare” is a popular traditional exemplum. A farmer is plowing his field when a hare unexpectedly dashes itself to death, giving him a free meal. The farmer then gives up his work, expecting this extraordinary event to be repeated. According to one interpretation, the story “warns us that everyone needs to work hard in order to reap the harvest. ‘No pay, no gain.’”4 In the transcription provided, the Chinese wordings are glossed in English in two steps. The first gloss is word for word, but arranged in word groups. As far as possible, grammatical items are glossed with English items or affixes, rather than with formal labels. The aim is to render the glossing as comprehensible as possible to readers of English. For example, completed aspect le is glossed here with the English past auxiliary ‘did,’ and continuous aspect zhe with ‘-ing.’ If the Chinese wordings are read aloud, each line will thus be comprehensible to English readers. Alternatively, reading the English gloss gives the flavor of the original grammar without hearing the original phonology.5 Following the English gloss is relatively easy with Chinese, as their grammars share some common structural patterns. For example, grammatical functions tend to be realized by words rather than affixes (so-called analytic languages), nominal group order is Modifier Head (but with some postpositions in Chinese), clause sequences tend to be similar, and once identities are presented they may be ellipsed in following clauses. Note that ‘being’ in

The Baboon and the Bee

283

English may be instantiated as you ‘have’ in Chinese. Also the whole clause is glossed with an English clause, but this is not a so-called ‘free translation,’ as the textual structure is maintained as closely as possible. Text: 1 Ren Tuzi “The Farmer and the Hare” 1 congqian you once

yige ren

zai tian li zhongdi

have one person at field in farm

Once there was a man farming in a field 2 huran

pao lai le

suddenly

run

yizhi tuzi

come did one hare

Suddenly out ran a hare 3 yitou

zhuang zai

headlong

tian bianir de da shu shang

strike be.at field side of

big tree upon

and dashed headlong against a big tree beside the field. 4 tuzi hare

zhuang si le strike die did

The hare dashed itself to death. 5 nage ren that person

feichang gaoxing very happy

The man was very happy. 6 ta ba tuzi shi qilai he disp hare pick up He picked up the hare 7 dai hui jia qu bring back home go and took it home. 8 cong na tian yihou from that day afterwards

ta jiu fang xia chutou he then put down hoe

From then on, he put down the hoe 9 zuo zai sit be.at

da shu dixia deng zhe big tree under wait -ing

and sat under the big tree, waiting, 10 xiwang zai hope

you

tuzi

pao lai

again have hare run come

hoping that another hare would come running, zhuang si zai strike die was

da shu shang big tree upon

and dash itself to death on the big tree.

284

David Rose

11 ta he

deng le hen jiu wait did very long

He waited very long, 12 tuzi hare

mei you zai lai not have again come

a hare never came again 13 ta de tiandi he of field

ke huangwu le but lay.waste did

but his fields lay waste. Source: Bittner, M. 2011 www.rci.rutgers.edu/~mbittner

2.1.1 Periodicity Story phases are realized in discourse as sequences of figures, that are organized in hierarchies of periodicity (waves of textual prominence). It is possible to display these patterns using only the English glosses without needing to repeat the original wordings. This is a further advantage of glossing with English words and affixes where possible. Analysis 1 displays the Themes of each message, up to and including the first participant identity. This is a discourse semantic perspective on Theme that enables us to see how identities are tracked through the Themes of messages, alongside the textual and circumstantial Themes that may precede thematic identities (Martin and Rose 2007; Rose 2001b).6

Analysis 1 Themes in “The Farmer and the Hare” textual/circumstantial

farmer

1

once

one person

2 3 4

suddenly headlong

5 6 7 8 9 10 11 12 13

one hare (it) hare that person he (he)

then from that day afterwards again

hare

disp hare7 (it)

he (he) (he) he

again

hare his field

The Baboon and the Bee

285

Each message includes a finite clause, along with dependent, projecting and projected clauses (Martin 1992: 235). The analysis shows the Theme of the first clause in each message. Phase shifts are indicated with lines. Double lines indicate more significant shifts between episodes or genre stages. The display shows the roles of Themes in structuring the sequence of phases (or method of development). Thematic identities are relatively constant within each phase, tracked by strings of reference or ellipsis. Shifts from one phase to the next are signaled by a) switching the identities of participants, from the farmer to the hare and back in lines 2 and 5, or b) re-identifying participants in lines 5, 8 and 11. Two major phase shifts are signaled by textual and circumstantial Themes in lines 2 and 8. The analysis does not show the functions of phases in the story’s field, but rather how they are signaled. From the speaker/listener perspective, these textual devices manage attention to patterns of expectancy, counterexpectancy and consequence that organize phases of a story (as discussed in 1.5). From a theoretical perspective, they provide evidence of phasal organization of stories. 2.1.2 Phases, Figures and Perspectives Analysis 2 displays the sequence of figures in each phase of Text 1 and nuclear relations within each figure. The nucleus of each figure includes an occurrence and an entity, or a relation between entities. The margin and periphery of a figure includes other entities and dimensions (Halliday and Matthiessen 1999; Martin 1992; Martin and Rose 2007; Rose 2001a). There is a column for conjunctions, and locutions are indicated with speech marks (“”). Phases are labeled to the left. In the analysis and following discussion, phase types are in lower case to clearly distinguish them from genre stages with initial caps. Lexical items instantiating occurrences are in bold to highlight their sequencing. In terms of genre, this story’s primary social function is interpretation of behavior. It is an exemplum, realized intrastratally by a counterexpectant Incident that starts in line 2. The interpretation is left implicit for listeners to infer. The Incident is realized interstratally in two episodes, including a problem (2–4) and a reaction (5–7), and a second problem (8–10) and result (11–13). These episodes involve significant shifts in field, signaled by circumstantial Themes. Within each phase in Analysis 2, figures follow each other expectantly, but relations between phases are either counterexpectant or consequential. The setting establishes field expectancy of farming activity. This expectancy is disrupted by the first problem, signaled by ‘suddenly.’ Unexpectedly, a wild hare runs, dashes against the tree and dies. Consequently, the farmer is happy, picks up the hare and takes it home. In the second episode, marked by ‘then from that day afterwards,’ he unexpectedly puts

286

David Rose

Analysis 2 Phases, Figure Sequences, Nuclear Relations and Perspective in Text 1 phase setting problem hare dies

conjunc

nucleus

margin/periphery perspective

1

one person have farm 2 suddenly one hare run come did 3 (hare) strike 4 hare strike

at field once land

narrator recounting

headlong upon big tree at side of field

die did reaction farmer happy

5 6 7

farmer feeling acting

8 then

that person he disp pick up (person) bring back go he put down

problem stops

very happy hare (hare) home hoe, from that day afterwards sit did big tree wait-ing underneath hope have run come again strike die big tree upon wait did very long not have again come lay.waste did

9

(person)

acting

working 10

(person) hare

result fields

11 12

he hare

wasted

13 but

his field

farmer

feeling narrator recounting

down his hoe and sits waiting for a hare do the same. The result is that he waited a long time and his field lay in waste. Analysis 2 also introduces another angle on the register of stories: the sources of expectancy either from the narrator or from characters, in other words their perspectives on the activities, shown in the right-hand column. The setting and first problem are simply recounted by the narrator, a so-called ‘observer perspective.’ The reaction on the other hand is the farmer’s; he is the source of happiness, which expects him to pick up the hare and bring it home. However, expectancy in the second problem is ambiguous. From the farmer’s perspective, putting down his hoe is expectant as he hopes a hare will come again. But from the listeners’ perspective it is counterexpectant; in a peasant economy, one unlikely incident should not cause a farmer to quit work. While the farmer did not expect the result, for listeners the result is expected by the problem. Hence when recounting the result, the perspective shifts from farmer to narrator. The implicit interpretation of the exemplum is shaped by these perspectives on expectancy. Members of a peasant community would be surprised at the hare dashing itself to death, empathize with the farmer’s reaction to his luck but then judge his hope as foolish and expect the bad

The Baboon and the Bee

287

result—and hence interpret the implicit theme. The perspective shifts to the narrator because the joke is on the farmer, implicitly shared by narrator and listeners. Character perspectives are connoted by signifying activities or properties, whose source is the characters’ consciousness. Here the farmer was ‘very happy’ (feichang gaoxing) and ‘hoping’ (xiwang). These are generalized as feeling, while other activities are generalized as acting, which is a non-signifying activity. Acting is included in characters’ perspectives where it is expected by their signifying. The farmer was happy so he picked up the hare, then he put down his hoe because he hoped a hare would come.8 Perspectives of characters and narrators is known in narrative theory as focalization, after Genette (1982). In critical theory, perspective is ideologically loaded as ‘reading position.’ Martin (2006: 276) examines sourcing and targets of attitude in stories to reveal how “texts are ideologically interested in divergent ways, and the different reading positions naturalised by each text” (taking ‘naturalisation’ from Barthes). The analysis here focuses on the sourcing of expectancy rather than attitude, which is less inscribed in traditional spoken stories than in contemporary written ones. The term ‘perspective’ is formalized to refer to expectancy sourcing (and could also be used for attitude sourcing).9 Crucially, the analysis shows how perspective is related to plot structuring, the phases through which a story unfolds. It is through the patterned coupling of perspectives with story phases that listeners are positioned in relation to ideological themes; in other words, that reading positions are naturalized. 2.2 Pitjantjatjara Story The next story is a ‘Dreaming story’ or myth of the Pitjantjatjara people of Australia’s Western Desert. This narrative is concerned with relations between men and women in this kinship structured society. It opens with two brothers marrying two sisters and camping together at a place called Piltati, hence the name of the story. The men hunt kangaroos in the hills, while the women collect plants and dig for burrowing animals on the plain. Expectancy is disrupted by a drought that forces the women to forage further away each day until they fail to return. The men search but cannot find them, wondering what has become of them and what they will do themselves. They resolve the problem by transforming themselves into giant serpents, rising into the sky from where they can see their wives, and then descend into the earth. In Pitjantjatjara, functions such as tense and circumstance type are realized by suffixes (‘agglutinative’). To enhance readability, tense suffixes are glossed with English auxiliary verbs past ‘go-did,’ and past durative ‘sit-were.’ Circumstance suffixes are glossed with prepositions, locative ‘place-at,’ and genitive ‘kangaroo-for.’ As in the

288

David Rose

Chinese story earlier, realis aspect (imperfective) is glossed with ‘–ing.’ Pitjantjatjara has dual and plural pronouns, glossed here as they2, we2 and they4 (3du, 1du, 3pl). Pronouns also show transitivity function (‘case’), but in this text they are all Medium (nominative). They may be full pronouns or clitic pronouns attached to the first clause element. Pitjantjatjara also uses two different additive conjunctions to track the Medium identity as the same or switched from the preceding clause, glossed and or andSWITCH. Participants tend to be presented at the start of each clause, with processes toward the end, sometimes followed by circumstances or other elements of new information. Text 2 Piltati 1 wati kutjara kunyu man two, it’s said

kuta-rara nyina-ngi brother-pair sit-were

It’s said that there were two men, who were brothers. 2 kungkawara kutjara young woman two

alti-ngu kangkuru-rara marry-did sister-pair

Two young women were married to them, who were sisters. 3 wati kutjara pula man two they2

a-nu malu-ku go-did kangaroo-for

Those two men went hunting for kangaroos. 4 kuka kanyila-ku game wallaby-for

tati-nu puli-ngka climb-did hill-on

For wallabies, that is, they climbed up in the hills, 5 munu pula and they2

kuka kanyila kati-ngu game wallaby bring-did

and they brought back wallaby meat to the camp. 6 ka pula andswitch they2

mai-ku tjaru-ukali-ngu vegetable.food-for down-descend-did

And the other two went down for vegetable foods, 7 munu pula and they2

mai ili ura-ningi food fig collect-were

and were collecting wild figs. 8 ngura-ngka alatjitu=ya place-at exactly=they4

nyina-ngi sit-were

It was right at that place (Piltati) that they were living. 9 munu and

kuka wiya-ringku-la ailuru-ri-ngu game finish-ing drought-become-did

Then as all the game finished a drought began. 10 putu unable

tjawa-ra pitja-ngi dig-ing come-were

Unable to dig anything up, the women were coming

The Baboon and the Bee 11 munu pula kunyu and they2, it’s said

pararitja-kutu a-nu far-towards go-did

Then it’s said they travelled far away, 12 munu pula and they2

ma-antjakari-ngu away-camp.out-did

and camped away overnight. 13 munu pula and they2

ngarin-tjanu-ngku sleeping-after

Then after sleeping, pungku-la kill-ing,

antjakaringku-la wirkati-ngu camp.out-ing arrive-did

killing and camping out further, they finally arrived 14 ngura kutjupa=lta place another=then

tjawaningi dig-were

Then they were digging in another place. 15 ka pula andswitch they2

putu nguri-ra nguri-ra unable search-ing search-ing

Meanwhile the other two were unable to find them. 16 pula kunyu nguri-ra nguri-ra minyma10 uti wirkan-ma they2, it’s said search-ing search-ing woman should arrive They kept on searching, thinking, “The women should arrive.” yaltjiri-ngu pula ai? what.happen-did they2 eh? “What’s happened to them, eh?” parari manti far probably

pula a-nu they2 go-did

“They must’ve gone far.” 17 ka pula andswitch they2

kuli-nu palya-nti think-did alright-maybe

and they thought “OK,” ka=li kuwari-mpa, so=we2 now,

putu nya-kula-mpa, yaltjiri-nku=li? unable see-ing-if, what.do-will=we2?

“so now if we can’t see them, what will we do?” 18 munu pula kulata kulpi-ngka tju-nu and they2 spear cave-in put-did Then they put their spears in a cave 19 munu=lta and=then

kuli-nu=lta think-did=then

Then they thought some more;

289

290

David Rose

20 paka-ra ngara-ngu rise-ing stand-did

karpi karpi twist twist

they leaped up into the sky, twisting around together. 21 kutjara panya tati-rampa kunyu wanampi-rampa nya-ngu=lta two that climb-ing, it’s said wanampi-becom-ing see-did=then Those two rising up, turning into Wanampi, saw, munta aha

nyaratja ngura nyara parari yonder place yon far

“Oh, there they are, that place far away.” 22 munu pula kunyu and they2, it’s said

unngu-wanu=lta a-nu inside-through=then go-did

And then they descended and entered the earth. Source: Rose, D. 2001b

2.2.2 Periodicity The Theme analysis in Analysis 3 shows some interesting parallels with the Chinese story. Thematic identities are again constant within phases, tracked by pronouns or ellipsis, and identities switch from phase to phase. New phases begin with an identity realized by a full nominal group or pronoun, or an identity switch with a switch conjunction. Major phase shifts are also signaled by circumstantial Themes at the end of the preceding phase, in lines 8 and 14 (in contrast to the start of a new phase in Chinese or English stories).

Analysis 3 Themes in the Piltati Story textual/circumst. 1 2 3 4 5

game wallaby-for and

6 7 8

andsw and place-at

9 10

and

11 12 13

and and and

men

women

man two (they2)11

young woman two

game

man two they2 (they2) they2 they2 they2 -they4 game (they2) they2 they2 they2

The Baboon and the Bee textual/circumst.

men

14

place another-at.that

15 16

andsw

they2 they2

17 18 19 20 21 22

andsw and and-at.that

they2 they2 (they2) (they2) two those they2

and

women

291 game

(they2)

2.2.3 Phases, Figure Sequences, Nuclear Relations and Perspective Analysis 4 shows the story’s phases, figure sequences, nuclear relations and perspective. In terms of genre, this story’s primary social function is resolution of a disruption in normality. It is a narrative, realized intrastratally by a counterexpectant Complication that is evaluated and resolved by the characters. The Complication is realized interstratally by a sequence of two problems, the Evaluation by a reaction and the Resolution by a solution. The analysis of nuclear relations and sequences again displays patterns of expectancy within phases, and counterexpectancy or consequence between phases. The settings include three activity series that are expectant within the desert culture: men and women marrying (1–2), men hunting game (3–5) and women gathering food (6–8). This normality is initially disrupted by the problems of game finishing and drought starting (9) and the women unable to dig up food (10). Problem 2 is travelling far to find food (11), camping out because it is too far to return each day, killing and camping out again and arriving and digging at another place, implicitly far from Piltati (12).12 At this point they have left their husbands, further disrupting the normal social order. The women’s unexpected departure expects the men’s reaction, in which they search for the women (15), expect their return and wonder what has happened (16). The solution is counterexpectant, as they ask themselves (17), put their spears in a cave (18), think (19), rise up, twisting around each other (20) and transform into wanampi serpents in the sky, from where they see the women in the distance (21). This solves the problem of their wives’ disappearance, and they dive into the earth (22). In terms of perspective, the source in the first three phases is the narrator simply recounting what happened. In contrast, the reaction is from the men’s perspective, searching for the women, speaking to each other, evaluating what should have happened and wondering what has happened. Likewise, the solution is also from the men’s perspective. They start by wondering what to do then put their spears in a cave, symbolically

6 7 8

9

women foraging

problem1 women cannot forage

problem2 women depart

3 4 5

men hunting

and and and

at.that

14

and

andsw and

and

conjunc

11 12 13

10

1 2

settings marrying

phase

they2 they2 they2

(woman2)

game

they2 (woman2) they2 they4

man two they2 (they2) they2

man two (they2)

nucleus

Analysis 4 Phases, Figures and Perspectives in the Piltati Story

go-did away-camp.out-did sleep-after kill-ing camp.out-ing arrive-did dig.up-were

finish-ing drought-become-did unable dig.up-ing come-were

down-descend-did collect-were sit-were

go-did climb-did bring-did

sit-were marry-did

place another

far-towards

(food)

vegetable.food-for food fig place-at exactly

kangaroo-for game wallaby-for hill-on game wallaby

brother-pair young woman two, sister-pair

margin/periphery

narrator recounting

narrator recounting

narrator recounting

perspective

17

solution men transform

they2

and-then

22

they2 we2 we2 they2

they2 they2 woman they2 they2

two those

andsw “so now if “ and and-at.that

“ “ “

andsw

21

18 19 20

15 16

reaction men worry

think-did unable see-ing what.do-will? put-did think-did-at.that ris-ing stand.up-did twist twist climb-ing wanampi-becom-ing see-did aha go-did

unable search-ing search-ing search-ing search-ing should arrive what.happen-did go-did

yonder place yon far inside-through (earth)

spear cave-in

OK-maybe (women)

far probably

perceiving acting

acting thinking acting

thinking evaluating

men perceiving saying evaluating

294

David Rose

burying this index of their manhood. Their activity then turns again from acting to thinking, whence they amazingly rise up and transform into wanampi serpents. Activities of searching and looking in 16 are generalized as perceiving, while saying is presumed by the men’s locutions in 17. The term evaluating is used to generalize register features that may be realized interstratally in appraisal systems as engagement, graduation, judgement and appreciation. Here the men evaluate events within the signifying activities of saying and thinking. Listeners in this desert culture would recognize the expectant sequence in the setting as an ideal of gender relations and economic activity. They would experience the drought with apprehension and interpret the women leaving as an inversion of normality. Hence they would empathize with the men’s reaction, searching and wondering what has happened. The men’s transformation into serpents is not a solution in itself, but it results in seeing their missing wives and sets the scene for the next episode. Entering the earth in the last line is expectant, as they have become giant snakes. This extract is actually only the first half of the story. The next episode begins from the women’s perspective, when they return to Piltati, discover the wanampi serpents’ burrow and attempt to dig them out, only to be swallowed by them and transformed into wanampi serpents themselves. It is a serial narrative with two Complications resolved through transformations. By transforming into the serpents and entering the earth, the men and women become the ancestral spirits of Piltati and its people. The implicit theme is the web of relations between gender, descent, place and Dreaming. 2.3 Oromo Story The final story is a myth of the Oromo people, traditionally cattle herders of Ethiopia and northern Kenya. The Oromo have been managing conflict over pasture and cattle for millennia, which is an implicit theme of this exemplum. A bee generously gives some honey to his guest a baboon, but the baboon covets the rest of the honey and challenges the bee to a battle. The battle takes place in a field of mogoree herbs, a food for cattle and people, but the bees defeat the baboons, killing many and mutilating their corpses. Like Pitjantjatjara, Oromo uses suffixes for tense and some circumstances but uses post-position words for other circumstances, as its nominal group order is Head Modifier (where English uses prepositions). Participants are also consistently presented at the start of the clause and processes at the end. Oromo also re-identifies the Medium in each verbal group, in the forms of tense suffixes, glossed as follows, past:3sg ‘did:he,’ past:3pl ‘did:they’ and pres:1sg ‘do:I.’ Aspect is glossed as imperfective ‘–ing’ and perfective ‘–to.’

The Baboon and the Bee

295

Text 3 Jaldeesi fa Jenaani, “The Baboon and the Bee” 1 gaaf toko jaldeesi worra kiniisa duf-e day one

baboon home bee

come-did:he

One day a baboon came to the home of a bee. 2 jenaani

kiniisa damma itti

then

bee

honey

kenn-e

him give-did:he

Then the bee gave him some honey. 3 jenaani

jaldeesi

then

guddo damma suni meef-at-e

baboon very

honey that

like-ing-did:he

Then the baboon having liked that honey very much, kiniis irra

fud-acu fed-e

bee from

take-to want-did:he

wanted to take it away from the bee. 4 ammo

waani ta-hu wolaal-e

but

what

be-to know.not-did:he

But he didn’t know what to do. 5 jenaani

waani kiniisa-ni jed-e

then

what

bee-to

an si

say-did:he I

had-u fed-a

you fight-to want-do:I

Then he said to the bee, “I want to fight you.” 6 worra keesani

hojaa tami

had-ani

jed-e gaaf-at-e

time which? fight-did:they say-did:he ask-ing-did:he

tribe your

“What time does your tribe fight?” he asked. 7 kiniisa

waani jed-e-ni

bee

what

say- did:he-to

The bee replied to him. 8 hojaa adduun baa-te time

sun

lafa mogorree keesatti had-ani

emerge-did:she field mogorree in

jed-e

fight-did:they say-did:he

“When the sun comes out, they fight in a field of mogorree plants,” he said. 9 jenaani then

jaldeesi

gal-e

baboon go.home-did:he

Then the baboon went home. 10 jaldeesi dibii hedduu baboon other many

yaam-e call-did:he

and called all the other baboons. 11 wolin together

duf-ani come-did:they

They all came together

296

David Rose

12 kiniisa

had-ani

bee

lafa mogorree keesatti

fight-did:they field mogorree in

and fought the bees in the field of mogorree plants, yoo

addunni guddo baa-te

when

sun

very

emerge-did:she

when the sun had fully risen. 13 ammo

jaldeesi

but

hin-dabs-at-ne

baboon not-win-ing-did:not

But the baboon wasn’t winning. 14 kiniisa

dabs-at-e

bee

win-ing-did:he

The bee was winning. 15 eegi then

lola hobbaas-ani war finish-did:they

Then the war finished. 16 jaldeesi baboon

nama isa kaa du-e hed-e man him who die-did:he

woli

gaaf-at-e

count-did:he each.other ask-ing-did:he

The baboons counted the men who had died, asking each other. 17 jenaani then

waani jed-ani what

say-did:they

Then they said, ilmani wayyu harko lama lacu harka irra son Wayyu Harko two

cir-ani=tti

both hand from cut-did:they=and

“The two sons of Wayyu Harko had both hands cut off karaa keesa

ciciis-ti

road on

lie-are:they

and they are lying in the road. ammalle

harkisoo gudubo fa harkambiyedarbo fa cufa-ya

fit-ani

also

Harkisoo Gudubo and Harkambiyedarbo and all-utterly

perish-did:they

“Also Harkisoo Gudubo and Harkambiyedarbo and all perished.” 18 jabeeni

cubbuu laafinna

strength evil

dugaa hin-dabs-at-u

weakness truth

not-winn-ing-are:not:it

Evil strength does not win against honest weakness. Source: Stroomer, K. 1995. A Grammar of Boraana Oromo. Koln: Rudigger Koppe Verlag

The Baboon and the Bee

297

2.3.1 Periodicity Analysis 5 shows that the explicit conjunction jenaani (‘then’) is frequently used to signal phases shifts (as is also common in English stories). Each phase begins with an explicit identity as Theme, but thematic identities may then switch within phases, more variably than in the Chinese and Pitjantjatjara examples. As in Pitjantjatjara, a major phase shift is signaled by a circumstantial Theme at the end of the preceding phase, in line 8. 2.3.2 Phases, Figure Sequences, Nuclear Relations and Perspective Analysis 6 shows phases, figures and perspectives. In terms of genre, this story’s primary social function is interpretation of behavior. It is an exemplum, realized intrastratally by a counterexpectant Incident that starts in line 3. The Interpretation is stated explicitly as a proverb. The Incident is realized interstratally through two episodes. The first includes the problems of the baboon coveting the bee’s honey and challenging him to a fight. The second episode is signaled by the marked Theme in 8. It includes the problem of rallying the other baboons, then being defeated and finally, the mutilation. Analysis 5 Themes in “The Baboon and the Bee” text/circumst

baboon

1 day one 2 then

baboon

3 then 4 but

baboon

other baboons

other

bee what (to do)

5 then 6 7 8 time sun emerge

what (they say) tribe your bee they

9 then 10 11 together 12

baboon

13 but 14 15 then

baboon

16 17 then

baboon

18

bee

baboon other many come-they:did bee bee war what (they say) strength evil

298

David Rose

Analysis 6 Phases, Figures and Perspectives in “The Baboon and the Bee” phase

conj

setting

1

bee gives

2

problem1 3

nucleus baboon

come-he:did

then

bee

give-he:did

then

baboon

very honey that like-ing-he:did take-to bee from want-he:did be-to what know. not-he:did

baboon covets

4

but

(baboon)

problem2 5 baboon

then “

(baboon) I

challenges 6 bee

“

7 8

problem3 9

“

say-he:did fight-to want-I:do tribe your fight-they:do say-he:did asking-he:did bee say-he:did-to (bees) fight-they:do

then

baboon

baboons 10 fight bees 11 12

13 but

baboon

losing

14

bee

result baboons

15 then 16

war baboon

comment proverb

home bee day one honey him

strength evil

recounting baboon feeling thinking

baboon saying

time which

feeling saying

what time sun saying emerge-she:did field mogorree in

not-winn-ingnot:did winn-ingthey:did

finish-they:did count-he:did man him who die-he:did ask-ing-he:did each.other 17 then (baboons) say-they:did what “ (bees) cut-they:did- son [name] two both hand from “and lie-they:do road on “also [names] perishand all they:did

18

narrator

what bee-to you

go. home-he:did baboon call-he:did other many (baboons) come-they:did together (baboons) fight-they:did bee field mogorree in when sun very emerge-she:did

solution

count the dead

margin/periphery perspective

not-winn-ing- weakness truth it:not:do

baboon saying acting

narrator recounting baboon (perceiving) saying

narrator evaluating

The Baboon and the Bee

299

Nuclear relations and sequences in Analysis 6 again display patterns of expectancy and consequence within and between phases. Within the setting, baboon visiting bee’s home expects the host giving honey to his guest. Although the baboon predictably likes the honey, tension is created when he unexpectedly wants to take the honey from the bee and wonders how. Tension is increased when he unexpectedly challenges the bee to a fight, asks him where the bees fight and accepts the bee’s reply. The second episode shifts the scene to the baboon’s home and the field of battle. Tension is further increased as he rallies the baboons to come together to fight the bees. In the solution, signaled with the counterexpectant ‘but,’ the baboons are unexpectedly losing. This is surprising, considering the many baboons, but it defuses the tension built through the preceding problems. The baboons’ loss then expects the result, of counting their dead. The disaster is amplified in message 17, signaled again with ‘then,’ as the bees have mutilated some named baboon kinsmen. This result strongly expects the concluding proverb. Analysis of sources shows that the perspective in each phase is primarily the baboon’s. In the first problem he likes the honey and wants to take it but knows not how. In problem 2 he tells the bee he wants to fight and asks the bee where and the bee replies. In problem 3 he rallies the other baboons. However in the solution, the perspective shifts to the narrator, as the baboons are losing and the bees are winning. But in the result, the perspective switches back to the baboons again as they count their dead and ask and answer each other. Finally it is the narrator who evaluates the result with the proverb. There is an apparent contradiction here between phases and perspectives. In terms of plot structuring, problems 1, 2 and 3 build tension that is released by the solution. But from the baboon’s perspective, problems 2 and 3 are solutions to problem 1—how to take the honey—and losing the battle is a problem for him, not a solution. The answer is that coveting the honey, challenging to fight and rallying the other baboons are problems for the bee, and the baboons’ defeat is a solution. The global perspective of increasing tension and solution is the bee’s, but it also mirrors the Oromo social order. The story attempts to resolve an inherent contradiction for warrior societies between negotiating resources through exchange (giving the honey) and negotiating through violence (trying to take it). It takes the baboon’s perspective because it is addressed to warriors. Like myths in general, it is a message from the ancestors to their descendants about strategies for survival, in particular, countering greed with generosity and instinct with ideology (Levi-Strauss 1978).

3. Discussion: Expectancy and Sourcing in Story Plots Phases were described earlier as structural units of field and as distinct types of activity organized in series. But phases also have values in tenor

300

David Rose

and mode. In terms of mode, a phase is a pulse of meaning presented in discourse as chains of identity in a wave of periodicity, as our Theme analyses show. In tenor, a phase is negotiated as an exchange, or as a prosody of attitude and engagement. Story phases are negotiated between narrator, characters and listeners. The negotiation is about the activity— how expectant or obligatory activities are and from whose perspective. Like moves in an exchange the role of each phase in the negotiation cannot be understood in isolation. A goal of each story is to negotiate a reading position around the story’s themes. A metaphor of ‘projection’ has been proposed for the relation between the twin fields of a story’s plot and its themes (Rothery and Stenglin 1997). But this metaphor ideationalizes a relation that the analyses here suggest is interpersonal. Reading positions are negotiated phaseby-phase through the sourcing of (counter) expectancy and consequence. They are the point at which the cline of instantiation meets the cline of individuation. That is, readings are afforded by the meanings instantiated in a text and by a person’s resources for interpreting them. Readings vary with speakers’ semiotic repertoires and with the values they embrace. Martin refers to these two dimensions of individuation as allocation (of repertoires) and affiliation, or “how users share attitude and ideation couplings . . . to form bonds, and how these bonds then cluster as belongings of different orders” (Martin and Matruglio 2013: 489). In our stories it is not so much attitude but expectancy that is shared (see also Mandela’s autobiographical The Meaning of Freedom analyzed in Martin and Rose 2007: 276). Stories negotiate sources of expectancy to guide listeners toward particular readings to naturalize reading positions. “The Farmer and the Hare” positions the farmer’s behavior outside the values of a responsible, hard-working peasant identity. The Piltati story positions the women’s’ behavior outside appropriate hunter-gatherer gender relations. “The Baboon and the Bee” positions the baboon’s behavior outside appropriate pastoralist group relations. In terms of the tenor variable of contact (Martin 1992; Martin and Rose 2007), the stories involve the listener, sharing values in the service of ideology. Despite wide variations between cultures, these functions of their stories are remarkably consistent, but the readings depend on listeners’ acculturation. Individuation thus gives us another useful perspective on typology, as expectancy and sourcing in story phases are multilingual potentials for affiliation, but resources for reading them are differently allocated to particular cultural formations. A significant resource for negotiating affiliation is types of counterexpectancy in problem phases. In each story, problems disrupt normalcy, in the sense of regular, predictable activities (what should happen), and propriety, or behavior considered proper to the social order (what should be done).13 In “The Farmer and the Hare,” the hare dashing to death is surprisingly unusual, as is the farmer putting down his hoe, as well as being improper for a peasant work ethic. In the Piltati story, the drought

The Baboon and the Bee

301

is unusual but not surprising in the desert climate, while the women departing is surprisingly unusual and improper for Pitjantjatjara gender relations. In “The Baboon and the Bee,” the baboon coveting the honey and fighting to take it are improper behaviors, but they not surprising for human nature. Other types of problems in this study’s corpus of traditional stories included threats to characters’ safety, difficulty in carrying out activities and loss of people or things. Each type may be more or less surprising. Each problem type also favors certain options for reactions such as sadness for loss, frustration for difficulty and fear for safety. Much more exploration is needed into variations in propriety, normalcy and surprise in problems, solutions and associated reaction types. Sourcing of perspectives is achieved by specific types of activities available to narrators and characters shown in the right column of Analyses 2, 4 and 6. Narrators may recount, describe and evaluate activities. But the perspectives of characters are sourced as signifying activities (saying, perceiving, thinking, feeling), the view from within. In some phases we are only allowed the narrator’s view from outside the events, while in others we experience the events through the eyes, ears, hearts, minds and voices of the characters. It is the switching back and forth from narrator to characters while manipulating expectancy and implication that guides us toward intended readings. Options for co-selecting phase types with perspective in story plots are outlined in Figure 9.6. In addition, the phase selection system enables phase types and perspectives to be selected recursively to build an unfolding story plot.

Figure 9.6 Expectancy and Sourcing in Story Plots

302

David Rose

There is a large literature on perspectives in literary narrative, such as Genette’s (1982) focalization types. Here my concern is with instantiations of perspective in these traditional oral stories as types of activities, such as signifying, acting, recounting, describing and evaluating. These activity types raise two theoretical issues for describing register and register/ language relations. The first is the relation between these types and register phases. Phases themselves are described earlier as structures realizing types of field. The activity types of signifying, acting, recounting, describing, evaluating are more general field options available to phase types, along with types of items such as people and things, and properties of time, place and qualities. Exploring science fields, Doran and Martin (2019) generalize activity types as cyclical or linear and describe activities that are more or less momented as ‘tiers’ of activity. It remains to be found whether the activity types identified for story plots here are more delicate categories or belong to parallel systems. Furthermore, phases are composed of one or more activities, or item/property relations. They are not less momented activities in Martin and Doran’s terms (although this would describe the synopses of each phase in the left column of Analyses 2, 4 and 6). This raises a question of whether the phase/activity relation can be described as tiered or another type of scalar relation. A second issue concerns stratal relations between activity types in field, figure sequences in discourse and process types in grammar. Recognition criteria at each stratum are both structural and lexical. In Halliday and Matthiessen’s (2014) grammar of transitivity, classification of process types appears to reflect activity types in fields such as material, mental, verbal and behavioral processes. Together with varying clause structures, criteria for process types include lexical items connoting field categories such as consciousness, emotion, behavior, symbol source and so on. Such relations between process types in grammar and activity, item or property types in register may be described as co-instantial. Through recurrent co-instantiation, grammatical types have evolved as default patterns for realizing activities such as perceiving, thinking, feeling or saying. From the perspective of grammar such categories appear to delineate the organization of meaning in a language, and embody a theory of reality, as Whorf (1956) proposed (cf. Halliday and Matthiessen 1999). However from the perspective of register, conscious activity may be instantiated in many ways. In story plots for example, perception may simply be implied by a description of what characters see or hear, once their perspective has been established. Thinking and saying may be implied by the signified activity. Fear may be implied by yelling, running away or lying still. Of course, metaphor also becomes a major factor in more written modes. In these cases, recognition of fields depends on the sequencing of activities within and between phases. Discourse semantic structures such as figure sequences are thus critical in the realization of

The Baboon and the Bee

303

fields, construing meanings that are reconstrued again as grammatical meanings. Beyond the structural contributions of grammar and discourse, however, recognition of fields in self-contextualizing texts like stories depends on familiarity with the connotations of lexical elements and inferential connections among them. I am using connotation here in a broad sense, not as a higher level of implicit interpretation beyond a sign’s face value but as a general semiotic relation between symbols and knowledge. This is consistent with its usage in logic (synonymous with intension), including all possible referents or attributes of a term, which inherently also includes meanings defined as ‘literal’ or ‘denotative.’ In fact any connotation always implicates individuals’ allocation of semiotic resources and affiliation with meaning groups. So-called ‘literal’ or ‘first-order’ meanings may be more widely allocated amongst a language community, while ‘second-order’ interpretations are more narrowly allocated and may invoke more specific affiliations. These are differences in degree, in contrast with commonly drawn distinctions in ‘orders’ of meaning (e.g. Barthes 1973). This may have implications in turn for how we interpret the connotative/denotative relation of language and context, and the place of lexis in the stratal hierarchy, if lexical meanings are connotative, realized in and co-instantiating with linguistic structures.

4. Conclusion: Shunting Between Language and Context, System and Text A central hypothesis of this chapter is that a phase is a unit of structure at the stratum of register. The types of phases likely to be instantiated in a text are conditioned from above at the stratum of genre. In story genres, the general functions of phases are contextualizing, sequencing, expecting and modulating each other. As these are relations between them, phases can only be identified in sequence. They can be instantiated in various ways that are realized as patterns of discourse configured in varieties of grammatical structures. However, relations between these strata afford far more varied options for register than grammar and discourse systems can account for. Ultimately, register must be described in its own terms, motivated but not determined by its realizations in language. As far as typology is concerned, systemic linguists are particularly interested in variations in meaning potential, but these variations lie well beyond the stratum of grammar. My own study of grammars from very different cultures, Pitjantjatjara and English, found markedly similar functional potentials despite numerous minor differences (Rose 2001b). What varies most widely is the registers that these grammars are deployed to realize. In this respect, Whorf’s (1956) hypothesis that the grammars of Native American and ‘Standard Average European’ languages encode radically different ways of meaning was premature. What his generation

304

David Rose

lacked was a stratified description of language in context, to adequately describe the cultural differences he perceived. Typological research now has much greater capacity to shunt fruitfully between langue and parole, between mapping the ‘silent highly-organized systems’ of langage and interpreting the ‘social, individual, heterogeneous and multiform’ voices of its sujets parlants.

Notes 1. Although the channel of fairy stories may be writing, the mode of meanings is more spoken than written. 2. Efforts at universalizing should themselves be seen as culturally specific practices (Mignolo 2012). 3. The term ‘event’ is used elsewhere in its non-technical sense to generalize ‘goings-on’ in a story. It is capitalized as Event to technicalize a type of story phase that is expectant and non-modulated. 4. https://sites.google.com/site/chinesefolktales/fable/farmer-and-rabbit 5. The rationale for this unusual glossing is the descriptive focus on register rather than grammar. 6. This Theme analysis extends Halliday and Matthiessen’s (2014) description of Theme in English that limits topical Themes to unmarked Subject/Theme or marked Themes that precede Subject. 7. Some languages have devices for including more than one participant in Theme (Rose 2001a, b). Chinese appears to use the ‘displacement’ particle ba for this function (termed ‘dispositive’ by Halliday and McDonald 2004). 8. Expectancy is a mutual, bi-directional relation in discourse. “Elements of structure, especially in grammatical relations, share a mutual expectancy in an order which is not merely a sequence” (Firth 1957: 17). 9. Martin 2006 uses ‘perspective’ informally for attitude sourcing, e.g. ‘Australian and Japanese perspectives.’ 10. The ‘y’ in minyma is not voiced but indicates a palatal ‘n’ in Pitjantjatjara orthography. 11. Pitjantjatjara can include multiple identities in Theme by means of clitic pronouns and ellipsis. 12. The place the sisters arrived at was about 800km to the southwest of Piltati, told in another version of the story. 13. ‘Normalcy’ is used here in contrast to [normality] in appraisal and [usuality] in modality.

References Barthes, R. (1973). Mythologies. London: Paladin. Caffarel, A., J.R. Martin and C.M.I.M Matthiessen (eds.). (2004). Language typology: a functional perspective. Amsterdam: Benjamins. Doran, Y.J. and J.R. Martin (2020). Field relations: understanding scientific explanations. In K. Maton, J.R. Martin and Y.J. Doran (eds.) Studying science: knowledge, language, pedagogy. London: Routledge. Firth, J.R. (1950). Personality and language in society. Sociological Review 42, 37–52. Firth, J.R. (1957). A synopsis of linguistic theory, 1930–1955. In Studies in linguistic analysis (Special volume of the Philological Society). London: Blackwell, 1–31. (reprinted in Palmer 1968, 168–205).

The Baboon and the Bee

305

Genette, G. (1982). Figures of literary discourse. New York: Columbia University Press. Halliday, M. A. K. (1978). Language as social semiotic: the social interpretation of language and meaning. Maryland: University Park Press. Halliday, M.A.K. and W. Greaves (2008). Intonation in the grammar of English. London: Equinox. Halliday, M.A.K. and C.M.I.M. Matthiessen (1999). Construing experience through meaning: a language-based approach to cognition. London: Cassell. Halliday, M.A.K. and C.M.I.M. Matthiessen (2014). An introduction to functional grammar. London: Edward Arnold. Hao, J. (2015). Construing biology: An ideational perspective. PhD thesis. University of Sydney, www.isfla.org/Systemics/Print/Theses/HAO_2015_PhD_ THESIS.pdf Hao, J. (2018). A ‘top-down’ approach to language description: the language of scientific explanations in Mandarin Chinese. Conference paper. Functional linguistics: descriptive and typological perspectives. Dept. of Linguistics, University of Sydney. Hao, J. (forthcoming 2019). Analysing scientific discourse from a systemic functional linguistic perspective: a framework for exploring knowledge building in biology. London: Routledge. Hao, J. and P. Wang. (2018). Describing relational clauses in Mandarin Chinese: a top-down approach. Conference paper. Functional linguistics: descriptive and typological perspectives. Dept. of Linguistics, University of Sydney. Levi-Strauss, C. (1978). The origin of table manners. London: Jonathan Cape. Malinowski, B. (1926). Myth in Primitive Society. New York: Norton and Co. Martin, J.R. (1992). English text: system and structure. Amsterdam: Benjamins. Martin, J.R. (2006). Genre, ideology and intertextuality: a systemic functional perspective. Linguistics and the Human Sciences 2(2), 275–298. Martin, J.R. (ed.). (2013). Interviews with Michael Halliday: language turned back on himself. London: Bloomsbury. Martin, J.R. and E. Matruglio (2013). Revisiting mode: context in/dependency in ancient history classroom discourse. In H. Guowen, Z. Delu and Y. Xinzhang (eds.) Studies in functional linguistics and discourse analysis. Beijing: Higher Education Press, 72–95. Martin, J.R. and G. Plum (1997). Construing experience: some story genres. Journal of Narrative and Life History 7(1–4), 299–308. Martin, J.R. and D. Rose (2007). Working with discourse: meaning beyond the clause. London: Continuum (1st edition 2003). Martin, J.R. and D. Rose (2008). Genre relations: mapping culture. London: Equinox. Martin, J.R. and D. Rose (2012). Genres and texts: living in the real world. Indonesian Journal of SFL 1(1), 1–21. Martin, J.R. and P.R.R. White (2005). The language of evaluation: appraisal in English. London: Palgrave. Matthiessen, C.M.I.M. (2015). The notion of a multilingual meaning potential: a systemic exploration. Mss. PolySystemic Research Group, Faculty of Humanities, PolyU, Hong Kong. Mignolo, W.D. (2012). Local histories/global designs: coloniality, subaltern knowledges, and border thinking. Princeton: Princeton University Press.

306

David Rose

Rose, D. (2001a). The western desert code: an Australian cryptogrammar. Canberra: Pacific Linguistics. Rose, D. (2001b). Some variations in theme across languages. Functions of Language 8(1), 109–145. Rose, D. (2005). Narrative and the origins of discourse: construing experience in stories around the world. Australian Review of Applied Linguistics Series S19, 151–173. Rose, D. (2018a). Sister, shall I tell you? enacting social relations in a kinship community. Functions of Language 25(1), 98–134. Rose, D. (2018b). Exploring register patterns across languages: the stories. Manuscript, www.readingtolearn.com.au/wp-content/uploads/2016/01/Exploringregister-patterns-across-languages-the-stories.pdf Rothery, J. and M. Stenglin (1997). Entertaining and instructing: exploring experience through story. In F. Christie and J.R. Martin (eds.) Genre and institutions: social processes in the workplace and school. London: Pinter (Open Linguistics Series), 231–263. Taverniers, M. (2008). Hjelmslev’s semiotic model of language: an exegesis. Semiotica 171(1–4), 367–394. Whorf, B.L. (1956). Language, thought, and reality: selected writings. With an introduction by John B. Carroll. Cambridge, MA: The MIT Press. Zhang, D. (2018). Interpersonal functionality bonding and consensus. Conference paper. Functional linguistics: descriptive and typological perspectives. Dept. of Linguistics, University of Sydney.

10 Language Shift Bilingual Exchange Structure in Classroom Interactions Harni Kartika-Ningsih

1. Introduction This chapter attempts to describe code-switching in an intervention program that systematically deploys students’ first language (L1) for the teaching of English (L2). The intervention is based on the Reading to Learn program (Rose, 2018a; Rose and Martin, 2012), which is extended to suit multilingual classrooms (Kartika-Ningsih, 2016). The aim of this chapter is to understand what variations in code-switching occur and what they do in the classroom interactions. The notion of code-switching has been extensively discussed across fields, including sociolinguistics (e.g. Auer, 1999) and psycholinguistics (e.g. Grosjean, 1982). The importance of understanding code-switching has been increasingly emphasized as bilingual education programs have gained a more prominent position in multilingual classroom settings (Garcia and Li Wei, 2014; Lin, 2013). It has been argued that code-switching provides assistance for L2 learners since it uses the language with which the students are most familiar (Lin, 2013; Cahyani et al., 2018). Among many functions of code-switching, it enables smoother knowledge exchanges during teaching and learning, such as translating difficult words or explaining some important grammatical elements (Creese and Blackledge, 2010; Lin, 2013). However, the current description of codeswitching is not sufficient to address the challenge in bilingual education. The reason is that most of those descriptions are not designed for pedagogical purposes, leaving the teachers to use code-switching ‘instinctively’ as needs arise in their teaching. A description of code-switching that derives from a social perspective of language is pertinent as it draws on an understanding of language learning, where classroom interaction involves both learning through language and learning about language (Halliday, 1993). This chapter will begin with a discussion of how Reading to Learn was adapted as the pedagogical basis for the bilingual teaching program. This program actively utilizes systematic code-switching to teach L2. Following this, a discussion of the discourse semantic system of negotiation

308

Harni Kartika-Ningsih

(Martin, 1992; Martin and Rose, 2007) will be used to investigate codeswitching in classroom interactions. This system enables us to attend to the dynamic of language interplay in bilingual classrooms as speakers exchange roles. Patterns of code-switching that emerge from analysis will be represented as a system network in order to clarify the ordered means through which code-switching is deployed.

2. Reading to Learn Bilingual Program The Reading to Learn bilingual program is an intervention program that extends monolingual Reading to Learn pedagogy (R2L) (Rose, 2018a) by systematically involving the first language of the students’ L1 for teaching the L2. The involvement of the Indonesian L1 is done in two ways (Kartika-Ningsih, in press). First, Indonesian reading texts are used along with one English reading text. The selected Indonesian texts are high stakes reading texts that are generically comparable to the English reading text. Also, classroom interactions between the teacher and students are designed to include code-switching between two languages. The examples in this chapter derive from an Indonesian classroom where English integrated with biology teaching is being taught and so, accordingly, involve shifts between Bahasa Indonesia and English (KartikaNingsih, 2016). While the focus of the classroom is on teaching English, Bahasa Indonesia is strategically used, and, to a limited degree, Sundanese is inevitably involved. Sundanese, though not included in the program design, occurs spontaneously between the teacher and students and among students to build solidarity; it generally involves humorous exchanges. The R2L bilingual program itself focuses on English and Bahasa Indonesia for pedagogical design purposes. The students learn science subjects in Bahasa Indonesia throughout their schooling, so the semiotic resources of field are more readily available in Bahasa Indonesia. Using Bahasa Indonesia is thus effective in reducing students’ learning load within the time frame provided by the schools to implement the program. R2L methodology is organized as a curriculum sequence, illustrated as three layers of concentric cycles in Figure 10.1 (Rose, 2018a; Rose and Martin, 2012). Each layer consists of three sequential teaching steps, reflecting the degree of support offered—the deeper the layer, the more intensive the support is provided. The three-tier cycles are designed to assist teachers and educators in carefully designing their lesson plans and fine-tune them according to the needs of their own classrooms. Each cycle consists of several stages, and each stage and cycle serve a different purpose. The first cycle (outer layer in Figure 10.1) deals with the context of the text—including the generic purpose of the text and its field. This outer layer consists of Preparing for Reading, Joint Construction and Individual Construction steps. The second cycle focuses in more detail on the text’s grammar and discourse. Here reading the text involves teachers identifying key words and sentence structures jointly with students. This

Language Shift

309

Figure 10.1 The Cycle of Teaching/Learning in R2L (Rose and Martin, 2012)

middle layer consists of Detailed Reading, Joint Rewriting and Individual Rewriting stages. The third cycle (the inner circle in Figure 10.1) is a more intensive one in that grammar and spelling are in focus. Its stages are Sentence Making, Spelling and Sentence Writing. All of these cycles involve sets of choices that can be selected from by teachers to suit the needs of their teaching and learning. In designing their teaching, teachers can choose to implement a sequence of steps in one tier or move between tiers. For example, if a teacher considers that her students are sufficiently skilled at reading and only requires a little support in writing, she can select the steps in the first tier—namely Preparing for Reading, Joint Construction and Individual Construction. The steps in the first tier assist students in accessing the reading texts and developing the writing skills at a more general level. However, if the teacher considers that the students need more intensive support to access reading texts, the teacher can move between tiers, selecting steps from both the first and second layers of the R2L cycle. The steps that are selected might, for example, unfold as Preparing for Reading, Detailed Reading, Joint Construction and Individual Construction. The teacher would start with Preparing for Reading (to unpack the genre and the field of reading text) followed by Detailed Reading (which further unpacks the discourse and the grammar in the text). Once the reading text has been unpacked sufficiently the students are prepared for jointly rewriting a text with the teacher (repacking the field knowledge in Joint Construction). Each step is designed in continuity with the previous step in relation to the overall three-tier cycle.

310

Harni Kartika-Ningsih

The bilingual intervention program was designed by selecting key steps from the R2L teaching cycle included here—namely Preparing for Reading ^ Detailed Reading ^ Joint Construction ^ Independent Construction (‘^’ indicates sequence). Each of the steps has its own specialized function in which code switching plays a role. All of these steps were used to focus on one reading text. Since three reading texts were involved, the intervention involved three iterations of the pedagogy, each of which consisted of the same sequence of steps. The first two iterations used L1 reading texts. The idea here was to establish familiarity with the discourse patterns and the field of the text for the students. The Preparing for Reading and Detailed Reading steps center on deconstructing the generic purpose and field of the text, primarily using the L1 and minimizing the use of L2. In the Joint Construction step the teacher and the students jointly wrote an L2 text using the L1 notes from the detailed reading step. This step allows for a scaffolded interaction, reinstantiating L1 as L2 reinstantiation. In the third iteration an L2 reading text was used at all stages, since familiarity with the context and discourse patterns of the teaching had been established. Familiarity with the teaching routine assists in reducing students’ learning load. The students can focus on key learning areas of the L2 text such as the field, discourse semantics and grammar, instead of juggling between a teacher’s instruction and the reading text at the same time. As this pedagogic sequence of iterations illustrates, throughout the curriculum cycle there is a gradual shift from primarily L1 use to primarily L2 use—although as we will see, at each stage there are ongoing shifts between the two languages as needs arise.

3. Analysis of Classroom Interaction: Learning Exchange and Exchange Structure To identify code-switching in the program two complementary perspectives are employed: the R2L learning exchange and the discourse semantic system of negotiation from Systemic Functional Linguistics. The R2L learning exchange is a tool for analyzing the teacher-student interactions used in the program that are deliberately planned by the teacher before the class. The discourse semantic system of Negotiation, on the other hand, is a more generalized system that focuses on exchanges in terms of the roles taken up by teachers and students. The complementarity of the two theoretical dimensions enables important insights in terms of both the function of code-switching in classroom instructions and the patterns of code-switching that emerge at different stages. We will briefly overview both the R2L learning exchange model and the system of negotiation here. 3.1 R2L Learning Exchange Exchanges in R2L are represented as a series of learning cycles. The core of the learning cycle is the fulfilment of a task that can only be done by

Language Shift

311

the students through the exchanges. The teacher can give various tasks, such as asking a question or ordering a student to write a math test, but the student’s answer is almost always evaluated. This idea is represented as an orbital structure, as shown in Figure 10.2 by placing Task in the middle of learning exchange. The Task stage in the exchange is often realized as a verbal answer that can be distinguished further depending on the source of the answer (Rose, 2014, 2018a). The Task can be either Propose, which derives from the students’ knowledge or experience, or Identify, which is sourced from an immediate medium being referred to. A Focus stage regularly occurs where the teacher directs the students to the task that needs to be done. This is complemented by an Evaluation where the teacher makes clear whether the task is fulfilled or not. Evaluate phases can be further distinguished as Affirm to indicate affirmation, or Reject to indicate disagreement (see Figure 10.3). This Focus ^ Task ^ Evaluate model resonates with the Initiation ^ Response ^ Feedback (IRF) model popularized by Sinclair and Coulthard (1975). These core elements are very common in classroom talk (see Cullen, 1998; Hargreaves, 2000).

Figure 10.2 Orbital Structure of R2L Learning Exchange

Figure 10.3 Options for Evaluate (Rose, 2014, p. 14)

312

Harni Kartika-Ningsih

In addition to the core phases, peripheral phases may be involved in the exchange. The teachers may provide preparation for the students before asking a question in the Focus phase (Prepare) and give elaboration with more explanation or discussion after the Evaluate phase (Elaborate). These peripheral phases are part of key elements in R2L pedagogic practice, particularly in several steps of the curriculum such as Detailed Reading. The following example shows the use of R2L learning exchange labels. In this excerpt, the teacher first gives a focus question for students, asking where the Javan Hawk-eagle can be found, which is followed by a student’s answer (Identify). The teacher then evaluates this positively with good. Example 1

Core Learning Exchange

spkr exchange

phase

T

Where can we find the Javan Hawk-eagle? Focus

S

Java.

Identify

T

Good.

Affirm

The teacher’s question is labeled Focus, the student’s answer is labeled Identify since the students are asked to locate the word in the text and the teacher’s approval is labeled Affirm indicating a praise good. This pattern of learning exchange is quite typical and can be found in most classroom interactions. But often the students’ answer does not fulfil the teacher’s expectation, either through silence as no student knows the answer or simply through a wrong answer. If this is the case, a rejection is often given as a form of evaluation as exemplified in the following. Example 2 A Rejected Answer spkr exchange

phase

T

What are they (stages in procedure text type)? Focus

S

Human agent.

Propose

T

No, no!

Negate

In this example, the teacher and the students are reviewing the stages of procedure text type that they learned in the previous lesson. The teacher’s question about the stages is a Focus phase. One student proposes Human agent as the answer (Propose), which is rejected by the teacher. Since the rejection is quite obvious, this phase is labeled as Negate. To negotiate and mitigate these instances where students may feel at a loss, or struggle to fulfil the learning task assigned, other peripheral elements are regularly added to the exchange. That is to enable students’ fulfilling the task, shifting the common ritual of quiz-like teaching to a more supportive environment. The core elements are expanded to add Prepare and Elaborate. That is by teacher providing preparation before

Language Shift

313

getting on Focus phase and elaborating after giving evaluation. The next example demonstrates how Prepare phase enables students to identify key words in the text and how Elaborate phase provides a space for the teacher to explain the key information. Example 1’ Learning Exchange With Peripheral Phases spkr exchange

phase

T

This sentence tells us the specific place to find the Javan Hawk-eagle.

Prepare

T

Where can we find the Javan Hawk-eagle?

Focus

Ss

Java.

Identify

T

Good.

Affirm

T

We can see the word Java is in the bird’s name too, which Elaborate tells us where it is from.

It can be seen the core learning exchange is encapsulated by peripheral phases. The teacher tells the students directly which sentence to focus on (Prepare) before asking the question (Focus). As the students answer the question they identify the word successfully (Identify). After the teacher affirms (Affirm) she goes on elaborating the relationship between the bird’s name and the place to find it (Elaborate). These peripheral phases are designed to ensure success for all students in the classroom. Thus in typical R2L classroom talk, teachers support students by providing a scaffolding before initiating the question and expanding the exchange after getting the students’ answers. 3.2 NEGOTIATION 3.2.1 Exchange, Role and Move As noted earlier, a complementary perspective on learning exchanges is through the discourse semantic system of negotiation. negotiation is a resource for “negotiating meaning in dialogue” (Martin, 1992) and examining interaction “as an exchange between speakers” (Martin and Rose, 2007). negotiation is a system at exchange rank whose elements of structure are as moves. At the level of exchange, a basic distinction can be made between speakers exchanging knowledge or action determining the role of the speakers. Speakers exchanging knowledge (e.g. providing information), play a knower role (symbolized as ‘K’) and exchange actions, such as providing service (symbolized as ‘A’). One exchange may involve one or more roles, and each role may be realized by one or more moves. Move is a functional unit of meaning that is realized in either a taggable clause or a minor clause. Knower roles moves are symbolized as K1 or K2 moves, and actor role realized moves are symbolized as A1 or A2 moves.

314

Harni Kartika-Ningsih

In a knowledge exchange there are a few potential structures that may occur (e.g. K1, K2 ^ K1 or dK1 ^ K2 ^ K1). The core role may only consist of one move—the K1 move since the speaker is giving information and hence playing the role of the primary knower. For example, a speaker sharing information where the Javan Hawk-eagle can be found that only consists of one K1 move. Example 3 spkr

Basic Knowledge Exchange exchange

move

Speaker: We can find the Javan Hawk-eagle in Java island. K1

An exchange may also involve two speaker roles, both primary and secondary knowers. The speaker who seeks for the information is called the secondary knower, realized in a K2 move; the speaker who provides the information plays the primary knower role, realized in a K1 move. An example of knowledge exchange involving K1 and K2 moves can be seen next. Speaker A asks Speaker B where they can find the Javan Hawkeagle, positioning him as the secondary knower in a K2 move. This move is responded by Speaker B’s answer in a K1 move, positioning her as the primary knower. Example 4

Knowledge Exchange With Two Roles

spkr

exchange

move

Speaker A: Where can we find the Javan Hawk-eagle? K2 Speaker B: In Java island, Indonesia. K1

This knowledge exchange yields a sequence of structural elements where K2 is followed by K1 (represented as K2 ^ K1 with ‘^’ indicating sequence). Potentially, the primary knower may delay the information for the secondary knower to answer first. In this case, the exchange is initiated with a delayed primary knower move symbolized by dK and structured as dK1 ^ K2 ^ K1. Here is an example illustrating potential delayed knowledge in an exchange. Example 5 Delayed Knowledge Exchange spkr exchange

move

Speaker A: Guess where we can find the Javan Hawk-eagle. dK1 Speaker B: In Java island, Indonesia. K2 Speaker A: You’re right! K1

In an action exchange, the exchange may potentially be realized in different structures such as A1, A2 ^ A1 or dA1 ^ A2 ^ A1. The core may consist of just an A1 move such as passing chili sauce, opening a door etc. In the exchange where two roles are involved, the speaker demands the action is secondary, realized by an A2 move, and the primary actor makes an A1 move. The exchange here demonstrates a command for Speaker B

Language Shift

315

to show where the Javan Hawk-eagle can be found (A2), and which location is pointed out by Speaker B (A1). Example 6 Action Exchange spkr

exchange

move

Speaker A: Show me in the map where I can find the Javan Hawk-eagle. A2 Speaker B: Here (pointing). A1

A delayed action may also take place. The next example involves Speaker A who offers to show the habitat of the Javan Hawk-eagle in a dA1 move. As Speaker B responds as the secondary actor (A2), Speaker A does the action by pointing, indicating his role in A1 move. Example 7

Delayed Action Exchange

spkr

exchange

move

Speaker A: Speaker B: Speaker A:

Let me show you where we can find the Javan Hawk-eagle. dA1 Sure. A2 Here (pointing). A1

In all examples each role is realized in one move. But a move may be complex, which means that it is realized in a series of interrelated moves performing the same function in the exchange. In this case, an equals sign ‘=’ is used to indicate moves in the complex. An example of move complex can be seen as follows. Example 8

Exchange With Move Complex

spkr

exchange

Speaker A: Guess where we can find the Javan Hawk-eagle! The famous Garuda bird. Speaker B: In Java island, Indonesia. Speaker A: Yeah. You’re right!

move dK1 =dK1 K2 K1 =K1

As illustrated, Speaker A initiates the exchange with two dK1 moves Guess where we can find Javan Hawk-eagle, The famous Garuda bird; this initiation is responded to by speaker B in a K2 move In Java island, Indonesia; and the exchange closes with B’s K1 move complex Yeah, You’re right. Note that each of these moves can be tagged. For example the dK1 move complex could have included the tagged move (1) Guess where we can find the Javan Hawk-eagle, will you?, and (2) (It’s) the famous Garuda bird, isn’t it? 3.2.2 Negotiating Meaning in Classroom In classroom settings, teachers tend to be the primary knower and are thus positioned to evaluate students’ answers; and students tend to be positioned as the secondary knowers whose answer is being evaluated

316

Harni Kartika-Ningsih

(see Martin and Rose, 2007; Rose and Martin, 2012; Rose, 2014). For this reason, if a teacher asks a question whose answer she will evaluate, her role is labeled as dK1 (delayed primary knower); her question is a kind of knowledge test rather than a genuine question. Students who answer the question play a K2 move, which the teacher needs to evaluate as valid or not. In a typical pedagogic exchange, the teacher’s K1 evaluation is anticipated but delayed by a question (dK1) and a learner’s response (K2), so the sequence is dK1 ^ K2 ^ K1. Example 1” Pedagogic Knowledge Exchange spkr

exchange

move

T S T

Where can we find the Javan Hawk-eagle? dK1 Java. K2 Good. K1’

In Example 1”, a knowledge exchange takes place. The teacher’s question Where can we find the Javan Hawk-eagle? is labeled as dK1 move. The student’s answer Java is a K2 move, which is followed by the teacher’s K1’ move in response to the student’s answer, indicating the teacher gives praise. Since a teacher’s evaluation is an important part of knowledge exchange, the use of an apostrophe (‘) will be given following K1 (K1’) to indicate the use of positive evaluation or to affirm or praise the student’s response. This is designed to correspond to the Affirm phase in the R2L learning exchange and distinguish from other evaluation phases (see Figure 10.2). Less often, a student may ask a question to the teacher, or the teacher may ask a student for information that she does not already know. Exemplified in Example 9, a student asks the teacher if the words Javan and Java are different (K2), followed by the teacher’s answer (K1). The sequence for this exchange is K2 ^ K1. Example 9

Student Asking for Information

spkr

exchange

S T

Ma’am, are Javan and Java different? K2 They are used in a sentence differently. K1

move

Turning to action exchanges, usually the teacher directs students’ activities or behavior, so the sequence is teacher in A2 move and student in A1 move. Less often, a student may ask permission for an action (dA1), the teacher gives permission (A2) and the student performs the action (A1), so the sequence is dA1 ^ A2 ^ A1. Example 10

Pedagogic Action Exchange

spkr

exchange

S

I want to scribe it. dA1

move

Language Shift T S

Sure. [scribing]

317

A2 A1

In Example 10, an action exchange takes place in a classroom. The student offers to scribe a sentence on the board I want to scribe it labeled as dA1 move. The teacher’s response Sure is an A2 move, which is followed by the student doing the action (scribing) as A1 move. In the next section, we will apply both the learning exchange notation and the negotiation analysis to a series of classroom exchanges in order to bring out the complementary perspectives they each bring (see Rose, 2014, 2018b; Kartika-Ningsih and Rose, 2018). From the R2L learning exchange perspective, the classroom interaction data can be observed in terms of its function in the teaching program. Negotiation then allows us to identify the dynamics of language shift in the interactions. From this, we will be able to observe the functionality of code-switching in teaching/learning exchanges.

4. Language Shift: Mapping Code-Switching 4.1 Overview of Code-Switching in R2L Bilingual Program As noted previously, in the R2L bilingual program L1 and L2 are used to varying degrees depending on the stage of the teaching. In the first iterations, that use L1 reading texts, the classroom interactions are designed to primarily use the L1 in order to establish the field-specific meanings and discourse patterns that the texts utilize. Once familiarity with the topic and discourse patterns have been established, the use of L2 is progressively increased. At the exchange level, this shift from L1 to L2 parallels that of the overall curriculum genre. In the early stages, L1 use tends to be prioritized in the initiating Prepare and Focus phases, while L2 use starts to take part in latter Evaluate and Elaborate phases. In the last iteration on the other hand, which uses an L2 reading text, L2 use starts up in the initiating phases and continues through the exchange. However, the teacher will turn to the L1 whenever needed to ensure students’ success in completing the tasks. For the rest of this chapter, the code-switching patterns that occur in the R2L bilingual program we will be described through a system termed language shift (Kartika-Ningsih and Rose, 2018). The system of language shift describes both (1) the interplay between the L1 and L2 that becomes the focus of the R2L bilingual program, and (2) the overarching meaning-making process realized by two or more languages. As discussed earlier, the data drawn from the program focused on two languages. Thus L1 here refers to the use of Bahasa Indonesia and L2 to English. 4.2 The System of Language Shift As noted previously, in classroom interactions secondary knower roles are typically taken up by the students who undertake tasks and primary

318

Harni Kartika-Ningsih

Figure 10.4 Options for Language Shift in Exchanges

knower roles by the teacher who focuses and evaluates the task (Martin and Rose, 2007; Rose, 2014). In formal terms the students undertake the Task, for example proposing an answer (Propose) or identifying words in a reading text or board (Identify); and the teacher adopts the primary knower roles in Prepare and Focus and Evaluate and Elaborate phases. The possibilities for language shift are represented paradigmatically in Figure 10.4. The system has two general options—interrole and intrarole. Interrole involves language shift between exchange roles, and intrarole involves language shift within exchange roles. Intrarole has a further option of intermove and intramove. Intermove involves language shift which occurs between moves, while intramove involves language shift within moves. A typical language shift occurs between roles. In the following exchange, for example, the teacher and student develop most of the learning exchange (Prepare ^ Focus ^ Identify) in Indonesian, before the teacher evaluates in English with Good, that’s very good! As this shift occurs between roles, this type of language shift will be termed interrole. Bold font is used to transcribe moves in Bahasa Indonesia, followed by italicized font for the English translation. Regular font indicates the use of English. Example 11

Interrole Language Shift

spkr

exchange

phase

T

Perhatikan di baris keenam. Pay attention to the sixth line. Ada satu kata yang artinya tempat asli dari huruf “E.” There’s one word which says original place starts from ‘E.’ Endemik. Endemic. Good, that’s very good!

Prepare A1

S1 T

Focus

move

dK1

Identify K2 Affirm

K1’

In this learning exchange, the teacher begins with an A1 move in Bahasa Indonesia (L1), asking the students to pay attention to a part of the text. This move realizes a prepare phase (Prepare) that prepares students for identifying the word. Following this, the teacher continues to speak in the L1 through the Focus phase, asking the students to find a word that starts

Language Shift

319

with E. As one student correctly identifies the word ‘endemic’ in the text in a K2 move (Identify), the teacher closes the exchange in a K1’ move in L2, affirming the student for her correct answer. Interrole language shift is an important routine in the early iterations of Detailed Reading stage in which L1 reading texts are used. Initiating the exchange in L1 reduces the students’ learning load and enhances the students’ success at identifying key words. Closing the exchange in L2 marks an introduction of L2 use through praise. This creates a positive environment for learning English and at the same time gives them familiarity with the discourse pattern. Intermove language shift is usually found either in the initiating or closing moves and can only occur in a complex of moves realizing the same role. Within the move complex, languages shift one move at a time. Next is an example of intermove language shift that occurs in the initiating moves of K1 and dK1 moves respectively. Example 12

Intermove Language Shift

spkr

exchange

phase

move

T

Still in the same sentence. Masih dalam kalimat yang sama. Still in the same sentence. I want you to find two words on the part of the body. Yang paling mendominasi. (The one) that dominates the most. Tail! Tail! (noisy) Mostly tail! (loud) Good one, that’s very good!

Prepare Prepare

K1 =K1

Focus

dK1

Some S S2 T

=dK1 Identify K2 Identify K1 Affirm K1’

In the initiating move of Prepare phases, the teacher uses a series of moves to assist students in identifying words in the text. She begins with a K1 move complex, using English (L2) for the first K1 move and Bahasa Indonesia (L1) for the second K1 move. Both moves are used to prepare students (Prepare) and orient them to the sentence they are reading. The following Focus phase follows a similar pattern, this time with a dK1 move complex that asks them to identify the words. The first dK1 move is in the L1 and is then continued in the L2. After the students’ correct answer in K2 move (Identify), the teacher closes the move in L2 realized in K1’ move as she gives praise (Affirm). Intermove language shift is used dominantly in the last iteration of Detailed Reading stage in which English reading text is used. As the students are familiar with the discourse pattern of the exchange, the teacher begins to involve English early on. The idea here is to re-instantiate a move in both languages within roles, familiarizing students with task setting in both L1 and L2.

320

Harni Kartika-Ningsih

Intramove languages shift occurs within one move. The shift may involve just one lexical item (such as a metalanguistic or technical term, or grammatical elements, such as tagging or particles), or several units of meaning. The following example presents an exchange that involves intramove language shift from the Joint Construction stage. As the teaching progresses, the teacher and the students jointly write a text on the board, and during the interactions intramove language shift occurs in almost each move. Example 13

Intramove Language Shift

spkr

exchange

T

Nah kita mau lanjutkan the male. Prepare K1 So we need to continue (the sentence) with the male. Kalau kita bilang the male, kita bisa kasih tahu ya, the Prepare =K1 male artinya the male purple crown fairy wren. If we say the male, we can tell that the male means the purple crown fairy wren male.

T

phase

move

In this exchange, the teacher enacts a series of K1 moves that prepare the students for writing a sentence they are discussing (Prepare). She begins with Bahasa Indonesia (L1): So we need to continue (the sentence), followed by shifting to English (L2) specifying the words the male (K1). She then continues preparing the students (Prepare—K1) by beginning with Indonesian then tracking the English words the male, to refer to the male purple crown fairy wren. In this intramove language shift, the move is basically done by using L1, but the language shift occurs when the lexical items used are in L2, despite the fact that those items that are technical terms have equivalent terms in L1.

5. Text Examples We will now look at how language shift works in longer examples. These texts were taken from both classes using the R2L bilingual program and from regular EFL classrooms. Each text sample serves as an illustration of where a particular type of language shift tends to occur. 5.1 L1 Based Detailed Reading In the R2L program, one iteration of the learning cycle uses the L1 reading text as the teaching material for teaching English (L2). Since it occurs toward the beginning of the lesson sequence, interrole shift is used more often, with initiating moves in L1 and closing moves in L2. Example 14 presents a series of interrole shifts of this kind. In this text, the teacher carefully assists the students in identifying the wordings in the L1 reading text, followed by

Language Shift

321

the activity of students identifying and highlighting the wordings. In doing so, she creates a consistent pattern of Prepare ^ Focus ^ Evaluate, realized as K1 ^ dK1 ^ K1, with interrole language shift used throughout. Example 14 spkr 1

2

3

4

5

6

T

Interrole Language Shift exchange

Sekarang nama yang kita kenal sebagai orang Indonesia. Now it’s about its name which we know in Indonesian. T Ada yang tahu, nggak? Does anyone know? S Me! Me! Elang jawa. T Good, that’s very good! T Elang jawa! Highlight Elang jawa. Ss (students highlighting their text) T In Indonesian the Javan Hawk-eagle is called Elang jawa. T Perhatikan tulisannya. Pay attention to the writing. Ss (looking at the text) T Elangnya “E”nya besar tapi “J” nya kecil. The “E” in Elang jawa is capital, but the ‘J’ is small. T Sama halnya dengan Nisaetus bartelsi, “N”nya besar, “B”nya kecil. It’s the same with Nisaetus bartelsi, “N” is capital, “b” is small one. Some S Oh iya ya. Oh (that is) right. T Perhatikan di baris keenam. Look at the sixth line. Ada satu kata yang artinya tempat asli dari huruf “E.” There’s one word which says original place starts from ‘E.’ S3 Endemik. Endemic. T Good, that’s very good!

phase

move language shift

Prepare

K1

Focus

dK1

Propose Evaluate Direct

K2 K1’ A2 =A2 A1

interrole

Elaborate K1

Direct

A2

Focus

A1 K1

Focus

K1

K2f Prepare

K1

Focus

dK1

Propose

K2

Evaluate

K1’

interrole

322 7

8

Harni Kartika-Ningsih T

Endemik artinya tempat asli di mana Elang jawa ditemukan. Endemic means the original place of where Elang jawa can be found. T Di mana Elang jawa bisa ditemukan? Where can Elang jawa be found? S6 Jawa. Java Some S Jawa. Java T Good, that’s very good. T Highlight ‘Jawa.’ Ss (highlighting Jawa)

Prepare

K1

Focus

dK1

Propose

K2

interrole

K2 Evaluate Direct

K1’ A2 A1

There are eight exchanges in this excerpt that focus on identifying and highlighting words in the text the students are looking at. In exchange 1, the teacher initiates the move by preparing them (Prepare) for identifying the words (K1) before asking a question (Focus) in another move (dK1). These two initiating moves are done in Indonesian (L1). It is followed by a student answering the question correctly (Identify) in a K2 move. The teacher then gives an affirmation (Affirm) in K1’ move in the English L2: Good, that’s very good. Exchange 2 involves the teacher asking the student to highlight the words Elang jawa in their reading text in a series of A2 moves, followed by students highlighting the words (A1). After the students highlight the words, the teacher elaborates on the answer by focusing on the way of writing Elang jawa, emphasizing the use of capitalization on the first word. All of the moves are done in L1. Exchanges 6–8 demonstrate another means of shifting language between roles. These exchanges follow the same pattern as the previous exchanges in that the activities in focus include identifying and highlighting words. In exchange 6, the initiating moves consist of the teacher preparing for the students to identify the words (A1) in L1, followed by the question move (dK1) still in L1. As a student answers successfully (K2), the teacher closes the move by giving affirmation (K1’) in L2. In exchange 7, the teacher initiates a move explaining what endemic is (K1), followed by a question (dK1). These initiating moves are in L1. When the students answer Jawa (K2), the teacher closes the move (K1) in L2: Good, that’s very good. In exchange 8, the teacher asks the students to highlight Jawa (A2), followed by the students highlighting the word (A1). These exchanges indicate the typical pattern of designed interrole language shift for this stage of the teaching/learning cycle. The teacher initiates a series of Prepare and Focus phases in L1, typically through A1 and

Language Shift

323

dK1 moves; and she closes the exchange with an Evaluation in the L2, usually through a K1 move. So we have language shift between roles. This pattern suits interactions where L1 reading texts are used.

5.2 L2 Based Detailed Reading We now move forward to observe classroom interactions where the language shift occurs within roles but between moves. In the following example, the teacher and students are involved in a detailed reading of an L2 text. In particular they are studying a report in English about the Purple-crowned Fairy Wren, an Australian bird. In this series of exchanges the teacher assists the students in identifying the length of the fairy wren. The learning exchanges are similar to the previous example, but the language shift here tends to be intermove in that L2 is used whenever an L1 equivalent has been modeled in previous moves or previous lessons. Example 15 spkr 1

T

Language Shift in Between Moves exchange

Selanjutnya. (It is) the next one. I want you to find how long the fairy wren is. S1, maybe you can find it? S1 [looking at the sentence] Lima belas Fifteen T Lima belas apa? Fifteen what? S1 Sentimeter. Centimeters. T Fifteen centimeters! That’s very good! 2 T Please highlight fifteen centimeters. Ss [highlighting] 3 T Still in the same sentence. Masih dalam kalimat yang sama. Still in the same sentence. I want you to find two words on the part of the body. Yang paling mendominasi. (The one) that dominates the most. Some S Tail! Tail! (noisy)

phase

move

language shift

Prepare

K1

intermove

Focus

dK1

Direct

A2 A1 K1

Identify

cf

intermove

rcf Evaluate Direct

K1 =K1’ A2

Prepare Prepare

A1 K1 =K1

intermove

Focus

dK1

intermove

=dK1

K2

324

4

S2 T T

5

Ss T

Harni Kartika-Ningsih Mostly tail! (loud) Good one, that’s very good! Now please highlight mostly tail. [highlighting] Now you can see mostly tail. Panjang tubuhnya didominasi oleh ekor. The length of its body is dominated by its tail. We call it ‘mostly tail.’

K1 K1’ A2 A1 Elaborate K1 Elaborate =K1

intermove

Elaborate =K1

In this text, the cycle consists of a sequence of exchanges focusing on identifying and highlighting text and elaborating further the meanings found in the text. In exchange 1 the teacher begins her K1 move in Indonesian “selanjutnya,” (it is) the next one, as a preparation to help the students to move to the next sentence (Prepare). The teacher then continues with a dK1 move in English, asking the students to identify the length of the fairy wren (Focus). In the next move, she asks a student (S1) directly in English to answer the question (Direct). Following the student’s incomplete answer, the teacher shifts to Indonesian in a tracking move so the student can give a complete answer “limabelas apa?” (fifteen what?). Finally, the teacher approves his answer in English, fifteen centimeters, and affirms it: That’s very good! (Affirm). Following this, an activity exchange is enacted in English as students are asked to highlight fifteen centimeters in their reading text. In exchange 3, a series of K1 moves are used to ensure that all students can follow, serving as a Prepare phase for the students. L1 and L2 are used interchangeably here involving intermove language shift. First, the teacher asks the students to keep paying attention to the same sentence in K1 move in L2: Still in the same sentence. This is then followed by another K1 move, but this time it is shifted to Indonesian “Masih di kalimat yang sama,” Still in the same sentence. The next moves are a series of dK1 moves realizing Focus but still using intermove language shift. The teacher asks the students to find two words in the text that indicate parts of the bird. The first dK1 move is in L2: I want you to find two words on the part of the body, and it is followed by another dK1 move in L1 to help students identify the words “Yang paling mendominasi”: The one that dominates the most. Several students provide an answer (K2), Tail, which is not fully correct. A student (S2) then steps up and answers loudly (K2): Mostly tail. This is then followed by the teacher’s praise (K1’) in L2: Good one, that’s very good. This exchange is followed by another highlighting activity in which the teacher gives the instruction in an L2 move: Now please highlight mostly tail (A1). In exchange 5, the teacher elaborates on the meanings developed

Language Shift

325

in the previous exchanges by again using intermove language shift to point out to students why the bird is described as being ‘mostly tail.’ The exchanges illustrated previously indicate two ways of using intermove language shift typically found in this cycle. The first is to use intermove shift one move at a time in different learning phases. As illustrated in exchange 1, the teacher initiates the exchange by using L1 in Prepare phase followed by shifting to L2 in Focus phase; each phase is realized in one move. The second is to shift languages in a move complex of the same phase. As shown in exchanges 3 and 5, Prepare and Focus phases consist of a series of dK1 and K1 moves, so the languages shift within the move complex of the same functional phase. This designed strategy of intermove shift allows the teacher to assist the students in attending to the task of reading in L2, and at the same time guide them to interpret the L2 instruction with less of a struggle. 5.3 L2 Based Joint Rewriting The following stage in the learning sequence is joint rewriting. In this stage, the teacher and students write a text together using the knowledge and patterns they have learned in the previous stage. They discuss sentences to be written on the board by using the highlighted words from detailed reading. At this point, the teacher uses intramove most of the time to help the students write the text. She uses L1 to explain what to write in a sentence and mentions the words to include in the sentence in L2. A series of examples shows the exchange during the joint rewriting session. Example 16

Intramove in Joint Rewriting

spkr exchange 1

T

T T

phase

Kita ingin membuat kalimat yang Prepare intinya mengenai the male yang tidak memiliki lilac crown, hanya punya lilac crown selama musim kawin. We want to create a sentence basically about the male which doesn’t have the lilac crown, (but) only has the lilac crown during the breeding season. Kita mulai dari subjeknya dulu ya. Prepare Let’s start from the subject. Ini kan secara umum jadi Prepare subjeknya ‘purple crown fairy wren’. This (the previous sentence) is in general, so the subject is ‘purple crown fairy wren.’

move language shift K1

intramove

=K1 =K1

intramove

326

Harni Kartika-Ningsih T

T

T

2

T

Ss

3

T T

S1 T T

Nah kita mau lanjutkan the male. So we need to continue on the male. Kalau kita bilang the male, kita bisa kasih tahu ya, the male artinya the purple crown fairy wren male. If we say the male, we can tell, right, that the male means the purple crown fairy wren male. Kita mulai dengan the male, the male not retain his lilac crown Let’s start with the male, (that is) the male—not retain his lilac crown. Masih ingat tidak artinya retain? Do you still remember what retain means? Memiliki. To retain (in BI). (nodding) Nah kita tidak ingin sama dengan retain, kita tidak ingin sama dengan kata yang sudah ada di teksnya. But we don’t want to keep ‘retain,’ (because) we don’t want to write the same (word) with what it’s in the (reading) text. Keeps. Keeps boleh! Keeps is possible! S1 bilang “keeps,” the male keeps his lilac crown S1 said ‘keeps,’ (so) the male keeps his lilac crown.

Prepare

=K1

intramove

Prepare

=K1

intramove

Prepare

=K1

intramove

Focus

dK1

intramove

Propose

K2

Affirm Focus

K1 dK1

intramove

Propose Affirm

K2 K1

intramove

Elaborate K1

Intramove language shift is used throughout each of the three exchanges. The first exchange involves a series of K1 moves that function to prepare the students for creating a sentence (Prepare). The teacher explains that the sentence needs to include the information about the male’s lilac crown, which is only donned during the breeding season, and that it should start with ‘the male lilac crown.’ The next moves give a reason that the words ‘purple crown fairy wren’ cannot be used since it is about the bird in general. She then suggests that the sentence can begin with ‘the male’ because it refers to the male species of the bird.

Language Shift

327

In exchange 2, the teacher asks if the students remember what retain means in L1 (Focus), with the students then proposing an answer (Propose) and her affirming through a nod. Following this, the teacher indicates that the same word retain cannot be used again through a Focus realized by a dK1 move. A student proposes the word keeps (Propose), and the teacher approves in an intramove shift (keeps is possible) as an Affirm realized by a K1 move. Through this series of intramove exchanges, the teacher builds up a model of L2 sentence writing by using the students’ knowledge of the field, in this case the male and retain. As students become familiar with the idea of the unique changes of the male lilac crown, she starts to repack the grammatical construction of new L2 sentence. In this case, the students learn that the words 1) lilac crown are not to be changed, 2) the male can be used again, and 3) retain should be changed to another word, which is all made possible through intramove language shift. Intramove language shift is effective in the joint rewriting stage for several reasons. First, since creating L2 sentences is a challenge for most students, the L1 is used to assist students in creating L2 sentences. In the previous stages, the field knowledge has been unpacked through detailed reading. This means that they are more prepared in both register knowledge and the discourse patterns realizing this. However, their grammatical knowledge, particularly for written language, still requires more work to be successful. Second, L2 words that create intramove language shift are the point of reference that becomes the focus of the L2 sentence teaching. In exchange 3, the student proposes keeps for replacing retain, and the teacher uses his answer as a stepping point to model how keeps is used in a sentence—as the male keeps his lilac crown. 5.4 EFL Classrooms In EFL classrooms, it is typical for teachers to use code-switching during teaching and learning despite often being encouraged to use L2 only (Lin, 2013; Mahboob and Lin, 2016). When EFL teachers explain, comment or give instructions to students, they tend to create a series of moves that inevitably involve language shift. The final example we will look at in this chapter involves a regular EFL class that is not undertaking the R2L bilingual pedagogy program. Here, the EFL class is concerned with the teaching of science in a CLIL (content and language integrated learning) context, and the teacher is giving comments on the students’ homework. The homework involves writing a lab report for their science experiments that were conducted in the previous lesson. The lab report includes a procedure, and so in this excerpt the teacher comments on the stage and phases necessary for a procedure genre.

328

Harni Kartika-Ningsih

Example 17

Intramove in Regular EFL Classroom

spkr exchange 1

2 3 4 5

6

7

8

T

phase

move language shift

Ada yang lengkap di bukunya. Prepare K1 Some parts are complete in their books. Oh ya goalnya ada =K1 Oh yes, (some have) the goal Titlenya ada. =K1 (Some have) the title. Stepnya ada =K1 (Some have) the step. Dan explanationnya ada, and also =K1 the language features of the text. And (some have) the explanation, and also the language features of the text Tapi yang Ibu lihat kayaknya dari =K1 from the evaluation, the text ada yang masih kebalik-balik. But what I’ve read I think from the evaluation, some texts are still confusing. Mana itu numbering and then =K1 temporal conjunction. Which part is ‘numbering,’ and then (which part is) temporal conjunction. Yang temporal conjunction itu kan =K1 yang kayak first What temporal conjunction is, for example, ‘first.’

L1

intramove intramove intramove intramove

intramove

intramove

intramove

This series of exchange consists of eight K1 moves, all part of the preparation for the students (Prepare) where the teacher comments on the students’ homework. Beginning with the use of L1, the teacher mentions that some of the students have completed the task. She further elaborates in another K1 move, mentioning the task completion by referring to the metalanguage in L2 within the move: goal, title, step, explanation and language features. In the next move, she comments on the rest of the students’ homework that she thinks is still confusing. She begins with giving comments in L1 “Tapi yang ibu lihat kayaknya” (But what I’ve read I think), then continues with using L2, from the evaluation, the text, in L2 within one move. The final two moves elaborate what she means as “masih kebalik-balik” (still confusing). She points out numbering and temporal conjunction as the types of conjunction that some students could not differentiate. She then gives an example of temporal conjunction which is first.

Language Shift

329

Throughout the exchange there is a tendency of shifting from L1 to L2 whenever metalanguage terms are necessary. Despite the fact that equivalent terms exist in L1, the teacher’s preference for using L2 lexis leads to the use of language shift within moves. It would appear that intramove language shift of this kind is ideationally motivated, concentrating on L2 technical terms, rather than the L2 grammar and discourse patterns needed to build a rich understanding of the field involved. Designed code shifting, or language shift, of the kind introduced in this chapter is needed to help teachers move beyond the limitations of this type of instinctive code shifting interaction.

6. Conclusion This chapter has examined how code-switching can be designed for enhanced teaching and learning in classroom interactions. It introduced the system of language shift, distinguishing shift in terms of whether the shift between languages is interrole, intermove or intramove. Interrole shift involves shift, which occurs between roles where the initiating move is done in a different language from the closing move. Intermove shift involves shift that occurs in a move complex, as the language shifts between moves playing the same functional role in an exchange. Intramove shift involves a shift that occurs within moves, which tends to be realized in shifts from one lexical item or grammatical element to another. This system maps the choices for language shifting that a teacher can employ during classroom talk. The three types of language shift were used in the R2L bilingual program to a different degree depending on the teaching steps and learning exchanges, among other pedagogic considerations. This kind of shifting languages is effective since it is built through reading and writing and the highly supportive talk-around-text accompanying it. It is carefully designed to enable students to succeed at fulfilling the demands of their learning task. Interrole is used in detailed reading to unpack L1 reading text; intermove is used in detailed reading to unpack L2 reading text; and intramove is used in joint rewriting mainly to assist in repacking the knowledge accumulated in the previous steps. The system of language shift offers a novel approach for designing metapedagogy in bilingual teaching programs. It purposely develops the common practice of code shifting and counters the notion of L2 only instruction by demonstrating precisely how L1 use can play an assigned role in building students’ L2 competence.

References Auer, P. 1999. From code-switching via language mixing to fused lects: toward a dynamic typology of bilingual speech. International Journal of Bilingualism, 3(4), 309–332. Retrieved from http://dx.doi.org/10.117/13670069990030040101

330

Harni Kartika-Ningsih

Cahyani, H., de Courcy, M., and Barnett, J. 2018. Teachers’ code-switching in bilingual classrooms: exploring pedagogical and sociocultural functions. International Journal of Bilingual Education and Bilingualism, 21(4), 465–479. Creese, A. and Blackledge, A. 2010. Translanguaging in the bilingual classrooms: a pedagogy for learning and teaching? The Modern Language Journal, 94(1), 103–115. Cullen, R. 1998. Teacher talk and classroom context. ELT journal, 52(3), 178–187. Garcia, O. and Li, W. 2014. Becoming bilingual and biliterate: sociolinguistic and sociopolitical considerations. In C. Stone, E. Siliman, B. Ehren, and G. Wallach (Eds.), Handbook of language and literacy (pp. 145–160). New York: The Guilford Press. Grosjean, F. 1982. Life with two languages: an introduction to bilingualism. Cambridge, MA: Harvard University Press. Halliday, M.A.K. 1993. Towards a language-based theory of learning. Linguistics and Education, 93–116. Hargreaves, A. 2000. Four ages of professionalism and professional learning. Teachers and Teaching: Theory and Practice, 6(2), 151–182. Kartika-Ningsih, H. 2016. Multilingual re-instantiation: Genre pedagogy in Indonesian classrooms. PhD thesis, Sydney University, www.isfla.org/Systemics/ Print/Theses/HKartika-Ningsih_thesis.pdf Kartika-Ningsih, H. and Rose, D. 2018. Language shift: analyzing language use in multilingual classroom interactions. Functional Linguistics, 5, 9. Retrieved from https://doi.org/10.1186/s40554-018-0061-0 Kartika-Ningsih, H. in press. Implementing the Reading to Learn Bilingual Program in Indonesia. In K. Rajandran and S. Abdul-Manan (Eds.), Discourses of Southeast Asia: A Social Semiotic Perspective. Singapore: Springer. Lin, A. 2013. Classroom code-switching: three decades of research. Applied Linguistics Review, 4(1), 195–218. Retrieved from http://hdl.handle.net/10722/184270 Mahboob, A. and Lin, A. 2016. Using local languages in English language classrooms. In W. A. Renandya and H. P. Widodo (Eds.), English language teaching today: building a closer link between theory and practice (pp. 35–49). New York, NY: Springer. Martin, J.R. 1992. English text: system and structure. Philadelphia and Amsterdam: John Benjamins Publishing Company. Martin, J.R. and Rose, D. 2007. Working with discourse: meaning beyond the clause. London: Continuum International Publishing Group. Rose, D. 2014. Analysing pedagogic discourse: an approach from genre and register. Functional Linguistics, 1, 11. Retrieved from www.functionallinguistics. com/content/1/1/11, https://functionallinguistics.springeropen.com/articles/ 10.1186/s40554-014-0011-4 Rose, D. 2018a. Pedagogic register analysis: mapping choices in teaching and learning. Functional Linguistics, 5(3), 1–33. Springer Open Access, Retrieved from http://rdcu.be/HD9G Rose, D. 2018b. Reading to learn: accelerating learning and closing the gap. Teacher training books and DVDs. Sydney: Reading to Learn. Retrieved from www. readingtolearn.com.au. Rose, D. and Martin, J.R. 2012. Learning to write, reading to learn: genre, knowledge and pedagogy in the Sydney school. South Yorkshire/Bristol: Equinox Publishing. Sinclair, J. and Coulthard, M. 1975. Towards an analysis of discourse. Oxford: Oxford University Press.

11 Academic Formalisms Toward a Semiotic Typology Y.J. Doran

Introduction Highly technical discourse regularly uses a range of formalisms for organizing its knowledge. Resources such as mathematics, system networks, tree diagrams and nuclear equations occur through varied disciplines and are crucial components of the texts students need to read and write to be successful in these areas. For the development of literacy pedagogy that can be embedded in disciplinary teaching and learning, an understanding of how these formalisms work and why they occur is vital. To date, however, multimodal discourse analysis informed by Systemic Functional Linguistics (hereafter SFL) has not provided comprehensive descriptions of academic formalisms, and the work that has been done has focused on specific resources in specific disciplines (such as mathematics; O’Halloran 2005; Doran 2017). This chapter extends this work by exploring a set of formalisms used in linguistics and physics to ascertain their role in building technical knowledge and to highlight some of the properties they share across disciplines. We begin by looking at some phases of technical discourse across physics and linguistics. In the first, a university physics textbook describes the behavior of gases (Young and Freedman 2012: 591–592; bold and italics in original): Measurements of the behaviour of various gases lead to three conclusions: 1. The volume V is proportional to the number of moles n. If we double the number of moles, keeping pressure and temperature constant, the volume doubles. 2. The volume varies inversely with the absolute pressure p. If we double the pressure while holding the temperature T and the number of moles n constant, the gas compresses to one-half of its initial volume. In other words, pV = constant when n and T are constant. 3. The pressure is proportional to the absolute temperature. If we double the absolute temperature, keeping the volume and

332 Y.J. Doran number of moles constant, the pressure doubles. In other words, p = (constant)T when n and V are constant. These three relationships can be combined neatly into a single equation, called the ideal-gas equation: pV = nRT

(ideal-gas equation)

Text 11.1 Young and Freedman (2012: 591–592)

For outsiders to physics, this text is likely impenetrable. It is written for those with significant training in physics and is introducing a highly technical component of its overall knowledge structure—the ideal-gas equation. What we will be concerned with in this chapter is why the text introduces the mathematical equation pV = nRT at the end of this phase, rather than just continuing with language. Later on in the book (pg. 1465), when introducing the nuclear reactions of nuclear fission (used in nuclear power and weapons) and nuclear fusion (used by the sun to produce energy), there is a shift from language to a different type of formalism referred to as nuclear equations. In the following excerpt, the nuclear equation (indented in the text) does not synthesize what comes before (as in the previous example), but it is used as a point of departure for a discussion of kinetic energy in fission: Fission Reactions You should check the following two typical fission reactions for conservation of nucleon number and charge: 235 1 92 U  0 n

o

236 92

U* o

144 89 1 56 Ba  36 Kr  3 0 n

235 1 92 U  0 n

o

236 92

U* o

140 54

Xe 

94 38 Sr 

2 01 n

The total kinetic energy of the fission fragments is enormous, about 200 MeV (compared to typical α and β energies of a few MeV). The reason for this is that nuclides at the high end of the mass spectrum (near A = 240) are less tightly bound than those nearer the middle (A = 90 to 145). Referring to Fig. 43.2 [not shown], we see that the average binding energy per nucleon is about 7.6 MeV at A = 240 but about 8.5 MeV at A = 120. Therefore a rough estimate of the expected increase in binding energy during fission is about 8.5 MeV – 7.6 MeV = 0.9 MeV per nucleon, or a total of (235)(0.9 MeV) ≈ 200 MeV. Text 11.2 Young and Freedman (2012: 1465)

The nuclear equations here indicate a series of events in the process of nuclear fission. Stepping through the first equation, it says that Uranium-235

Academic Formalisms

333

1 ( 235 92 U ) collides with a neutron ( 0 n ) to produce (→) a highly energized 236 * Uranium-236 atom ( 92 U ) that then decays (→) into three distinct com89 ponents: a Barium-144 atom ( 144 56 Ba ), a Krypton-89 atom ( 36 Kr ) and 1 three neutrons ( 3 0 n ). The nuclear equation distils this quite technical event into a single symbolic formula. Moving from physics to linguistics, we see similar shifts. In his seminal description of English transitivity, Halliday (1967: 39–40), describes a set of process types in English before encapsulating the description in a system network:

Two of the three process types are each associated with only one participant, non-directed action with actor and ascription with attribuant; structurally, that participant is the subject in each case. The third, directed action, is associated with two participants, actor and goal, either of which may be the subject. The four examples could thus be grouped as follows: Process types: (S=actor) (S=goal) (i) (i) she washed the clothes the clothes were washed non-directed action (iii) (S=actor) the prisoners marched ascription (iv) (S=attribuant) she looked happy directed action

Let us now represent these in terms of grammatical features of the clause using the following labels: extensive effective operative receptive descriptive intensive

clause with ‘action’ process-type clause with ‘directed action’ process-type clause with ‘directed action’, subject as actor clause with ‘directed action’, subject as goal clause with ‘non-directed action’ process-type clause with ‘ascription’ process-type

These features may be organized in systems ordered in delicacy as follows:—

Text 11.3 Halliday (1967: 39–40)

334 Y.J. Doran Here, the description of process types is reworked in language until it is synthesized into the culminating system network. Read from left to right, the network indicates that there are two types of major clause: extensive and intensive. Within extensive there are two further types, effective and descriptive, and within effective there are two further types, operative and receptive. The system network synthesizes these types of clause in one diagram. In another example from linguistics, after explaining context-sensitive grammars and exploring some example sentences, Lyons (1967) chooses to represent an example—the chimpanzee eats the bananas—through a tree diagram. A similar diagram to this is drawn in Figure 11.1. Read from top to bottom, this says that a sentence (∑) includes a singular noun phrase (NPsing) and a singular verb phrase (VPsing). Looking at just the NPsing branch on the left, the tree also specifies that the noun phrase includes an article (T) and a singular noun (Nsing), which in turn includes a noun stem (N) and a zero suffix (ø). Finally, the dotted line indicates that the noun is substituted with the word chimpanzee.

Figure 11.1 Tree Diagram

Academic Formalisms

335

These four formalisms all occur in highly technical discourse and are used to synthesize key aspects of the knowledge of their disciplines. But at first glance it is not entirely clear exactly why they are used or how we can understand them in terms of their meaning-making. To see these as a simple short-hand for language—some sort of abbreviation or acronym— is to miss the functionality of these resources for distilling large amounts of intersecting technical meaning (Martin 1993: 252). Neither does it account for their role in deriving new technical meanings (Doran 2017, 2018a; Martin 2013). The fact that formalisms are used so extensively across different disciplines suggests they form a crucial component of much academic discourse. But the fact that there are many diverse types of formalism—system networks, tree diagrams, nuclear equations and mathematics being only four of innumerable relatively discipline-specific formalisms—suggests they each maintain their own functionality for their own particular disciplinary context. This chapter is concerned with how to describe these formalisms, taking into account what they all share as formalisms used to organize technical knowledge, and how they all differ in terms of their specific functionality. In broad terms, this chapter takes a step toward developing a semiotic typology of academic formalisms. The description in this chapter arises from three broad research programs. The first has to do with a long running educational linguistics concern with exploring the nature of knowledge across disciplines—understandings that are fundamental to the development of linguistically informed pedagogy and curriculum embedded in disciplines (Rose and Martin 2012). This action research program has been considerably enhanced through dialogue between Systemic Functional Linguistics (SFL), which underpins this chapter, and a sociological framework known as Legitimation Code Theory (LCT) (Maton et al. 2016, 2020; Martin et al. 2019; Maton and Doran 2017). In work arising from this dialogue, it is clear that disciplines that utilize a range of formalisms tend to be associated with the sciences (e.g. physics and chemistry) or social sciences (e.g. linguistics and psychology), in comparison with the humanities (e.g. cultural studies and art history), which tend not to use formalisms (Parodi 2012; Lemke 1998).1 In terms of Legitimation Code Theory (LCT), this marks a broad distinction between the knowledge codes tending to underpin the sciences and the knower codes that are more often prevalent in the humanities (Maton 2014). From this perspective, we can ask why academic formalisms tend to be used more in knowledge codes than in knower codes. The second research program relevant to this chapter is the expanding field of multimodal semiotics. This research is concerned with the full range of meaning-making resources used in human communication, and how they come together to organize our lives. From the perspective of multimodality, academic formalisms are pervasive in particular registers; but as noted prior, aside from mathematical symbolism (see Lemke 1998, 2003; O’Halloran 2005; Doran 2017, 2018a, b, 2020), they have

336 Y.J. Doran not yet received a great deal of attention. These semiotic resources have been considered more language-like when compared to resources such as images, animation, architecture etc. (cf. Lemke’s 2003 mathematics in the middle interpretation of mathematics in relation to language and image in science, and O’Halloran’s 2005 language-based approach to mathematics). Questions about the inherent similarity and difference among the meanings afforded by different modalities of communication have yet to be resolved, in part because an adequate theory of broader semiosis is yet to be developed. So from this perspective, we can ask about the specific functionality of academic formalisms and their joint communicative affordances. The final research program contextualizing this chapter is the focus on Systemic Functional language description and typology explored in this book. In particular, it addresses the question of what we mean by functional language typology (see Martin and Quiroz this volume), and how we describe languages in ways that enable us to compare and contrast their functionality. In the spirit of the previous two research programs, this concern with functional language typology can be expanded into asking how we can go about developing a functional semiotic typology of academic formalisms in particular and semiosis in general. With reference to these three research programs, this chapter will explore four academic formalisms—system networks and tree diagrams in linguistics, and nuclear equations and mathematical symbolism in physics. It will address both their grammatical organization and the technical meanings they organize. The analysis will focus on two main points. First, each formalism construes a distinct set of meanings that are crucial for the development of knowledge in the discipline they are used in. Importantly, the particular meanings that each formalism construe is discipline specific. Unlike language that can organize an enormously wide range of different ideational, interpersonal and textual meanings for use across a range of registers and genres, each formalism is much more specialized— targeting one or two types of ideational meaning. We will explore this through an evolving model of field in SFL presented in Doran and Martin (2020). Second, although each formalism construes different meanings, they are all able to iterate these meanings to an indefinite extent. That is, no matter the particular meanings they construe, each formalism can make as many of these meanings as necessary for purposes of knowledge building. This potential for repetition underpins the general functionality of academic formalisms, as it enables highly complex and elaborated meanings to be integrated relatively efficiently. Iteration is enabled by the grammatical organization of each resource. Although at first glance each formalism appears very different, the need for repetition of field-specific meanings means that they are all organized around a particular type of structure that Halliday (1965) has referred to as a univariate structure, which he defines as repetitions of a single variable.

Academic Formalisms

337

These two points, specialization and iteration, will shape the chapter. We begin by focusing on the differences between these formalisms in terms of the particular field-specific meanings they construe. Following this, we will come back to their similarities in terms of their ability for iteration and the grammatical organization enabling this.

Academic Formalisms and Field Systemic networks, tree diagrams, nuclear equations and mathematical symbolism all construe different types of meaning. To get a sense of their differences, we will step through each from the perspective of field, which can be viewed in common sense terms as considering the ‘content’ meanings of educational knowledge. More strictly speaking, in SFL terms, field is a resource for construing phenomena—for building knowledge across all walks of life. In terms of Doran and Martin (2020), system networks and tree diagrams both construe relatively static relations between items. Beginning with system networks, as noted earlier these networks make distinctions in terms of type and subtype. In the network in Text 3, extensive and intensive are types of major clauses, while effective and descriptive are types of extensive clauses and so on. This organization of clauses into types and subtypes is known as a classification taxonomy. Tree diagrams also organize taxonomic relations between items. However whereas system networks organize their meanings into type-subtype relations, tree diagrams organize them into part-whole relations—a composition taxonomy. In Figure 11.1, for example, singular noun phrases (NPsing) and singular verb phrases (VPsing) are not types of sentence (∑), but rather they are parts of the sentence. Similarly, suffixes such as -ø are -s are not a type of noun (N) or verb (V), but they are parts of these word classes. In this sense, the two linguistic formalisms of system networks and tree diagrams both construe taxonomic relations, but one construes taxonomies of classification while the other construes those of composition.2 Whereas the two linguistic formalisms construe a relatively static perspective, nuclear equations in physics symbolize a dynamic perspective on unfolding events. Each arrow in these equations construes a shift from one set of particles to another. In terms of field, rather than establishing taxonomic relations between items, nuclear equations realize unfolding activities. In equation (1), the nuclear reaction is organized in terms of two activities, each specified by →: (1)

235 1 92 U  0 n

o

236 92

U* o

144 89 1 56 Ba  36 Kr  3 0 n

1 236 * 235 1 The first activity, 235 92 U  0 n o 92 U , specifies that 92 U  0 n (Uranium* 235 and a neutron) shifts to 236 92 U (an excited uranium-236 atom). The

338 Y.J. Doran * 144 89 1 second activity 236 92 U o 56 Ba  36 Kr  3 0 n specifies that the excited 89 1 Uranium-236 atom splits into 144 56 Ba  36 Kr  3 0 n (a Barium-144 atom, a Krypton-89 atom and three neutrons). By placing both of these activities in a single equation, the formalism indicates that they form moments of a broader activity, which is known in language as nuclear fission. This means in Doran and Martin’s (2020) terms that the full equation involving multiple arrows realizes a momented activity. In addition to these activities, nuclear equations also realize a set of compositional relations. The symbols on either side of the arrows 1 236 * 144 89 1 ( 235 92 U  0 n , 92 U , 56 Ba  36 Kr  3 0 n ) specify the particles involved 1 at each step of the activity. At the highest level, the expression 235 92 U  0 n specifies that in the system at this step of the activity, there is a Uranium-235 1 atom ( 235 92 U ) and a neutron ( 0 n ); in the second step, the system comprises * just a single Uranium-236 atom ( 236 92 U ); and in the third step, the system comprises a Barium-144 atom, a Krypton-89 atom and three neutrons 89 1 ( 144 56 Ba  36 Kr  3 0 n ). Whereas the linguistic formalisms realize taxonomic relations and nuclear equations realize activities, the symbols in mathematics construe a third type of meaning modeled by Doran and Martin (2020) as properties. This accounts for the fact that unlike the elements in the linguistic formalisms or the arrows in nuclear equations, each symbol in mathematical equations is gradable and/or measurable. In pV = nRT, for example, we are able to have more or less volume or pressure (symbolized by V and p) and are able to measure these numerically as, say, V = 0.0224m3 or p = 1.013 × 105. In system networks on the other hand we are not able to have more or less extensive or intensive;3 neither can we have more or less of the symbols NPsing or -s in tree diagrams, or more or less → (glossed as ‘becomes’) in nuclear equations. Properties are gradable and potentially measurable meanings and are crucial components of many academic fields.4 Symbols are organized into mathematical statements through a set of relations such that if a change in the value of one symbol occurs, then this will affect all other symbols. For example in pV = nRT, assuming all else stays the same, if p increases, then V must decrease or the product of n, R and T must increase (or both). Doran and Martin (2020) refer to relations between properties as interdependency relations.5 Each mathematical statement sets up a vast network of interdependency relations that construe the possibilities of the physical systems they describe. The overview so far gives a sense of each formalism in terms of the main meanings they realize in field, as shown in Table 11.1. Importantly for our understanding of these formalisms, these meanings constitute the vast majority of the possible meanings each can make. For example, although mathematics can make very small distinctions in classification through subscripts on symbols (e.g. x1, x2, x3 etc.), this is minor

Academic Formalisms

339

Table 11.1 Field Meanings Realized by Academic Formalisms Formalism

Field meanings realized

System networks Tree diagrams Nuclear equations Mathematics

classification taxonomies composition taxonomies momented activities and composition taxonomies interdependencies between properties

compared to its ability to realize interdependencies; and algebraic mathematics has no way of construing composition or activity (Doran 2018a). Similarly, although tree diagrams construe composition, any construal of classification is relatively minor. Looking more broadly, although most of the formalisms have some small variations that may organize informational prominence (giving textual meaning), such as the vertical arrangement of system networks, the choice of the left or right side of the equation in mathematics and the ordering of particular symbols in nuclear equations, they do not appear to show any variation that can be considered ‘interpersonal’ in terms of organizing dialogue, evaluation or nuanced social relations of any sort; seen from a linguistic perspective, they do not display relatively independent variation that suggests meanings of mood (Halliday and Matthiessen 2014), nor do they suggest any appraisal or negotiation (Martin and White 2005), nor do they indicate variations in status or contact (Martin 1992) (see Doran 2018b for detailed argumentation concerning this in mathematics). All this is to say that each formalism is heavily devoted to a very specific set of field-relations. This explains the relatively mundane observation that we do not converse in mathematics, nor share our feelings with system networks. In general terms then, although the particular meanings they construe differ, each formalism shares this feature of construing only a very constrained set of field-specific meanings. In the next section we will show that this is not all they share. In addition to having a relatively constrained set of meanings they realize, these formalisms share the ability to iterate these meanings through a highly recursive grammatical organization.

The Grammatical Organization of Academic Formalisms So far we have seen that each formalism construes very particular fieldspecific meanings. This section will explore these formalisms in terms of their overall grammatical organization (building upon the final chapter of Doran 2018a). In particular, we will show that they all maintain a highly iterative grammar that enables an indefinite repetition of their field-specific meanings. As defined by Halliday (1965), univariate structures involve indefinite iteration of the same structural relation. In

340 Y.J. Doran English, for example, a typical univariate structure involves the complexing of groups or clauses where multiple elements of the same type are presented in sequence: (1) Dylan was the first to go, then (2) Frank left, and then (3) Josie went home.

Here, each of the clauses perform a similar function to the others in terms of their temporal unfolding, and importantly, an indefinite number of clauses can be added into this sequence. This is in contrast to a typical multivariate structure that involves distinct functions that can only occur once. The functional organization of English transitivity structures exemplifies a multivariate structure (Halliday and Matthiessen 2014): Twenty-two working-class northern clubs formed the Northern Rugby Football Union Actor Process Goal

Here, the Actor, Process and Goal each perform distinct functions in terms of their potential grammatical patterns and can only occur once (we cannot have two Processes in a single clause, two Actors or two Goals, for example). This is quite distinct from the patterns we will see in the formalisms that follow, where most options can occur an indefinite number of times. Starting with system networks, we will step through each formalism in turn, building up our picture of how they work. System Networks System networks are organized around sets of choices placed in systems, with the minimum network being a system with two choices as in Figure 11.2. Here the choices are shown through the square ‘or’ bracket ‘[‘. This indicates that in the system of mood either indicative or imperative is a possibility. As discussed in the previous section, from the perspective of field, each choice realizes an item in a classification taxonomy. Note that for this description, a choice includes both the feature, shown in lower case (indicative or imperative) and its realization statement—marked by ↘ and including functions (+ Subject; + Finite).6

Figure 11.2 Minimal System Network

Academic Formalisms

341

In such systems, each choice plays the same role; their vertical ordering higher or lower is not meaningful, ideationally speaking. Systems can be expanded to involve more than two choices, as in Figure 11.3:

Figure 11.3 System With Three Choices

In principle, systems can be expanded indefinitely to involve any number of choices. This opens the potential for networks to realize an indefinitely broad classification taxonomy with innumerable subtypes. Grammatically speaking, as each choice plays the same role and there can be any number of them, these systems involve univariate structure organized as a complex of choices.7 Although in principle systems may include any number of choices, in practice they tend toward only two. This enable generalizations to be captured at multiple levels of delicacy. Any further choices at a greater delicacy occur by adding systems to the right of each choice in the network, as in Figure 11.4.

Figure 11.4 System Network With Two Levels of Delicacy

342 Y.J. Doran Here, each of the choices of indicative and imperative have two subtypes: declarative and interrogative for indicative, and jussive and optative for imperative. Networks can specify any degree of delicacy, meaning there is potential for indefinite iteration along this dimension of the formalism (alongside the potential for an indefinite number of features in systems noted earlier). This means there may be an indefinite number of layers of systems, with each system potentially including an indefinite number of features. The indefinite expansion of choices within a system and the ability for indefinite layering through increased delicacy offer two paths for iteration in the grammar of system networks. From the perspective of field, this means that system networks enable the realization of both extra breadth in taxonomies (indefinite numbers of items in a single layer of a classification taxonomy—i.e. an indefinite number of co-classes) and extra depth (an indefinite number of levels in a classification taxonomy— i.e. extra iterations of subclasses within subclasses). There is one further means for complexing that dramatically increases the degree of field-specific meaning a network can realize. This is the possibility for simultaneous systems where multiple systems can be crossclassified. This is shown in system networks through a curly bracket ‘{‘, that indicates an ‘and’ relation. In Figure 11.5, the network indicates that if a major clause is chosen, then one must choose from each of mood and theme and transitivity. Thus a clause will take choices from each of these systems. As with the previous expansions, there can be an indefinite number of simultaneous systems. In terms of field, the use of simultaneous systems enables multiple classification taxonomies to be realized at once. Grammatically speaking, this again involves a univariate structure; there may be any number of simultaneous systems and each performs the same broad ideational function. In Figure 11.5 these taxonomies are all relatively independent: each choice in theme can occur with each choice in transitivity with each choice in mood. But system networks also enable more specific interdependencies to be shown, as in Figure 11.6 (from further on in Halliday 1967). This system network sets up multiple, interdependent classification taxonomies. At its most general, it sets up a single taxonomy of extensive vs. intensive. But within extensive, it realizes two distinct taxonomies: one with two choices, effective and descriptive, and another with three choices: operative, middle and receptive. At a more delicate level, it specifies much more intricate interdependencies. For example the intersection of both effective from one taxonomy and operative from another, opens up a third taxonomy with two types (goal-transitive and goalintransitive), while the intersection of effective and receptive allows for another taxonomy that distinguishes two further types (agent-oriented

Figure 11.5 Simultaneous Systems

Figure 11.6 System Network Realizing Multiple Interdependent Classification Taxonomies

344 Y.J. Doran and process-oriented). System networks enable a precise realization of these classification taxonomies in one synoptic snapshot. And its grammar enables there to be any number of these taxonomies that may be indefinitely deep or indefinitely wide (for good examples of large networks along these lines, see Halliday and Matthiessen 2014, especially Fig.4–13, and Matthiessen 1995). The functionality of system networks is summarized in Table 11.2. Table 11.2 Grammatical Organization and Field Relations in System Networks Univariate structures in the grammar

Field-specific meanings realized

choices in a system systems ordered in delicacy (layers of univariate structure) simultaneous systems

breadth of a classification taxonomy depth of a classification taxonomy number of classification taxonomies

Tree Diagrams Like system networks, tree diagrams are organized around a relatively small number of elements. However unlike system networks, tree diagrams have been taken up by a wide range of different schools of linguistics. This means there is not just one type of tree diagram; what is possible and what regularly occurs in this formalism varies across the field. For this reason, here we will focus only on general features that associated with basic phrase structure grammars (e.g. Chomsky 1957). Constituency tree diagrams are organized around branches that emanate from nodes. A basic tree will involve a node with two or more branches, as in Figure 11.7:

Figure 11.7 Basic Tree Diagram

This tree diagram is read as saying an NP is composed of a Det and an N. The Det is substituted lexically by the and the N is substituted by ball. As noted prior, the relationship between the NP and the Det and N is one of composition. Like system networks, this relationship can be repeated

Academic Formalisms

345

indefinitely for each branch.8 In Figure 11.8, for example, there are six levels of the composition taxonomy realized by a series of branches from S to the Det and N of the park.9

Figure 11.8 Tree Diagram With Increased Depth

Iterating branches down the page realizes increasing levels of depth in the composition taxonomy. In the previous examples the trees set up binary branches where no more than two branches occurred for each node (e.g. an NP split into two branches, a Det and an N). In some schools of linguistics, there can also be as many branches as needed for any level in the constituency hierarchy. That is, one can also expand indefinitely the breadth of compositional taxonomies, such as in Figure 11.9.10 Tree diagrams are thus potentially organized around a univariate structure on two dimensions. First, they enable an indefinitely iterative number of levels in hierarchy, developing depth in the composition taxonomy. Second, depending on the sub-field, they enable an indefinite number of branches to occur in each level, allowing increasing breadth of the taxonomy. These two possibilities for expansion are summarized in Table 11.3.11

346 Y.J. Doran

Figure 11.9 Tree Diagram With Increased Breadth

Table 11.3 Grammatical Organization and Field Relations in Tree Diagrams Univariate structures in the grammar

Field-specific meanings realized

levels in the tree branches in each level

depth of a classification taxonomy breadth of a classification taxonomy

Mathematical Symbolism Moving from linguistic formalism to mathematical symbolism, there is a shift from construing taxonomic relations between items to establishing interdependencies between measurable properties. Like system networks and tree diagrams, mathematical symbolism can establish innumerable relations between properties through a highly univariate grammar at multiple levels (the overview given here is based upon the more detailed description given in Doran 2018a). Here we will focus on mathematical symbolism most closely associated with elementary algebra, often used in high school science (see Doran 2018a for details). A basic mathematical statement involves two expressions linked by a Relator. In equation (2) which describes kinetic energy, the energy of motion, the two expressions are KE and 1 mv 2 and the 2 Relator is = (2) KE

1 mv 2 2

Academic Formalisms

347

Ideationally speaking, the two expressions perform the same role. This can be shown by the fact that they can be swapped in order with only a change in the information organization (i.e. ideationally, (2) and (3) are the same): (3)

1 mv 2 2

KE

This basic mathematical statement can be expanded to include more expressions and Relators, such as the following equations that have three and four respectively: (4) E1 (5) O

E2 v f

k 3 u 108 104.1 u 106

2 .9 m

In principle, statements can include any number of expressions linked by Relators and can generally be rearranged in any order without affecting their ideational meaning.12 As there is the potential for indefinite expansion and each expression performs the same ideational function, we can again describe this area of mathematics’ grammar as a univariate structure. Although in principle there can be an indefinite number of expressions in a statement, for textual reasons, there is a strong tendency to have only two, with three or four occurring at times and any more being unusual. To allow for further expansion of relations between symbols, then, there is a second avenue for indefinite iteration involving the number of symbols within an expression. As the left-hand side of equations (2), (4) and (5) show, it is common for one of the expressions to include only a single symbol. But in other expressions in these equations, there are many more. Every symbol that is added must necessarily be linked by an arithmetic operator such as + (addition), − (subtraction), ÷ (division) and × (multiv plication).13 For example, the second expression in equation (5) ( ) has f 1 two, while the second expression in (2) ( mv 2 ) has five, and in the fol2 lowing equation, there are 16 across three expressions: (6) r

4SH 0

n2 =2 mZe 2

n2 a0 Z

Like expressions, symbols can also be iterated indefinitely. Indeed it is here that most of the expansion occurs.

348 Y.J. Doran We can thus again consider this type of formalism to be organized around a univariate structure. Like system networks and tree diagrams, mathematics offers multiple avenues for indefinite iteration. In this case, both levels of iteration realize the same field relation: interdependencies between itemized properties. Table 11.4 summarizes these possibilities for mathematical symbolism.

Table 11.4 Grammatical Organization and Field Relations in Mathematical Symbolism Univariate structures in grammar

Field-specific meanings realized

expressions within statements symbols within expressions

interdependencies between itemized properties

Nuclear Equations As with the other three formalisms, nuclear equations are also organized through multiple components with a univariate structure. Like system networks and tree diagrams, but unlike mathematics, the two univariate components organize different meanings at field. However whereas each component in system networks and tree diagrams organize slightly different facets of the same broad area of field—the depth, breadth and number of classification taxonomies in system networks, and the depth and breadth of composition taxonomies in tree diagrams—the two avenues for expansion in nuclear equations organize quite different field relations: the momenting of activities and the breadth of composition taxonomies. At the highest level, a basic nuclear equation is organized through two expressions on either side of an arrow →. In the following equation the 14 209 two expressions are 223 88 Ra (signifying Radium-223) and 6 C  82 Pb (Carbon-14 plus Lead-209): (7)

223 88 Ra

o

14 209 6 C  82 Pb

As we have already seen, nuclear equations can have more than two expressions, with each new expression being linked by another →, as in the three expressions in: (8)

235 1 92 U  0 n

o

236 92

U* o

140 54

Xe 

94 38 Sr 

2 01 n

Just like each of the formalisms we have seen so far, the number of expressions can be repeated indefinitely. A well-known decay chain, known as the actinium chain, for example, can be represented as follows:14

(9)

­ 98.62%E  227 ½ D  o Th  o° ° . . d 21 773 y 18 718 235 231 231 227 ° 90 219 ° 223 D E D D U  o Th  o Pa  o Ac ® Ra  o Rn ¾ 11.434 d 25.52 h 7.308u108 y 3.276u104 y  92 90 91 89 223 8 8 86 1.38%D E °  ° o Fr  o 21.8 min °¯ 21.773 y 87 °¿ ­ 99.73%D 207 ½ E  o Tl  o° ° 2 . 13 min 4 . 77 min 81 219 215 211 211 ° ° 207 D D E Rn  o Po  o Pb  o Bi ® Pb (stable) ¾ 3 . 96 s 1 . 778 ms 36 . 1 min 83 ° 0.27%E  211 86 84 82 82 D °  o Po  o 0.516 s °¯ 2.13 min °¿ 84



350 Y.J. Doran As discussed earlier, by expanding nuclear equations in this way, the text is able to indefinitely moment the activities of nuclear reaction and decay. The discipline is able to specify in as much or as little detail the steps between two states. The second avenue for expansion occurs within each expression and has to do with what particles occur at each stage of the activity. In the following equation, each state is represented by a set of symbols that rep1 resent particles. The first involves two symbols 235 92 U and 0 n , the second 236 * involves just a single symbol 92 U and the third involves three symbols 140 94 1 54 Xe  38 Sr  2 0 n . (10)

235 1 92 U  0 n

o

236 92

U* o

140 54

Xe 

94 38 Sr 

2 01 n

The number of symbols in each expression can again in principle be indefinitely expanded. And each perform the same function in the sense that they can be rearranged in any order without change to their ideational meanings (only textual meanings). In this sense, this provides a second avenue for univariate expansion. In this case, the expansion does not moment an activity, but rather expands the breadth of the compositional taxonomy at this point of the activity. It is important to note that unlike tree diagrams, nuclear equations cannot expand in depth; they cannot iterate further parts of parts of parts. Any depth is specified precisely by the numbers to the side of the symbols, such that 235 92 U indicates the Uranium involves 235 nucleons and 92 protons. The nuclear equation cannot specify what constitutes these nucleons or protons (i.e. quarks), nor can it group different sets of protons or nucleons together in other levels of a compositional taxonomy. The only iterative expansion available here is an expansion in the breadth of the compositional taxonomy at a single level.15 Finally, there is one further small dimension of iteration that occurs in nuclear equations. If we look again at equation (9) reproduced here, the expression following 227 89 Ac (Actinium-227) provides two alternative pathways for the decay. It may decay to 227 90 Th (Thorium-227) or 211 to 233 84 Bi (Bismuth-211), 87 Fr (Francium-233). Similarly, following 207 two alternative pathways are given via 81 Tl (Thallium-207) or 211 84 Po (Polonium-211). These alternatives are shown through the brace { and their vertical arrangement. From the perspective of field, these realize alternative activities construing the particular pathways of the decay. In principle multiple alternatives can be shown, however in practice, this is restricted to the specific possibilities for decay that any particular isotope has, and so any more than two is unusual. Nonetheless, the formalism enables a realization of multiple parallel activities (similar to system networks enabling multiple classification taxonomies).

(9)

­ 98.62%E  227 ½ D  o Th  o° ° . . d 21 773 y 18 718 227 ° 90 235 231 231 219 ° 223 D E D D U  o Th  o Pa  o Ac ® Ra  o Rn ¾ 11.434 d 25.52 h 7.308u108 y 3.276u104 y  89 223 8 92 90 91 8 86 1.38%D E °  ° o Fr  o 21.8 min °¯ 21.773 y 87 °¿ ­ 99.73%D 207 ½ E  o Tl  o° ° 2 . 13 min 4 . 77 min 81 219 215 211 211 ° ° 207 D D E Pb (stable) Rn  o Po  o Pb  o Bi ® ¾ 3 . 96 s 1 . 778 ms 36 . 1 min 83 ° 0.27%E  211 86 84 82 82 D °  o Po  o 0.516 s °¯ 2.13 min °¿ 84



352 Y.J. Doran The possibilities for nuclear equations are summarized in Table 11.5. Table 11.5 Grammatical Organization and Field Relations in Nuclear Equations Univariate structures in the grammar

Field-specific meanings realized

expressions in the equation alternative expressions symbols in an expression

momenting of activities number of activities breadth of a compositional taxonomy

Field and Grammar of Academic Formalisms This overview has shown that each resource construes a small set of specific relations in field, but that their grammar has evolved to iterate these meanings indefinitely. This enables the functionality of each resource for the particular technical knowledge that their respective discipline needs, while at the same time enabling the expansion and integration of this technical knowledge into a coherent whole. As far as our concern with semiotic typology is concerned, this overview opens a series of questions that could be used to organize a more elaborated typology in this area (paralleling the questions presented in Martin and Quiroz in this volume). From this perspective we may ask of any formalism: •

Does the formalism involve iterative (univariate) structures? •

• •

If so, how many?

Do distinct iterative structures realize the same meanings at field? Or do they realize different meanings? What meanings do these iterative structures organize? •

Taxonomy? • • •

•

Activity? • •

•

Only one or many? Composition or classification? Breadth (co-type/co-part) or depth (type-subtype/whole/part)?

Only one or many? Do they moment these activities or not?

Property? • •

Only one or many? Do they realize independencies between these properties?

Such questions are foundational for the development of a typology of formalisms, and, more broadly, a typology of semiotic resources in general—

Academic Formalisms

353

by providing a principled set of possibilities that go beyond the surface features of resources (such as whether they are images or symbols, language-like or not etc.).

Non-Iterative Elements This chapter has primarily concerned itself with the iterative structures that form the core of academic formalisms. Each resource also includes a small component of elements that are not iterative but realize crucial field-specific meanings for their discipline. System networks, for example, include realization statements that are linked to individual choices, such as: optative ↘ + Subject; Subject:: let’s Here, the choice optative leads to the insertion of a Subject and the lexicalization of this Subject as let’s. This example is best read as realizing a small relation of composition, formalizing as it does that optative clauses are composed of (amongst other things) a Subject lexicalized as let’s. However these realization rules are not iterative in the sense that one cannot specify further depth in composition within the Subject through multiple realization rules for a single feature. That is, in the optative example earlier, one could not put another arrow ↘ below the realization rule to put forward another element within the Subject or let’s and then indefinitely repeat this down to the morpheme. To do this would require another set of system networks (Martin 2013).16 Similarly, as we’ve seen, tree diagrams include a non-iterative component whereby the final node is replaced by a lexical item. In the diagram here, the N is replaced with ball and the Det is replaced with the. However one cannot insert another lexical item once ball or the has been inserted.

Figure 11.10 Lexical Substitution in Tree Diagrams

354 Y.J. Doran The same applies for mathematics and nuclear equations. Both include small components that construe meaning that cannot be repeated. In the case of mathematics, small modifications of symbols can indicate distinctions in classification. In the following example taken from a high school physics classroom, the subscripts indicate three types of E (energy): Eemitted (energy emitted), Ei (initial energy) and Ef (final energy): (11) Eemitted

Ei – Ef

Again, these subscripts cannot be iterated (a symbol such as Eemittedi f2 cannot occur). In the case of nuclear equations, the numbers to the left of symbols, such as 227 90 Th indicate a small compositional taxonomy. We will step this through using 2 01 n , the final symbol in the equation 235 1 92 U  0 n

o meanings are: • •

•

236 92

U* o

140 54

Xe 

94 38 Sr 

2 01 n. The three numbers and their

The subscript 0, known as the atomic number, gives the number of protons in the particle (in this case zero). The superscript 1, known as the mass number, gives the total number of nucleons in the particle (which group together the protons and neutrons) (in this case one). The number on the left 2, gives the total number of these particles (in this case two).

This establishes a small composition taxonomy whereby the group of particles include a particular number of nucleons, which in turn comprises a particular number of protons. But this cannot be iterated; the symbol cannot have superscripts on superscripts or subscripts on subscripts that indicate what is in the protons. So, like the other formalisms, nuclear equations complement their iterative components with small non-iterative components that realize distinct field-specific meanings. From the perspective of developing a semiotic typology, this allows us to ask of any formalism: • • •

Does the resource couple iterative structures with non-iterative structures? If so, what type of structures are used and how many are there? What meanings are they being used for? Taxonomy? Activity? Property?

Semiotic Typology The questions raised in the previous sections by no means exhaust the possibilities for variation across semiosis. They deal in particular with academic formalisms and the meanings they organize for technical

Academic Formalisms

355

disciplines. One obvious question not asked previously is why some formalisms seem to use an imagic mode (such as system networks and tree diagrams) and others use a more symbolic mode (mathematics, nuclear equations). For reasons of space, we cannot explore this question here; however it does flag work to be done to understand the range of meaningmaking systems used in social life. Recent decades of research in Social Semiotics and Systemic Functional Semiotics have dramatically pushed our understanding of semiotic resources to the point where we can now begin to compare and contrast them and develop a typology. For this to be successful, we need a principled means of comparison that allows us to see both the similarities and the differences and does not assume that certain categories such as metafunction, rank etc. will necessarily occur as they do for language.

Note 1. Although this is a strong tendency, a clear exception to this is the formal logic used in some branches of philosophy. 2. The tree diagram does make some small distinctions in classification through subscripts on the symbols, such as Nsing and Nplur distinguishing between singular and plural nouns, however importantly, this cannot be repeated indefinitely in the way the branches showing composition can. 3. In some system networks, descriptions have involved clined systems that show gradations between ‘more’ or ‘less’ of a feature, such as van Leeuwen’s (2009) ‘parametric’ systems for voice quality, which would here be described in terms of property. In these instances, the network is no longer realizing a classification taxonomy. There have also been attempts at complementing the typologies of system networks with topologies, generally shown through cartesian planes (e.g. Martin and Matthiessen 1991 for various areas of English lexicogrammar, and Martin and Rose 2008 for genre). However in general, most system networks conform to the description in this chapter. 4. More specifically, mathematical symbols are instances of itemized properties (Doran and Martin 2020; Doran 2020). Properties in general describe elements that are gradable, such as warm : warmer; big : bigger. Itemized properties are also gradable but are construed as items in and of themselves, either through grammatical metaphor such as warmth, or technicalization of a dimension such as temperature (Hao 2020a, b). For example, warm is a property and can be graded, but when discussed as temperature, it is an itemized property that can be both graded and classified into absolute temperature, relative temperature etc. This distinction does not affect the argument in this chapter. 5. In Doran (2018a) these relations were called implication complexes and were described as being realized by co-variate structures. With the revision of the model of field in Doran and Martin (2020), this is now better described in terms of interdependency relations. 6. For reasons of space, we will not explore realization statements in detail here. See the final chapter of Doran (2018a) for a more in-depth discussion. 7. It must be stressed that here we are talking about the structure of the system network formalism itself, not the component of language the system network describes (which can be either multivariate or univariate). 8. In contrast, the relation between Det and the, and N and ball, known as lexical substitution, is not one of composition; the is not part of Det and ball

356 Y.J. Doran

9.

10. 11.

12.

13. 14. 15.

16.

is not part of N here. Rather, it is better described in terms of an elaboration, where the lexical item (ball/the) and its grammatical category (N/Det) are equated. Importantly for our description, this distinction between lexical substitution and the composition relations of the tree is paralleled by the fact that whereas the compositional relations can be repeated, lexical substitution cannot. In general, once a branch has had a lexical item substituted (e.g. N has become ball), one cannot substitute another lexical item for ball and then another and then another indefinitely. For examples of how many iterations constituency trees can show, see those used in the early years of generative semantics, especially Ross and Lakoff (such as one reproduced in Harris (1993: 144) with no fewer than 23 tiers for the sentence, Floyd broke the glass). For examples of trees that significantly iterate the number of branches in each level—so emphasizing breadth over depth—see Appendix B of Fawcett (2000). The order of the lexical items at the bottom of the diagram also reflect, to a certain degree, their ordering in language. In this sense, the tree diagrams realize what Doran and Martin (2020) call a spatio-temporal property; they specify the location of each lexical item in the sentence. Utilizing both the vertical and horizontal axes enables the diagram to show these multiple fieldspecific meanings, depth and breadth of composition and spatio-temporal properties. Indeed this is one of the main affordances of these types of images over phrase structure rules. However a discussion of the functionality of imagic vs. symbolic formalisms is beyond the scope of this chapter. This analysis is slightly complicated by inequations that use Relators such as > (larger than), < (smaller than), ≥ (larger than or equal) and ≤ (smaller than or equal to). These can also be rearranged without affecting their ideational meanings, however it requires a small alternation in the direction of the Relator. That is y > 7 construes the same ideational meanings as 7 < y. See Doran (2018a: 80–88) for more detailed argumentation. In the case of multiplication, the × is typically elided, so that an expression 1 1 mv 2 means u m u v 2 . 2 2 This particular nuclear equation is taken from the Wikipedia page for Uranium 235. https://en.wikipedia.org/wiki/Uranium-235 Accessed 11/02/19 Similarly, unlike tree diagrams, the order of the symbols does not specify any 94 1 spatial ordering or location; 140 54 Xe  38 Sr  2 0 n does not mean that there is spatial layout that orders Xenon-140 and then Strontium-94 then two neutrons. The expression simply notes that these particles comprise the system at that point. Though see Bateman (2008) for a suggestion that SFL consider using tree fragments as realization rules, which would enable an iteration of composition along these lines—essentially combining system networks with tree diagrams.

References Bateman, J. A. (2008) Systemic functional linguistics and the notion of linguistic structure: Unanswered questions, new possibilities. In J. Webster (ed), Meaning in context: Implementing intelligent applications of language studies. London: Continuum International Publishing Group. 24–58. Chomsky, N. (1957) Syntactic structures. The Hague: Mouton. Doran, Y. J. (2017) The role of mathematics in physics: Building knowledge and describing the empirical world. Onomázein. Special Issue of Systemic Functional Linguistics and Legitimation Code Theory. 210–226.

Academic Formalisms

357

Doran, Y. J. (2018a) The discourse of physics: Language, mathematics and image. London: Routledge. Doran, Y. J. (2018b) Intrinsic functionality of mathematics, metafunctions in systemic functional semiotics. Semiotica. 225. 457–487. Doran, Y. J. (2020) Multimodal knowledge: Language, mathematics and images in physics. In K. Maton, J. R. Martin and Y. J. Doran (eds), Studying science: Knowledge, language and pedagogy. London: Routledge. Doran, Y. J. and Martin, J. R. (2020) Field relations: Understanding scientific explanation. In K. Maton, J. R. Martin and Y. J. Doran (eds), Studying science: Knowledge, language, pedagogy. London: Routledge. Fawcett, R. (2000) A theory of syntax for systemic functional linguistics. Amsterdam: John Benjamins. Halliday, M. A. K. (1965) Types of structure. In J. R. Martin and Y. J. Doran (eds) (2015), Grammatics. Volume 1 in systemic functional linguistics: Critical concepts in linguistics. London: Routledge. 227–241. Halliday, M. A. K. (1967) Notes on transitivity and theme in English: Part 1. Journal of Linguistics. 3:1. Halliday, M. A. K. and Matthiessen, C. M. I. M. (2014) Halliday’s introduction to functional grammar. 4th ed. London: Routledge. Hao, J. (2020a) Analysing scientific discourse from a systemic functional perspective: A framework for exploring knowledge building in biology. London: Routledge. Hao, J. (2020b) Building taxonomies: A discourse semantic model of entities and dimensionality in biology. In K. Maton, J. R. Martin and Y. J. Doran (eds), Studying science: Knowledge, language and pedagogy. London: Routledge. Harris, R. A. (1993) The linguistics wars. Oxford: Oxford University Press. Lemke, J. L. (1998) Multiplying meaning: Visual and verbal semiotics in scientific text. In J. R. Martin and R. Veel (eds), Reading science: Critical and functional perspectives on discourses of science. London: Routledge. Lemke, J. L. (2003) Mathematics in the middle: Measure, picture, gesture, sign, word. In M. Anderson, A. Saenz-Ludlow, S. Zellweger and V. Cifarelli (eds), Educational perspectives on mathematics as semiosis: From thinking to interpreting to knowing. Ottawa: Legas Publishing. 215–234. Martin, J. R. (1992) English text: System and structure. Amsterdam: John Benjamins. Martin, J. R. (1993) Life as a noun: Arresting the universe in science and the humanities. In M. A. K. Halliday and J. R. Martin (eds), Writing science: Literacy and discursive power. London: The Falmer Press. 242–306. Martin, J. R. (2013) Systemic functional grammar: A next step into the theory: Axial relations. [Chinese translation and extensions by Wang Pin and Zhu Yongsheng]. Beijing: Higher Education Press. Martin, J. R., Maton, K. and Doran, Y. J. (eds) (2019) Academic discourse: Systemic functional linguistics and legitimation code theory. London: Routledge. Martin, J. R. and White, P. R. R. (2005). Language of evaluation: Appraisal systems in English. London: Palgrave. Maton, K. (2014) Knowledge and knowers: Toward a realist sociology of education. London: Routledge. Maton, K. and Doran, Y. J. (2017) SFL and code theory. In T. Bartlett and G. O’Grady (eds), The Routedge handbook of systemic functional linguistics. London: Routledge. 605–618. Maton, K., Martin, J. R. and Doran, Y. J. (eds) (2020) Studying science: Knowledge, language, pedagogy. London: Routledge.

358 Y.J. Doran Maton, K., Martin, J. R. and Matruglio, E. (2016) LCT and systemic functional linguistics: Enacting complementary theories for explanatory power. In K. Maton, S. Hood and S. Shay (eds), Knowledge-building: Educational studies in legitimation code theory. London: Routledge. 93–114. Matthiessen, C. M. I. M. (1995) Lexicogrammatical cartography: English systems. Tokyo: International Language Sciences Publishers. O’Halloran, K. L. (2005) Mathematical discourse: Language, symbolism and visual images. London: Continuum International Publishing Group. Parodi, G. (2012) University genres and multisemiotic features: Accessing specialized knowledge through disciplinarity. Fórum Linguístico. 9:4. 259–282. Rose, D. and Martin, J. R. (2012). Learning to write/reading to learn: Genre, knowledge and pedagogy in the Sydney school. London: Equinox. Young, H. D. and Freedman, R. A. (2012) Sears and Zemansky’s university physics with modern physics. 13th ed. San Francisco: Addison Wesley.

Index

activity 16–17, 190, 198, 204, 238–242, 244–246, 253, 262–268, 277, 302, 352; expectancy activity 241; implication activity 239, 241; momented activity 241, 244, 277; unmomented activity 241 agnation 5–7, 106–109, 111–112 Anterior 244–248, 255, 263–267 appliability 25 appraisal 12, 198, 201, 203 aspect 49–52, 65–67 axis 10, 189, 192, 195, 229, 275; axial argumentation 35, 129; axial functionality 22 bilingual education 307 Chinese 209, 213, 258, 262, 268–272 class 1–5 Classical Tibetan 73–101 clause complex 194–195, 200, 211, 213, 224, 249, 252, 254, 258, 261, 267–269 code-switching 307, 317–320 Comment 143, 163–165, 169, 173–175 communicative dynamism 162–164 complexing 239, 249–250, 255, 258, 265 conjunction see connexion conjunction 231, 243–245, 249, 258–261, 268 connexion 12, 198, 200, 222, 224, 231, 250, 239, 242–249, 263–268 connotative semiotic 275 context 14, 23, 190, 196, 228–234, 239, 241, 268; see also genre; register contextualization 134, 136–137

co-text 197, 229 co-textual functionality see discourse semantics covert category see cryptotype cryptogrammar 5–7, 113–124 cryptotype 103–104, 112–113 curriculum genre 308–309 Daneš 162, 182 declarative 91–96 denotative semiotic 275 Detailed Reading 309–310; L1 based 320–323; L2 based 323–325 directionality 148, 155–156 discourse semantics 10–14, 23 downranking 239 educational linguistics 335 EFL 327 element 199 embedded 249–258, 263–269 enabling 130, 135, 142, 145–147, 153, 158 enation 5, 106 engagement 12 entity 190–196, 199, 202, 204, 205, 209, 213, 222, 226–231, 242, 243 ergative 74–77 exchange 36–38, 80–81, 313–317 exemplum 14–16 expectancy 299–301 experiential meaning 39–41, 74–77, 102–124 fact 117, 119 field 16–17, 190, 195–196, 205–206, 213, 222, 224, 229–230, 238–242, 268–269, 275, 276–277, 278–279, 336, 337

360

Index

figure 39, 117, 198–201, 241–248, 255, 263–268 figure sequence 286, 292, 298 Firbas 162–163, 182 Firth 273 focalization 302 formalism 331, 334; and field 337–339, 352 Fries 166, 177, 182 function 2–6 functionality 189–190, 196, 229 functional language typology see language typology Functional Sentence Perspective 162, 185 functional structure 162, 164 genre 14–18, 274, 275 Given-New 162, 165, 169, 172 Gleason 5, 106–109, 196, 232 grammatical metaphor 11–12 graphology 10, 13 Halliday 8–9, 11, 104 hyperNew 131, 134–135 hyperTheme 130, 133–135, 140, 147–148 hyperwave 133–134, 141, 147 hypotaxis 201, 249, 258–262, 265, 267 idea 117, 120 ideation 12, 39, 198–199, 222–225 identification 12, 194, 196, 198, 201, 204, 207, 213, 228–233 imperative 83–88 individuation 273, 274 Indonesian 308 instantiation 189–190, 273, 274–276 Interactional Linguistics 166–167 interdependency 190, 249–250, 257 interrogative 88–91 iteration 336, 350; non-iterative elements 353–354 Joint Re-writing 309–310; L2 based 325–327 Khorchin Mongolian 35–72 knower code 335 knowledge code 335

321–322; intra-move 320, 325–327, 328–329 language typology 189–191, 193, 195–196, 229, 232, 234, 235, 268, 273, 274, 336; semiotic typology 331 langue 273, 304 learning exchange 310–313 Legitimation Code Theory (LCT) 335 logical meaning 203, 239, 242–249, 258–263, 267–268; logical relations 239, 240, 263, 268 macroNew 131, 135, 139, 147, 148 macrophenomena 117 macroTheme 130, 131, 133, 135, 138, 147, 148, 158 macrowave 133, 134–135, 147 Malinowski 279 marked focus 164, 177 mathematical symbolism 332, 338, 346–348 Mathesius 162–165, 171, 185 mental clause 113–124 message 129; continuing 135; culminative 135; discontinuing 140; initial 135 metafunction 7–10, 22–23, 74–78, 190, 196–198, 206, 229 metaphenomena 117–119, 121–123 metaredundancy 13 methodology 22–23 modality 11–12, 52–56; metaphor 11; modulation 53–54; probability 11 modal verb 239, 245, 264–268 mode 17–18, 275 mood 8, 36, 80–98, 274, 241–243 move 313–317 multimodal semiotics 18, 24, 335–336 multilinguality 24 negotiation 12, 24, 80–81, 190, 198, 201–202, 228–229, 310, 313–317 New 131, 160, 163–165, 169, 172, 176–177, 183–186, 206, 232–235, 270 nuclear equation 332–333, 337–338, 348–352 nuclear relations 286, 292 overt category see phenotype

langage 273, 304 language shift 318–320, 329; intermove 319, 323–325; inter-role 318,

paradigmatic axis 6, 10 parataxis 249, 258–265, 267

Index parole 273, 304 particle 79–80 Peirce 165, 186 periodicity 12, 130–134, 137, 148, 172, 196–198, 204–207, 212–213, 228–231, 284–285, 290–291, 297 person 60–65 perspective 286–287, 292, 298, 301–302 phase 277, 280–282, 286, 292, 298, 299 phenotype 5, 104 phonology 10, 13 physics 331–332 Pike 129–130, 161 polarity 42–45 Posterior 244–248, 255, 263–267 Prague School 162–164, 173 preselection 130, 136, 157 process type 112–113, 333–334 property 352 quality 199 rank 2, 274 reactance 6, 105 reading position 299 Reading to Learn 307–310 realization 275 realization statement 7 register 14, 16–18, 274 Rheme 139–140, 142, 144, 164–165, 168, 172, 174–175 Role and Reference Theory 163 Saussure 273 Schegloff 167, 179, 186 scientific knowledge 238, 239 sentence focus 164, 174, 176, 184 signology 10, 13 social semiotics 18; see also multimodal semiotics Spanish 113–124, 191–196, 207–214, 222–237, 271 specialized knowledge 238 story 274, 277–278; Chinese story 282–287; Oromo story 284–299;

361

Pitjantjatjara story 287–294; story phases 280–282 stratification 10–18, 194–195, 240, 273, 274–276 structure 3–7, 189, 190, 194, 226, 228, 230, 232–233, 249–250, 257 syntagm 2–6, 189, 231 system see paradigmatic axis system network 6, 9, 334, 337, 340–344 taxonomic relations 190, 241 taxonomy 337, 342; classification taxonomy 17, 337, 338, 339, 340–345, 352; composition taxonomy 337, 338, 339, 348–350, 352 teaching/learning cycle 309, 310–312 tenor 17–18, 275, 339 tense 45–49; relative tense 56–59 text reference 10–11 textual metafunction 77–78 theme 9, 274, 342 Theme 77–78, 129, 162–165, 284–285, 290–291, 297; interpersonal 142; mood orientation 145; selection 144; textual 141; transitive orientation 147 Topic 162–165, 169–177, 180–181, 183, 185 transitive 74–77 transitivity 9, 40, 274, 342 tree diagram 2, 4, 334, 337, 344–346 trinocular perspective 35 typology see language typology univariate structure 339–340, 345, 348, 352 unmarked focus 164 valeur 191 verbal group 36–72 wave 129–135, 139, 145, 172, 190, 194, 204–207, 217–218, 222, 225–227 Whorf 5–6, 103–105