105 14 4MB
English Pages 186 [177] Year 2023
Peking University Linguistics Research 7
Jian Hu
A Constructional Approach to Interpersonal Metaphor of Modality
Peking University Linguistics Research Volume 7
Series Editors Rui Guo, Chinese Department, Peking University, Beijing, China Wangqi Jiang, School of Foreign Languages, Peking University, Beijing, China
Peking University Linguistics Research (PKULR) is a cooperation project between Springer Nature and Peking University Press. This series presents the latest discoveries and developments of significance in linguistic research conducted by famous Chinese scholars. Titles in this series are carefully evaluated, examined and selected by Peking University (which ranks No. 10 in the world and No. 1 in China in the QS World University Rankings-Linguistics 2016) and Peking University Press (which was honored as the most influential publisher in linguistics according to Chinese Book Citation Index, 2016), covering all major aspects of linguistics— phonetics, phonology, pragmatics, semantics, morphology, syntax, theoretical linguistics, applied linguistics and inter-disciplinary studies. PKULR aims to provide an invaluable guide to the very nature of language. On the one hand, it tries to offer a thorough grounding in the fundamental concepts of linguistics; on the other hand, it also attaches great importance to the practical application of these concepts, esp. in Chinese context.
Jian Hu
A Constructional Approach to Interpersonal Metaphor of Modality
Jian Hu School of Foreign Languages Beihang University Beijing, China
This book has been funded by Beijing Social Science Fund, project no. 16YYB012, at a later stage of the research. ISSN 2662-3129 ISSN 2662-3137 (electronic) Peking University Linguistics Research ISBN 978-981-99-2763-0 ISBN 978-981-99-2764-7 (eBook) https://doi.org/10.1007/978-981-99-2764-7 Jointly published with Peking University Press The print edition is not for sale in China (Mainland). Customers from China (Mainland) please order the print book from: Peking University Press. Translation from the Chinese language edition: “情态隐喻的认知构式研究” by Jian Hu, © Peking University Press 2020. Published by Peking University Press. All Rights Reserved. © Peking University Press 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publishers, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publishers nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publishers remain neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore
Acknowledgments
I am grateful to Prof. Mark Turner of CWRU and Prof. Wang Zhenya of BLCU for providing me the knowledge and challenges necessary to work on this project. I am also very thankful to Prof. Wei Naixin of Beihang University for precious encouragement and support. Gratitude also goes to Prof. Vera Tobin, Prof. Li Fuyin, Prof. Wen Jun, Prof. Feng Zengxin, Prof. Gao Mingle, Prof. Dong Qiming, and Prof. Liu Linjun for readings. Finally, I would like to thank my parents and above all to my wonderful family, Mr. Guo Mi and Miss Guo Ruoxi, for their understanding and support over the years.
v
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Form-Meaning Mapping and Grammatical Metaphor . . . . . . . . . . . . 1.2 Defining Interpersonal Metaphor of Modality . . . . . . . . . . . . . . . . . . . 1.3 Goals of This Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Theoretical Orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.5 Methodology Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.6 Outline of the Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1 1 2 3 3 4 5 5
2 GM Theory and Interpersonal Metaphor of Modality . . . . . . . . . . . . . . 2.1 The Theory of Grammatical Metaphor . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1 Clarification of Key Notions . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.2 The Development of GM Theory . . . . . . . . . . . . . . . . . . . . . . . 2.1.3 Ideational Metaphor and Interpersonal Metaphor . . . . . . . . . 2.1.4 The Current State of GM Studies . . . . . . . . . . . . . . . . . . . . . . . 2.2 The Systemic Functional Approach to Modality . . . . . . . . . . . . . . . . . 2.3 SFL Approach to Interpersonal Metaphor of Modality . . . . . . . . . . . 2.4 The Semiotic Grammar Alternative to Metaphor of Modality . . . . . 2.5 The Problems with GM Theory and MM Studies . . . . . . . . . . . . . . . . 2.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7 7 7 8 12 15 17 18 20 22 24 25
3 Towards a Constructional Approach to Metaphor of Modality . . . . . . 3.1 General Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Constructional Approaches to Grammar . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Cognitive Grammar as a Construction Grammar . . . . . . . . . . 3.2.2 Cognitive Construction Grammar . . . . . . . . . . . . . . . . . . . . . . . 3.3 Subjectivity from a Constructional Perspective . . . . . . . . . . . . . . . . . . 3.3.1 Modality and Subjectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 Subjectivity and Intersubjectivity from a Constructional Perspective . . . . . . . . . . . . . . . . . . . . . .
29 29 30 31 32 34 34 37
vii
viii
Contents
3.4 Towards the Constructional Approach to Metaphor of Modality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 The Usage-Based Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6 The Corpus-Linguistic Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
40 43 45 47 47
4 The Semantic Motivation of MM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Theoretical Basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 The Proposal of a Refined Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 The Semantic Analysis of MM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
51 51 56 59 63 64
5 The Lexical and Syntactic Properties of MM . . . . . . . . . . . . . . . . . . . . . . 67 5.1 The Usage-Based Model and Corpus-Based Method . . . . . . . . . . . . . 67 5.2 The Syntactic Properties of MM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 5.2.1 Problems with Constituency Analysis . . . . . . . . . . . . . . . . . . . 71 5.2.2 The Syntagmatic Structure of MM . . . . . . . . . . . . . . . . . . . . . . 74 5.3 [It be ADJ that P] Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 5.3.1 The Productivity of the Construction . . . . . . . . . . . . . . . . . . . . 76 5.3.2 The Lexical Properties of the Construction . . . . . . . . . . . . . . . 80 5.4 [I VMental that P] Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 5.5 [I be ADJ that P] Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 5.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 6 The Pragmatics and Discourse Functions of MM . . . . . . . . . . . . . . . . . . 6.1 Discourse as the Research Instrument . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 [I VMental that P] Constructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.1 Functions of the Construction in Discourse . . . . . . . . . . . . . . 6.2.2 The Distributional Features of the Construction . . . . . . . . . . . 6.3 [I VVerbal that P] Constructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 [It be ADJ that P] Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.1 The Functions of This Construction in Discourse . . . . . . . . . 6.4.2 The Distributional Features of the Construction . . . . . . . . . . . 6.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
105 105 107 107 112 114 117 118 128 136 137
7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 Summary of Major Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Implications and Contributions of the Study . . . . . . . . . . . . . . . . . . . . 7.3 Limitations and Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
139 139 141 142 143
Contents
ix
Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Appendix B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Appendix C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Appendix D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Appendix E . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Abbreviations and Acronyms
A ADJ Adv. Agt AM BCSN CG COCA COHA COMP-clause CT-clauses DM EM GM H ICE-GB IS LOB MM MS NP OBJ P Rec S SEM SFG SFL SG SUBJ SYN
Addressee Adjective Adverb Agent Affective modality Basic communicative space network Cognitive grammar Corpus of Contemporary American English Corpus of Historical American English Complement clause Complement-taking clauses Deontic modality Epistemic modality Grammatical metaphor Hearer International Corpus of English, Great Britain section Immediate scope Lancaster-Oslo/Bergen corpus Metaphors of Modality Maximal scope of conception Noun phrase Object Proposition Recipient Speaker Semantics Systemic functional grammar Systemic functional linguistics Semiotic grammar Subject Syntax xi
xii
V Vmental Vverbal VP
Abbreviations and Acronyms
Verb Mental verb Verb of verbal process (speech act verb) Vantage point
List of Figures
Fig. 2.1 Fig. 2.2 Fig. 2.3 Fig. 3.1 Fig. 3.2 Fig. 3.3 Fig. 4.1 Fig. 4.2 Fig. 4.3 Fig. 4.4 Fig. 4.5 Fig. 4.6 Fig. 4.7 Fig. 4.8 Fig. 5.1 Fig. 5.2 Fig. 5.3 Fig. 6.1
Ravelli’s model of grammatical metaphor as a semantic compound (From Ravelli, 1988: 137) . . . . . . . . . . . . . . . . . . . . . . . System network of modality (Halliday & Matthiessen, 2004: 150) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . System of orientation types in modality (Halliday & Matthiessen, 2004: 619) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Construal configuration and its basic elements, (From Verhagen, 2005: 41) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Construal configuration for impersonal complementation constructions (From Verhagen, 2005: 134) . . . . . . . . . . . . . . . . . . . Symbolic structure of a construction (Croft & Cruse, 2004: 258) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . “Onstage” versus “Offstage” elements in construal (Langacker, 2013: 3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Canonical viewing arrangement (Langacker, 1987: 139) . . . . . . . . Basic communicative space network (from Ferrari & Sweetser, 2012: 49) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The refined basic communicative space network . . . . . . . . . . . . . . Dimensions of VIEWPOINT space . . . . . . . . . . . . . . . . . . . . . . . . . Mental space construal configuration for utterance with modal adjunct (14a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mental space construal configuration for first-person modal expression (14b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mental space construal configuration for impersonal complementation construction (14c) . . . . . . . . . . . . . . . . . . . . . . . . Hierarchic organization of modal expressions . . . . . . . . . . . . . . . . Simplified VerbNet entry for a verb class (https://uvi.col orado.edu/uvi_search_anywhere) . . . . . . . . . . . . . . . . . . . . . . . . . . . Mental space construal configuration for [I VVerbal that P] . . . . . . Succession of space configuration in discourse . . . . . . . . . . . . . . . .
9 18 18 38 39 40 53 53 55 58 58 59 60 61 69 89 99 122
xiii
List of Tables
Table 2.1 Table 4.1 Table 5.1 Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 5.6 Table 5.7 Table 5.8 Table 5.9 Table 5.10 Table 5.11 Table 6.1 Table 6.2 Table 6.3
Congruent pattern between the semantic and grammatical categories in SFL (From Halliday, 2004, 1998): 40) . . . . . . . . . Values of modality with verbs . . . . . . . . . . . . . . . . . . . . . . . . . . . System of mental processes type . . . . . . . . . . . . . . . . . . . . . . . . . Structure of it-extraposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Predicative adjectives used in [It be ADJ that-clause] pattern in COCA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Types of adjectives with extraposed it . . . . . . . . . . . . . . . . . . . . . Matrix ADJ predicates of extraposed that-clauses (Herriman, 2000: 597–598) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Semantic classification of adjectival predicates in [It be ADJ that-clause] construction . . . . . . . . . . . . . . . . . . . . . . . . . . . Verbs and verb classes with the syntactic frame [I Verb that P] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Examples of verbs serving as process in mental clauses (Adapted from Halliday & Matthiessen, 2004: 210) . . . . . . . . . Verbs, verb classes, and process types with the syntactic frame [I Verb that P] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Working definitions of semantic categories adopted in classification of modality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Classification of [I Verb that P] constructions . . . . . . . . . . . . . . Distribution of [I VMental that P] construction among different genres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Distribution of [I VVerbal that P] construction among different genres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Distribution of [It be ADJ that P] construction in COCA among different genres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12 61 68 73 78 80 81 83 91 92 94 95 96 112 115 128
xv
Chapter 1
Introduction
1.1 Form-Meaning Mapping and Grammatical Metaphor How do language forms relate to meanings? The answer to this question is probably where different schools of linguistics diverge from each other. A focal question in functional linguistic study is the mapping between forms and meanings. In systemic functional linguistics (henceforth referred to as SFL), the relationship between semantics and lexicogrammar is based on the principle of rank-based constituency, in other words, primarily realizational correspondence of a figure, a proposition (proposal), and a message to a clause (Halliday & Matthiessen, 2004). The underlying assumption is that for any given semantic configuration, there is one congruent realization in the lexicogrammar (Halliday, 1985). However, such congruent mapping, albeit prototypical, only partially defines the relationship between semantic units and grammatical units. There are realignments in the realizational relationship between semantic and grammatical units; this is the phenomenon of grammatical metaphor. Metaphor here is defined as “variation in the expression of meaning” (Halliday, 1985: 320), and then grammatical metaphor is brought in where the variation is essentially in the grammatical forms. Grammatical metaphor (henceforth GM) in systemic notion is a powerful resource to foreground meanings in nominal groups, background personal and subjective voice, and build cohesion. The significance of grammatical metaphor studies has been widely acknowledged in Australia, Europe, and Asia.
© Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7_1
1
2
1 Introduction
1.2 Defining Interpersonal Metaphor of Modality GM, differentiated in terms of metafunctions, has two basic types: ideational metaphor and interpersonal metaphor. The role of GM in mood and modality is examined within interpersonal metaphor. Interpersonal Metaphor of Modality1 has been subsumed under the theoretical rationale of GM as discussed above in SFL. The underlying assumption is that modal meanings are most congruently expressed by modal elements in the clause. Then, if a modal proposition or proposal is realized as if it was a projection sequence by a nexus of two clauses, rather than by a single clause, it is identified as Metaphor of Modality (Halliday & Matthiessen, 2004: 626). Some typical examples of metaphorically versus congruently expressed modality are shown in (1): (1)
a.
I think Mary knows
b.
Mary will know
c.
It is likely that Mary knows
d.
Mary probably knows (adapted from Halliday & Matthiessen, 2014: 693)
Examples (1b) and (1d) are considered ‘congruent’ realization of modality, while examples (1a) and (1c) are ‘metaphorical’. Behind this systemic concept of ‘congruency’ is the assumption that semantic categories are constant, and for each semantic category, there is a corresponding grammatical category to map with it. Modality as a semantic feature is originally said by Halliday to apply at clause rank; in other words, this is the congruent realization. The SFL rationale underlying Metaphor of Modality (henceforth MM) is that the “explicitly subjective and explicitly objective forms of modality […] represent the modality as being the substantive proposition” (Halliday & Matthiessen, 2014: 698). This claim is partly based on the matrix clause (of a declarative sentence) status they accredit to patterns like I think or It is likely, which, of course, is traditionally considered as the matrix clause. However, a reanalysis is put forward by Thompson (2002) in that “the issue around which the talk centers” (p. 133) in such sentences is contained in the so-called complement clause and not in the matrix clause (cf. Verhagen, 2005). This shows the controversy over the matrix clause status of patterns like I think or It is likely taking a complement clause. To address the controversy and beyond, it is now to offer a definition of interpersonal Metaphor of Modality (MM) in clearer terms. MM refers to the linguistic phenomenon where a modal meaning is not encoded in a modal auxiliary or adjunct, but in the rank of clause.
1
Metaphor of Modality is capitalized in this book as it is the central key term.
1.4 Theoretical Orientation
3
1.3 Goals of This Book The analysis above raises at least two questions of theoretical interest. First, what is the fundamental distinction between ‘metaphorical’ modal expressions and modal elements within a clause in terms of meaning? In other words, is there a motivation for the fact that modality is realized by a projection sequence by a nexus of two clauses instead of modal adjuncts in the clause? Second, can this linguistic phenomenon be theorized alternatively without the controversial notion of ‘congruency’? My book will, above all, try to solve these two puzzles. It proposes a constructional approach to account for Metaphor of Modality in an aim to achieve a more finegrained analysis of the semantic structure of this linguistic phenomenon, with inputs from the cognitive linguistic approach, in which semantic aspects receive focused attention.
1.4 Theoretical Orientation In functional approaches to linguistic investigation, functional considerations are taken as being foundational to the problem of describing language form. A functional approach to language is any approach that places particular emphasis on the communicative and social functions of language and attempts to explain the grammatical properties of language in terms of the way language is actually used in communicative context. Therefore, grammatical phenomena are characterized in terms of discourse, pragmatic, sociolinguistic, and cultural properties. Broadly speaking, “cognitive linguistics belongs to the functionalist tradition” (Langacker, 1999a: 13). Both cognitive and functional linguists believe that language is shaped and constrained by the functions it serves and by a variety of related factors: from environmental to sociocultural. Cognitive linguistics subsumes a variety of concerns and broadly compatible theoretical approaches with a common basic outlook that language is an integral facet of cognition. It seeks insofar as possible to explicate language structure in terms of the other facets of cognition on which it draws, as well as the communicative function it serves. Its guiding assumption is that language is symbolic in nature, that is, a language relates a meaning or conceptualization with a phonological/written/signed form. Another guiding principle is that language is shaped by functional aspect and by general cognitive abilities as well. Its objective is to analyze language in order to find the underlying conceptualizations motivating its emergence. The cognitive and the functional dimensions of language are not just two separate issues. Both functional and cognitive approaches attend to the pairing of form and meaning. Lexicon and grammar form a continuum in cognitive linguistics; that is why Panther and Thornburg (2009: 14) consider it reasonable to replace the term ‘grammar’ with ‘lexicogrammatical system’, or, “lexicogrammar”, which is the same term used in systemic functional linguistics. Both view lexicon and grammar as
4
1 Introduction
consisting of meaningful units, and hence of linguistic signs. Of course, there is certainly not theoretical uniformity between the two approaches, but a broad compatibility of outlook promises integration to some extent. The current study is, in some degree, theoretically grounded on the integration of the cognitive and functional approaches to the study of language structure. The theoretical and descriptive claims in this study will react to a number of longstanding issues in systemic notion of grammatical metaphor and syntactic analysis as well and throw new light from a cognitive constructional approach.
1.5 Methodology Prospects The constructional approach will be adopted to account for Metaphors of Modality. I, to a large extent, follow the research tradition of cognitive linguistics, particular that of construction grammars, in terms of methodology. I’m trying to model the cognition that lies behind language and goes far beyond it, like which language reflects in certain ways, and which in turn supports the dynamics of language use and language organization. This modeling is based on well-established constructs in cognitive linguistics and developed from previous models. The approach adopted in this study is essentially a usage-based one, for which constructions can be appropriately modeled, combined with abductive reasoning. The term “usage-based” was coined by Langacker (1987), but the roots of this view of language can be found among the functional-typological approach to language and American functionalist approach. From a broader view of usage-based theoretical approach, constructions can be seen as processing units or chunks—sequences of words that have been used often enough to be accessed together. Along with usage-based approach is relatively novel corpus-linguistic methodologies that permit a data-driven identification of routinized patterns—the assumed building chunks of speaker’s linguistic knowledge. The combination of usage-based model with data-intensive methodologies is promising for research proposals of constructions. In line with this approach, corpus observations, both qualitative and quantitative, will be used. Therefore, the descriptive claims in this study are based closely on the observation and interpretation of patterns in authentic data. The main data source of language facts is Corpus of Contemporary American English, henceforth referred to as COCA (Davies, 2008). In my quest for a sufficiently large corpus to get abundant examples of patterns and frequency data, COCA has been a natural choice. I searched it via the interface: http://corpus.byu.edu/coca/. Please note that presently the website of COCA is at https://www.english-corpora.org/coca/. The corpus comprises about 450 million words of text and is equally divided among spoken, fiction, popular magazines, newspapers, and academic genres. It includes 20 million words each year from 1990–2012, and the corpus is also updated regularly (the most recent texts are from Summer 2012). Because of its design, it is perhaps the only corpus of English that is suitable for looking at current, ongoing
References
5
changes in the language (http://corpus.byu.edu/coca/). When relevant constructions are qualitatively analyzed, larger discourse contexts are taken into account. Above all, the aim is to explore as far as possible in authentic language data the range of formal properties and semantic or discourse functions found with the linguistic facts in question and to best account for them.
1.6 Outline of the Book This chapter has presented an introduction to Metaphor of Modality and a large number of topics that will be discussed more fully in later chapters. I have rationalized the integration of functional and cognitive approaches to language and proposed a constructional approach to MM. Chapter 2 reviews the systemic functional studies of grammatical metaphor and the studies of modality in major schools of linguistics, with a focus on Metaphor of Modality, which is a crossover of GM and modality. In Chap. 3, I argue for the constructional approach to describe and account for MM. Under the theoretical framework of constructional grammar, Chap. 4 models the cognitive structure and process that lie underneath variant modal constructions and are reflected in its language form and language organization in certain ways and provides a fine-grained analysis of the semantic structure of explicit modal constructions. Chapter 5 investigates the lexical and syntactic properties of MM, based on the large corpus, COCA. Chapter 6 incorporates the SFL emphasis on discourse into the constructional framework and addresses the distinction of metaphorical and congruent modes of modality in terms of pragmatics and discourse function. In each of these chapters, the adopted method is presented, targeting its share of the research questions. In the final chapter, the main theoretical claims and empirical findings of this study are summed up, with their theoretical and empirical values. Then this book rounds up with a few extending areas for further research.
References Davies, M. (2008). The Corpus of Contemporary American English (COCA). Available online at https://www.english-corpora.org/coca/. Halliday, M. A. K., & Matthiessen, C. (2004). An introduction to functional grammar. Hodder Arnold. Halliday, M. A. K., & Matthiessen, C. (2014). Halliday’s introduction to functional grammar. Routledge. Langacker, R. W. (1987). In Foundations of cognitive grammar, volume 1, Theoretical prerequisites. Stanford University Press. Langacker, R. W. (1999a). In Grammar and conceptualization. (Cognitive Linguistics Research 14.) Mouton de Gruyter. Panther, K. U., & Thornburg, L. (2009). Introduction: On figuration in grammar. In Metonymy and metaphor in grammar (pp. 1–44).
6
1 Introduction
Thompson, S. A. (2002). “Object complements” and conversation towards a realistic account. Studies in language. International Journal sponsored by the Foundation “Foundations of Language”, 26(1), 125–163. Verhagen, A. (2005). Constructions of intersubjectivity: Discourse, syntax, and cognition. Oxford University Press.
Chapter 2
GM Theory and Interpersonal Metaphor of Modality
Interpersonal Metaphor of Modality, a subcategory of grammatical metaphor in systemic functional linguistics, represents a SFL position in a large and complex area of linguistic study on modality from a unique perspective. In this chapter, I first review the previous literature around grammatical metaphor theory and SFL perspective on modality and then focus on (interpersonal) Metaphor of Modality and other work particularly addressing this linguistic phenomenon.
2.1 The Theory of Grammatical Metaphor This section will focus on the particular theoretical framework of grammatical metaphor, of which Metaphors of Modality is a part. GM theory has undergone substantial development and elaboration since Halliday (1985) put forward the systemic notion of grammatical metaphor. The prominent themes in its development include the seemingly constant clarification of its systemic notion and the exploration of GM subtypes.
2.1.1 Clarification of Key Notions The term “metaphor” comes from the Greek metaphora and signifies ‘transfer, transposition’. Aristotle in his poetics viewed metaphor as part of rhetoric and poetics. To him, it consists of naming an entity with a word referring to something different, the transference being on the grounds of analogy (Divasson & León, 2006: 6517). Traditionally, metaphor is one major type of figurative usage. Typical examples are like:
© Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7_2
7
8
2 GM Theory and Interpersonal Metaphor of Modality
(2)
a.
Juliet is the sun. (Shakespeare)
b.
I was lost in a sea of nameless faces.
Probably following Aristotle’s tradition, metaphor is seen as the general term for verbal transference of various kinds, and usually, as variation in the use of words. Halliday (1985) then, defined metaphor, perspectivized ‘from above’, as “variation in the expression of meaning” (p.320). As observed by him, in rhetorical transference, there is strong grammatical element and “the variation is essentially in the grammatical forms although often entailing some lexical variation as well” (p. 320). For example: (3)
a.
Protests flooded in.
b.
Very many people protested.
As can be seen, not only are words varied but also grammatical forms are transferred. The systemic notion of ‘metaphor’ is different from the current and prominent cognitive approach to metaphor. In the conceptual theory of metaphor, metaphor is not a property of individual linguistic expressions and their meanings, but of whole conceptual domains. A metaphor is a conceptual mapping between the source domain—the domain supporting the literal meaning of the expression, and the target domain—the domain the sentence is actually about. Metaphor is not simply a stylistic feature of language, but metaphoric is fundamentally the nature of thought itself. In other words, metaphor is the “cognitive mechanism whereby one experiential domain…is partially mapping onto a different experiential domain, the second domain being partially understood in terms of the first one” (Barcelona, 2003: 211). A metaphor is a mapping of a source conceptual schema onto a target conceptual schema (Turner, 1991: 52). In summary, the conceptual theory of metaphor claims that conceptual organization is in large part metaphorical, and consequently thought itself is metaphorical, whereas the systemic notion of metaphor, at least initially, focuses on verbal transference from the perspective of how the meaning is expressed.
2.1.2 The Development of GM Theory Halliday (1985) proposes a different type of perspective on metaphor, defined as variation in the expression of meaning. Then grammatical metaphor is brought in where the variation is essentially in the grammatical forms, based on his assumption that for any given semantic configuration, there is one congruent realization in the lexicogrammar. For example:
2.1 The Theory of Grammatical Metaphor
(4)
a.
Her eyes are brown [congruent]
b.
She has brown eyes [metaphorical]
9
The concept of incongruence, in fact, already turned up in earlier work by Halliday (1978: 180): ‘incongruent’ means “not expressed through the most typical (and highly coded) form of representation”. Halliday (1985: 321) further made it clear that: This is not to say that the congruent realization is better, or that it is more frequent, or even that it functions as a norm; there are many instances where a metaphorical representation has become the norm, and this is in fact a natural process of linguistic change.
He added that the selection of metaphor is itself a meaningful choice, and further semantic features are added to GM even though the variants are synonymous in certain respects. What mostly determines the extent of GM in a text is whether that text is spoken or written. Through his initial description of grammatical metaphor, Halliday (1985) expands the traditional notion of metaphor to include metaphorical processes in the grammar. However, he did not give any explicit definition of GM. Halliday’s theory is taken up by Ravelli (1985), whose definitions of GM center on “an alternative lexicogrammatical realization of a semantic choice” (p. 55) and involve “a realization of a meaning in an atypical lexicogrammatical form” (p. 4). It is suggested that there must be a typical or expected realization, which is the congruent one. This initial model is later refined by Ravelli (1988), who proposes that metaphor also involves semantic variation. She presents the new model as illustrated in Fig. 2.1. In this interpretation, a congruent lexicogrammatical realization derives from a simple, straightforward semantic choice, whereas grammatical metaphor is seen as a combination of semantic features. Thus, GM also involves a meaning difference from congruent choice, since the grammatical category itself has a feedback effect
S – Semantic choice; c- congruent form; m- metaphorical form Fig. 2.1 Ravelli’s model of grammatical metaphor as a semantic compound (From Ravelli, 1988: 137)
10
2 GM Theory and Interpersonal Metaphor of Modality
into the semantics, though their meanings are close because of “shared experiential content” (Ravelli, 1988: 134). This analysis effectively accounts for meaning expansions by demonstrating that metaphorical modes lead to compound semantic choices rather than one single meaning. However, this model seems to suggest that the semantic product of metaphor is merely the sum of the meanings chosen. This weakness might be due to the difficulty with illuminating the exact nature of the semantic difference in the then state of systemic theory. Dead or frozen grammatical metaphor is excluded as certain metaphorical expressions enter into the system of language and become the typical or expected way of realizing that meaning. This also shows that being “typical” might not be an appropriate criterion for congruent realizations. Grammatical metaphor is later taken by Halliday (1989) as a process, a substitution of one grammatical class, or one grammatical structure, by another. In his view, GM is part of the reason why scientific English is difficult to understand, as it presents a different view of the world.1 Yet the aspect of meaning is not exactly treated in this early framework. A focus is on the analysis of GM, in which a metaphorical realization is analyzed by unpacking it into a congruent agnate form and comparing its structure to the congruent one (Taverniers, 2003). In later justification for grammatical metaphor, congruent realization is seen as the prototypical realization of any given semantic configuration. For instance, a verb is the prototypical realization of a process, a noun the prototypical realization of a thing, and an adjective the prototypical realization of a property. GM involves the transformation from one grammatical class to another. GM construes (meaning ‘symbolically constructs’ in SFL) a phenomenon of one kind in a way typically representing a phenomenon of some other kind. In other words, grammatical metaphor is “a substitution of one grammatical class, or one grammatical structure, by another” (Halliday, 1993: 79). Later views capture the stratal tension between the semantics and the lexicogrammar for more explicit description. According to Halliday (1998), language, as the semiotic of higher order consciousness, is stratified: it has a stratum of lexicogrammar in between the meaning and the expression. The relationship between semantics and lexicogrammar, according to systemic functional grammar, is prototypically natural; and the phenomenon whereby a set of agnate forms is present in the language having different mappings between the semantic and the grammatical categories is referred to as grammatical metaphor (Halliday & Matthiessen, 1999: 7). Grammatical metaphor is then redefined as “a realignment between a pair of strata: a remapping of the semantics onto the lexicogrammar” (Halliday, 1998), in
1
The notion grammatical metaphor (GM) was often used when ideational metaphor was actually referred to in a long period from 1990s to early 2000s, which reveals the focused SFL attention on ideational metaphors rather than interpersonal metaphors then.
2.1 The Theory of Grammatical Metaphor
11
other words, “the expression of meaning through a lexicogrammatical form which originally evolved to express a different kind of meaning” (Thompson, 1996: 165). In Halliday’s theorizing, the congruent construal of meaning is the presupposition of grammatical metaphor. Halliday explains the term “congruent” as the pattern in which the grammatical categories evolved in the semiotic transformation of experience, construing the elements and figures. Halliday (2004) further justified the conception of ‘congruence’ in that it “captures the inherent directionality of the human semogenic process” (p. 37), in the three time frames, i.e., phylogenesis, ontogenesis, and logogenesis: The reason we can interpret one form of construal as congruent is that it is historically prior: in the grammar’s construction of reality, the mapping of process into verb and of entity into noun precedes the mapping of process into noun. (Halliday, 2004/1999: 107)
It was found that the evolution of grammatical metaphor, as far as the West is concerned, began with the explosion of process nouns in emergent scientific Greek from 550 BC onwards (Halliday, 1987: 146; Banks, 2003: 129). As another time frame, logogenetic time is the time of the unfolding of the text. As said by Halliday (1999), a text tends to move from the more congruent to the more metaphorical. A phenomenon, through gateway of analogy, can be propelled from one category to another, where the original interpretation is combined with the new one into a more complex whole. GM is a distinct construing of experience in which there is junction of semantic features (Halliday & Matthiessen, 1999: 273), rather than simply an alternative realization of the same meaning. The shift in the rank and the class status becomes metaphorical when a semantic junction with the original is created. The metaphoric elements embody a junction of two semantic categories. Therefore, whether there is a junction of two semantic categories could be used as an index for identification of GM. The key defining motifs of GM are transformation within the semiotic mode, semantic compounding, and stratal tension. Halliday’s approach to cognition is utterly language-based. In his view, the grammatical system of language is able to construe experience by imposing categories on people’s perceptions of phenomena (Halliday, 1998). GM is reinterpreted as “the reconstrual of one domain in terms of another domain”, for example: the domain of figures, congruently construed (symbolically constructed) as clauses, is reconstrued in terms of the domain of elements/things as nominal groups. Furthermore, mode of language is taken into account in the GM conceptualization. The grammar of spoken language, where congruent patterns are laid down, construes a world that is fluid and transitory just as speaking itself; the grammar of written language construes a world that is solid, lasting and clearly bounded just like writing itself (Halliday, 1999).
12
2 GM Theory and Interpersonal Metaphor of Modality
2.1.3 Ideational Metaphor and Interpersonal Metaphor In SFL, language features functional diversity—ideational, interpersonal, and textual metafunctions—simultaneously. According to Halliday (2004: 53), ‘meaning’ is a parallel mode of semiotic activity that simultaneously both construes experience and enacts the social process. In this respect, GM is subdivided into ideational metaphors and interpersonal metaphors (and textual metaphor included by other scholars but excluded by Halliday himself). According to Halliday and Mattiessen (1999), the ideational semantic resources construe our experience of the world that is around us and inside us. Halliday adopts a language-based approach to cognition in his account of GM. In his view, the grammatical system of language is able to construe a universe of things and relations, imposing categories on people’s perceptions of phenomena. The clause is the central construct to construe happenings/figures. A process embodies the temporal properties of a figure unfolding in time. A participant is a thing that can carry or ascribe attributes. As the different kinds of elements, process, participants, and circumstances are construed respectively, as verbs/ verbal group, nouns/nominal group, and adverbial group or prepositional phrase, this pattern is ‘congruent’ in the sense being the way language evolved. The congruent pattern between the grammatical and the semantic categories is specified in Table 2.1. Within the ideational metafunction, grammatical metaphor involves firstly in rank a ‘remapping’ of sequences, figures, and events in the semantics onto clause nexuses, clauses, and groups in the lexicogrammar (Halliday, 2004: 58). The whole set of mappings seems to be shifted downwards: a sequence is realized by a clause; a figure is realized by a group. Primarily, it is a strategy enabling us to transform our experience of the world. For instance, nominal groups realize figures construed as entities, so the ideational potential of classification and characterization becomes available. GM, then, is interpreted as the reconstrual of one domain in terms of another domain, for example: the domain of figures, congruently construed (symbolically constructed) as clauses, is reconstrued in terms of the domain of elements/things as nominal groups. On the principle of congruence between semantic and grammatical categories, GM, consequently, shifts both the rank and the class status: for instance, the sequence, from being realized by a clause nexus to a single clause, and the Table 2.1 Congruent pattern between the semantic and grammatical categories in SFL (From Halliday, 2004, 1998): 40)
Congruence in rank
Congruence in status
Semantic Grammatical Semantic
Grammatical
Sequence Clause nexus Entity Clause Quality Figure Element Group/phrase Process Circumstance Relator
Noun/nominal group Adjective Verb/verbal group Adv. /Adv. Group/ Prep. Phrase Conjunction
2.1 The Theory of Grammatical Metaphor
13
circumstances, from being realized by adverbial groups or prepositional phrases to adjectives. The metaphoric shift—from sequence to figure, from figure to element, and so on—creates a greater distance from ordinary experience. The metaphorical mode of construal makes it possible to recast everyday experience, retaining selectively some features from the congruent wording and adding others. GM adds a further perspective on the phenomenon being represented, without displacing the congruent perspective. In other words, our everyday construal of the conscious experience, that is, our folk theory of sensors sensing phenomena or projecting ideas, is the congruent folk model; the scientific model is metaphorical (Halliday & Matthiessen, 1999). Halliday and Matthiessen (2004: 646) classified three types of ideational metaphor according to where this downgrading starts: (1) sequence, (2) figure, or (3) element. On the element status, Halliday (1998, 1999) enumerates 13 types of grammatical metaphor. In discourse, instances of grammatical metaphor tend to come in favorite clusters or syndrome. Metaphoric syndrome could be best represented by the ‘favorite clause type’, that is, one clause ‘relational’ of three elements: nominal group, verbal group, and another nominal group, construing a sequence of two figures, linked by a logicalsemantic relation. Nominalizing is the most powerful resource for creating ideational metaphor. This kind of nominalizing metaphor plays a dual role: on the one hand, it enables constructing hierarchies of technical terms, and on the other hand, it realizes a stepwise development of argument, with what would congruently construe message/ proposition/figure packaged in nominal form as themes. What’s more, when any phenomenon of experience is construed as a thing, it has the maximum potential for semantic elaboration, because qualities can be attached to things for expansion and they can have expansion of their own. In a series of papers published in late 1990s, Halliday (1999) particularly justified the conception of ‘congruence’ in that it captures the inherent directionality of the human semogenic process, in the three time frames, i.e., phylogenesis, ontogenesis, and logogenesis. In terms of phylogenesis, the reason one form of construal is interpreted as congruent is that it is historically prior: in the grammar’s construction of reality, the mapping of process into verb and of entity into noun precedes the mapping of process into noun. Ontogenesis is the process of the growth and maturation of the user of the language. Halliday suggests that grammatical metaphor is a feature that is more characteristic of the language of adolescence than of earlier childhood. That is, the congruent is ontogenetically prior to the metaphorical. Painter (1993) states that meaning and lexical class are congruent with one another in young children’s speech and “control of experiential grammatical metaphor is a late development” (p. 111). Children are initiated into grammatical metaphor in the upper years of primary school. Derewianka (1995: 198) also holds that “grammatical metaphor is intimately involved in the development of experiential meanings and is particularly implicated in the shift from common sense to uncommon sense”.
14
2 GM Theory and Interpersonal Metaphor of Modality
Based on the data gathered in English metalinguistic texts, Liu (2007: 117) groups ideational grammatical metaphors into four categories: nominalization, verbalization, adjectivation, and prepositionalization. Firstly, a nominal group/phrase may be derived either from an adjective encoding a property, or from a verb encoding a process, a prepositional phrase encoding a circumstance, or from a relator encoding the paratactic or hypotactic relations; all these are nominalization. Secondly, verbalization refers to the semantic transference in which non-process is taken as process, which is lexicogrammatically realized by the substitution of a verb phrase for a non-verbal group/phrase. Thirdly, adjectivation refers to the transference from a verb encoding a process, from a conjunction encoding the paratactic or hypotactic relations, or from an adverb encoding a circumstance into an adjective phrase representing a quality. Last, prepositionalization refers to the transference in which the paratactic or hypotactic relation congruently encoded by a conjunction is realized metaphorically as a circumstance encoded by a prepositional phrase. The other major category of GM is interpersonal metaphor. In Halliday’s (1994) version of SFL, the interpersonal organization of a clause is structured in terms of a Mood + Residue pattern. The different systems are specified on two different levels: the level of speech function on the one hand, and the level of mood and modality on the other hand (Taverniers, 2008). The systems of mood and modality deal with the lexicogrammatical formation of an utterance. Metaphor of mood is based on the relationship between the speech functional categories of ‘statement’, ‘question’, ‘command’, and ‘offer’ on the one hand, and the expression of these categories through different MOOD types on the other hand. They are based on the traditional distinction between clause types in terms of, for example, declarative, indicative, interrogative, and imperative moods. In systemic functional grammar, statements are expected to be realized congruently as declaratives, questions as interrogatives and commands as imperatives (Martin, 1997: 26). Thus, the shift in the realizational domain of proposal (commands) from ‘imperative’ to ‘indicative’ clauses is a type of mood metaphor (Halliday & Matthiessen, 2004: 632). In addition, there is another prominent kind of mood metaphor, based on projection. The proposition or proposal can be realized by a clause nexus of projection rather than congruently by a simple clause. Consequently, this expands the meaning potential of speech function in two aspects. In other words, metaphors of mood make it possible for the semantic system of speech function to be further elaborated in delicacy (Halliday & Matthiessen, 2004: 631). The potential for negotiation in dialogue created by metaphors of mood is related to the contextual variables of tenor often in terms of formality and politeness, or more generally, the social distance between the speaker and the addressee. Further options that are available for expressions with [indicative] mood consist in the system of MODALITY. Then interpersonal Metaphor of Modality (MM) is brought in as another type of interpersonal metaphor. As MM is the research focus of the current study, it will be reviewed in more details in the next section. Interpersonal metaphor is crucial in terms of the ability to express point of view, argue a position, and negotiate with others (Torr & Simpson, 2003: 181).
2.1 The Theory of Grammatical Metaphor
15
Based on a general semiotic-functional characterization of the interpersonal sign in terms of scoping and grounding, Taverniers (2008) defines interpersonal metaphor as involving a doubling of semiosis, viz. a doubling of scoping in its structuralrealizational dimension, and a doubling of grounding in its semantic-functional dimension. To round up, Halliday has studied ideational grammar in depth and tried to justify the conception of ‘congruence’ in that it captures the inherent directionality of the human semogenic process in the three time frames of phylogenesis, ontogenesis, and logogenesis. However, this justification of ‘congruence’ does not apply to interpersonal metaphor, as evident from his own words in Introduction to Functional Grammar (Halliday & Matthiessen, 2004: 636): Unlike interpersonal metaphor, the other type of grammatical metaphor, ideational metaphor, is learned later by children and is not part of the grammar of ordinary, spontaneous conversation that children meet in the home and neighbourhood…
In other words, interpersonal metaphor is not learned later by children and may be part of the grammar which children get access to in the home around. Therefore, justification of ‘congruence’ with respect to interpersonal metaphor is not as substantive as that of ideational metaphor, though the latter also has its problems.
2.1.4 The Current State of GM Studies More recently, it is clearly stated by SFL scholars that the choice of a more metaphorical grammatical realization construes a different meaning from the choice of a more congruent one. Therefore, GM has the effect of expanding the meaning potential of the language, through realignments in the realizational relationship between semantic units and grammatical ones (Halliday & Mattiessen, 2004: 592). Grammatical metaphor expands the meaning potential of the language because it combines two ways of meaning. Furthermore, it represents a different way of thinking and talking about the world (Thompson, 2004: 223). This shows GM is closely related with human cognition. Thompson (2004: 224) sees that there will be many factors influencing the extent to which either metaphorical or congruent wordings might be preferred in particular contexts. However, he does not specify the factors involved. Grammatical metaphor, as further clarified by Halliday and Mattiessen (2004: 658), is not “the most straightforward coding of the meaning selected originally”. The choice of a more metaphorical wording, which combines two ways of meaning, construes a different meaning from what a congruent wording would do (Thompson, 2004: 224). In a sum, systemic functional grammar tends to focus on the lexicogrammatical stratum that lies between the strata above it (semantics) and below it (phonology). The stratal relationship between semantics and lexicogrammar has been a key motif for grammatical metaphor throughout.
16
2 GM Theory and Interpersonal Metaphor of Modality
The significance of grammatical metaphor studies has been acknowledged by Chinese systemic functionalists ever since 1990s. They closely track the development of GM theories with new framework and breakthrough, while questioning Halliday’s vague definition of ‘congruence’ from time to time (Hu, 2000). A considerable number of scholars have attempted to develop and elaborate the GM theory to strengthen its rationale and rescue its weaknesses. The motif of GM theories is proposed by Liu (2005) to reveal the complex relations among language forms, meanings and contexts and their roles in the process of language evolution. Zhang and Zhao (2008) investigate the structural relations between the congruent and the metaphorical forms at the lexicogrammatical level and divide the congruent form into two types: the one with formal similarity and the other without formal similarity. As a step further, Cong and Wang (2013: 33) argue that the congruent and the metaphorical forms have both same and different meanings: the same meanings refer to the same lexical meanings, whereas different meanings refer to different grammatical meanings and new meanings made through the interaction between grammatical meanings and lexical meanings. Yet lexical meaning is not an autonomous phenomenon; words in general are signs for unspecific complexes of mental representation (Geeraerts, 2010: 298). In this sense, this semantic analysis still has its problem. In terms of the principle of formal similarity and the criteria of semantic changes, Cong and Wang (2013) categorize grammatical metaphors into three types, viz., semantic reduction, semantic expansion, and semantic restructuring. For example, they take nominalization as a typical subtype of semantic reduction. However, this is not always the case. Let us see a classical example of metaphoric reconstrual given by Halliday (2004: 57): Clause: the driver was driving the bus too fast down the hill Nominal group: the driver’s overrapid downhill driving of the bus
In this example, the original meaning is not reduced in the nominalizing metaphor. Therefore, this new categorization of GM still has its problem. There are also studies on application of GM to academic discourse analysis and second language acquisition. Hu (2015)’s study finds relative more use of GM in China English academic texts than their Singapore English counterparts. Liardét (2018) investigates Chinese EFL learner’s use of evaluation and stance in academic texts by exploring their deployment of interpersonal grammatical metaphors. The systemic notion of grammatical metaphor is continuously made up by various scholars in attempts to solve its theoretical problems, for example, the vague definition of ‘congruence’, inadequacy in illuminating the difference in meaning between the congruent and the metaphorical due to more emphasis onto the lexicogrammatical level. All these make the identification what is congruent and what is metaphorical problematic. Zhang and Dong’s (2014) recent review of its developmental models is a comprehensive summary and evaluation of GM. They divide the development of GM theory into three stages: functional model, stratified functional model, and stratified systemic functional model. Moreover, they point out the necessity of studying GM in terms of motivations, functions, and realizations.
2.2 The Systemic Functional Approach to Modality
17
2.2 The Systemic Functional Approach to Modality Much functional work on modality is descriptive, relying on other theoretical concepts to define and explain features of modal meaning. Systemic functional treatment of modality is no exception. According to SFL, functionality is intrinsic to language. It consists of three integral components, each corresponding to one of the metafunctions: the ideational, the interpersonal, and the textual, with multi-functionality in the clause as one fundamental feature of functional grammar. In SFG, mood plays a special role in carrying out the interpersonal functions of the clause. The mood element consists of two parts: the subject and the finite operator, which is part of a verbal group. Any finite is inherently positive or negative in polarity. Then modality encapsulates the immediate degrees between the positive and the negative poles of polarity. modality is referred to as “the speaker’s judgment, or request of the judgment of the listener, on the status of what is being said” (Halliday & Matthiessen, 2014: 172). It is a major exponent of the interpersonal function. Modality is classified into two major types, based on that the modal space has a different significance for propositions and proposals. In a proposition, “two kinds of intermediate possibilities” (ibid: 177) are identified: (i) degrees of probability (which corresponds to epistemic modality); (ii) degrees of usuality. Halliday and Matthiessen refer to this type as modalization. In a proposal, there are also two kinds of intermediate possibility. In a command, the intermediate points represent degrees of obligation (which corresponds to deontic modality); in an offer, they represent degrees of inclination. These scales of obligation and inclination are categorized as modulation. This paradigm also indicates that modality is organized as a system of three values: median, high, and low. In SFG, the options of modality are present throughout the system, as shown in Fig. 2.2. In this systemic network of modality, a system of orientation is differentiated with examples such as it is certain (that) that is true and I’m certain (that) that is true. With these examples, according to Halliday and Matthiessen (2004: 149), the speaker is explicitly stating the source of the conviction: it is ether being said to be objective, or presented as a subjective judgment on the speaker’s part. In contrast, the versions like that must be true and that’s certainly true leave implicit the source of the conviction. Halliday (1985, 1994) includes probably the first proposal to add a dimension of explicit versus implicit in terms of the source of the conviction. It is intimately combined with the subjective or objective dimension. Halliday formulated a comprehensive system of Modality, as demonstrated by Fig. 2.2. He offers a functional framework to account for the various expressions of modality. Moreover, his study of modality can be expanded to the level of pragmatics and provides an analytic tool for discourse analysis from a sociolinguistic perspective (Feng, 2011).
18
2 GM Theory and Interpersonal Metaphor of Modality
Modality
Modalization
usuality
Type
Modulation
obligation inclination subjective
Orientation
objective
Modality
explicit Implicit Value
Polarity
Median
high
Outer
low
Positive
direct
Negative
transferred
Fig. 2.2 System network of modality (Halliday & Matthiessen, 2004: 150) Fig. 2.3 System of orientation types in modality (Halliday & Matthiessen, 2004: 619)
ORIENTATION
Subjective objective Explicit Implicit
2.3 SFL Approach to Interpersonal Metaphor of Modality In system network of modality, the basic diction that determines how each type of modality, viz. probability, usuality, obligation, or inclination, will be realized is the orientation: the distinction between subjective and objective modality, and between the explicit and implicit variants. The system of orientation is shown in Fig. 2.3. In the ‘subjective’ orientation, “the modal assessment is presented as one given by the speaker, or demanded from the addressee” (Matthiessen et al., 2016: 236). Nonetheless, the notion of ‘objective’ is not defined, though it apparently contrasts with ‘subjective’. This is partly due to the research status quo that the distinction between subjective and objective modality is still heatedly debated and not settled. However, this deficiency leaves problems for subsequent studies. For instance, Yang (2021) suggests that the distinction between “implicit subjective” and “implicit objective” orientations in English may not be necessary and “apparently both modal verbs and modal adverbs indicate the speakers’ subjective assessment” (p. 169). This reveals the need to clarify the notion of ‘subjectivity’. Halliday (1985, 1994) claims that modal meanings are most congruently expressed by modal elements in the clause. The metaphorical realization of modality is a clause that projects or embeds the clause to which a modal value is assigned. According
2.3 SFL Approach to Interpersonal Metaphor of Modality
19
to Halliday, the metaphorical realization of modality diverges from the congruent pattern, by a clause that projects or embeds the clause to which a modal value is assigned. Here, a modal meaning is construed outside the proposition to which modal appraisal is assigned. Typical examples given by Halliday and Matthiessen (2004: 615) are as follows: (5)
a.
I think Mary knows.
b.
It is likely that Mary knows.
In order to accommodate metaphorical expressions in the system of modality, Halliday introduces an additional subsystem with the options [implicit] and [explicit]. Thus, the semantic domain of modality is extended through grammatical metaphor to include explicit indication of subjective and objective orientation by a nexus of two clauses. This is made possible in two types of constructions: projecting mental processes, as illustrated in examples (5a) above, and relational processes, as illustrated in the example (5b). With explicit subjective modality, GM presents the evaluation as a projection of the writer’s stance; with explicit objective modality, GM is used to construe the evaluation as fact rather than opinion. By creating new patterns of structural realization, GM opens up new systemic options in meaning and leads to an expansion of the meaning potential. According to Halliday and Matthiessen (2004: 625), “we only say we are certain when we are not”, which implies a tendency of objectification of the subjective propositions (Li, 2007). Lexicogrammatically, interpersonal Metaphor of Modality is realized by a clause complex, in which a clause such as I think or It is likely is followed by a thatclause as in (5). According to A Comprehensive Grammar of the English Language (Quirk et al., 1985), both sentences can be analyzed into the matrix sentence plus a complementation by a finite clause. According to Halliday (1994: 355) and Halliday and Matthiessen (2014: 698), encoding an appraisal with the impersonal pronoun it extraposition rather than a subjective projecting clause, such as I think, makes it possible for the speaker to disguise the source of the opinion and thereby makes it appear as if it was not his or her point of view at all. Halliday and Matthiessen (2004: 605) also indicate that projection covers several types of modal assessment: not only probability, usuality, inclination, and obligation, but also naturalness, obviousness, certainty, prediction, presumption, and desirability. They also give an indication of where these types of modal assessment correspond to categories of appraisal described by Martin and White (2005): affect, appreciation, and judgment. Compared with other types of grammatical metaphor, the treatment of interpersonal Metaphor of Modality is very brief in Halliday (1985, 1994) and Halliday and Matthiessen (2004, 2014). Chinese scholars further expound Metaphors of Modality from aspects of grammatical forms and types of modality. Zhou (2003) analyzes the special effect of discourse created by MM from their explicit orientations and discourse cohesion. A contrastive study is conducted by Wei (2008) on MM between
20
2 GM Theory and Interpersonal Metaphor of Modality
English and Chinese in terms of syntactic variations, nominalization and transitivity alternations, subjective and objective preferences of modality. His research shows that both English and Chinese have incongruent forms of modality. The incongruent forms of modality in English usually exist in the hypotactic clauses, whereas the Chinese counterparts normally appear in the paratactic clauses. Li (2007) tries to show the implication of objectification in MM in terms of pragmatics.
2.4 The Semiotic Grammar Alternative to Metaphor of Modality Like Halliday, McGregor’s (1997) semiotic grammar also differentiates proposition and proposal, but he theorizes modality in a different way and casts doubts on Halliday’s (1985, 1994) interpersonal Metaphor of Modality. Fundamental to semiotic grammar is the notion that not only is language a semiotic system in the Saussurean sense but also that its organizing system, its grammar is a semiotic system. McGregor (1997) stresses that language is a system of signs in which form and function are inseparably connected. Syntactic forms are one side of linguistic signs, which have a meaning on the other side. Although morphology, syntax, and semantics are regarded as part of a single enterprise, the primary subject matter of semiotic grammar is the form of linguistic signs—the form of the signifier and the form of the signified. Semiotic grammar presumes constituency, dependency, conjugational, and linking relationships as four distinct types of syntagmatic relationship. Constituency has figured prominently in the linguistic theorizing of the twentieth century, to the virtual exclusion of all other syntagmatic relationships. It is characterized by McGregor (1997) as “the semiotic relationship which a part may discharge in the whole to which it belongs” (p. 58). Linking relationships obtain between a linguistic entity and one or more others, or a non-linguistic entity. Dependency is described as part-part relationships, whereas conjugational relationships are referred to as wholewhole relationships. McGregor differentiates a binary contrast between SCOPAL and FRAMING conjugational relationships. In framing, the encompassed whole is delineated from the surrounding linguistic context and marked off as to be viewed as depiction of its referent, typically as a speech or thought. Scopal relationships, by contrast, do not involve this sort of marking off. Conjugational relationships always involve a modified unit, the whole over which the relationship applies. These four types of relationships in turn define four semiotic types: experiential, textural, logical, and interpersonal. From the experiential perspective, a clause represents or constructs a situation, an experiencable phenomenon that is spatially and temporally located in some referent world. The syntagmatic relationships that characterize the interpersonal semiotic, according to semiotic grammar, fall into conjugational relationships, which are
2.4 The Semiotic Grammar Alternative to Metaphor of Modality
21
further classified along two dimensions: one more syntactic, the other more semantic in nature. The first dimension involves a binary contrast between SCOPING in which a unit applies a certain domain, leaving its mark on the entirety of this domain and FRAMING in which a unit delineates the domain over which it applies, marking off from everything else. The other dimension involves a distinction between three types of modification: illocutionary, attitudinal, and rhetorical. McGregor’s (1997) semiotic grammar theorizes that modal auxiliaries like may, must, and can, in fact, hold the full clause within their scope though distributionally strongly associated with the verb. They are referred to as modal operators, rather than auxiliaries, to reinforce the fact that they serve to modify the whole clause interpersonally, through a scopal relationship. Within the same framework, sentences like ‘It was fortunate that I survived the crash’ or ‘It was good that you came’ instantiate scopal relationship of objective evaluation. The relative clause with dummy it subject is the SCOPING clause. In SCOPING a unit applies a certain domain, leaving its mark on the entirety of this domain, which is the case with the above two sentences. The two main types of modality are epistemic and deontic; other types are three less widespread modalities. In epistemic modality, the speaker indicates that they cannot definitely attest to the truth values of the proposition expressed by the clause. In deontic modality, the speaker evaluates the status of the clause in terms of moral, social, and legal norms. Clearly, this type of modification can apply only to proposal, as opposed to proposition. The advantages of the framing analysis are discussed in some detail concerning the grammatical structures involved in represented speech and thought. What’s more, McGregor (1997: 266) casts doubts on Halliday’s (1985, 1994) Metaphor of Modality. The following examples are taken as illustration. (6)
(a)
I think they will come tomorrow.
(a’)
They will probably come tomorrow.
(b)
I don’t believe that pudding ever will be cooked.
(b’)
That pudding probably never will be cooked.
Halliday takes the former ones (6a) and (6b) as metaphorical variants of the latter ones. McGregor points out that there is no need to invoke metaphor to explain the semantic commonality between the two types of variants. In both cases, interpersonal modification of the status kind of the propositions is involved. Nevertheless, there are differences in the nature of this interpersonal modification: in the former, (6a) and (6b), it is framing, whereas in the latter, (6a’) and (6b’) it is scopal. There is no need to invoke grammatical Metaphor of Modality to explain the similarity in meaning. The marking off of framing is illustrated by the Chinese-box diagram below.
22
2 GM Theory and Interpersonal Metaphor of Modality
I believe
That pudding ever will be cooked not Neg
McGregor’s notion of framing refers to a syntagmatic relationship involving two ‘wholes’, one encompassing the other, indicating that it is to be taken not as a first-order state of affairs, but rather as a second-order ‘framed’ reality, set apart from direct experiential description (Vandelanotte, 2009: 131). However, McGregor (1997: 266) thinks that the above congruent and metaphorical types of variants are roughly “synonymous” and their differences lie in scopal versus framing, which is syntactic in nature. Nevertheless, McGregor’s tackles only one type of modal metaphor—sentences such as I think they will come tomorrow, in comparison with modal adjuncts. As for relational clause with dummy it subject, such as “It was fortunate that he survived the crash”, it is analyzed by McGregor (ibid: 242) as the scoping clause. This type of clause is treated as objective modal metaphor in SFL. This is where invites further clarification. In a word, McGregor’s proposal of conjugational relationships is illuminating with respect to syntagmatic relationship, though his questioning on interpersonal Metaphor of Modality is not completely grounded.
2.5 The Problems with GM Theory and MM Studies In a sum, grammatical metaphor is a further layer of system networks to complement the existing networks of transitivity, mood and modality, theme and rheme to make SFL model of grammar adequate. While grammatical metaphor has been interpreted as one of the most significant innovations in SFL (Martin, 1992: 490), its inherent problems have been pointed out by a number of scholars. Grammatical metaphor is built on the assumption that there is congruent mapping between lexicogrammatical units and semantic units. Behind the concept of ‘congruency’ is the assumption that semantic categories like ‘process’ and ‘entity’ are constant and for each semantic category, there is a corresponding grammatical category to map with it. That is why this mapping is called congruent mapping. Doubts persist on drawing a clear line between metaphor and congruency. What makes things more intrigue is that mapping identified by the systemic notion to be metaphorical may, in practice, have been entrenched and become ‘congruent’, as proposed by Holme (2003). Take ‘education’ for an example: it is metaphorical in the systemic notion, as it combines the semantic features of process and entity. However, according to word frequency data of COCA, ‘education’ has a frequency of 113,731, which is 10 times higher than that of its ‘congruent’ counterpart ‘educate’, a token frequency of 10,171. The expression ‘education’ is so entrenched that it can hardly
2.5 The Problems with GM Theory and MM Studies
23
be said to be incongruent. The same trend is observed in attention versus attend and service versus serve. Chinese scholars also question Halliday’s vague definition of ‘congruence’ from time to time (Hu, 2000). The root cause for the weaknesses of GM theory is also explored: Halliday is criticized for being too focused on the lexicogrammatical level without illuminating the difference in meaning between the congruent and the metaphorical (Hou, 2008). Metaphors of Modality is a case in point. What is exactly the semantic difference between the congruent and the metaphorical modes of modality except for implicit versus explicit orientations? This remains clouded. Amid the criticism on GM theory, its possible dissolution is proposed by a Chinese scholar, who has extended Fawcett (2006, 2008)’s challenges on this theoretical construct. The researcher Wang (2013) approaches the dissolution of the grammatical metaphor theory from two levels: term level and theoretical level: (i) core terms like grammatical metaphor and congruity, with their theoretical problems, could be replaced by some other more accurate terms; (ii) the detailed study of the meaning level can result in the deconstruction of grammatical metaphor theory. A shortcoming of this article is that she does not propose exactly what theoretical notions could replace “grammatical metaphor” and “congruence” and how the meaning level can be described in a more detailed way. In addition, contrary to what she has found, Halliday (1999) does provide some theoretical criterion for congruency, in terms of the human semogenic process. In systemic functional grammar, the congruent form of modality is an adjunct to a proposition. In the explicitly subjective and explicitly objective forms of modality, “the speaker construes the proposition as a projection and encodes the subjectivity (I think), or the objectivity (it is likely), in a projecting clause” (Halliday & Matthiessen, 2004: 615). These forms “represent the modality as being the substantive proposition” and therefore are metaphorical (ibid: 624). For one thing, whether a projecting clause such as I think or It is likely represents “a proposition in its right” (ibid: 615) is dubious. The above proposed motivation for what is metaphorical is not unproblematic. For the first-person psychological predicates (I think) followed by a clause, in Modern English, the expected hierarchical relationship is reversed, with the original matrix clause having a subsidiary role and the original subordinate clause being the proposition proper (Benveniste, 1971: 228–29; Brinton, 1996: 244). That [I Mental-V] like I think has lost its status as a main clause can be shown by the tag: (7)
a
I think it’s going to rain, isn’t it?
b*.
I think it’s going to rain, don’t I?
Utterance (7a) is acceptable and the proposition is in fact ‘it’s going to rain’, whereas (7b) is hardly acceptable. Clause P (it’s going to rain) becomes the proposition proper, while I think takes a subsidiary role. Therefore, one could argue that we should accept that ‘I think’ now takes a subsidiary role and is not the proposition proper and it represents modality. Why such expressions should be construed as metaphorical in present-day English?This categorization of Metaphor of Modality
24
2 GM Theory and Interpersonal Metaphor of Modality
heavily relies on traditional linguistic concepts such as constituency, including the matrix/main clause, for explanation. For another, the underlying assumption here is that for modality there is one corresponding form, that is, an adjunct to a proposition. In fact, the construct of modality is very complex, encompassing semantically and functionally distinctly categories. The heterogeneity of the linguistic phenomenon under the heading of modality makes it unlikely that one form type could be designated as the congruent form of modality. Although Halliday has justified the conception of ‘congruence’ in that it could be located in the inherent directionality of the human semogenic process in the three time frames, i.e., phylogenesis, ontogenesis, and logogenesis, all the evidence and illustrations are exclusively concerned with ideational metaphor. We can infer from his own words that this justification of ‘congruence’ does not apply to interpersonal metaphor (cf. Sect. 2.1). This shows the directionality of ontogenesis cannot consistently justify the distinction of ‘congruence’ and grammatical metaphor. For Metaphor of Modality particularly, even Halliday (1994) himself considers it “not always possible to say exactly what is and what is not a metaphorical representation of modality”. (p. 355). Modality is defined by Halliday and Matthiessen (2004) as the area of meaning that lies between positive and negative polarity. However, since this definition is conceptualized on the clause level, it does not perfectly apply to modality realized by a nexus of two clauses. In a nutshell, Interpersonal Metaphor of Modality, based on the systemic notion of grammatical metaphor, is built on the assumption that there is congruent mapping between lexicogrammatical units and semantic units. Doubts persist on drawing a clear line between metaphor and congruency. Moreover, the volume of study on Metaphor of Modality is comparatively very small and leaves lots of questions unsettled.
2.6 Summary In this chapter, I have reviewed the previous literature around GM theory and especially interpersonal Metaphor of Modality and other work particularly addressing this linguistic phenomenon. In answer to the questions raised on grammatical metaphor theory and interpersonal Metaphor of Modality, I propose that these problems can be solved with insights from the theoretical framework of cognitive construction grammar. It remains to be explored how construction grammar accounts for MM with fine-grained corpus data, yet it is promising that constructions as ‘vertical’ structures that combine semantic, discourse-functional, and morphosyntactic information will enable an adequate description of this linguistic phenomenon. In systemic functional grammar, language is modeled to have various levels, viz. strata as metalanguage. According to Halliday (2004), in language as the higher order
References
25
stratified semiotic, meaning is created across a semiotic space which is defined “by the semantic stratum and the lexicogrammatical stratum, a new, purely abstract level of semiotic organization which interfaces only with the two material interfaces” (P. 54). The semantic stratum is language interfacing with the eco-social world, and the stratum of lexicogrammar is in between the semantic one and the expression one. That is why Halliday needs to bring in the notions of congruency and metaphor to account for the stratal relationship between semantics and lexicogrammar. Halliday’s argument for interpersonal Metaphor of Modality is based on his presupposition: a projecting clause like I think or it is likely represents a proposition. This claim seems to be rather dubious, as revealed in the previous section. In addition, very few SFL study has successfully illuminated the difference in meaning between the congruent and the metaphorical modes of modality except for a rather vague notion of explicitness. Targeting these problems concerning Metaphor of Modality and GM at large, I formulate some general research questions: How can we explicate the rather controversial notion of “congruency”, possibly from an alternative constructional approach? How can we account for the ‘congruent’ and ‘metaphorical’ modes of modality in terms of semantics, discourse functions, and morphosyntax? Which factors motivate interpersonal MM? In the next chapter, I develop a usage-based construction grammar model to account for interpersonal Metaphor of Modality, coupled with a corpus-linguistic methodology.
References Banks, D., et al. (2003). The evolution of grammatical metaphor in scientific writing. In S. Vandenbergen (Ed.), Grammatical metaphor: Views from systemic functional linguistics (pp. 127–148). John Benjamin Publishing Company. Barcelona, A. (2003). Clarifying and applying the notions of metaphor and metonymy within cognitive linguistics: An update. Metaphor and Metonymy in Comparison and Contrast, 208– 277. Brinton, L. J. (1996). Pragmatic markers in english: Grammaticalization and discourse functions. Mouton de Gruyter. Cong, Y., & Wang, H. 丛迎旭、王红阳. (2013). Recategorization of ideational grammatical metaphors based on semantic change基于语义变化的概念语法隐喻模式与类型. Modern Foreign Languages 现代外语 , 1, 33–39. Derewianka, B. (1995). In Language development in the transition from childhood to adolescence: The role of grammatical metaphor. Macquarie University, PhD thesis. Divasson, L., & León, I. (2006). Metaphors in English, French, and Spanish medical written discourse. In Brown, Keith et al. (Eds.), Encyclopedia of language and linguistics (2nd ed., pp. 6517–21). Elsevier Ltd. Fawcett, R. P. (2006). Establishing the grammar of” typicity. English: An Exercise in Scientific Inquiry, 159–262. Fawcett, R. P. (2008). In Invitation to systemic functional linguistics through the Cardiff grammar: An extension and simplification of Halliday’s systemic functional grammar. Equinox.
26
2 GM Theory and Interpersonal Metaphor of Modality
Feng, Z. 封宗信. (2011). The system of Modality in systemic functional linguistics: Logic, semantics, and pragmatics 系统功能语言学中的情态系统: 逻辑、语义、语用. Foreign Language Education 外语教学 , 6, 1–5. Geeraerts, D. (2010). Theories of lexical semantics. Oxford University Press. Halliday, M. A. K. (1978). Language as social semiotics. Edward Arnold. Halliday, M. A. K. (2004). The language of science. Continuum. Halliday, M. A. K., & Matthiessen, C. (1999). Construing experience through meaning: A languagebased approach to cognition. Continuum. Halliday, M. A. K., & Matthiessen, C. (2004). An introduction to functional grammar. Hodder Arnold. Halliday, M. A. K., & Matthiessen, C. (2014). Halliday’s introduction to functional grammar. Routledge. Halliday, M. A. K., et al. (1987). Language and the order of nature. In N. Fabb (Ed.), The linguistics of writing: Arguments between language and literature (pp. 35–154). Manchester University Press. Halliday, Michael A. K. (1989). In Some grammatical problems in scientific English. ERIC Clearinghouse. Halliday, M. A. K. (1993). The analysis of scientific texts in English and Chinese. In Halliday and Martin (Eds.), Writing science: Literacy and discursive power. Falmer. Halliday, M. A. K. (1998). Things and relations: Regrammaticizing experience as technical knowledge. In J. R. Martin & Robert Veel (Eds.), Reading science: Critical and functional perspectives on discourses of science. Routledge. Halliday, M. A. K. (1999).The grammatical construction of scientific knowledge: The framing of the english clause. In R. R. Favretti et al. (Eds.), Incommensurability and translation: Kuhnian perspectives on scientific communication and theory change. Edward Elgar. Holme, R. (2003). Grammatical metaphor as a cognitive construct. In Simon-Vandenbergen et al. (Eds.), Grammatical metaphor: Views from systemic functional linguistics (pp. 391–416). John Benjamin Publishing Company. Hou, J. 侯建波. (2008). From the stratum of semantics to stratum of lexicogrammar: On Fawcett’s questioning of Grammatical Metaphor 从语义层到词汇语法层: 评Fawcett对语法隐喻的质 疑. Foreign Languages and Their Teaching 外语与外语教学 , 5, 5–8. Hu, J. (2015). A comparative study of China English and Singapore English: The case of grammatical metaphor in academic discourse. Journal of World Languages, 1, 50–61. Hu, Z. (胡壮麟) (2000). 语法隐喻的韩礼德模式. Foreign Languages and Their Teaching, 2, 88–94. Li, X.-Q. 李向秋. (2007). Q principle. Quantity principle and implication of objectification in metaphors of modality Q原则、数量原则与情态隐喻的客观化含义. Journal of Sichuan International Studies University 四川外语学院学报, 3, 72–74. Liardét, C. L. (2018). ‘As we all know’: Examining Chinese EFL learners’ use of interpersonal grammatical metaphor in academic writing. English for Specific Purposes, 50, 64–80. Liu, C. 刘承宇. (2007). In A functional-cognitive stylistic approach to grammatical metaphor: A case study of english metalinguistic texts语法隐喻的功能-认知文体学研究——以英语元语 言语篇为例. PhD dissertation, College of Foreign Languages and Cultures, Xiamen University. Liu, C. 刘承宇. (2005). Rankshift in opposite direction between ideational metaphor and interpersonal metaphor 概念隐喻与人际隐喻级转移的逆向性. Foreign Language Teaching and Research 外语教学与研究, 4, 289–293. Martin, J. R., & White, P. R. R. (2005). The language of evaluation: Appraisal in English. Palgrave Macmillan. Martin, J. R. (1992). In English text: System and structure. John Benjamins Publishing. Martin, J. R. (1997). Analyzing genre: functional parameters. In F. Christie & J. R. Martin (Eds.), Genre and institutions: Social processes in the workplace and school (pp. 3–39). Continuum. Matthiessen, C. M. et al. (2016/2010). In Key terms in systemic functional linguistics. Foreign Language teaching and Research Press. McGregor, W. B. (1997). Semiotic grammar. Clarendon Press.
References
27
Painter, C. (1993). Learning language: A functional view of language development. In R. Hasan & J. R. Martin (Eds.), Language development: Learning language, learning culture. Ablex. Quirk, R., et al. (1985). A comprehensive grammar of the english language. Longman. Ravelli, L. (1988). Grammatical metaphor: An initial analysis. In E. H. Steiner & R. Veltman (Eds.), Pragmatics, discourse and text: Some systemically-inspired approaches (pp. 133–147). Ablex. Ravelli, L. J. (1985). In Metaphor, mode, and complexity: An exploration of co-varying patterns. B.A. thesis, Department of Linguistics, University of Sydney. Taverniers, M. (2008). Interpersonal grammatical metaphor as double scoping and double grounding. Word-Journal of the International Linguistic Association, 59, 1–31. Taverniers, M. (2003). Grammatical metaphor in SFL. In Simon-Vandenbergen et al. (Eds.), Grammatical metaphor: Views from systemic functional linguistics (pp. 5–31). John Benjamin Publishing Company. Thompson, G. (1996/2004). In Introducing functional grammar. Routledge. Torr, J., & Alyson, S. (2003). The emergence of grammatical metaphor. In Simon-Vandenbergen et al. (Eds.), Grammatical metaphor: Views from systemic functional linguistics (pp. 169–183). John Benjamin Publishing Company. Turner, M. (1991). Reading minds: The study of english in the age of cognitive science. Princeton University Press. Vandelanotte, L. (2009). Speech and thought representation in english: A cognitive-functional approach. Mouton de Gruyter. Wang, F. 王馥芳. (2013). Grammatical metaphor theory: Possible deconstruction语法隐喻理论可 能“消解”论. FLLTP 外语教学理论与实践, 1, 22–29. Wei, Z. 魏在江. (2008). Metaphor s of modality in English and Chinese: A contrastive study基于 功能的英汉语情态隐喻对比研究. Modern Foreign Languages 现代外语 , 3, 264–272. Yang, S. (2021). In A systemic functional study of modality in modern Chinese. Springer Nature Singapore Pte Ltd. Zhang, D., & Zhao, J. 张德禄、赵静. (2008). On the formal similarity principle in ideational grammatical metaphor between congruent and metaphorical forms论语法概念隐喻中一致式 与隐喻式的形似性原则. Journal of Foreign Languages 外国语, 6, 25–32. Zhang, D., & Dong, J. 张德禄、董娟. (2014). On the developmental models of grammatical metaphor语法隐喻理论发展模式研究. Foreign Language Teaching and Research 外语教学 与研究, 1, 32–44. Zhou, D. 周大军. (2003). Metaphors of modality in English and its explanatory power of discourse 英语的情态隐喻及其语篇解释力. Journal of Sichuan International Studies University 四川 外语学院学, 6, 100–104.
Chapter 3
Towards a Constructional Approach to Metaphor of Modality
This chapter will discuss the need to incorporate the cognitive constructional approach to account for (interpersonal) Metaphor of Modality and how this theoretical approach can solve some problems of grammatical metaphor and adeptly describe the linguistic phenomenon brought to attention by MM. Section 3.1 enumerates general research questions in response to these problems. Section 3.2 briefly surveys major constructional approaches to grammar, as not a few of my research ideas are inspired by these approaches. The rest of the chapter gradually builds up the argument for the constructional approach to MM, followed by the presentation of the corpus-linguistic method.
3.1 General Research Questions In systemic functional grammar language is modeled to have various levels, viz. strata as metalanguage. According to Halliday (2004), in language as the higher order stratified semiotic, meaning is created across a semiotic space which is defined “by the semantic stratum and the lexicogrammatical stratum, a new, purely abstract level of semiotic organization which interfaces only with the two material interfaces” (P. 54). The semantic stratum is language interfacing with the eco-social world, and the stratum of lexicogrammar is in between the semantic one and the expression one. That is why Halliday needs to bring in the notions of congruency and metaphor to account for the stratal relationship between semantics and lexicogrammar. Unfortunately, these notions invite substantive criticism among linguists. As revealed in Chap. 2, Halliday’s argument for Metaphors of Modality based on somewhat dubious presumption. There is a pressing need to illuminate the differences in meaning between the congruent and the metaphorical modes of modality.
© Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7_3
29
30
3 Towards a Constructional Approach to Metaphor of Modality
Targeting these problems concerning Metaphor of Modality and GM at large, we formulate some general research questions as follows: i. How can the form-meaning mappings of modality be theoretically approached without the controversial notion of ‘congruency’? ii. What is the semantic difference between the ‘congruent’ and ‘metaphorical’ modes of modality? iii. Which factors motivate interpersonal Metaphor of Modality?
3.2 Constructional Approaches to Grammar Different from SFG, a cognitive model of grammar takes a form-meaning pairing or symbolic unit as the fundamental unit of grammar. It is called ‘construction’ in construction grammar. An expression that once has been used sufficiently frequently and has become entrenched, acquiring the status of a cognitive routine will become a symbolic unit. From this perspective, a unit is a symbolic entity that is stored and accessed as a whole. Furthermore, the symbolic units represented in the speaker’s grammar are conventional. The conventionality of a linguistic unit relates to the idea that linguistic expressions become part of the grammar of a language by virtue of being shared among members of a speech community. A symbolic unit, viz. a form-meaning pairing, is stored and accessed as a whole. Moreover, constructions as form-meaning pairings are ‘vertical’ structures that combine semantic, syntactic, and even phonological information (Croft & Cruse, 2004: 247). The semantic interpretation rule that maps the syntactic structure to its corresponding semantic structure is unique to that schematic construction. As an illustration, representations in Goldberg’s construction grammar model involve at least two levels: SEM and SYN. For instance, the ditransitive construction is defined as syntactic [SUBJ V OBJ1 OBJ2] linked with an agent understood to cause or intend transfer of possession, which is shown as follows (cf. Goldberg, 2006: 20): SEM: SYN
Intend-CAUSE-RECEIVE
Agt
Rec
Theme
|
|
|
|
Verb
SUBJ
OBJ1
OBJ2
This approach renders the differentiation of what is congruent and what is metaphorical mode of expression unnecessary. We could avoid the vague lotion of ‘congruency’ in Grammatical Metaphor theory. Among constructional approaches to grammar, two current theories of construction grammar will be explicated in detail because of their likely insights into my research project. They are cognitive grammar and cognitive constructional grammar.
3.2 Constructional Approaches to Grammar
31
3.2.1 Cognitive Grammar as a Construction Grammar Cognitive grammar is a theory of linguistic structure that has been developed by Langacker since 1976. Cognitive grammar is a detailed, carefully worked-out theory of syntax and semantics. In the cognitive grammar representation of a construction, the symbolic unit itself must link the form and meaning of the construction (Croft & Cruse 2004: 279). Langacker describes the link as a symbolic correspondence. Langacker views any unit with complex symbolic structure as a construction, regardless of whether its structure or meaning can be predicted from the properties of its subparts. This research paradigm emphasizes appropriate semantic tools, without which grammar cannot be effectively studied. In cognitive grammar, meaning is equated with conceptualization, consisting of both conceptual content and a particular way of construing that content (Langacker, 2008). Langacker (1999a) views grammar largely as a matter of imposing and symbolizing particular ways of “construing conceptual content” (P. 46). Cognitive grammar belongs to the overall enterprise of ‘natural grammar’. There is a natural affinity between cognitive grammar and several contemporary school of linguistic research, especially the functional tradition. For Langacker and his followers, the study of grammar is an important stimulus and source of evidence for semantic investigation. Therefore, it is considered most productive for semantic and grammatical investigation to proceed in parallel as mutually dependent facets. According to Langacker (1987), grammar is symbolic in nature, consisting in the conventional symbolization of semantic structure rather than constituting an autonomous formal level of representation. Langacker (1987) considers the nature of meaning and how to deal with it as the most fundamental issue in linguistic theory. As meaning is seen as a cognitive phenomenon in cognitive grammar, it is equated with conceptualization or mental experience, explicated as cognitive processing. A detailed and explicit account of the semantic component is considered fundamental to adequate description and understanding of grammar. A conceptual structure that functions as the semantic component of a linguistic expression is termed as a semantic structure. It is conceptualization tailored to the specifications of linguistic convention. Conceptualization, mental experience, and cognitive processing are proper concerns of semantic analysis in cognitive linguistics. Cognitive grammar embraces a ‘subjectivist’ view of meaning (Langacker, 1988: 6). The semantic value of an expression crucially involves the way subjects of conceptualization choose to think about this entity or situation and mentally portray it. Central to Langacker’s model is the language user’s ability to construe a conceived situation in alternate ways by means of alternate images for the purpose of thought or expression, which is fundamental to lexical and grammatical variability (Langacker, 2009: 6). In choosing a particular expression or grammatical form, a speaker construes the conceived situation in a certain way. The same situation can be described by a variety of semantically distinct expressions that embody different ways of construing or structuring it.
32
3 Towards a Constructional Approach to Metaphor of Modality
Construal involves perspective, specifically vantage point and direction of mental– scanning. The overall viewing arrangement, vantage point, the viewing asymmetry between the subject and object of conception, and the sequence of mental access are all subsumed under perspective (Langacker, 2008: 1). Within Langacker’s (1987: 123) conceptual framework of perspective, viewpoint subsumes the two more specific notions, vantage point and orientation. In the default arrangement, the vantage point is the actual location of the speaker and hearer (Langacker, 2008: 26). A particular choice of vantage point imposes on the scene an alignment into foreground and background. Numerous scholars (e.g., Langacker, 1987; Talmy, 1978) have accorded the importance of the perspective taken on a scene to semantic and grammatical structure. In Langacker’s model, subjectification relates to perspective. For example, speaker and hearer are usually subjectively construed or ‘off-stage’, and only become objectively construed or ‘on-stage’ when linguistically profiled, for example by the use of pronouns such as I or you. Langacker argues that subjective construal is immanent in (subsumed by) objective construal, because whether or not the conceptualizer is on-stage (objectified), his or her perspective in terms of participation in scanning is part of the conceptualization process. The idea behind immanence is that subjective construal is there all along, but only comes to the fore when objective construal is attenuated or weakened (c.f. Evans & Green, 2006: 729). According to cognitive grammar, the full conceptual or semantic value of a conceived situation is a function of not only its content, but also how we structure this content with respect to such constructs as attention, selection, perspective, and level of schematicity. In this structuring, we are capable of making adjustments, thereby transforming one conceptualization into another that is roughly equivalent in terms of content but differs in how this content is construed. Many of Langacker’s ideas are foundational to the concept of constructions as a network of symbolic units and the cognitive approach I adopt in this research project, as will become clear in subsequent chapters.
3.2.2 Cognitive Construction Grammar Goldberg’s constructional approach has come to be known as cognitive construction grammar (henceforth referred to as CCxG) since the publication of her book Constructions at Work in 2006. Goldberg (1995, 2006) built a construction grammar on the patterns she found on sentences. It embraces the structuralist notion that each construction is understood as one potential option among others. The meaning assigned to one construction on this view is not necessarily an inherent aspect of the form the construction takes, but may be in part a pragmatic inference made upon the rejection of other alternatives (Levinson, 1983). CCxG seeks to provide motivation for each construction that is posited (Goldberg & Suttle, 2010). There has been a great deal of attention paid to marked constructions within the theory.
3.2 Constructional Approaches to Grammar
33
The central thesis of Goldberg’s theory is that “constructions themselves carry meaning, independently of the words in the sentence” (Goldberg, 1995: 1). According to Goldberg (2006: 5): Any linguistic pattern is recognized as a construction as long as some aspect of its form or function is not strictly predictable from its component parts or from other constructions recognized to exist.
Goldberg also adds that patterns “are stored as constructions even if they are fully predictable as long as they occur with sufficient frequency” (p. 5). Since establishing what level of frequency is sufficient for pattern storage and entrenchment is problematic, Traugott and Trousdale (2013:5) tackle ‘sufficient frequency’ with replication and conventionalization in the textual record. Given that the central status of constructions blurs the boundaries between lexicon and syntax, Goldberg assumes the lexicon-grammar continuum. CCxG emphasizes that there are often interactions between parts that lead to emergent properties that can only be described at the level of the whole. Representations in Goldberg’s model involve at least two levels: SEM and SYN. For instance, the ditransitive construction is defined as syntactic [SUBJ V OBJ1 OBJ2] linked with an agent understood to cause or intend transfer of possession, which is shown as follows (cf. Goldberg, 2006: 20): SEM: SYN
Intend-CAUSE-RECEIVE
Agt
Rec
Theme
|
|
|
|
Verb
SUBJ
OBJ1
OBJ2
Relationships between constructions are captured in terms of motivation and inheritance. Four different kinds of inheritance links—polysemy links, subpart links, instance links, and metaphorical extension links—are discussed by Goldberg (cf. Croft & Cruse, 2004: 273–5). Similarities between constructions play a significant role in this model. The semantic structure of a construction is assumed to be potentially complex, made up of semantic components among which certain semantic relations hold. Because they specify a surface form and a corresponding function, constructionist approaches provide a direct way of accounting for these facts. Subtle facts about the functions of constructions are emphasized, including their semantic properties, their information structure or discourse properties, and their conditions of use, such as register and genre (Goldberg & Suttle, 2010: 469). Goldberg (1995) claims certain aspects of the meaning of a sentence, as well as certain restrictions upon its structure, arise directly from the properties of the skeletal grammatical construction rather than from the properties of the lexical verb. Goldberg (2006: 98) hypothesizes that speakers have not only item-specific knowledge about particular expressions, but also generalized or schematic knowledge about them.
34
3 Towards a Constructional Approach to Metaphor of Modality
3.3 Subjectivity from a Constructional Perspective 3.3.1 Modality and Subjectivity Halliday’s differentiation of subjective and objective orientation in modality is in a way more or less equivalent to Lyons’s (1977) distinction of subjective and objective modality, though Halliday does not clarify it. A starting point for thinking about subjectivity is Lyon’s definition of subjectivity: The term subjectivity refers to the way in which natural languages, in their structure and their normal manner of operation, provide for the locutionary agent’s expression of himself and his own attitudes and beliefs. (Lyons, 1982: 102)
The key distinction between subjective and objective modality is mainly discussed in terms of epistemic modality, which is concerned with the interpretation of the world via the laws of human reason, with the expression of the degree or nature of the speaker’s commitment to the truth of what he says, roughly corresponding to Halliday and Matthiessen (2004)’s probability and usuality modality types. According to Lyons (1977: 797), while objective epistemic modality expresses an objectively measurable chance that the state of affairs under consideration is true or not, subjective epistemic modality involves a purely subjective guess regarding its truth. Subjective epistemic modality can be accounted for in terms of the speaker’s qualification of the I-say-so component of his utterance. Objective modality, in contrast, can be described as having an it-is-so component that expresses an objectively measurable chance for the truth of the state of affairs under consideration (Lyons, 1977: 800; Nuyts, 2006: 13). Nevertheless, for all its keenness and wealth of observations, Lyons’ discussion still does not lead to a thorough understanding of the differences between subjective and objective modalities. The dimension of subjectivity is not limited to epistemic modality. Lyons (1977) argues that it also applies to deontic modality. Lyons’ definition of subjectivity versus objectivity also draws on differences in the quality of the evidence leading to the judgment. Heltoft (2005: 86) considers subjectivity in terms of ‘modal factor position’: subjectivity refers to modal factor location in the speaker, whereas objectivity refers to modal factor position different from the speaker. Lyons has not suggested a link between his modality types and the modal adverbs versus modal adjectives issue, but the connection is made by Hengeveld (1988), who claims that modal adverbs always expressive subjective modality and modal adjectives objective modality. Nevertheless, there are other diverse opinions on the subjectivity/objectivity of the modal adverbs and adjectives: for example, Perkins (1983) assumes that both express objective modality, while Watts (1984) suggests that both express subjective modality. All these illustrate the problems regarding how the distinction between subjective and objective modality is analyzed. Different from Lyons (1977), an alternative way to define the distinction between more and less subjective modal expressions is in terms of who is responsible for the modal evaluation. This distinction is termed subjectivity versus intersubjectivity.
3.3 Subjectivity from a Constructional Perspective
35
According to Nuyts (2006: 14), subjectivity applies if the issuer presents an evaluation as being his/her own responsibility. He prefers to interpret the objective pole of the scale as intersubjectivity, which applies if the issuer indicates that an evaluation is shared with a wider group of people, including himself/herself and possibly the hearer. A first-person subject then codes subjectivity, and an impersonal subject codes intersubjectivity: (8)
a.
I think they have forgotten to take the key. (subjective) SUBJ
b.
It is quite probable that they have forgotten to take the key. (Intersubjective)
This dimension seems structurally present in epistemic and deontic modality, and in boulomaic attitude, but not in dynamic modality. In general, research on modality within functional linguistics and cognitive semantics has been very interested in the notions of subjectivity and intersubjectivity (Coates, 1983; Langacker, 1985, 1999b; Narrog, 2005; Traugott, 1989, 2006). In the last two decades, there have been many studies highlighting subjectivity and subjectification. The former relates to ways in which speakers use language to express their perceptions, feelings, and opinions in discourse; and the latter concerns how such expressive motivations and strategies conventionalize and interact with linguistic structure (Scheibman, 2001: 61) In current linguistic theorizing, the two most prominent exponents of subjectivity are probably Langacker and Traugott who extended it to subjectification according to the premises of their respective theoretical frameworks. Both have treated modal verbs as prominent examples of subjectification and grammaticalization. Subjectivity and intersubjectivity, as synchronic notions, can be theorized in many ways, from cognitive construal to human interaction and the procedure producing and understanding talk (Schifrin, 1990; Traugott, 2010). Langacker’s (1991a) theory, developed within cognitive semantics, positions the notions of ‘subjectivity’ and ‘objectivity’ as an aspect of perspective. Moreover, subjectivity/objectivity is often a matter of degree. Subjectification is defined as a shift from an objective construal of some entity to a more subjective one. His focus is on the event and the role of event participants, and the comparison of related constructions in synchrony constitutes the material basis for analysis. An entity is construed objectively to the extent that it is put onstage as a focus object of conceptualization. The grounding predication, on the other hand, is a combination of the reference-point construction and subjectification (Langacker, 1999b). Langacker’s (1990, 1998, 1999b, 2003) proposals on the subjectification of modal verbs make crucial reference to the force-dynamic model of modal meanings adopted by Talmy (1988), Sweetser (1990) and Johnson (1987), which integrated into his own. The development of the English modals according to Langacker can be summarized as follows. The evolution of the present-day English modals is seen by Langacker (1991: 334) as primarily a matter of progressive subjectification. The modals in English historically derive from main verbs with a purely objective meaning yet force dynamic
36
3 Towards a Constructional Approach to Metaphor of Modality
notions that designate some kind of potency, which is defined as “a physical or mental force” that, once launched, tends to cause the process’s occurrence (ibid: p. 333). The crucial development is that the conception of directed potency, abstract force such as obligation or permission, loses its profiled status, and the source of potency is no longer identified with the grammatical subject, but is implicit and subjectively construed. The source can be the speaker or some individual from the context, but also some general authority. Then the resulting expressions have become grounding predications. The conception of a directed potency loses its profiled status. In the past nearly fifteen years, a number of scholars associated with the framework of cognitive grammar have taken up Langacker’s ideas to various degrees and suggested several modifications. Yet a problem for the cognitive semantic theory of subjectivity is lack of a real pragmatics. Unlike Langacker, Traugott’s approach is historically informed. Traugott’s approach develops the traditional perspective of pragmatics as focusing on meaning which arises from the interaction among speakers, addresses, and the context. Subjectivity is the relevance of the speaker to meaning, and intersubjectivity is the relevance of the speaker-addressee relationship to meaning (Porter, 2009: 131). Then, expressions the prime semantic or pragmatic meaning of which is to index speaker attitude or viewpoint and speaker’s attention to addressee self-image are linguistic markers and expressions of subjectivity and intersubjectivity. Subjectification is the mechanism by which meanings are recruited by the speaker to encode and regulate attitudes and beliefs. ‘Subjectification in grammaticalization’ is the development of a grammatically identifiable expression of speaker belief or speaker attitude to what is said (Traugott, 1995: 32). The development of epistemic meaning is conceived as a typical instance of subjectification in modal meaning, since the evolution deontic > epistemic has often been cited as part of the tendency to recruit meaning into increasingly subjective semantics (Traugott & Dasher, 2002: 149; Narrog, 2010: 391). According to Traugott (1989), with meanings becoming increasingly based on the speaker’s subjective belief state/attitude toward the proposition, the modals acquired epistemic meanings that are primarily focused on the internal world of the speaker’s belief and knowledge states. In a general sense, language can be said to be strongly marked by subjectivity in that any selection from the lexical and/or grammatical repertoire passes through the speaker/writer. In contrast to this general subjectivity, De Smet and Verstraete (2006) distinguish two types of what they call ‘semantic subjectivity’: Following a distinction made by Halliday and Hasan (1976) and Halliday (1985, 1994), one type of subjectivity functions in the ‘ideational’ component of the utterance, and another type functions in the ‘interpersonal’ component. Ideational subjectivity involves the description of a content “situated in the speaker’s subjective belief-state/attitude toward the situation”. It corresponds roughly to Traugott’s (1989: 34) ‘internal propositional’ meaning, that is, meanings based on the evaluative, perceptive, or cognitive described internal situation. Conversely, the term interpersonal provides a useful generalization over the non-ideational types of subjectivity: it refers both to the speaker’s attitudes and judgments and to his or her interaction with the interlocutor.
3.3 Subjectivity from a Constructional Perspective
37
Although the two types are jointly called ‘semantic subjectivity’, the ideational and interpersonal expressions of subjectivity show divergent syntactic behavior. According to Traugott (2010: 31), Halliday’s term ‘interpersonal’ includes both subjective and intersubjective elements. She also matches Halliday’s terminology ‘ideational’ and ‘interpersonal’ to ‘less subjective’ and ‘subjective’/ ‘intersubjective’ respectively, along a continuum. A speech situation not just involves a speaker but an interpersonal relationship between speaker and addressee, and thus it is subjective as well as intersubjective in a general sense. In general, Traugott’s historically based empirical approach and Langacker’s conceptual perspective on subjectification complement each other, as suggested by Evan and Green (2006: 733) and Narrog (2010: 392). Subjectivity and intersubjectivity, which implicate a continuum, are more current and possibly better terms than the dichotomy of subjectivity and objectivity. Different from Lyons (1977) and Halliday (1985, 1994), an alternative way to define the distinction between more and less subjective modal expressions is in terms of who is responsible for the modal evaluation. This distinction is termed subjectivity versus intersubjectivity. According to Nuyts (2006: 14), subjectivity applies if the issuer presents an evaluation as being his/her own responsibility. He prefers to interpret the objective pole of the scale as intersubjectivity, which applies if the issuer indicates that an evaluation is shared with a wider group of people, including himself/herself and possibly the hearer. A first-person subject then codes subjectivity, and an impersonal subject codes intersubjectivity: (9)
a.
I think they have forgotten to take the key. (subjective)
b.
It is quite probable that they have forgotten to take the key. (Intersubjective)
This dimension seems structurally present in epistemic and deontic modality, and in boulomaic attitude, but not in dynamic modality.
3.3.2 Subjectivity and Intersubjectivity from a Constructional Perspective In Verhagen (2005)’s study of intersubjectivity, complementation constructions may be viewed as general grammaticalized expressions for intersubjective coordination. It is apparent that speakers/writers can manage their coordination relationship with the addressee by modifying the content of the matrix clause. What Verhagen’s framework appears to propose is an approach which seeks continuity of underlying construal across different constructions (Dancygier, 2008). Every utterance relies on a construal configuration. Verhagen (2005) has further developed Langacker’s construal theory: the ‘ground’ of any linguistic usage event comprises two conceptualizers, the first performing the role of being responsible for the utterance, the second performing the role of interpreting it in a particular way.
38
3 Towards a Constructional Approach to Metaphor of Modality
O: Object of conceptualization
S: Subject of conceptualization (Ground)
1
2
Fig. 3.1 Construal configuration and its basic elements, (From Verhagen, 2005: 41)
These subjects of conceptualization engage in cognitive coordination by means of utterance, with respect to some object of conceptualization. The ground furthermore contains the knowledge that the two conceptualizers mutually share. The configuration is represented in Fig. 3.1. Verhagen argues that the complement-taking-clauses (referred to by him as CT-clauses) operate at level S and its relation to level O, rather than object of conceptualization. The conceptualizer that is explicitly mentioned in the matrix clause has been called by Verhagen (2005: 97) as the ‘onstage conceptualizer’. For instance, the use of I or we can be seen as foregrounding, or bringing into explicit parts of the ground. In other words, this highlights the ground participant who is saliently invoked in the argument. Some other cases, in contrast, do not contain an explicit onstage conceptualizer, such as the utterance below: (10)
It was obvious that his opponent was closing in
This type is called impersonal complementation constructions, which do not attribute a thought directly to one of the conceptualizers in the ground, as there is no onstage conceptualizer. An important category of CT-clauses of impersonal constructions consists of CT-clauses with copular verbs and a nominal that-clause indicating evaluation or epistemic stance. Examples of copular epistemic CT-clauses include It is likely that…, It is clear that… or It seems that…, It appears that…. Although not immediately obvious, the stance expressed evokes the idea of some conceptualizer, without being anchored to a particular subject of conceptualization. This general character of impersonal complementation construction is represented in Fig. 3.2. Impersonal complementation constructions invite the addressee to entertain the object of conceptualization in a particular way. The difference between the CTclause of an impersonal complementation construction and that of a personal one is that identification of the latter’s subject of conceptualization is constrained linguistically, while this may left completely open in the former. Different communicative
3.3 Subjectivity from a Constructional Perspective
39
O: Object of conceptualization
S: Subject of conceptualization (Ground)
1
2
Fig. 3.2 Construal configuration for impersonal complementation constructions (From Verhagen, 2005: 134)
acts differ in the selection of the aspects of the construal put on stage, but they rely on the same configuration. Complementation constructions differ from direct speech or thought in that the speaker/writer is held responsible for the form and content of a complement clause. According to Verhagen (2005: 114), not just the view of the onstage conceptualizer, but also the argumentative orientation of the author is also an important factor in the way complement clauses contribute to the coherence of discourse. In conclusion, the content of a CT-clause does not designate an object of conceptualization, but rather instructs the addressee to construe it in a particular way and thus to engage in a cognitive coordination with another subject of conceptualization, ultimately the speaker/writer. The basic function of complementation constructions is that of connecting the object-dimension and the intersubjective dimension (Verhagen, 2005: 150). What Verhagen’s (2005) framework suggests is that there are also clusters of constructions which have similar construal-management goals, while using less specific or more varied formal signals (Dancygier, 2008). In other words, a basic mental space configuration can be used in a variety of constructions to achieve, in this case, a similar kind of intersubjective management of construal. As his approach has, in a way, dealt aptly with syntactic constructions such as It is likely that X and I believe X, it can provide valuable insights to my account of metaphoric modality in terms of cognition, syntax, and discourse.
40
3 Towards a Constructional Approach to Metaphor of Modality
3.4 Towards the Constructional Approach to Metaphor of Modality The survey in the above sections shows that cognitive constructional approaches abandon the duality between syntactic and semantic components to represent grammatical knowledge. These approaches emphasize that speakers’ knowledge of language consists of systematic collections of form-function pairings, or constructions, at varying levels of generality and complexity. Constructions are fundamentally symbolic units, as represented in Fig. 3.3. As Fig. 3.3 illustrates, the form of a construction can be associated with linguistic relevant information, be it syntactic, morphological, or phonological. The term ‘meaning’ encompasses all of the conventionalized aspects of a construction’s function, which may include not only properties of the situation described by the utterance, but also properties of the discourse context and of the pragmatic situation of the interlocutors (Croft & Cruse, 2004: 258). The notion that constructions are regarded as learned pairings of form and meaning has important implications for the architecture of grammar. Several themes explicate how constructional knowledge is organized in cognitive construction grammar (CCxG); they are motivation, constructional taxonomies, plus frequency, and productivity. Above all, the existence of any construction in CCxG is thought to be largely motivated by properties of human cognition and interaction. Motivation aims to explain why it is at least possible and at best natural that a particular form-meaning correspondence should exist in a given language. Functional and historical generalizations count as explanations. Motivation is provided for each construction that is posited. Motivation is also often used to account for the fact that formally similar constructions are also often semantically similar, in other words, semantic and formal overlaps of constructions. Fig. 3.3 Symbolic structure of a construction (Croft & Cruse, 2004: 258)
3.4 Towards the Constructional Approach to Metaphor of Modality
41
Our knowledge of language is understood to form a network of interrelated constructions; these constructions are related to one another via a variety of inheritance links, based on the idea that constructional organization follows similar principles as conceptual categories (Goldberg & Suttle, 2010: 468; Boas, 2013: 245). Only once an expression has been used sufficiently frequently and has become entrenched, acquiring the status of a cognitive routine, it will become a symbolic unit. From this perspective, a unit is a symbolic entity that is stored and accessed as a whole. Furthermore, the symbolic units represented in the speaker’s grammar are conventional. The conventionality of a linguistic unit relates to the idea that linguistic expressions become part of the grammar of a language by virtue of being shared among members of a speech community. In CCxG, frequency is important for accounting for a construction’s productivity. CCxG takes a strong usage-based view of frequency role and item-specific instance status. Determining a construction’s type frequency is important because it has been found that increased type frequency directly correlates with a construction’s ability to occur with novel items. Other important aspects influencing a construction’s productivity are token frequency, degree of openness, and statistical pre-emption (Goldberg, 2006: 93). More recently, “productivity is best regarded as a function of type frequency, semantic coherence, and the inverse correlation between the two” (Barðdal, 2008: 173). This section initiates the argument for the cognitive constructional approach to account for interpersonal Metaphor of Modality proposed by Halliday (1985, 2004). In his account, MM falls into two types: explicit subjective modality and explicit objective modality. Typical examples in turn are I think (that) Mary knows and It is likely (that) Mary knows. I will focus on these two types, but as subtypes of construction, each as one potential option among variant modes of modal expression. Treating the ‘metaphorical’ realization of modality by a clause that projects or embeds the clause to which a modal value is assigned (Halliday, 1985, 1994) as constructions has the following advantages: with unique semantic interpretation rules for any independent construction, their semantic differences and syntactic distinction can be adequately described and explained. There is no need to bring in what is congruent realization of modality and then identify metaphorical modes of modal expression, which has been found to be theoretically clumsy and problematic, as our critical review of grammatical metaphor in Chap. 2 shows. Besides the classical examples I think (that) Mary knows and It is likely that Mary knows, there are quite lot members of both types, including those in (11a–11f) and (12a–12f) culled from COCA: (11)
a.
I believe that he is a liberal.
b.
I feel that my privacy has been violated.
c.
I guess that when you go to Harvard, as Barbara did, you tend to pick up connections.
d.
I find that adding in a bit of pleasure tends to make work both more enjoyable and more successful, Ms. D’Alessandro.
42
3 Towards a Constructional Approach to Metaphor of Modality
e.
I suspect that there’s nothing unconstitutional about it.
This set of constructions can be skeletally represented as [I VMental that P], where VMental is short form for Mental Verb and P codes the proposition expressed by the complement that clause. Verbs such as believe, feel, and think are concerned with our experience of the world of our own consciousness, and they are categorized as Mental Verbs. Mental Verbs are also referred to as private verbs (Biber, 1988), psychological verbs and psychological predicates (Leech, 1983), and cognitive verbs (Fetzer, 2014). For the purpose of this study, I adopt Diessel and Tommsello’s (2001) definition of the term proposition as “the meaning or content of a clausal construction denoting some state of affairs (e.g., a state, an event, or an activity)” (p. 101). (12)
a.
it is essential that the nurse fully inform parents of both the risks and the appropriate preventative measures
b.
it is clear that such groups are not fully under the government’s control
c.
it is important that people are able to talk about it
d.
It’s likely that my genes predisposed me to have a mood disorder
e.
it was imperative that we got a chance to hear it again
f.
It’s great that “Newsweek” has decided to do the story
From the constructional approach, the set of constructions in (12) will be represented as [It be ADJ that P] construction. In general, constructional approaches provide a direct way of accounting for subtle facts about both surface forms and corresponding functions, including their semantic properties, their discourse properties, and their conditions of use such as genre. The parts of the syntactic structure are called elements and parts of the semantic structure are components. A symbolic link joins a syntactic element of a construction to a semantic component of that construction. There is also a symbolic link joining the whole syntactic structure to the whole semantic structure. There are two other relations apart from the semantic link: one semantic relation that holds between semantic components and one syntactic relation that joins syntactic elements. More broadly, much work in cognitive and functional linguistics was concerned with the identification of meaningful syntactic structures, viz. constructions, and characterization and explanation of their semantic and distributional properties. Given the assumed close connection between syntax and semantics, many studies in this field routinely use syntactic arguments to support semantic claims and vice versa (Gries, 2006). Semantic aspects receive focused attention in cognitive linguistics. One aim of the current project is to achieve a more fine-grained analysis of semantic structure, accounting for the subtle facts about similarities and distinctions among variant constructions of modality.
3.5 The Usage-Based Model
43
To sum up, the constructional meaning, including not only its semantic structure but also its discourse and pragmatic functions, will allow us to account for the similarities and differences in the syntactic and behavioral properties of variant modal expressions. Based on this approach, the current study can set out to answer a number of specific research questions: i. What motivates the use of modal clausal constructions like I think or it is likely from cognitive perspective? What is the semantic difference between clausal modal constructions, namely, [It be ADJ that P] construction and [I VMental that P] construction, and modal adjuncts? ii. Are there any other functional explanations of the particular form-meaning correspondence in either construction in addition to semantic accounts? iii. What sorts of relations can be found between these constructions and beyond?
3.5 The Usage-Based Model Like most cognitive constructional approaches, the current study adopts a usagebased model for language use, in which patterns of language use are taken as evidence for the independent representation of grammatical information. The usage-based model is a model of grammatical representation in which language use determines grammatical representation. The roots of this view of language can be found with the functional-typological approach to language starting in the 1960s. This approach encapsulates the usagebased premise that linguistic structure is formed by repetition of certain linguistic patterns in language use. The importance of repetition to grammaticization has been emphasized in Haiman’s (1994) discussion of the parallels between the general cultural phenomenon of ritualization and the process of grammaticization in language and in Boyland’s (1996) examination of the effects of repetition on the cognitive representation of grammaticizing constructions. Decades later cognitive approaches enrich the usage-based proposal by the recognition of the cognitive processes that feed into linguistic structure, such as categorization, construal, entrenchment, and schematization. Specifically, frequency of use and similarity of form and meaning are the determining factors for the structure of grammatical knowledge in the mind. Following the usage-based model, my study allows both instances and generalizations to be captured in terms of fully articulated schematic networks including low-level schemas (Barlow & Kemmer, 2000; Boas, 2013; Langacker, 1987). In this view, patterns occurring with sufficient frequency are possibly stored as constructions alongside more general linguistic generalizations. According to Bybee (2013), exemplar representation is central to a usage-based approach to language, since it reflects the way that linguistic structure emerges when human cognition deals with tokens of experience with the world. Exemplars are categories formed from tokens of experience that are judged to be the same (Pierrhumbert, 2001). Exemplars may differ in strength depending upon the number of tokens that
44
3 Towards a Constructional Approach to Metaphor of Modality
comprise them. The exemplars themselves are grouped together by similarity. My study takes an exemplar view of constructions. As claimed by Bybee (2006, 2013), an exemplar usage-based model is essential to explaining how constructions acquire idiosyncratic semantic, pragmatic, and phonetic properties. Another role for exemplars in a construction’s representation is seen in the way the schematic slots in constructions expand. A schematic slot in a construction might consist of a list of all the items that have occurred in that slot. Or a slot might be considered a set of abstract semantic features that constrains the slot, as usually proposed. For instance, as found by Goldberg (1995) and Israel (1996), the set of lexical items that can occupy a slot in a construction may be constituted of two or more clusters of closely related items. The semantic or formal properties of a slot determine its schematicity. Still another important argument for an exemplar model of linguistic representations is that exemplar, by their very nature, provides a record of token and type frequency in linguistic experience for constructions. In CCxG, frequency is also important for accounting for a construction’s productivity, because type frequency, which counts how many different items occur in the schematic slots of constructions, relates directly to productivity. High type frequency correlates to productivity partly in that the greater the number of types in a construction, the more cases there are for the item-based analogy that creates novel instances of the construction. Linguists like Bybee (2006, 2013) prefer to keep the effects of token frequency and type frequency separate, since they are found to have distinct effect. In an exemplar model, tokens of words or constructions are represented in memory along with the situations they have been associated with in experience. The usage-based model allows grammar to accommodate dynamic aspects of language. Seen from the aspect of ontogenesis, children begin with very narrow construction types, even specific to individual verbs and nouns, and gradually build more schematic grammatical constructions over time. Seen from the aspect of phylogenesis, a construction emerges and spreads in an incremental fashion, as many historical linguists have observed in detailed studies. In line with the usage-based model, my analyzes and theoretical claims are based on data from observed actual language use, including single instances, text segments, and frequency data. Similarity of form and meaning is observed and abstracted from instances of language. Patterns of language use are taken as evidence for the independent representation of grammatical information in constructions. The frequency data of variant modal grammatical patterns are collected and analyzed from large corpora such as COCA, as the conventionality of a linguistic unit relates to the idea that linguistic expressions are shared among members of a speech community. Thus, sufficient frequency of a symbolic unit (construction) can be an indication of its conventionality.
3.6 The Corpus-Linguistic Methodology
45
3.6 The Corpus-Linguistic Methodology A usage-based model of language cognition has to be based on actual use in the first place. It is exactly here that corpus data may play a major role in refining cognitive constructional grammar and increasing its usage-basedness: “corpora are samples of ‘actual use of the linguistic system’; the ‘schematic networks’, ‘low-level schemas’, and ‘linguistic conventions’ correspond largely to the lexicogrammatical patterns and routines that can be identified by drawing on corpus data” (Mukherjee, 2004: 87). Methodologically, research in construction grammar has emphasized the importance of attested data, gathered from discourse or corpora (Goldberg, 1997: 3). Constructionist approaches also share a view that both lexicon and grammar form a continuum of constructions, viz. pairs of form and meaning. This view of language makes the study of grammar more tractable to investigation by corpus-linguistic methods (cf. Stefanowitch & Gries, 2003). In order to answer the research questions formulated above, the present study tries to merge two currents in modern linguistics, one theoretical and one methodological. The theoretical current is made up of constructional theories that view syntactic structures as meaningful elements, while the methodological current is that of corpus linguistics. When investigating the relationship between words and grammatical structures, researchers typically focus on the preferences or restrictions associated with slots in the construction. In constructional approaches to grammar, language as a whole is recast as a system of uniform elements—constructions, any form-meaning pattern that is cognitively routinized. Corpus-linguistic methodologies permit a data-driven identification of these routinized patterns—the assumed building blocks of speakers’ linguistic knowledge. This methodology is adopted in our identification and analysis of the syntactic pattern. It is surprising that, in spite of the prominence of the term usage-based in recent years, corpus-linguistic approaches have not enjoyed a particular prominence with a few exceptions. In general, the methodological orientation of cognitive linguistics has so far been relatively qualitative. However, much work within this paradigm has started to adopt a more balanced perspective, incorporating many kinds of evidence. The main argument for using a corpus is that it could be a more reliable guide to language use than native intuition is. There was a lot of evidence that linguistic intuition, so-called, is not always reliable, but what one finds in a corpus can more or less be taken as authentic (Fillmore, 1992: 3). The essence of corpus linguistics, according to Barlow (2013), “is the analysis of a corpus or corpora to obtain information, often quantified, concerning the formal, semantic, or textual aspects of lexicogrammatical patterns contained in the corpus data” (p. 443). Such analyses are interpreted as revealing something about language, whether viewed from a social or cognitive perspective. Taking a corpus to be a sample of usage, we can obtain data precisely relevant to usage events, viz. situated instances
46
3 Towards a Constructional Approach to Metaphor of Modality
of language use that is contextually and culturally embedded, especially with respect to the form and frequency of language expressions. The observation and interpretation of patterns in authentic data are the basis of the descriptive claims in the present study. The main data source of language facts is Corpus of Contemporary American English (COCA). In my quest for a sufficiently large corpus to get abundant examples of patterns and frequency data, COCA has been a natural choice. I searched it by means of the Brigham Young University interface: http://corpus.byu.edu/coca/. Please note that presently COCA is accessible via https:/ /www.english-corpora.org/coca/. The corpus comprises about 450 million words of text, with equally divided spoken, fiction, popular magazines, newspapers, and academic genres. It is claimed to be the only corpus of English that is suitable for looking at current, ongoing changes in the language (http://corpus.byu.edu/coca/). When relevant constructions are qualitatively analyzed, larger discourse contexts from the texts are taken into account. The corpus data are mainly used to describe two major constructions of modal variants, namely [I VMental that P] construction and [It be ADJ that P] construction. With either construction, the first step is to identify the lexis that has the schematic pattern and count both the type frequency and token frequency. Then these lexical items are classified according to their semantic features. Large online databases such as FrameNet and VerbNet assist in lexical-semantic analysis. The online FrameNet lexical database currently consists of more than 10,000 lexical units, that is, words or expressions paired to meanings, and illustrated by more than 135,000 annotated sentences taken from corpora. The semantic functions in the FrameNet lexicon are defined relative to a frame. VerbNet (Kipper-Schuler, 2006) is the largest online verb lexicon currently available for English. It is a hierarchical domain-independent, broad-coverage verb lexicon with mappings to other lexical resources such as WordNet and FrameNet (Baker et al., 1998). VerbNet (VN) is organized into verb classes extending Levin (1993) classes through refinement and addition of subclasses to achieve syntactic and semantic coherence among members of a class. Each verb class in VN is completely described by selectional restrictions on the arguments, and frames consisting of a syntactic description and semantic predicates. Using VN, we can find the verb class to which each verb form belongs and the syntactic frames with corresponding semantics which feature the particular verb class and its subgroups. In most cases, more co-text for each concordance line needs to be seen. Furthermore, text excerpts from the corpus are used to illustrate the discourse-functional properties of the constructions. They can only be examined by looking at an extended piece of discourse. To sum up, corpus linguistics can fruitfully complement cognitive construction grammar in developing a genuinely usage-based model for description and account of language phenomenon, including variant modal expressions.
References
47
3.7 Summary With cognitive grammar, cognitive construction grammar (CCxG), and cognitive theories of grammaticalization/constructionalization as theoretical background, this chapter has argued the advantages of treating variant expressions of modality as constructions over Halliday’s (1985, 1994)’s grammatical Metaphors of Modality. Constructions as form-meaning pairings are ‘vertical’ structures that combine semantic, syntactic, and even phonological information. Since meaning directly maps to form, their semantic differences and syntactic distinction can be adequately described and explained. There is no need to first bring in what is congruent realization of modality and then account for metaphorical modes of modal expression. Constructional approaches provide a direct way of accounting for subtle facts about both surface forms and corresponding functions, including their semantic properties, their discourse properties, and their conditions of use such as genre. The cognitive, lexicogrammatical, and functional, dimensions of the MM phenomenon will be explored in the subsequent Chaps. 4, 5, and 6, respectively. In accordance with the nature of each dimension, appropriate methods are adopted, primarily under the methodology of cognitive linguistics. Therefore, methodology is presented along with each of these chapters.
References Baker, C. F. et al. (1998). FrameNet. http://framenet.icsi.berkeley.edu/fra-menet. Barðdal, J. (2008). Productivity: Evidence form case and argument structure in Icelandic. John Benjamins. Barlow, M., & Kemmer, S. (Eds.). (2000). Usage-based models of language. Center for the study of language and information. Barlow, M. (2013). Individual differences and usage-based grammar. International Journal of Corpus Linguistics, 4, 443–478. Biber, D. (1988). Variation across speech and writing. Cambridge University Press. Boas, H. C. (2013). Cognitive construction grammar. In T. Oxford (Ed.), Hoffmann and Trousdale (pp. 233–253). Oxford University Press. Boyland, J. T. (1996). Morphosyntactic change in progress: A psycholinguistic approach. University of California at Berkeley. Bybee, J. L. (2006). From usage to grammar: The mind’s response to repetition. Language, 82, 711–733. Bybee, J. L. (2013). Usage-based theory and exemplar representations of constructions. In T. Oxford (Ed.), Hoffmann and Trousdale (pp. 49–69). Oxford University Press. Coates, J. (1983). The semantics of the modal auxiliaries. Croom Helm. Croft, W., & Alan Cruse, D. (2004). In Cognitive linguistics. Cambridge University Press. Dancygier, B. (2008). Arie Verhagen: Constructions of intersubjectivity: Discourse, syntax, and cognition. Linguistics, 3, 658–667. De Smet, H., & Verstraete, J.-C. (2006). Coming to terms with subjectivity. Cognitive Linguistics, 17(3), 365–392. Diessel, H., & Tomasello, M. (2001). The acquisition of finite complement clauses in English: A corpus-based analysis. Cognitive Linguistics, 2, 97–141.
48
3 Towards a Constructional Approach to Metaphor of Modality
Evans, V., & Green, M. (2006). Cognitive linguistics: An introduction. Edinburgh University Press. Fetzer, A. (2014). I think, I mean and I believe in political discourse: Collocates, functions and distribution. Functions of Language, 1, 67–94. https://doi.org/10.1075/fol.21.1.05fet. Fillmore, C. J. (1992). Corpus linguistics versus computer-aided armchair Linguistics. In J. Svartvik (Ed.), Directions in corpus linguistics (pp. 35–60). Mouton de Gruyter. Goldberg, A. E. (1995). Constructions: A construction grammar approach to argument structure. The University of Chicago Press. Goldberg, A. E. (2006). Constructions at work: The nature of generalization in language. Oxford University Press. Goldberg, A. E., & Laura, S. (2010). Construction grammar. https://doi.org/10.1002/wcs.22 Goldberg, A. E. (1997). The relationships between verbs and constructions. Lexical and syntactical constructions and the construction of meaning, 383–398. Gries, S. T. (2006). Introduction. In Gries & Stefanowitsch (Eds.), Corpora in cognitive linguistics: Corpus-based approaches to syntax and lexis (pp. 1–18). Mouton de Gruyter. Haiman, J. (1994). Ritualization and the development of language. In W. Pagliuca (Ed.), Perspective on grammaticalization (pp. 3–28). John Benjamins. Halliday, M. A. K. (2004). The language of science. Continuum. Halliday, M. A. K., & Hasan, R. (1976). Cohesion in english. Longman. Halliday, M. A. K., & Matthiessen, C. (2004). An introduction to functional grammar. Hodder Arnold. Heltoft, L. (2005). Modality and subjectivity. In A. Klinge, & H. H. Muller (Eds.), Modality: Studies in form and function (pp. 81–101). Equinox. Hengeveld, K. (1988). Illocution, mood and modality in a functional grammar of Spanish. Journal of Semantics, 1, 227–269. Israel, M. (1996). The way constructions grow. In A. Goldberg (Ed.), Conceptual structure, discourse and language. CSLI, 217-230. Johnson, M. (1987). The body in the mind: The bodily basis of meaning, imagination, and reason. University of Chicago Press. Kipper-Schuler, K. (2006). VerbNet. http://verbs.colorado.edu/verb-index/index.php. Langacker, R. W. (1988). An overview of cognitive grammar. In B. Rudzka-Ostyn (Ed.), Topics in cognitive linguistics (pp. 3–48). John Benjamin. Langacker, R. W. (1990). Subjectification. Cognitive Linguistics, 1, 5–38. Langacker, R. W. (1998). On subjectification and grammaticalization. In J.-P. Koenig (Ed.), Discourse and cognition: Bridge the gap (pp. 71–89). CSLI. Langacker, R. W. (1999b). Losing control: Grammaticisation, subjectification, and transparency. In A. Blank & P. Koch (Eds.), Historical semantics and cognition (pp. 147–175). Mouton de Gruyter. Langacker, R. W. (2008). Cognitive grammar: A basic introduction. Oxford University Press. Langacker, R. W. (1985). Observations and speculations on subjectivity. In J. Haiman (Ed.), Iconicity in syntax (pp. 109–150). John Benjamins. Langacker, R. W. (1987). In Foundations of cognitive grammar, volume 1, Theoretical prerequisites. Stanford University Press. Langacker, R. W. (1991). Concept, image, and symbol: The cognitive basis of grammar. (Cognitive Linguistics Research 1). Mouton de Gruyter. Langacker, R. W. (1999a). In Grammar and conceptualization. (Cognitive Linguistics Research 14.) Mouton de Gruyter. Langacker, R. W. (2003). Extreme subjectification: English tense and modals. In H. Cuyckens et al. (Eds.), Motivation in language: Studies in honor of Gunter Radden (pp. 3–26). Benjamins. Langacker, R. W. (2009). In Investigations in cognitive grammar (Vol. 42). Walter de Gruyter. Leech, G. N. (1983). Principles of pragmatics (Vol. 285). Longman. Levin, B. (1993). English verb classes and alternations: A preliminary investigation. University of Chicago Press. Levinson, S. C. (1983). Pragmatics. Cambridge University Press.
References
49
Lyons, J. (1977). Semantics. Cambridge University Press. Lyons, J. (1982). Deixis and subjectivity: Loquor, ergo sum? In R. J. Jarvella, & W. Klein (Eds.), Speech, place, and action: Studies in deixis and related topics (pp. 101–124). Wiley. Mukherjee, J. (2004). Corpus data in a usage-based cognitive grammar. Language and Computers, 1, 85–100. Narrog, H. (2005). On defining modality again. Language Sciences, 2, 165–192. Narrog, H. (2010). (Inter)subjectification in the domain of modality and mood-concepts and crosslinguistic realities. In: Davidse, Kristin et al. (Eds.), Subjectification, intersubjectification and grammaticalization (pp. 385–430). De Gruyter Mouton. Nuyts, J. (2006). Modality: Overview and linguistic issues. In W. Frawley (Ed.), The expression of modality (pp. 1–25). Mouton de Gruyter. Perkins, M. R. (1983). Modal expressions in english. ABLEX Publishing Corporation. Pierrehumbert, J. B. (2001). Exemplar dynamics: Word frequency, lenition and contrast. In J. Bybee & P. Hopper (Eds.), Frequency and the emergence of linguistic structure (pp. 137–157). John Benjamins. Portner, P. (2009). Modality. Oxford University Press. Scheibman, J. (2001). Local patterns of subjectivity in person and verb type in American English conversation. In Bybee & Hopper (Eds.), Frequency and the emergence of linguistic structure (pp. 61–90). John Benjamins Publishing Company. Schifrin, D. (1990). The management of a co-operative self during argument: The role of opinions and stories. In A. D. Grimshaw (Ed.), Conflict talk: Sociolinguistic investigations of arguments in conversations (pp. 241–259). Cambridge University Press. Stefanowitch, A., & Gries, S. T. (2003). Collostructions: Investigating the interaction of words and constructions. International Journal of Corpus Linguitstics, 2, 209–243. Sweetser, E. (1990). From etymology to pragmatics: Metaphorical and cultural aspects of semantic structure. Cambridge University Press. Talmy, L. (1988). Force dynamics in language and cognition. Cognitive Science, 1, 49–100. Talmy, L. (1978). The relation of grammar to cognition: A synopsis. In D. Waltz (Ed.), Proceedings of TINLAP-2: Theoretical issues in natural language processing. University of Illinois Coordinated Science Laboratory. Traugott, E. C. (1989). On the rise of epistemic meanings in english: An example of subjectification in semantic change. Language, 1, 31–55. Traugott, E. C., & Dasher, R. (2002). Regularity in semantic change. Cambridge University Press. Traugott, E. C., & Trousdale, G. (2013). Constructionalization and constructional changes. Oxford University Press. Traugott, E. C. (1995). Subjectification in grammaticalization. Subjectivity and subjectivisation, 31– 54. Traugott, E. C. (2006). Historical aspects of modality. The expression of modality, 1107–139. Traugott, E. C. (2010). (Inter)subjectivity and (inter)subjectification: A reassessment. In D. Kristin et al. (Eds.), Subjectification, intersubjectification and grammaticalization (pp. 29–74). De Gruyter Mouton. Verhagen, A. (2005). Constructions of intersubjectivity: Discourse, syntax, and cognition. Oxford University Press. Watts, R. J. (1984). An analysis of epistemic possibility and probability. En-Glish Studies, 65, 129–140.
Chapter 4
The Semantic Motivation of MM
In the last chapter, we developed a constructional approach to account for Metaphor of Modality initiated by Halliday (1985, 1994). The principal objective is to have an in-depth and broader understanding of this linguistic phenomenon. This chapter addresses the pressing need to clarify the semantic motivation of modal clausal constructions like I think or it is likely and to illuminate how they, MM in SFL, differ semanticallly from modal adjuncts, which is considered as the congruent mode of modality in SFL, as very few SFL study has successfully illuminated this problem. The aim of this chapter is to analyze the cognitive structure and process that lie underneath Metaphor of Modality and are reflected in its language form and language organization in certain ways. Note that a “conceptual view of meaning does not imply that semantic structure is directly accessible to introspective awareness” (Langacker, 2008: 38). Rather it is only through careful linguistic analysis with appropriate tools that we can arrive at a principled and revealing characterization of semantic structure. Section 4.1 introduces the theoretical basis on which to analyze metaphorical modality from a cognitive semantic perspective. Section 4.2 propose a semantic analysis of variant modal constructions via a Construal Mental Spaces Model of Modality, based on the semantics apparatuses needed to provide a plausible conceptual analysis of the modality system, including definitions for a proposed set of analytic constructs. This chapter rounds up with laying out the advantages of the Construal Mental Spaces Model for analyzing the dynamic expressions of modality.
4.1 Theoretical Basis As mentioned in the last chapter, Halliday’s differentiation of subjective and objective orientation is in a way more or less equivalent to Lyons’s (1977) distinction of subjective and objective modality. In the last two decades, there have been many © Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7_4
51
52
4 The Semantic Motivation of MM
studies attempting to theorize subjectivity. Langacker is one of the most prominent exponents of subjectivity, and he extended it to subjectification, which concerns how the motivations and strategies of speaker’s expressing their perceptions, feelings, and opinions in discourse conventionalize and interact with linguistic structure (Scheibman, 2001). Langacker (1987, 2006) uses the terms subjective and objective in reference to the subject and object of conception. The subject (S) is the locus of neural activity through which it engages some facet of the world, the object (O). The degree of ‘subjectivity’ or ‘objectivity’ with which the conceptualizer construes a particular entity or situation is a focal aspect of perspective. It needs to be noted that Langacker uses the terms subjective, objective, and their derivatives in a different, technical sense. Rather than referring to the semantic content of a linguistic expression, they should be understood in terms of the viewing arrangement or perspective assumed to construe a particular conceptual content. The term ‘objective’ is understood as functioning solely and prominently as the OBJECT OF PERCEPTION in contrast to the perceptual apparatus itself; the term ‘subjective’, functioning exclusively as part of the SUBJECT OF PERCEPTION, viz. one component of the perceiving apparatus. In line with Langacker, subjectivity and objectivity refers to the viewing arrangement or perspective to construe a particular conceptual content. Objectivity aligns to the object of conceptualization, whereas subjectivity aligns to the subject of conceptualization, namely, the conceptualizer. The conceptualizer can be identified primarily with the speaker, secondarily with the addressee, and derivatively with some other individual whose perspective is adopted or otherwise considered (Langacker, 1991: 318). Subjectivity/objectivity in relation to the conceptualizer, as one facet of the perspective taken on a scene, represents an important factor in both semantic and grammatical structure. A profiled content can occupy a number of different positions vis-à-vis the Ground. A wholly objective construal in which the ground lies outside the scope of the predication is represented in Fig. 4.1a. Another option is for a particular facet of the ground to be put onstage as one of the relational participants, as in 4.1c and d. Metaphorically, we can speak of S being the offstage viewer and O the onstage entity being viewed. Status concerning this asymmetry is one facet of an element’s construal: S is construed subjectively, and O objectively (Langacker, 2006). In addition to its salient onstage elements, an expression’s meaning includes a viewing arrangement or viewpoint. Its canonical form is sketched in Fig. 4.2, which introduces two additional perspectival factors: vantage point and scope. A vantage point is the location from which a situation is apprehended. In the canonical arrangement, the interlocutors are together in space and time, viewing the objective scene from the same vantage point offstage. Scope pertains to conceptual content: the extent of the content invoked and degrees of centrality imposed by viewing. The maximal scope is all the content that figures in the meaning of an expression, even if only peripherally. Within that, the immediate scope is the portion being attended to (the onstage region). Moreover, within the immediate scope, the expression’s profile is maximally prominent as the specific focus of attention.
4.1 Theoretical Basis
53
Fig. 4.1 “Onstage” versus “Offstage” elements in construal (Langacker, 2013: 3)
Fig. 4.2 Canonical viewing arrangement (Langacker, 1987: 139)
Multiple viewers and vantage points figure in all but the simplest expressions (Langacker, 2008). The interlocutors recognize other viewers (including each other) and to some extent simulate their experience. Their actual position in the Ground therefore functions not only as the default-case vantage point, but also as point of departure for invoking and simulating others. ‘Subjectification’ is a type of semantic shift in which an originally objective notion is transferred to the subjective axis of the construal relation itself. In CG, conceptualization makes pervasive use of imaginative abilities like mental space construction and management, and blending (Broccias, 2013: 196). Mental Spaces Theory is particularly well adapted to formalize such concepts as Ground, Content, and Viewpoint. Mental Spaces Theory (henceforth referred to as MST) is a highly influential cognitive theory of meaning construction. A mental space is “a partial and local conceptual representation, which can be mapped onto or combined with other such spaces to build complex conceptual structure”. According to MST, since a given mental space is always attached to some perceiver or cognizer, mental spaces necessarily structure viewpoints. Any space configuration will include a Base, a Viewpoint, a Focus and an Event. The Viewpoint is the space from which the discourse is currently being viewed and from which other spaces are currently being built and accessed. Viewpoint is commonly used by scholars in a very broad
54
4 The Semantic Motivation of MM
sense. In this study, it adopts a narrower sense within the Mental Spaces framework. It refers to a relevant cognizer’s mental standpoint from which other mental spaces are viewed, including mainly epistemic, affective, social and even discursive standpoint. The Focus is the space where new content is being added, accessed from the current viewpoint space, and the Event represents the time associated with the event being described. Base, Viewpoint, and Focus need not be distinct; more often than not, the same space are found serving as Viewpoint and Focus, or Base and Focus, or Base and Viewpoint, or a11 three: Base, Viewpoint, and Focus. Spaces can be introduced explicitly by space-builders or implicitly on pragmatic grounds. Elements in different spaces are linked by connectors that set up mappings between counterpart elements. The Mental Spaces framework provides a way to represent viewpoint phenomena. Mental spaces also have relative status, i.e. hierarchical, with relation to other spaces: “mental spaces can be embedded within other mental spaces” (Tobin & Israel, 2012: 25). Embedded spaces are assumed to entail a restriction of the validity or factuality of the embedded material (Fauconnier, 1985; Sanders & Redeker, 1996). Moreover, grammar guides a variety of mental tasks in the treatment of the incoming information such as the management of viewpoint, focus and base structure in the constructed network. Within MST framework, subjectivity is considered in terms of distance from the speaker’s internal mental space in the mental space network. Any communicative use of language necessarily involves the presumption that the speaker has mental states, and that he/she is expressing some content of his/her mental states, in some speech act setting, using some set of linguistic forms. This being the case, any communicative speech act rests on the presumed presence of Content, Epistemic, Speech Act, and Metalinguistic spaces. . On this basis, Sanders et al. (2009) formalize the concept of a Basic Communicative Space Network (henceforth BCSN), which in essence is a specification in mental space terms of the minimum basic structures involved in a speech-interaction Ground, including the construal of a Content Space network relative to a Ground network. A more recent Ferrari and Sweetser (2012)’s analysis shows that a mental space model involving a BCSN with added internal elaboration of the Ground is helpful in modeling viewpoint, subjectivity, and subjectification. In subjectification processes, implicit viewpoint structure can develop into explicit viewpoint structure. Halliday (1985, 1994) formalizes Metaphor of Modality as explicit indication of subjective and objective orientation by a nexus of two clauses, but this criterion is mainly in formal terms. Can we define MM in terms of cognitive structure so that its distinction in meaning can be revealed? Viewpoint structure in subjectification processes appears to be a promising construct. The distinction between profiling the expressed content and profiling the Ground is a matter of degree. Following Langacker, Ferrari and Sweetser (2012) agree that no utterance is referential without Grounding. Furthermore, there is a cline from minimal profiling of Ground elements to full overt profiling, with every possible degree of relative profiling of Ground Content in between the opposite ends. Ferrari and Sweetser (2012: 48) give a more elaborate mental space instantiation of Langacker’s Ground as the Ground network, which involves: (1) a Ground
4.1 Theoretical Basis
55
Base Space: the real Speaker and Hearer in their spatiotemporal setting; (2) one or more Epistemic Spaces: understandings of S’s and H’s beliefs and reasoning processes, which are also background to the communication; (3) a Speech Act Space of performed speech interactions; (4) a Metalinguistic Space of shared linguistic forms that can be referenced. Their version of BCSN model is shown in Fig. 4.3. Within the Ground, the Ground Base Space is taken to be less subjective, and the Speech Act and Epistemic Spaces, more subjective (Ferrari & Sweetser, 2012: 51). Indeed, mental spaces, grounding and viewpoint have been found to motivate a multitude of linguistic structures, such as definiteness markers, the will future, and “A knows P” (Ferrari & Sweetser, 2012: 47–61). Based on this line of research, I will propose a refined mental space structural model to highlight viewpoint phenomena in language. What follows is to provide a finer-grained semantic account of variant modal expressions referred to as Metaphor of Modality by Halliday, in terms of viewing
Fig. 4.3 Basic communicative space network (from Ferrari & Sweetser, 2012: 49)
56
4 The Semantic Motivation of MM
arrangement to construe a particular conceptual content, developed mainly in line with Langacker (2006)’s constructs shown below. • Scope: the extent of the conceptual content invoked and degrees of centrality imposed by viewing • The maximal scope: all the content that figures in an expression’s meaning, even if only peripherally • The immediate scope: the portion being attended to, that is, the onstage region, within the maximal scope • Profile: the specific focus of attention within the immediate scope. • Ground network: a network encompassing the speech event, its participants with the extent of their involvement, and its immediate circumstances. Langacker’s frameork of viewing arragement will be incorporated into Ferrari and Sweetser (2012)’s concept of Basic Communication Space Network, which is based on the presumption concerning communicative use of language: the speaker has mental states, and that he/she is expressing some content of his/her mental states, in some speech act setting, using some set of linguistic forms. They delineate understandings of S’s and H’s beliefs and reasoning processes as Epistemic Space.
4.2 The Proposal of a Refined Model There is something missing in Ferrari and Sweetser (2012)’s presumption concerning communication: in order to express some content, as a default arrangement, the speaker and the interlocutor are together in a location, from which they observe and describe actual occurrences in the world around them. In other words, their viewing and attention to worldly occurrences is also a precondition for their expression. Cognitive perspective starts with bodily viewpoint within a real physical Ground of experience. There is evidence that communication activates the embodied neural system’s representation of viewpoint (Sweetser, 2012: 3). Viewpoint permeates human cognition and communication (ibid: 1). Therefore, it makes sense to include viewpoint in Basic Communication Space Network. Since a given mental space is always attached to some perceiver or cognizer, mental spaces necessarily structure viewpoints. Verbs such as think, which explicitly marks cognitive states, necessarily build mental spaces, as Fauconnier (1985, 1994) has said. In communication, such as an exchange, the participants directly involved, as well as the subjectivities evoked by the participants, are all engaged in viewpointspecific construals negotiated in discourse (Dancygier, 2012a, 2012b: 219). According to Wellman and Miller (2008), in everyday reasoning, belief-desire and obligation-permission/deontic concerns interrelate. Both epistemic reasoning and deontic reasoning are early developing, universal, and inseparable in children’s understanding. In addition, “affective judgments are inescapable” (Zajonc, 1980: 156). Psychological experiments have supported that organism is equipped with a neurochemical
4.2 The Proposal of a Refined Model
57
apparatus capable of telling not only the new from the old and but also the good from the bad, and of making all these decisions rapidly (cf. Zajonc, 1980). It is agreed that affective phenomena deserve far more attention than they have received from cognitive psychologists. EPISTEMIC is relative to the spheres of knowledge of the speaker (S’), the hearer (H’), metaphorically speaking, cold cognitions of S’, H’. AFFECTIVE is relative to the spheres of affective states and judgments, in other words, hot cognitions of S’, H’. Affective judgments and epistemic cognition are under the control of partially independent systems that can influence each other in a variety of ways, and both constitute independent sources of effects in information processing in communication. Cognitive scientists tend to focus on epistemic grounding but neglect the affective and social aspects of human organism. Nevertheless, psychological, linguistic and sociolinguistic studies have demonstrated the prevalence of affective judgments and socially mediated judgments and their critical place in the phenomenon viewpoint. Therefore, Epistemic Spaces is too narrow a term to encompass not only epistemic but affective and socially mediated reasoning processes. I am replacing it with Viewpoint Spaces, where the term ‘Viewpoint’ refers to relevant cognizers’ mental standpoints from which other mental spaces are viewed, including epistemic, affective, social and even discursive standpoints. As a result, the following elements are deemed necessary in the Ground network: – Ground Base Space: the real speaker (writer) and the addressee in their spatiotemporal settings. The ‘addressee (A)’ is replacing the hearer (H) as both spoken and written discourses are taken into consideration. – Viewpoint Space: the space from which the situation is currently being viewed, involving the epistemic, affective, deontic states and processes of the speaker (writer) and the addressee. – Speech Act Space: Speech interactions performed by speaker, or writer to a lesser degree, and the addressee. They are acts that are intended to be performed be a speaker in uttering a linguistic expression and produce a certain effect in or exert a certain influence on the addressee. Thus, a refined Basic Communication Space Network model is proposed and graphically represented in Fig. 4.4. Following Verhagen (2005: 41), graphically CONTENT, which is Object of Conceptualization, is up, since it is also the onstage region, and GROUND is down. In the VIEWPOINT Space, the speaker/writer and addressee are simultaneously Subjects of Conceptualization. This model reinforces that meaning resides in a spatial network of mental spaces, a point clarified in Ferrari and Sweetser (2012)’s approach to subjectivity and Tobin and Israel (2012)’s study of irony. Following Verhagen (2005), both the speaker (S) and the addressee (A) are the subjects of conceptualization engaging in cognitive coordination by means of the utterance, with respect to some object of conceptualization in CONTENT Space. Davidson (2001) argues that thought requires a kind of triangular interaction between (at least) two subjects and objects in the world. The interaction is twofold from the viewpoint of each of two subjects: each is interacting
58
4 The Semantic Motivation of MM
Fig. 4.4 The refined basic communicative space network
simultaneously with the world and with the other subject. The relation of joint attention is widely believed to hold between the two subjects. However, one can only have joint attention “if one has the capacity to ‘join’ another person—which means that one needs to be able to share experiences with others, registering intersubjective linkage” (Hobson, 2005: 201). The conceptual structure of Fig. 4.4 is inherent in any linguistic usage event. What is more, EPISTEMIC, AFFECTIVE or DEONTIC standpoints may present in VIEWPOINT spaces, as shown in Fig. 4.5. (Other kinds of standpoint might also input into VIEWPOINT spaces.) Fig. 4.5 Dimensions of VIEWPOINT space
4.3 The Semantic Analysis of MM
59
4.3 The Semantic Analysis of MM In this section, I will define MM conceptually with reference to the Refined Basic Communication Space Network proposed in Sect. 4.2. This section will demonstrate the likely advantages of a definition with reference to this model. My demonstration begins with some typical examples given in previous studies as congruent and metaphorical modal expressions in (14). Consider two interactants are in the exchanges on the probability of Mary knowing about something. Each of the three utterances in (14) is possible. (14)
a.
Mary will know.
b.
I think Mary knows.
c.
It is likely that Mary knows.
The fact that (14a), (14b), and (14c) are all agnate ways of construing the same focal situation can be represented, on the basis of the Refined Basic Communication Space Network demonstrated in Fig. 4.4, in terms of the construal configuration as shown in Figs. 4.6, 4.7 and 4.8. To some extent, they represent the reanalysis of construal configuration for utterance with modal adjunct, first-person perspective, and impersonal complementation constructions, based on the refined Basic Communication Space Network model. Not all aspects of the construal configuration have to be marked by means of some linguistic unit. Figure 4.6 represents the construal configuration in Example (14a): Mary will know.
Fig. 4.6 Mental space construal configuration for utterance with modal adjunct (14a)
60
4 The Semantic Motivation of MM
Fig. 4.7 Mental space construal configuration for first-person modal expression (14b)
As shown in Fig. 4.6, VIEWPOINT Space involving the epistemic, affective, deontic states and processes of the speaker and the addressee is held offstage. Meanwhile it operates in the Ground as the source of the potency. The speaker and the addressee, plus their speech act, are implicit. Consider the modal adjunct will, where the potency amounts to a kind of probability. The modal adjunct itself is only a prompt cuing a Viewpoint space in the Ground. Expressions such as 1 think are naturally treated as stance expressions; ‘stance’ represents the kind of subjectivity where the speaker’s viewpoint is directly coded in the expression (Dancygier, 2012a, 2012b: 74). In other words, the speaker’s viewpoint is put onstage as part of CONTENT profiled in the utterance. Therefore, the mental space construal configuration for utterance with first-person perspective is different from the one shown in Fig. 4.6. Figure 4.7 represents the mental space construal configuration for utterance with first-person perspective in which VIEWPOINT space in the GROUND is mapped into the CONTENT Space. Meanwhile, the speaker, mediated through VIEWPOINT space as the subject of conceptualization holding the viewpoint of P (P stands for the proposition embedded within the complement clause), is put on stage, formalized by the first-person pronoun. Moreover, this up-moved VIEWPOINT space is distinct from Object of Conceptualization, as linguistically marked by the fact that they are realized by two clauses: the projecting clause and the projected clause. In terms of
4.3 The Semantic Analysis of MM
61
Fig. 4.8 Mental space construal configuration for impersonal complementation construction (14c)
cognitive structure, the borderlines of these two spaces are very distinct; iconicly, Object of Conceptualization, P, is framed within the higher content space. This, then, allows the speaker/writer to further qualify his viewpoint of P explicitly. For example, various values, high, median or low, can be attached to the epistemic judgment of P using different Verbs. These different values are illustrated in Table 4.1. It is likely Mary knows (14c) is a typical example of that complement clauses controlled by an adjective. It is considered as one major grammatical device used to express stance, viz., linguistic coding of viewpoint. In this respect, it is similar to utterance with first-person perspective, as VIEWPOINT space in the GROUND is also mapped into the CONTENT Space. It is different in that the speaker/writer chooses not to put or specify the Subject of Conceptualization on stage; therefore, Table 4.1 Values of modality with verbs
Mental verbs
Examples
High
Believe
I believe Mary knows I find Mary knows
Median
Think
I think Mary knows
Low
Guess
I guess Mary knows I suspect Mary knows
62
4 The Semantic Motivation of MM
it is desubjectified. Figure 4.8 represents the mental space construal configuration for utterance with impersonal complementation construction. As can be seen in this Figure, VIEWPOINT space in the GROUND is mapped into CONTENT Space. This up-moved VIEWPOINT space is linked to P, Object of Conceptualization, in the discourse. Similar to Fig. 4.7, the borderlines of these two spaces are very distinct; iconicly, P is framed within the higher CONTENT space. The distinction of VIEWPOINT space and Object of Conceptualization is formally marked: they are represented by two clauses, instead of one clause. Subsequently this allows the viewpoint of P to be elaborated explicitly. Different from Fig. 4.7, the speaker/writer chooses not to put or specify the Subject of Conceptualization on stage; therefore, it is desubjectified. Subject of conceptualization is off stage, while the viewpoint itself is on stage. In other words, ‘it’ indicates that this view may be shared with a wider group of people, including possibly the speaker and the hearer, in this sense intersubjectivity applies. As the above analysis of the conceptual structure shows, whether there is a cognitive process of putting VIEWPOINT space onstage involved can aptly account for the linguistic phenomena from utterance with modal adjunct to first-person perspective and impersonal complementation construction, in other words, from usage event involving modal adjuncts within the clause to explicit clausal representation of modality. This is the crucial difference between these two modes of expression in their cognitive structure. In cases where VIEWPOINT space is only grounding the predication and held off stage, it is represented by modal adjuncts within the clause, only implicitly cuing subjectivity involved. In other cases where VIEWPOINT space is put onstage, it is represented by a projecting clause taking the complement that profiles Object of Conceptualization. What these modal usage events share is that there is VIEWPOINT space operating in the Ground as the source of the potency. Therefore, in terms of conceptual structure, modality can be defined in terms of Viewpoint space grounding the profiled Content, as the source of the potency. Moreover, whether there is a cognitive process of putting VIEWPOINT space onstage involved motivates the variation in expressive modes. In other words, modality is an operation on the way a focused proposition is accessed and viewed: it is a way of construing an expressed proposition or an observed scene. Grammatical Metaphor of Modality has been criticized for its inadequacy in illuminating the difference in meaning between the congruent and the metaphorical, namely, between modal adjuncts within the clause to explicit clausal representation of modality. The benefit of our alternative approach, the mental space approach to subjectification, is evident from the above analysis: their difference is illuminated in terms of whether VIEWPOINT space is only grounding the predication and held off stage or VIEWPOINT space is put on stage. Because there is naturally variation among the construed conceptual and/or deictic distance between the speaker/writer, these varied degrees of subjectivity will map onto the form. When the speaker/writer consciously disintegrates his subjective construal of the focal situation from his object of conceptualization, through representing it with the explicit mark of ‘I’ as the conceptualizer, his level of commitment
4.4 Summary
63
to the factuality of the information is lower. After all, it indicates that this focal situation is accessed from ‘my VIEWPOINT space’ and is not necessarily the fact. Meanwhile, through the selection of different Verbs, the speaker/writer can more finely moderate his level of commitment. In other words, Metaphor of Modality with firstperson perspective represents the linguistic consequence of the cognitive process in which VIEWPOINT Space, with its first-person Subject of Conceptualization along, projects out of GROUND onto CONTENT. In some other contexts when the speaker/ writer consciously disintegrates his subjective construal of the focal situation from his object of conceptualization, he is conscious that the construed conceptual and/ or deictic distance of the addressee and others may be different from his subjective construal. Communicatively he needs to manage the coordination relationship between speaker/writer, addressee and others potentially involved. The strategy he adopts is to invite the addressee to entertain the object of conceptualization in the same way as marked by the impersonal complementation construction. The subject of conceptualization, as the source of potency, is kept fuzzy. Therefore, the speaker/ writer’s level of commitment to the truth of the information is even lower. To put it another way, the impersonal type of MM represents the linguistic consequence of the cognitive process in which VIEWPOINT Space projects out of GROUND onto CONTENT with the subject of conceptualization kept fuzzy. In terms of cognitive structure, the two subtypes of MM have something in common: both involve the cognitive process of putting VIEWPOINT Space, the source of potency, on stage. They are different from canonical modal expression with modal adjuncts in that in the latter the source of potency, viz. VIEWPOINT Space, is off stage and remains in the ground. The above analysis indicates that modality can be approached cognitively as a VIEWPOINT phenomenon.
4.4 Summary Like much functional work on modality, Halliday’s model is descriptive. Its interpretation on features of modal meaning relies on an understanding of other theoretical concepts such as subjective and objective modality, and some concepts such as source of conviction that are introduced without clear definition. The cognitive dimension of MM analyzed in this chapter is theoretical approached in terms of subjectification and mental space. The refined Basic Communicative Space Network model can explain the behaviors of relevant linguistic structures, like the representation of modality in the projecting clause instead of modal adjunct, first person pronoun in the projecting clause or impersonal complementation construction. In this chapter, we approach Metaphor of Modality through construal operation of putting Viewpoint Space from Ground Network on stage. In cases where VIEWPOINT space is grounding the predication and held off stage, it is represented by modal adjuncts within the clause, and the speaker’s viewpoint remains implicit. In other cases where VIEWPOINT space is put on stage, it is represented by a projecting
64
4 The Semantic Motivation of MM
clause taking the complement that profiles the focal situation. Meanwhile, the elaboration of Ground network in terms of mental spaces captures the complexity of dynamic viewpoint representation in discourse, which might not be able to be accurately accounted for by simplified equation of Ground to subject of conceptualization. As evident from the analysis in this section, modality and variant modal expressions can be accounted for in terms of the grounding of VIEWPOINT Space with or without putting it on stage. In summary, the difference in meaning between congruent and metaphorical modes of modality can be illuminated by the analysis of their underlying cognitive structure. Modal adjuncts within the clause only cue that a VIEWPOINT space is grounding the predication and held off stage, in other words, the viewpoint structure remains implicit. A projecting clause taking the complement that profiles the focal situation, in contrast, represents the VIEWPOINT space being put onstage, in other words, the explicit viewpoint structure. According to Langacker’s CG model, the language user’s ability to construe a conceived situation in alternate ways by means of alternate images for the purpose of thought or expression is fundamental to lexical and grammatical variability. In the subsequent chapter, my focus will be the detailed linguistic consequences of the cognitive structure of MM. In other words, their lexical and syntactic properties will be explored, based on the large corpus, COCA.
References Broccias, C. (2013). Cognitive grammar. In Hoffmann & Trousdale (Eds.), The oxford handbook of construction grammar (pp. 191–210). Oxford University Press. Dancygier, B. (2012a). Negation, stance verbs, and Intersubjectivity. In B. Dancygier & E. Sweetser (Eds.), Viewpoint in language: A multimodal perspective (pp. 69–96). Cambridge University Press. Dancygier, B. (2012b). Conclusion: Multiple viewpoints, multiple spaces. In B. Dancygier & E. Sweetser (Eds.), Viewpoint in language: A multimodal perspective (pp. 219–230). Cambridge University Press. Davidson, D. (2001). Subjective, intersubjective, objective. Clarendon Press. Fauconnier, G. (1985). Mental spaces: Aspects of meaning construction in natural language. MIT Press. Fauconnier, G. (1994). In Conceptual projection and middle spaces. University of California, Department of Cognitive Science Technical Report 9401 (available online at blending.stanford.edu and mentalspace.net). Ferrari, L., & Sweetser, E. (2012). Subjectivity and upwards projection in mental space structure. In B. Dancygier & E. Sweetser (Eds.), Viewpoint in language: A multimodal perspective (pp. 47– 68). Cambridge University Press. Hobson, R. P. (2005). What puts the iointness into ioint attention?. In Joint attention: Communication and other minds: Issues in philosophy and psychology: Issues in philosophy and psychology, (pp. 185). Langacker, R. W. (2008). Cognitive grammar: A basic introduction. Oxford University Press. Lyons, J. (1977). Semantics. Cambridge University Press. Langacker, R. W. (1987) In Foundations of cognitive grammar, volume 1, Theoretical prerequisites. Stanford University Press.
References
65
Langacker, R. W. (2006). Subjectification, grammaticization, and conceptual archetypes. In A. Athanasiadou, C. Canakis, & B. Cornillie (Eds.), Subjectification: Various paths to subjectivity (Cognitive Linguistics Research 31) (pp. 17–40). Mouton de Gruyter. Langacker, R. W. (1991). In Concept, image, and symbol: The cognitive basis of grammar. (Cognitive Linguistics Research 1.) Mouton de Gruyter. Langacker, R. W. (2013). In Essentials of cognitive grammar. Oxford University Press. Scheibman, J. (2001). Local patterns of subjectivity in person and verb type in American English conversation. In Bybee & Hopper (Eds.), Frequency and the emergence of linguistic structure (pp. 61–90). John Benjamins Publishing Company. Sanders, J., & Redeker, G. (1996). Perspective and the representation of speech and thought in narrative discourse. In Spaces, worlds and grammar (pp. 290–317). Sanders, T., Sanders, J., & Sweetser, E. (2009). Causality, cognition and communication: A mental space analysis of subjectivity in causal connectives. In Causal categories in discourse and cognition (pp. 20–59). Sweetser, E. (2012). Introduction: Viewpoint and perspective in language and gesture, from the ground down. In B. Dancygier & E. Sweetser (Eds.), Viewpoint in language: A multimodal perspective (pp. 1–24). Cambridge University Press. Tobin, V., & Israel, M. (2012). Irony as a viewpoint phenomenon. In B. Dancygier & E. Sweetser (Eds.), Viewpoint in language: A multimodal perspective (pp. 25–45). Cambridge University Press. Verhagen, A. (2005). Constructions of intersubjectivity: Discourse, syntax, and cognition. Oxford University Press. Wellman, H. M., & Miller, J. G. (2008). Including deontic reasoning as fundamental to theory of mind. Human Development, 2. Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American Psychologist, 2, 151–175.
Chapter 5
The Lexical and Syntactic Properties of MM
Surprisingly, the study of Metaphor of Modality in systemic functional grammar did not devote much research space to their lexicogrammatical features, as its focus is, arguably, the notional aspect. This chapter will investigate their lexical and syntactic properties of MM, based on the large corpus, COCA. Section 5.1 presents the usagebased model as the basis for our principled analysis. Section 5.2 argues for the fit of the corpus-linguistic methodology with our theoretical orientation and research question and outlines research procedures. Section 5.3 presents a somewhat unique perspective on the syntagmatic structure of MM. Sections 5.4 and 5.5 focus respectively on the lexical properties of two major metaphorical modal constructions: [It be ADJ that P] construction and [I VMental that P] construction.
5.1 The Usage-Based Model and Corpus-Based Method As we argued in Chap. 3, the usage-based model, adopted in cognitive grammar, is compatible with functional linguistics. The usage-based thesis holds that the language user’s knowledge of language is formed by the abstraction of symbolic units from situated instances of language use: an utterance, namely, a usage event. Following this model, patterns of language use are taken as evidence for the representation of grammatical information in my analysis. Most of the expressions we employ are symbolically complex, to some degree analyzable into smaller symbolic elements. Grammar consists of the patterns for constructing such expressions. Such expressions and their patterns are referred to by Langacker (2008) as constructions. Two usage-based properties are particularly assumed to affect grammatical representation: the frequency of occurrence of particular grammatical forms and structures, and the meaning of the words and constructions in use. In the usage-based model, word forms are not the output of rules but instances of schemas. There is evidence for product-oriented schemas in morphology and © Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7_5
67
68
5 The Lexical and Syntactic Properties of MM
syntax, which cannot be easily modeled by rules. Most syntactic constructions are schematic to some degree. Type frequency determines the degree of entrenchment of a schema. The usage-based model represents schema abstracted inductively from actually occurring utterances. In Corpus of Contemporary American English (COCA), besides the classical examples [I think that P] and [It is likely that P], there are quite lot members of both types, including those in (15a–15f) and (16a–16f): (15)
a.
I believe that he is a liberal.
b.
I feel that my privacy has been violated.
c.
I guess that when you go to Harvard, as Barbara did, you tend to pick up connections.
d.
I find that adding in a bit of pleasure tends to make work both more enjoyable and more successful, Ms. D’Alessandro.
e.
I suspect that there’s nothing unconstitutional about it.
This set of constructions can be skeletally represented as [I VMental that P], where VMental is short form for verb of mental processes and P codes the proposition expressed by the complement that clause. For the purpose of this study, I adopt Diessel and Tommsello’s (2001) definition of the term proposition as “the meaning or content of a clausal construction denoting some state of affairs (e.g., a state, an event, or an activity)” (p. 101). Verbs of mental processes are also referred to as private verbs (Biber, 1988), psychological verbs and psychological predicates (Leech, 1983), and cognitive verbs (Fetzer, 2014). Mental processes, if used with firstperson self-reference, are assigned the status of metaphorical modality (Halliday & Matthiessen, 2004: 208). In SFG, the primary process types settle down to six primary types in Halliday (1985), ‘material’, ‘behavioral’, ‘mental’, ‘verbal’, ‘relational’, and ‘existential’ (cf. Halliday & Matthiessen, 2004, 2014). The mental processes are shown with a further step in delicacy in Table 5.1, which is adapted from the account presented in Matthiessen (2014: 147). Besides prototypical examples like [It is likely that P], it remains to see what types of adjectives can be put into the slot of this pattern.
Table 5.1 System of mental processes type Process type 1
Process type 2
Systemic range
Example of metaphorical modality
Mental
Cognitive
[know, think, understand, doubt …]
I think that …
Desiderative
[decide, desire, prefer, want I want your father to buy a …] house
Perceptive
[see, hear, feel, taste …]
I hear/see that …
Emotive
[fear, admire, worry …]
I fear/regret that …
5.1 The Usage-Based Model and Corpus-Based Method
(16)
69
a.
It is essential that the nurse fully inform parents of both the risks and the appropriate preventative measures
b.
It is clear that such groups are not fully under the government’s control.
c.
It is important that people are able to talk about it
d.
It’s likely that my genes predisposed me to have a mood disorder.
e.
It was imperative that we got a chance to hear it again.
f.
It is great that “Newsweek” has decided to do the story.
This set of constructions will be represented as [It be ADJ that P] construction, and its lexical properties will be explored later in this chapter. We can use degree of semantic similarity to rank the syntactic distinctions and thus form a hierarchy, as illustrated in Fig. 5.1 for modes of modal expression: Figure 5.1 is a visual means to represent semantic distance between constructions. A usage-based model of language cognition has to be based on actual use in the first place. It is exactly here that corpus data may play a major role in refining cognitive grammar and increasing its usage-basedness: “corpora are samples of ‘actual use of the linguistic system’; the ‘schematic networks’, ‘low-level schemas’ and ‘linguistic conventions’ correspond largely to the lexicogrammatical patterns and routines that can be identified by drawing on corpus data” (Mukherjee, 2004: 87). Methodologically, research in construction grammar has emphasized the importance of attested data, gathered from discourse or corpora (Goldberg, 1997: 3). Constructionist approaches also share a view that both lexicon and grammar form a continuum of constructions. This view of language makes the study of grammar more tractable to investigation by corpus-linguistic methods (cf. Stefanowitch & Gries, 2003). In order to answer some of the research questions formulated in Chap. 3, efforts are exerted to merge two currents in modern linguistics, one theoretical and one methodological. The theoretical current is made up of cognitive-functional theories that view syntactic structures as meaningful elements, while the methodological current is that of corpus linguistics. When investigating the relationship between
Modal expressions
Clausal Modal Constructions
Modal Adjuncts
Modal Verbs
may
[I VMental that P]
Modal Adv.
etc.
probably
etc.
I think that P
Fig. 5.1 Hierarchic organization of modal expressions
[It be ADJ that P]
etc.
It is likely that P
etc.
70
5 The Lexical and Syntactic Properties of MM
words and grammatical structures, researchers typically focus on the preferences or restrictions associated with slots in the construction. In constructional approaches to grammar, language as a whole is recast as a system of uniform elements—constructions, any form-meaning pattern that is cognitively routinized. Corpus-linguistic methodologies permit a data-driven identification of these routinized patterns—the assumed building blocks of speakers’ linguistic knowledge. This methodology is adopted in our identification and analysis of the syntactic patterns. It is surprising that, in spite of the prominence of the term usage-based in recent years, corpus-linguistic approaches have not enjoyed a particular prominence in cognitive grammar. In general, the methodological orientation of cognitive linguistics has so far been relatively qualitative. However, much work within this paradigm has started to adopt a more balanced perspective, incorporating many kinds of evidence. The main argument for using a corpus is that it could be a more reliable guide to language use than native intuition is. There was a lot of evidence that linguistic intuition, so-called, is not always reliable, but what one finds in a corpus can more or less be taken as authentic (Fillmore, 1992: 3). The essence of corpus linguistics, according to Barlow (2013), “is the analysis of a corpus or corpora to obtain information, often quantified, concerning the formal, semantic, or textual aspects of lexicogrammatical patterns contained in the corpus data” (p. 443). Such analyses are interpreted as revealing something about language, whether viewed from a social or cognitive perspective. Taking a corpus to be a sample of usage, we can obtain data precisely relevant to usage events, viz. situated instances of language use that is contextually and culturally embedded, especially with respect to the form and frequency of language expressions. The observation and interpretation of patterns in authentic data are the basis of the descriptive claims in this study. The main data source of language facts is Corpus of Contemporary American English (COCA). In my quest for a sufficiently large corpus to get abundant examples of patterns and frequency data, COCA has been a natural choice. I searched it by means of the Brigham Young University interface: http://corpus.byu.edu/coca/. Please note that presently the website of COCA is at https://www.english-corpora.org/coca/. The corpus comprises about 450 million words of text, with equally divided spoken, fiction, popular magazines, newspapers, and academic genres. It includes 20 million words each year from 1990–2012, and the corpus is updated regularly (the most recent texts are from summer 2012). It is claimed to be the only corpus of English that is suitable for looking at current, ongoing changes in the language (http://cor pus.byu.edu/coca/). When relevant constructions are qualitatively analyzed, larger discourse contexts from the texts are taken into account. The corpus data are mainly used to describe two major constructions of MM, namely, [I VMental that P] construction and [It be ADJ that P] construction. With either construction, the first step is to identify the lexis that has the schematic pattern and count both the type frequency and token frequency. Then these lexical items are classified according to their semantic features. Large online databases such as FrameNet and VerbNet assist in lexical-semantic analysis. Neale (2006)’s paper has
5.2 The Syntactic Properties of MM
71
demonstrated that the second-level use of corpora allows access to a broad coverage of language. The online FrameNet lexical database currently consists of more than 10,000 lexical units, that is, words or expressions paired to meanings, and illustrated by more than 135,000 annotated sentences taken from corpora. The semantic functions in the FrameNet lexicon are defined relative to a frame. Another database used is VerbNet, which is organized into verb classes extending Levin (1993) classes through refinement and addition of subclasses to achieve syntactic and semantic coherence among members of a class. Each verb class in VN is completely described by selectional restrictions on the arguments, and frames consisting of a syntactic description and semantic predicates. Using VN, we can find the verb class to which each verb form belongs and the syntactic frames together with corresponding semantics that feature the particular verb class and its subgroups. In most cases, more co-text for each concordance line is read carefully for semantic coding. To sum up, corpus linguistics can fruitfully complement cognitive grammar in developing a genuinely usage-based model for description and account of language phenomenon, including Metaphor of Modality. This approach and method particularly fit the account of [I Vmental that P] construction and [It be ADJ that P] construction.
5.2 The Syntactic Properties of MM 5.2.1 Problems with Constituency Analysis In the formal model, the syntactic component mediates between form and meaning. Formal models view linguistic elements as having a componential structure: elements are viewed as having a complex internal structure, which may consist of structural ‘building blocks’ like morphemes or grammatical categories, or which may consist of semantic ‘building blocks’ like semantic primitives or meaning components. Because of its emphasis on economy and generalization, the formal model places strict constraints upon grammatical constructions and processes. For example, the transformational model assumes that all clause types are constructed according to the same general principles, sharing a similar underlying structure. Furthermore, it is assumed that non-canonical clause types like interrogative clauses, passive clauses, and cleft clauses are related to, and therefore derived from, more basic underlying clause structures (Evans & Green, 2006: 778). According to A Comprehensive Grammar of the English Language (Quirk et al., 1985), the syntactic structure of [I VMental that P] construction can be analyzed into the matrix clause plus a complement clause. Complement clauses are commonly defined as subordinate clauses functioning as an argument of a predicate (Noonan,
72
5 The Lexical and Syntactic Properties of MM
1985: 42): they may serve as subject or object of the matrix clause. In [I VMental that P] construction, the complement clause is characterized as objects of the Verb. Quirk et al. (1985) characterize the complement clause as object of the ‘governing’ verb in the matrix clause in sentence pattern like [I VMental that P]. For example, in the following sentence, (17)
I think that there should be a ban on tobacco advertising.
that-clause is analyzed as the object of “think”. If complement clauses are clausal constituents filling a syntactic argument position, then a major prediction will be that in all positions in which a clause can appear as an argument of a verb, a noun phrase should also be possible. However, there is compelling argument against it, when the types of verbs used are investigated. Consider the following example: (18)
(a)
I reflected for the first time that I had offered to kill a man because he had been somewhat overbearing.
(a’)
* I reflected for the first time the idea of offering to kill a man because he had been somewhat overbearing.
Verbs like reflect are indisputably intransitive predicates that do not take any objects, so it is not possible to reflect something (*i.e., the idea). As evident from this example, complement clauses can occur in environments in which a noun phrase is impossible. Note some other examples: (19)
(a)
He insists that he fell instantly in love.
(a’)
*He insists the fact of falling instantly in love.
(b)
I hope that you’ll agree that Ryan Widmer is not guilty of any wrongdoing.
(b’)
*I hope your agreement that Ryan Widmer is not guilty of any wrongdoing.
Verbs like insist and hope are intransitive in certain usage and cannot take a direct object. There is also cross-linguistic evidence; consider the following example in Chinese: (20)
(a)
我认为—一切是天意。 I think—everything is the will of heaven.
(a’)
*我认为——一切是天意的看法。 * I think—the opinion that everything is the will of heaven.
The concordance of 我认为(I think) shows: it is followed by complement clauses rather than noun phrase. In generalization, calling the complement clause a direct object is grammatically prohibited. Therefore, constituency analysis of [I VMental that P] into matrix clause plus the complement clause as object has its problems.
5.2 The Syntactic Properties of MM
73
Then what is the case with [It be ADJ that P]? In syntax [It be ADJ that P] is a typical pattern of it-exposition. The term it-extraposition refers to the syntactic process of shifting a clause out of subject position to the right of the superordinate predicate while replacing it by the dummy pronoun anticipatory it (Quirk et al., 1985: 1391–1393; Kaltenböck, 2005: 119). The syntax of it-extraposition, especially its syntactic derivation, has received considerable attention in the generative literature (cf. Emonds, 1976; Rosenbaum, 1967). In general, it-extraposition is assumed to be derived from a syntactically more basic non-extraposed counterpart with canonical subject-predicatecomplement order by an extraposition movement that shifts the subject clause to the right and inserts anticipatory it in the vacated position. According to Quirk et al. (1985), extraposition operates almost exclusively on subordinate nominal clauses. In general, it-extraposition presents itself as an essentially bipartite structure consisting of a matrix clause, which is composed of anticipatory it and a matrix predicate, and a complement clause, as shown in Table 5.2. Its non-extraposed counterpart is claimed to be ‘That there’s an explanation for all this is possible’. This view of complement clauses as extraposed subject emphasizes similarities between the matrix clauses of complex sentences and simplex clauses. This is a manifestation of the preference for abstractness in grammatical analysis, emphasizing similarity and viewing differences as superficial, secondary. However, from a functional perspective, overlap between it-extraposition and the so-called non-extraposed counterpart is minimal, and they show a very different behavior in their actual use in texts (Kaltenbock, 2004). In terms of their communicative function, it-extraposition and non-extraposition do not display a one-to-one correspondence that would allow easy exchange of the two constructional types. It has been shown that it-extraposition displays a number of very specific discourse functions, which differ substantially from those for non-extraposition. As found by scholars like Kaltenbock (2004), it-extraposition and non-extraposition are considerably constrained by functional considerations such as given vs. new information structure and weight distribution, which tie each type to a specific discourse context, allowing rarely free variation. Based on a large corpus of naturally occurring data, Miller (2001: 683) investigates the information structure of it-extraposition and nonextraposition. It is shown that non-extraposition requires that the content of the subject be discourse old or directly inferable. If the content is discourse-new, then extraposition is necessary. The division of matrix and complement clause is believed to allow for the separate expression of some state of affairs (the complement clause) and some speaker Table 5.2 Structure of it-extraposition It
Is possible
That
Anticipatory it
Matrix predicate
Complementizer
Matrix clause it-extraposition
Complement clause
There’s an explanation for all this
74
5 The Lexical and Syntactic Properties of MM
comment (the matrix predicate) (Kaltenbock, 2004: 213). With reference to Halliday’s (1970) framework, Kaltenbock (2004) argues that the complement clause has experiential function, whereas the matrix clause conveys interpersonal meaning in that it is typically concerned with the speaker’s angle, including his attitudes and judgments (Halliday & Hasan, 1976: 26–27). It appears that ‘It-extraposition’ clearly separates the interpersonal component and the experiential component syntactically, using the interpersonal element as theme. Furthermore, based on corpus data, overall, it-extraposition is the most frequent variant by far with a ratio of roughly 8:1, compared with its non-extraposed counterpart. All these reveal the problems with the constituency analysis in which itextraposition is assumed to take the vacated subject position as derivation from the non-extraposition. Therefore, [It be ADJ that P] construction is not derived from a syntactically non-extraposed counterpart with canonical subject-predicatecomplement order. There must be motivated explanations otherwise.
5.2.2 The Syntagmatic Structure of MM Constituency, partly revisited above, has figured prominently in the linguistic theorizing of the twentieth century, to the virtual exclusion of all other syntagmatic relationships. Semiotic grammar (SG) adds three other types for necessary and sufficient description of grammatical phenomena: conjugational, dependency, and linking relations (McGregor, 1997). In conjugational relationships, one unit encloses and thus shapes the other, indicating how it is intended to be taken or viewed interpersonally. Conjugational relationships are taken as the syntagmatic relationships which characterize, define, and are defined by the interpersonal semiotic. Conjugational relationships may be classified along the dimension involving a binary contrast between SCOPING and FRAMING. FRAMING refers to “a relationship characterized by the fact that the encompassed whole is marked off as a demonstration rather than a description,” (ibid: 252). Differently, in SCOPING, a unit applies over a certain domain, yet without involving the sort of marking off. According to McGregor (1997)’s grammatical analysis, in complementation constructions such as I think (that) Mary will know, the matrix clause “I think” and the complement that-clause are two ‘wholes’; the former encompasses the latter, which is marked off as a demonstration of the thought contents. In this way, the matrix clause will be called the framing clause, while the complement that-clause will be called the framed clause. Sentences like ‘It was fortunate that I survived the crash’ and ‘It’s likely (that) Mary knows’ are categorized by McGregor (1997) as instantiation of scopal relationship. In his analysis, the relative clause with dummy it subject is the SCOPING clause. In SCOPING, a unit applies a certain domain, leaving its mark on the entirety of this domain. However, SCOPING does not involve the sort of marking off as FRAMING does. According to McGregor, the scoping clause “It’s likely” could be replaced by a scoping adverbial. Note the following sentences:
5.2 The Syntactic Properties of MM
(21)
a.
It’s likely (that) the business format appeals to a very narrow slice of the audience.
b.
The business format likely appeals to a very narrow slice of the audience.
75
To treating (21a) and (21b) as similar would fail to account for the difference between these two utterances: in the former, that-clause is marked off from the clause “It’s likely”, whereas in the latter “likely” is integrated very closely within the simple sentence. It appears a marking-off exists between the clause “It’s likely” and that-clause as two ‘wholes’. Another difference is that in (21a) the expression of modality is explicit whereas in (21b) it is implicit, as pointed out by Halliday (1985, 1994). McGregor’s (1997) categorizing both as scopal fails to account for this distinction. In semiotic grammar, modality is also described in terms of SCOPING syntagmatic relationship, in which a unit applies over a certain domain, leaving its mark on the entirety of this domain. The difference between modality and sentences such as ‘It’s likely (that) Mary knows’ lies in the distinction between RHETORICAL and ATTITUDINAL type of modification. This distinction does not seem to be watertight since modality can also involve epistemic stance and [It be ADJ that P] construction can also perform rhetorical function. This construction can be a way of avoiding the use of a personal pronoun (I think that …), thereby signaling reduced personal involvement (Kaltenbock, 2005: 137). In a way, [It be ADJ that P] allows the speaker/ writer to express a personal comment toward some state of affairs. Halliday (1994: 355) claims, encoding an appraisal with the impersonal pronoun it extraposition rather than a subjective projecting clause, such as I think, makes it possible for the speaker to disguise the source of the opinion and thereby make it appear to be objective. Therefore, analysis of [It be ADJ that P] in terms of SCOPING clauses is not without its problems. The syntactic account of this construction could be adequate if we consider its cognitive structure analyzed in Chap. 4. The cognitive process of putting VIEWPOINT Space onstage motivates the employing a separate full clause to represent some kind of viewpoint or attitude. In Framing, the encompassed whole is delineated from the surrounding linguistic context and marked off as to be viewed as depiction of its referent, typically as a speech or thought. The matrix clause “It be ADJ” and the complement that-clause are two ‘wholes’; the former encompasses the latter, which is marked off as a demonstration of the thought contents. In this way, the matrix clause is the framing clause, while the complement that-clause is the framed clause. As revealed in Chap. 4, the two types of Metaphor of Modality are semantically similar in that VIEWPOINT space is mapped onto the CONTENT space and the borderlines of VIEWPOINT space and Object of Conceptualization P are very distinct. Iconicly, P is framed within the higher CONTENT space. Likewise, they are formally similar: a projecting modal clause is framing the complement clause—[that P]. In other words, the syntactic properties of MM are motivated by the corresponding
76
5 The Lexical and Syntactic Properties of MM
cognitive structure, which, in turn, can account for the semantic and formal overlaps between [It be ADJ that P] and [I VMental that P].
5.3 [It be ADJ that P] Construction In this section, we are concerned with the construction schematically represented as [It be ADJ that P], where P codes the proposition (or proposal) expressed by the complement that clause.
5.3.1 The Productivity of the Construction A question that may be raised is whether [It be ADJ that P] really represents a construction. In other words, is it an independent symbolic unit or a lexical rule? This may be a type of question that in principle need not always be answered, as argued in Croft (2003). My study still considers this question and finds that the claim that [It be ADJ that P] represents an independent symbolic unit is supported by the fact that it exhibits substantial productivity. This fact will be amply illustrated by authentic usage examples from COCA: (22)
a.
it was transparent that at least part of that censure was driven by the fact that the American team was all African-American.
b.
it’s hypocritical that the company’s magazine Traveler shows really nice nature scenes and yet is printed on virgin paper, which destroys the types of scenery it shows.
c.
“Don’t you think it’s idiosyncratic that yours is the only observer group complaining about credential problems?”
d.
I think it’s idiotic that the bailiffs have guns.
e.
it was humiliating that Israel could not make its own way in the world.
f.
it is imminent that we take a long hard look at the real potential for gymnastics and how it can be developed in a way that it can regain its status
g.
it was handy that the rooftops were flat
h.
it was serendipitous that a test came to hand so easily.
i.
It was salutary that they should tell the genocidal government in Sudan and its proxies, who’ve killed 200,000, that they’re making a huge mistake if they think they will benefit by running out the clock on the Bush administration.
Such diverse examples show that we are dealing here with a very productive pattern. Take handy for example, according to Collins English dictionary, this adjective is just used as Epithet, Complement to Noun phrase, or Predicative to human subject (meaning ‘useful’; ‘nearby and therefore easy to get or reach’; ‘good at using something’). The handy example appears in a novel of 2011. The fact that this is
5.3 [It be ADJ that P] Construction
77
unproblematically understood in the sense that the content of the complement clause is viewed by the character as easy cannot be strictly based on any general, conventional use of this adjective. Thus, it presumably derives from the general recognition of the general pattern as a conventional way of expressing one’s viewpoint or evaluation on a situation. This supports the claim that it is entrenched as an independent processing unit. The ideas of frequency and predictability have played a fundamental role in models of human language processing for well over a hundred years. Relatively recent models (cf. Bybee & Scheibman, 1999; Jurafsky et al., 2001; Seidenberg & MacDonald, 1999) have proposed that probabilistic information about words and other linguistic structure is represented in the minds of language users and plays a role in language comprehension, production, and acquisition. Bybee (1995) has argued that high type frequency leads to entrenchment of the pattern rather than that of specific instances. Repetition leads to the automatization of a sequence of units and the reanalysis of the sequence as a single processing chunk, with formerly separate units losing their individual meaning. The resulting schema with an open slot is applicable to a category of items, including elements that are recognized as members of the category even if they have not been used with the pattern before. In other words, such a schema is productive. In a word, type frequency correlates with productivity. For instance, if large quantities of adjective types can occur in a construction, there would be high probability for this construction to license new adjectives. In corpus-linguistic presentation, the pattern is ‘it followed by BE followed by an adjective, followed by a that-clause’. With it in the concordance as the starting point, followed by the copular [be] and adjective represented by [j*], examples of the predicative adjectives of that-clauses were collected. The researcher also checks the index lines and makes a judgment on the accuracy of the identification of the grammatical frame—it [be] ADJ that-clause. Table 5.3 lists the frequencies of predicative adjectives used in [It be ADJ thatclause] pattern in COCA. It is apparent that the type frequency of [It be ADJ thatclause] pattern is substantially high: total 535 types, with 33,150 token instances of this pattern among 450 million words. In other words, the type frequency is 535, and the normalized token frequency is 73.67 per million. Therefore, this pattern should be well-entrenched as a schema and productive as well. Note that we do not count the frequency of zero that-clause, since extraposed zero that-clause is rare in written and spoken text, as the previous study by Kaltenbock (2004) reported that extraposed zero that-clause only has a frequency of 0.17 per 10,000 words. The current study also checked this finding with the sample data from COCA. Since zero that-clause has to be manually identified, it is not feasible to read all the concordance lines of [It be ADJ] in COCA, which is estimated to be at least more than 33,150 instances. The sample data is taken completely at random from COCA and can be downloaded on http://corpus.byu.edu/full-text/samples.asp, containing 1.7 million words. As explicit grammatical marker is lacking in zero that-clause, there will be too many ‘noisy’ index lines. Therefore, we put each of the above 535 adjectives into the slot of [It be ADJ] and check their frequencies in the sample data. Take likely as an example:
78 Table 5.3 Predicative adjectives used in [It be ADJ that-clause] pattern in COCA
5 The Lexical and Syntactic Properties of MM
Tokens per adjective
Adjectives (types)
4700
Clear
Total no. of tokens 4700
3218
True
3218
2946
Possible
2946
2105
Important
2105
1586
Likely
1586
1257
Obvious
1257
1164
Unlikely
1164
834
Interesting
834
543
Imperative
543
537
Apparent
537
487
Evident
487
466
Essential
466
435
Unfortunate
435
372
Ironic
372
355
Good
355
352
Inevitable
352
326
Doubtful
326
317
Critical
317
290
Significant
290
276
Amazing
276
267
Conceivable
267
250
Noteworthy
250
247
Great
247
236
Understandable
236
213
Crucial
213
201
Vital
< 200
506ADJs
Total no. of types
535
201 9169 33,150
Line 1 is coded as zero that-clause, whereas Line 2–6 fit the [It be ADJ that-clause] pattern. See another example with possible:
5.3 [It be ADJ that P] Construction
79
Line 1, 2, 4, 6, 13–15 fit the [It be ADJ that-clause] pattern, while Line 3 and Line 7 are coded as zero that-clause. It can be seen that not only complement clause, to-infinitive can also collocate with [It be ADJ], and [It be ADJ] can also stand on its own as an independent clause, which are, of course, ‘noise’ with respect to our task of studying the pattern involving complement clause. In this way, we check each type, and find that in total the ratio of that-clause to zero that-clause is larger than 12: 1. Considering the rare occurrence of zero thatclause and the likely accompanying noises if it were taken into consideration in the search, we choose to concentrate on the [It be ADJ that-clause] pattern. The insight that [It be ADJ that-clause] pattern is entrenched as a productive construction could account for the problematic phenomenon noted in Sect. 5.1. Since overlap between it-extraposition and non-extraposition is minimal from a functional perspective, and they show a very different behavior in their actual use in texts (Kaltenbock, 2004: 298), it works better if this pattern is treated not as derived from non-extraposition but an independent construction. Based on corpus data, overall, itextraposition is the most frequent variant by far with a ratio of roughly 8:1 and as such can be taken as the statistically unmarked form compared with its non-extraposed counterpart. I have also surveyed Corpus of Historical American English (COHA). In 1815, the early stage of [It be ADJ that-clause] pattern emergence, this construction is limited to only a few adjectives (8): observable, evident, self-evident, true, probable, strange, wonderful, and sufficient. Comparing them to the predicative adjectives in Table 5.3 which is retrieved in 2010s, I find that this construction, over time, come to be used with a larger number of collocates (535). As a summary of related analyses in last and current chapters, a variety of facts justify us in viewing [It be ADJ that-clause] pattern as a grammatical construction, that is, as a conventional pairing of form and meaning (Fillmore et al., 1988; Goldberg, 1995, 2006). First, the construction assigns an idiomatic interpretation to sentences having the general form [It be ADJ that-clause] pattern: in all cases a viewpoint toward the proposition P represented by that-clause is entailed and profiled. Moreover, despite this idiomatic interpretation and unusual argument structure, the construction is used productively with a diverse array of predicates. One way or another, this construction must be listed as a conventional part of English grammar.
80
5 The Lexical and Syntactic Properties of MM
5.3.2 The Lexical Properties of the Construction The ADJs used in [It be ADJ that-clause] pattern in COCA are assigned to semantic categories based on categorizations by other studies. For the most part, despite the use of different terminology by researchers, there is consensus on the semantic roles of lexical items. Quirk et al., (1985: 1224) distinguish subtypes of adjectives with extraposed it distinguished by in A Comprehensive Grammar of the English Language, reproduced with some modification in Table 5.4. Herriman (2000) had conducted a semantic classification of the matrix predicates, including predicative adjectives, of extraposed that clauses in the Lancaster-Oslo/ Bergen corpus (LOB). The chief semantic category represented by the matrix predicates of extraposed that-clauses, found by Herriman (2000) in LOB, which consists of a million words of written British English, is epistemic modality, i.e., the speaker’s opinion of the truth-value of the extraposed clause. About a third of the matrix predicates of extraposed that-clauses represent evaluation, i.e., various value judgments of the content of the extraposed clause. The representation of deontic modality, which involves the intention, desire, will of the speaker or a participant in the clause as regards the content of the extraposed clause, is rather unusual, and the representation of dynamic modality is extremely rare. His operation with an elaborate system of semantic categories and subcategories provides a useful framework for the semantic classification of predicative ADJs. What follows is Herriman (2000)’s semantic classification of the matrix predicates of extraposed clauses in LOB (Table 5.5). In this table, the number of occurrences is given in brackets after each item that occurs more than once in the corpus. In Herriman (2000)’s classification, Evaluation as a rather broad term concerning Appropriateness, Significance, Frequency, Responsibility, Emotive Reaction, and even General. This shows it requires more systemic categorization. Martin and White (2005: 35) has labeled Evaluation as attitude that “is concerned with our feelings, including emotional reactions, judgments of behavior and evaluation of things”. Their system of appraisal is the result of linguistic Table 5.4 Types of adjectives with extraposed it Type
Semantic feature
Examples
A
Deontic modality
Essential necessary compulsory obligatory imperative appropriate proper improper
B
Epistemic modality
Apparent evident likely unlikely possible true untrue certain clear implicit obvious well-known plain impossible
C
Emotive
Curious odd peculiar dreadful sad silly fortunate unfortunate tragic alarming annoying pleasing depressing shocking disappointing surprising embarrassing frightening irritating admirable
5.3 [It be ADJ that P] Construction
81
Table 5.5 Matrix ADJ predicates of extraposed that-clauses (Herriman, 2000: 597–598) Epistemic modality
Truth: AdjPs: true (52), clear (35), possible21 (32), obvious (30), evident (23), likely (20), probable (18), certain (13), unlikely (9), apparent (8), impossible (6), inevitable (6), arguable (4), inconceivable (3), conceivable (2), improbable (2), incredible (2), plain (2), known, unthinkable
Evaluation
Appropriateness: AdjPs natural (12), right24 (6), appropriate (3), enough (3), odd (3), sufficient (3), understandable (3), ironical (2), strange (2), curious, fatuous, ironic, logical, odd, peculiar, reasonable, suitable, unthinkable Significance: AdjPs: important (13), significant (4), noteworthy (4), far (from our minds), fundamental, insignificant, minor, remarkable, worthy of note. General: AdjPs: unfortunate (4), good (3), better (2), lamentable (2), distasteful, fortunate, marvelous, shameful, welcome Emotive Reaction: AdjPs: surprising (18), interesting (5), astonishing (2), encouraging (2), amazing, gratifying, puzzling, sobering Frequency: AdjPs: common (2), standard, widely held Responsibility: AdjPs: accidental (2)
Deontic modality
Obligation: AdjPs: essential (10), necessary (6), expedient, imperative, mandatory, vital. Volition: AdjPs: desirable, undesirable
Dynamic modality
Human Attribute: AdjP: characteristic of someone’s generosity of spirit
research over a period of 20 years, and the subclassification of affect, judgment, and appreciation was established in this theory based on extensive discourse analysis. Affect systems characterize phenomena by reference to emotion, and it corresponds to affective modality in my study. Judgment systems consist of resources for morally evaluating human people and their behavior, by reference to a set of norms, which roughly correspond to deontic modality, which is a well-received term and is included in the current study. Appreciation is the institutionalization of feelings with a view to assessing the value and social significance of things. It is organized around three variables: reaction, composition, and valuation. Reaction concerns the degree to which the text/process in focus captures our attention and the emotional impact it has on us (Martin, 2010). Composition concerns perceptions of proportionality/balance and detail in a text/process. Valuation has to do with our assessment of the social significance of the text/process. This class will be defined as evaluative modality in my study, and of course, here it adopts a narrow sense of evaluation, that is, appreciation. There is another extensive study of subject it-extraposition in the one-millionword ICE-GB corpus (short for International Corpus of English, Great Britain section) by Kaltenbock (2004). Semantically, the predicative adjectives governing extraposed that-clause are divided into four different classes: (a) adjectives referring to the truth or transparency of the embedded proposition; (b) adjectives referring to the likelihood/possibility of a state of affairs; (c) adjectives pointing at the importance or newsworthiness of an event; and (d) adjectives expressing value judgments of a more specific nature. Even Kaltenbock (2004) himself admits that the borderlines between his categories are “of a fuzzy nature” (p. 87). Class (c) expresses some kind
82
5 The Lexical and Syntactic Properties of MM
of value judgment on the content of the complement clause and thus represents a subgroup of (d). Of course, classifications based on semantic criteria involve a number of problems. The criteria by which semantic categories may be distinguished from one another are no foolproof, clear-cut. Besides referring to Herriman (2000)’s and Martin and White (2005)’s framework with some kind of modification as shown above, I have used a semantic tool—the online FrameNet lexical database. The information may be accessed by starting from lexical units (or starting from the frames). We can get a definition of the frame, together with a detailed description of the relevant grammatical patterns. To illustrate, I will present an example of frame analysis, the lexical unit ‘predictable’. Starting from the lexical unit ‘predictable’ as a search word, we find it belongs to the frame “Expectation”. Then we can get the definition of the frame “Expectation” and the relevant grammatical patterns, which are shown as follows: Expectation Definition: Words in this frame have to do with a cognizer believing that some phenomenon will take place in the future. Some words in the frame (e.g., foresee.v) indicate that the phenomenon is asserted also to be true, while others do not Lexical Units: anticipate.v, await.v, expect.v, expectation.n, foresee.v, foreseeable.a, predict.v, predictable.a, prediction.n, premonition.n, unexpected.a, unforeseeable.a, unpredictable.a, wait.v Annotation predictable.a • 010-s20-it-extrapos-that 1. It was [Degree almost] PREDICTABLE Target [Phenomenon that a perfectionist like Pérignon would not rest until he found a bottle strong enough to withstand the internal pressure exerted by a sparkling wine] .[Cognizer INI] 2. It was PREDICTABLE Target [Phenomenon that another important single issue could bring the students back out of their classrooms and onto the streets] .[Cognizer INI] (https://framenet.icsi.berkeley.edu/luIndex. Note: the lexical unit ’predictable’ is presented in capitalization within the annotating examples; only a small relevant portion is reprinted here.) From frame analysis of “predictable”, we get that “predictable” can be semantically tagged as “Expectation”; and judging from the definition of “Expectation”, we categorize it into “epistemic modality”. In an attempt to be systematic and exhaustive, COCA, probably the only large and balanced corpus of American English is searched for the pattern It be ADJ that.
5.3 [It be ADJ that P] Construction
83
With it in the concordance as the starting point, followed by the copular [be] and adjective represented by [j*], examples of the predicative adjectives of that-clauses were collected. I set the minimum frequency as 10 per 450 million words, which is the size of COCA. The purpose is not to overlook adjectives of relatively low frequency in this pattern, as high type frequency leads to entrenchment of the pattern rather than that of specific instances (Bybee, 1995). The full classification of the predicative adjectives used in [It be ADJ that] pattern in the COCA corpus is given in Appendix 1, which is summarized in Table 5.6. The chief semantic category represented by the ADJs in [It be ADJ that-clause] construction in the corpus is epistemic modality, i.e., the speaker’s opinion concerned with the interpretation of the world via the laws of human reason and with the expression of the degree or nature of the speaker’s commitment to the truth of what he says. This type of opinion is represented by nearly two-thirds (65.9%) of the corpus examples. The subtypes of this category are exemplified as follows in descending order of their frequency in the corpus. Note that only subtypes that have a minimum frequency of 1% are summarized here.
Table 5.6 Semantic classification of adjectival predicates in [It be ADJ that-clause] construction Semantic category
Subtype
Examples
Epistemic modality
Possibility
Possible, impossible, likely
Certainty
Certain, undeniable, doubtful
Actuality
True, correct, incredible
Usuality
Rare, unusual, rare
Obviousness
Obvious, apparent, clear
Perception
Understandable, unbelievable, unthinkable, incomprehensible
Expectation/Expectedness
Natural, predictable, ironic
Affective
Deontic modality
Evaluative modality
Deduction
Logical, reasonable, paradoxical
Intensifying
Great, fantastic, fascinating
Un/happiness
Sad, tragic, distressing
Dis/satisfaction
Regrettable, disappointing, outrageous
In/security
Surprising, funny, encouraging
Obligation
Mandatory, urgent
Necessity
Necessary, essential, imperative
Acceptability
Good, okay, unacceptable
Morality
Wrong, shameful, unconscionable
Impact
Interesting, impressive, exciting
Valuation
Important, significant, critical
Quality
Neat, nice, wonderful
Composition
Appropriate, incomprehensible, proper
84
5 The Lexical and Syntactic Properties of MM
Obviousness (24.3 %) (23)
Still, it is clear that such groups are not fully under the government’s control. (120616, Newspaper)
Possibility (21.9 %) (24)
it is possible that some participants may have underreported smoking for reasons of social desirability. (2012, Academic)
Actuality (11.4 %) And it is true that his weaknesses remain. (120313, Spoken)
(25)
Expectation/Expectedness (3.9%) (26)
In today’s global society, musical recontextualization is a reality, and as global media allows music to be shared with astonishing ease, it is natural that fusion genres emerge. (2011, Magazine)
Certainty (2.1%) (27)
In 20 years it’s certain that leadership will be a continued need. (2011, Magazine)
Usuality (1.8%) it’s rare that you should ever do something like that. (110914, Spoken)
(28)
About one-fifth (18.9%, with a token frequency of 13.9 per mil) of the ADJs in [It be ADJ that-clause] construction in the corpus represent Evaluative Modality, which involves an institutionalized view to assessing the value and social significance of things. The subtypes of this category are exemplified (in descending order of their frequency in the corpus) as follows: Valuation (12.6%) (29)
As this study highlights, it is important that assessments be used only for those purposes for which they were designed and for which there are data validating their use. (2010, Academic)
Impact (3.1%) (30)
But it is interesting that obesity and diabetes are increasing as autism incidence is. (2012, Magazine)
Quality (2.4%) (31)
It’s nice that he’s cool. (120315, Spoken)
A somewhat small portion (8%) of the ADJs in [It be ADJ that-clause] construction in the corpus represent Deontic modality, which involves a gradual scale going from absolute moral necessity via the intermediary stages of desirability, acceptability, and undesirability, to absolute moral unacceptability. In other words, the token frequency of this category is only 5.9 per million. The subtypes of this category are exemplified as follows (in descending order of their frequency in the corpus): Necessity (5.7%)
5.3 [It be ADJ that P] Construction
(32)
85
It is necessary that administrators demonstrate effective leadership at education institutions where human resource is effectively educated. (2012, Academic)
Acceptability (1.9%) (33)
it’s good that we will never be able to compose ourselves into a neat little still-life, but couldn’t we, for God’s sake, begin to understand what we meant to each other? (2006, Fiction)
Finally, an even smaller proportion (6.1%) of the ADJs in [It be ADJ that-clause] construction in the corpus represent Affective Evaluation, which is concerned with the emotional side of human behavior and involves the fundamental constructs of receiving, responding, and valuing in hot cognition. The subtypes of this category are exemplified as follows (in descending order of their frequency in the corpus): In/security (2.8%) (34)
For such a difficult case, it’s surprising that the jury was out just four hours. (120428, Spoken)
Un/happiness (2.5%) (35)
I think it is sad that we make movies and we can’t show them. (110724, Newspaper)
In summary, based on the token frequencies of [It be ADJ that-clause] construction types, we find that the chief semantic category is epistemic modality, i.e., the speaker’s opinion concerned with the interpretation of the world via the laws of human reason and with the expression of the degree or nature of the speaker’s commitment to the truth of what he says. This type of opinion is represented by about two-thirds of the corpus examples: 65.9% in the current study, compared with 64% in Herriman (2000). Nearly one-fifth (about 18.9%) of the ADJs in [It be ADJ that-clause] construction in COCA represent evaluative modality. The types of modality in Halliday and Matthiessen’s (2004: 618) classification include probability, usuality, obligation, and inclination, which can be put under the headings of epistemic and deontic modality. We believe that Evaluative Modality should be included as well. Example (29) is a case in point: ‘it is important that assessments be used only for those purposes for which they were designed and for which there are data validating their use’ indicates that the proposition represented in that-clause is not a fact; in other words, it is intermediate between the positive and the negative poles of polarity. The framing clause “it is important” expresses the speaker (writer)’s judgment on the status of what is being said. In this way, utterances that realize evaluative modality like example (29) should be included in the systemic categories of modality. There is none modal verb expressing evaluative modality. In contrast, [It be ADJ that-clause] construction, as a major construction type of Metaphor of Modality, can realize evaluative modality. We believe, this is very much one of the motivations for MM, expanding the semantic domains of
86
5 The Lexical and Syntactic Properties of MM
modality. Nearly one-fifth of [It be ADJ that-clause] constructions representing Evaluative Modality clearly demonstrates the linguistic consequence of this motivation of semantic expansion. In terms of Deontic modality, a smaller proportion is found in his corpus than in COCA: 4 versus 8%. Herriman (2000) defines deontic modality as the intention, desire, will of the speaker or a participant in the clause as regards the content of the extraposed clause. This seems incomplete and incongruent with well-received definition. In the current study, deontic modality is defined in terms of social or institutional laws, which involves a gradual scale going from absolute moral necessity via the intermediary stages of desirability, acceptability, and undesirability, to absolute moral unacceptability (Nuyts, 2005: 9), which is not that narrow. In this sense, our findings do not contradict with Herriman’s (2000), and are, in a way, more comprehensive and refined. Finally, another 6.1% represent affective modality, which also expands the semantic domains of modality. There are also constraints on what types of adjectives can be put into the slot of this construction. As revealed from the above-sorted statistics, certain semantic groups of adjectives instead of all adjectives are compatible with this construction. For example, safe occurs in somewhat similar patterns: it [be] safe, it [be] safe to-inf, and it [be] safe for X to Y, but does not occur in [It be ADJ that p] construction, as exemplified below: (36)
(37) (38)
They are overworked and the equipment is a bit old and possibly inadequate, and it might not be as safe as we’d like. Is that fair? Mr. NANCE: Well, it is safe, I want to make sure that everybody understands that. The system, by any reasonable measure, is safe. (19910802, Spoken) a.
It’s safe to surf this beach. (20010806, Spoken)
b.
?* It’s safe that we surf this beach.
a.
He certainly wants to remain until the situation in Somalia settles down enough so that it’s safe for him to return. (19930625, Spoken)
b.
?* He certainly wants to remain until the situation in Somalia settles down enough so that it’s safe that he returns.
Adjectives usually specify both the property and the scalar position, that is, situating its trajector vis-à-vis a scale representing the degree to which it exhibits a certain property. Adjectives that just specify the property and the degree exhibiting a certain property of a particular situation, be it be state, an activity, or some focal entity, like safe, are not adequately compatible with [It be ADJ that P] construction. Only adjectives involving the mental states and processes of the speaker (writer), the addressee, or the others whose perspective is adopted or otherwise considered, in terms of the laws of human reason and the degree or nature of the speaker’s commitment to the truth of what he says (epistemic modality), our assessment of the impact, quality, significance of the proposition (Appreciation), shared social or institutional laws (Deontic modality), or the human emotional side and the fundamental constructs of receiving, responding, and valuing in hot cognition (Affective judgment) are compatible with this construction. In other words, the adjectives tend to involve
5.4 [I VMental that P] Construction
87
the epistemic, affective, deontic states and processes of the speaker (writer) and the addressee, or involve a subjective assessment concerning the status of a proposition. These preferences and constraints provide the evidence for status of the pattern [It be ADJ that p] as a schema rather than a rule. In other words, it is an independent construction, not a generalized rule. The members of the adjective category combined with the copula be are generally interpreted as ascribing a property to a referent. However, observe the following utterances in (39) from COCA: (39)
a.
it is essential that the nurse fully informs parents of both the risks and the appropriate preventative measures
b.
it is clear that such groups are not fully under the government’s control.
c.
it is important that people are able to talk about it
d.
It’s likely that my genes predisposed me to have a mood disorder
e.
it was imperative that we got a chance to hear it again.
f.
It’s great that “Newsweek” has decided to do the story.
The predicate in (39a–f) does not ascribe a property to the referent, the proposition represented by that-clause. Instead, it should be interpreted as encoding the epistemic, affective, deontic stance or evaluation held by some unspecified viewers toward the proposition, and this semantic interpretation is achieved via the interaction between the adjectives and the syntactic structure [it be ADJ that P].
5.4 [I VMental that P] Construction In Chap. 4, we have analyzed the semantic structure of [I VMental that P] construction. Note the following example sentences from COCA that are identical in syntactic specification: (40)
a.
I think that for Hugo Chavez, it’s a problem that George Bush is not in office.
b.
I mention that I’ve done speech and debate for about three years and it’s helped me a lot.
c.
I pray that all comforts have been provided you and your fellow travelers.
d.
I explain that our work is research in progress that has yet to be accepted by the scientific community.
e.
I ask that Congress tell our children why declaring war against Iraq is necessary for peace in the world.
f.
I say that I appreciate the sentiment, but there’s a greater issue here.
88
5 The Lexical and Syntactic Properties of MM
It becomes evident that not only Mental Verbs, verbs of a ‘say’ type can also be put into the same slot. All these examples can be represented as [I Verb that P] construction, which is a more schematic construction at a higher level of the constructional network, in which [I VMental that P] construction is a subtype. We also distinguish this performative use from the formulaic use of a CT-clause, short for complement-taking clause. The latter is historically related to the performative use, from which it developed through grammaticalization. The following examples illustrate the formulaic use of a CT-clause: (41)
You’re right, I guess.
(42)
She left I think.
(43)
I mean, I think we are confused right now.
In these examples, a CT-clause is a holistic formula functioning as an epistemic marker or attention getter that is only loosely neighbored to the COMP-clause (short for complement clause), which is really an independent utterance. This is evidenced by a number of features that characterize the formulaic use (c.f. Hooper, 1975; Thompson & Mulac, 1991; Diessel & Tomasello, 2001): a. The CT-clauses are always short and formulaic. b. Since the COMP-clause is non-embedded both formally and conceptually, it does not include a that-complementizer. c. The order of CT-clause and COMP-clause is variable: the CT-clause may precede or follow the COMP-clause or may even be inserted into it. Since what we intend to study is the pattern of [I Verb] taking an embedded COMP-clause rather than the formulaic use of a CT-clause, so the constructions we study have that-complementizer. Still, the performative use and the formulaic use may form a continuum with many intermediate cases. Nevertheless, the formulaic use of CT-clauses is not included from the present study, since its unique pragmatic properties and formal features suggest the status of another independent construction, which needs to be treated separately in some other future projects. The first-person pronoun is the most basic expression of speaker subjectivity in English. “I signifies ‘the person who is uttering the present instance of the discourse containing I’” (Benveniste, 1971:218). Subjectivity is characterized by Benveniste as the ability of speakers to view themselves as subjects and he discusses how grammatical categories like person in the form of personal pronouns contribute to this expressive capacity of speakers. The first-person singular pronoun is a rich source of subjectivity in language in that it explicitly refers to the speaker.
5.4 [I VMental that P] Construction
89
When the issuer presents a proposition as being his/her own responsibility, the first-person pronoun is used. It is also the most explicit type of anchoring the speaker’s stance in discourse. [I Vmental that P], such as I think that for Hugo Chavez, it’s a problem that George Bush is not in office, typically qualifies the speaker’s level of commitment to the truth of the information. Working in the general description of English, a number of researchers have developed accounts of verb patterns and classification of verbs in delicacy. These include the work by Fred Householder and his team from the 1960s and the work by the Lexicon Project at the MIT Center for Cognitive Science 1983–1987 distilled and presented by Levin (1993). Martha Palmer has developed VerbNet based on Levin (1993), adding classes, verb senses, and links to both WordNet and FrameNet. VerbNet (VN) (Kipper-Schuler, 2006) is the largest online verb lexicon currently available for English. It is a hierarchical domain-independent, broad-coverage verb lexicon with mappings to other lexical resources such as WordNet and FrameNet (Baker et al., 1998). VerbNet is organized into verb classes extending Levin (1993) classes through refinement and addition of subclasses to achieve syntactic and semantic coherence among members of a class. Each verb class in VN is completely described by selectional restrictions
Fig. 5.2 Simplified VerbNet entry for a verb class (https://uvi.colorado.edu/uvi_search_anywhere)
90
5 The Lexical and Syntactic Properties of MM
on the arguments, and frames consisting of a syntactic description and semantic predicates. Using VerbNet, we can find the verb class to which each verb form belongs and its syntactic frames together with corresponding semantics. Below is a simplified VerbNet entry (Fig. 5.2). As for the lexical items that are used in the open slot in constructions, sometimes this slot is highly generalized, for example, taking all nouns or all verbs of the language. At other times, the slot is semantically constrained. Bybee and Eddington (2006) found evidence that high-frequency lexical instances of constructions can act as the central members of the category formed by members of the open slot in a construction. In order to investigate the lexical properties of [I Vmental that P] constructions on a comprehensive empirical base, we first survey which verbs can be put into the verb slot of [I Verb that P]. Our approach is bottom-up, which is complimentary to the perspective ‘from above’ adopted in SFG. The syntactic frame of [I Verb that P], in the metalanguage of VerbNet, fits into NP V THAT S. Therefore, we first search VerbNet and find all the verb classes and their members that have this syntactic frame. As a result, we get a list of verbs. I use this customized list as part of the pattern: I [[email protected]: Verb] that. [CST] ([email protected] identifies the user; CST is the part of speech tag for that as conjunction) and put it as search string in the large corpus COCA. It turns out that some verbs and verb classes have the syntactic frame NP V THAT S but do not occur in the pattern [I Verb that-P] in COCA. Take “cable” as an example; it has syntactic frame NP V THAT S as shown below. (44)
Heather cabled that the party would be tonight.
However, there is no record of [I cable] being followed by that-clause. Finally, the dataset of verbs and verb classes having the syntactic frame of [I Verb that P] is summarized in Table 5.7.
5.4 [I VMental that P] Construction
91
Table 5.7 Verbs and verb classes with the syntactic frame [I Verb that P] Class
Semantics
Members
Learn-14–2
Transfer_ information
Learn, read,
Declare-29.4
Declare
Admit, affirm, assert, aver, claim, confirm, discover, grant, guarantee, maintain, observe, reaffirm, show, prove, declare, find, judge
Conjecture-29.5–1
Believe
Allow, anticipate, believe, conjecture, deny, determine, doubt, expect, feel, foresee, guess, hold, infer, know, mean, presume, realize, recognize, speculate, surmise, suspect, swear, trust
Consider-29.9
Consider
Acknowledge, deem, reckon, report, view, know, recognize, see, consider, imagine, remember, believe, think, feel, posit, suppose, suspect
See30.1
Perceive
See, feel, sense, notice, perceive, hear, detect
Admire-31.2
Emotional state
Admire, adore, affirm, appreciate, believe, dread, fancy, lament, prefer, reaffirm, resent, respect, stand, support, tolerate, treasure, trust, worship
Estimate-34.2
Assess
Count, estimate, guess, judge
Say-37.7–1
Transfer information
Advise, announce, claim, confess, confide, declare, mention, note, observe, propose, remark, repeat, report, reveal, say, state, volunteer, recommend, suggest
Manner_ speaking-37.3
Transfer information
Call, mumble, murmur, scream, shout, whisper, yell
Complain-37.8
Transfer information
Brag, complain, object,
Advise-37.9–1
Transfer information
Advise, warn
Confess-37.10
Transfer information
Acknowledge, admit, reveal
Promise-37.13
Transfer information
Ascertain, guarantee, promise
Order-60–1
Command
Ask, command, demand, order, require
Wish-62
Desire
Dream, expect, hope, imagine, intend, mean, plan, propose, think, wish
Allow-64
Allow
Allow
Accept-77
Approve
Accept, understand
Discover-84
Discover
Ascertain, deduce, determine, discover, find, guess, rationalize, realize, hear, learn, read
Comprehend-87.2
Understand
Grasp, understand
Care-88.1
Emotional state
Care, mind, wonder, worry
Deduce-97.2
Conclude
Conclude, deduce, gather, infer, prove, reason, rationalize
92
5 The Lexical and Syntactic Properties of MM
In traditional grammar, the system of the verb is defined as transitivity and was thus concerned with categories such as transitive verb, intransitive verb, and copula verb; but in the last 50 years or so, linguists have upgraded their accounts of transitivity from the domain of the verb to that of the clause and they have as a result been able to develop more holistic and explanatory accounts of transitivity (e.g., Halliday, 1967; Hopper & Thompson, 1980). Looking at this system from above, from the vantage point of semantics, systemic linguists characterize it as a resource for construing a quantum of change in our experience of the flow of events around us and inside us. The transitivity system construes the world of experience into a manageable set of process types (Halliday & Matthiessen, 2004: 170). A key part of the SFL account of transitivity is the description of the system of process type. The entry condition is ‘major clause’, the terms are ‘material’, ‘behavioral’, ‘mental’, ‘verbal’, ‘relational’, and ‘existential’ in the description presented fully first in Halliday (1985). This system is located at the least delicate end of the lexicogrammatical cline from grammar to lexis. These primary process types are further divided into subtypes, with a further step in delicacy, with examples of verbs serving in the different types. The examples of verbs serving in the subtypes of mental clause, for instance, are given in Table 5.8, which are adapted from Halliday and Matthiessen (2004: 210; 255). Matthiessen (2014) explores the possibility of linking the SFL description of process type to large-scale classifications of process configurations of the kind developed in lexical resources such as WordNet and FrameNet, as well as Levin’s (1993) verb classes and alternations, and reports on his mapping of process type onto Levin’s verb classes. His project has demonstrated that each verb class can be interpreted at clause rank as a transitivity configuration consisting of a process, and it is possible to classify Levin’s verb classes under the terms of the system of process type. The differentiation of verb classes and the number of verbs allocated to each process type Table 5.8 Examples of verbs serving as process in mental clauses (Adapted from Halliday & Matthiessen, 2004: 210) Type
Semantic feature
Examples
1
Cognitive
Think, believe, suppose, expect, consider, know, understand, realize, appreciate, imagine, dream, pretend, guess, reckon, conjecture, hypothesize, wonder, doubt, remember, recall, forget, fear, strike, convince, remind, escape, puzzle, intrigue, surprise
2
Desiderative
Wish, hope, decide, agree, want, desire, intend, plan, resolve, determine, comply, refuse
3
Perceptive
Perceive, sense, see, notice, feel, glimpse, hear, overhear, taste, smell
4
Emotive
Like, fancy, love, adore, dislike, hate, detest, despise, loathe, abhor, rejoice, exult, grieve, mourn, bemoan, bewail, regret, deplore, fear, dread, enjoy, relish, marvel, allure, attract, please, displease, disgust, offend, repel, revolt, gladden, delight, gratify, sadden, depress, pain, alarm, startle, frighten, scare, horrify, shock, comfort, reassure, encourage, amuse, entertain, divert, interest, fascinate, bore, weary, worry
5.4 [I VMental that P] Construction
93
can be taken as a measure of its systemic elaboration—the degree to which each process type is extended further in delicacy. In our study, we also attempted a classification of verb classes under the terms of process types, which then is shown in Table 5.9. As Table 5.9 shows, there seems a minor mismatch between verb classes in VerbNet and the systemic functional account of the system of process type. The notion of process type is basically situated along the ideational line of meaning in the clause, and therefore, it alone may not be able to describe the lexical properties of [I Verb that P] in adequate delicacy, since [I Vmental that P] construction of this pattern, as a major construction type of Metaphor of Modality, is also operating on the interpersonal line of meaning. Appraisal theory, an extension to the basic ideas of interpersonal meaning in SFL, is an appropriate complement. With insights from this more recent framework, the class 34.2 ‘Estimate’ verbs can be coded as ‘Evaluative’, which also belongs to mental process. In terms of modal semantics, the types of meanings represented by [I Verb that P] construction could be grouped into four chief semantic categories: Epistemic, Affective, Boulomaic, and Verbal. The working definitions of each category are shown in Table 5.10. Table 5.11 is the classification of [I Verb that P] constructions. The numbers in the brackets indicate type frequency, which counts how many different items occur in the Verb slot of [I Verb that P] construction. I also make a few additions. For instance, in VerbNet, “argue” does not have the syntactic frame NP V THAT S, which turns out not to be the case when we search COCA, as demonstrated in the following examples: (45)
a.
I argue that political science is not a science.
b.
Anyway, I argued that Noah was faking it.
c.
and I argued that that was not true.
d.
I argue that he was incredibly engaged with his family, friends, and music
e.
I argue that he wasn’t as dependent on her as it seemed.
f.
Through his character portrayals, I argue that Faulkner seeks, but never finally answers, that basic question.
g.
I argue that it’s just as important, maybe even more important, to understand what it is to be human.
We can find 369 occurrences of [I argue that-clause] in COCA. Not only argue, other verbs like reason, mean, demonstrate, rationalize, prove, reason, show can also be put into the Verb slot of [I Verb that P] construction, and they all belong to the semantic frame of reasoning, meaning “argue, demonstrate, or prove that something is the case” (https://framenet.icsi.berkeley.edu/fndrupal/index.php?q=frameIndex). Therefore, I add “Argue (reasoning)” as a subtype of Epistemic Verbs, with the above verbs as members. Another revision concerns the split in the classification of certain verbs. A case in point is “note”. Besides fitting the semantic frame of statement and
94
5 The Lexical and Syntactic Properties of MM
Table 5.9 Verbs, verb classes, and process types with the syntactic frame [I Verb that P] Class
Semantics
Members
Process type 2
Process type 1
Learn-14–2
Transfer_ information
Learn, read,
Cognitive
Mental
Declare-29.4
Declare
Admit, affirm, assert, aver, claim, confirm, discover, grant, guarantee, maintain, observe, reaffirm, show, prove, declare, find, judge
indicating
Verbal
Conjecture-29.5–1
Believe
Allow, anticipate, believe, conjecture, deny, determine, doubt, expect, feel, foresee, guess, hold, infer, know, mean, presume, realize, recognize, speculate, surmise, suspect, swear, trust
Cognitive
Mental
See30.1
Perceive
See, feel, sense, Perceptive notice, perceive, hear, detect
Mental
Admire-31.2
Emotional state
Admire, adore, affirm, appreciate, believe, dread, fancy, lament, prefer, reaffirm, resent, respect, stand, support, tolerate, treasure, trust, worship
Emotive
Mental
Estimate-34.2
Assess
Count, estimate, guess, judge
*
*
Say37.7–1
Transfer information
advise, announce, claim, confess, confide, declare, mention, note, observe, propose, remark, repeat, report, reveal, say, state, volunteer, recommend, suggest
indicating
Verbal
Manner_speaking 37.3
Transfer information
Call, mumble, murmur, scream, shout, whisper, yell
quoting
Verbal
(continued)
5.4 [I VMental that P] Construction
95
Table 5.9 (continued) Class
Semantics
Members
Process type 2
Process type 1
Complain-37.8
Transfer information
Brag, complain, object,
Indicating
Verbal
Advise-37.9–1
Transfer information
Advise, warn
indicating
Verbal
Confess-37.10
Transfer information
Acknowledge, admit, reveal
indicating
Verbal
Promise-37.13
Transfer information
Ascertain, guarantee, promise
indicating
Verbal
Order-60–1
Command
Ask, command, demand, order, require
Imperating
Verbal
Wish -62
Desire
Dream, expect, hope, imagine, intend, mean, plan, propose, think, wish
Desiderative
Mental
Allow-64
Allow
Allow
Desiderative
Mental
Accept-77
Approve
Accept, understand
Emotive
Mental
Discover-84
Discover
Ascertain, deduce, determine, discover, find, guess, rationalize, realize, hear, learn, read
Cognitive
Mental
Comprehend-87.2
Understand
Grasp, understand
Cognitive
Mental
Care-88.1
Emotional state
Care, mind, wonder, worry
Emotive
Mental
Deduce-97.2
Conclude
Conclude, deduce, gather, infer, prove, reason, rationalize
Cognitive
Mental
Table 5.10 Working definitions of semantic categories adopted in classification of modality Category
Working definition
Epistemic
Concerned with the interpretation of the world via the laws of human reason and with the expression of the degree or nature of the speaker’s commitment to the truth of what he says
Affect
Concerned with the emotional side of human behavior and involves the fundamental constructs of receiving, responding, and valuing in hot cognition
Boulomaic Relating to the human desire Speech act Refers to the uttering of a linguistic expression whose function is to say things and to perform acts as well
96 Table 5.11 Classification of [I Verb that P] constructions
5 The Lexical and Syntactic Properties of MM
Category (Type F.)
Process type 2 Process type 1 Verb class (Type Frequency)
Epistemic modality (74)
Cognitive
Mental
Consider (16)
Cognitive
Mental
Conjecture/ Believe (22)
Cognitive
Mental
Argue (reasoning) (7)
Cognitive
Mental
Discover (learn) (10)
Cognitive
Mental
Comprehend (2)
Perceive
Mental
See (9)
Evaluative
Mental
Estimate (4)
Emotive
Mental
Emotional state (16)
Emotive
Mental
Emotional state _care (4)
Affective (22)
Mental
Accept (2)
Boulomaic Desiderative modality (11)
Emotive
Mental
Wish/Desire/ allow (11)
Speech act (56)
Indicating
Verbal
Say (11)
Quoting
Verbal
Manner_ speaking (7)
Indicating
Verbal
Declare (18)
Indicating
Verbal
Complain (3)
Indicating
Verbal
Advise (5)
Indicating
Verbal
Confess (4)
Indicating
Verbal
Promise (3)
Imperating
Verbal
Order (5)
therefore falling into the “say” subtype, it has another sense: observe, perceive (http:/ /verbs.colorado.edu/html_groupings/note-v.html). So it is also grouped into the “see (perceive)” subtype. As seen from Table 6.3, [I Verb that P] construction is highly productive, as a large variety of verbs being compatible with it. As noted by Bybee and Thompson (1997: 7) and Poplack (2001: 422), type frequency determines degree of productivity, since the more lexical items co-occur with a pattern or construction, the less likely it will be associated with any one item in particular. The largest, most productive category, that is, the one with the highest type frequency, is epistemic category, with 74 verb members. It has eight subtypes: Consider, Conjecture, See, Argue, Discover, Deduce, Estimate, and Comprehend.
5.4 [I VMental that P] Construction
97
This category may probably also have the highest token frequency, since, in addition to having more members than the others, certain members in this category have token frequency enormously higher than those in the other categories in COCA: the token frequency of [I think that P] is 31264; that of [I know that P] is 7506; and that of [I believe that P] is 5934. The subtypes of this category are exemplified as follows: Conjecture/Believe (22) I believe that Jon Stewart is a loose cannon. I believe that he’s a liberal. I believe that he has it in for conservatives. (120115, Spoken)
(46)
Consider (16) (47)
I think that anxiety and fear of some kind are part and parcel of being human. (2012, Magazine)
Discover (10) (48)
I discover that knee socks and a miniskirt are welcome in Alphabet City but suspect on the Upper East Side.
See (9) (49)
I see that we’re not necessarily male and female. (120227, Spoken)
Argue (7) (50)
I argue that it’s just as important, maybe even more important, to understand what it is to be human. (2012, Magazine)
Deduce (4) (51)
From this, I deduced that there had been a problem, but I let that pass as well and went upstairs to the apartment. (2007, Fiction)
Estimate (4) (52)
I estimate that my activities on social-media sites contribute less than 5 percent to my annual writing income. (2011, Academic)
Comprehend (2) (53)
So I understand that my competitors did what they thought they had to do. (120112, Spoken)
The second largest, productive category is verbal category, with 56 Speech Act verbs. With Austin’s (1962) seminal work on speech act theory and Searle’s (1969) development and elaboration of it, speech act verbs have been widely recognized as forming an established group of verbs which express illocutionary force in their first person, present tense forms. We sort out eight subtypes: Declare, Say, Manner_Speaking, Advise, Order, Complain, Admit, and Promise. The subtypes of this category are exemplified as follows: Declare (18) (54) Say (11)
I declare that I have no conflicts of interest. (2011, Academic)
98
(55)
5 The Lexical and Syntactic Properties of MM
I say that I appreciate the sentiment, but there’s a greater issue here. (091113, Spoken)
Manner_Speaking (7) (56)
I shouted that the sea was rough and would put sudden strains on the wire. (1991, Fiction)
Advise (5) (57)
Unofficially, I advise that we erase the records and forget the whole thing. (1991, Magazine)
Order (5) (58)
I ordered that no effort or resource be spared in prosecuting this war. (021211, Spoken)
Admit (4) (59)
I never say that it was terrific and I admit that we crossed lines. (110415, Spoken)
Promise (3) (60)
I promise that we won’t do it. (111114, Spoken)
Complain (3) (61)
I complained that he was canceling at the last minute, and he said,’ I never hear you complain about my ripped chest. How do you think it got that way?’ (2011, Magazine)
In general, verbs that can be put into the slot in [I Verb that P] fall into two process types: mental and verbal. This means two parallel categories of constructions can be abstracted from these construction instances: they are [I Vmental that P] construction and [I Vverbal that P] construction. The cognitive structure of the latter can also be explained under the same model of refined basic communication Space network that is proposed in Chap. 4, as demonstrated in Fig. 5.3. As Fig. 5.3 shows, within GROUND, there is cross-space mapping between VIEWPOINT space and SPEECH ACT space: the speaker S” in SPEECH ACT space is also subject of conceptualization in VIEWPOINT space, and both spaces attend to the same CONTENT. The two spaces are blended and S mapped onto the CONTENT Space. Both SPEECH ACT space and VIEWPOINT space connect to the blended space, which serves as CONTENT BASE/VIEWPOINT space in Onstage Region. Meanwhile, the speaker, simultaneously as the subject of conceptualization, is put onstage, formalized by the first-person pronoun. Within CONTENT, the emergent blended space is also distinct from Object of Conceptualization, as linguistically marked by the fact that they are realized by two clauses: the projecting clause and the projected clause. In a sum, the cognitive structure of [I Vverbal that P] construction appears analogous to [I Vmental that P], with the only difference that in the former SPEECH ACT space is blended with VIEWPOINT space and mapped onto the CONTENT Space while in the latter only VIEWPOINT space is mapped onto the CONTENT, with SPEECH ACT space held offstage within ground. Their cognitive structure is also directly mapped
5.4 [I VMental that P] Construction
99
Object of Conceptualization p
CONTENT BASE /VIEWPOINT S’s Viewpoint CONTENT
VIEWPOINT
SPEECH ACT
Subject of Conceptualization
A’
S’
S”
A”
S A GROUND BASE GROUND Fig. 5.3 Mental space construal configuration for [I Vverbal that P]
onto their syntactic structure: they have the same schematic structure [I Verb that P], but different categories of lexical verbs to fill in the slots. These linguistic facts can be taken into account of the key psychological principle of language organization in cognitive linguistics: If construction A is related to construction B syntactically, then construction A is motivated to the degree that is related to construction B semantically (cf. Goldberg, 1992; Haiman, 1985; Lakoff, 1987)
Considering speech act modality is proposed by Sweetser (1990) as part of her tripartite modal structure, [I Vverbal that P] construction could be considered an extended type of Metaphor of Modality. We can also find that when cognitive, perceptive, and evaluative verbs are put into the slot of [I Verb that P], the construction expresses epistemic modality. This epistemic modality construction is a good illustration of the interaction between lexemes and structural frames in which they appear, especially as in I see that we’re not necessarily male and female. The basic sense of see is assumed to be ‘perceive/ view by sight’. This means see is not prototypically united with epistemic modality,
100
5 The Lexical and Syntactic Properties of MM
which is concerned with the interpretation of the world via the laws of human reason. The integration of see with the epistemic modality construction is possible through the categorizing relationship of extension. Through repeated use, the pattern [I see that …] (it has a token frequency of 940 in COCA) can become conventionalized so that see can now be said to have new epistemic sense. Among the verbs that can be put into the slot in [I Verb that P] construction, there are 83 verbs (some verbs have multiple senses and therefore appear in different subtypes) belonging to verbs of mental processes. In other words, these 83 verbs can be put into slot in [I Vmental that P] construction. It can be said that [I Vmental that P] is a construction with considerable productivity and its productivity is constrained by the semantic properties of the schematic slot: only mental verbs are sanctioned in this construction. In [I Vmental that P] construction, epistemic category is the most productive. Affective category is only fairly productive, with 20 verb members, some exemplars of which are illustrated with context below: (62)
I admire that some women really like to “pink” things up, but I’d rather have a bike that performs ably than a “feminine” color. (2012, Magazine)
(63)
I care that I’m fat. (111227, Spoken)
Then it is followed by the category of boulomaic modality, which is lower in productivity, with lower type frequency (10) and stronger semantic coherence, as all the verbs have the sense of wish/desire and cannot be divided into subtypes. A prototypical exemplar is “wish”, as illustrated in the following utterance: (64)
I wish that I had been more sensitive to my parents, but I certainly don’t regret coming out. (120327, Spoken)
In general, Mental Verbs and verbal verbs are the major types of verbs in the slot of this construction. Mental Verbs mainly cluster into three subtypes: epistemic, affective, and boulomaic, while speech act verbs cluster into three subgroups.
5.5 [I be ADJ that P] Construction To some extent, [I be ADJ that P] construction could be considered as a variant to [I Vmental that P] construction. [Vmental ] and [be ADJ] can be thought of as variants of the predicate to the subject I in the FRAMING clause. Accordingly, both share similar syntactic properties. The research procedures are almost the same as what are carried out in the study of [It be ADJ that P] construction. With I in the concordance as the starting point, followed by the copular [be] and adjective represented by [j*], examples of the predicative adjectives of that-clauses were collected, with a minimum frequency of
5.5 [I be ADJ that P] Construction
101
5, from COCA. The researcher makes a judgment on the accuracy of the identification of the grammatical frame—[I be ADJ that-clause]. The full classification of the predicative adjectives used [I be ADJ that-clause] pattern in the COCA corpus is given in Appendix 4. Semantic Classification of ADJs in Sect. 5.4, of course, has offered very helpful reference. Still, I have also used the semantic tool—the online FrameNet lexical database if the semantic tagging of an adjective is not that straightforward. The types of meanings represented by the ADJs in the corpus examples could be roughly divided into four chief semantic categories: epistemic, affective, evaluative, and boulomaic. Semantically, [I be ADJ that-clause] construction can be considered as a variant of [I VMental that P] construction. In [I be ADJ that-clause] construction, the category of affective modality is clearly the most productive, with a type frequency of 41, which is almost 70% of the numbers of adjectives (minimum frequency is 5) that can be put into the ADJ slot. The subtypes of this category include un/happiness, dis/satisfaction, in/security, and anger. The subcategory of anger is not found in either [It be ADJ that P] construction or [I Vmental that P] construction, which is rather unique to this construction, which is illustrated below: (65)
I’m angry that he took advantage of me the way he did. (110527, News)
(66)
I ’m outraged that they would think I’m racist. (080318, Spoken)
(67)
I am mad that every time there’s a high-profile rape case in this country, we beat the hell out of the victim. (060501, Spoken)
(68)
I am furious that this is being used by groups that never have supported women’s rights, that never have intervened on behalf of victims of sexual harassment. (19940825, Spoken)
It is followed by category of epistemic modality, with lower type frequency (15). Nevertheless, this category contains constructions with much higher token frequency; for example, [I be sure that] has token frequency of 1932, which is more than 2 times of the second one. Constructions representing this category are illustrated in the following utterances: (69)
I ‘m sure that they will come to an easy answer of what is going on. (120,314, Spoken)
(70)
I am convinced that White House pressure for good news was the main reason for the Challenger disaster. (2011, Magazine)
(71)
I am aware that it will not clear my name of scandal; my intention is only to show that I acted in the best interests of Britain at all times. (2012, Fiction)
(72)
I ‘m hopeful that once we fix the remaining structural issues, we can earn the confidence of the voters to then fix the initiative process. (110,128, Newspaper)
Finally, there are only two items involving some evaluative sense and one item involving boulomaic sense. As a result, these two categories are far from productive, and thus peripheral to [I be ADJ that P] construction.
102
5 The Lexical and Syntactic Properties of MM
Globally, there is not much difference between [I be ADJ that P] construction and [I VMental that P] construction. The major differences lie in that the former can express extensive affective modalities in a highly productive way while the latter more aptly represents many kinds of epistemic modality.
5.6 Summary The observations presented in this chapter can be summarized as follows. Approached bottom-up, the major constructions of Metaphor of Modality are found to have idiosyncratic syntactic and lexical properties. (1) The syntagmatic relationship between the projecting modal clause and [that P] is that of FRAMING; (2) There are preferences (constraints) on what types of adjectives can be put into the slot of [It be ADJ that-clause] construction. Based on the token frequencies of its construction types, we find that the chief semantic category is epistemic modality. Then nearly onefifth of [It be ADJ that-clause] construction in COCA represents evaluative modality and more that 5% represent affect. The findings indicate that one of the motivations for MM is to expand the semantic domains of modality. (3) Likewise, among [I Vmental that P] construction, the most productive category is epistemic modality. Although only fairly productive, the representation of affect also demonstrates the expansion of semantic domains. (4) In addition to [I Vmental that P] construction, a parallel category of constructions can be abstracted: it is [I Vverbal that P] construction. [I Vverbal that P] construction is considered to be an extended type of Metaphor of Modality because its cognitive structure and syntactic behaviors are topological to metaphorical modes of modality, especially [I Vmental that P] construction. Analyzing the lexical and syntactic properties of MM not only reveals the idiosyncratic characteristics of its forms but also shows how the semantic domains of modality are expanded via the interaction of lexical resources and syntactic frames.
References Austin, J. L. (1962). How to do things with words. Oxford University Press. Baker, C. F., et al. (1998). FrameNet. http://framenet.icsi.berkeley.edu/fra-menet Barlow, M. (2013). Individual differences and usage-based grammar. International Journal of Corpus Linguistics, 4, 443–478. Biber, D. (1988). Variation across speech and writing. Cambridge University Press. Benveniste, E. (1966/1971). Subjectivity in language. In Benveniste, E. (ed.), Problems in General Linguistics. Coral Gables, FL: University of Miami Press, 223–230. Bybee, J. L., & Eddington, D. (2006). A usage-based approach to Spanish verbs of ‘becoming’. Language, 323–355. Bybee, J. & Thompson, S. (1997). Three frequency effects in syntax. In Proceedings of the TwentyThird Annual Meeting of the Berkeley Linguistics Society (pp. 378–388). Bybee, J. L., & Scheibman, J. (1999). The effect of usage on degrees of constituency: The reduction of don’t in English. Linguistics, 4, 575–596.
References
103
Bybee, J. L. (1995). The development of past tense modals in English. In L. B. Joan & S. Fleischman (Eds.), Modality in grammar and discourse (pp. 503–517). John Benjamins. Croft, W. (2003). Lexical rules versus constructions: a false dichotomy. In H. Cuyckens et al. (Eds.), Motivation in language: Studies in honor of Gunter Radden (pp. 49–68). Benjamins. Diessel, H., & Tomasello, M. (2001). The acquisition of finite complement clauses in English: A corpus-based analysis. Cognitive Linguistics, 2, 97–141. Emonds, J. (1976). In A transformational approach to english syntax. Root. Evans, V., & Green, M. (2006). Cognitive linguistics: An introduction. Edinburgh University Press. Fetzer, A. (2014). I think, I mean and I believe in political discourse: Collocates, functions and distribution. Functions of Language, 1, 67–94. https://doi.org/10.1075/fol.21.1.05fet Fillmore, C. J. (1992). Corpus Linguistics versus computer-aided armchair linguistics. In J. Svartvik (Ed.), Directions in corpus linguistics (pp. 35–60). Mouton de Gruyter. Fillmore, C. J., Kay, P., & O’Connor, M. C. (1988). Regularity and idiomaticity in grammatical constructions: The case of let alone. Language, 64, 501–538. Goldberg, A. E. (1992). Argument structure constructions. University of California, Berkeley. Goldberg, A. E. (1995). Constructions: A construction grammar approach to argument structure. The University of Chicago Press. Goldberg, A. E. (2006). Constructions at work: The nature of generalization in language. Oxford University Press. Goldberg, A. E. (1997). The relationships between verbs and constructions. In Lexical and syntactical constructions and the construction of meaning (pp. 383–398). Haiman, J. (1985). Natural syntax: Iconicity and erosion. Cambridge University Press. Halliday, M. A. K. (1967). Intonation and grammar in British English. Mouton. Halliday, M. A. K., & Hasan, R. (1976). Cohesion in English. Longman. Halliday, M. A. K. (1985/1994). An Introduction of Functional Grammar. London: Edward Arnold. Halliday, M. A. K., & Matthiessen, C. (2004). An introduction to functional grammar. Hodder Arnold. Halliday, M. A. K., & Matthiessen, C. (2014). Halliday’s introduction to functional grammar. Routledge. Halliday, M. A. K. (1970). Functional diversity in language as seen from a consideration of modality and mood in English. In Foundations of language (pp. 322–361). Herriman, J. (2000). Extraposition in English: A sudy of the interaction between the matrix predicate and the type of extraposed clause. English Studies, 6, 582–599. Hooper, J. B. (1975). On assertive predicates. In J. P. Kimball (Ed.), Syntax and semantics (Vol. 4, pp. 91–124). Academic Press. Hopper, P., & Thompson, S. A. (1980). Transitivity in grammar and dis-course. Language, 56, 251–299. Jurafsky, D., et al. (2001). Probabilistic relations between words: Evidence from reduction in lexical production. In J. Bybee & P. Hopper (Eds.), Frequency and the emergence of linguistic structure (pp. 229–254). John Benjamins Publishing Company. Kaltenbock, G. (2004). In It-extrapositon and non-extraposition in English: A study of syntax in spoken and written texts. Braumuller. Kaltenbock, G. (2005). It-exposition in English: A functional view. International Journal of Corpus Linguistics, 2, 119–159. Kipper-Schuler, K. (2006). VerbNet. http://verbs.colorado.edu/verb-index/index.php Lakoff, G. (1987). Women, fire, and dangerous things: What categories reveal about the mind. University of Chicago Press. Langacker, R. W. (2008). Cognitive grammar: A basic introduction. Oxford University Press. Leech, G. N. (1983). Principles of pragmatics (Vol. 285). Longman. Levin, B. (1993). English verb classes and alternations: A preliminary investigation. University of Chicago Press. Martin, J. R., & White, P. R. R. (2005). The language of evaluation: Appraisal in english. Palgrave Macmillan.
104
5 The Lexical and Syntactic Properties of MM
Martin, J. R. (2010). Sense and sensibility: Texturing evaluation. In W. Zhenhua (Ed.), Discourse semantics (语篇语义研究) (pp. 341–375). Shanghai Jiao Tong University Press. Matthiessen, C. M. (2014). Extending the description of process type within the system of transitivity in delicacy based on Levinian verb classes. Functions of Language, 2, 139–175. https://doi.org/ 10.1075/fol.21.2.01mat McGregor, W. B. (1997). Semiotic grammar. Clarendon Press. Miller, P. H. (2001). Discourse constraints on (non)extraposition from subject in English. Linguistics, 39(4), 683–701. Mukherjee, J. (2004). Corpus data in a usage-based cognitive grammar. Language and Computers, 1, 85–100. Neale, A. (2006). Matching corpus data and system networks: Using corpora to modify and extend the system networks for transitivity in English. In G. Thompson & S. Hunston (Eds.), System and corpus: Exploring connections (pp. 143–163). Noonan, M. (1985). Complementation. In T. Shopen (Ed.), Language typology and syntactic description: Complex constructions (2nd ed., Vol. 2, pp. 52–150). Cambridge University Press. Nuyts, J. (2005). The modal confusion: On terminology and the concepts behind it. In Modality: Studies in form and function (pp. 5–38). Poplack, S. (2001). Variability, frequency, and productivity in the irrealis domain of French. Typological Studies in Language, 45, 405–430. Quirk, R., et al. (1985). A comprehensive grammar of the english language. Longman. Rosenbaum, P. S. (1967). Phrase structure principles of English complex sentence formation. Journal of Linguistics, 1, 103–118. Searle, J. R. (1969). In Speech acts: An essay in the philosophy of language (Vol. 626). Cambridge University Press. Seidenberg, M. S., & MacDonald, M. C. (1999). A probabilistic constraints approach to language acquisition and processing. Cognitive Science, 23, 569–588. Stefanowitch, A., & Gries, S. T. (2003). Collostructions: Investigating the interaction of words and constructions. International Journal of Corpus Linguitstics, 2, 209–243. Sweetser, E. (1990). From etymology to pragmatics: Metaphorical and cultural aspects of semantic structure. Cambridge University Press. Thompson, S. A. & A. Mulac. (1991). A quantitive perspective on the grammaticalization of epistemic parentheticals in english. In E.C. Traugott & B. Heine (Eds.), Approaches to grammaticalization (Vol. 2. pp. 313–329). John Benjamins.
Chapter 6
The Pragmatics and Discourse Functions of MM
Targeting the linguistic phenomenon of MM, the core question is how and why metaphorical and congruent modes of modality differ from each other. In Chap. 4, we have explored what cognitive structure and process motivate the systemic contrast between them. In Chap. 5, we have investigated in detail how MM is realized in terms of lexicogrammar, that is, what idiosyncratic lexical and syntactic properties MM has. In the present chapter, combining the SFL tradition with cognitive framework of construction, we tackle the question of motivation in terms of pragmatics and discourse functions: what are the functional and pragmatic motivations of metaphorical modal constructions?
6.1 Discourse as the Research Instrument SFL has a long standing interest in discourse analysis, deriving historically in part from Firth’s (1957) concern with meaning as function in context (Martin, 2009: 154). The term ‘discourse’ (or ‘text’ often interchangeably) refers to any instance of language in any medium that makes sense to someone who knows the language (Halliday & Matthiessen, 2004: 3). In this study, we focus on discourse not as an object in its own right, but as an instrument for finding what the discourse (text) reveals about the system of the language, especially the system of modality, which is where my study is positioned. In usage-based model, which is applied in last chapter, the frequency of occurrence of particular grammatical forms and structures is assumed to affect grammatical representation. Likewise, many discourse grammarians feel that text frequency is vital to understanding the discourse motivations for particular grammatical constructions.
© Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7_6
105
106
6 The Pragmatics and Discourse Functions of MM
The importance of using natural discourse data in linguistics is well established in grammars that acknowledge the discourse basis of linguistic phenomena. The distinctive methodological characteristics associated with studies in the discoursefunctional mold are an insistence on carefully recorded natural data and an interest in quantitative and distributional information about grammatical patterns. As a corpus represents a systematic sample of text collected according to clearly stated criteria, corpus-linguistic method has facilitated our analysis. The corpuslinguistic tools, of course, have been used in last chapter, where sentence level instances plus frequency data are gathered to analyze the lexical and syntactic properties of Metaphors of Modality. In the current chapter, text excerpts from the corpus are used to conduct functional analysis and interpretations. Functions can only be examined by looking at an extended piece of discourse. This can be illustrated by the phrase I must admit in a corpus of spoken English by Hunston (2002: 59). Several concordance lines of I must admit show that the speaker uses I must admit when saying something that is not uncomplimentary. In the following example, the concordance line itself does indicate that what is said is not uncomplimentary. However, looking at a larger co-text shows that the speaker is saying something negative about a book which overall he has enjoyed. (73) … toward the end of the book, I must admit it goes rather fanciful and he traveled abroad and slew dragons and things like that … But er up until that time it was rather a rather interesting story.
There are a number of explanatory themes, or functional dimensions, which have been especially important in discourse-functional work. One large body of research in studies of discourse and grammar is related to the theme of ‘information flow’, which has to do with the way information is distributed within and across clauses. Halliday (1967) produced characterization of given/new in terms of speaker-expectations. Given information is specified as being treated by the speaker as recoverable either anaphorically or situationally. Another prominent theme is the range of phenomena associated in the literature with ‘interpersonal metafunction’, ‘perspective’, and ‘subjectivity/intersubjectivity’, which has been held to influence a wide range of aspects of linguistic form. A somewhat related theme is lexicalized clauses as epistemic or evaluative discourse markers. Recently, many discourse grammarians have focused on the nature of biclausal (multi-clausal) constructions. The Metaphor of Modality phenomenon will be shown to be related to these themes, which have been illuminated by a discourse-functional approach.
6.2 [I VMental that P] Constructions
107
6.2 [I VMental that P] Constructions 6.2.1 Functions of the Construction in Discourse In discourse-functional studies, an important body of work has centered on subjectverb combination such as the English I think and I guess. It has been argued that, unlike what was traditionally assumed to function as the main clause found with their subordinate clause in examples like I think that anxiety and fear of some kind are part and parcel of being human, the main function of these expressions is to index subjectivity and stance (Karkkainen, 2003) and explicit subjective modality as a metaphorical mode of modality (Halliday, 1985, 1994; Halliday and Mattiessen 2004, 2014). In Chap. 5, we have analyzed syntagmatic structure of this construction as [I Vmental ] framing the complement that-clause, and I think and I guess function as the framing clause. According to Karkkainen (2003), I with mental verbs such as think, believe, and guess typically express the speaker’s stance regarding a subsequent piece of discourse P, in other words, they are index of subjectivity. The concepts of ‘subjectivity’ and ‘subjectification’ have become increasingly popular in the cognitive-functional literature for nearly thirty years. Broadly speaking, subjectivity involves expression of self and the representation of a speaker’s perspective or point of view in discourse. Subjectivity is the relevance of the speaker to meaning, and intersubjectivity, then, is the relevance of the speaker-addressee relationship to meaning (Porter, 2009: 131). In a general sense, language can be said to be strongly marked by subjectivity in that any selection from the lexical and/or grammatical repertoire passes through the speaker/writer. In contrast to this general subjectivity, De Smet and Verstraete (2006) distinguish two types of what they call ‘semantic subjectivity’: Following a distinction made by Halliday and Hasan (1976) and Halliday (1985, 1994), one type of subjectivity functions in the ‘ideational’ component of the utterance, and another type functions in the ‘interpersonal’ component. In SFL, ideational metafunction deals with the representation of content, in other words, experience with the world (Halliday & Mattiessen, 1999: 7), while the interpersonal metafunction deals with the positioning of the speaker with respect to this representation and his or her interaction with the interlocutor. This distinction, as remarked by De Smet and Verstraete (2006: 385), roughly corresponds to the distinction between content and grounding in cognitive grammar (Langacker, 1990). De Smet and Verstraete (2006: 385) think it is necessary to propose a further distinction between ideational and interpersonal types of subjectivity. Ideational subjectivity involves the description of a content “situated in the speaker’s subjective belief-state/attitude toward the situation”. Conversely, the term interpersonal provides a useful generalization over the non-ideational types of subjectivity: it refers both to the speaker’s attitudes and judgments and to his or her interaction with the interlocutor. This dual perspective seems correspondent to the distinction between subjectivity and intersubjectivity recently developed.
108
6 The Pragmatics and Discourse Functions of MM
Although the two types are jointly called ‘semantic subjectivity’, the ideational and interpersonal expressions of subjectivity show divergent syntactic behavior. They think a meaning situated in the speaker’s beliefs and attitudes may figure either in the way the speaker takes an interactive position with respect to that content or in the content of the utterance, i.e., what is described. Among various [I Vmental that P] constructions, there is one item having very high token frequency. The token frequency of [I think that-clause] is 31,263 in COCA. It may have formed a more autonomous chunk. I think is the single most frequent I + verb pattern in the spoken components of COCA and the British National Corpus (BNC) and [I think that-clause] is the most frequent among [I Vmental that P] construction in COCA. I think has been examined as a parenthetical expression and as a pragmatic marker, and its distribution and function, related complementizer dropping, and patterned co-occurrences with argumentative discourse connectives and pragmatic markers are considered (cf. Aijmer, 1997; Fetzer & Johansson, 2010; Karkkainen, 2003; Thompson & Mulac, 1991). My analysis starts with a transcribed discourse excerpt of NPR’s TALK OF THE NATION radio program. The background to this issue is that in an op-ed in Slate magazine, writer Jessica Olien says she wants to raise a child alone, without a husband to complicate the process. The following exchange is from Jessica Olien and John Donvan, the host. (74)
DONVAN: Is there something lost to not having in a single-mother household a father figure and in a single-father household a mother figure? Do you want to have the some—do you want to have a father figure in the son—in the life of a son or daughter who is raised with a single mom in some way? OLIEN: I mean, I kind of—to me that kind of depends on if you, like, look at, you know, your parent—like my mother, I feel like she fulfilled both roles pretty well. I don’t think—I think it depends on how you view gender roles and how important those are to you in raising a child. I don’t think that you necessarily need a father figure. I think that a child should be exposed to both men and women, and I don’t think that that is necessary, though, to have a father and a mother in a home raising a child.
The utterance I think that a child should be exposed to both men and women especially catches my attention. Modal verbs or adjuncts like should, will and may, as claimed by Halliday (1985, 1994), are congruent representation of subjective modality, whereas ‘I think’, being a projecting mental clause, is a metaphorical mode of modality. Then a question arises: Why does the speaker choose to express modality congruently and metaphorically, or first explicitly subjective and then implicitly subjective simultaneously within one usage event? The modal verb ‘should’ expresses the speaker’s reasoning, implicitly in SFL terms, that it is necessary for children to be exposed to both men and women. Then why “I think” is used beforehand to indicate explicitly this is what I subectively think? This suggests that there is likely other motivations to use [I think]. As analyzed in Chap. 4, it is
6.2 [I VMental that P] Constructions
109
mapped onto the domain of CONTENT. In other words, it represents the VIEWPOINT space mapped onto CONTENTS and has a blended ideational plus intersubjecitve meaning. This is not a single instance. For instance, in Olien’s turn, all of her utterances begin with constructions such as I mean, I think that, or I don’t think that, which are quite prominent. (75)
OLIEN:: I was just—it’s a—it’s just an essay. It was about my own choices and what I think would work well for my situation. But I’m certainly not saying that it’s—all women should become single parents now, and that we should, you know, men are expandable, and we should no longer have nuclear families. I just think that this has worked really well for me. And I feel very satisfied and fulfilled as an adult because of how I was raised.
This analysis shows that, whereas modal verbs like should expresses deontic modality, metaphorical modal expressions such as I mean, I think that, or I don’t think that represent interpersonal subjectivity with input ideational meaning. Through the integration of ideational resources, Metaphor of Modality has the potential to represent meanings richer and finer-grained than modal adjuncts. In addition, an intrinsic disparity in salience seems evident between the members of various oppositions like explicit versus implicit. In this respect, metaphorical modal expressions are more salient or prominent than modal verbs. In other words, MM (of the subjective type) render subjectivity prominent. (76)
And I think that that choice should at least be able to be discussed and open to them. (120,227, Spoken)
The pattern [I VMental that P] or a reduced variant of it, according to Nuyts (2001: 118), exhaust the syntactic possibilities of and meanwhile typify the qualificational reading of all mental state predicates. The qualificational reading is impossible in any of the other syntactic patterns in which the non-qualificational reading occurs. The construction typically occurs in contexts in which the speaker voices personal views belonging in the realm of strictly individual experiences or attitudes or involving an opposition between the speaker and the hearer, for example: (77)
I suspect that there’s nothing unconstitutional about it (Medicaid provision).
(78)
NEAL-CONAN: It’s hard to argue that Hitler made the country what it is today. The country… DOUGLAS: Well… NEAL-CONAN:… when he left it, was in ashes. DOUGLAS: I mean that he built it up. After World War II, there was a lot of problems from the reign of Wilhelm.
110
6 The Pragmatics and Discourse Functions of MM
(79) COLMES: Well, again, and I think the military also—if someone in the military wants to comment and thinks they were offended by Charlie Rangel, that’s something for them to say. But, by the way, in terms of actually hanging, would you consider hanging barbaric? HOLMES: I consider that it is an appropriate punishment for a mass murderer. COLMES: We don’t hang in this country. We don’t that. HOLMES: Mass murderer. Saddam Hussein, we know that it was… COLMES: Hanging you don’t think is a barbaric act? HOLMES: We know in Iraq that that was in keeping with their traditions of carrying out the death penalty. I don’t think that we need to be shedding one single tear for Saddam Hussein. (CROSSTALK)
In utterances (77), (78), and (79), the speaker expresses, in turn, his/her view on US’s Medicaid provision, Hitler, and hanging. All these issues are very controversial in their sociocultural contexts. In the first case, a lot Americans are skeptical about Medicaid provision, but the speaker thinks differently; in the second, the other speaker holds the view that Hitler destroyed Germany, which is a common belief, but the speaker thinks “he built it up”; Finally, the speaker considers hanging “an appropriate punishment for a mass murderer”, although a lot people consider hanging barbaric, especially since death penalty has been abolished in many states of America. In this sense, [I Vmental that P] is considered as a discourse strategy, more specifically, a facesaving device, in Nuyts (2001)’s opinion, when the speaker reacts against something held by common people or something which has been said by the conversation partner. Note the following example: (80) IRA-FLATOW: Sure. But shouldn’t I be given that option to take that risk if I want to? LAURA-NEWMAN: You mean to take - have a surgery or… IRA-FLATOW: Have a surgery if I… LAURA-NEWMAN: Well, here’s… IRA-FLATOW:... I’m facing the choice—why shouldn’t I be given that option? LAURA-NEWMAN: Well, you know, we live in a culture that actually treasures choice, but I feel that we’re dealing with a health care system that is collapsing. And there are choices and there are choices, and why not put our limited health care dollars into health problems that could be managed well where people will actually feel better and live longer? IRA-FLATOW: Let’s go to the phones, to Bob in St. Louis. Hi, Bob. BOB: Hi. Yes, I’m a physician. I think maybe she’s painting the picture with too broad a brush. (120427, Spoken)
This excerpt is from Science Friday, a weekly discussion of the latest news in science, technology, health, and the environment hosted by Ira Flatow. Ira interviews
6.2 [I VMental that P] Constructions
111
scientists, authors, and policymakers, and listeners can call in and ask questions as well. This discussion is titled “How Buffett’s Cancer Is Shaping National Dialogue”. IRA argues that people should be “given that option to take that risk”. LAURANEWMAN, the interviewee, argues against it with “we’re dealing with a health care system that is collapsing”. She uses “but I feel that” as face-saving discourse strategy. Again, BOB, the listener calling in, uses “I think” to say something against LAURANEWMAN. In this way, [I Mental Verb that P] construction enables the speaker to voice personal views, which may be against common belief or the interlocutor’s viewpoints, with face-saving. It is a very powerful discourse and rhetorical strategy. The function of mental verbs with first-person subjects in discourse has been examined by Karkkainen (2003), who assigns them the status of a personalized stance marker. Hooper (1975) classifies them as weak assertives. Nuyts (2001: 165) also points out that the epistemic meaning of mental state predicate suggests that the speaker is saying tentative things about which he may be wrong in fact. Thus, the mental state predicate weakens or mitigates the force of the claim. This, however, is only one side of the story. Previous functional analysis shows that first-person-singular Mental Verb-based clauses are structurally independent from the clause over which they have scope. They add subjectively qualified information, which, according to Fetzer (2014: 72), may upgrade or downgrade the value of epistemicity expressed by the speaker. Observe the following utterances from COCA: (81)
I believe that diplomacy should be the cornerstone of any foreign policy. (100,915, Spoken)
(82)
TONY BLAIR: Well, let us wait and see exactly what happens. And as I say, I don’t think it would be very sensible to start canvassing too openly what the military options are. But you can rest assured that we are looking at every single part of this to make sure that the action that we take is effective. And I believe that it can be effective. (19,980,205, Spoken)
(83)
I suspect that the Kosovars could do it with many fewer casualties than NATO can do it with ground troops. (19,990,419, Spoken)
(84) I think that what it means is that Dr. Lee has testified in a very cautious manner. He has qualified every answer to a hypothetical question put to him by the defense. I suspect that the prosecution will make some great points on their cross examination of Dr. Lee. And I don’t know, even though it has been billed as Dr. Lee making great strides for the defense here, I believe that once the prosecution is completed, Lee will- Dr. Lee will probably say it is consistent with another fabric.. (19,950,823, Spoken)
It can be seen that I believe generally performs the boosting function, which is signified by the salient discourse pattern with expressions of modality coding necessity such as the modal should, can and will; I suspect may boost or attenuate the pragmatic force of an utterance, expressing possibility and probability on the one hand, and certainty on the other, depending on its local context. To conclude, the mental state predicate may weaken or boost the force of the claim, depending on its
112
6 The Pragmatics and Discourse Functions of MM
discourse context and which particular Mental Verb used. Through the integration of ideational resources from mental verbs plus the syntactic pattern and its interpersonal meaning, Metaphors of Modality have the potential to represent meanings richer and finer-grained than those of modal adjuncts.
6.2.2 The Distributional Features of the Construction To get a picture of [I VMental that P] constructions as a whole, I have investigated the distribution of these constructions in at the macro-level of discourse types, viz. genres, which is shown in Appendix 2 and summarized in Table 6.1. It is clear that this construction has much higher frequency (449.3 per million) in spoken genre than in other genres, which ranges from 42 to 63 per million. In spoken communication, which is characterized by spontaneity, the speaker naturally conceptualizes the scene from his/her own viewpoint or deictic center. Furthermore, [I VMental that P] construction is used to invite or instruct the hearer to see the scene from the speaker’s viewpoint for smooth interaction or accomplishment of communication goals. In written discourse, from fiction to academic genre, it is expected and even ideal that the writer should not conceptualize the reality, which is actually not the real reality but projected reality, solely from his/her personal viewpoint. When a writer attempts to represent multiple viewpoints, different means are required. Viewpoint shifts through the choice of deictics, besides “I”, is such a case. This may partly explain the relatively lower frequency of [I VMental that P] construction in written genres. Observe an exemplar [I expect that P] from COCA, which only has 21 occurrences in academic section but much higher frequency in spoken section: 129, nearly five times of its average frequency in other genres. Some examples are shown below:
Table 6.1 Distribution of [I VMental that P] construction among different genres Section
ALL
SPOKEN
FICTION
MAGAZINE
NEWSPAPER
ACADEMIC
FREQ
62,282
42,937
5694
5041
4774
3836
PER MIL
134.13
449.30
62.97
52.75
52.05
42.12
6.2 [I VMental that P] Constructions
(85)
As a historian I expect that most people will not find my research very exciting (2009, Academic)
(86)
FINALLY, I expect that some readers of this essay will say that, though I have done a service to the students who might not ordinarily want to read and discuss literature, I have shortchanged the others, the ones like myself, who deserved to get more out of a course than the sort of carefully pruned readings and measured assignments that I offered. (2011, Academic)
(87)
Therefore, I expect that there will be a correlation between testing methodology and student-evaluation results. (1993, Academic)
(88)
I expect that the international community, the pressure will mount. (120,528, CBS_NewsEve)
(89)
SCOTT-PELLEY: When you first approach a hospital or a doctor to ask them for potentially, hundreds of thousands of dollars in free medical care, what’s your pitch? What do you tell them? ELISSA-MONTANTI: I tell them this true story. Here’s a child that’s—that’s battered. I just tell them the reality. I expect them to help. I’m grateful because they don’t have to help. But I expect that they would, because how could you not? (110327, CBS 60 MINUTES 7:17 PM EST)
(90)
Mr. BRINKLEY: Here in the studio with us are George Will and Sam Donaldson, both of ABC News. Now, we’ve seen all kinds of events this morning which we have covered for the American people and you’ve seen them along with us. What do you expect to see there next? Mr. LEKOTA: I expect that in the near future, the movement will make contact with the government and I expect that Comrade Mandela will be at the head of the negotiating team of the liberation movement. (19,900,211, ABC_Brinkley)
113
As shown in Example (85–87), the writer is very cautious in expressing personalized stance. In (85), the writer makes it clear that “I” does not equate to the physical speaker/writer but “I” as “a historian”. In (86), there are viewpoint blends between the “I” perspective and readers. In addition, in (87), the writer’s personalized stance is based on sound reasons given in the previous context. A writer normally thinks more carefully about opinions voiced. In spoken section of COCA, in contrast, more often than not, interlocutors tend to seek personal responses or comments from each other. This may also motivate the use of construction to lead a personalized response, as shown in (89) and (90). Moreover, the somewhat ungrammaticality in (88) shows that the speaker is pressed to formulate immediate, unprepared reactions, which is more likely to come from the speaker’s viewpoint or deictic center than otherwise. In general, discourse strategy seems a quite frequent factor in our corpus occurrences of [I VMental that P] construction. This may thus be another element motivating their remarkably high frequency in spoken language as compared to other written genres. On the one hand, face work surely is more at stake in oral interaction. On the other hand, as the claim made in spoken communication is immediate and unprepared, and it is very hard for the speaker to control its argumentative force. Thus, the speaker needs to modify the force of the claim, and [I VMental that P] construction is a very convenient device to weaken or boost it. In written discourse, the reader, after
114
6 The Pragmatics and Discourse Functions of MM
all, is not present at the real time, real space with the writer. Therefore, the task of face-saving and management of argumentative force is not that pressing. To sum up, [I VMental that P] construction is used much more frequently in spoken communication than written genres, largely as a discourse strategy. It is used to invite or instruct the hearer to see the scene from the speaker’s viewpoint. Moreover, it may weaken or boost the force of the claim, depending on its discourse context and the Mental Verb used.
6.3 [I VVerbal that P] Constructions All languages seem to have means of marking stretches of language as the spoken or written utterances in a particular speech situation, ways of representing utterances produced in a certain speech situation as belonging to that situation—the represented speech situation (McGregor, 1997: 252). The term ‘represented speech’ designates the grammatical means available in a language for making such distinctions, that is, for framing an utterance. It effects interpersonal modification of the represented. [I VVerbal ] is a type of framing clause, as a way of indicating how the utterance P should be interpreted. In terms of this function, [I VVerbal that P] construction is no different from [I VMental that P] construction. The use of this construction is constrained by the pragmatic properties of different speech act verbs. It has often been pointed out that many speech act verbs can be used performatively, i.e., that they can be used in the first person, present tense, to indicate the illocutionary force of the utterance, for example: (91)
(BEGIN-VIDEO-CLIP) BARACK-OBAMA-PRES: I say that it’s a time when the tax burden on the wealthy is at its lowest level in half a century. The most fortunate among us can afford to pay a little more. UNIDENTIFIED-MALE: Once again President Obama makes his case for higher taxes to bring down the federal debt. But almost everyone admits that will not solve the problem. So why is the President going down this road? Dick Morris and Congressman Dennis Kucinich will analyze. (110415, Fox_Baier)
‘I say that’ in Obama’s utterance carries stronger illocutionary force than the otherwise I think that. It is easier to understand if we look at the context: “almost every admits that will not solve the problem”, indicating implicitly that Obama is facing strong opposition. He must use a language that carries weight and force. The use of “I say that” also adds a tone of assurance. Besides, this utterance also has its perlocutionary force: the direct consequence is the case of ‘the wealth’s paying more taxes. I have also investigated the distribution of [I VVerbal that P] construction in at the macro-level of discourse types, viz. genres, which is shown in Appendix 3 and summarized in Table 6.2. In these spoken, fiction, magazine, newspaper, and academic genres, [I VVerbal that P] construction adds an illocutionary ‘operating instruction’ as to how the framed
6.3 [I VVerbal that P] Constructions
115
Table 6.2 Distribution of [I VVerbal that P] construction among different genres Section
All
SPOKEN
FICTION
MAGAZINE
NEWSPAPER
ACADEMIC
FREQ
4250
1313
758
877
517
785
PER MIL
9.15
13.74
8.38
9.18
5.64
8.62
clause has to be taken, as with I suggest, I maintain or I show in academic discourse, which is illustrated below: (92)
I suggest that we continue to serve those already included in music education, but pause to consider those we may unintentionally exclude. (2011, Music Educators Journal, Academic)
(93)
Still, I maintain that hospitals do generate an exceptional social space. (2009, Anthropological Quarterly, Academic)
(94)
First, I show that the debate over the effectiveness of psy-war can not be fully resolved. (2010, War in History, Academic)
The illocutionary force of these utterances varies with the use of different speech act verbs. In spoken communication, which is characterized by spontaneity, the speaker naturally conceptualizes the scene from his/her own viewpoint or deictic center. Furthermore, a blending of VIEWPOINT space and SPEECH ACT space in involved in the cognitive structure of [I VVerbal that P] construction so that the hearer is also instructed to see the scene from the speaker’s viewpoint for smooth interaction or accomplishment of communication goals. What’s more, the speaker’s intention to exert a certain influence on the addressee through the uttering of a linguistic expression is more at stake in spoken communication. To achieve a real-time influence, a best strategy would to put the speech act onstage explicitly and render it prominent. All these constitute the main explanations why [I VVerbal that P] construction should occur more frequently in spoken genre. Another reason may be that in spoken interaction a speaker is often asked to make comments on the issue, for example:
116
6 The Pragmatics and Discourse Functions of MM
(95)
PINSKY: What do you say to that, Lisa? BLOOM: I say that no one is talking about the middle ground, which is requiring by law, parents to lock up guns and to keep the ammunition separate, requiring—(120315, CCN_DrDrew)
(96)
BAIER: But for the people who say—and they look at the exit polls tonight and other nights and say, the one person who wants you to stay in this race is really Mitt Romney, what do you say to them? GINGRICH: I say that we just got two out of every three delegates in Mississippi and Alabama for somebody other than Mitt Romney. I don’t think that’s what he wanted. (120,313, Fox_Baier)
The occurrence of [I VVerbal that P] construction is lower in written discourse types such as Newspaper and academic genres. Since lots of such usage is found in quotations within these written discourses, the frequency of this construction in the real written sense would be even lower. Besides, this construction tends to occur in certain subtypes of these genres. Example (97) below is taken from letters in a Newspaper and (98) is taken from reviews. These contexts are either interactive in nature or where writers are expected to make comments on a topic. (97)
As a former kindergarten teacher, I say that we need to let children be children. (100,516, LETTERS, Newspaper)
(98)
Once again, I do not think I exaggerate when I say that the shallowness of this culture has been repressingly brought on and perpetuated by those who have the power, whether it is political, commercial, ecclesiastical, or pedagogical. And once again, I say that the grief being done to our young by a culture of shallowness-a culture of pretended depth created by political, educational, and commercial pretenders-should cause everyone who has ever been near a child’s mind to bend over with sadness and anger. (2000, Arts Education Policy Review, Academic)
In general, the pragmatic properties of [I VVerbal that P] construction are not very distinct from those of [I Vmental that P] construction. The framing clause or the reporting clause in Vandelantte (2009)’s term, fulfills the prerequisites of performativity and functions interpersonally. The framing clause has grammaticalized and subjectified, serving no longer to construe representationally a mental or verbal process, but to provide a modal or illocutionary comment on the speaker’s own discourse produced in the speech act. The reporting clauses are highly formulaic, which means they are entrenched, becoming a cognitive routine. Their differences concern mainly the performativity, which is only a matter of degree. This can be explained by their syntactic similarity and their affinity in semantic structure. In syntax, both have the same syntactic pattern [I Verb that P] with verbs of different semantic feature. In semantic structure, the putting onstage of SPEECH ACT space is analogous to that of VIEWPOINT space, and then both serve as CONTENT BASE/VIEWPOINT space in onstage region. The speaker, as the subject of conceptualization, is also put onstage in both constructions. Both can be said to be parallel constructions under the higher, more schematic construction.
6.4 [It be ADJ that P] Construction
117
Nevertheless, through the integration of ideational resources from speech act verbs plus the syntactic pattern and its interpersonal meanings from the blending of SPEECH ACT space and VIEWPOINT space, [I VVerbal that P] construction expands the meaning potential of language.
6.4 [It be ADJ that P] Construction As Halliday (1994: 355) claims, encoding an appraisal with the impersonal pronoun it extraposition rather than a subjective projecting clause, such as I think, makes it possible for the speaker to disguise the source of the opinion and thereby make it appear to be objective. This view is further, albeit from a different approach, elaborated and supported by Kaltenbock (2004)’s analysis of it-extraposition and non-extraposition. He demonstrates the importance of supplementing syntactic description with a functional analysis to get a more comprehensive picture of a grammatical category. One general characteristic of it-extraposition is that of being an impersonal construction, which converges with Halliday’s claim on the impersonal pronoun it extraposition. It is a way of avoiding the use of a personal pronoun (I think that …), thereby signaling reduced personal involvement (Kaltenbock, 2005: 137). Thus, it-extraposition allows the speaker/writer to express a personal comment toward some state-of-affairs, but present it as if it were some accepted view rather than his/her personal judgment. If it is true that [It be ADJ that-clause] construction is marked for explicit objective orientation and a way of avoiding the use of a personal pronoun such as I think, I believe, and I know, then its use collocatively with the personal pronoun I as I find in COCA would be hard to explain. Some corpus examples are retrieved below: (99)
… I think that it is true he is going to start releasing systematically all of the hostages, and our hostages and they will be out by the end of March if there hasn’t been a war. (19,901,214, Spoken)
(100)
I believe that it is unlikely that he would have put his wife at risk by transporting her and the stolen gold in the same vehicle. (2009, Fiction)
(101)
I think that it’s essential that there’s regulation on infertility treatments. (19,930,226, Spoken)
(102)
And I think that it’s horrible that with one blanket executive order that he has released from death row all of these individuals instead of approaching it on a case by case basis. (20,030,112, Spoken)
(103)
I believe that it is imperative that the public lands remain under the control of Congress. (19,960,128, News)
(104)
… I think that it is important that the public understand that I am not simply The Wizard of Oz. (20,000,609, Spoken)
(105)
I know that it’s terrible that he sold those junk bonds to little old ladies. (19,901,121, Spoken)
118
6 The Pragmatics and Discourse Functions of MM
If I think/believe/know represents personal and subjective modality and [It be ADJ that-clause] construction expresses impersonal and objective modality, why can both constructions occur in the same utterance? Won’t they discord with each other? The occurrence of the above utterances challenges the dichotomy of objective and subjective modality proposed by Halliday.
6.4.1 The Functions of This Construction in Discourse When we analyze the larger context of this construction, we often find the evidence that justifies the proposition P. Take the extended context of example (106) as an illustration: (106)
And he in fact said he was looking at it and he said that he thought that was probably right and he was going to try and find, it had been 90 days and we haven’t been attacked and he said he would try to find a way and within a week he had made an announcement and I think that it is true he is going to start releasing systematically all of the hostages, and our hostages and they will be out by the end of March if there hasn’t been a war. That is something to look forward to and in the meantime we are getting them out every day. They have been coming out every week a bunch of them coming out. (19901214, Spoken)
In the co-text, we find that both the verbal reaction that ‘he’ made and what have already occurred are evident toward the truth of this expectation. Therefore, it is more feasible to interpret the construction ‘it is true’ in terms of evidentiality rather than objective modality. This pattern becomes more apparent in the extended context of (107): (107)
“… Nash was a different breed of criminal than most that flourished during those days. Yes, he was a thief, but he also was a comparatively honorable man. I believe that it is unlikely that he would have put his wife at risk by transporting her and the stolen gold in the same vehicle. That, of course, is merely conjecture on my part. However, it is supported by the fact that Nash did not have the gold with him when he was apprehended by federal agents six days later in Hot Springs, Arkansas”. (2009, Fiction)
It is clear that the construction ‘it is unlikely’ could not represent objective modality as evident from the next sentence: That, of course, is merely conjecture on my part. Furthermore, the judgment of Nash as a person and ‘the fact’ in the co-text all support that the possibility ‘that he would have put his wife at risk by transporting her and the stolen gold in the same vehicle’ is very low. Evidentiality, again, at least partly motivates the use of [It be ADJ that P] construction. According to Herriman (2000: 582), the function of the matrix predicate in itextraposition is to express the speaker’s opinion toward the content of the extraposed item. As analyzed in Chap. 4, [It be ADJ that P] construction encodes a viewpoint toward the content of the framed clause. Here the analysis of discourse excerpts shows
6.4 [It be ADJ that P] Construction
119
that this construction is distinct from I think or I believe in terms of evidentiality, rather than objectivity versus subjectivity. Evidentiality concerns the speaker’s indication of the nature of the evidence invoked for the state of affairs expressed in the utterance (Nuyts, 2001: 27). Bybee (1985) and Palmer (1986), for example, include evidentials in the category of epistemic modality. In their analysis, on the one hand, there is the speaker’s evaluation of the probability of the situation, i.e., the epistemic qualification; on the other hand, there is his/her characterization of the status or quality of the evidence for that qualification, i.e., evidentiality. The evidential dimension explicates the status of the information on which the epistemic qualification is based, thus interacting with and affecting the epistemic qualification. According to Nuyts (2001: 66), modal adjectives in the standard form of expression [It be ADJ that-clause] do systematically involve an additional evidential meaning, which is at least one of the motives for using them as found in several of his corpus cases. My corpus cases retrieved from COCA also support this point. (108)
According to documents obtained by the Monitor, it is clear that Perry does have two passports, but the second one was issued after the first one ran out of pages.
(109)
And it is true that his weaknesses remain. He’s got the same weaknesses with the same demographics.
(110)
Bipolar disorder is known to have a strong genetic component-so It’s likely that my genes predisposed me to have a mood disorder.
When it is clear that, it is true that, or it’s likely that is used, evidence could be found in the context, such as “documents”, “demographics”, or knowledge concerning “bipolar disorder”. It can be inferred that the motive for the use of [It be ADJ that-clause] construction has something to do with evidentiality. And the nature of the evidence invoked for representing the state of affairs tends to be objective. So Halliday’s explicit objective orientation does not equate to objective modality; if it is called objective, it should be in the sense of objective evidentiality, which is a dimension of modality. I agree with Nuyts (2001: 306) that it is more reasonable to handle modality and subjectivity as two different categories. [It be ADJ that-clause] construction mainly operates on the dimension of evidentiality, while expressing a view toward the content of that clause. Besides evidentiality, is there any other function that this construction plays in discourse? Consider, for example, discourse segment (111) below:
120
(111)
6 The Pragmatics and Discourse Functions of MM
When Oleg Kashin, a prominent journalist and outspoken Kremlin critic, was severely beaten in Moscow in 2010 -- both his legs were broken and he lost a finger -- suspicion immediately fell on soccer hooligans hired by government friendly forces. No one has been charged. Still, it is clear that such groups are not fully under the government’s control. The rioting and racist attacks that occurred in December 2010 began after gangs of youths began throwing signal flares and other projectiles at the police, eventually forcing them to back down. The incident clearly caught the Kremlin off guard. Vladimir V. Putin, who was prime minister at the time and is now president, took the unusual step of meeting with the leaders of soccer fan clubs to defuse tensions. (120616, Newspaper)
One primary function of language is to convey information from the speaker to the addressee. From the addressee’s perspective, information is thought in terms of whether it is more or less expected or predictable given the setting and the previous discourse (Givon, 1983). It is expected that information which is relatively accessible or predictable to be coded with less linguistic work; conversely, information which is relatively surprising should be coded with special or marked linguistic mechanisms. If this is the case, Still, it is clear that indicates that the information “such groups are not fully under the government’s control” is discourse-new. A look at the linguistic context supports this motivation. From the previous sentence, the readers understand that there was “suspicion” that “soccer hooligans” may be under the government control, somewhat indirectly. Immediately afterward, the conjunction “still” indicates an adversative relation with the previous utterance, and the readers could infer that what follows will be in contrast to the expectation derived from the content just said. The information being ‘contrary to expectation’ or in contrast seems a conventional context for this construction. Let us examine more examples: (112)
BAIER: Steve, final thoughts? HAYES: #1 Yes, I have a little bit different take. #2 I mean, I agree with Charles and A.B. and Kirsten that the big picture, Mitt Romney moves ahead in delegates slightly. #3 We don’t know the exact number right now, but he’ll be ahead at the end of the night in delegates from tonight. #4 He’ll add to his lead in that sense. #5 And it is true that his weaknesses remain. #6 He’s got the same weaknesses with the same demographics. (120313, Spoken)
This is the transcription excerpt of an interacted news commentary from Fox_ Baier program titled “Santorum Wins Both Mississippi and Alabama Presidential Primaries”. From the context, we can know that Hayes’s utterances are all about his viewpoints or commentary on the news “Santorum Wins Both Mississippi and Alabama Presidential Primaries”. In Sentence #1 and #2, “I” is the conceptualizer is on stage. Then in Sentence #3 and #4, the object of conceptualization is blended with the subjectivity, which is cued implicitly by will in “he’ll” indicating a sense of inclination viewed by the speaker. Therefore, the speaker’s degree of commitment for the validity of the Content is rather high. Part of the reason is that this expectation is in agreement with Charles and A.B. and Kirsten (other news commenters), and there is no need for extra rhetoric effort. Sentence #5, “And it is true that his weaknesses
6.4 [It be ADJ that P] Construction
121
remain” is followed by evidence “He’s got the same weaknesses with the same demographics”, presented to substantiate this claim. This clearly shows evidentiality can be one of the motive for using [it be ADJ that P] construction. According to Nuyts (2001), adjectival constructions as such systematically involve an additional evidential meaning (p. 66). From the utterance before “And it is true that”, the readers understand that the speaker thinks “Mitt Romney” will “add to his lead”. Therefore, the speaker’s bringing up of “his weaknesses” is quite contrary to what would be expected. In other words, the information led by it is true that construction is contrastive to the previous utterance. [It be ADJ that P] construction may be used in cases where there is the need to weaken or mitigate the contrastive or adversative force of the proposition P. We will analyze more examples systemically. Case I: it is true that p Let us observe the following paragraph from an article in the academic journal MIDDLE EAST QUARTERLY (SPRING 2012), with the use of it is true that p construction: (113)
#1 Some argue that sanctions have the same effect of rallying the Iranian people behind the regime, but conversations with Iranians have not borne this out. #2 Whether an Iranian is likely to place the blame for the sanctions on Ahmadinejad’s hostile statements or U.S. and European hawkishness tends to depend on their preexisting political views. #3 It is true that sanctions cannot do much to hinder the activities of the likes of the Quads Force, the external Iranian intelligence agency, or the “millionaire mullahs”, but their loosening or tightening can be an invaluable pressure card against the regime. (Middle East Quarterly. 2012: COCA)
A cognitive construction compatible with this piece of discourse proceeds as follows: [Sentence #1] Some argue that sanctions have the same effect of rallying the Iranian people behind the regime, but conversations with Iranians have not borne this out. The space builder Some argue sets up a VIEWPOINT space V that will partition off information about some others’ perspective on “sanctions” which is being considered. The complement clause will structure this new space, viz. FOCUS space M. The second clause but conversations with Iranians have not borne this out establishes the Base space B. The word but is an explicit pragmatic signal: the structure of B differs from that of M, showing that it is incompatible with its counterpart in space M. The succession of mental space construction and connection with the unfolding of discourse can be represented in Fig. 6.1. [Sentence #2] Whether an Iranian is likely to place the blame for the sanctions on Ahmadinejad’s hostile statements or US and European hawkishness tends to depend on their preexisting political views. This is a proposition but it is modalized in that the predicate “tends to” cues a degree of usuality. As indicated in Fig. 6.1, there is a Ground Base/Viewpoint space
122
6 The Pragmatics and Discourse Functions of MM
Fig. 6.1 Succession of space configuration in discourse
grounding the content of this proposition, and it is not profiled, in other words, it is off stage. Here, the cuing of the speaker’s viewpoint is too implicit to be noticed if not for this type of linguistic analysis. This utterance, similar to ‘Mary will know’, which has been analyzed in Chap. 3, falls into the pattern of non-perspectivized utterances. [Sentence #3] It is true that sanctions cannot do much to hinder the activities of the likes of the Quads Force, the external Iranian intelligence agency, or the “millionaire mullahs, but their loosening or tightening can be an invaluable pressure card against the regime.
6.4 [It be ADJ that P] Construction
123
As revealed in Fig. 6.1, the space builder ‘It is true’ sets up a new VIEWPOINT space V’ that will partition off information about desubjectified perspective on “sanctions” which is being considered and acknowledged. The term ‘desubjectified’ is used because this pattern does not attribute the thought directly to one of the conceptualizers in the ground, as there is no onstage conceptualizer. In a word, this view has its epistemic grounding of unidentified others, probably the readers as well, though all implicitly. It is also a strategy to pave the way for the readers to accept the author’s view immediately afterward—“but their loosening or tightening can be an invaluable pressure card against the regime”, whose viewpoint grounding, though, is offstage. Meanwhile, embedding profiled content (sanctions cannot do much to hinder the activities of the likes of the Quads Force, the external Iranian intelligence agency, or the “millionaire mullahs”) in the Viewpoint space built by “It is true that” indicates the author’s epistemic distance from the Content, for a modal adjunct such as can is usually blended within the propositional content on the same plane, whereas “It is true that” and the proposition P are operating on two planes: the former interpersonal, the latter ideational. Therefore, this construction reflects the author’s epistemic distance from P and the author’s degree of commitment for its validity is much reserved. It has the function of managing the intersubjective coordination between participants in the ground, which conforms to Verhagen’s (2005: 132) argument. This sentence has the pattern or colligation: it is true that p, but…, in which “but” expresses an adversative relation and means ‘contrary to expectation’ (Halliday & Hasan, 1976: 250). This suggests that it is true that p may also have the function of introducing a viewpoint that has an adversative relation to other opinions in the context. Similar discourse pattern with this construction is observed in COCA: (114)
Senator Barack Obama, who announced his candidacy for the U.S presidency on Feb. 10, has been lauded time and again for his remarkable charisma. It is true that when he is on the stage, you have trouble looking at anybody else. Yet charisma gets you only so far in politics. (2007, Magazine)
(115)
It is true that compared with the century-old telephone system, the Internet is a paragon of deregulation and decentralization. In four critical areas, however, it requires oversight and coordination in order to operate smoothly. (2005, Academic)
Likewise, the construction it is true that reflects the author’s epistemic distance from the proposition expressed by the complement clause and the author’s degree of commitment for its validity is much reserved. It is also a strategy to manage the intersubjective coordination between the speaker/writer and the addressee. The transitional expressions yet and however express an adversative relation and means ‘contrary to expectation’. This colligation suggests that it is true that p may also have the function of leading to a viewpoint that has an adversative relation to the speaker’s intentionality. However, being different from the speaker’s viewpoint is not equal to objective modality; instead ‘desubjectified’ or ‘intersubjective’ is more likely to be the feature of this construction.
124
6 The Pragmatics and Discourse Functions of MM
Case II: it is clear that P The pattern it is clear that p, similar to it is true that p, has the semantic feature of epistemic modality. Do they perform the same pragmatic functions? The following example will help give an idea of how it is clear that P is used in discourse. (116)
#1 When Oleg Kashin, a prominent journalist and outspoken Kremlin critic, was severely beaten in Moscow in 2010 -- both his legs were broken and he lost a finger -- suspicion immediately fell on soccer hooligans hired by government friendly forces. #2 No one has been charged. #3 Still, it is clear that such groups are not fully under the government’s control. #4 The rioting and racist attacks that occurred in December 2010 began after gangs of youths began throwing signal flares and other projectiles at the police, eventually forcing them to back down. #5 The incident clearly caught the Kremlin off guard. #6 Vladimir V. Putin, who was prime minister at the time and is now president, took the unusual step of meeting with the leaders of soccer fan clubs to defuse tensions. (MICHAEL SCHWIRTZ, New York Times. 2012: COCA)
The it is clear that P construction resembles it is true that p in two aspects: (1) this pattern does not attribute the thought directly to one of the conceptualizers in the ground, as there is no onstage conceptualizer; (2) the sentence is followed by evidence presented to substantiate this claim, which is narrated from Sentence #4 to #6. In other words, it is followed by rather elaborated depiction of the evidence. So, evidentiality can be one of the motive for using it is clear that P construction, similar to it is true that construction. A pattern it is clear that p, but Y is also detected, attested by corpus examples from COCA: (117)
a.
It is clear that it is our right to respond, but if we respond militarily, then we will be acting in accord with the Israeli agenda, which we are not interested in doing. (2008, academic)
b.
According to documents obtained by the Monitor, it is clear that Perry does have two passports, but the second one was issued after the first one ran out of pages. (20020624, Newspaper)
c.
When Sir Thomas and Tom return from Antigua it is clear that there is something wrong between them, but neither explains what happened. (1999, Magazine)
d.
It is clear that she likes Ben, but when his face gets close to hers she smells the alcohol on his breath and she turns away. (1995, Fiction)
e.
It’s clear that they know that. But they just decided they weren’t going to call the tune. (19951004, Spoken)
f.
Taking all these trials into consideration, it is clear that PS is not a single-nutrient therapy for dementia, but it does offer some degree of value (2005, Academic).
These examples are from various genres: both spoken and written, including academic, news, fiction, and magazine as well, and span from 1990 to 2000s, which are good representatives of contemporary American English. This shows it is clear
6.4 [It be ADJ that P] Construction
125
that p as well as it is true that p shares the same feature of introducing something adversative to the speaker’s viewpoint. In other words, both represent desubjectified or intersubjective viewpoints. Case III: it is important that P The following example will help give an idea of how it is important that P, a prototypical [It be ADJ that P] construction representing appreciation, is used in discourse. (118)
#1 When an SCI occurs to a baby, toddler, child, or young teenager, sexuality is an equally important topic, but the issues are somewhat different. #2 Parents often report that soon after injury, they are bothered by questions about their child’s eventual ability to enjoy sexual intimacy and to have children. #3 These questions are a natural result of parents trying to come to grips with all of the implications of the injury for their child’s future life, even if the child is young at the time. #4 Because some parents are shy about raising the topic of sexuality with their physicians, it is important that healthcare professionals initiate such discussion as part of the acute treatment and initial rehabilitation. #5 Parents must be reassured that injury to a premenarchal female will not affect age of menarche or on their daughter’s ability to become pregnant or have children. #6 Parents whose sons have sustained an SCI must be reassured that although there may be problems with erections, ejaculation, and fertility, there are treatments that are available and that will continue to improve. (Orthopaedic Nursing. 2004, Vol. 23 Issue 5, p300–308)
Pragmatically, it is important that P construction in sentence #4 resembles it is true that p, it is clear that P, and it is possible that P in two aspects: (1) this pattern does not attribute the thought directly to one of the conceptualizers in the ground, as there is no onstage conceptualizer; (2) the content filled in the context of it is important that P construction is in contrastive relation to “some parents are shy about raising the topic of sexuality” in the previous context. What’s more, [it is important that] is independent from the proposition to which it is accessing, which may upgrade the potency or illocutionary force for uttering something potentially Adversative. Notice in sentence #5 and #6, “must” is used, which represents deontic modality. In other words, it is the obligation/duty of “healthcare professionals” to reassure parents (and parents “whose sons have sustained an SCI”). But also note that passive voice is used and as a result the subjects of this obligation go implicit. Rhetorically, the writers intend not to impose something on healthcare professionals, who are target readers of the academic journal Orthopaedic Nursing. Likewise, it is important that P construction is used to distance the writers from the viewpoint, from imposing something on their colleagues. In turn, it achieves a sense of professionalism. In addition, the premise leading to the conclusion P, that is, Because some parents are shy about raising the topic of sexuality with their physicians, is available to the addressees. Therefore, a dimension of evidentiality is also involved.
126
6 The Pragmatics and Discourse Functions of MM
Case IV: it is essential that P The following example will help give an idea of how it is essential that P, a prototypical [It be ADJ that P] construction representing Deontic Modality, is used in discourse. It is a discourse segment of a research article published on Journal of Emotional & Behavioral Disorders. (119)
#1 Findings from this, when interpreted in conjunction with the findings of other treatment-outcome studies (see Ayllon &; Roberts, 1974; Coie &; Krehbiel, 1984; Lane, 1999) provide preliminary evidence to warrant implementation of this type of practical, cost-effective, academic-oriented intervention on a larger scale, particularly given that the outcomes might ultimately result in long-term benefits to the children and to society as a whole. #2 Clearly, it is essential that interventions be conducted early in a child’s educational career (Walker &; Severson, in press). #3 Although the outcomes of this study are encouraging in that they provide direction for how best to intervene with this most challenging population, the findings are, at best, preliminary. (2001, Academic)
It is essential that P construction resembles it is true/clear that p in two aspects: (1) this pattern does not attribute the thought directly to one of the conceptualizers in the ground, as there is no onstage conceptualizer; (2) circumstances of evidence are provided for this claim, though here it is in the form of references in the bracket. What’s more, deontic modality is defined in terms of social or institutional laws, which may be imposing on the addressees. The proposition in sentence #2 implicates that educators or trainers relevant should conduct interventions early in a child’s educational career, which is something potentially imposing. Speakers naturally would adopt pragmatic strategies to dress something imposing. Negative politeness, for instance, is oriented toward addressee(s)’ negative face, involving his/her desire to be left free to act as they choose (Thomas, 1995: 172). The strategies like ‘point of view distancing’ and impersonality are among Brown and Levinson’s list of negative politeness strategies. In it is essential that interventions be conducted early in a child’s educational career, as a case in point, by using an impersonal construction [It be ADJ that P], the speaker/writer distanced himself/herself from the viewpoint expressed. Likewise, it is important that P construction as analyzed above is also used to distance the writers from the viewpoint, from imposing something on their colleagues. What about the other constructions like it is clear that P? The example (116) is a report from the newspaper The New York Times, with the sentence: it is clear that such groups are not fully under the government’s control. Newspaper correspondents in general will distance themselves from explicit evaluation, as being neutral and objective is news value, and they do not want to appear partial to one side, in this case, the Kremlin government. Meanwhile, saying something adversative to the ‘suspicion’, which could be public opinion, is also seen as an act of negative face. [It be ADJ that P] construction is used, as a result. “It is true that” also indicates the author’s epistemic distance from the Content. As interpreted from the context, we can see the writer himself is quite reserved about, or disagree on, the proposition sanctions cannot do much to hinder the activities of the likes of the Quads Force.
6.4 [It be ADJ that P] Construction
127
Therefore, he projects it onto a clause on the ideational level, separated from the projecting clause of viewpoint. Case V: it is sad that P The following example will help give an idea of how it is sad that P, a prototypical [It be ADJ that P] construction representing affective modality, is used in discourse. It is a discourse segment of a letter to the editor, published in a magazine. (120)
#1 Thank you for the article, “Whitney At 35!” (Dec. 1998). It was a very moving and soul -- searching article. #2 It is sad that she must always defend her personal and professional life. #3 I am glad she is such a deep Sister who seems to have things the way she wants them. #4 As a Christian, you are taught to build each other up, and if everyone was doing that, everyone would be standing tall. (1999, Magazine)
The [it is sad that P] construction resembles [it is true/clear/essential that p] in two aspects: (1) this pattern does not attribute the thought directly to one of the conceptualizers in the ground, as there is no onstage conceptualizer; (2) although evidence is not explicitly provided for P (she must always defend her personal and professional life), but we can see from the context that the tenor of the letter involves a reader of the magazine expressing his/her view on the article “Whitney At 35!” and the proposition P is probably based on this article. So evidentiality is still involved in the use of [It be ADJ that P] construction. What’s more, intersubjectivity is involved in sentence #2 (It is sad that she must always defend her personal and professional life). Intersubjectivity is a well-received notion in both cognitive science and linguistics now, yet it turns out to be understood in the restricted sense of ‘shared’ or ‘mutual’ understanding between the speaker and the hearer. Actually, in Husserl’s (1960) original writings intersubjectivity includes the role of tools and other artifacts in evoking other minds and other lives, the sense of belonging to a community or to a particular relationship even when others are not co-present, the access to and use of human languages and other semiotic resources. Therefore, as I see it, intersubjectivity should be extended to encompass not only the speaker and the addressee, but also the others whose minds and/or semiotic resources are evoked; they all can be inputs into VIEWPOINT spaces or SPEECH ACT space, even though the others may not co-present in Ground Base space. What’s more, if we apply the distinction between subjectivity and intersubjectivity recently developed, instead of the dichotomy of subjective and objective orientation, to differentiate [I Vmental that P] construction and [It be ADJ that P] construction, it would pose little problem to explain the use of [It be ADJ that-clause] construction in colligation with [I Vmental that P] construction. Subjectivity and intersubjectivity are on a cline, and language users can make most use of the two constructions to adapt the degrees of subjectivity. Besides, sentence #3 introduced by “I am glad” has adversative relation with sentence #2 (It is sad that she must always defend her personal and professional life) This shows it is sad that p, as well as it is clear that p, is used as a prompt to something adversative afterward.
128
6 The Pragmatics and Discourse Functions of MM
In summary, the distinction between subjectivity and intersubjectivity can better explain the discourse behaviors of [I Vmental that P] construction and [It be ADJ that P] construction than the dichotomy of subjective and objective orientation. In addition to intersubjectivity, evidentiality and information flow such as prompting something adversative or contrastive may also motivate the use of [It be ADJ that P] construction.
6.4.2 The Distributional Features of the Construction The distribution of [It be ADJ that P] construction is uneven among different genres. We can clearly see that the distribution of [It be ADJ that P] construction is much higher in academic genres than other genres, as Table 6.3 shows: The token frequency of this construction in academic texts is roughly two times of the average, and about four times of fiction. Genre helps with grouping together texts that have similar purposes, structures and contexts. In SFL model, genre is positioned as an abstract level of analysis coordinating field, mode, and tenor, which are collectively realized in turn through language, including discourse semantics and lexicogrammar. Genre is responsible for coordinating the recurrent configurations of meaning in a culture (Martin & Rose, 2008: 232). Swales (1990) has made clear that we need to see discourse community and genre together to offer a framework of how meanings are socially constructed in writing. The idea of discourse community draws together a number of key aspects of context that are crucial to the production and interpretation of spoken and written discourse: knowledge of a cultural and interpersonal situation, knowledge of interlocutors, knowledge of the world, and knowledge of texts and conventions for saying things. Table 6.3 Distribution of [It be ADJ that P] construction in COCA among different genres Section
All
SPOKEN
FICTION
MAGAZINE
NEWSPAPER
ACADEMIC
FREQ
29,533
6317
2942
5045
4257
10,972
PER MIL
63.60
66.10
32.53
52.79
46.41
120.48
6.4 [It be ADJ that P] Construction
129
Looking at specific genres tells us a lot about how we create meanings. Take academic genre as a case, the general principles underlying academic discourse include economy, the emphasis upon clarity, rational argument supported by evidence, caution and restraint. Evidence is a basic part of any academic exploration as it is often the means and even the ends of most systematic scholarly studies. Bennett’s (2009) survey of English academic style manuals finds that a number of authors stress the empirical basis of the academic project by insisting that arguments presented be based upon evidence or facts. The norm of academic writing is not to say something for which one lacks adequate evidence. As a rule of thumb, experienced authors exercise caution to live up to the accuracy requirement largely achieved through evidentiality in research articles. Evidentiality involves the speaker’s/writer’s characterization of the status or quality of the evidence for that qualification, which can be represented by [It be ADJ that P] construction. Look at a discourse excerpts from an academic journal: (121)
One of the museums’ greatest strengths can be seen as their ability to replicate themselves: the original founders of the union, Cuauhtmoc Camarena and Teresa Morales, continue as spokespeople for the union, but it is clear that they are training the next generation -- names of new representatives are appearing at some of the international conferences at which the union is represented. (American Indian Quarterly Summer/Fall 2006, Vol. 30 Issue 3/4, p441–460, 20p)
The dimension of evidentiality is clearly involved with the use of this construction: the immediate context “– names of new representatives are appearing at some of the international conferences at which the union is represented” provides the evidence for this qualification. It appears that evidence or empirical circumstances can be one of main motives for using [It be ADJ that P] construction, which is believed to part of its meaning in Perkins (1983). Note another example: (122)
Critical to all mastery attempts by children is the opportunity to gain rewards and to learn to set mastery goals that are optimally challenging yet realistic. This model has been widely applied to the sports and physical activity domain with able-bodied children, and researchers have investigated the motivational constructs among children with disabilities. Children who are visually impaired have often been denied the opportunity to set their own mastery goals, and their parents and teachers may protect them from attempting to perform certain physical activities. Thus, it is likely that they do not enjoy certain physical activities because they have not experienced successful mastery attempts in the physical domain. (Journal of Visual Impairment & Blindness Dec2006, Vol. 100 Issue 12, p726-736, 11p)
In example (122), evidence for the very proposition is also provided in the context, this time being based on reference to literature. Besides, logical argumentation leads to the proposition embedded with [It be ADJ that P] construction. This is also the case with the following example:
130
6 The Pragmatics and Discourse Functions of MM
(123)
China produces the largest source of sulfur dioxide (SO 2) and sulfur emissions in East Asia. Since China remains heavily dependent on coal as an energy source, it is unlikely that this trend will be reduced significantly in the near future. (International Social Science Review 2006, Vol. 81 Issue 1/2, p43–57, 15p)
The discussion of data is an important part in research articles. The data is typically displayed in a table, graph, figure, or some other kind of non-verbal illustration. The construction such as [it is clear that] is used in data commentary, like the commentary on a table: From table 3, it is clear that F calculated value of 94.521 is statistically significant (Education, Winter2006, Vol. 127 Issue 2, p176-186). Sometimes, a series of [It be ADJ that P] constructions are used in data commentary on a table, as the following example illustrates: (124)
The methionine (M)/valine polymorphism at codon 129 of the PRNP gene has been examined in populations with and without CJD in many countries; results have varied (Table 3). Overall, it is clear that the M allele bestows substantial susceptibility to the sporadic and the iatrogenic forms of CJD; in consequence, the proportion of persons with MM homozygous genotype is overrepresented in both categories of disease (the sole exception occurred in UK growth hormone recipients, which led to speculation that a different strain of the pathogenic agent might have been disseminated) (10). It is also clear that, as a group, persons with heterozygous genotype had longer incubation periods than did those with homozygous genotype, particularly in France. Notwithstanding this statistical conclusion, it is noteworthy that several persons with MM homozygous genotype had incubation periods > 30 years, including a patient with recently diagnosed CJD, whose incubation period was 42 years, the current world record for any type of iatrogenic disease.
This is a paragraph of a research article published in the journal Emerging Infectious Diseases • June 2012, in which the writers discuss the data displayed in Table 3. All together three constructions of [It be ADJ that P] are used in four sentences. These constructions are used by the writers to instruct the addressees to pay joint attention to certain aspects of the data. Their functions are primarily intersubjective, that is, instructing the addressees to jointly view something. Moreover, the data or evidence can be displayed along with the text itself, or only mentioned based on intertextual evidence, as illustrated in the following example: (125)
From existing documents it is evident that she did communicate oral instructions to Nicolo regarding the decorations and provisions for her foundation. (Art Bulletin Dec94, Vol. 74 Issue 4, p644, 20p, 8bw)
In the sciences, new knowledge is accepted by experimental proof. The humanities, on the other hand, rely on case studies and narratives while claims are accepted on strength of argument. The social sciences fall between these extremes. In whatever fields of study, proof or argument is essential; no wonder evidentiality is an important dimension in academic representation.
6.4 [It be ADJ that P] Construction
131
Over the past twenty years, academic writing has gradually lost its traditional tag as an objective and impersonal form of discourse and come to be seen as “a persuasive endeavor involving interaction between writers and readers” (Hyland, 2005: 173). Writers seek to offer a credible representation of themselves and their work by acknowledging alternative views and evaluating their material, so that controlling the level of personality in a text becomes central to building a convincing argument. This context of academic culture would sanction desubjectified representation of viewpoints, which also implicates caution and restraint. The following is such a case: (126)
While it is unfortunate that “Convention, Context, Change” excluded painting and graphics and retained the racial barrier, the level playing field -- upon which “power-knowledge plays” will nevertheless continue unabated -- sometimes shimmers on the horizon. (African Arts, Winter94, Vol. 27 Issue 4, p54, 14p, 1 diagram, 11c, 12bw)
The use of [It be ADJ that P] construction enables the academic writer to express his/her hedged negative viewpoint toward “Convention, Context, Change” excluded painting and graphics and retained the racial barrier in a desubjectified, cautious and restrained manner. Persuasion the academic genre involves the use of language to relate independent beliefs to shared experience. Writers fan support, express collegiality, and negotiate disagreement through patterns of rhetorical choices that connect their texts with their disciplines (Hyland, 2008: 549). For instance: (127)
As the focus shifts in education to the test scores of unsatisfactory achievers or partially proficient students, it is imperative that our advanced students receive appropriate instruction as well. (Roeper Review, Fall2006, Vol. 29 Issue 1, p28–31, 4p)
In this educational review article, the teacher expresses his different viewpoint on No Child Left Behind: our advanced students should receive appropriate instruction as well. The construction “it is imperative that” is used to negotiate disagreement in a way that is impersonal and professional. It is very common for academic writers to view a phenomenon from a perspective different from previous studies, as novelty is valued in research. The academic world is actually severely competitive, yet academic writers need to maintain harmonious collegiality as their peers serve as reviewers, gatekeepers, and target readers of their research articles. The key, then, is expressing different viewpoints, which are potentially face threatening toward their colleagues in the same discourse community, in a mitigated way to fan support. [It be ADJ that P] construction is one of such mitigating rhetorical means, as further illustrated in the following example.
132
(128)
6 The Pragmatics and Discourse Functions of MM
Emphasis on winning is certainly one of the reasons cited for some parental stress in coaching. However, it is possible that coaches contribute to their own dilemma. They may be oblivious to what most parents want from sport participation. Perhaps it is the “ over involved “ parent who is the vocal minority in sport. A minority, who in over emphasizing winning, stifles or “ drowns out “ the expectations of the majority of parents. (Physical Educator, Fall94, Vol. 51 Issue 3, p 130, 8 p)
Apparently adding the layer of [It be ADJ that] to a proposition seems to violate the general principle of economy, but there is the competing iconic motivation, for example, the linguistic distance between expressions corresponds to the conceptual distance between them (Haiman, 1983: 782). The layer of [It be ADJ that] symbolizes the conceptualizer’s epistemic distance toward the proposition, indicating he is distant from the commitment to the validity of the claim. [It be ADJ that P] construction tends to occur in local contexts where considerations for evidentiality, epistemic distance and pragmatic needs outweigh other features. In order for claims made in research articles and academic papers to be accepted by disciplinary gatekeepers, in the form of journal referees and examiners, they need to be phrased in ways that gatekeepers find persuasive. The rhetorical strategies employed by authors in these academic genres reflect power symmetries between authors and gatekeepers and constitute the means by which academic authors display awareness of the discourse community’s expectations, and attempt to establish credibility and membership in the community (Koutsantoni, 2006: 33). Achieving persuasiveness of their articles toward journal referees and addressees, experienced authors exercise caution to live up to the accuracy requirement, through evidentiality and epistemic distance in research articles. Since [It be ADJ that P] construction is often thought to be associated with objectivity, it is possible that some academic authors exploit it in order to package their opinions as facts. The following is a case in point: (129)
Although the Office of Applied Studies (National Survey on Drug Use and Health, 2006) report indicates that the incidence of drinking among Black adolescents is less than that found among Whites, it is possible that at least some Black adolescents may consume alcohol for reasons related to ongoing discrimination and its concomitant anger. That is, it is possible that, without being aware of it, in order to cope with widespread and continuous exposure to racism, some Black adolescents may engage in a form of self-medication consisting of alcohol abuse. (Adolescence, Fall 2006, Vol. 41 Issue 163, p 485–492, 8p)
The use of modal verb ‘may’ indicates subjective modality, that is, the proposition is the writer’s opinion, or speculation. Nevertheless, the writer embeds this subjective proposition with the construction it is possible that, making it seemingly factual. Usage like this could be writers’ strategy to exploit [It be ADJ that P] construction: packaging their opinions as factual or objective.
6.4 [It be ADJ that P] Construction
133
To summarize, there are a number of discoursal or pragmatic factors motivating the use of [It be ADJ that P] construction in the academic genre: (i) evidentiality or data-basedness; (ii) desubjectificaition or epistemic distance; (iii) disagreement negotiation in an impersonal and professional style; (iv) rhetoric mitigation; and (v) strategic exploitation. They reflect what the academic discourse community value and expect, which might account for the highly frequent occurrence of this construction in the academic genre. The search in COCA finds that the token frequency of [It be ADJ that P] construction in academic texts is about four times of fiction. This section explores why its occurrence in fiction is much lower than that of academic texts. The list of [It be ADJ that P] construction types in academic versus fiction genres can be found in Appendix II. We focus on those whose ratio (the relative percentage in the two sets of sections) is larger than 1.0. There are 177 construction types whose A/F (academic: fiction) ratio is larger than 1.0 and 141 types whose F/A (fiction: academic) ratio is larger than 1.0. As for construction types with a ratio of no less than 4.0, there are 62 types in terms of A/F ratio but only 18 types in terms of F/A ratio. The highest F/A ratio is 7.74, whereas the highest A/F ratio is 238, with 26 construction types having A/F ratio of more than 10. All these add up to the trend of much higher presence of this construction in academic genre than fiction. In academic genre, for instance, [It be ADJ that P] construction is frequently used in data commentary, which surely is not the way the world or events are construed in fiction. In the fiction, there is the slanting of the fictional world toward ‘reality’ as apprehended by a particular participant or set of participants. A fiction writer can voluntarily limit his omniscience to those things that belong to one person’s model of reality. He can also vary the fictional point of view, sometimes claiming authorial omniscience, sometimes presenting one character’s version of events, sometimes that of another (Leech & Short, 2007: 140). In general, fiction is an artistic form of narrative. In academic writing, the writer is doing a different job: argumentation. The writer draws on interactional resources with the aim to involve the readers in the argument of the text. In this way, the writer conducts more or less overt interaction with the addressee. [It be ADJ that P] construction invites the addressee to view the proposition P in a particular way and facilitates the addressee’s cognitive coordination with the writer, who, nevertheless, is just grounding the predication covertly. Only through maintaining the addressee’s cognitive coordination, with the outcome that the addressee is following the writer’s line of argumentation, the writer can successfully persuade the addressee to take his arguments or claim. Therefore, in academic texts, the writer frequently takes an intersubjective viewpoint, managing the mutual cognitive states of the writer and the addressee. This partly explains the high frequency of [It be ADJ that P] construction in the academic genre. In the fiction, in contrast, the point of view is from either the writer, or one character, or any other character. In order to create the mock reality of the fictional world, the author usually does not interact with the reader directly. Constructions
134
6 The Pragmatics and Discourse Functions of MM
of intersubjectivity between the writer and the reader such as [It be ADJ that P] construction are not favored in the fiction. When this construction does occur in the fiction, it often appears in conversation in the novel. Take [it be cool that] as an example, all of its occurrences are in the context of represented speech in conversation: (130)
Addie: I think we’re getting a little off the --Bobby: It’s cool that you’re not saying the Pledge, Addie, I mean it’s cool that you’re standing up for your principles and all, but --Addie: Thank you. Bobby: But what difference does it make? I mean, just because you sit there and don’t say the words with everybody else, that’s not going to help some poor guy hundreds of miles downstate in New York City who gets beaten up just because he’s black or poor or something. Addie: I contend that it does make a difference. (COCA FICTION. 2003. The Misfits. New York: Aladdin Paperbacks)
(131)
“I think it’s cool that you bake, Gordon”, Stephanie said. # “Lion, you love his lunches, “Zee prompted”. Gordo made those lunches?” Lion said, dumbfounded. (COCA FICTION. 2003. Weirdest of all. Highlights for Children Vol. 58, Iss. 12; pg. 10)
(132)
“Don’t sweat it”, she said. “I think it’s cool that you cuss and stuff when you’re mad. Everybody was paying attention. What you said was right, these guys just think when they use something up there’s always going to be more”. “I shouldn’t have cussed”, I said. (COCA FICTION. 1991. Animal Dreams)
Conversation in fiction mocks reality of the interaction between subjects. In example (130), the speaker uses ‘It’s cool that’ as a rhetorical strategy to pave the way for the addressee to accept his view in which he disagrees with the latter. In example (131) and (132), the speaker is using the construction for evaluation or comment on an event. The motif of using [It be ADJ that P] construction does not appear to come from objective modality. Rather, this construction has expanded the meaning potential of modality: modal verbs can express epistemic, deontic, and dynamic modalities; by [It be ADJ that P] construction, evaluative modality is added to the meaning potential. In general, the two-way interaction of conversation sanctions such constructions of intersubjectivity as [It be ADJ that P] construction. As shown in the above examples, [It be ADJ that P] construction is also frequently used in colligation with [I Mental Verb], like I think and I mean. Besides their contribution in terms of subjectivity and intersubjectivity, a major motivating factor is to expand the meaning potential of modality, with the lexical resources of the adjectives and verbs interacting with the syntactic pattern. The contextual feature of conversation in fiction is prominent in other [It be ADJ that P] constructions besides It’s cool that, as shown in the following examples: (133)
“Weird?” Bernice said, frowning. “Uh, yeah -- it’s weird that I should come into what looks like an ordinary coffee shop and get the best blueberry pancakes I’ve ever eaten”. (COCA FICTION. 1991. Cold Fire. New York: G. P. Putnam’s Sons)
6.4 [It be ADJ that P] Construction
(134)
ACANIT: There is a lot of pain in this house. # CLAIRE: Look, dear... I think it’s fine that you’re appearing to me, and that you know things about me, in essence it’s normal, you’re a secondary effect, in a manner of speaking, surely caused by... I don’t know, and the staging is very... effective, with your attire and the dark skin and the hard times, to be sure. But I’ll tell you here and now, I can’t take you seriously if you start talking in that way. (COCA FICTION. 2011. MARBURG. Theatre Forum—International Theatre Journal)
(135)
Remembering seeing her again, on the Fourth of July among the palms and marigolds of Brand Park near San Fernando, tall and majestic among her shorter sisters, he recalled his brother told her, “Hey, why don’t you call Nico? Nico wants to see you, girl! “and she said,” No, I better not, I still love Nico but it’s better that I don’t see him “ while her sisters scuffed at the dirt, grimacing. (COCA FICTION. 2005. To Reign after Death. Southwest Review Vol. 90, Iss. 4; pg. 555, 13 pgs)
135
The conversation between these characters could easily occur in real life. It is observed through the concordance lines of [It be ADJ that P] in fiction that the majority appear in the context of conversation. So we can say that this construction is preferred in certain context in fiction, or in other words, there is contextual constraints on the use of [It be ADJ that P] construction in fiction. This partly explains the much lower frequency of this construction in fiction than that in academic genre. The token frequency of this construction in spoken genre is 66.10 per mil, which is the only one above average next to academic genre. Note the following discourse excerpts from different radio programs in COCA. (136)
SUSAN STAMBERG: Usual trouble spots prevail. Chechnya, Bosnia - a lot goes on. Nothing changes. DANIEL SCHORR: Chechnya, Bosnia—you know, it’s funny that almost any Saturday we can come up to this point and say,’ Meanwhile, in Chechnya,’ and’ Meanwhile, in Bosnia, a little more despair, a few more people dead. (19950204, NPR_Weekend)
(137)
SCHIEFFER: The president keeps saying that -- that he didn’t know anything about it. How, after all that has come out, can anyone reasonably conclude that the president was out of the loop about all of this, Mr. Secretary? Mr-KEMP: I’ll be honest: I don’t know. I certainly wasn’t in the loop as the housing secretary or in the Congress in the last administration. But look, it is possible that Cappie Weinberger, Secretary of Defense Weinberger could have marked a, you know, ballot that he thought that George Bush knew? But is that really indictable? Are we going to impeach the president on that evidence? (19921101, CBS_FaceNation)
(138)
(on camera): Welcome back. It’s clear that history is not on the side of any civilian leader in Pakistan, but the fact that Nawaz Sharif was not sentenced to death is being seen by some observers as a compromise (20000407, CNN_ Insight)
From the context of these utterances, expressions such as you know, but look and Welcome back are clearly indicating the speaker is managing the cognitive states between the interlocutor or the listeners. In other words, intersubjective coordination
136
6 The Pragmatics and Discourse Functions of MM
may be the discourse motivation for using the constructions like it’s funny that, it is possible that and It’s clear that, which is partly determined by the interactive features of spoken genre, especially in interviews and other speeches involving the interaction between the speaker and the listeners. To sum up, [It be ADJ that P] construction exhibits certain contextual and generic features, and its use is likely to be motivated by its functions in discourse, especially as a means of intersubjective coordination.
6.5 Summary As revealed in the above qualitive analysis of MM usge in discouse, the distinction between subjectivity and intersubjectivity is demonstrated to better explain the discourse behaviors of [I Vmental that P] construction and [It be ADJ that P] construction on the one hand and [It be ADJ that P] construction on the other hand than the dichotomy of subjective and objective orientation. This distinction of subjectivity and intersubjectivity categorizes these two major types of Metaphor of Modality in terms of interpersonal metafunction. In terms of textual considerations, evidentiality and information flow such as prompting something adversative or contrastive may motivate the use of [It be ADJ that P] construction. [It be ADJ that P] construction exhibits certain generic features, with highest distribution in academic genre but lowest in fiction. There are a number of textual or pragmatic factors motivating the use of [It be ADJ that P] construction in the academic genre: (i) evidentiality or data-basedness; (ii) desubjectificaition or epistemic distance; (iii) disagreement negotiation in an impersonal and professional style; (iv) rhetoric mitigation; and (v) strategic exploitation. In contrast, [I VMental that P] construction is largely used as a discourse strategy, which occurs much more frequently in spoken communication than written genres. It is mainly used to invite or instruct the hearer to see the scene from the speaker’s viewpoint. Moreover, it may weaken or boost the force of the claim, depending on its discourse context and the Mental Verb used. [I be ADJ that P] construction is a variant of [I VMental that P] construction. Its outstanding feature lies in the expressivity of a wider range of affective modality in a highly productive way. The subcategory of Anger is rather unique to this construction, as this subcategory is not found in either [It be ADJ that P] construction or [I VMental that P] construction. Generally, the pragmatic properties of [I VVerbal that P] construction are not very distinct from those of [I VMental that P] construction and therefore can be considered as a variant of the latter. Their differences concern mainly the performativity, which is only a matter of degree. This can be explained by their syntactic similarity and their affinity in semantic structure. In syntax, both have the same syntactic pattern [I Verb that P] with Verbs of different semantic feature. In terms of cognitive structure, the putting onstage of SPEECH ACT space is analogous to that of VIEWPOINT space, and then both serve as CONTENT BASE/VIEWPOINT space in onstage region. The speaker, as the subject of conceptualization, is also put onstage in both constructions.
References
137
To conclude, the global syntactic patterning of the constructions including [It be ADJ that P] and [I VMental that P] appears determined by an interaction of the cognitive process of putting VIEWPOINT space onstage with projecting it from the interpersonal domain to the ideational domain, yet retaining a blending of both functional meanings. We have observed a remarkable systematicity in the syntactic pattern in which the two different expression types occur. The ‘matrix’ clause in both constructions can be characterized as framing clause, from which the complement clause [that P] is enclosed and shaped. The syntax of this paradigm appears molded by an iconic force, that is, a tendency to preserve in surface structure the conceptual relations existing between the chunks of information contained in the utterance, in this case particularly the status of the up-moved VIEWPOINT qualification as a meta-operator over the state of affairs. The choice of two clauses reflects the distinction and separation between the conceptual status and metafunctional status of the modal qualification and the state of affairs.
References Aijmer, K. (1997). I think–An English modal particle. In S. Toril & O. J. Westvik (Eds.), Modality in Germanic languages : Historical and comparativ perspectives (pp. 1–48). Mouton de Gruyter. Bennett, K. (2009). English academic style manuals: A survey. Journal of English for Academic Purposes, 8, 43–54. Bybee, J. L. (1985). Morphology: A study of the relation between meaning and form (Vol. 9). John Benjamins Publishing. De Smet, H., & Verstraete, J.-C. (2006). Coming to terms with subjectivity. Cognitive Linguistics, 17(3), 365–392. Fetzer, A., & Johansson, M. (2010). Cognitive verbs in context: A contras-tive analysis of English and French argumentative discourse. International Journal of Corpus Linguistics, 15(2), 240– 266. https://doi.org/10.1075/ijcl.15.2.05fet Fetzer, A. (2014). I think, I mean and I believe in political discourse: Collocates, functions and distribution. Functions of Language, 1, 67–94. https://doi.org/10.1075/fol.21.1.05fet. Firth, J. R. (1957). Ethnographic analysis and language with reference to Malinowski’s views. Man and Culture: An evaluation of the work of Bronislaw Malinowski, 93–118. Givon, T. (Ed.). (1983). Topic continuity in discourse: A quantitative cross-language study. Benjamins. Haiman, J. (1983). Iconic and economic motivation. Language, 59(4), 781–819. Halliday, M. A. K. (1967). Intonation and grammar in British english. Mouton. Halliday, M. A. K., & Hasan, R. (1976). Cohesion in English. Longman. Halliday, M. A. K., & Matthiessen, C. (2004). An introduction to functional grammar. Hodder Arnold. Halliday, M. A. K., & Matthiessen, C. (2014). Halliday’s introduction to functional grammar. Routledge. Halliday, M. A. K. (1999).The grammatical construction of scientific knowledge: The framing of the english clause. In R. R. Favretti et al. (Eds.), Incommensurability and translation: Kuhnian perspectives on scientific communication and theory change. Edward Elgar. Herriman, J. (2000). Extraposition in English: A sudy of the interaction be-tween the matrix predicate and the type of extraposed clause. English Studies, 6, 582–599. Hooper, J. B. (1975). On assertive predicates. In J. P. Kimball (Ed.), Syntax and semantics (Vol. 4, pp. 91–124). Academic Press.
138
6 The Pragmatics and Discourse Functions of MM
Hunston, S. (2002). Corpora in applied linguistics. Cambridge Univeristy Press. Husserl, E. (1960). Uncovering of the sphere of transcendental being as monadological intersubjectivity. Cartesian Meditations: An Introduction to Phenomenology, 89–151. Hyland, K. (2005). Stance and engagement: A model of interaction in academic discourse. Discourse Studies, 3, 173–192. https://doi.org/10.1177/1461445605050365 Hyland, K. (2008). Genre and academic writing in the disciplines. Language Teaching, 4, 543–562. https://doi.org/10.1017/S0261444808005235 Kaltenbock, G. (2005). It-exposition in English: A functional view. International Journal of Corpus Linguistics, 2, 119–159. Kaltenbock, G. (2004). It-extrapositon and non-extraposition in English: A study of syntax in spoken and written texts. Braumuller. Karkkainen, E. (2003). Epistemic stance in English conversation. A description of its interactional functions, with a focus on ‘I think’. Benjamins. https://doi.org/10.1075/pbns.115 Koutsantoni, D. (2006). Rhetorical strategies in engineering research articles and research theses: Advanced academic literacy and relations of power. Journal of English for Academic Purposes, 5, 19–36. Langacker, R. W. (1990). Subjectification. Cognitive Inguistics, 1, 5–38. Leech, G., & Short, M. (2007). In Style in fiction. Pearson Education Limited. Martin, J. R. (2009). Discourse studies. Continuum Companion to Systemic Functional Linguistics, 154–165. Martin, J. R. & Rose, D. (2008). Genre relations: mapping culture. Equinox Publishing Ltd. McGregor, W. B. (1997). Semiotic grammar. Clarendon Press. Nuyts, J. (2001). Epistemic modality, language, and conceptualization: A cognitive-pragmatic perspective. John Benjamins Publishing Company. Palmer, F. R. (1986). Mood and modality. Cambridge University Press. Portner, P. (2009). Modality. Oxford University Press. Perkins, M. R. (1983). In Modal Expressions in English. ABLEX Publishing Corporation. Swales, J. M., & Swales, J. (1990). Genre analysis: English in academic and research settings. Cambridge University Press. Thomas, J. (1995). Meaning in interaction: An introduction to pragmatics. Longman. Thompson, S. A. & A. Mulac. (1991). A quantitive perspective on the gram-maticalization of epistemic parentheticals in english. In E.C. Traugott & B. Heine (Eds.), Approaches to grammaticalization (Vol. 2, pp. 313–329). John Benjamins. Verhagen, A. (2005). Constructions of Intersubjectivity: Discourse, syntax, and cognition. Oxford University Press. Vandelanotte, L. (2009). In Speech and thought representation in English: A cognitive-functional approach. Mouton de Gruyter.
Chapter 7
Conclusion
This work has been concerned with approaching Metaphor of Modality within an integrated cognitive-functional framework. It has analyzed the cognitive structure, functional properties, as well as syntactic and lexical properties of variant MM constructions. This final chapter summarizes the major findings, discusses the contribution of my work, and points out its limitations and future research as well.
7.1 Summary of Major Findings To sum up, this study has illuminated how and why Metaphor of Modality differs from the ‘congruent’ mode of modality, namely modal adjuncts, in terms of lexical and syntactic properties, cognitive structure, and functional motivations. Approached bottom-up, the major constructions of Metaphor of Modality are found to have idiosyncratic syntactic and lexical properties. (1) The syntagmatic relationship between the projecting modal clause and [that P] is that of FRAMING; (2) There are preferences (constraints) on what types of adjectives can be put into the slot of [It be ADJ that-clause] construction. Based on the token frequencies of its construction types, we find that the chief semantic category is epistemic modality. Then nearly one-fifth of [It be ADJ that-clause] construction in COCA represent evaluative modality and more that 5% represent Affect. The findings support that one of the motivations for Metaphor of Modality is to expand the semantic domains of modality. (3) Likewise, among [I Vmental that P] construction, the most productive category is epistemic modality. Although only fairly productive, the representation of Affect also demonstrates the expansion of semantic domains. (4) In addition to [I Vmental that P] construction, a parallel category of constructions can be abstracted: it is [I Vverbal that P] construction. [I Vverbal that P] construction is considered to be an extended type of Metaphor of Modality because its cognitive structure and syntactic behaviors are topological to metaphorical modes of modality, especially © Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7_7
139
140
7 Conclusion
[I Vmental that P] construction. Analyzing the lexical and syntactic properties of MM not only reveals the idiosyncratic characteristics of its forms, but also how the semantic domains of modality are expanded via the interaction of lexical resources and syntactic frames. More specifically, in accordance with a usage-based model, I have found the lists of all the lexical items that have occurred in the slots of these clausal modal constructions via corpus-linguistic methodology, together with information on their frequency and distribution in different genres. Based on the lists, I generalized the preferences (or restrictions) associated with slots in the constructions. These empirical data are more comprehensive as well as finer-grained than the examples provided in the literature on interpersonal Metaphor of Modality. Furthermore, the difference in meaning between congruent and metaphorical modes of modality has been illuminated by the analysis of their underlying cognitive structure. Modal adjuncts within the clause only cue that a VIEWPOINT space is grounding the predication and held off stage, in other words, the viewpoint structure remains implicit. Metaphorical modes of modality, in contrast, represent the VIEWPOINT space being put onstage, namely, the explicit viewpoint structure. Moreover, we have demonstrated that the distinction between subjectivity and intersubjectivity can better explain the discourse behaviors of [I Vmental that P] construction and [I be ADJ that P] construction on the one hand and [It be ADJ that P] construction on the other hand than the dichotomy of subjective and objective orientation. This distinction of subjectivity and intersubjectivity categorizes these two major types of Metaphor of Modality in terms of interpersonal metafunction. In terms of textual considerations, it seems that Metaphor of Modality also contributes much more textually than modal adjuncts within the clause. It is here that [It be ADJ that P] construction and [I VMental that P] construction exhibit divergent discourse behaviors. On the one hand, evidentiality and information flow such as prompting something adversative or contrastive may motivate the use of [It be ADJ that P] construction. [It be ADJ that P] construction exhibits certain generic features, with the highest distribution in academic genre but lowest in fiction. There are a number of textual or pragmatic factors motivating the use of [It be ADJ that P] construction in the academic genre. In contrast, [I VMental that P] construction is largely used as a discourse strategy, which occurs much more frequently in spoken communication than written genres. It is mainly used to invite or instruct the hearer to see the scene from the speaker’s viewpoint. Moreover, it may weaken or boost the force of the claim, depending on its discourse context and the Mental Verb used. [I be ADJ that P] construction is a variant of [I VMental that P] construction. Its outstanding feature lies in the expressivity of a wider range of affective modality in a highly productive way. To conclude, the global syntactic patterning of the constructions including [It be ADJ that P] and [I VMental that P] is largely motivated by the cognitive process of putting VIEWPOINT space onstage and adding a layer of ideational meaning to their interpersonal meaning. The syntax of this paradigm appears molded by an iconic force, that is, a tendency to preserve in surface structure the conceptual relations existing between the chunks of information contained in the utterance, in
7.2 Implications and Contributions of the Study
141
this case particularly the status of the up-moved VIEWPOINT qualification as a metaoperator over the state of affairs. The choice of two clauses reflects the distinction and separation between the conceptual status and metafunctional status of the modal qualification and the state of affairs.
7.2 Implications and Contributions of the Study In this study, I have proposed a refined Basic Communication Space Network model, on which the cognitive analysis of variant modal expressions is theoretically based. In the refined model, VIEWPOINT Spaces, which involve Epistemic, Affective, and Deontic standpoints, are replacing EPISTEMIC Spaces in the previous model. This refined model can explain a lot more facts about the behaviors of linguistic structures of an interpersonal nature, for instance, the representation of modality in the projecting clause vs. modal adjunct, first-person pronoun in the projecting clause vs. impersonal complementation construction, and the lexical properties of the words in the slots of clausal modal constructions. In addition, the colligation of [I VMental that P] and [It be ADJ that P] in the corpus such as I think that it’s true that challenges SFL’s claim that [I VMental that P] expresses subjective modality and [It be ADJ that P] expresses objective modality. In our analysis, [It be ADJ that P] construction expresses intersubjective viewpoint, which indicates that the view may be shared with a wider group of people, including possibly the speaker and the hearer. A speech situation naturally not just involves a speaker but an interpersonal relationship between speaker and addressee; thus, it is subjective as well as intersubjective in a general sense. Subjectivity and intersubjectivity implicate a continuum, and they are matters of degree, so it is natural and not problematic that [I VMental that P] and [It be ADJ that P] are used together in the same utterance. Furthermore, both constructions, with their finer-grained modal meanings, are devices for the speaker/writer to dynamically modulate the degree of subjectivity and intersubjectivity. With an additional layer of the viewpoint construction, the view toward the proposition is, iconicly, more distancing. My book abandons the dichotomy of subjective and objective orientations to account for variants of modal constructions. Instead, subjectivity and intersubjectivity are applied to account for modality as a viewpoint phenomenon. Speakers can create complex viewpoint structures because they have the cognitive ability to simultaneously evoke a network of mental spaces. On this basis, I have elaborated GROUND, adding VIEWPOINT spaces, which include subject of conceptualization and his/her vantage point and orientation, as well as sharing and negotiation of viewpoint between speaker and addressee. The speaker’s own viewpoint, along with the addressee’s and even others’ viewpoints are always relevant in contributing to how the speaker chooses to express viewpoint linguistically. To a certain extent, my study has enriched our understanding of the mental spaces network involved in communication and the factors affecting the linguistic expression of viewpoint.
142
7 Conclusion
One problem with grammatical metaphor framework of modality is its failure to illuminate the difference in meaning between the congruent and the metaphorical. What differentiates the congruent and metaphorical modes of modality in SFL literature is, after all, the ‘explicit’ versus ‘implicit’ orientation. Our study has clarified what this ‘explicitness’ means. In terms of cognitive structure and process, it refers to putting VIEWPOINT space onstage; in terms of lexicogrammar, it refers to the representation of modality in the projecting clause, framing the proposition; in terms of pragmatic functions, it can refer to the encoding of subjectivity and intersubjectivity in the projecting clause. Therefore, the metaphorical modes of modality are more salient than modality realized by modal adjuncts within the proposition. In terms of methodological approach, I explore Metaphor of Modality based on the more recent usage-based model. My analyses and theoretical claims are based on data from observed actual language use, including single instances, text segments, and frequency data. In this way, this project provides a reliable and empirical record of MM in real use, with detailed frequency and distribution data. The prototypicallity or innovativeness of exemplar constructions can be consulted with ease and clarity, which lead to a higher level of reliability than that of the examples, either invented or searched, offered in most MM literature. My research findings also have implications for how to develop students’ critical thinking. Through treating variant modal expressions as a viewpoint phenomenon, instructors can expose to students how utterances are construed and represented from different viewpoints. Then the instructor can present the refined Basic Communication Space Network model as a starting point for critical thinking. This is to raise students’ awareness of the subjects of conceptualization and VIEWPOINT space underlying opinions and propositions. Gradually, through analysis of variant viewpoint constructions in discourse, students could understand that any linguistic form choice is evidence concerning the relevant cognizer’s viewpoint. This facilitates them in identifying relevant people’s positions and arguments, evaluating the possibility and evidence for alternative points of view, and presenting a point of view in a structured, clear, convincing way. All these help develop students’ critical thinking as both awareness and as a process.
7.3 Limitations and Future Research The present study also has some limitations. Unfortunately, due to time and feasibility constraints, the formulaic use of CT-clauses is not included from the present study. Its unique pragmatic properties and formal features suggest the status of another independent construction, which needs to be treated separately. The formulaic use of a CT-clause does not appear to manifest the framing syntagmatic relationship between itself and its complement clause. Nevertheless, the formulaic use of CTclauses does form a continuum with the constructions studied in this work. Therefore, future work could add this set of construction, which does not formally include a
Reference
143
that-complementizer. The likely findings would offer a broader-spectrum picture of these related linguistic phenomena. Another limitation, due to time limits, lies in that we focus on the modalization type of metaphor but haven’t covered the modulation type of metaphor, such as I want John to go and it’s expected that John goes. The construction types like [it’s expected that] could be considered as the variants of [It be ADJ that P]. The construction types like [I want John to] deserve separate description of their lexical and syntactic properties, although their cognitive structure can also be said to be putting VIEWPOINT space (represented by I want) onstage. In another aspect, inevitably, the lower-level semantic classification of constructions has been based on the researcher’s subjective interpretation of the examples. Although I have referred to relevant literature as comprehensively as possible, and have used online semantic analytic tools, such as FrameNet and VerbNet databases, the classification cannot be said to be free of my intentionality. Consequently, I suspect linguists of a different school might have different views toward the semantic classification. Chen (2014) revisits the phenomenon of ideational metaphor to seek a cognitivepragmatic account, identifying it as a unique way of securing the presumption of optimal relevance and highlighting its specific cognitive effects not available from its non-metaphorical counterpart. My cognitive-functional approach to Metaphor of Modality could be extended with a cognitive-pragmatic account, which could be pursued further by subsequent studies. I have mentioned in the previous section that my refined basic communication space network model can help develop students’ critical thinking as both awareness and as a process. This means we can apply this model to critical reading and writing course and incorporate various constructions of viewpoint into teaching materials. Subsequent work could be an empirical study with an aim to prove this hypothesis, testing whether this application will enhance students’ critical awareness and competence. In a sum, given so far limited functional or cognitive accounts of interpersonal Metaphor of Modality, this book is, hopefully, contributing substantively to this particular area of research. Meanwhile, it provokes more research questions to be sorted, which is part of the reason why the complexity of the modality phenomenon has fascinated generations of linguists. I anticipate, with optimism, that the findings of this work, both theoretical and empirical, will have rich implications for further explorations and applications, particularly in the research fields of modality and mismatch in the architecture of grammar.
Reference Chen, X. 陈新仁. (2014). A pragma-cognitive account of grammatical metaphor语法隐喻的认知 语用解读. Journal of Foreign Languages 外国语, 2, 33–41.
Appendix A
In the following semantic classification of the adjectival predicates of that clauses in [It be ADJ that-clause] in the COCA, the patterns are presented in descending order of their frequency. S. No. Construction type
Semantic tag
Category
1
[IT] [BE] [CLEAR] [THAT]
Obviousness
Epistemic 4700
2
[IT] [BE] [TRUE] [THAT]
Actuality
Epistemic 3218
Frequency
3
[IT] [BE] [POSSIBLE] [THAT]
Possibility
Epistemic 2946
4
[IT] [BE] [IMPORTANT] [THAT]
Valuation
Evaluative 2105
5
[IT] [BE] [LIKELY] [THAT]
Possibility
Epistemic 1586
6
[IT] [BE] [OBVIOUS] [THAT]
Obviousness
Epistemic 1257
7
[IT] [BE] [UNLIKELY] [THAT]
Possibility
Epistemic 1164
8
[IT] [BE] [INTERESTING] [THAT]
Impact
Evaluative
834
9
[IT] [BE] [IMPERATIVE] [THAT]
Necessity
Deontic
543
10
[IT] [BE] [APPARENT] [THAT]
Obviousness
Epistemic
537
11
[IT] [BE] [EVIDENT] [THAT]
Obviousness
Epistemic
487
12
[IT] [BE] [ESSENTIAL] [THAT]
Necessity
Deontic
466
13
[IT] [BE] [UNFORTUNATE] [THAT]
Un/happiness
Affective
435
14
[IT] [BE] [IRONIC] [THAT]
Expectation
Epistemic
372
15
[IT] [BE] [GOOD] [THAT]
Acceptability
Deontic
355
16
[IT] [BE] [INEVITABLE] [THAT]
Expectation
Epistemic
352
17
[IT] [BE] [DOUBTFUL] [THAT]
Certainty
Epistemic
326
18
[IT] [BE] [CRITICAL] [THAT]
Valuation
Evaluative
317
19
[IT] [BE] [SIGNIFICANT] [THAT]
Valuation
Evaluative
290
20
[IT] [BE] [AMAZING] [THAT]
Quality
Evaluative
276
21
[IT] [BE] [CONCEIVABLE] [THAT]
Possibility
Epistemic
267 (continued)
© Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7
145
146
Appendix A
(continued) Semantic tag
Category
22
S. No. Construction type [IT] [BE] [NOTEWORTHY] [THAT]
Valuation
Evaluative
Frequency 250
23
[IT] [BE] [GREAT] [THAT]
Intensifying
Affective
247
24
[IT] [BE] [UNDERSTANDABLE] [THAT]
Expectation
Epistemic
236
25
[IT] [BE] [CRUCIAL] [THAT]
Valuation
Evaluative
213
26
[IT] [BE] [VITAL] [THAT]
Valuation
Evaluative
201
27
[IT] [BE] [PROBABLE] [THAT]
Possibility
Epistemic
199
28
[IT] [BE] [SURPRISING] [THAT]
In/security
Affective
181
29
[IT] [BE] [RARE] [THAT]
Usuality
Epistemic
175
30
[IT] [BE] [REMARKABLE] [THAT]
Valuation
Evaluative
157
31
[IT] [BE] [CERTAIN] [THAT]
Certainty
Epistemic
148
32
[IT] [BE] [SAD] [THAT]
Un/happiness
Affective
145
33
[IT] [BE] [NECESSARY] [THAT]
Necessity
Deontic
131
34
[IT] [BE] [NICE] [THAT]
Quality
Evaluative
127
35
[IT] [BE] [NATURAL] [THAT]
Expectation
Epistemic
113
36
[IT] [BE] [FUNNY] [THAT]
In/security
Affective
112
37
[IT] [BE] [INCONCEIVABLE] [THAT]
Possibility
Epistemic
110
38
[IT] [BE] [NOTABLE] [THAT]
Valuation
Evaluative
107
39
[IT] [BE] [WONDERFUL] [THAT]
Quality
Evaluative
105
40
[IT] [BE] [APPROPRIATE] [THAT]
Composition
Evaluative
104
41
[IT] [BE] [ODD] [THAT]
Usuality
Epistemic
98
42
[IT] [BE] [UNDENIABLE] [THAT]
Certainty
Epistemic
92
43
[IT] [BE] [STRANGE] [THAT]
Usuality
Epistemic
88
44
[IT] [BE] [PLAUSIBLE] [THAT]
Possibility
Epistemic
82
45
[IT] [BE] [BETTER] [THAT]
Quality
Evaluative
77
46
[IT] [BE] [PLAIN] [THAT]
Obviousness
Epistemic
73
47
[IT] [BE] [ARGUABLE] [THAT]
Actuality
Epistemic
68
48
[IT] [BE] [OUTRAGEOUS] [THAT]
Dis/satisfaction Affective
67
49
[IT] [BE] [FORTUNATE] [THAT]
Un/happiness
63
Affective
50
[IT] [BE] [IMPOSSIBLE] [THAT]
Possibility
Epistemic
63
51
[IT] [BE] [CURIOUS] [THAT]
Usuality
Epistemic
58
52
[IT] [BE] [OK/OKAY] [THAT]
Acceptability
Deontic
56
53
[IT] [BE] [LOGICAL] [THAT]
Deduction
Epistemic
54
54
[IT] [BE] [RIDICULOUS] [THAT]
In/security
Affective
53
55
[IT] [BE] [UNFAIR] [THAT]
Composition
Evaluative
51
56
[IT] [BE] [FAIR] [THAT]
Composition
Evaluative
51
57
[IT] [BE] [ENCOURAGING] [THAT]
In/security
Affective
50
58
[IT] [BE] [WELL-KNOWN] [THAT]
Expectation
Epistemic
49 (continued)
Appendix A
147
(continued) Semantic tag
Category
59
S. No. Construction type [IT] [BE] [TRAGIC] [THAT]
Un/happiness
Affective
Frequency 48
60
[IT] [BE] [BEST] [THAT]
Quality
Evaluative
48
61
[IT] [BE] [INCREDIBLE] [THAT]
Actuality
Epistemic
46
62
[IT] [BE] [REGRETTABLE] [THAT]
Dis/satisfaction Affective
41
63
[IT] [BE] [TERRIBLE] [THAT]
In/security
38
Affective
64
[IT] [BE] [UNTHINKABLE] [THAT]
Perception
Epistemic
38
65
[IT] [BE] [FASCINATING] [THAT]
Impact
Evaluative
37
66
[IT] [BE] [DEAR] [THAT]
Valuation
Evaluative
36
67
[IT] [BE] [DISAPPOINTING] [THAT]
Dis/satisfaction Affective
35
68
[IT] [BE] [ASTONISHING] [THAT]
In/security
Affective
35
69
[IT] [BE] [UNCLEAR] [THAT]
Obviousness
Epistemic
34
70
[IT] [BE] [UNUSUAL] [THAT]
Usuality
Epistemic
34
71
[IT] [BE] [WEIRD] [THAT]
Usuality
Epistemic
33
72
[IT] [BE] [URGENT] [THAT]
Obligation
Deontic
30
73
[IT] [BE] [WRONG] [THAT]
Morality
Deontic
30
74
[IT] [BE] [COOL] [THAT]
Quality
Evaluative
30
75
[IT] [BE] [EXTRAORDINARY] [THAT]
Usuality
Epistemic
29
76
[IT] [BE] [AXIOMATIC] [THAT]
Obviousness
Epistemic
28
77
[IT] [BE] [INDISPUTABLE] [THAT]
Certainty
Epistemic
27
78
[IT] [BE] [UNBELIEVABLE] [THAT]
Perception
Epistemic
27
79
[IT] [BE] [SHAMEFUL] [THAT]
Morality
Deontic
25
80
[IT] [BE] [SELF-EVIDENT] [THAT]
Obviousness
Epistemic
24
81
[IT] [BE] [WELL] [THAT]
Quality
Evaluative
24
82
[IT] [BE] [FINE] [THAT]
Acceptability
Deontic
24
83
[IT] [BE] [REASONABLE] [THAT]
Deduction
Epistemic
23
84
[IT] [BE] [INSTRUCTIVE] [THAT]
Quality
Evaluative
21
85
[IT] [BE] [APPALLING] [THAT]
In/security
Affective
21
86
[IT] [BE] [DESIRABLE] [THAT]
Acceptability
Deontic
20
87
[IT] [BE] [IMPROBABLE] [THAT]
Possibility
Epistemic
20
88
[IT] [BE] [INTRIGUING] [THAT]
Impact
Evaluative
20
89
[IT] [BE] [NORMAL] [THAT]
Usuality
Epistemic
20
90
[IT] [BE] [SHOCKING] [THAT]
In/security
Affective
20
91
[IT] [BE] [CORRECT] [THAT]
Actuality
Epistemic
19
92
[IT] [BE] [ABSURD] [THAT]
Dis/satisfaction Affective
19
93
[IT] [BE] [BAD] [THAT]
Acceptability
Deontic
18
94
[IT] [BE] [LUCKY] [THAT]
Un/happiness
Affective
18
95
[IT] [BE] [CRAZY] [THAT]
In/security
Affective
17 (continued)
148
Appendix A
(continued) S. No. Construction type
Semantic tag
Category
Frequency
96
[IT] [BE] [DISTURBING] [THAT]
Dis/satisfaction Affective
17
97
[IT] [BE] [NOTICEABLE] [THAT]
Obviousness
17
Epistemic
98
[IT] [BE] [PREDICTABLE] [THAT]
Expectation
Epistemic
17
99
[IT] [BE] [PARAMOUNT] [THAT]
Valuation
Evaluative
16
100
[IT] [BE] [HORRIBLE] [THAT]
In/security
Affective
16
101
[IT] [BE] [EXCITING] [THAT]
Un/happiness
Affective
15
102
[IT] [BE] [FRUSTRATING] [THAT]
Dis/satisfaction Affective
14
103
[IT] [BE] [LUDICROUS] [THAT]
Dis/satisfaction Affective
14
104
[IT] [BE] [KEY] [THAT]
Valuation
Evaluative
14
105
[IT] [BE] [PARADOXICAL] [THAT]
Deduction
Epistemic
14
106
[IT] [BE] [PROPER] [THAT]
Composition
Evaluative
14
Evaluative
107
[IT] [BE] [RELEVANT] [THAT]
Impact
108
[IT] [BE] [REFRESHING] [THAT]
Dis/satisfaction Affective
14 14
109
[IT] [BE] [UNAVOIDABLE] [THAT]
Certainty
Epistemic
14
110
[IT] [BE] [UNACCEPTABLE] [THAT]
Acceptability
Deontic
14
111
[IT] [BE] [UNCONSCIONABLE] [THAT]
Morality
Deontic
14
112
[IT] [BE] [UNIMAGINABLE] [THAT]
Perception
Epistemic
13
113
[IT] [BE] [SUFFICIENT] [THAT]
Valuation
Evaluative
13
114
[IT] [BE] [GRATIFYING] [THAT]
Dis/satisfaction Affective
13
115
[IT] [BE] [IMPRESSIVE] [THAT]
Impact
Evaluative
12
116
[IT] [BE] [FANTASTIC] [THAT]
Quality
Evaluative
12
117
[IT] [BE] [FEASIBLE] [THAT]
Possibility
Epistemic
12
118
[IT] [BE] [DISTRESSING] [THAT]
Un/happiness
Affective
11
119
[IT] [BE] [NEAT] [THAT]
Quality
Evaluative
11
120
[IT] [BE] [MANDATORY] [THAT]
Obligation
Deontic
11
121
[IT] [BE] [PATHETIC] [THAT]
Quality
Evaluative
10
122
[IT] [BE] [STUNNING] [THAT]
Expectation
Affective
10
123
[IT] [BE] [TYPICAL] [THAT]
Usuality
Epistemic
10
124
[IT] [BE] [IRRELEVANT] [THAT]
Impact
125
[IT] [BE] [INCOMPREHENSIBLE] [THAT] Composition
Evaluative
10
Evaluative
10
Appendix B
The distribution of [It be ADJ that-clause] in Academic genre versus Fiction in COCA (Only those with a ratio above 1.0 is reproduced here) SEC 1: Academic (vs. Fiction), 91,066,191 WORDS
1
Word/ Phrase
Tokens 1
Tokens 2
PM 1
PM 2
Ratio
[IT] [BE] [NOTEWORTHY] [THAT]
217
0
2.38
0.00
238 79.06
2
[IT] [BE] [NOTABLE] [THAT]
72
0
0.79
0.00
3
[IT] [BE] [PLAUSIBLE] [THAT]
61
0
0.67
0.00
66.98
4
[IT] [BE] [ARGUABLE] [THAT]
44
1
0.48
0.01
43.69
5
[IT] [BE] [SIGNIFICANT] [THAT]
195
5
2.14
0.06
38.73
6
[IT] [BE] [ENCOURAGING] [THAT]
26
1
0.29
0.01
25.82
7
[IT] [BE] [CRITICAL] [THAT]
167
7
1.83
0.08
23.69 18.67
8
[IT] [BE] [DESIRABLE] [THAT]
17
0
0.19
0.00
9
[IT] [BE] [LIKELY] [THAT]
948
51
10.41
0.56
18.46
10
[IT] [BE] [REGRETTABLE] [THAT]
15
0
0.16
0.00
16.47
11
[IT] [BE] [REASONABLE] [THAT]
16
1
0.18
0.01
15.89
12
[IT] [BE] [IMPROBABLE] [THAT]
15
1
0.16
0.01
14.90
13
[IT] [BE] [INSTRUCTIVE] [THAT]
13
0
0.14
0.00
14.28
14
[IT] [BE] [PROBABLE] [THAT]
138
10
1.52
0.11
13.70
15
[IT] [BE] [PARADOXICAL] [THAT]
12
0
0.13
0.00
13.18
16
[IT] [BE] [INTRIGUING] [THAT]
12
0
0.13
0.00
13.18
17
[IT] [BE] [UNCLEAR] [THAT]
12
0
0.13
0.00
13.18
18
[IT] [BE] [UNDENIABLE] [THAT]
52
4
0.57
0.04
12.91
19
[IT] [BE] [SUFFICIENT] [THAT]
12
1
0.13
0.01
11.92
20
[IT] [BE] [EVIDENT] [THAT]
341
30
3.74
0.33
11.29
21
[IT] [BE] [PARAMOUNT] [THAT]
10
0
0.11
0.00
10.98
© Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7
(continued) 149
150
Appendix B
(continued) SEC 1: Academic (vs. Fiction), 91,066,191 WORDS
22
Word/ Phrase
Tokens 1
Tokens 2
[IT] [BE] [DEAR] [THAT]
22
2
PM 1
PM 2
Ratio
0.24
0.02
10.92
23
[IT] [BE] [ESSENTIAL] [THAT]
261
24
2.87
0.27
10.80
24
[IT] [BE] [REMARKABLE] [THAT]
63
6
0.69
0.07
10.43
25
[IT] [BE] [INTERESTING] [THAT]
218
21
2.39
0.23
10.31
26
[IT] [BE] [CONCEIVABLE] [THAT]
132
13
1.45
0.14
10.08
27
[IT] [BE] [IMPERATIVE] [THAT]
310
31
3.40
0.34
9.93
28
[IT] [BE] [PREDICTABLE] [THAT]
10
1
0.11
0.01
9.93
29
[IT] [BE] [INDISPUTABLE] [THAT]
9
0
0.10
0.00
9.88
30
[IT] [BE] [APPARENT] [THAT]
311
32
3.42
0.35
9.65
31
[IT] [BE] [LOGICAL] [THAT]
29
3
0.32
0.03
9.60
32
[IT] [BE] [DOUBTFUL] [THAT]
149
16
1.64
0.18
9.25
33
[IT] [BE] [UNLIKELY] [THAT]
537
59
5.90
0.65
9.04
34
[IT] [BE] [FEASIBLE] [THAT]
8
0
0.09
0.00
8.78
35
[IT] [BE] [CRUCIAL] [THAT]
88
11
0.97
0.12
7.94
36
[IT] [BE] [RELEVANT] [THAT]
8
1
0.09
0.01
7.94
37
[IT] [BE] [SELF-EVIDENT] [THAT]
8
1
0.09
0.01
7.94
38
[IT] [BE] [SURPRISING] [THAT]
86
11
0.94
0.12
7.76
39
[IT] [BE] [DISTURBING] [THAT]
7
1
0.08
0.01
6.95
40
[IT] [BE] [UNCERTAIN] [THAT]
6
0
0.07
0.00
6.59
41
[IT] [BE] [NECESSARY] [THAT]
87
14
0.96
0.15
6.17
42
[IT] [BE] [VITAL] [THAT]
86
14
0.94
0.15
6.10
43
[IT] [BE] [POSSIBLE] [THAT]
1557
259
17.10
2.86
5.97
44
[IT] [BE] [CURIOUS] [THAT]
30
5
0.33
0.06
5.96
45
[IT] [BE] [PROPER] [THAT]
6
1
0.07
0.01
5.96
46
[IT] [BE] [UNAVOIDABLE] [THAT]
6
1
0.07
0.01
5.96
47
[IT] [BE] [CONCERNED] [THAT]
5
0
0.05
0.00
5.49
48
[IT] [BE] [INDICATIVE] [THAT]
5
0
0.05
0.00
5.49
49
[IT] [BE] [UNTRUE] [THAT]
5
0
0.05
0.00
5.49
688
5.30
50
[IT] [BE] [IMPORTANT] [THAT]
129
7.55
1.43
51
[IT] [BE] [DISAPPOINTING] [THAT] 5
1
0.05
0.01
4.97
52
[IT] [BE] [IRONIC] [THAT]
21
1.09
0.23
4.68
99
53
[IT] [BE] [PREFERABLE] [THAT]
4
0
0.04
0.00
4.39
54
[IT] [BE] [PROMISING] [THAT]
4
0
0.04
0.00
4.39
55
[IT] [BE] [CONFIDENT] [THAT]
4
0
0.04
0.00
4.39
56
[IT] [BE] [COMMENDABLE] [THAT] 4
0
0.04
0.00
4.39
(continued)
Appendix B
151
(continued) SEC 1: Academic (vs. Fiction), 91,066,191 WORDS Word/ Phrase
Tokens 1
Tokens 2
57
[IT] [BE] [CHARACTERISTIC] [THAT]
4
58
[IT] [BE] [INCONTROVERTIBLE] [THAT]
59 60
PM 1
PM 2
Ratio
0
0.04
0.00
4.39
4
0
0.04
0.00
4.39
[IT] [BE] [FORESEEABLE] [THAT]
4
0
0.04
0.00
4.39
[IT] [BE] [WELL-ESTABLISHED] [THAT]
4
0
0.04
0.00
4.39
61
[IT] [BE] [AXIOMATIC] [THAT]
13
3
62
[IT] [BE] [CLEAR] [THAT]
1812
449
63
[IT] [BE] [UNEXPECTED] [THAT]
4
1
0.14
0.03
4.30
19.90
4.97
4.01
0.04
0.01
3.97
64
[IT] [BE] [NOTICEABLE] [THAT]
10
3
0.11
0.03
3.31
65
[IT] [BE] [MANDATORY] [THAT]
3
0
0.03
0.00
3.29
66
[IT] [BE] [UNDISPUTED] [THAT]
3
0
0.03
0.00
3.29
67
[IT] [BE] [QUESTIONABLE] [THAT] 3
0
0.03
0.00
3.29
68
[IT] [BE] [AWARE] [THAT]
3
0
0.03
0.00
3.29
69
[IT] [BE] [ALARMING] [THAT]
3
0
0.03
0.00
3.29
70
[IT] [BE] [DISTRESSING] [THAT]
3
0
0.03
0.00
3.29
71
[IT] [BE] [IMPLAUSIBLE] [THAT]
3
0
0.03
0.00
3.29
72
[IT] [BE] [IMPRESSIVE] [THAT]
3
0
0.03
0.00
3.29
73
[IT] [BE] [WELL-RECOGNIZED] [THAT]
3
0
0.03
0.00
3.29
74
[IT] [BE] [CERTAIN] [THAT]
70
23
0.77
0.25
3.02
75
[IT] [BE] [APPROPRIATE] [THAT]
21
7
0.23
0.08
2.98
76
[IT] [BE] [FASCINATING] [THAT]
3
1
0.03
0.01
2.98
77
[IT] [BE] [KEY] [THAT]
3
1
0.03
0.01
2.98
78
[IT] [BE] [WELL-DOCUMENTED] [THAT]
3
1
0.03
0.01
2.98
79
[IT] [BE] [UNDERSTANDABLE] [THAT]
71
24
0.78
0.27
2.94
80
[IT] [BE] [WELL-KNOWN] [THAT]
20
8
0.22
0.09
2.48
81
[IT] [BE] [UNFORTUNATE] [THAT]
54
24
0.59
0.27
2.23
82
[IT] [BE] [SUGGESTIVE] [THAT]
2
0
0.02
0.00
2.20
83
[IT] [BE] [SYMPTOMATIC] [THAT]
2
0
0.02
0.00
2.20
84
[IT] [BE] [TENABLE] [THAT]
2
0
0.02
0.00
2.20
85
[IT] [BE] [REASONED] [THAT]
2
0
0.02
0.00
2.20
86
[IT] [BE] [SHOCKING] [THAT]
2
0
0.02
0.00
2.20
87
[IT] [BE] [LAUDABLE] [THAT]
2
0
0.02
0.00
2.20
(continued)
152
Appendix B
(continued) SEC 1: Academic (vs. Fiction), 91,066,191 WORDS Word/ Phrase
Tokens 1
Tokens 2
PM 1
PM 2
Ratio
88
[IT] [BE] [INTUITIVE] [THAT]
2
0
0.02
0.00
2.20
89
[IT] [BE] [MANIFEST] [THAT]
2
0
0.02
0.00
2.20
90
[IT] [BE] [NOTORIOUS] [THAT]
2
0
0.02
0.00
2.20
91
[IT] [BE] [PERPLEXING] [THAT]
2
0
0.02
0.00
2.20
92
[IT] [BE] [INCONSISTENT] [THAT]
2
0
0.02
0.00
2.20
93
[IT] [BE] [INESCAPABLE] [THAT]
2
0
0.02
0.00
2.20
94
[IT] [BE] [CONSPICUOUS] [THAT]
2
0
0.02
0.00
2.20
95
[IT] [BE] [DUBIOUS] [THAT]
2
0
0.02
0.00
2.20
96
[IT] [BE] [DECISIVE] [THAT]
2
0
0.02
0.00
2.20
97
[IT] [BE] [UNMISTAKABLE] [THAT] 2
0
0.02
0.00
2.20
98
[IT] [BE] [UNIMAGINABLE] [THAT] 4
2
0.04
0.02
1.99
99
[IT] [BE] [WORRISOME] [THAT]
4
2
0.04
0.02
1.99
100
[IT] [BE] [GRATIFYING] [THAT]
4
2
0.04
0.02
1.99
101
[IT] [BE] [EXCITING] [THAT]
2
1
0.02
0.01
1.99
102
[IT] [BE] [DEFINITE] [THAT]
2
1
0.02
0.01
1.99
103
[IT] [BE] [COMMON] [THAT]
2
1
0.02
0.01
1.99
104
[IT] [BE] [CORRECT] [THAT]
2
1
0.02
0.01
1.99
105
[IT] [BE] [AMUSING] [THAT]
2
1
0.02
0.01
1.99
106
[IT] [BE] [INEVITABLE] [THAT]
81
48
0.89
0.53
1.68
107
[IT] [BE] [TRUE] [THAT]
787
518
8.64
5.73
1.51
108
[IT] [BE] [ASTONISHING] [THAT]
6
4
0.07
0.04
1.49
109
[IT] [BE] [UNQUESTIONABLE] [THAT]
3
2
0.03
0.02
1.49
110
[IT] [BE] [NATURAL] [THAT]
31
21
0.34
0.23
1.47
111
[IT] [BE] [FAIR] [THAT]
5
4
0.05
0.04
1.24
112
[IT] [BE] [INCONCEIVABLE] [THAT]
23
20
0.25
0.22
1.14
113
[IT] [BE] [INCONTESTABLE] [THAT]
1
0
0.01
0.00
1.10
114
[IT] [BE] [INARGUABLE] [THAT]
1
0
0.01
0.00
1.10 1.10
115
[IT] [BE] [INFERABLE] [THAT]
1
0
0.01
0.00
116
[IT] [BE] [INFORMATIVE] [THAT]
1
0
0.01
0.00
1.10
117
[IT] [BE] [INDISPENSABLE] [THAT] 1
0
0.01
0.00
1.10
118
[IT] [BE] [INDUBITABLE] [THAT]
1
0
0.01
0.00
1.10
119
[IT] [BE] [FEARFUL] [THAT]
1
0
0.01
0.00
1.10
120
[IT] [BE] [INADMISSIBLE] [THAT]
1
0
0.01
0.00
1.10
(continued)
Appendix B
153
(continued) SEC 1: Academic (vs. Fiction), 91,066,191 WORDS Word/ Phrase
Tokens 1
Tokens 2
PM 1
PM 2
Ratio
121
[IT] [BE] [HOPEFUL] [THAT]
1
0
0.01
0.00
1.10
122
[IT] [BE] [HYPOTHESIZE] [THAT]
1
0
0.01
0.00
1.10
123
[IT] [BE] [IMMATERIAL] [THAT]
1
0
0.01
0.00
1.10
124
[IT] [BE] [IMMINENT] [THAT]
1
0
0.01
0.00
1.10
125
[IT] [BE] [ENGLISH] [THAT]
1
0
0.01
0.00
1.10 1.10
126
[IT] [BE] [FANTASTIC] [THAT]
1
0
0.01
0.00
127
[IT] [BE] [EXPLETIVE] [THAT]
1
0
0.01
0.00
1.10
128
[IT] [BE] [EXPLICIT] [THAT]
1
0
0.01
0.00
1.10
129
[IT] [BE] [DISGUSTING] [THAT]
1
0
0.01
0.00
1.10
130
[IT] [BE] [FUNDAMENTAL] [THAT]
1
0
0.01
0.00
1.10
131
[IT] [BE] [HEARTENING] [THAT]
1
0
0.01
0.00
1.10
132
[IT] [BE] [HELPFUL] [THAT]
1
0
0.01
0.00
1.10
133
[IT] [BE] [FELL] [THAT]
1
0
0.01
0.00
1.10
134
[IT] [BE] [APT] [THAT]
1
0
0.01
0.00
1.10
135
[IT] [BE] [AMOUR-PROPRE] [THAT] 1
0
0.01
0.00
1.10
136
[IT] [BE] [ALTRUISTIC] [THAT]
1
0
0.01
0.00
1.10
137
[IT] [BE] [ACCIDENTAL] [THAT]
1
0
0.01
0.00
1.10
138
[IT] [BE] [COMFORTING] [THAT]
1
0
0.01
0.00
1.10
139
[IT] [BE] [CHILLING] [THAT]
1
0
0.01
0.00
1.10
140
[IT] [BE] [COUNTERINTUITIVE] [THAT]
1
0
0.01
0.00
1.10
141
[IT] [BE] [CREDIBLE] [THAT]
1
0
0.01
0.00
1.10
142
[IT] [BE] [CONVINCED] [THAT]
1
0
0.01
0.00
1.10
143
[IT] [BE] [DAUNTING] [THAT]
1
0
0.01
0.00
1.10
144
[IT] [BE] [DEPRESSING] [THAT]
1
0
0.01
0.00
1.10
145
[IT] [BE] [DEBATABLE] [THAT]
1
0
0.01
0.00
1.10
146
[IT] [BE] [MELANCHOLY] [THAT]
1
0
0.01
0.00
1.10
147
[IT] [BE] [LIGHT] [THAT]
1
0
0.01
0.00
1.10
148
[IT] [BE] [IRRATIONAL] [THAT]
1
0
0.01
0.00
1.10
149
[IT] [BE] [INSUFFICIENT] [THAT]
1
0
0.01
0.00
1.10
150
[IT] [BE] [INTEGRAL] [THAT]
1
0
0.01
0.00
1.10
151
[IT] [BE] [PERTINENT] [THAT]
1
0
0.01
0.00
1.10
152
[IT] [BE] [PHENOMENAL] [THAT]
1
0
0.01
0.00
1.10
153
[IT] [BE] [NOTE-WORTHY] [THAT]
1
0
0.01
0.00
1.10
154
[IT] [BE] [OBJECTIONABLE] [THAT]
1
0
0.01
0.00
1.10
(continued)
154
Appendix B
(continued) SEC 1: Academic (vs. Fiction), 91,066,191 WORDS Word/ Phrase
Tokens 1
Tokens 2
PM 1
PM 2
Ratio
155
[IT] [BE] [PRESUMABLE] [THAT]
1
0
0.01
0.00
1.10
156
[IT] [BE] [PROBLEMATIC] [THAT]
1
0
0.01
0.00
1.10
157
[IT] [BE] [QUADRUPLE] [THAT]
1
0
0.01
0.00
1.10
158
[IT] [BE] [TROUBLESOME] [THAT]
1
0
0.01
0.00
1.10
159
[IT] [BE] [THOUGHT-PROVOKING] [THAT]
1
0
0.01
0.00
1.10
160
[IT] [BE] [TIMELY] [THAT]
1
0
0.01
0.00
1.10
161
[IT] [BE] [SOBERING] [THAT]
1
0
0.01
0.00
1.10
162
[IT] [BE] [UNANIMOUS] [THAT]
1
0
0.01
0.00
1.10
163
[IT] [BE] [UNCANNY] [THAT]
1
0
0.01
0.00
1.10
164
[IT] [BE] [STRAIGHTFORWARD] [THAT]
1
0
0.01
0.00
1.10
165
[IT] [BE] [SALIENT] [THAT]
1
0
0.01
0.00
1.10
166
[IT] [BE] [SECURE] [THAT]
1
0
0.01
0.00
1.10
167
[IT] [BE] [SELF-DECEPTION] [THAT]
1
0
0.01
0.00
1.10
168
[IT] [BE] [RED] [THAT]
1
0
0.01
0.00
1.10
169
[IT] [BE] [REFRESHING] [THAT]
1
0
0.01
0.00
1.10
170
[IT] [BE] [REALISTIC] [THAT]
1
0
0.01
0.00
1.10
171
[IT] [BE] [REALIZABLE] [THAT]
1
0
0.01
0.00
1.10
172
[IT] [BE] [REQUISITE] [THAT]
1
0
0.01
0.00
1.10
173
[IT] [BE] [UNQUESTIONED] [THAT] 1
0
0.01
0.00
1.10
174
[IT] [BE] [UNREMARKABLE] [THAT]
1
0
0.01
0.00
1.10
175
[IT] [BE] [USEFUL] [THAT]
1
0
0.01
0.00
1.10
176
[IT] [BE] [WELL-UNDERSTOOD] [THAT]
1
0
0.01
0.00
1.10
177
[IT] [BE] [OBVIOUS] [THAT]
330
304
3.62
3.36
1.08
TOKENS 1
PM 2
SEC 2: Fiction (vs. Academic), 90,429,400 WORDS WORD/ PHRASE
TOKENS 2
PM 1
Ratio
1
[IT] [BE] [WEIRD] [THAT]
7
0
0.08
0.00
7.74
2
[IT] [BE] [COOL] [THAT]
7
1
0.08
0.01
7.05
3
[IT] [BE] [FINE] [THAT]
6
0
0.07
0.00
6.64
4
[IT] [BE] [BETTER] [THAT]
32
6
0.35
0.07
5.37
(continued)
Appendix B
155
(continued) SEC 2: Fiction (vs. Academic), 90,429,400 WORDS WORD/ PHRASE
TOKENS 2
TOKENS 1
PM 2
PM 1
Ratio
5
[IT] [BE] [GREAT] [THAT]
16
3
0.18
0.03
5.37
6
[IT] [BE] [STRANGE] [THAT]
37
7
0.41
0.08
5.32
7
[IT] [BE] [WELL] [THAT]
10
2
0.11
0.02
5.04
8
[IT] [BE] [INCREDIBLE] [THAT]
5
1
0.06
0.01
5.04
9
[IT] [BE] [LUCKY] [THAT]
9
2
0.10
0.02
4.53
10
[IT] [BE] [RIDICULOUS] [THAT]
4
0
0.04
0.00
4.42
11
[IT] [BE] [INCOMPREHENSIBLE] [THAT]
4
0
0.04
0.00
4.42
12
[IT] [BE] [HOT] [THAT]
4
0
0.04
0.00
4.42
13
[IT] [BE] [SAD] [THAT]
13
3
0.14
0.03
4.36
14
[IT] [BE] [GOOD] [THAT]
90
21
1.00
0.23
4.32
15
[IT] [BE] [NICE] [THAT]
21
5
0.23
0.05
4.23
16
[IT] [BE] [ABSURD] [THAT]
4
1
0.04
0.01
4.03
17
[IT] [BE] [BEST] [THAT]
24
6
0.27
0.07
4.03
18
[IT] [BE] [FUNNY] [THAT]
17
5
0.19
0.05
3.42
19
[IT] [BE] [ODD] [THAT]
30
9
0.33
0.10
3.36
20
[IT] [BE] [OUTRAGEOUS] [THAT] 3
0
0.03
0.00
3.32
21
[IT] [BE] [SILLY] [THAT]
3
0
0.03
0.00
3.32
22
[IT] [BE] [SWEET] [THAT]
3
0
0.03
0.00
3.32
23
[IT] [BE] [CHEAP] [THAT]
3
0
0.03
0.00
3.32
24
[IT] [BE] [BAD] [THAT]
3
0
0.03
0.00
3.32
25
[IT] [BE] [AWFUL] [THAT]
3
0
0.03
0.00
3.32
26
[IT] [BE] [CONVENIENT] [THAT]
3
0
0.03
0.00
3.32
27
[IT] [BE] [CRAZY] [THAT]
3
0
0.03
0.00
3.32
28
[IT] [BE] [TERRIBLE] [THAT]
3
1
0.03
0.01
3.02
29
[IT] [BE] [SHAMEFUL] [THAT]
3
1
0.03
0.01
3.02
30
[IT] [BE] [WRONG] [THAT]
3
1
0.03
0.01
3.02
31
[IT] [BE] [OKAY] [THAT]
8
3
0.09
0.03
2.69
32
[IT] [BE] [UNFAIR] [THAT]
9
4
0.10
0.04
2.27
33
[IT] [BE] [PERFECT] [THAT]
2
0
0.02
0.00
2.21
34
[IT] [BE] [OK] [THAT]
2
0
0.02
0.00
2.21
35
[IT] [BE] [OBLIGATORY] [THAT]
2
0
0.02
0.00
2.21
36
[IT] [BE] [SURE] [THAT]
2
0
0.02
0.00
2.21
37
[IT] [BE] [CRIMINAL] [THAT]
2
0
0.02
0.00
2.21
38
[IT] [BE] [DISGRACEFUL] [THAT] 2
0
0.02
0.00
2.21
39
[IT] [BE] [ACCEPTABLE] [THAT]
0
0.02
0.00
2.21
2
(continued)
156
Appendix B
(continued) SEC 2: Fiction (vs. Academic), 90,429,400 WORDS WORD/ PHRASE
TOKENS 2
TOKENS 1
PM 2
PM 1
Ratio
40
[IT] [BE] [HORRIBLE] [THAT]
2
0
0.02
0.00
2.21
41
[IT] [BE] [FRENCH] [THAT]
2
0
0.02
0.00
2.21
42
[IT] [BE] [EMPTY] [THAT]
2
0
0.02
0.00
2.21
43
[IT] [BE] [WONDERFUL] [THAT]
10
5
0.11
0.05
2.01
44
[IT] [BE] [TYPICAL] [THAT]
4
2
0.04
0.02
2.01
45
[IT] [BE] [UNBELIEVABLE] [THAT]
4
2
0.04
0.02
2.01
46
[IT] [BE] [PATHETIC] [THAT]
2
1
0.02
0.01
2.01
47
[IT] [BE] [PRUDENT] [THAT]
2
1
0.02
0.01
2.01
48
[IT] [BE] [FORTUITOUS] [THAT]
2
1
0.02
0.01
2.01
49
[IT] [BE] [INCUMBENT] [THAT]
2
1
0.02
0.01
2.01
50
[IT] [BE] [INEXPLICABLE] [THAT]
2
1
0.02
0.01
2.01
51
[IT] [BE] [AMAZING] [THAT]
32
18
0.35
0.20
1.79
52
[IT] [BE] [UNUSUAL] [THAT]
5
3
0.06
0.03
1.68
53
[IT] [BE] [PLAIN] [THAT]
21
13
0.23
0.14
1.63
54
[IT] [BE] [RARE] [THAT]
29
19
0.32
0.21
1.54
55
[IT] [BE] [URGENT] [THAT]
8
6
0.09
0.07
1.34
56
[IT] [BE] [IMPOSSIBLE] [THAT]
20
16
0.22
0.18
1.26
57
[IT] [BE] [UNTHINKABLE] [THAT]
8
7
0.09
0.08
1.15
58
[IT] [BE] [FORTUNATE] [THAT]
18
16
0.20
0.18
1.13
59
[IT] [BE] [FAT] [THAT]
1
0
0.01
0.00
1.11
60
[IT] [BE] [FULL] [THAT]
1
0
0.01
0.00
1.11
61
[IT] [BE] [HANDY] [THAT]
1
0
0.01
0.00
1.11
62
[IT] [BE] [HEALTHY] [THAT]
1
0
0.01
0.00
1.11
63
[IT] [BE] [HEART-RENDING] [THAT]
1
0
0.01
0.00
1.11
64
[IT] [BE] [EERIE] [THAT]
1
0
0.01
0.00
1.11
65
[IT] [BE] [EMBARRASSING] [THAT]
1
0
0.01
0.00
1.11
66
[IT] [BE] [ENDEARING] [THAT]
1
0
0.01
0.00
1.11
67
[IT] [BE] [EZRY] [THAT]
1
0
0.01
0.00
1.11
68
[IT] [BE] [FALSE] [THAT]
1
0
0.01
0.00
1.11
69
[IT] [BE] [IMPLICIT] [THAT]
1
0
0.01
0.00
1.11
70
[IT] [BE] [HUMID] [THAT]
1
0
0.01
0.00
1.11
71
[IT] [BE] [HILARIOUS] [THAT]
1
0
0.01
0.00
1.11
(continued)
Appendix B
157
(continued) SEC 2: Fiction (vs. Academic), 90,429,400 WORDS WORD/ PHRASE
TOKENS 2
TOKENS 1
PM 2
PM 1
Ratio
72
[IT] [BE] [INCONVENIENT] [THAT]
1
0
0.01
0.00
1.11
73
[IT] [BE] [INAPPROPRIATE] [THAT]
1
0
0.01
0.00
1.11
74
[IT] [BE] [ANNOYING] [THAT]
1
0
0.01
0.00
1.11
75
[IT] [BE] [ALONE] [THAT]
1
0
0.01
0.00
1.11
76
[IT] [BE] [ACCU-READ] [THAT]
1
0
0.01
0.00
1.11
77
[IT] [BE] [ADMIRABLE] [THAT]
1
0
0.01
0.00
1.11
78
[IT] [BE] [BARBARIC] [THAT]
1
0
0.01
0.00
1.11
79
[IT] [BE] [BEAUTIFUL] [THAT]
1
0
0.01
0.00
1.11
80
[IT] [BE] [CLEAN] [THAT]
1
0
0.01
0.00
1.11
81
[IT] [BE] [COINCIDENTAL] [THAT]
1
0
0.01
0.00
1.11
82
[IT] [BE] [DESPERATE] [THAT]
1
0
0.01
0.00
1.11
83
[IT] [BE] [DIFFERENT] [THAT]
1
0
0.01
0.00
1.11
84
[IT] [BE] [DISCONCERTING] [THAT]
1
0
0.01
0.00
1.11
85
[IT] [BE] [CUSTOMARY] [THAT]
1
0
0.01
0.00
1.11
86
[IT] [BE] [CUTE] [THAT]
1
0
0.01
0.00
1.11
87
[IT] [BE] [DARLEEN] [THAT]
1
0
0.01
0.00
1.11
88
[IT] [BE] [DEMONSTRABLE] [THAT]
1
0
0.01
0.00
1.11
89
[IT] [BE] [CREEPY] [THAT]
1
0
0.01
0.00
1.11
90
[IT] [BE] [CRUEL] [THAT]
1
0
0.01
0.00
1.11
91
[IT] [BE] [REAL] [THAT]
1
0
0.01
0.00
1.11
92
[IT] [BE] [QUEER] [THAT]
1
0
0.01
0.00
1.11
93
[IT] [BE] [SIMPLE] [THAT]
1
0
0.01
0.00
1.11
94
[IT] [BE] [STUNNING] [THAT]
1
0
0.01
0.00
1.11
95
[IT] [BE] [STUPID] [THAT]
1
0
0.01
0.00
1.11
96
[IT] [BE] [RISKY] [THAT]
1
0
0.01
0.00
1.11
97
[IT] [BE] [QUIET] [THAT]
1
0
0.01
0.00
1.11
98
[IT] [BE] [SELF-SLAUGHTER] [THAT]
1
0
0.01
0.00
1.11
99
[IT] [BE] [SENSIBLE] [THAT]
1
0
0.01
0.00
1.11
100
[IT] [BE] [SERENDIPITOUS] [THAT]
1
0
0.01
0.00
1.11
101
[IT] [BE] [SERIOUS] [THAT]
1
0
0.01
0.00
1.11
(continued)
158
Appendix B
(continued) SEC 2: Fiction (vs. Academic), 90,429,400 WORDS
102
WORD/ PHRASE
TOKENS 2
TOKENS 1
PM 2
PM 1
Ratio
[IT] [BE] [UN-ARGUABLE] [THAT]
1
0
0.01
0.00
1.11
103
[IT] [BE] [UNBEARABLE] [THAT] 1
0
0.01
0.00
1.11
104
[IT] [BE] [UNCONSCIONABLE] [THAT]
1
0
0.01
0.00
1.11
105
[IT] [BE] [UNENDURABLE] [THAT]
1
0
0.01
0.00
1.11
106
[IT] [BE] [SUSPICIOUS] [THAT]
1
0
0.01
0.00
1.11
107
[IT] [BE] [SOUND] [THAT]
1
0
0.01
0.00
1.11
108
[IT] [BE] [STEEP] [THAT]
1
0
0.01
0.00
1.11
109
[IT] [BE] [SUPER] [THAT]
1
0
0.01
0.00
1.11
110
[IT] [BE] [POINTLESS] [THAT]
1
0
0.01
0.00
1.11
111
[IT] [BE] [PREDICTIVE] [THAT]
1
0
0.01
0.00
1.11
112
[IT] [BE] [PITIFUL] [THAT]
1
0
0.01
0.00
1.11
113
[IT] [BE] [PROFITABLE] [THAT]
1
0
0.01
0.00
1.11
114
[IT] [BE] [NUGUGI] [THAT]
1
0
0.01
0.00
1.11
115
[IT] [BE] [PAST] [THAT]
1
0
0.01
0.00
1.11
116
[IT] [BE] [PATENT] [THAT]
1
0
0.01
0.00
1.11
117
[IT] [BE] [MARVELOUS] [THAT]
1
0
0.01
0.00
1.11
118
[IT] [BE] [LOVELY] [THAT]
1
0
0.01
0.00
1.11
119
[IT] [BE] [MAGIC] [THAT]
1
0
0.01
0.00
1.11
120
[IT] [BE] [MIRACULOUS] [THAT] 1
0
0.01
0.00
1.11
121
[IT] [BE] [MOONLESS] [THAT]
1
0
0.01
0.00
1.11
122
[IT] [BE] [INSANE] []
1
0
0.01
0.00
1.11
123
[IT] [BE] [INSUPPORTABLE] [THAT]
1
0
0.01
0.00
1.11
124
[IT] [BE] [JEALOUS] [THAT]
1
0
0.01
0.00
1.11
125
[IT] [BE] [JUST-JUST] [THAT]
1
0
0.01
0.00
1.11
126
[IT] [BE] [LIQUID] [THAT]
1
0
0.01
0.00
1.11
127
[IT] [BE] [UNHEARD-OF] [THAT]
1
0
0.01
0.00
1.11
128
[IT] [BE] [UNHOLY] [THAT]
1
0
0.01
0.00
1.11
129
[IT] [BE] [WORRIED] [THAT]
1
0
0.01
0.00
1.11
130
[IT] [BE] [WARM] [THAT]
1
0
0.01
0.00
1.11
131
[IT] [BE] [UNIMPORTANT] [THAT]
1
1
0.01
0.01
1.01
132
[IT] [BE] [IRRELEVANT] [THAT]
1
1
0.01
0.01
1.01
133
[IT] [BE] [LUDICROUS] [THAT]
1
1
0.01
0.01
1.01
(continued)
Appendix B
159
(continued) SEC 2: Fiction (vs. Academic), 90,429,400 WORDS WORD/ PHRASE
TOKENS 2
TOKENS 1
PM 2
PM 1
Ratio
134
[IT] [BE] [SYMBOLIC] [THAT]
1
1
0.01
0.01
1.01
135
[IT] [BE] [TRAGIC] [THAT]
1
1
0.01
0.01
1.01
136
[IT] [BE] [UNACCEPTABLE] [THAT]
1
1
0.01
0.01
1.01
137
[IT] [BE] [IMAGINABLE] [THAT]
1
1
0.01
0.01
1.01
138
[IT] [BE] [EXTRAORDINARY] [THAT]
1
1
0.01
0.01
1.01
139
[IT] [BE] [FRIGHTENING] [THAT] 1
1
0.01
0.01
1.01
140
[IT] [BE] [ADVISABLE] [THAT]
3
3
0.03
0.03
1.01
141
[IT] [BE] [NORMAL] [THAT]
3
3
0.03
0.03
1.01
Appendix C
The distribution of [I V Mental that P] construction in different genres Submit
All
Spoken
Fiction
Magazine
Newspaper
Academic
1
I THINK THAT
31,264
27,777
560
944
1293
690
2
I KNOW THAT
7506
4188
1438
817
723
340
3
I BELIEVE THAT
5934
3399
310
647
699
879
4
I HOPE THAT
2437
1521
150
233
347
186
5
I FEEL THAT
1973
1017
150
259
319
228
6
I UNDERSTAND THAT
1798
1198
267
119
168
46
7
I FIND THAT
1062
393
131
299
126
113
8
I SUSPECT THAT
1049
414
120
234
104
177
9
I REALIZE THAT
1008
214
301
269
132
92
10
I SEE THAT
940
279
376
155
89
41
11
I REMEMBER THAT
621
161
266
106
51
37
12
I DOUBT THAT
613
181
168
99
80
85
13
I GUESS THAT
604
309
202
35
45
13
14
I READ THAT
483
235
77
84
73
14
15
I MEAN THAT
479
195
130
52
39
63
16
I WISH THAT
467
243
78
57
69
20
17
I ARGUE THAT
369
15
3
15
10
326
18
I SUPPOSE THAT
344
108
149
32
28
27
19
I HEAR THAT
333
167
95
33
28
10
20
I NOTICE THAT
291
66
111
78
22
14
21
I IMAGINE THAT
274
95
104
37
17
21
22
I WORRY THAT
244
45
58
64
47
© Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7
30 (continued) 161
162
Appendix C
(continued) Submit
All
Spoken
Fiction
Magazine
Newspaper
Academic
23
I EXPECT THAT
240
129
27
34
29
21
24
I RECOGNIZE THAT
200
74
21
41
29
35
25
I GATHER THAT
179
107
50
14
3
5
26
I PROPOSE THAT
167
18
21
12
34
82
27
I APPRECIATE THAT
156
69
41
15
27
4
28
I GATHER THAT
143
104
24
10
2
3
29
I TRUST THAT
127
34
37
19
19
18
30
I SENSE THAT
126
38
32
29
13
14
31
I LEARN THAT
84
5
25
30
11
13
32
I NOTE THAT
84
23
19
17
6
19
33
I PRESUME THAT
71
45
10
3
6
7
34
I CONCLUDE THAT
69
1
10
13
9
36
35
I CONSIDER THAT
67
27
11
12
11
6
36
I DISCOVER THAT
55
2
15
23
7
8
37
I ACKNOWLEDGE THAT
52
15
4
14
5
14
38
I ANTICIPATE THAT
44
25
2
3
7
7
39
I SHOW THAT
40
4
3
7
26
40
I RESPECT THAT
39
22
7
6
4
41
I ESTIMATE THAT
38
3
7
18
5
42
I SWEAR THAT
36
4
23
8
1
43
I DREAM THAT
31
5
17
4
3
44
I ADMIRE THAT
30
7
10
7
6
45
I WONDER THAT
28
5
20
3
5 2
46
I PREFER THAT
28
6
6
7
5
47
I RESENT THAT
20
8
8
2
2
4
48
I SURMISE THAT
18
2
7
3
2
4
49
I SPECULATE THAT
15
2
3
3
1
6 9
50
I HOLD THAT
15
2
3
1
51
I CARE THAT
15
6
5
1
3
52
I RECKON THAT
13
5
5
3
6
2
3
1
2
3
2
5
53
I VIEW THAT
12
54
I DENY THAT
12
(continued)
Appendix C
163
(continued) Submit
All
Spoken
Fiction
55
I PERCEIVE THAT
12
3
4
56
I POSIT THAT
12
57
I SUPPORT THAT
11
58
I DEMONSTRATE THAT
11
Magazine
Newspaper 2
2 7
1
Academic 3 10
3 11
59
I FORESEE THAT
8
1
2
3
60
I INFER THAT
7
2
3
1
1
1
1
61
I OBSERVE THAT
6
1
1
3
1
62
I ASCERTAIN THAT
6
1
2
1
2
63
I REASON THAT
5
1
1
1
64
I PROVE THAT
5
2
1
2
65
I JUDGE THAT
5
2
66
I DETERMINE THAT
5
3
2 3
1
1
67
I DEEM THAT
4
3
1
68
I CONJECTURE THAT
4
3
1
69
I MIND THAT
3
1
70
I INTEND THAT
3
2
71
I DETECT THAT
2
72
I FANCY THAT
2
73
I LAMENT THAT
2
1 1
74
I TREASURE THAT 2
1
I WORSHIP THAT
1
1
76
I TOLERATE THAT 1
77
I STAND THAT
1
I RATIONALIZE THAT
1
79
I PLAN THAT
1
80
I GRASP THAT
1
81
I DREAD THAT
1
82
I DEDUCE THAT
1
83
I ADORE THAT
1
TOTAL
62,282
1 1
1 1
75
78
1
1
1 1 1
1 1 1 1 1 1 1 42,937
5694
5041
4774
3836
Appendix D
The distribution of [I Vverbal that P] construction in different genres Context
All
Spoken Fiction Magazine Newspaper Academic
1
[I] [FIND] [THAT]
1062 393
131
299
126
113
2
[I] [SAY] [THAT]
969
537
181
105
85
61
3
[I] [SUGGEST] [THAT]
669
68
71
116
112
302
4
[I] [PROPOSE] [THAT]
167
18
21
12
34
82
5
[I] [RECOMMEND] [THAT]
149
28
20
58
20
23
6
[I] [ADMIT] [THAT]
147
25
54
36
17
15
7
[I] [CONFESS] [THAT]
141
12
49
44
11
25
8
[I] [ASK] [THAT]
133
72
23
9
21
8
9
[I] [MENTION] [THAT]
133
19
40
54
15
5
10 [I] [NOTE] [THAT]
84
23
19
17
6
19
11 [I] [MAINTAIN] [THAT]
62
8
7
7
8
32
12 [I] [GUARANTEE] [THAT] 58
12
8
18
16
4
13 [I] [DISCOVER] [THAT]
2
15
23
7
8
55
14 [I] [CALL] [THAT]
55
25
9
11
8
2
15 [I] [ACKNOWLEDGE] [THAT]
52
15
4
14
5
14 (continued)
© Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7
165
166
Appendix D
(continued) Context 16 [I] [PROMISE] [THAT]
All
Spoken Fiction Magazine Newspaper Academic
51
8
32
9
2
17 [I] [SHOW] [THAT]
40
4
3
7
18 [I] [SWEAR] [THAT]
36
4
23
8
1
26
19 [I] [DEMAND] [THAT]
22
5
10
2
4
1
20 [I] [CLAIM] [THAT]
22
2
2
6
1
11
21 [I] [GRANT] [THAT]
21
6
4
1
3
7
22 [I] [ASSERT] [THAT]
16
2
1
1
3
9
1
23 [I] [DECLARE] [THAT]
14
4
4
24 [I] [REPEAT] [THAT]
13
7
2
25 [I] [CONFIRM] [THAT]
11
2
3
3
26 [I] [REPORT] [THAT]
7
27 [I] [STATE] [THAT]
6
28 [I] [OBSERVE] [THAT]
6
1
29 [I] [JUDGE] [THAT]
5
30 [I] [PROVE] [THAT]
5
2
31 [I] [ANNOUNCE] [THAT]
5
1
32 [I] [AFFIRM] [THAT]
4
33 [I] [REVEAL] [THAT]
4
1 1
1
1
1
1
1
3
1 1
3 1
2
1
2
2
2
2
1
3 2
36 [I] [AVER] [THAT]
3
1
2
1
1
37 [I] [ADVISE] [THAT]
2
38 [I] [COMPLAIN] [THAT]
2
2
39 [I] [OBJECT] [THAT]
2
1
40 [I] [ORDER] [THAT]
2
1
41 [I] [WHISPER] [THAT]
2
42 [I] [WARN] [THAT]
1
1
1 1 2 1
43 [I] [VOLUNTEER] [THAT] 1
1 1 1
46 [I] [MURMUR] [THAT]
1
1
47 [I] [MUMBLE] [THAT]
1
1
48 [I] [COMMAND] [THAT]
1
1
4250 1313
758
TOTAL
1
2
3
1
2
2
3
45 [I] [REAFFIRM] [THAT]
3
1
34 [I] [REQUIRE] [THAT]
1
3 3
2
35 [I] [REMARK] [THAT]
44 [I] [SAY] [THAT-TO]
2 1
877
517
785
Appendix E
In the following semantic classification of the adjectival predicates of that clauses in [I be ADJ that-clause] in the COCA corpus, the patterns are presented in descending order of their frequency. S. No.
ADJ within the pattern
1
[I] [BE] [SURE] [THAT]
Certainty
Epistemic
1932
2
[I] [BE] [AFRAID] [THAT]
In/security
Affective
894
3
[I] [BE] [CONVINCED] [THAT]
Certainty
Epistemic
685
4
[I] [BE] [GLAD] [THAT]
Un/happiness
Affective
586
5
[I] [BE] [SURPRISED] [THAT]
In/security
Affective
417
6
[I] [BE] [CONFIDENT] [THAT]
Certainty
Epistemic
409
Un/happiness
Affective
340
Epistemic
311
Semantic tag
Category
Frequency
7
[I] [BE] [SORRY] [THAT]
8
[I] [BE] [CERTAIN] [THAT] Certainty
9
[I] [BE] [CONCERNED] [THAT]
In/security
Affective
290
Perception
10
[I] [BE] [AWARE] [THAT]
Epistemic
259
11
[I] [BE] [HOPEFUL] [THAT] Expectation/Expectedness
Epistemic
206
12
[I] [BE] [AMAZED] [THAT] In/security
Affective
179
13
[I] [BE] [HAPPY] [THAT]
Un/happiness
Affective
175
14
[I] [BE] [DISAPPOINTED] [THAT]
Dis/satisfaction
Affective
168
15
[I] [BE] [GRATEFUL] [THAT]
Un/happiness
Affective
150 (continued)
© Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7
167
168
Appendix E
(continued) S. No.
ADJ within the pattern
Semantic tag
Category
Frequency
16
[I] [BE] [WORRIED] [THAT]
In/security
Affective
124
17
[I] [BE] [PLEASED] [THAT] Dis/satisfaction
Affective
112
18
[I] [BE] [PROUD] [THAT]
Dis/satisfaction
Affective
108
19
[I] [BE] [DELIGHTED] [THAT]
Un/happiness
Affective
82
20
[I] [BE] [THANKFUL] [THAT]
Dis/satisfaction
Affective
76
21
[I] [BE] [ANGRY] [THAT]
Anger
Affective
73
22
[I] [BE] [LUCKY] [THAT]
Un/happiness
Affective
56
23
[I] [BE] [SATISFIED] [THAT]
Dis/satisfaction
Affective
55
24
[I] [BE] [SCARED] [THAT]
In/security
Affective
52
25
[I] [BE] [OPTIMISTIC] [THAT]
Certainty
Epistemic
50
26
[I] [BE] [SHOCKED] [THAT]
In/security
Affective
46
27
[I] [BE] [ASHAMED] [THAT]
In/security
Affective
41
28
[I] [BE] [APPALLED] [THAT]
In/security
Affective
40
29
[I] [BE] [POSITIVE] [THAT] Certainty
Epistemic
40
30
[I] [BE] [SAD] [THAT]
Un/happiness
Affective
33
31
[I] [BE] [FORTUNATE] [THAT]
Un/happiness
Affective
31
32
[I] [BE] [OUTRAGED] [THAT]
Anger
Affective
27
33
[I] [BE] [UNAWARE] [THAT]
Perception
Epistemic
26
34
[I] [BE] [SKEPTICAL] [THAT]
Certainty
Epistemic
23
35
[I] [BE] [EMBARRASSED] [THAT]
In/security
Affective
22
36
[I] [BE] [ASTONISHED] [THAT]
In/security
Affective
21
37
[I] [BE] [FEARFUL] [THAT] In/security
Affective
20
38
[I] [BE] [INTERESTED] [THAT]
Evaluative
16
39
[I] [BE] [EXCITED] [THAT] Un/happiness
Affective
15
40
[I] [BE] [ASTOUNDED] [THAT]
Affective
14
Impact
In/security
(continued)
Appendix E
169
(continued) S. No.
ADJ within the pattern
Semantic tag
Category
Frequency
41
[I] [BE] [MAD] [THAT]
Anger
Affective
12
42
[I] [BE] [NERVOUS] [THAT]
In/security
Affective
11
43
[I] [BE] [COMFORTABLE] [THAT]
In/security
Affective
11
44
[I] [BE] [CONSCIOUS] [THAT]
Perception
Epistemic
10
45
[I] [BE] [FURIOUS] [THAT] Anger
Affective
9
46
[I] [BE] [JEALOUS] [THAT] Dis/satisfaction
Affective
9
47
[I] [BE] [SUSPICIOUS] [THAT]
Certainty
Epistemic
9
48
[I] [BE] [UNHAPPY] [THAT]
Un/happiness
Affective
8
49
[I] [BE] [ELATED] [THAT]
Un/happiness
Affective
8
50
[I] [BE] [DISMAYED] [THAT]
Un/happiness
Affective
8
51
[I] [BE] [DISGUSTED] [THAT]
Dis/satisfaction
Affective
7
52
[I] [BE] [FRIGHTENED] [THAT]
In/security
Affective
7
53
[I] [BE] [MINDFUL] [THAT]
Perception
Epistemic
7
54
[I] [BE] [RIGHT] [THAT]
Actuality
Epistemic
7
55
[I] [BE] [PUZZLED] [THAT] In/security
Affective
6
Evaluative
6
Epistemic
5
Affective
5
Boulomaic
5
56
[I] [BE] [CURIOUS] [THAT] Impact
57
[I] [BE] [CLEAR] [THAT]
58
[I] [BE] [AMUSED] [THAT] Un/happiness
59
[I] [BE] [ADAMANT] [THAT] Total
Obviousness Will
8354
Bibliography
Goldberg, A. E. (1996). Construction grammar. In Brown & Miller (Eds.), Concise encyclopedia of syntactic theories. Pergamon. Halliday M A K. (2009). Methods—techniques—problems. In Continuum companion to systemic functional linguistics (pp. 59–86). Halliday, M. A. K. (1985/1994). In An introduction to functional grammar. Edward Arnold. Martin, J. R. (2010/2000). Beyond exchange: Appraisal systems in english. In W. Zhenhua (Ed.), Discourse semantics (语篇语义研究) (pp. 203–245). Shanghai Jiao Tong University Press. Shao, X., & Xu, X. (邵新光 徐秀芹). (2015). 多媒体环境下语法隐喻在英语阅读中的认知作用 研究 [Cognitive Effects of Grammatical Metaphor on English Reading in Multi-media Setting]. TEFLE, 1, 57–61.
© Peking University Press 2023 J. Hu, A Constructional Approach to Interpersonal Metaphor of Modality, Peking University Linguistics Research 7, https://doi.org/10.1007/978-981-99-2764-7
171