Referring in Language: An Integrated Approach 1107143470, 9781107143470

Referential expressions include terms such as determiners, proper names, noun phrases, pronouns, and all other expressio

277 83 6MB

English Pages 256 Year 2023

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Referring in Language: An Integrated Approach
 1107143470, 9781107143470

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Referring in Language

Referential expressions include terms such as determiners, proper names, noun phrases, pronouns, and all other expressions that we use to make reference to things, beings, or events. The first of its kind, this book presents a detailed, integrated account of typical and atypical uses of referential expressions, combining insights from discourse, cognitive, and psycholinguistic literature within a functional model of language. It first establishes a foundation for reference, including an overview of key influences in the study of reference, the debates surrounding (in)definiteness, and a functional description of referring expressions. It then draws on a variety of approaches to provide a comprehensive explanation of atypical uses, including referring in an uncollaborative context, indefinite expressions used for definite reference, reference by and for children, and finally metonymic reference with a special focus on metonymy in medical contexts. Comprehensive in scope, it is essential reading for academic researchers in syntax, discourse analysis, and cognitive linguistics. lise fontaine is a reader at Cardiff University. She has published many edited volumes, including The Oxford Companion to the English Language (2018) and The Cambridge Handbook of Systemic Functional Linguistics (2019). katy jones is a senior lecturer at Cardiff University. She has published a variety of papers on reference and discourse. She is the director of the Writing Centre in the School of English Communication and Philosophy at Cardiff University. david scho¨nthal is a lecturer in academic writing at Cardiff University. He is co-editor of The Cambridge Handbook of Systemic Functional Linguistics (2019).

Referring in Language An Integrated Approach Lise Fontaine Cardiff University

Katy Jones Cardiff University

David Schönthal Cardiff University

Shaftesbury Road, Cambridge CB2 8EA, United Kingdom One Liberty Plaza, 20th Floor, New York, NY 10006, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 103 Penang Road, #05-06/07, Visioncrest Commercial, Singapore 238467 Cambridge University Press is part of Cambridge University Press & Assessment, a department of the University of Cambridge. We share the University’s mission to contribute to society through the pursuit of education, learning and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107143470 DOI: 10.1017/9781316534625 © Lise Fontaine, Katy Jones and David Schönthal, 2023 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press & Assessment. First published 2023 A catalogue record for this publication is available from the British Library. Library of Congress Cataloging-in-Publication Data Names: Fontaine, Lise, author. | Jones, Katy (senior lecturer), author. | Schönthal, David, author. Title: Referring in language : an integrated approach / Lise Fontaine, Katy Jones, David Schönthal. Description: Cambridge, United Kingdom ; New York, NY : Cambridge University Press, 2023. | Includes bibliographical references and index. Identifiers: LCCN 2022053488 (print) | LCCN 2022053489 (ebook) | ISBN 9781107143470 (hardback) | ISBN 9781316507797 (paperback) | ISBN 9781316534625 (epub) Subjects: LCSH: Reference (Linguistics) Classification: LCC P325.5.R44 F66 2023 (print) | LCC P325.5.R44 (ebook) | DDC 401/.456–dc23/eng/20221123 LC record available at https://lccn.loc.gov/2022053488 LC ebook record available at https://lccn.loc.gov/2022053489 ISBN 978-1-107-14347-0 Hardback Cambridge University Press & Assessment has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Contents

List of Figures List of Tables Acknowledgements

Part I 1

2

page viii ix x

The Nature of Referring and Referring Expressions

The Multidisciplinary Field of Referring

3

1.1 Introduction 1.2 The Foundations of Reference 1.2.1 Philosophy of Language 1.2.2 Influences from Computational Linguistics 1.2.3 Psycholinguistic Perspectives on Referring 1.3 On Text and Discourse 1.3.1 Reference as Textual 1.3.2 Discourse Perspectives 1.4 Cognitive Perspectives: Reference in the Mind 1.4.1 Summary 1.5 Volume Aims and Structure 1.5.1 Aims of the Book 1.5.2 Volume Structure

3 5 6 8 10 13 13 16 17 20 20 20 21

The Great Debate: Definiteness and Indefiniteness

23

2.1 The The 2.2 Reference and Referring 2.2.1 Textual Reference 2.2.2 Discourse Reference 2.2.3 Summary 2.3 (In)Definiteness 2.3.1 Definiteness 2.3.2 Indefiniteness 2.3.3 Identifiability 2.4 What’s in a Name? 2.5 Summary

23 25 27 29 31 32 33 36 41 43 46

v

vi

3

Contents

A Functional Grammar for Referring Expressions

48

3.1 A Non-headed Noun Phrase 3.2 A Cognitive Approach to the Grammar of Referring 3.2.1 Signals of Definiteness and Indefiniteness 3.2.2 The Grounded Expression 3.3 The Basic Organisation of the Noun Phrase 3.3.1 A Cognitive Functional Description of the Noun Phrase 3.3.2 Categorisation: Referential Type 3.3.3 Determination and Modification 3.3.4 The Fourth Zone: Qualification 3.3.5 Four Zones: A Summary 3.4 Summary

48 49 50 51 53 54 55 58 63 67 68

Part II 4

5

6

Typical Reference

Typicality and Atypicality

73

4.1 On Categories 4.2 What Is Typicality and Atypicality? 4.2.1 Register 4.2.2 Typicality and Prototypicality 4.2.3 Markedness 4.2.4 Norms and Exploitations 4.2.5 Typicality and Atypicality as Usage-Based and Register-Dependent 4.3 Typical Reference 4.3.1 The Accessibility Scale 4.3.2 The Givenness Hierarchy 4.4 Towards an Account of Atypicality in Reference 4.4.1 The (In)Stability of Identifiability 4.4.2 Atypical Reference as an Exploitation of the Norm 4.5 Summary of the Chapter

73 74 76 77 78 79 80 81 82 84 87 87 89 91

Referring in Spontaneous Discourse

92

5.1 5.2 5.3 5.4

92 93 99 106

Introducing Spontaneity Mode: Defining Spontaneous Language Overview of Referring Expressions in Spontaneous Communication Concluding Remarks

Referring in Prepared Discourse 6.1 6.2 6.3 6.4 6.5 6.6

Time and a Word Mode and Prepared Writing How We Make a Jointly Constructed Text ‘Remotely’ Referring Expressions in Written Texts Written Genres Summary

Part III 7

108 108 109 114 115 120 123

Atypical Reference

Referring in Non-collaborative Contexts 7.1 Introduction 7.2 Problematic Reference

127 127 131

Contents 7.3 Referential Choice 7.3.1 Type Specification and Rephrasing 7.3.2 Rejected Referring Expressions 7.4 Concluding Remarks

8

Indefinite Expressions for Referring Definitely 8.1 Introduction 8.2 Referring and (In)Definiteness 8.2.1 Late Indefinites 8.2.2 Lexical Cohesion 8.3 Co-extensional Analyses 8.3.1 Background to Co-extensional Analyses and Texts 8.3.2 Co-extensional Analysis 1: The Reader’s Digest Text 8.3.3 Co-extensional Analysis 2: The David Beckham Text 8.3.4 Co-extensional Analysis 3: The Jewish National Fund Text 8.3.5 Co-extensional Analysis 4: The Greg Dyke Text 8.3.6 Concluding Comments on the Analyses 8.4 Discussion: ‘Dual’ Reference 8.5 Conclusion

9

Child’s Play 9.1 Introduction 9.2 Children and Reference 9.2.1 The Earliest Stage: Protolanguage 9.2.2 Developing Awareness of Common Ground 9.3 Under/Overspecification 9.4 Creative Licence: Playing with Reference 9.5 Conclusion

10

Reference and Metonymy 10.1 A Name by Any Other . . . 10.2 Metonymy, of the Referential Kind 10.2.1 Metonymy and Metaphor 10.3 Metonymic Expressions in Healthcare 10.3.1 Nurse Shift-Change Handover Discourse 10.3.2 Staff Attitudes towards Metonymy 10.4 Metonymic Reference 10.5 Typical and Atypical Metonymic Reference

11

Epilogue 11.1 Synthesis 11.2 Beyond This Volume

References Index

vii 136 139 140 148

150 150 154 156 161 165 166 168 169 170 170 171 172 174

175 175 175 176 178 185 191 196

197 197 199 205 206 207 210 212 215

217 217 220

222 240

Figures

2.1 3.1 3.2 3.3 4.1 4.2 5.1 5.2 5.3 7.1 7.2 8.1 8.2 8.3 8.4 8.5 9.1 9.2

viii

Givón’s coding gradation of reference (1993a: 225) page 40 Individuative reference (Radden and Dirven, 2007: 89) 52 Generic reference (Radden and Dirven, 2007: 106) 53 A functional-cognitive model of the English NP (adapted from Ghesquière, 2014: 24) 54 Accessibility Marking Scale (Ariel, 1990: 73) 83 The Givenness Hierarchy (Gundel et al., 2019: 68) 85 Categories of cohesive reference (adapted from Halliday and Hasan, 1976: 33) 102 Classification of person reference (Halliday and Hasan, 1976: 44) 102 Types of phoric relations (adapted from Martin, 1992: 124) 103 ‘Spatula and Fat’ by Andrew Mason (2005) licensed under CC BY 2.0. www.flickr.com/photos/a_mason/7251819 137 Rephrasing referring expressions (based on Clark and Wilkes-Gibbs, 1986) 142 Cohesive links (Hasan, 1985) 164 Similarity chain for the Reader’s Digest text 168 Similarity chain for the David Beckham text 169 Similarity chain for the Jewish National Fund text 170 Similarity chain for the Greg Dyke text 171 Principle of quantity (Givón, 1985) 187 Tom’s story as originally written 190

Tables

3.1 3.2 3.3 3.4

Various perspectives on the noun phrase page 54 Examples of the three main functional zones of the English NP 56 Ghesquière’s determination zone with examples (2014: 44) 63 A functional description of the referring expression (adapted from Ghesquière, 2014) 68 5.1 Distribution of type of referring expressions 105 9.1 Specification values of initial referring expressions for children and adults (adapted from Uzundag and Küntay, 2018: 84) 188

ix

Acknowledgements

We would like to thank Helen Barton at Cambridge University Press not only for her constant support and encouragement throughout the preparations for this volume, but also for her patience since this volume took much longer than anticipated. Life is like that sometimes, and we are grateful for her understanding. We are also grateful to her editorial team at Cambridge University Press, especially Isabel Collins, who was incredibly helpful to us in preparing the manuscript. We have many people to thank in terms of discussions of the content in many of the chapters, and we would like to single out those who commented on drafts of the manuscript. Our sincere thanks go to Dr Anne McCabe, who provided constructive comments and suggestions on the manuscript. We also want to thank Dr Michelle Aldridge for sharing her expertise in child language development and in health communication, and for her valuable comments on drafts of the related parts of the volume. We are especially grateful for the positive comments and helpful suggestions that we received from the anonymous external reviewer of the final draft. We would also like to thank the anonymous reviewers of the proposal itself for the support they gave to this project and for the very useful comments that helped us in directing our plans. We are very grateful to a dear friend, J. Shallow, for generously contributing a custom-designed illustration (the way i see things) for one of our examples. Thank you! Our thanks also go to Emma Collier for her detailed and careful proofreading of our work. Last but certainly not least, we want to thank Ceri Tanswell for her work on the index for this volume. Writing a volume such as this takes its toll, and it takes time. We want to express our gratitude to our families and friends who supported us throughout these past many years as we devoted so much time to referring. Lise Fontaine was supported by Cardiff University’s research leave award on two separate occasions during the writing of this book. I (Lise) am very grateful for the opportunity to have had dedicated time for this work. I am especially grateful to my colleagues for their support. x

Part I

The Nature of Referring and Referring Expressions

1

The Multidisciplinary Field of Referring

One of the oldest problems in the philosophy of language, indeed, one of the oldest problems in Western philosophy: the nature of linguistic reference. (Searle, 1990: xiii)

1.1

Introduction

The motivation for this volume began several years ago when we started to notice certain types of referring expressions which did not, in our view, fit comfortably with the established literature on reference. We started discussing these outlier types of expressions in terms of atypical reference. As will be clear later in this volume, we have continued to use the term ‘atypical’ despite it being not entirely satisfactory, for reasons we explain in Chapter 4. For now, we can simply think of typical reference as conventional and atypical reference as divergent uses. Since the field of study related to referring and reference is broad – spanning multiple disciplines, including philosophy, psychology, linguistics, and computer science – understanding it is a multidisciplinary task. Some disciplines combine more readily than others for a variety of reasons. We also find that even within one discipline, there are many different approaches and perspectives. This kind of variation is certainly true of linguistics. While this volume is anchored within linguistics, we build on multiple perspectives to provide an integrated approach to referring in language with a focus on atypical expressions. At this point you might be wondering what exactly an atypical example of referring is. We want to differentiate these expressions from ones that are errors, or that are related to language development (see Chapter 9). The expressions that caught our attention often appeared to be unusual uses of definite and indefinite articles, or ones that seemed to be playing with the conventions of reference. There is, for example, in the general literature on referring, an assumption that most uses of the definite article, the, signal a known discourse referent. In fact, in the previous sentence, we used the in the

3

4

The Multidisciplinary Field of Referring

noun phrase (NP)1 the definite article to signal that there is in fact only one such article in the English language and most people interested in language know this. The definite article is called definite because of its typical use in specifying an identifiable, or definite, referent. We will return to the concepts of definiteness and identification in various places in this chapter and later chapters (specifically in Chapter 2). For now, and with the idea of the definite article in mind, let us consider examples (1) and (2).2 The uses of the in these two examples seem atypical in the sense that they do not represent the most common or most frequent uses of the; in other words, the highlighted expressions are doing more than what we typically think of as the job of the definite article. (1)

That’s not to say I’m not feeling some inflation; as a consumer I’ve begun really noticing that my money isn’t going as far as it did, say, about five years ago when the wife and I relocated to California.

(2)

I’m tired of doing things that I know I SHOULDN’T do and I don’t WANT to do but can’t seem to have control over. I hate being the wife that freaks out in the grocery store or the mom who can’t play with her kids because I can’t pull myself away from the anxiety of everything.

In example (1), the could easily be replaced by my as in my wife. Some of us might not appreciate the use of the in this expression because it has an objectifying connotation. There is something about using this expression, the wife, to refer to a specific woman that is atypical, or certainly not the norm for referring to people, other than by their position (e.g. the teacher or the lawyer). The juxtaposition of the with a common noun implies a kind of relational position. The use of the in example (2), while also an atypical expression, is different from its use in example (1). In (2), the speaker is referring to herself, and the implication is that there are types of wife (the kind of wife that . . .), which construes a qualitative description similar to I am smart or I am sensitive, for example.3 However, here, it is very clearly a self-assessment of behaviour that the speaker finds undesirable; she is referring to herself, as signalled by the use of I in a chain of self-reference (the wife, the mom, I, myself). It is as though the speaker steps out of definite reference to herself as

1 2

3

The linguistic term noun phrase or NP will be explained in detail in Chapter 3. We would like to point out that we are only commenting here on the use of the. Referring to close personal relationships using expressions such as the wife, the daughter, etc. is common in many dialects of English. Unlike example (1), the phoricity of the here is referred to as esphoric reference by Martin (1992: 123), although in this case the expression is not being used to introduce a participant to the ongoing discourse, which is the common use of esphora (see also Du Bois, 1980 for a discussion of esphora).

1.2 The Foundations of Reference

5

the speaker to identify a type of referent. Thus, these uses of the stand out and, because such uses are typically infrequent, they carry different, or additional, referential meanings. It is not only definite NPs that are used in atypical ways. The literature on referring is also clear that indefinite expressions such as someone are inherently non-specific since there is no discourse referent that can be identified. If you return home and see your front door is open and there appears to have been a break-in, you are likely to say someone broke in. The use of someone is useful for such occasions because you may know that one or more humans are involved, but you may have no idea specifically which individual or individuals this includes. In examples (3) and (4), we present instances of the use of someone where this is not the case. (3)

Someone didn’t put the garbage in the bin and now there’s garbage all over the deck.

(4)

I’m guessing someone took the car and someone took the truck.

Example (3) is one that some parents may recognise as a useful strategy to let a child know that they have made a mistake. These examples illustrate, then, that there are contexts in which a speaker will use an indefinite or general expression despite knowing the specific identity of the referent. In example (4), the speaker is referring to her son and daughter-in-law who had been visiting one afternoon but left, each in one of their jointly owned vehicles. The speaker, not knowing which one drove off in the car and which one left in the truck, opts for someone. These examples illustrate how a speaker can have a specific referent in mind and still choose an indefinite expression. The question raised is whether these instances are referring expressions or not. We will come back to this point later in the chapter and in more detail in Chapter 2. These four examples provide a glimpse into the kinds of expressions that have motivated this volume. In the remainder of this chapter, we will provide an overview of the approaches that have informed the integrated perspective we present in the volume. Towards the end of this chapter, we will set out the aims of the volume and briefly describe the structure of the book.

1.2

The Foundations of Reference

The study of reference is a rich field with significant and historical foundations in philosophy which have spread to computational approaches and psycholinguistics as well as general linguistics. In this section, we very briefly review some of the key contributions from philosophy, computational linguistics, and psycholinguistics to give us the background of influences that are still relevant to our understanding of reference and referring.

6

The Multidisciplinary Field of Referring

1.2.1

Philosophy of Language

The earliest studies of referring expressions and reference stem from philosophy. It was then developed outwards and into linguistics via two principal paths: semantics (e.g. Lyons, 1977) and computational linguistics (e.g. Appelt, 1985). Searle’s work on reference has led indirectly from philosophy to studies of discourse and pragmatics (e.g. Searle, 1969). The long-lasting influence of philosophy of language on linguistics is not surprising since, in comparison, linguistics is a young science. It is worth noting briefly, as we do here, some of the key features that we have inherited from philosophers. For philosophy, the central notion has been that of unique reference – the use of an expression to uniquely identify some object or person. Referring expressions generally include ‘proper names, definite descriptions (“the tallest man in the world”); demonstrative terms (“this man”) and some pronouns’ (Evans, 1982: 1). In this view, our examples (2) to (4) would not be referring expressions since they cannot be said to uniquely identify an object or person. Within the field of philosophy, however, there has been some dispute as to what constitutes a referring expression. Part of the problem lies in interpretations of the (non-)existence of the referent and/or truth value of the expression in question; for example, whether an expression referring to something that does not exist can be said to have reference. The first real treatment of reference is attributed to Frege (1892/1993), although according to Searle (1969: 77), it can be traced back as far as Plato. For Frege, a referring expression is an expression that refers to a unique object, in other words, a definite description. In his view, a referring expression is a type of sign, and by sign Frege (1892/1993: 24) means ‘any designation representing a proper name, which thus has as its reference a definite object (this word taken in the widest range)’. One of Frege’s greatest contributions, or at least his longest-lasting contribution, is his idea that any given referring expression will have two different types of meaning: sense and reference. Frege (1892/1993) demonstrated the need for this distinction most famously with his now classic morning star/evening star instance, given here in example (5). (5)

The morning star is the evening star

Frege’s claim was that while each nominal expression in (5) has a different sense, they both share the same reference (i.e. both NPs are used to refer to Venus). Reference, for Frege, is the actual object being designated, that is, ‘an object perceivable by the senses . . . an internal image’ (1892/1993: 26). Sense is part of the meaning of the sign but separate from reference. He also claimed that an expression may have a sense but no reference, as in example (6) from Frege (1892/1993: 25), where there is no unique object to be designated. This expression is not a proper name (definite referring expression) in Frege’s

1.2 The Foundations of Reference

7

terminology. While the distinction between sense and reference may be welcome to some linguists, most of us would still want to consider examples such as (6), and our invented example (7), as referring expressions despite the absence of a uniquely identifiable referent. (6)

the least rapidly convergent series

(7)

the largest number in the world

Frege’s work led to considerable debate and various developments in relation to reference, but we will jump to Strawson (1950) because his contributions are important to linguistic views on reference. He argued that existence should not be a criterion for defining a referring expression, which led to his theory of presupposition, a standard notion in the field of pragmatics. This work may well have formed the first functional approach to the study of referring expressions. In Strawson’s view, a referring expression is ‘an expression with a uniquely referring use’ (ibid.: 324); by ‘uniquely referring use’, Strawson intends the use of expressions ‘to mention or refer to some individual person or single object or particular event or place or process, in the course of doing what we should normally describe as making a statement about that person, object, place, event, or process’ (ibid.: 320). These types of expressions include proper names, singular demonstrative pronouns, singular personal and impersonal pronouns, and ‘phrases beginning with the definite article followed by a noun, qualified or unqualified in the singular’ (ibid.). However, he insists on the importance of the use of the expression. He makes a three-way distinction among an expression, the use of an expression, and the utterance of an expression (ibid.: 325–326). Both the context of the utterance of an expression and its function are very important according to Strawson. The following two examples, (8) and (9), are given by Strawson to make this point. (8)

the king of France is wise

(9)

the greatest French soldier

In (8), Strawson explains (1950: 325) that the context of the utterance the king of France is important since if uttered during the reign of Louis XIV, some would say the sentence is true, but if it were said during the reign of Louis XV, many would say it is false. Similarly, the expression the greatest French soldier, in example (9), may or may not be uniquely referring, depending on the function of its use in a given context. If uttered in the sentence Napoleon was the greatest French soldier, then the expression is not referring since it is not being used to refer to a particular individual; instead, it is being used to say something about a particular individual (Napoleon). However, if the same

8

The Multidisciplinary Field of Referring

expression is used in the sentence the greatest French soldier died in exile, then it is being used to refer to an individual (Napoleon). The realisation of the significance of context enabled Strawson to see referring not as something that an expression does, but rather as a resource that can be used by a speaker. Linguistic approaches to reference have inherited a great deal from philosophical studies of reference. Traces of its influences can be found in most, if not all, linguistic theories, and this is most notably relevant in the use of concepts such as definiteness, sense, and reference. Although much of what we glean from philosophical studies of language is important, some of it carries with it concerns and definitions that are not relevant for linguistic studies, as we will see in our discussion of (in)definiteness in Chapter 2. Nevertheless, any study of referring expressions must understand the roots of its development, and very clearly these roots lie in the earliest philosophical studies of linguistic reference. Comprehensive treatments of reference within the philosophical traditions are presented in Bach (1987) and Sainsbury (2005). For more recent contributions to the study of reference from a philosophical perspective, see work by Clapp et al. (2019) on negative existentials, and also Hanks (2019) on speech acts. 1.2.2

Influences from Computational Linguistics

Although we are not including computational approaches to referring in this volume, it is worth briefly considering contributions that have been made from this field. Referring expressions, as a domain of study, have received the greatest attention from the field of natural language processing, which covers various computational linguistic tasks, such as machine translation, speech recognition, text-proofing, automated summarisation, information retrieval and extraction, question answering, natural language understanding and natural language generation, and, increasingly, robotics. In language generation, many approaches were developed to solve the problem of generating referring expressions, for example, from a pragmatic perspective (e.g. Appelt, 1985; Appelt and Kronfeld, 1987; Dale and Reiter, 1995), from a discourse perspective (e.g. Paris and McKeown, 1987), from a syntactic perspective (e.g. Power, 1999), or from the perspective of the computational architecture (e.g. Horacek, 1997). Researchers working on referring expressions for the purposes of language generation have generally defined the generation of a referring expression as ‘the task of selecting words or phrases to identify domain entities’ (Reiter and Dale, 1997: 68). Domain entities are an infinite set of entities (things) which belong to a particular domain, including everything concrete, abstract, and imaginable. For any given computer generation system, the set of domain entities is usually constrained to a specific context, for example, weather

1.2 The Foundations of Reference

9

reports (Reiter et al., 2005), automatic route descriptions (Dale et al., 2005), or recipes (Dale, 1989). As it is widely accepted that a referring expression will be realised as a noun phrase, referring expression generation is, then, a procedure for producing an NP that ‘identifies an entity . . . to the hearer in the current discourse context’ (Reiter and Dale, 1997: 79). While it is perhaps more common for computational work to take up concepts from linguistics than the other way around, there are two contributions we feel are worth mentioning briefly here. One stems from the work of Dale and Reiter (1995), which focussed on how to incorporate Gricean maxims (Grice, 1975) while avoiding the potential to create false implicatures. Example (10) from Dale and Reiter (1995: 251) illustrates this. Here, a referring expression is generated that provides an overspecification of the referent, that is, it is not as brief as it could be (cf. Sit by the table), where the additional information expresses the communicative goal of warning the addressee not to touch the table as they sit down (because of the fresh paint). (10)

Sit by the newly painted table

A key finding in their work was that, at least in computational terms, adhering strictly to Grice’s maxim of ‘Brevity’ is very costly. Psycholinguistic evidence suggests that speakers do not adhere to this maxim strictly when referring (see Levelt, 1989, for example). The conclusion drawn by Dale and Reiter (1995: 247) concerning the task of generating referring expressions is that ‘the Gricean maxims should not be interpreted too literally’ (ibid.: 249). Another, perhaps more significant, view of referring expressions that was gaining importance at the same time was the view of referring as action and plan rather than as description, that is, the speaker is planning an expression, rather than describing an object (Dale and Reiter, 1995) in the computational architecture. A major contribution comes from Kronfeld, who argued that ‘a computational model of referring must show how the successful use of a referring expression in a given context is due to the solving of a planning problem – given also a goal, various rationality assumptions, and relevant linguistic institutions’ (Kronfeld, 1990: 9). Kronfeld understood that referring was not only about identification and that the speaker had to plan the best approach to achieving the desired effects (however unconsciously). Planning is required, according to Kronfeld, because of the need to consider ‘the effects on the hearer’s mental state’ (ibid.: 76). In summary, computational models, in the very least, offer contributions that help us separate cognitive, psycholinguistic, and pure linguistic behaviour, even if there is some degree of overlap. Both the domains of the philosophy of language and the computational modelling of language come (although via very different routes) to find that there is a kind of separation, or rather a distinction to be made, between the beliefs that a speaker has and the linguistic

10

The Multidisciplinary Field of Referring

choices available. Psycholinguistics is the field that tries to maintain the balance between what is going on in the brain and what is happening in the language system. In what follows, we turn our attention to more experimental evidence from psycholinguistics. 1.2.3

Psycholinguistic Perspectives on Referring

Despite the significant contributions made by philosophers, Clark and WilkesGibbs (1986: 3) argue that philosophy offered a limited account of referring. They claim that the main reason for this lies in what they call a ‘literary model’ of reference, which holds four tacit assumptions (ibid.): (i) reference is only expressed by certain types of noun phrase – proper noun, definite description, or a (certain) pronoun; (ii) noun phrases identify the referent uniquely against the common ground; (iii) speaker intention is met simply by issuing that noun phrase; and (iv) the referring process is controlled by the speaker alone. When compared against what speakers actually do during conversation, Clark and Wilkes-Gibbs (1986) found that these assumptions failed to account for ‘real’ instances of referring. In what follows, we will very briefly consider what referring looks like from a psycholinguistic processing perspective. One of the most important contributions from psycholinguistic research into referring expressions has been in relation to the role of collaboration between speaker and addressee, which helps participants to refer successfully (Clark and Wilkes-Gibbs, 1986; Schober and Clark, 1989; Brown, 1995; Brennan and Clark, 1996). The ‘collaborative view’ proposed by Schober and Clark (1989: 211–232) claims that participants in a conversation actively work together to ensure that understanding takes place and do not proceed with the conversation until they are satisfied that they have mutually understood each other. We will expand on this issue throughout the volume but in particular in Chapter 5. For successful reference to take place, there needs to be a level of mutual understanding between the speaker and addressee. Evans (1982: 315) suggests that the interlocutors not only need to be thinking of the same referent, but also thinking of it ‘in the right way’, proposing that this necessitates ‘thoughts that are pretty similar on the part of the speaker and hearer’ (ibid.: 316). Brown echoes this (1995: 65) but distinguishes between correct interpretation and adequate interpretation. The former requires that the listener must have the same thought, or thoughts that are ‘pretty similar’ to the speaker, while the latter takes into consideration the relationship between ‘the utterance, the context and the listener’s intentions in interpreting the utterance’ (ibid). The question here, of course, is to what extent the addressee needs to retrieve the intended referent and think of it in a sufficiently similar way to the speaker for successful understanding to take place. Clark and Wilkes-Gibbs’ notion of understanding ‘a criterion sufficient for current purposes’ (1986: 34) goes

1.2 The Foundations of Reference

11

some way to explaining how closely aligned interlocutors have to be. This criterion states that for a new contribution (i.e. ‘a unit of conversation’ (ibid: 35)) to proceed, the old, previous contribution must have been mutually accepted. Clark and Wilkes-Gibbs’ (1986: 33) principle of ‘mutual responsibility’ between speaker and addressee is also important since referring is a collaborative process requiring action on the part of both speaker and addressee in order to allow for successful reference to take place. They suggest that during the referring process, the speaker and addressee may repeatedly repair, expand on, or replace the referent until they arrive at a mutually acceptable version (see our discussion of rephrasing in Chapter 7). So, before the conversation can move on, both interlocutors must be satisfied that they have reached a sufficient level of mutual understanding. In Chapter 7, we consider how this collaboration works even when speaker and addressee are in a more adversarial than collaborative relation. As long as the speaker and addressee are satisfied that mutual understanding has taken place, then it is sufficient. In ongoing, interactive discourse, it is fairly easy to stop and check understanding, especially if there is any doubt about the identity of the referent. In written modes, the principle of mutual responsibility may be weakened or modified and becomes instead a principle of ‘distant responsibility’ (Clark and Wilkes-Gibbs, 1986: 34), which means that writers have to ensure that the reader understands their meaning ‘to a criterion sufficient for current purposes’ (ibid: 36). It could be that the writer needs to provide other pointers to help the reader: for example, certain co-textual or semantic clues, such as the ongoing sequence of references to the same referent or the fact that the heading or title is the discourse topic of the text. Clark and Wilkes-Gibbs (ibid: 36) add, however, that some writers may ‘retain a sprinkling of provisional noun phrases, repairs, expansions and replacements apparently to affect a spontaneous style or for other rhetorical effects’. This suggests that the principles of both mutual and distant responsibility may be somewhat violated in order to have a particular impact on the discourse (cf. Fox, 1987). The identification of a referent is not the only function of referring expressions; the speaker or writer might be ‘doing something special’ (ibid.: 17). This matter is revisited in Chapter 8, where the function of the expression-type is discussed. When a speaker is intentionally over- or under-explicit, this overspecification or underspecification could be seen as a breach of the ‘communicative contract’ (Givón, 1993a: 232) or ‘conceptual pact’ (Brennan and Clark, 1996: 1491), that is, a temporary agreement between interlocutors about how they are to conceptualise a particular entity, since the speaker could be considered to be acting in an uncooperative manner. For instance, as we saw in example (10), the modifier newly painted was over-explicit for referential purposes but still a successful reference. The overspecification does not

12

The Multidisciplinary Field of Referring

necessarily mean that these expressions are not collaborative. In the case of a written text, the writer might assume that they have put sufficient cues in place for the reader to be able to successfully identify the referent. Thus, the principle of distant responsibility to the addressee remains, at least to some extent, intact (see Chapter 6). The role of the addressee is central to referring, not just in terms of a shared conceptualisation of the referent, but because for acts of referring to work, the addressee has to implicitly or explicitly agree to the expression. As explained by Brennan (2000: 4), ‘referring expressions are provisional until ratified by addressees’, who play a crucial role in the successful resolution of a referring expression. In instances where the identity of the referent could be potentially unclear or the intentions of the speaker are uncertain, it is the addressee’s role which is the most decisive. We will look at examples of this in some detail in Chapter 7, but for now we will briefly consider evidence of how speakers include their addressee’s knowledge about the referent and surrounding context. A study which is particularly pertinent to Brennan’s above mentioned claim is that of Brown’s shared map task (1995), in which she reports that addressees are not particularly concerned with the definiteness status of referring expressions. Brown discovered that listeners, when collaborating in a shared map task, interpreted both definite and indefinite expressions as referring. The listener was able to identify the referent successfully in both cases. In the map task, it did not appear to matter whether an indefinite or definite expression was used to describe a shared feature; the listener took it as intending to identify a particular referent (ibid.: 70–72). Brown found that indefinite expressions are treated as referring to ‘the one unique example of a palm beach on their map’ (ibid.: 70). Wright, in his analysis of a similar map task, supports this observation, stating that ‘the occurrence of definite and indefinite articles is not a reliable indicator as to whether a speaker is treating an item as shared or not’ (1990: 71). Further, he (ibid.: 78) concludes that ‘[i]t does indeed seem that more skilled users of the language are able to tailor the referential expressions they use to the informational conditions in which they occur [. . .] [and] that speakers take account not only of information in the referential domain, but also of their hearer’s state of knowledge of that domain’. This comment is noteworthy since Wright (1990) is suggesting that it is not the formal marking of referring expressions which is crucial to the mutual understanding of the interlocutors, but rather their judgement and interpretation of the context, as well as how they evaluate each other’s awareness of the context of utterance. Returning to Brown’s map task experiment, it must be noted that she concludes that the speaker is simply indicating that ‘the epistemic status of the referent is not (or may not be) well-grounded for both participants’ by

1.3 On Text and Discourse

13

using an indefinite expression (1995: 75). However, the comment in parenthesis is significant; it is not necessarily the case that the referent is not wellgrounded, it simply ‘may not be’, which accordingly means that it may be well-grounded also. This is an issue which is revisited in Chapter 8, where we consider the extent to which the referent is grounded for the participants and what implications this has for the referring status of indefinite noun phrases. 1.3

On Text and Discourse

In this section, we first consider what reference looks like when examined from a textual perspective and the concepts of co-reference and anaphora. Since this perspective has some important shortcomings in helping us understand reference fully, we then turn our attention to perspectives from discourse studies. 1.3.1

Reference as Textual

Linguistics has largely been built on so-called armchair theorising which, until fairly recently, has underpinned the approach taken by linguists to study reference and anaphora. These theories tend to be formed by introspection using invented two-sentence examples (Cornish, 1999: 1). In the spirit of tradition, let us begin this discussion with the invented example in (11) and our introspective comments on the expression of interest, they, which is highlighted in bold. (11)

Katy has three little boys. They are playing over there.

Interpreting this pair of sentences should not be too problematic as there is nothing inherently ambiguous about the expressions. They cannot refer to anything other than three little boys as it is the only plural noun phrase in the first sentence. The second sentence clearly elaborates on the first, creating a certain coherence between the two sentences. Co-textual approaches are wellequipped to deal with such instances. More traditional accounts of reference and anaphora are largely concerned with the formal and semantic constraints which allow or prohibit the pairing or matching of certain linguistic elements. In this way, these approaches focus on pairs of constructed sentences at best, and do not concern themselves with the wider context or discourse. Traditional accounts of referring have tended to view reference as a coreferential procedure, that is, as a co-textual relationship between a lexical antecedent and a referent which is dependent on the antecedent for its semantic interpretation. This view is broadly supported by some functionalists, such as Halliday and Hasan (1976) and Martin (1992), as well as in computational approaches, where formal constraints are prioritised (as noted by Cornish,

14

The Multidisciplinary Field of Referring

2010: 212). However, formal approaches to reference have received considerable criticism for not considering the communicative event as a whole (e.g. by Du Bois, 1980; Brown and Yule, 1983; Garnham, 2001; Schiffrin, 2006; Cornish, 2010). For example, the two invented sentences in (12) pose some degree of difficulty for a co-textual framework. Out of context, this pair of sentences is problematic. An obvious connection or coherence between the two sentences is more difficult to determine, so a logical interpretation would be difficult. If we now consider a slightly modified set of sentences in (13), we might agree that there is a somewhat tighter connection between this pair of sentences. The addition of the concessive but and the prosody of the speaker help the addressee interpret the utterance. (12)

They are playing over there. Katy has three little boys.

(13)

They are playing over there. But Katy actually has THREE little boys.

We could imagine that the two speakers are watching some children play, and (13) is in response to the question Where are Katy’s children? The concessive clause, the inclusion of the adverb actually, and the emphatic THREE all work jointly to suggest that only two of Katy’s boys are present, and the first speaker may not be aware that Katy has three boys. Further, consideration of the context in which the utterance is made is crucial to the successful interpretation of the referring expression. These three examples suggest that there are more factors involved in the process of referring other than the various formal expressions involved in the referring process and their relative position in the co-text. Not only does there need to be some coherence between the clauses, but the context and discourse event as a whole are crucial to the full pragmatic interpretation of the utterances. In reality, a neat and tidy pairing with a lexical antecedent seems to be the exception rather than the rule. Another example of a type of anaphor which does not have an explicit linguistic antecedent is that of a ‘conceptual’ anaphor (Oakhill et al., 1992: 257). These anaphors construct their antecedent using cultural knowledge, as illustrated in (14). (14)

I need a knife. Where do you keep them?

In this situation, them means ‘the knives that I presume you have in your house’ (ibid: 258). According to Gernsbacher (1991), this kind of anaphor is often used to refer to members of a collective set, as in (14), and is in fact rated very ‘natural’ and readily understood by readers. The fact that they are easily understood by addressees demonstrates the important role that cultural knowledge plays in the interpretation of reference. A co-textual interpretation would not be able to account for these types.

1.3 On Text and Discourse

15

A further illustration of the notion of ‘antecedent’ not being able to account for the realities of language in use is ‘associative anaphora’ (also known as inferable reference (Gundel, 1996: 141), bridging reference (Sanford and Garrod, 1981: 94; Martin, 1992: 124), or an association set (Hawkins, 1991: 409)). Haviland and Clark (1974: 514) use example (15) to illustrate this phenomenon. (15)

Mary unpacked the picnic things. The beer was warm.

In this example, the anaphoric expression the beer does not refer directly to the antecedent the picnic things, but an addressee would be unlikely to have any difficulty in interpreting the referent the beer as referring to part of the picnic. Givón (1995: 351–352) suggests that this is a hybrid type of grounding, which he calls ‘double-grounded frame-based reference’. Accessing the referent, then, is partly through an anaphoric connection to the referent’s ‘episodic trace in the episodic representation of the current text, and partly through connections to generic-lexical knowledge’ (‘episodic memory’ refers to the longerterm recall of a mental text-trace (ibid: 344)). In example (15), the definite expression the beer receives its anaphoric grounding in part from the preceding textual antecedent the picnic things and in part from generic-lexical knowledge of the frame picnic things4 and its sub-component beer. Further, the same explanation can be applied to metonymic anaphora (see Chapter 10), where the referent does not relate directly to the thing it represents, but rather to something closely related to it, such as in (16). (16)

Table 4’s waiting for their bill.

Here, the anaphoric antecedent for their is not available from the preceding text, but rather the definite expression Table 4 receives its grounding from the speech situation, the discourse representation set up contextually by the interlocutors. The inference works because, in the context of a restaurant, it is unlikely that servers know the names of their customers, and even less likely that they write the names on top of the bills in order to identify which bill is required; restaurant tables are often numbered for this reason. In examples (14) to (16), there are no canonical textual antecedents. Cornish (2010: 228) suggests that it is not the antecedent which provides the in-context interpretation of the anaphor, but rather the ‘antecedent trigger’, which is ‘a percept [. . .], an utterance token or a semiotically relevant non-verbal signal’ which is outside the co-text but in the discourse representation of the participants. This notion of an antecedent trigger has resonance with cognitive accounts of reference, such as Ariel’s Accessibility Theory, in which the

4

We use small caps throughout to indicate labels of concepts.

16

The Multidisciplinary Field of Referring

antecedent is considered to be a mental representation and not part of the cotext (Ariel, 1996: 17). Ariel indeed argues that the ‘accessibility of mental entities is not presumed to remain constant’ (ibid: 21), an obvious illustration of which can be seen with evolving reference, as we will see in Chapter 4. Givón (1995: 350) echoes this with the idea that the interlocutors have a preexisting mental structure of the referent, and the antecedent is accessible in ‘some extant mental representation’ (ibid: 376). 1.3.2

Discourse Perspectives

The discourse-functional approach considers reference to be a discourse-level referring procedure, involving a dynamic three-way relationship between text, discourse, and context, and rather than relying on a lexical antecedent for resolution, the anaphor is identified through a mental representation of the entity in question. That is, the discourse-functional approach emphasises the tracing of referents in the interlocutors’ respective discourse models of the communicative event, which are constantly evolving (Cornish, 2010: 208). This can be contrasted, to some extent, with Martin’s ‘participant identification’ and ‘referent tracking’ (1992), which largely depends on pairs of textinternal relations between an anaphor and an antecedent, as well as Halliday and Hasan’s particular notion of reference, which involves items making reference to something else for their interpretation (1976: 31). 1.3.2.1 Text, Context, and Discourse The discourse-functional account puts forward the idea that there are three ‘interdependent, interactive and interdefining’ concepts which play a crucial role in the successful interpretation of anaphora (Cornish, 2010: 210). These are text, discourse, and context, which Cornish describes as follows. Text refers to the connected sequence of spoken and non-spoken signs and signals produced by the interlocutors. The text serves as a holding bay of cues which act as instructions to the addressee to develop ‘a conceptual model of the situation being evoked by the speaker, a model which, [they intend], will be identical to the one which the speaker is building’ (Cornish, 1999: 34). In his account, discourse represents the mentally represented product of the ‘hierarchical, situated sequence of utterance, indexical, propositional and illocutionary acts’ carried out jointly by the interlocutors as the communication unfolds; the communicative event which is integrated within a specific context (Cornish, 2010: 209). Finally, Cornish sees context as the domain of reference of a specific text, the sociocultural environment and genre of the current speech event, and the surrounding co-text. The context is constantly revised and (re)constructed as the discourse is updated (ibid.). In Cornish’s account of reference, discourse is dependent on text and context. It is the discourse event that is likely to be stored in the long-term

1.4 Cognitive Perspectives: Reference in the Mind

17

memory of the participants for possible recovery at a later point. The text is a temporary construct, likely to be forgotten once the corresponding discourse has been created. This claim is supported by evidence from Sachs (1967: 437), who found that ‘the original form of the sentence is stored only for the short time necessary for comprehension to occur’, but recognition memory for the meaning of a sentence declines less rapidly. 1.3.2.2 Anaphoric Predication Anaphoric predication (Cornish, 1999, 2010) refers to the idea that what follows a referent downstream helps point to a particular referent, and at the same time disambiguates from any potential distractors. Compare the implicatures drawn from the use of the man in examples (17) and (18), from Wilson (1992) cited in Cornish (2010: 230). (17)

Sean Penn attacked a photographer. The man was quite badly hurt.

(18)

Sean Penn attacked a photographer. The man must be deranged.

In (17), we interpret the man as referring to the photographer, as the predicating component expresses the result of an action, in this case an attack, and can only apply to the receiver of that action. In the case of (18), however, there is an evaluation (must be deranged) on the part of the speaker about the cause of the attack (Cornish, 2010: 230). Therefore, it is the entire anaphoric predication which picks up the relevant discourse representation and enables the integration of the two discourse units and thus the successful interpretation of the anaphor. This is what Cornish suggests is the discourse contribution to anaphoric reference (ibid). Brown (1995: 64) supports this view, maintaining that the referring expression must be connected to whatever is predicated of it, that ‘the new information contained in the predicate must be filed under the correct “address in memory”’. So, in short, by incorporating discourse and taking a functional approach to chains of reference, the antecedent trigger, the predication of the anaphor, as well as the denotation of the anaphor within its immediate context all interact to provide the appropriate conditions for the successful interpretation of the expression. It is not simply a matter of a relation between two co-occurring expressions, as more formal approaches have argued. 1.4

Cognitive Perspectives: Reference in the Mind

The study of reference has also received considerable attention from a cognitive perspective. Reference, under this conception, examines whether the discourse partners have a similar mental representation of the entity in question at any one point in the discourse. Chapter 8 closely examines Ariel’s Accessibility Theory (1990) and Gundel et al.’s Givenness Hierarchy (1993),

18

The Multidisciplinary Field of Referring

so what follows is a brief overview of some of the main considerations of the study of reference from a cognitive perspective. The following is intended to provide a context in which Accessibility Theory and the Givenness Hierarchy can be understood. The various statuses that expressions may have are discussed by Chafe (1976: 27) in terms of ‘the speaker’s assessment of how the addressee is able to process what [they are] saying against the background of a particular context’. He suggests that not only are we able to store a great deal of knowledge, but we also have ‘temporary states’ with relation to that knowledge (ibid). Therefore, speakers must adapt what they are saying to fit in with what they assume the addressee is thinking at that moment, and only then will the message be assimilated. The notion of ‘givenness’, for Chafe, is tightly connected to the idea of consciousness (1976: 30). Given (or old) information is ‘that knowledge which the speaker assumes to be in the consciousness of the addressee at the time of utterance’ (Chafe, 1994: 53), whereas new information is what the speaker assumes they are introducing into the consciousness of the addressee. This needs to be seen in contrast with what the addressee is expected or not expected to know; that is, the addressee may have knowledge of the entity but may not be thinking about it at the time of speaking, thus it is not considered to be ‘activated’. Finally, if speakers consider the idea to have been ‘semiactive’ in the consciousness of the addressee, then they are likely to verbalise it as ‘accessible’ information, but if an idea is considered to be previously ‘inactive’, then it will be verbalised as ‘new’ (ibid: 74). While Chafe refers to the status of the referring items in terms of activation, for Ariel, it is a matter of the degree of accessibility. That is, the speaker needs to ensure that the level of accessibility indicated by their referring expression corresponds to the level of accessibility associated with the addressee’s mental model of the discourse under construction. Further, the speaker must carry out a new assessment of the degree of accessibility each time a referral is made, even when it is to the same object (Ariel, 1996: 21). This is because it cannot be assumed that the accessibility of mental entities remains constant (indeed, we have already seen that the antecedent does not remain static). Thus, once speakers have made a judgement on the current degree of accessibility, they will then select an appropriate referring expression. If the referent is assumed to be at the forefront of the addressee’s mind and there are no distractors, then an accented pronoun (i.e. a pronoun with intonational prominence, e.g. Tom hit Joe, then HE hit Jacob) or zero anaphor (i.e. a reference position that is filled with a morphologically unrealised form, e.g. He said he loves her and ø wants to be with her forever)5 will be used. In this case, the level of cognitive

5

The ø symbol represents the absent he.

1.4 Cognitive Perspectives: Reference in the Mind

19

accessibility is presumed to be high. On the other hand, if the speaker judges that the intended referent is not in focus but is familiar to the addressee at that specific point in the discourse and is therefore ‘activated’, then this is considered to have an intermediate level of accessibility, and a demonstrative pronoun or accented third person pronoun is likely to be used. Finally, if the speaker deems the referent to be inactive, it may have been active earlier in the discourse and succeeded by another referent, or it may be in the long-term memory of the addressee. In this case, a long definite description or full name will be used (Ariel 1996: 21–23; see also Chafe, 1976 and Givón, 1995: 50). Two features dictate the level of saliency or accessibility of the mental entity acting as a linguistic antecedent in Ariel’s (1996) hierarchy: i. Some mental entities are more salient: the interlocutors, sentence and discourse topics, repeated topics, and human and animate objects. ii. The relation between the mental representation of the antecedent and the referring expression. For example, the distance between the last mention of the antecedent and the anaphor, with recent mentions being more accessible. Also, the level of cohesion plays a role; the more cohesion there is between the two discourse units, the more accessible the referring expressions are (ibid: 22). Significantly, none of the referring expressions in Ariel’s hierarchy need to be paired with a co-textual antecedent in order to be interpreted successfully, as traditional approaches to anaphora claim they do. Rather, an antecedent is a mental representation which is ‘stored in the addressee’s memory in different degrees of accessibility’ (Ariel, 1996: 17). However, there appears to be an inconsistency here because Ariel seems to view all referring expressions as anaphoric, irrespective of the context of their use, and argues that the ‘referential-anaphoric distinction’ should be abandoned (ibid: 7). This position appears to disregard the basic referring ability of referring expressions, which is illustrated by her reliance on distance as a measurement of accessibility, despite stating that there are four criteria which contribute to the accessibility of an antecedent. Antecedents are, as a result, things that are mentioned in the linguistic context and are therefore linguistic entities. It is unclear how Accessibility Theory would account for references to indexicals in the physical context (e.g. pass me that) since accessibility cannot be measured in terms of distance from the previous mention. This not only seems to contradict her claim that antecedents are mental representations, but also limits her approach to coreference (see Reboul (1997) for a detailed critique of Accessibility Theory). Further, one possible effect of relying on factors such as distance and disambiguation to predict the wording of referring expressions is to imply that identifying the referents of referring expressions is the only function performed by these noun phrases. We will see in Chapter 4 that this is not the case.

20

The Multidisciplinary Field of Referring

Another influential contribution from a cognitive-pragmatic perspective comes from Gundel et al. (1993), who propose a Givenness Hierarchy of six cognitive statuses relating to the form of referring expressions. What is different about this hierarchy is that it is implicational, and each status on the hierarchy ‘is a necessary and sufficient condition for the appropriate form or forms’ (ibid: 275). That is, when using a specific form, the speaker indicates that they assume the relevant cognitive status is met, and since each status entails all lower statuses, the speaker also indicates that all lower statuses have been met. So, for example, if an entity is ‘in focus’, that is, the speaker believes the addressee knows exactly which referent is being referred to, they will formally represent this status by a pronoun, for example, it. In this model, if the referent is ‘in focus’, then it is also ‘activated’, ‘familiar’, ‘uniquely identifiable’, ‘referential’, and ‘type identifiable’. We will discuss these two models in greater detail in Chapter 8. 1.4.1

Summary

It is clear from this overview that reference has attracted a great deal of scholarly interest from various perspectives. We saw that the acts of producing and interpreting referring expressions are complex and dynamic and cannot be fully understood from just one position. By combining insights from the distinct, but related, approaches examined above, a more balanced and credible account of what speakers and addressees actually need to know and do when involved in the referring process can be achieved. By including text, context, and discourse event together with cognitive and psycholinguistic factors, a more comprehensive understanding can be achieved. 1.5

Volume Aims and Structure

1.5.1

Aims of the Book

The aim of this volume is to provide an integrated approach to referring expressions, one that considers the lexicogrammar of referring expressions alongside cognitive and contextual factors. Current literature on referring tends to be restricted to typical reference and only pays passing attention to marked or atypical uses of referring expressions. The volume bridges the gap by offering an integrated perspective on referring where atypicality is examined in a multilayered coherent approach to reference. In order to gain a full understanding of the nature of referring, a multifaceted approach needs to be taken; any account of referring expressions should take into consideration insights from the discourse, cognitive, and psycholinguistic literature because each of these perspectives can provide a level of understanding which helps

1.5 Volume Aims and Structure

21

develop the whole picture of the complex process of reference production and interpretation. As with many other accounts of referring, we view reference as addressee orientated, collaborative, and context dependent. Both participants in the exchange (speaker/writer and addressee) have a responsibility to negotiate meaning and understanding, but this does not always work, and it does not always work in a typical or expected way. After establishing our framework and what we mean by typical reference and referring expressions, this book will focus on instances of referring in use where the encoding is incongruent with the intended use of the expression and where ratification of the expression is not always predictable or successful. 1.5.2

Volume Structure

The chapters making up this book have been grouped into three parts. The first part provides a foundation for the study of reference and referring expressions. Following this introductory chapter, Chapter 2 examines the main debates related to definite versus indefinite expressions. As we show, a given expression does not always correspond directly to the conventional referential status for the expression. While an expression may be marked as definite or indefinite, specificity and identifiability are meanings construed by the speaker. This chapter will also establish the use of key terms in the volume. Chapter 3 provides a lexicogrammatical description of referring expressions from a functional and structural perspective. Here, we consider the elements of the noun phrase and their role in acts of referring, for example determiners and categorisation. The purpose of the second part is to examine the main features of typical referring expressions and their use. We use Chapter 4 to establish what we mean by typical and atypical expressions by examining related terms from various perspectives. The remaining two chapters in this part focus, respectively, on spontaneous discourse in Chapter 5 and prepared discourse in Chapter 6. These two chapters provide an empirical basis for establishing typical referring in spontaneous informal language use and in planned written genres, respectively. In Chapters 5 and 6, we argue that the degree of spontaneity in language production plays a significant role in determining the nature of referring expressions. The third and final part provides a collection of chapters detailing various ways in which speakers exploit referential conventions to satisfy a particular communicative goal. First, in Chapter 7, we focus specifically on problems in referring, including how these problems get resolved. The main focus of the chapter is on uses of reference in contexts which are discursively noncollaborative, and here, we will be looking at examples from legal discourse,

22

The Multidisciplinary Field of Referring

with a particular emphasis on type specification. Chapter 8 provides evidence for the specific use of a particular type of indefinite expression in journalistic discourse. We show how writers can create a textual and contextual shared space with their readers which allows an indefinite expression to be understood as uniquely identifying. Chapter 9 focusses on the use of referring expressions both by and with children, including a brief overview of how reference develops in children’s language. In addition to examining the role of overspecification and underspecification in referring expressions, we also consider creative atypical uses of referring expressions as found in children’s literature. In Chapter 10, we consider metonymy as a particular atypical referring expression. Using examples of metonymy from medical discourse contexts, including nurse handover discourse and medical students’ self-reflections, we reveal the complex landscape of metonymic reference, which as we show, is essential in some contexts but stigmatising in others. The volume concludes with an epilogue which summarises the volume and raises some questions about where this volume might lead.

2

The Great Debate Definiteness and Indefiniteness

In London, a person gets mugged every ten minutes. And he’s getting mighty sick of it!

2.1

The The

The humour in the above joke lies in the expectations set up through the choice of referring expression in the first sentence; more specifically, the ambiguity inherent in the indefinite article a. We do not expect to be able to identify who gets mugged every ten minutes in London. We expect it to mean any person, and so we are surprised when we learn that one particular (rather ill-fated) person is the sole victim of these regular attacks. We know this because we unexpectedly encounter the definite pronoun, he, which can only refer back to the indefinite nominal expression in the previous sentence. The above joke illustrates the complexities involved in the field of reference, and this is perhaps one of the reasons why the study of reference and (in)definiteness has attracted considerable attention from various academic disciplines that are engaged with the study of language and communication, ranging from linguistics and philosophy to language acquisition, psycholinguistics, and more recently computer science. Moreover, reference is at the heart of human communication: when acquiring language, one of the first things children do is learn how to refer to objects (Matthews et al., 2007; also see Chapter 9). Humour is not the only site for our sensitivity to definiteness and indefiniteness. We are generally very good at managing when we should use an indefinite expression or a definite expression. For example, in a conversation with someone, if the speaker intends to refer to another person, they cannot use their name unless they know the addressee also knows this person’s name. If we compare the invented examples (19) and (20), and if we know that the addressee has no idea who Jane is, then we can easily imagine the different responses that the addressee would give. For example (19), the addressee will almost certainly say ‘who’s Jane?’ and for (20), the addressee would likely offer a suggestion about a good real estate agent. 23

24

The Great Debate: Definiteness and Indefiniteness

(19)

Jane is looking for a good real estate agent.

(20)

Someone I work with is looking for a good real estate agent.

To illustrate this point with naturally occurring conversation, let us consider the two excerpts presented below. Both are transcripts from an interview with author Margaret Atwood (MA), who appeared on the BBC programme HardTalk on 3 October 2016, hosted by Stephen Sackur (SS). In Excerpt 1, Stephen Sackur is asking her about her position on environmental issues, given her involvement, over many years, as a climate change activist. When he uses the expressions the Canadians and your people, we might assume he intends to refer to all Canadians all over the world, or perhaps only those in Canada. She immediately takes issue with the expression your people, which, by all accounts, would be viewed as a definite expression. Despite the use of the possessive pronoun (your), she is rejecting the expression as not specific enough. Her people are a diverse collection of different people, as she very clearly points out. What might Stephen Sackur have asked instead? He might have asked, “how do various groups of Canadians react to you?” Had he done so, he would have produced an indefinite expression by using a plural noun without a specifying determiner. Excerpt 1 There isn’t a ‘the people’ ss:

ma:

When you talk about these issues inside Canada, I just wonder how the Canadians react to you. Do you feel at one with your people? Or someone out of sync with your people? Well, which people? No, uh, there’s isn’t a ‘the people’. So there are these people here and those people there and these people up here and those other people over here. If we are talking about mowing through indigenous people’s rights in order to do this stuff, I would be with the indigenous people. If we’re talking about we have to shut down all production and consumption of oil immediately, that’s actually just not practicable. So how are you going to do it? And I think – ‘at one with my people’ – I think most of ‘my people’ would accept both of those points of view, or at least 80% of them would.

We find something similar in Excerpt 2, which occurs towards the end of the interview. Here, Stephen Sackur is interested in finding out what Margaret Atwood might predict about the future. Note that he does not actually say ‘the future’ but rather implies it with the expression, a picture of human society in the rich world a generation from now. His use of the indefinite article should have suggested that his intended referent is not specific, and that any example of what life might be like in years to come would suffice. However, Margaret Atwood wants to clarify before responding that she rejects the notion that there is such a thing as ‘the future’. We cannot know why she does this. Perhaps it follows from the exchange moments earlier that we presented as Excerpt 1, or

2.2 Reference and Referring

25

perhaps she simply wants to emphasise her point. Irrespective of her motivation and intentions, the two excerpts combined show clearly how sensitive we are to the interplay between definite and indefinite reference. Excerpt 2 There is no ‘the future’ ss:

ma:

In previous books you’ve speculated about what’s going to happen over 20, 30, 40 years, what’s go– . . . Paint me a picture of human society in the rich world a generation from now. OK. So, there is no ‘the future’. There isn’t any one ‘the future’. There are an infinite number of possible futures, and as Donald Rumsfeld said, and it was probably about the only thing he said that I agree with, it’s the unknown unknowns that get you. So we don’t know what the unknown unknowns are because they are unknown, but leaving them aside, should we continue down the road that we are on, the biggest threat to us as a species would be the death of the oceans. And the reason that that is the biggest threat for us, ‘cause we’re not plants, we breath oxygen, and 60 to 80% of the oxygen that we breathe is created by the marine algaes, as it was created in the beginning, this did not used to be an oxygen atmosphere. So kill the oceans, you will choke to death.

Reference and reference-related concepts can mean different things to different people. That may seem like a vague and somewhat unhelpful way to start a chapter claiming to discuss, and perhaps get to the bottom of, what reference is, but it also highlights the main problem. That is, there are no truly agreed-upon definitions and uses of the myriad terms and concepts that accompany the study of reference. All theories of reference would agree that there is an important distinction between indefinite and definite expressions, but the issues at the boundary between the two often require some clarification. As such, the aim of this chapter is to unpick the complex, confusing but often colourful stitching that makes up the reference tapestry, and then put it all back together in a way that is hopefully less complex and confusing but maintains its vibrancy. This chapter is organised as follows. In the next section, we establish how we view the terms reference and referring. Section 2.3 is concerned with the key areas of debate around the concepts of definiteness and indefiniteness, and after discussing these we examine how these concepts interact with the concept of identification. Given that proper names are generally assumed to be the classical definite expression, one that is uniquely identifying, we briefly discuss in Section 2.4 some variation in terms of their use, including metonymic reference and identification of a type of referent. We end the chapter with a brief summary in Section 2.5. 2.2

Reference and Referring

Before moving to our discussion of definiteness and indefiniteness, we would like to clarify what we mean in this volume by reference and referring. We saw

26

The Great Debate: Definiteness and Indefiniteness

in Chapter 1 that different disciplines view reference differently, but we also find differences within disciplines. This is true for functional linguistics, which is a broad term for a general area of linguistics, including various functional and cognitive frameworks. Throughout the volume, we will be drawing principally on functional accounts, supplementing where appropriate with work from related disciplines such as psycholinguistics. Functional linguistics is particularly well-suited to the study of reference because, as Butler (2005: 5) explains, for functionalists, ‘both the language system as a whole and the formal structures it uses are, in large part, motivated by external factors such as cognitive structures, processing constraints and social factors’. Of course, different functional theories prioritise certain factors over others, for example cognitive linguistics (e.g. Langacker, 1987) prioritises cognitive structures, whereas for systemic functional linguistics (e.g. Halliday, 1985) social factors dominate. Functional theories of language are those which ‘are mainly concerned with examining the question of why language structure is the way it is and with finding explanations in language use’ (Bischoff and Jany, 2013: 1). See Butler and Gonzálvez-García (2014) for a detailed comparison of sixteen different functional frameworks. In the previous chapter, we explained that an understanding of reference requires a multidimensional approach which considers the expression itself, the discourse in which it functions, and the cognitive representation it has for the speaker and the addressee. In this sense, reference can be seen as a fourway relation between (a) a speaker, using (b) an expression to identify (c) an entity in order that (d) an addressee is able to recognise the entity in question. For example, David (speaker) might say ‘I would love to have that beautiful pink and purple neck adornment’ to identify a particular tie so that his friend (addressee) can recognise it (and perhaps buy it as a present for his birthday). Under this conception, reference is assumed to be largely a pragmatic phenomenon as it concerns the speaker’s use of linguistic expressions to identify the entities in question (Abbott, 2010: 2). Linguistic, cognitive, and discourse factors contribute to the speaker’s choice of the form of the linguistic item selected to carry out this task, as well as to the addressee’s interpretation of the expression. In the philosophical tradition, as we saw in Chapter 1, reference is said to concern the relationship between linguistic terms and entities that exist in the ‘real world’. The sentences presented in examples (21) and (22), taken from Givón (2015: 308)1, provide a good illustration of the problem of the validity of the statement. According to this tradition, the subject of (21) genuinely refers as it connects the linguistic term the King of England to something that 1

We have modified the original example because the current monarch in the UK is now King Charles III.

2.2 Reference and Referring

27

exists in the real world (i.e. the current king, Charles III). On the other hand, the King of France in (22) does not refer to an existing entity so cannot be said to be truly referring. (21)

The King of England is bald

(22)

The King of France is bald

As was discussed briefly in Chapter 1, the validity or truth value of the expression is not a concern for linguists, especially since language is particularly designed to allow us to make reference to things that might be imaginary or invented and we can say things that are not true. There could be a context, fictional or not, where there is a King of France and from a linguistic perspective we do not need to care whether the king is bald or not. The expression itself allows the speaker to share a conceptualisation of an intended referent with the addressee. The nature of the relationship plays a role in how reference is defined. 2.2.1

Textual Reference

Perhaps the best-known work on reference within functional linguistics is found in Halliday and Hasan’s (1976) seminal work on cohesion. However, reference as used by Halliday and Hasan is much narrower than how we have been using the term so far. For Halliday and Hasan, reference is to be distinguished from lexical items which have a semantic interpretation ‘in their own right’ as opposed to those items, that is, reference, which are ‘directives indicating that information is to be retrieved from elsewhere’ (ibid.: 31). This use of reference includes pronouns, demonstratives (including non-selective the), and comparatives, which as function words are considered to have no semantic content of their own. We can see this distinction clearly in example (23), from Fontaine (2008), where co-referential expressions are marked respectively with the index i or j. Not every use of he in this example has the same referent, but it is clear that the information required to identify the referent is given earlier in the text. There is one instance where the referent is unclear (in he also said, he could refer to either John or the doctor), but in this example it is not particularly important; it is clear who did the tests, who has a sinus infection, and who finally went (to the doctor). While all instances of he would be considered instances of reference, John and the doctor would not. This type of reference is best viewed as cohesive reference since it is a much more restricted type of reference than what we are interested in here. We will return to cohesive reference in Chapter 5, but for now we would like to highlight some of the issues that arise from this framework and clarify how it informs our understanding of reference more broadly.

28

The Great Debate: Definiteness and Indefiniteness

(23)

Johni went to see the doctorj and hej did some tests and hei?j? also said hei had a sinus infection. I’m glad hei finally went.

Co-referential forms in Halliday and Hasan’s (1976) sense thus direct the addressee to look elsewhere for their interpretation; if this is done within the text, as in (23), it is said to be an endophoric reference. Their framework also includes exophoric reference, that is, reference to the situational context, for example reference to the speaker (e.g. I, me) or an object (e.g. that). To explain endophora, Halliday and Hasan (1976: 52) state that ‘[i]t is characteristic of third person forms that they may be cumulatively anaphoric. One occurrence of John at the beginning of a text may be followed by an indefinitely large number of occurrences of he, him, or his all to be interpreted by reference to the original John’. This claim is highly debatable since we know that we do not typically find a long series of pronouns following a single proper name, whether in written or spoken form. As Brown and Yule (1983: 200) point out, this view appears to be ‘the view of an analyst who has worked and reworked relatively small chunks of text which are capable of being displayed on a single page’. Halliday and Hasan appear to be claiming that reference is a one-to-one connection between a content word and a function word, and all subsequent mentions of an established entity must be interpreted with reference to the original expression. We could also debate whether a proper name, such as John, should be considered more contentful than a pronoun, or have the same level of content as a lexical noun phrase (NP) such as the doctor. While Halliday and Hasan are very clear that they are using the term reference to capture one device that contributes to the textual meaning of the text, its texture, it is not a concept that is suitable for reference in the broad sense of referring. Reference is far more complex than this. The novel A Tale for the Time Being by Ruth Ozeki provides an interesting illustration from which to base this discussion. It traces the lives of two characters: a Japanese schoolgirl, Naoko Yasutani, and a novelist called Ruth. On the first page and a half alone, Nao refers to herself twenty times using a variety of expressions (Nao, I, me, every one of us, my present, my future, my kind of time being, we, Naoko Yasutani, myself, and so on). Over 432 pages, 163 footnotes, and 6 appendices, the characters’ lives unfold as we learn more about Nao and Ruth. If we take Halliday and Hasan’s claim to an extreme, then at the end of the 432 pages, when the reader reaches the final reference to Naoko Yasutani, the reader will still be interpreting the anaphoric expression by reference to the original form on page 1. This type of textual or cohesive reference will not be able to account for chains of referring expressions such as those found in Ozeki’s novel. As we have hinted at earlier in this volume, a convincing account of reference needs at least some input from processing models.

2.2 Reference and Referring

29

Indeed, Brown and Yule (1983: 200) suggest that if someone is telling another person about an individual they recently met, and then after first mentioning the individual, they continue to refer to this individual using a long chain of pronouns (e.g. she), it is very unlikely that by the end the addressee will be able to recall the exact wording of the introducing expression. The question they ask is whether this would mean that the addressee would not be able to interpret the references. Brown and Yule (1983: 200) argue that it is rather that the referent is established through a mental representation and not the wording (see also Van Langendonck, 2007: 86). The view of the referent as a mental representation rather than something anchored in the text or in the ‘real world’ is also the position of several functional linguists. Fawcett argues that a referent is ‘a more or less welldefined mental construct to which the performer (speaker) wishes to refer’ (1980: 90). This position is also echoed by Rijkhoff (2002: 27), who explains that the referent is in the mind of the speaker and ‘not entities in the external physical world’. As such, ‘referents of NPs are rather mental representations of entities as they are created, stored, and retrieved in the minds of the speech participants’ (ibid.). A recurring perspective that comes along with this view is that ‘speakers rather than NPs are capable of referring’ (ibid.: 28). We will now explore this perspective a bit further in the next section with a focus on the notion of discourse reference, but we do return to the cognitive representation of the referent in later chapters. 2.2.2

Discourse Reference

For Du Bois (1980: 208), ‘a noun phrase is referential when it is used to speak about an object as an object, with continuous identity over time’ (emphasis in original). Du Bois’ use of the term ‘object’ is not meant to include only concrete things: it could also be abstract: it could be known or unknown, or in an imaginary or hypothetical world. Further, referential noun phrases ‘may be thought of in cognitive terms as either activating a mental “file” for some object [. . .] or referring back to a previously opened file’ (ibid.: 209). In contrast, something is non-referential if it is the ‘quality defined by the noun’ that is exploited, where the attributes of the particular noun are abstracted, as in ‘he looks like a uh. . . Chicano American’ (ibid.). In Du Bois’ example, there is no intention to actually give concrete meaning to the noun phrase ‘a Chicano American’, but rather to the qualities of appearing to have the physical qualities of a person with these particular origins, whatever they might be. This definition then establishes that the focus is on the object itself, but also that speakers establish then modify a mental representation of the referential item, rather than make reference to the original form of the expression.

30

The Great Debate: Definiteness and Indefiniteness

A similar position is taken by Martin (1992), who also distinguishes between nominal expressions which can be referring (i.e. can be used by a speaker to refer) and those which cannot. Rather than referent, Martin uses the term participant, which he defines as ‘a person, place, or thing, abstract or concrete, capable of functioning as Agent or Medium in transitivity’ (ibid.: 129). We can gloss this in our terms as follows: an expression can be said to be referring if it can be used to express an agentive subject of a clause. Martin’s definition would clearly exclude attributive expressions such as gentle in a clause such as the puppy is gentle since adjectives cannot express an agentive subject, at least not as the only item in the expression.2 However, as Martin points out (ibid.: 129), ‘all participants are realised through nominal groups but not all nominal groups realise participants and some nominal groups realise more than one’.3 Following Du Bois (1980), Martin considers certain noun phrases as not having any referential potential in the sense that they do not ‘realise participants’ (1992: 129). Excluded are nominal expressions such as attributive expressions, structural, or empty third person pronouns such as it, idioms, and certain indefinite nominals ‘under negative scope’ since ‘they do not introduce participants to a discourse’ (ibid.: 130). Martin’s example (ibid.) to support this latter claim is given here as example (24), which is somewhat of an exception to his definition of participant since ‘anyone’ can express an agentive subject (e.g. Anyone can open this jar). However, the question raised here is whether the expression involves a discourse referent. The classification of the expression used here, as well as the discourse in which it occurs, is significant. For the purposes of this discussion, we can compare invented examples (25) and (26) with Martin’s example. In (25), we would want to consider the nominal expression as referential, that is, the expression the boy is being used to refer to a specific and identifiable person. Despite the use of the indefinite pronoun anyone, (26), with a bit more co-text, suggests that anyone includes John and is, in this sense, co-referential with John. (24)

He didn’t see anyone there

(25)

He didn’t see the boy there

(26)

He ran into the house looking for John but he didn’t see anyone there.

From the discourse perspective, and based on the discussion of textual reference above, it is clear that there is more involved with reference than a 2

3

We exclude from this expressions such as the poor which clearly can, with the definite article, be used to refer and can express an agentive subject of a clause, for example, The poor have raised awareness about the risks of higher gas prices (invented example). Within the framework of systemic functional linguistics, the term ‘nominal group’ is used to refer to the nominal unit. Throughout this volume, we will however use the term ‘noun phrase’ unless it is within a citation where ‘nominal group’ is used.

2.2 Reference and Referring

31

relation between an expression and an entity in the ‘real world’. According to Schiffrin (2006: 35), referring is ‘multi-faceted’ and involves some kind of interactive coordination between speaker production and addressee interpretation. This view corresponds well to the picture of reference and referring that we have been building so far. Referring is an act; it is something a speaker does. Schiffrin (2006: 36) uses the term ‘referrals’ to capture ‘communicative attempts by a speaker to evoke a referent (the idea a speaker has of something in the world) through a referring expression (a linguistic expression that can represent and evoke an entity)’. Referring can thus be seen as a cumulative and jointly constructed process which works together with previously gathered information and develops alongside the upcoming discourse (ibid). Other scholars in the discourse view of reference seem to agree with this way of seeing reference and referring. For example, Schouten and Vonk (1995: 2) explain reference as ‘the function of an expression in the discourse of identifying individuals or objects which may but need not be present in either the knowledge base of the speaker or the knowledge base of both speaker and hearer’. 2.2.3

Summary

The main difference between the textual and discourse approaches to reference relates to the object of study, for example, whether the perspective is trying to account for what gives a text cohesion or to explain the function of the expression in the process of referring within an ongoing discourse. As we stated in Chapter 1, our main interest in this volume is on those acts of referring that are, in some way, atypical, and our objective in this section was to establish what we mean by reference and referring. Given the emphasis made above on the nature of the referent being a mental construct, which is shared in an act of referring, we can modify the description of reference we gave earlier to account for this view. We are now in a position to refine our characterisation of reference as a fourway relation between (a) a speaker, who uses (b) an expression to refer to (c) a mental construct of a referent in order that (d) an addressee can form a sufficiently similar mental representation which enables the recognition of the speaker’s intention. Having now established reference as a particular function (cf. Schouten and Vonk, 1995), we can define the terms referent, referring, and referring expression as follows, adapted from Fontaine (2012: 45): Referent: a mental construct, or conceptualisation, of an entity (cf. Du Bois’s (1980) term ‘object’) that the speaker intends to refer to; in other words, “something that is brought into focus or attention by the speaker”, a concept in the mind of the speaker rather than an object in the world.

32

The Great Debate: Definiteness and Indefiniteness

Referring: a process created by a speaker who uses a linguistic expression to indicate, introduce, or identify a referent for the addressee. Referring expression: a linguistic representation4 (formal expression) used by a speaker to refer to a referent, most commonly expressed by a noun phrase.

Trying to explain what reference is can quickly become cumbersome. Reference is the relationship, referring is the act, and referring expression is the linguistic expression used by a speaker to refer. When we say ‘reference’, we are fundamentally taking the speaker’s perspective, even though the speaker will, in turn, be taking the addressee’s knowledge and role into account. We will be looking at many examples of the relationship between the speaker5 and addressee throughout the volume. We have hinted at the fact that reference can be definite or indefinite, and while these terms would appear to differentiate two discrete categories, like many other aspects of language, this is not actually the case. The distinction between the two terms is not mutually exclusive; there are areas between that are more difficult to classify with precision. The next section examines what constitutes the notion of definiteness and identification in more specific detail. 2.3

(In)Definiteness

As we mentioned earlier, for an expression to be used to refer, the speaker needs to have something (in mind) to refer to. Most theories of reference agree that indefinite expressions are used primarily for establishing or introducing a referent into the discourse, which may then be followed by definite expressions for subsequent mentions. This perspective leads to an association of indefinite with unknown and unidentifiable, and definite with known and identifiable. Schiffrin (2006: 36) puts it in slightly different terms, that of ‘accessing’ (first mention) and ‘maintaining’ (subsequent mentions) a referent (cf. Martin’s (1992) terms presenting vs presuming). However, it is also largely acknowledged that there are exceptions to this pattern, as we will see below. One key area of challenge is how the features of definiteness and indefiniteness seem to vary according to the ways in which they are used in the literature. In what follows we take a closer look at these two concepts. Much has been written about (in)definiteness, which could be seen as hardly surprising given that the is the most frequently used word in the English language. On the other hand, English is somewhat in the minority when it comes to explicitly marking indefiniteness; most languages do not in fact 4 5

Referring expressions are not restricted to linguistic expressions, but here we are only concerned with those expressed by the English language. We are using the term speaker irrespective of mode of production (see Chapter 5), that is, whether the mode is spoken, written, or signed.

2.3 (In)Definiteness

33

overtly encode indefiniteness (Lyons, 1999: 48). That said, languages which do not encode indefiniteness might rely on other distinctions to communicate similar functions. For example, Japanese can mark topic-hood (which is correlated with, but not equal to, definiteness) through the use of the particle wa, and further definiteness can be marked through the use of demonstrative pronouns (kore, sore, are – this, that, that (over there)) or demonstrative adjectives (kono, sono, ano + noun – this, that, that (over there) + noun), among other features. Finnish has no grammatical definiteness marking at all. It does, however, mark topic, focus, and telicity (i.e. the property of marking an event as finished or complete), and this gives enough information to work out whether a referent is definite or indefinite. Welsh has no indefinite article, and indefiniteness is signalled by the absence of any preceding article. What this suggests is that a distinction between the form of an expression and the function of the expression may help us better understand (in)definiteness as it relates to reference. In what follows, we will first discuss features of definiteness, followed by consideration of indefiniteness before bringing this section to a close with a brief discussion of identification as a potential feature of reference. 2.3.1

Definiteness

The range of interpretations of ‘definiteness’ extends from uniqueness (Russell, 1930), mental accessibility and identifiability (Givón, 1993a), identifiability, uniqueness and inclusiveness (Hawkins, 1978), unique identifiability (Gundel et al., 1993), and familiarity (Christopherson, 1939) to particularisation (Fawcett, 1980), specificity and identifiability (Halliday and Hasan, 1976), and recoverability (Martin, 1992). These notions themselves are notoriously difficult to define in any consistent way. For example, as Hawkins (1984: 649) points out, ‘[d]efining what it means for something to be “identifiable” is, however, no easy matter [. . .] an adequate definition of identifiability covering every single use of a definite description is probably doomed from the start’. There is also overlap in these notions. For example, Birner and Ward (1994: 96) consider familiarity and identifiability to have overlapping features. That is, for something to be identifiable, it typically must also be familiar. So, rather than being unique features of definiteness, familiarity and identifiability are most likely to be frequently expressed by the same expression. For example, a speaker who notices a shiny new car driving past can easily say to their companion, that car is amazing. It will be clear to the addressee that the speaker has a specific car in mind and that the speaker believes the addressee will be able to identify the referent despite the fact that this referent is not familiar to either the speaker or addressee. In what sense then is this expression an instance of definite reference?

34

The Great Debate: Definiteness and Indefiniteness

For Givón, definiteness ‘is inherently about knowledge by one mind of the knowledge of another mind’ (1989: 206), as well as how certain the speaker is about knowing it. So, definiteness is about the assumptions the speaker makes about what the addressee knows, believes in, is familiar with, or is able to identify. The speaker then chooses an appropriate referring expression that fits these assumptions (cf. Kibrik’s (2011: 15) concept of referential choice). This knowledge (or set of beliefs (Fawcett, 1980)) influences the formal nature of the expression, that is, the lexicogrammatical features which are best suited for successful reference. Chafe (1976: 39) argues that a speaker will use a definite expression if they assume the addressee knows and is able to identify the particular referent, adding that ‘identifiable would be a better term than definite’. The need for this distinction is because if the speaker thinks the addressee knows and is able to pick out the correct referent, then the speaker will encode it as definite; thus, identifiability is an inherent feature of definiteness. However, Chafe also acknowledges that ‘we are stuck with the traditional label’ (ibid.: 39). He integrates definiteness with identifiability, which he divides into three components: shared knowledge between the speaker and hearer, which may be either direct or inferable; the use of language which is sufficient to identify the referent; and the saliency of the referent within the specific context (ibid.: 93–94). Another feature frequently associated with definiteness is the status of given information. Definiteness and givenness often go together (as in anaphoric reference chains, for example), but equally it is possible to find instances where the item is both definite and new, as in I talked with the carpenter yesterday, where definiteness is established through shared cultural knowledge rather than by a prior mention. Givenness tends to be associated with clause initial positions, typically as the subject of the clause. This point of departure for the clause is referred to as the theme element for Halliday (1985), where the post-theme portion of the clause is called rheme, generally corresponding to new information. As Fries (1983: 144) notes, ‘[t]he theme-rheme organisation of the sentence forms part of a larger pattern which governs the flow of information’, and most accounts of theme (e.g. Halliday, 1985) suggest that there are observable relations between theme and definiteness (Crompton, 2008; see also Xu, 2023). That is, referents within the theme of a particular syntactic unit will tend to be identifiable or recoverable for the addressee (see Chapter 5 for discussion of Du Bois’ (2003b) preferred argument structure). On the other hand, referents in the rheme have a tendency to be nonidentifiable or recoverable. In other words, themes tend to be familiar and given whereas rhemes are unfamiliar and new, which means that there are likely to be more markers of definiteness in utterances which are thematic, and more markers of indefiniteness in rheme elements (Crompton, 2008). However, in a corpus analysis of forty texts, Crompton found exceptions, suggesting that there is not necessarily a typical pattern of use.

2.3 (In)Definiteness

35

Before moving on to our discussion of indefiniteness, we would like to consider an example of personal pronoun use which normally is considered as both recoverable and identifiable but most of all as a definite reference. The example of interest is taken from Fontaine and Jones (2020), which analysed shifting reference in reconciliation talk6. The excerpt given in example (27) is taken from a video7 narrative by Chief Dr Robert Joseph, a Hereditary Chief of the Gwawaenuk First Nation and a residential school survivor. The video is part of the Canada is . . . online series, which showcases the different facets of Canadian identity. In the last section of the narrative, we argue that Chief Joseph’s use of we in the instances highlighted in bold are specifically reconciliation-related since he uses the pronoun we to show he is able to speak both as a member of the Aboriginal peoples, as shown by the double underscored instances (we), contrasted with the reference to non-Aboriginal Canadians (Canadians highlighted in grey), and as a member of all Canadians (we). However, the first we used in this excerpt (so we’re trying . . .) is arguably ambiguous. Who is the intended referent? (i.e. who is trying to look through a new lens?) Is it all Canadians or Aboriginal peoples? The complexity of the reference in this case blurs the boundaries, making the referent less definite. The ambiguity appears to be picked up in the next use of we, where Chief Joseph makes clear that we and Canadians refers to two groups, and then clearly these two distinct referents are brought together as one. The definite nature of the referent may not always be an advantage; some degree of indefiniteness may allow for the ‘notion of “togetherness” [that] is highly salient in Chief Joseph’s text’ (Fontaine and Jones, 2020: 199). We therefore cannot say that all definite expressions are instances of definite (identifiable) reference. (27)

Canada, by the way, is the only Western country that has had a Truth and Reconciliation Commission, so we’re trying to look through a new lens. We and Canadians, we as an Aboriginal, we celebrate each other, everybody cheering each other up as we move toward a more equal prosperous future for all of us.

In a sense, the flip side of the definiteness coin is indefiniteness, and we will turn our attention to a lack of definiteness in the next section. However, as we showed in our discussion of example (27), there is some middle ground that clouds our understanding of these two concepts which are generally thought of 6

7

Reconciliation talk stems from the Truth and Reconciliation Commission which was established in 2008 as the result of a very large class-action settlement brought about by the Aboriginal peoples of Canada in relation to the atrocities of the residential school system. The video is called Namwayut: we are all one. Truth and reconciliation in Canada and is available online at www.youtube.com/watch?v=2zuRQmwaREY (last accessed 12 March 2022).

36

The Great Debate: Definiteness and Indefiniteness

as mutually exclusive. We need to be clear in our use of definiteness whether we are making a statement about a definite marker, for example, the definite article, or a proper name, or about a type of reference. As we will see in the next section, the same is true for indefiniteness. 2.3.2

Indefiniteness

While it is actually quite difficult to address the notions of definiteness and indefiniteness as two separate items because of their inherent interconnection, we will attempt to do so here. This is because we feel that indefiniteness merits a closer examination than it would receive if bundled together with its definite counterpart. Traditional accounts of reference argue that a speaker cannot be said to be referring if they use an indefinite expression. For example, de Haan (1987: 172) firmly states that ‘no indefinite NP can refer to an object which is or has been identified’. Further, indefiniteness is considered to be semantically undefined, non-specific, and non-referential (Halliday and Hasan, 1976: 70). Indefinite expressions usually introduce new topics (Givón, 1983: 10), do not refer to identifiable entities (Givón, 1993b: 174), and are not available (Givón, 1983: 10); nor are they recoverable (Martin, 1992: 98). They are nonanaphoric (Hawkins, 1978: 17–18), but as Du Bois (1980: 208) points out, they can be referential-specific, that is, the speaker is referring to a specific entity but cannot expect the hearer to be able to identify which one (cf. epistemic status of the referent). We can see an example of this in (28), a statement made by the character Watson from the BBC programme Sherlock. This example highlights nicely the difference between specific and non-specific uses of the indefinite article. (28)

I’m looking for a friend. A very specific friend. I’m not just browsing.

Clearly Watson knows who he is looking for (as do the viewers), but the addressee is not expected to know. Interestingly, Spanish, which has a distinct way of encoding this kind of specific indefinite reference, can translate I’m looking for a friend in two different ways. The first estoy buscando un amigo has a non-specific reading, whereas the second, estoy buscando a un amigo has a specific reading; the speaker has a particular friend in mind. At this point, we are perhaps able to start drawing a distinction between the terms specific and identifiable: if something is specific, the speaker knows the identity of the thing being referred to, but the addressee is not expected to, but if something is identifiable, both speaker and addressee are thought to be able to pick out the referent. We will return to this later in our discussion. Bach (2008: 28) argues that indefinite descriptions do not refer but merely allude to something. To illustrate this point, he provides the following

2.3 (In)Definiteness

37

scenario: a group of ‘unsavoury men’ crashes a party one night, and an elderly reveller reports the statement given in (29) to the police. (29)

A big hoodlum had a concealed weapon.

Even though she has a specific hoodlum in mind, she does not specify which one. According to Bach, this would still be true even if she was mistaken about which ‘hoodlum’ was carrying a weapon. So, rather than this utterance being a ‘singular proposition’ (i.e. a proposition about a particular individual) about the hoodlum the speaker had in mind, it instead semantically expresses a general proposition. This is because ‘the police could understand her perfectly well without having any idea which hoodlum she has in mind. They understand that she has a certain hoodlum in mind, the one she is alluding to’ (Bach, 2008: 18). But, in this example, the addressee, that is, the police, ‘have no idea which hoodlum she has in mind’; they do not know the specific hoodlum and cannot identify them. But, we would argue that as the elderly reveller does have a specific hoodlum in mind, this would count as specific reference. As another example, consider the utterance in (30). (30)

An agency which should be protecting the natural world appears to have identified and aligned itself with people damaging it. (Guardian, 07/06/2012)

The question here is whether the addressee could know from (30) which agency is being referred to. Out of context, it is possible (if the addressee has knowledge of the particular environmental organisation in question, for example, but unlikely. But what if this sentence were not isolated and instead occurred halfway through a piece about that particular agency? Given the very specific context in which this expression actually occurs, the addressee may well be able to identify the referent of the expression as the previously mentioned, fully identified referent Natural England. It must be questioned whether the writer in instances such as (30) is merely alluding to a nonidentifiable entity or if the act is more like an act of referring. We will return to this discussion in Chapter 8. Indeed, Karttunen (1976: 367) maintains that indefinite NPs can have both a specific and non-specific interpretation. His example is presented here as (31). (31)

Bill didn’t see a misprint.

The noun phrase, a misprint, can be interpreted as either specific, as in there is a known misprint that Bill did not see, or non-specific, as in Bill saw no misprints (i.e. there weren’t any). Karttunen (ibid.: 368) further suggests that certain verbs lend themselves to either a specific or non-specific interpretation. Consider examples (32) and (33).

38

The Great Debate: Definiteness and Indefiniteness

(32)

John tried to find a piano [but he didn’t succeed in finding one]

(33)

John tried to lift a piano [but he didn’t succeed in lifting it]

As Karttunen suggests, there is something inherent in the meaning of the verb lift that leads us to a specific interpretation of the object in question, whereas find leads us to interpret the object as an example of the kind of thing John was trying to find. Interestingly, compare (34) with (32) and (33). (34)

John went to the librarian because he couldn’t find a book

It is unlikely this means just any book, so we also need to consider the role of the contextual semantics of the verb–noun phrase relationship as well as differences in the meaning of find. Nevertheless, it is the specific interpretation which allows for the establishment of a discourse referent (and hence sets up the possibility of continuity of reference), whereas the non-specific reading does not. Moreover, Chafe (1976: 42) suggests that indefiniteness and ‘newness’ tend to go together. The reason for this is that if an addressee is not assumed (by the speaker) to be able to identify the referent, then it is unlikely that the referent is already in the addressee’s consciousness. Therefore, the information cannot be considered to be ‘given’ (i.e. in the consciousness of the addressee) and needs to be introduced. However, in cases where the indefinite referent is different from the referent which established ‘givenness’, indefiniteness does not necessarily entail newness, such as in the examples (35) and (36) from Chafe (1976: 42). (35)

I saw an eagle this morning

(36)

Sally saw one too

Chafe maintains that the expression one in (36) is clearly given, but it is not the same referent as in (35). It is more likely that if Sally had seen the same bird, rather than the same type, then the utterance in (36) would be Sally saw it too. The given status of the expression stems from the fact that it is ‘a particular which is categorised in the same way as the other’ (ibid.: 42). This resonates to some extent with Hawkins’ (1978: 184) view that an indefinite NP can refer, but only when there is ‘at least one more such object in that [shared] set which the reference can exclude’. Thus, it is not possible to say ‘I saw a Queen Elizabeth II today’ because there are no other choices of referent in the particular set. Givón (1993a: 215) uses the same indefinite NP in two different examples to discuss the contrast between referring and non-referring expressions, which are given here as (37) and (38). (37)

John married a rich woman, though he didn’t know her well

2.3 (In)Definiteness (38)

39

? John married a rich woman, though he didn’t know any well

The difference between these two examples is that of ‘referential intent’. In uttering (37), by using the referring pronoun her, the speaker is committed to ‘the existence in the universe of discourse of some rich woman that John married’ (ibid.: 215). On the other hand, using the non-referring pronoun any in the same frame does not work. Now consider the utterances in examples (39) and (40). (39)

John wanted to marry a rich woman though he didn’t know her well

(40)

John wanted to marry a rich woman though he didn’t know any (well)

Once again, by uttering either (39) or (40), the speaker may or may not be committed to identifying a specific woman in the universe of discourse. Two interpretations of a rich woman are possible: a referring interpretation using her or a non-referring interpretation with any. Radden and Dirven (2007: 94) attribute the differences in interpretation with these utterances to having factual or non-factual reality. Like Kartunnen, Givón (1993a) proposes that a speaker can use an indefinite expression to refer, depending on the referential intent of the speaker. That is, whether the speaker is committed to the existence in the universe of discourse of a particular entity; the speaker need not only encode perceived hearer knowledge but can also encode their own position in terms of givenness/ newness as well as their own knowledge about the referent. This position is one we follow in this book. Givón (ibid.: 224) further argues that there is ‘a continuum of referential intent’. That is, the grammar of English is able to systematically encode referential intent. To illustrate this, let us consider the set of utterances in (41), taken directly from Givón (1993a: 225). (41)

Did you see anything there? Did you see anybody there? Did you see any man there? Did you see some man there? Did you see a man there? Did you see a tall man there? Did you see a tall man wearing a blue shirt there? Did you see a tall man there wearing a blue shirt and sitting on a red barrel and twirling a silver baton in his left hand?

Givón (1993a: 225) explains this continuum from both the speaker’s and addressee’s perspective. From the speaker’s perspective, the dimension is psychological and relates to how strongly the speaker intends to refer to a particular entity. From the addressee’s perspective, the dimension is probabilistic, that is, it relates to the probability that the entity being referred to is a

40

The Great Debate: Definiteness and Indefiniteness

Grammatical

Likelihood of specific reference

device

More likely……………………………………………………… less likely

Indefinite a

articles

>

some

>

any

Restrictive modification Lexical noun specification

more modification

Specific noun

>

>

person

less modification

>

thing

Figure 2.1 Givón’s coding gradation of reference (1993a: 225)

particular individual. He concludes by outlining three grammatical devices which appear to encode the continuum of referential intent, represented in Figure 2.1. It is important to note that the gradation of referential intent that Givón outlines relates to specific reference, that is, the speaker’s intention to refer to a particular individual. Givón does not suggest that the addressee is expected to be able to identify the particular entity, but simply acknowledge the existence of such. We will return to this notion of referential intent and indefinite expressions which may refer to identifiable entities in Chapter 8. An interesting distinction between (in)definiteness and (in)explicitness is made by Schiffrin (1994), who takes the position that expressions introducing referents into the discourse could be indefinite and explicit (e.g. a woman I work with), but then expressions which are used for subsequent reference to the entity could be definite and inexplicit (e.g. she) (1994: 198–199). Schiffrin views definiteness in much the same way as Givón, in that the speaker expects the addressee to be able to identify the entity ‘from whatever clues are available’, whereas explicitness ‘has to do with the presentation of information that actually enables H [the hearer] to correctly identify a referent i.e. the lexical clues that allow H to single out whom (or what) S [the speaker] intends to differentiate from other potential referents’ (ibid.: 199). Further, explicitness should be seen as a continuous process rather than a discrete distinction between what is explicit and inexplicit, and (in)definiteness and (in)explicitness can ‘crosscut’ each other; expressions can be definite and explicit (the father of my children), definite and inexplicit (he), indefinite and explicit (a man I met 19 years ago), and indefinite and inexplicit (someone I met when I was younger).

2.3 (In)Definiteness

41

Lexical explicitness seems to play an informative role, perhaps stepping in where the grammar does not help disambiguate potential referents. It could be the case that definiteness relates to the speaker’s intentions and assumptions about the addressee’s anticipated knowledge, but explicitness is perhaps motivated by the speaker’s cooperative intentions, providing information which facilitates the addressee’s identification of the referent. We have seen above several arguments for the possibility of specific or speaker reference, which suggests that an indefinite expression can be said to be referring if the speaker has a particular entity in mind when making the utterance. However, these descriptions indicate that indefinite descriptions do not have a specific and identifiable referent. In other words, in contrast to definite expressions, there is a distinction to be made in terms of the shared knowledge between the speaker and addressee. We will return to this theme in Chapter 8. It can be concluded that definiteness is not an inherent property of the expression but of the beliefs of the speaker vis-à-vis their addressee. Indefiniteness suggests that the speaker does not believe the addressee will be able to identify the referent, or indeed that there is no specific referent to be identified. On the other hand, we have seen that both definite and indefinite expressions can contain additional information which enables the addressee to identify an intended referent, and this ‘explicitness’ and definiteness can easily interact to reflect the intentions and assumptions of the speaker. 2.3.3

Identifiability

Before leaving the discussion of (in)definiteness, we should consider how noun phrases are categorised as definite or indefinite in English. Whilst ‘it is absurd to draw boundaries around phenomena under study and then use these boundaries to justify one’s intellectual stance’ (Halliday, 2000: 378), this is what seems to inevitably happen, and the phenomenon of (in)definiteness is no exception. Therefore, the criteria for the categorisation of an NP are generally based on the type of determiner that is included in the expression (see Chapter 3) and in some cases whether there is phoric element to the expression (e.g. anaphoric items, including the). Martin (1992: 98) suggests that in English, all noun phrases encode the identifiability (or ‘recoverability’ in Martin’s terms) of the referent ‘as explicitly recoverable from the context or not’. For Martin indefinite expressions do not encode the participant as recoverable, but ‘pronouns, demonstratives, the definite article and proper names signal that the participant’s identity is in some sense known’ (ibid.). In contrast, Schiffrin (1994: 198) states that a formal differentiation between definite and indefinite referring terms ‘is relatively easy’, but that it is the functional differences which are difficult to define. That is, the problem lies in

42

The Great Debate: Definiteness and Indefiniteness

revealing the conditions ‘under which different terms can be appropriately used’ (ibid.). Further, Chafe (1994: 93) acknowledges that although identifiability (and therefore definiteness) is often associated with the use of the definite article, it is ‘by no means always’ the case. This idea is echoed by Du Bois (1980: 203), who notes that simply defining these expressions in terms of the presence of a definite or indefinite determiner is not sufficient. According to Givón (1993a: 216), ‘the mere presence of the indefinite article a(n) in English does not guarantee either a referring or non-referring interpretation of the noun phrases. The indefinite article is indeed irrelevant to this feature of meaning’. Whilst the latter point may be considered a little extreme, these comments suggest that it is in fact the context and the conditions the speaker finds themselves in which can be the biggest indicator of (in)definiteness, not the mere presence or absence of a definite or indefinite determiner. According to Lyons (1999: 33–36), what makes an expression indefinite is the absence of a definite determiner, rather than the presence of an indefinite determiner. That is, a, as a putative indefinite article, does not encode indefiniteness, but it can signal it. This is made clear in examples (42) and (43), from Lyons (1999: 33). (42)

I bought three books this morning

(43)

I wonder if Helen has read many books

Although the indefinite NPs highlighted in bold in examples (42) and (43) do not contain the indefinite article, they are nevertheless considered to be indefinite because they do not involve either identifiability or inclusiveness. They instead contain ‘cardinality terms’ which do not necessarily encode indefiniteness. Lyons (ibid.) maintains that this is because they can co-occur with definite determiners, as can be seen in examples (44) and (45). (44)

Pass me those three books.

(45)

I’ve only read a few of the many books she’s written.

This claim is also illustrated by the fact that a only occurs with singular countable nouns, but not necessarily with all singular countable nouns; consider one cup. One is a cardinality term, not an indefinite determiner, and like other cardinality terms can occur with definite determiners, as in the one cup. Lyons concludes that cardinality terms are neutral in terms of (in)definiteness, but what makes the NPs in (44) and (45) indefinite is the absence of a definite determiner rather than the presence of something indefinite (cf. languages, such as Welsh, which do not have indefinite articles). These various types of determiners are discussed in Chapter 3.

2.4 What’s in a Name?

43

It seems useful, then, to make a distinction between the function of the expression and the formal categorisation of its structure. Lyons (1999: 33–36) is suggesting that for something to be definite, there must be some marker or indicator of definiteness. Consequently, NPs cannot be formally identified as indefinite; they can only be encoded for +/ definiteness. The formal coding of (in)definiteness is, in some cases at the very least, somewhat irrelevant to the (in)definite status of an expression; rather it is the textual, contextual, and discoursal conditions which perhaps indicate how an expression is to be interpreted. Before ending this chapter, we want to give some attention to proper names, which are traditionally viewed as the prototypical definite expression. 2.4

What’s in a Name? ‘What’s your name,’ Coraline asked the cat. ‘Look, I’m Coraline. Okay?’ ‘Cats don’t have names,’ it said. ‘No?’ said Coraline. ‘No,’ said the cat. ‘Now you people have names. That’s because you don’t know who you are. We know who we are, so we don’t need names.’

— Neil Gaiman, Coraline

The quotation above from Gaiman’s novel Coraline captures, to some extent, and with some humour, the philosophical view of proper names.8 These linguistic forms are unique identifiers; in other words, we use them to identify specific individuals. In this sense they seem incontrovertibly definite, and indeed they are considered ‘definite’ in most theories of reference. But are they even words? Are they always definite? In what follows, we will consider some instances that suggest the picture is not as clear as we might think when it comes to proper names. We might like to use a dictionary as the authority on words, including proper names. Proper names must surely be part of the mental lexicon. For Hanks (2013: 64) they are ‘special kinds of words, with special rules governing their role as conventional units of a language’. While it is likely impossible to estimate the portion of the mental lexicon that is taken up by proper names on an individual level, we do have a sense of relative size from very large databases, and, as Hanks (ibid.: 35–36) reports, ‘over 70% of lexical entries already are proper names, and this percentage continues to increase’. It is worth noting that proper names do not stick around in the language in the same way as nouns, for example; ‘most proper names tend to be ephemeral’ 8

See Van Langendonck (2007) for a useful discussion of the philosophical theories of proper names.

44

The Great Debate: Definiteness and Indefiniteness

(ibid.: 36). The perspective from lexicography is described very nicely in the following citation from Hanks (ibid.: 34, emphasis in original). Dictionaries that do not include names seem to be operating with some definition of word that implies that a word denotes a class of entities, events, or functions, rather than an individual though generally such dictionaries do not say this explicitly, either in the front matter or at the entry for word. In such dictionaries, names are excluded because they denote individuals, not classes. Dictionaries that include proper names, on the other hand, seemed to be operating, informally, with some definition of word that implies that a word denotes all items of shared knowledge in a culture, which includes at least items like Shakespeare and London. If you do not know that London is the capital of England and that Shakespeare is a famous poet and dramatist of the late 16th and early 17th century, you do not count as a fully-fledged member of the Englishspeaking community. Such knowledge is regularly exploited by speakers and writers using the proper name of a famous individual or place as a generic term for a set of individuals. The person or place exemplifies a salient member of a set, having the property most associated with that individual.

We can see an example of a proper name shifting temporarily into a class of entity in example (46), where the first instance of Charles is used in the traditional understanding of a proper name, that is, to identify an individual, but the second use is non-referential since it is used in an attributive description (see discussion above in relation to non-referential expressions). There is a tendency for proper names to become descriptors, that is, adjectives, as in Cartesian or Darwinian for example. As Hanks (ibid.: 35) explains, ‘proper names can also be the root of dynamic adjectival coinages’, but most of the time these expressions (e.g. a Charles) rarely last longer than the instance in which they are used (cf. Hanks’ notion of exploitation, which is discussed in Chapter 4). (46)

amy:

Jake, we need to focus on Cheddar. Charles is a grown man. He can take care of himself. jake: No, he’s not. He’s a Charles, all right? And he’s blind! Oh, God, he’s probably already dead. (Brooklyn 99, Cheddar, Season 3, Episode 18, aired 1 March 2016)

Given that proper names are not thought to have any meaning independent of the individual that they name, they are viewed as not having any semantic content. The uniqueness of proper names can be seen in the fact that we do not need to qualify or restrict their reference with the addition of restrictive modifiers – the proper name being used as such is sufficient. However, speakers of English certainly do modify proper names and sometimes treat them as attributes, as classes of entity, and sometimes a proper name is modified. Their implied uniqueness of the referent is reflected in the fact that proper names, when used as such, may not occur with restrictive adjectives, restrictive relative clauses, or other restrictive modifiers. Non-restrictive adjectives (e.g. clever) and relative clauses, as in (47) for example, are of course

2.4 What’s in a Name?

45

acceptable but do not serve to restrict the reference, which is already assumed by the use of a proper name. (47)

Clever Maria, who brought her own utensils, did not have to wait for the porters.

Prince (1992: 301) suggests that the use of a proper name generally supposes a hearer-old referent, that is, one presumed to be already known to the addressee under the name used. Presumably this is a result of the lack of overt descriptive content shown by proper names. The main point here is that proper names are assumed to be definite. As we will point out below, this assumption does not hold. For now, let us turn to example (48), where the proper names highlighted in bold are used referentially but no longer uniquely identifying. In expressing Imperial, Cambridge, and Oxford as plural common nouns, combined with the definite article, the, we find evidence to support Anderson’s claim (2003: 359) that ‘plural names are also notionally deviant in not necessarily ensuring singular uniqueness of identification’. (48)

Britain’s higher education system is not a research project into the limits of the market. It is a source of world-leading research and also of urgently needed skilled workers and technicians. At every level, from the members of the global elite, the Imperials, Cambridges and Oxfords, to Lincoln’s groundbreaking co-operation with Siemens and Hereford’s embryonic New Model in Technology & Engineering, these are places of innovation as well as education, a precious well of community empowerment and opportunity. They have a public value that goes far beyond the number of students they can recruit. And that is what should be at the heart of the government’s review. (source: www.theguardian.com/commentisfree/2018/nov/02/universitybankrupt-inevitable-market-forces, Friday 2 November 2018, written by Anne Perkins)

Further evidence that proper names are not solely used by speakers to uniquely refer comes from work by Dancygier (2009: 168), who presents a convincing case for the rich contextual frames represented by proper names. We can see an illustration of this in the use of Canada in examples (49) and (50), which exploits certain aspects of its frame and suppresses others, for example, the spatial reference remains but any institutional reference is suppressed. In both instances, Canada occurs in the noun phrase as a common noun, denoting a class rather than an individual entity. In other words, the use of my Canada allows for the mental representation of more than one Canada, or more than one kind of Canada, construing a class of thing that can be denoted by Canada and including some types of Canada while excluding others. Dancygier (2009) has shown that proper names used this way are not uniquely referring as would be expected; they instead ‘activate a unique frame of knowledge associated with the name in question’ (ibid.: 166). The effect of the possessive determiner

46

The Great Debate: Definiteness and Indefiniteness

in these instances contributes, according to Dancygier (2011: 231), ‘an experiential or attitudinal perspective of a specific individual’. We might therefore paraphrase examples (49) and (50) as ‘X’s version of Canada’. (49)

Being a Canadian means different things to different people. I am proud to be a Canadian because my Canada is about inclusion. (2011) www.mykawartha .com/opinion-story/3686565-my-canada-includes-everyone/

(50)

Trudeau’s Canada has plenty of critics, too. (2017) www.battleofideas.org .uk/2017/session/trudeaus-canada-a-liberal-nirvana/

Once we begin delving deeper into the use of proper names, we begin to see their semantic potential is much broader and more diverse than we might have assumed. Proper names can defy the limited role we assign to them by restricting them to unique, identifying, definite reference. Increasingly, the view of proper names is that they are ‘words just as others’ (Van Langendonck, 2007: 67; cf. Hanks, 2013). While we will not take space here to detail the various classifications of proper names, we do want to make one final point concerning the implications of a restricted referential view of proper names. The general assumption among scholars and teachers is, according to Klassen (2018, 2021), that proper names are known, given, and highly accessible. She explains the fact that this is not always the case raises questions for first and additional language learning. The approach she proposes (2021: 6) is as follows: Rather than assuming names are known, a more cautious approach would be to categorise proper names in terms of frequency, as is done with other vocabulary in lexical analyses. Given that names are often culturally specific as well as ephemeral, many proper names will appear as ‘off list’, that is, not occurring in any of the frequency bands. Thus, when using this approach, certain proper names will have to be treated as UNKNOWN.

Klassen (2018) focussed on understanding the lexical load of learners in reading, so her comments need to be viewed in that light. However, the point is well taken – we cannot simply assume that all instances of proper names can be treated as uniquely referring or as given. What we have been proposing in this chapter is that the formal expression (e.g. common noun, pronoun, proper name) and the referential function (e.g. specific, definite, indefinite) are best viewed as interacting with each other, rather than as a strict association. If we want to know if the use of a proper name is an instance of definite reference, the answer is yes, sometimes, but not always. 2.5

Summary

This chapter has shown that reference is a dynamic process which involves the knowledge and understanding of both speaker and addressee. The study of

2.5 Summary

47

reference is better carried out using an approach which reflects this two-way interaction as well as the discourse event as a whole. The functional approach to reference seems to be able to take on board these requisites and offers a way of explaining reference provided it takes into consideration the discourse, cognitive, and lexicogrammatical requirements of the referring process. It was also noted that (in)definiteness is not simply a matter of the presence or absence of a definite determiner or pronoun. Definiteness appears to be more to do with the way in which speakers view their intended referent, and whether they consider the addressee able to view it in a similar way. Thus, the entire discourse event contributes to the establishment of an entity as definite or not, not just its formal structure. We have also argued for the separation of reference (function) from the expression (form). An indefinite expression (form) can be used for definite reference (function), and a definite expression (form) can be used for indefinite reference (function). There is no one-to-one relationship between the lexicogrammatical realisation of the expression and its function in an act of reference. In the next chapter, we take a close look at the form and internal functions of referring expressions.

3

A Functional Grammar for Referring Expressions

Whales do not really belong with the fish they superficially resemble, since the similarity of form and behaviour conceals radical differences of structure and function. (Evans, 1982: 1)

3.1

A Non-headed Noun Phrase

The previous chapter focussed on what we mean by reference and referring and provided an overview of the concept of a referring expression. Referring expressions are, at least within the language system, dominantly a linguistic form, and this form is by far most frequently expressed by the noun phrase (NP) in English. We might even say that referring is what the noun phrase is designed for. It is reasonable to assume, as do others working in this area, ‘that the semantic goal of a noun phrase is presentative in the sense that, by uttering a noun phrase, the speaker is activating a referent in current discourse’ (Martínez-Insua and Pérez-Guerra, 2011: 203). Most functional approaches to language description take the position that language is how it is because of the functions it has come to serve (Halliday, 1973), and this applies to the noun phrase as well. In this chapter we will consider the description of referring expressions in terms of lexicogrammatical patterning. We will use the term noun phrase for this grammatical unit, but it is worth explaining our choice before getting into the detail of the description. The NP has various other names depending on the theoretical approach taken. For example, it is at times labelled the nominal group (Halliday, 1985) or the determiner phrase (see Hudson, 2004), where the terminology reflects the emphasis placed on the class of element which forms the basis for the group. The terms used for this unit typically assume that one element is the pivotal element, or the ‘head’ of the unit, and all other elements somehow depend on this one key element. The use of noun phrase assumes the unit is headed by a noun (or a nominal in the case of the nominal group) and, similarly, the use of ‘determiner phrase’ assumes the determiner element is the head. A full exploration of the theoretical positions involved would be a 48

3.2 A Cognitive Approach to the Grammar of Referring

49

distraction since we only need one description to inform our discussions of referring expressions. We avoid the theoretical conundrum surrounding the concept of headedness by following Dryer (2004), who argues convincingly for a non-headed description of the noun phrase. It is not that nouns and determiners do not play important roles in referring expressions, they certainly do, but we do not need the concept of headedness to have a useful description of the grammar of referring expressions. This description will allow us to have some consistency throughout the discussions presented in this volume and, where it seems appropriate, we will direct the curious reader to relevant literature where the debates can be pursued. This area of study is underrepresented relative to the verb phrase and clause, for example. As MartínezInsua and Pérez-Guerra (2011: 218) point out so aptly, ‘there is room for further investigation on nouns, noun phrases, and their structure and function in clause, discourse(s) and genres’. Through this chapter, we offer our own contribution to nominal expressions and their function in discourse. In the remainder of this chapter, then, we present our curated view on what we feel are useful complementary approaches to the lexicogrammar of referring expressions. We will begin with perspectives from cognitive grammar since, after all, referring expressions are first and foremost cognitive in nature. However, as they are also interactive and designed to serve a social purpose, we draw on relevant approaches from functional grammar, for example, Halliday (1985), who views language as a socially oriented meaning-making system. This chapter forms a basis for the terminology needed to discuss lexical referring expressions. Consequently, we will not be discussing items such as personal pronouns and proper names, for reasons which will become clear. 3.2

A Cognitive Approach to the Grammar of Referring

In Chapter 2, we had a look at referring from two complementary perspectives: definiteness and indefiniteness. The literature on definiteness is not as definite (pardon the pun!) as we might have hoped. The terms definite and indefinite apply equally to the formal expression and to the type of reference, but the two may diverge. In Chapter 1, we emphasised the interactive and collaborative nature of successful reference, which suggests that for reference to work well, ‘speaker and addressee [have to] put in extra effort, generally together, to make sure the reference has been understood’ (Clark and Wilkes-Gibbs, 1986: 6). What we have learnt from psycholinguistic experiments is that speakers do not simply utter any random referring expression; they are demonstrably aware that the addressee has to be able to identify the referent. When identification fails, the addressee will ask for clarification, as we will see in Chapter 7. We know from Chapter 2 that, by convention, a definite expression will also

50

A Functional Grammar for Referring Expressions

function as an identifying (definite) reference and an indefinite expression will function as an indefinite reference, or be non-referential. While there is some truth to this association, we will see in Chapter 8 that this is not always the case. The noun phrase generally holds the clues to how well the speaker and addressee manage the dance of reference. 3.2.1

Signals of Definiteness and Indefiniteness

Part of successful reference involves referential choice (Kibrik, 2011: 15), or the selection of the most appropriate referring expression. It is generally in the best interests of both the speaker and the addressee to get it right because the point of the communicative exchange is not to discuss the best expression for the intended reference but to move through each instance of reference seamlessly so that it does not detract from the discourse goals. With this in mind, let us consider the four main types of reference proposed by Radden and Dirven (2007: 87–112). There are two sub-classes of reference, called individuative and generic. Individuative corresponds to ‘things that are occurring, have occurred, or may occur’ (ibid.: 90), whereas generic reference, for Radden and Dirven, is used ‘when we talk more generally about the world as we perceive it to be structured’ (ibid.). Both individuative and generic reference can be either definite or indefinite, and each has an association with a particular feature of the noun phrase. For example, individuative definite reference is associated with the use of deictic determiners such as this (e.g. this lake is beautiful), anaphoric uses of pronouns such as it, and some uses of the definite article, the. This association is not exclusive as concerns the definite article because we find it used for both individuative and generic definite reference, as illustrated, respectively, in examples (51) and (52), taken from Radden and Dirven (ibid.: 91), where the expressions in question are highlighted in bold. In example (51), the underscored expression a gendarme introduces the referent to the discourse using an indefinite expression, and the expression itself is considered indefinite due to the use of the indefinite article. Instances such as this, ‘are not yet accessible to the hearer, the speaker has to open a mental space for them in the hearer’s mind’ (ibid.: 90). Having activated the referent in the addressee’s mind, a definite expression is used, the gendarme, to continue talking about the referent. However, in example (52) the definite article is used, but this noun phrase, although technically definite due to the definite article, is treated as generic definite reference and hence not identifiable. If we consider example (53), also from Radden and Dirven (ibid.: 91), we find another indefinite expression, a tiger, which on the level of form appears similar in structure to a gendarme from example (51), and yet this reference is a generic indefinite reference.

3.2 A Cognitive Approach to the Grammar of Referring

51

(51)

A Belgian found a monkey and asked a gendarme what he should do with it. The gendarme told him to take the animal to the zoo.

(52)

The tiger hunts by night.

(53)

A tiger has a life-span of about 11 years.

What we can glean from these examples is that while some resources from the grammar of the English noun phrase are typically associated to certain types of reference, there are some less frequent associations. The use of the article in English appears to be more polysemic in function than we might have otherwise believed. In the next section, we consider the concept of grounding, which is used to account for the instantiation of the referent in the NP. It also gives us the central connection between the referent, which is a mental construct, and its realisation in form, that is, a noun phrase. 3.2.2

The Grounded Expression

If we accept that the referent is a mental construct, we might wonder about the relationship between the referent and the linguistic form it takes in discourse. As we mentioned above, referring is what the noun phrase is designed for. An act of reference requires a full nominal (Langacker, 1987: 53, emphasis in original) since it ‘presupposes instantiation of the type in question and designates one or more instances’. Although this expression is more commonly referred to as a noun phrase, the noun itself is not referential. It is true that a noun phrase may consist of only a noun, but there are additional semantic features that must be expressed, for example, quantity, including number (e.g. dogs) or mass (e.g. water). The function of the noun in the expression is to provide a type specification of the referent, in other words to classify the referent as a type of thing. In addition to type specification, Langacker (1987: 53, emphasis in original) identifies two additional bits of information that must be provided by the expression: First a nominal provides some indication of quantity, either in absolute terms (e.g. three cats) or proportionally (most cats). Second, a nominal effects the grounding of the designated instances, that is, it indicates how they relate to the speech event and its participants (the ground). Grounding pertains primarily to whether, within some frame of reference, an instance of the type (or a set of instances) is uniquely apparent to both the speaker and hearer.

In this view, the referent is construed ‘as an instance of some type and further incorporates some specification of quantity and grounding’ (Langacker, 1991: 54). We will take a closer look at these functions in the next section, but for now let us consider a scenario, adapted from Radden and Dirven (2007: 88–89), which will allow us to consider how grounding works.

52

A Functional Grammar for Referring Expressions

We can imagine that two friends, Harriet and Luke, are having a phone conversation and Harriet is about to tell Luke where they will be meeting for dinner. Harriet has planned for them to meet at a particular pub, and now she has to come up with a referring expression that will let Luke know where to go. There are a variety of ways she could do this. For example, she could use the name of the pub, but maybe she does not know its name. She will use the noun pub and not restaurant, café, or bar because it provides the most culturally appropriate way to categorise the referent for Luke. However, there are several pubs which are quite close together, so Harriet will have to provide more specific information. In other words, since she wants him to be able to identify a particular instance, or ‘token’ of the type pub1, she has to specify information that will help Luke identify the referent. Harriet is unlikely to utter the sentence in example (54) because, even though it signals that Luke should know which pub she is referring to, it will not help him know which of the set of pubs in the area it is. He might have to go in and out of several pubs to find Harriet. What we would expect to happen if she did utter (54) is that Luke would reject the expression and ask ‘which one?’ Although these examples are invented, given the scenario we have constructed, Harriet is more likely to use the definite article along with visual description and location as in example (55). (54)

Let’s meet at the pub.

(55)

Let’s meet at the tiny Tudor-style pub directly opposite the castle entrance.

For Radden and Dirven (2007), the act of referring can be visualised as shown in Figure 3.1 (ibid.: 89) and Figure 3.2 (ibid.: 107). In each case, the speaker and addressee (hearer) need to share a sufficiently similar conceptualisation of the referent (indicated by the dotted arrows), and this is shown by the bold circle in each figure, respectively, representing the referent that the speaker wants the addressee to attend to. In Figure 3.1, which corresponds to example Thing Speaker Reference mass Hearer Grounding: referring expression

the pub

Figure 3.1 Individuative reference (Radden and Dirven, 2007: 89) 1

We will use small caps to indicate the label of a category.

3.3 The Basic Organisation of the Noun Phrase

53

Thing Speaker Generic referent: class of thing Hearer

Grounding: referring expression

the tiger

Figure 3.2 Generic reference (Radden and Dirven, 2007: 106)

(54), the type specification (i.e. type of thing) is the noun pub, and given that this is a class of thing, there are theoretically an infinite number of instances of the class of pub in the world: the reference mass (ibid.: 88). The instance being referred to is indicated by the circle in bold. The referent is grounded by the full referring expression (the pub). If we now consider Figure 3.2, we see a different representation of the conceptualisation of the referent. The expression illustrated here is the generic (non-individuative) reference which was presented above as example (53). As Radden and Dirven (2007: 106) explain, in this case, ‘[t]he speaker thinks of a category, such as “tiger”, singles out an instance of the category, which in generic reference is the class of tigers, and grounds this instance in the current discourse by means of a generic referring expression so that speaker and hearer can establish mental contact to the same referent’. The instance in this case is that of the class (indicated by the circle in bold), rather than an individual or a set of individuals as is the case with individuative reference, as we saw in Figure 3.2. Given that our understanding of an act of referring remains quite abstract in many ways, these visualisations hopefully provide a useful representation that will help as we move to the more concrete realisation of reference in linguistic form. In the next section, we will map these concepts onto the functional elements of the noun phrase. 3.3

The Basic Organisation of the Noun Phrase

There is quite a bit of agreement on the basic grammar of the noun phrase, and we find ‘a considerable amount of consistency in nominal group descriptions’ (Fontaine, 2017: 268). The terms may differ, but what appears to be consistent is that the unit of the noun phrase is built by a core element, typically a noun, which corresponds to what Langacker (1991) refers to as type specification,

54

A Functional Grammar for Referring Expressions

Table 3.1 Various perspectives on the noun phrase

Fries (1970) Halliday (1985 onwards) Fawcett (1980, 2000) Martin (1992) Fontaine (2012)

a

little

house

with a yard

determiner 2 Deictic Premodifier quantifiying determiner deictic deictic determiner

looseknit modifier Epithet modifier

head Thing Head head

restrictive modifier Qualifier Postmodifier Qualifier

epithet modifier

thing thing

Qualifier Qualifier

Figure 3.3 A functional-cognitive model of the English NP (adapted from Ghesquière 2014: 24)

and which represents for Halliday (1985: 167) the ‘semantic core’ of the noun phrase. As we can see in Table 3.1, the analysis of example (56), taken from Fontaine (2017: 268), shows slightly different terms used by various scholars to describe the functional elements expressed by the example. Irrespective of the terms, there are four clear functional positions in all descriptions: deictics/ determiners, modifiers, head/thing, postmodifier/qualifier. (56)

a little house with a yard

In what follows, we provide a unifying account, drawing on cognitive grammar and systemic functional grammar broadly, and adapting the approach developed by Ghesquière (2014). 3.3.1

A Cognitive Functional Description of the Noun Phrase

Ghesquière (2014) proposes four functional zones for the English noun phrase in order to account for the instantiation of a type of entity. A simplified view of her model is presented in Figure 3.3, which shows roughly the equivalent areas that we see in Table 3.1, that is, determiners, modifiers, head/thing and qualifiers (or postmodifiers). The following description of each zone is summarised from Ghesquière (2014). The determination zone includes three sub-zones, including a primary determiner and a zone for a secondary determiner to the left and right of the primary determiner. The items in this zone ‘give either identifying or

3.3 The Basic Organisation of the Noun Phrase

55

quantifying information in a broad sense’. The modification zone is divided into various types, similar to Fries (1970), and indeed many functional descriptions make a similar distinction between types of modifiers. Ghesquière divides modifiers into those which ‘attribute properties and qualities to the entity referred to by the head’ and those which ‘measure the degree of gradable properties referred to by the descriptive modifiers (adjective-intensifiers) and/ or the head of the NP (noun-intensifiers)’ (Ghesquière 2014: 25). The categorisation zone is perhaps the most novel aspect of the model since it recognises the important role of certain lexical items to contribute to type specification beyond a single noun. For Ghesquière, the categorisation zone is a grouping of the traditional head of the noun phrase and any classifiers which sub-categorise the type (or thing). The fourth zone in this model is not labelled as a zone per se but instead is described as ‘multi-functional’. This zone captures all postnominal modifiers, such as those given in Table 3.1. We will exclude discussion of this fourth zone for now and return to it in the relevant section below, where we will propose calling it the qualification zone, for reasons which will become clear in Section 3.3.4. Before moving on to discuss each zone in turn, Table 3.2 presents some illustrative examples of noun phrases with different combinations of the zones. The examples in the table were taken from the enTenTen corpus through the SketchEngine platform (Kilgarriff et al., 2014). Note that we have simplified the description somewhat in this table, for example, we have not detailed Ghesquière’s further distinction between the subtypes of descriptive modification (bleached vs non-bleached adjectiveintensifiers). The remainder of our discussion of the noun phrase is organised by each zone. We begin with the categorisation zone, which allows us to focus on the expression of the type specification of the referent in an act of referring. We then look at modification and determination, and the overlap between these two zones. The final zone, which occurs to the right of categorisation, is discussed in terms of particular complexities in the noun phrase. We appreciate that these examples and the basic description present the English NP in an overly simplistic way. As Ghesquière (2014: 45) has shown, this grammatical unit ‘has a more complex syntagmatic structure than is traditionally assumed’, and we feel we are only now beginning to understand some of that complexity. 3.3.2

Categorisation: Referential Type

According to Ghesquière (2014: 26) ‘[a]ll the elements contributing to the delineation of the class of entities referred to are then viewed as constituting one functional-structural zone, irrespective of word class membership’. In most standard grammars, this would simply be considered the head of the noun phrase, particularly in approaches which take a head-driven account but

56

Table 3.2 Examples of the three main functional zones of the English NP Instantiation of a type of entity Modification Determination Secondary

all

such

Primary a those a your a her two those

Degree modification Secondary

Noun intensifier

Adjective intensifier

Descriptive modification Subjective

Objective

boring beautiful

Categorisation Classifier

Thing

train

journey hills mess choice evening hair stories experts

total other wonderful very

long short

same

3.3 The Basic Organisation of the Noun Phrase

57

also in some functional accounts (e.g. Fries, 1970; Fawcett, 1980; Halliday, 1985; Quirk et al., 1985; Langacker, 2008). Attempts to use more functional terminology involve terms such as ‘thing’ (Halliday, 19852; Fontaine, 2012) or ‘primary information-bearing unit’ (Croft, 2001). To simplify the discussion, rather than using ‘thing’ or ‘head’ we will opt for the term ‘type’, which allows us to make a direct connection with type specification in an act of referring. We might have chosen ‘class’ since the function of this ‘semantic core’ (Halliday and Matthiessen, 2004: 325) is not to identify a referent but to denote its class (cf. Fawcett’s (1980) concept of ‘cultural classification’). However, to use ‘class’ risks confusion with the function of classifier which is already in use in many frameworks. The use of ‘thing’ is also somewhat problematic since within cognitive linguistics it has its own defined use, as explained by Radden and Dirven (2007: 42): The term ‘thing‘ [will be used] for a conceptual unit that is expressed in language as a noun. Things appear in our conceptual world as autonomous, or independent, conceptual units that have a certain stability in space and time. The most prototypical things in our experiential world relate to physical objects such as cows and computers. Less prototypical things are abstract units such as illnesses and joy.

What we want to capture in labelling the main element of the noun phrase is its function rather than its structural position or relation. As Langacker (2008: 310) explains, ‘[t]here is a natural tendency for the internal structure of nominals to straightforwardly reflect the semantic functions that characterize them. It is normal for the type to be specified by a lexical noun selected from a very large inventory’. We might not be doing much by shifting away from the term ‘head’, which is the common term for the core of the noun phrase, but ‘head’ is generally a structural term and one that brings with it its own set of problems and complexities (see Schönthal (2016) for a detailed discussion of headedness in NPs). The function of ‘type’ (what Ghesquière (2014) calls ‘head’) is not restricted to a single orthographic noun, nor should it be since it would not make sense to build a description of language solely on the basis of written text. If we really stop to think about it, we do not yet really know what a word is (cf. Wray, 2015). Traditionally, the so-called head of the NP has been considered only as a single orthographic word. Compound nouns, whether written as two words or one, are considered as a single lexeme. For example, in the NP a new tea towel, tea towel, as a compound, is a single lexeme despite being written as two orthographic words. In working out the type element, identifying its expression can be made challenging due to compounds and also due to subclassifying modifiers, which we will consider now. 2

Halliday (1985) maintains both terms, ‘head’ and ‘thing’ (see Table 3.1); however, this is not unproblematic, as discussed by Fontaine (2017).

58

A Functional Grammar for Referring Expressions

Following Ghesquière (2014), we have included (sub-)classification in the categorisation zone. Although treated as a type of modification by Halliday and Matthiessen (2004: 319), because classifiers ‘indicat[e] a particular subclass of the thing’, we can view them as contributing to type specification. By including classifiers in the categorisation zone, we eliminate, to a large extent, the difficulty in distinguishing between a compound noun and a classifier-noun combination. As pointed out by Ghesquière (2014: 27), ‘classifier-noun combinations can become so entrenched in speakers’ minds that the contrast with compounds is reduced to a minimum’. There are some guidelines for differentiating between the two, and since in some instances the classifier can be expressed by an adjective, it can be difficult to identify, for example, an objective descriptive modifier from a classifier. However, often the distinctions do not hold, especially those based on the notion of an orthographic word, which is particularly unreliable in English (e.g. living room but bathroom). A useful distinction for classifiers and compounds has been put forward by Ghesquière (ibid.): ‘semantically, classifiers tend to be organized in mutually exclusive and exhaustive sets (taxonomies), e.g. electric trains/steam trains, whereas compounds are not’. The use of a classifier, then, generally implies a contrast with other ‘kinds of’ relations. Ghesquière (ibid.) compares criminal law with blackbird to show how criminal law offers a contrast with other types of law (e.g. family law, civil law), forming mutually exclusive and exhaustive sets, but blackbird, as a compound, does not. A blackbird is a type of bird, but it does not contrast with other kinds of birds (e.g. yellow bird or small bird). However, even this distinction becomes difficult for some lexical items such as tea towel, which might not set up a contrast with other kinds of towel, but rather other types of materials used in a kitchen (e.g. dish cloth). There is a sense in which the distinction between classifier + noun and compound noun, where both are written as two orthographic words, is not sufficiently important to justify a strict division. It may be, rather, that we need a model that is ‘more tolerant of underspecification, variation, and inherent vagueness [where] the word may find it occupies a lesser place’ (Wray, 2015: 750). This last point is worth pondering since most grammatical descriptions of English are based on word class distinctions. 3.3.3

Determination and Modification

In this section, we will begin with a discussion of the types of modifiers we find in the modification zone, as these are the elements which are semantically most closely associated to the class of ‘thing’, that is, categorisation of the referent. Following this, we will then move on to a discussion about various types of determiners, including quantification.

3.3 The Basic Organisation of the Noun Phrase

59

3.3.3.1 Modifiers in the Modification Zone While it is possible for some modifiers to contribute to type specification, we agree with Ghesquière (2014: 28) and many other functional linguists (Halliday, 1985; Fawcett, 2007) that different modifiers contribute differently. However, ‘the distinction between descriptive and classifying modifiers is not a discrete one’ (Ghesquière, 2014: 28). Fawcett (2007: 172; emphasis in original) states that the function of modifiers ‘is to describe the referent’, and he makes the following distinction: while ‘all modifiers can be used to classify the referent, [. . .] some simply depict [. . .] and some are affective’. We can recognise the potential for some modifying items to be classifiers, as discussed above. The types of modifiers that appear in the modification zone are similar to what Halliday (1985) refers to as epithets. Halliday and Matthiessen (2004: 318–319) further distinguish between experiential and interpersonal epithets, although as they state, ‘[t]here is no hard and fast line between these two’. The basic division can be thought of as objective versus subjective (see Ghesquière, 2014: 30). For example, in the long train, the adjective long has an experiential function of adding an intrinsic property to train, whereas in the stupid train, the adjective stupid has an interpersonal function in that it expresses the speaker’s attitude towards the train in question (see Halliday and Matthiessen, 2004: 319). However, we should note that again the categorisation of the two types of modifier is not discrete; we find, instead, that they ‘represent values on a cline or continuum’ (Ghesquière, 2014: 29). Concerning objective and subjective modifiers, they are seen as two types of descriptive modifiers. Ghesquière (2014: 29) separates these descriptive modifiers from modifiers which do not describe the referent but rather ‘modify or measure the degree of qualities conveyed by a descriptive modifier (e.g. very nice) and/or a gradable head noun (e.g. a complete idiot, such a big mess)’ (ibid.: 34). These degree modifiers are treated separately from the descriptive modifiers as they serve different functions. Just as the descriptive modifiers can be divided into objective and subjective types, the degree modifiers are also divided by the type of intensification they express, whether that is in relation to the lexical noun (expressing type specification) or to a descriptive modifier. Ghesquière (2014) refers to these respectively as noun-intensifiers and adjective-intensifiers (e.g. a total mess and her very long hair from Table 3.2). 3.3.3.2 Determiners We have, to some extent, already discussed the role of determiners in an act of referring since, in our discussion of (in)definiteness in Chapter 2, we saw that one of the main distinctions concerns the nature of the article used in the expression (e.g. the or a). However, the category of determiners is much more complex than the use of an article. The determination zone includes ‘elements [which] deictically and/or phorically anchor the instances of the type referred to by the NP in terms of such notions as givenness

60

A Functional Grammar for Referring Expressions

(the), relative quantity (most), etc. to the information the hearer has already built up from the previous discourse’ (Ghesquière, 2014: 45). Not all determiners are equal, and some are considered more central or primary than others within this zone. For Halliday, the main determiner element is referred to as deictic, but he also recognises a pre-deictic element (Halliday, 1985), as does Martin (1992). The deictic element, according to Halliday and Matthiessen (2004: 312), ‘indicates whether or not some specific subset of the Thing is intended; and if so, which’. Quantity is expressed separately by the numerative element. In Ghesquière’s (2014) model, these two functions are combined in the primary determiner position. As she explains, ‘determining elements occur either in the core or primary determiner position, or as secondary determiners in predeterminer (such a) or postdeterminer (the other) position’ (ibid.: 44). Primary determiners function to express ‘identifying and quantifying information’, typically expressed by ‘definite determiners, indefinite determiners, relative quantifiers and absolute quantifiers’ (Ghesquière, 2014: 46). Among determiners are included articles, demonstrative determiners, and possessive determiners. Halliday and Matthiessen (2004: 314) note that, unlike demonstratives and possessives, the, as in the train, ‘means simply “you know which train”’, but does not specify any further which train exactly the speaker is referring to. Langacker (2004: 99) describes the definite article in terms of ‘directive force’ in that it ‘[implies] that there is only one instance of the specified type within the relevant scope of consideration’. Demonstratives make ‘reference to some kind of proximity to the speaker’ (Halliday and Matthiessen, 2004: 314; emphasis in original). While the demonstratives this and these are used for things near the speaker, that and those are used for things that are not in the speaker’s proximity. So by uttering this train, the speaker tells the listener ‘you know which train: – the one near me’ (Halliday and Matthiessen, 2004: 314). By uttering that train, on the other hand, the speaker indicates that though the listener should still know which train is being referred to, it is not to be found near the speaker. Possessives work in a similar way as demonstratives, but they mark ‘possession, that is, [they make] reference to [a] person as defined from the standpoint of the speaker’ (Halliday and Matthiessen, 2004: 314; emphasis in original). Thus, similar to demonstratives, possessives, in a noun phrase such as my train, tell the listener ‘you know which train: – the one I own’ (Halliday and Matthiessen, 2004: 314). Unlike demonstratives, however, possession ‘manifests our very general cognitive ability to evoke one conceived entity as a way of establishing mental contact with another’ (Langacker, 2004: 100). In other words, possessive constructions introduce a second referent into the equation. A noun phrase with a possessive determiner is thus grounded in the current discourse situation by the means of the two conceptual entities, possessor and possessed.

3.3 The Basic Organisation of the Noun Phrase

61

Unlike possessive pronouns (e.g. my, your, our), a possessive determiner may also be expressed by a genitive phrase3, such as John’s or the town’s, which always consists of a possessor, expressed by a nominal expression and a genitive element (‘s) (see Fawcett, 2000). While a possessive determiner is generally considered definite, this is not always the case. Its (in)definite status depends on the status of the possessor. In an example such as ‘the student’s complaint’ (Langacker, 2004: 100–101), both nominal expressions are definite, is definite. However, Langacker (ibid.) explains that we find examples where both are indefinite, as in ‘The teacher was responding to some student’s complaint’. A further consideration with possessive constructions, such as John’s picture, is their ambiguity, as pointed out by Hawkins (1981: 247–248). The expression John’s picture can be interpreted in the following three ways: (i) picture belonging to John; (ii) picture depicting John; (iii) picture produced by John. This ambiguity vanishes when the possessive element is moved into a postnominal position as in the picture of John, which can only be interpreted to mean the second of the three meanings above, whereas a picture of John’s would be used for the first and third meaning. We will discuss the postnominal position in Section 3.3.4. As was mentioned above, Halliday separates the functions in the noun phrase occurring before modification into deictic and numerative elements, which are combined in the model we are advocating here. Quantity does seem to be part identifying and part describing, and for that reason it might be tempting to treat such meanings as modifiers or as their own kind of element, or indeed as more closely related to type specification (Langacker, 1991). We are following Ghesquière (2014: 45), who argues that ‘core determiners are themselves inherently quantifying’. Within systemic functional grammar, Fawcett (2000) includes quantity as a type of determiner, although separate from the deictic determiner. Quantification for Halliday and Matthiessen (2004: 317–318) is either exact or inexact and either quantitative or ordinative, as illustrated in examples (57) to (60). Ghesquière (2014) prefers an absolute/relative contrast and points out that both ‘absolute and relative quantifiers can function either as primary or secondary determiners’ (ibid.: 47), e.g. most children versus most of those children. (57)

Three trains (exact quantitative)

(58)

Several trains (inexact quantitative)

(59)

The first train (exact ordinative)

3

Radden and Dirven (2007: 158) refer to this type of determiner as qualification by possessive relation as opposed to by property.

62

A Functional Grammar for Referring Expressions

(60)

The preceding train (inexact ordinative)

As indicated earlier, Langacker (2004) states that the deictic the can, in some contexts, have the directive force to mean that there is only one single instance of the thing that can be the referent. Consider example (61), taken from Langacker (2004: 99), which supports the argument made by Ghesquière (2014) that core determiners also express quantity. (61)

It’s a nice house, but the roof leaks

Since houses usually only have one roof, the deictic the in this example can be interpreted to mean the one roof that belongs to the house mentioned earlier. Note that a possessive pronoun would work here as well, for example, its roof leaks. The use of secondary determiners allows the speaker an opportunity to provide the addressee with information that is ‘more semantically diverse and more complex than the primary determiners they are bound by’ (Davidse et al., 2008: 500). In some circumstances, the referent of a noun phrase represents a kind of abstraction rather than a particular instance of a thing (Breban, 2011: 512), as illustrated by example (62). (62)

Three times during the class, a student complained

For Breban (2011: 512), ‘[t]his example talks about three events involving three different individuals, e.g., one student complained, then a second student complained, then a third student complained. However, the wording a student complained expresses a generalization over those events’. Yet, this generalisation is triggered by an element outside the noun phrase, namely three times. In this sense, the example provides an instance of generalisation at the level of the clause. Breban then goes on to say that the same phenomenon also exists at the level of the noun phrase, and she identifies (ibid.: 515) four different secondary determiners which express generalised instantiation: ‘the adjective of multiple exposure usual [. . .], the adjectives of identity same and identical [. . .], the adjectives of similarity similar and comparable [. . .], and the binominal type noun constructions sort of X/kind of X/type of X’. Breban (2011: 514) briefly compares the adjectives of generalised instantiation with descriptive adjectives and claims that secondary determiners do not describe a property. Instead, they supplement the determining information given by the primary determiner by adding more specific referential information. For example, [. . .] usual in the usual Christmas lunch, adds information necessary to successfully conceptualize and identify the referent. It specifies the level of abstractness of the instance, which is underspecified by the ordinary indefinite and definite determiners, viz. that the hearer has to conceive of a

3.3 The Basic Organisation of the Noun Phrase

63

Table 3.3 Ghesquière’s determination zone with examples (2014: 44) Determination zone Secondary determiner

most of such

Primary determiner

Categorisation zone Secondary determiner

Head

main

problems problems problem problem problem

those those the a a

similar

generalized instance with the characteristics of the previous Christmas lunches [they] attended.

Thus, not only deictics but also certain adjectives fulfil the determiner function of telling the hearer which particular instance of a thing the speaker is referring to. Fundamentally, ‘secondary determiners are typically textually intersubjective in that their specific function is to guide the hearer in [their] interpretational task and to help establish a joint focus of attention between speaker and hearer’ (Ghesquière, 2014: 48). To summarise this brief discussion of primary and secondary determiners, we have reproduced Ghesquière’s (2014: 44) set of examples illustrating the determination zone in Table 3.3. Some of these examples are illustrative of a degree of complexity in the noun phrase given the potential for more than one noun in the full nominal expression, for example, the vast majority of those problems. To close this discussion, then, we would like to direct readers interested in more detailed accounts of secondary determiners to the following works in addition to Ghesquière (2014): Breban and Davidse (2003); Davidse et al. (2008); Breban (2010); and Kavalir (2017). 3.3.4

The Fourth Zone: Qualification

Although we have largely adopted Ghesquière’s (2014) description of the noun phrase with some minor modifications, she did not provide the details we would need to similarly discuss the functional elements of this fourth zone since this area of the noun phrase was not involved in her area of study, specifically (inter)subjectification in prenominal elements. As we have done above, we will draw on cognitive and functional perspectives about postnominal modification, or what are generally called postmodifiers, to develop this fourth zone. It is, in a sense, here where it matters what theoretical position is taken concerning the notion of head in the noun phrase because one person’s secondary determiner may be another person’s head. In nominal expressions

64

A Functional Grammar for Referring Expressions

with more than one noun, it can be challenging to work out what is a determiner and what is a postmodifier. In the set of examples given in (63) and (64), we could have a secondary determiner in (63) but a postmodifier in (64), as indicated in bold. See Fontaine and Schönthal (2020) for a detailed discussion of this problem. (63)

I drank a bottle of wine

(64)

I broke a bottle of wine

3.3.4.1 Elements in Postnominal Position In the following, the nature of postnominal elements within the English noun phrase will be presented. ‘Unlike the elements that precede the Thing, which are words (or sometimes word complexes, like two hundred, very big [. . .]), what follows the Thing is either a phrase or a clause’ (Halliday and Matthiessen, 2004: 323). In some cases, the Thing (or type) can also be followed by a simple adjective or an adjective phrase (e.g. the stars visible tonight). In the postmodifier position, we typically find various types of clauses (both finite and non-finite) and prepositional phrases (as in (56), a little house with a yard), as well as various types of nominal and adjectival modifiers that present modification after categorisation. Qualifiers are relatively infrequent, and they generally add a degree of complexity to the nominal expression. For this reason, they tend to be far less frequent in spoken language as compared to written and far more frequent in more technical or lexically dense texts. For example, Fontaine (2008) found that out of a total of 2,957 noun phrases analysed, there were 336 qualifiers, which represents 11.36 per cent of all noun phrases. This collection of noun phrases came from two different text types, and when the instances are separated by text type, there are clear differences. On the one hand, the collection of personal email conversations had a frequency of 6.45 per cent (161 instances of a total of 2,491), while in the educational brochures analysed, the frequency rose to 37.55 per cent (175 of 466). In both cases, the majority of all qualifiers was expressed by prepositional phrases for both types of text. These results are in line with Biber and Clark’s (2002: 46) diachronic study, which showed that ‘prepositional phrases (rather than relative clauses) are overwhelmingly the most common type of postmodification, accounting for 80 percent of all postmodifiers in news and academic prose’. Relatively little work has been done on this area of the noun phrase in relation to reference. Within systemic functional linguistics, the fourth zone is referred to as the qualifier position (Fawcett, 1980; Halliday, 1985). While qualifiers are treated as postnominal modifiers, as Halliday and Matthiessen (2004: 324) explain, ‘the characterization is in terms of some process within

3.3 The Basic Organisation of the Noun Phrase

65

which the Thing is, directly or indirectly, a participant. It may be a major process, i.e. a relative clause; or a minor process – a prepositional phrase’. The key aspect of interest here, as explored in Fontaine (2008), is that in an act of referring, the speaker ‘uses’ the qualifier for the purposes of referring by setting up an additional situation to the one in which the referent has a functional role. This relation is explicitly marked in the case of relative clauses, such as the underscored clause in example (65), taken from Fontaine (2008), where it is clear the intended referent is being involved in two distinct situations: (i) as working (e.g. they are working) and (ii) as what she takes (e.g. she takes them). (65)

The chemo pills she takes are working

For Fawcett (2004; see also Fontaine, 2008), the description of the referent by qualification can be captured by three main relations: ‘by role in other situation’, which covers relative clauses (as in (65); ‘by relation with other thing’, where the qualifier is expressed by a prepositional phrase (as in (56)); ‘by alternative specification of thing’, which captures nominal apposition (e.g. my friend, John); or ‘by deictic place’, where the postmodifier is a deictic expression of place (e.g. the book over there). Unfortunately, further detail on these relations was never published, so we will turn to Radden and Dirven (2007: 157–166), who propose similar distinctions, although in slightly different terms. They include qualification by property, by relation, and by situation. Qualification by relation is further divided into three subtypes: possessive, intrinsic, and schematic relation. We will not discuss qualification by property at this point because this primarily concerns modification which we discussed above. Qualification by possessive relation has also already been discussed in connection with the determination zone. Therefore, in what follows, we will consider functional distinctions for qualification by intrinsic relation, schematic relation, and situation. 3.3.4.2 Qualification by Intrinsic Relation Radden and Dirven (2007: 159) state that ‘[i]n English, intrinsic relations are expressed by the preposition of, which has no other function than relating the two noun phrases’. Consider the examples (66) to (69), taken from Radden and Dirven (2007: 159). (66)

the end of the tunnel

(part/whole relation)

(67)

the father of the bride

(relational term)

(68)

the review of the book

(reified thing)

(69)

the issue of unemployment

(identifying relation)

66

A Functional Grammar for Referring Expressions

What these four examples have in common is that the first noun is intrinsically related to the second noun within the noun phrase. In other words, an end is always an end of something, a father is always the father of someone, a review is always written on something in particular and thus always a review of something, and an issue is always related to something, that is, an/the issue of something. Expressions such as these are the topic of much debate in the literature. For an excellent overview of these issues, see Schönthal (2016) and for a novel discussion of the textual functions of these expressions, Schönthal (2022). 3.3.4.3 Qualification by Schematic Relation Qualification by schematic relation ‘relates two entities in a highly abstract, typically image-schematic way [. . .] typically expressed by prepositions. For example, the preposition under evokes the image schema of verticality, on the contact schema, in the container schema, behind the front-back schema, etc.’. (Radden and Dirven, 2007: 159). Examples of schematic relations are given in (70) to (73), taken from Radden and Dirven (2007: 160). (70)

the plates on the dinner table

(71)

a book on Prince Charles

(72)

a new play by Alan Ayckbourn

(73)

a donation to the church

Qualification by schematic relation corresponds to Fawcett’s (2004) description ‘by relation with other thing’ and Halliday and Matthiessen’s (2004: 324) view of the qualifier as characterising the referent in a minor process. These approaches highlight the relationship between the type specification of the referent and that of another referent. 3.3.4.4 Qualification by Situation Qualification by situation involves relative clauses, which ‘may qualify definite or indefinite referents and may or may not restrict them’ (Radden and Dirven, 2007: 161). This type of qualification is very clearly similar to Fawcett’s (2004) description by role in other situations as well as Halliday and Matthiessen’s (2004: 324) functional account of relative clauses. As with most explanations of relative clauses, Radden and Dirven describe three subtypes, each of which is illustrated respectively in (74) to (76), from Radden and Dirven (2007: 161). (74)

I have friends all over the world. The friend who lives in Tokyo is coming to see me.

(75)

A friend who lives in Tokyo is coming to see me.

3.3 The Basic Organisation of the Noun Phrase (76)

67

My friend, who lives in Tokyo, is coming to see me.

Examples (74) and (75) both contain a restrictive relative clause, which qualifies a definite and an indefinite referent, respectively. Example (76), on the other hand, contains a definite referent with a non-restrictive relative clause. The restrictive relative clauses add further information necessary to identify which friend exactly is coming for a visit. The relative clause in (76) is non-restrictive because the referring expression my friend already provides enough information (depending on the immediate context) for the addressee to identify the intended referent. However, importantly, and as marked by the relative pronoun who in each example, the referent is also expressed as a participant in the relative clause. Consequently in (24), for example, the referent is participating in both a situation of ‘coming’ and a situation of ‘living’. Since none of the abovementioned approaches to qualification by situation discuss non-finite postmodifying clauses in any detail, we turn to de Haan’s (1989: 64) study of postmodifying clauses. In his relatively large (130,000 words) corpus study of postmodifying clauses, de Haan identified 2,430 examples of postmodifying clauses. While his findings show that finite clauses are more frequent than non-finite clauses (ibid.: 62), the distribution was more equal in Fontaine’s (2008) study, although it is important to note that the number of postmodifying clauses in her study was much lower (only 115 postmodifying clauses in a corpus of 13,907 words). However, when text type was considered, the email conversational texts had a higher frequency of finite clauses (61.7 vs 38.3 per cent), whereas the reverse trend was true of the educational brochure (52.7 vs 47.3 per cent); see example (77), from Fontaine (2008). This example illustrates a non-finite past participle clause, which according to de Haan (1989: 64) is the most frequent type of non-finite postmodifying clause. (77)

an era driven by environmental awareness and social change at the grassroots level

There is no reason to exclude these non-finite clauses from what we are calling qualification by situation. The main difference related to the amount of explicit information encoded in the qualifier, for example, non-finite clauses do not encode tense. 3.3.5

Four Zones: A Summary

One of the difficult aspects of studying nominal expressions is that much of the work tends to be focussed solely on the nominal expression itself or on its role in discourse without consideration of its lexicogrammar or indeed on its cognitive aspects. Very few approaches attempt an integrated account, yet this

68

A Functional Grammar for Referring Expressions

Table 3.4 A functional description of the referring expression (adapted from Ghesquière, 2014) Modification zone

Determination zone

Primary determiner Secondary determiner

Degree Noun

Description

Noun intensifier Adjective intensifier

Subjective Objective

Categorisation zone

Qualification zone

Classifier

By intrinsic relation By schematic relation By situation

Type

is really needed and what we have tried to do here. Fries (2001) is notable for trying to maintain such an integrated approach. As he explains (ibid.: 87), ‘we regularly choose the wordings of our nominal groups (including Heads and Modifiers) to establish features of the referents which are relevant to the discourse’. This spirit of Fries’ comment is valid, although we have steered away from head-modifier relations for this description. This is not to say that such dependency relations are not relevant, since it is clear that there are some dependencies that are best accounted for by logical relations (see Kavalir, 2017). By way of a summary of the above discussion of the functional grammar of referring expressions, we have brought everything together in Table 3.4. This account has drawn heavily from the impressive work by Ghesquière (2014). We have complemented her approach by integrating a cognitively oriented account of qualification. We also replaced her use of ‘thing’ with ‘type’ because we felt it would be a clearer term. However, as is often the case, there are no perfect terms, and the true test is how well they hold up both in theory and in practice. 3.4

Summary

This chapter brings to a close the first section of this volume, which, through these first three chapters, was designed to provide an introductory overview of existing approaches to reference and the main theoretical debates from grammatical, cognitive, and functional perspectives. In this chapter, we focussed on developing a cognitive functional account of the noun phrase that is specifically intended to integrate its use in acts of referring and, as such, its use as a referring expression. We have provided a description of the functional elements that comprise this unit and the structures that serve to realise them.

3.4 Summary

69

We mentioned earlier that this chapter would not include discussion of pronominal reference or simple proper nouns. The main reason for this might be obvious at this point, and this is because their use is as a full nominal expression, that is, there can be no determination, modification, or qualification and furthermore they are not, typically at least, used to categorise a type. However, we know from our discussion in Chapter 2 that proper names are more complex than a simple account of their use would reveal. These more atypical uses shift the proper name into the categorisation zone and open up the full potential of the functional zones of the noun phrase. The term noun phrase is not ideal. We have used it more for convenience since it is so well established, and we felt it was a ‘good enough’ fit. However, Langacker (2008: 310) prefers the term ‘nominal’ because these expressions ‘are not always phrases, nor do they always contain a noun (as traditionally understood). An expression does not qualify as a nominal because it exhibits any particular structural configuration. Instead, the crucial factors are meaning and function’. His use of ‘nominal’ is more encompassing and would be preferable if we wanted a term to include pronouns and proper nouns, irrespective of any categorisation function. A truly meaning-driven framework, as outlined by Polguère (2015), does need to be driven by meaning. The speaker’s production of these expressions for the purposes at hand must be the leading force. After all, as Fawcett (1993: 630) has argued: language is not, as it has traditionally been taken to be, directionally neutral, but that the meanings that it is organized to express are those of the performer rather than those of the addressee. . . . the ‘theme’ of the clause is the ‘point of departure’ of the performer, deixis in language reflects the standpoint of the Performer; modality expresses the confidence level of the Perfomer, etc. That is, the basic model of language is that of production.

There is possibly no better place in the grammar to develop this approach than in the study of referring expressions due to the need for an integrated approach which combines cognitive, discourse, and grammatical perspectives. We hope that we have been convincing in this regard over these first three introductory chapters. In the next section of the volume, we set out our view of typical reference. First, in Chapter 4, by examining what we mean by the concepts of typicality and atypicality and how they relate to reference. The following two chapters each represent an aspect of typical reference. Chapter 5 presents an account of how reference works in spontaneous discourse, and Chapter 6 looks at reference in written discourse where the writer/reader relationship and the mode of production shapes referential choice.

Part II

Typical Reference

4

Typicality and Atypicality

There will always be more than one way to skin a category. (Halliday, 1997: 266)

4.1

On Categories

There is a sense in which all of linguistics is a categorisation problem of one kind or another; ‘[c]ategorization applies at every level of form and meaning’ (Bybee, 2013: 50). Categories are, however, problematic, and linguistics is no different from other disciplines in being confronted with categorisation problems; ‘[w]e find entities that cannot be properly classified, either because they would simultaneously belong to two or more traditional categories or because they are too underspecified to put them in one of the traditional categories’ (van Lier and Rijkhoff, 2013: 6). We have to categorise our objects of study, and yet the process of categorising is inherently problematic. However, categorisation should be an explicit part of what we do as linguists. In the previous chapter, we suggested that perhaps some of our grammatical categories need to tolerate a degree of underspecification and greater fluidity. According to Halliday and Matthiessen (1999: 193), ‘[t]he grammar imposes a categorization that is compromising, fluid, indeterminate and constantly in process of change, along with changes in the human condition and in the interaction of humans with their environment’. As humans, we do like categories, and they are essential in our everyday lives and our development both individually and socially. Categories involve boundaries, and therein lies the challenge. There are many areas of the grammar which have boundaries, but these boundaries are, to some extent, artificial even though there are normally good reasons for where and how a boundary is established. As Croft and Cruise (2004: 89) explain, ‘a boundary is arguably the most basic of all the properties of a category. A category is like a container: one of its major functions is to divide the objects in the world into those things that are in it and those things that are not in it. This function cannot be fulfilled without a boundary’. The title of this chapter suggests two categories: typical and atypical. While it is true that we want to make a distinction between the two, we are not 73

74

Typicality and Atypicality

proposing two closed, mutually exclusive sets. Much of the literature on reference has established our core understanding of acts of referring, and this body of work has made a significant contribution to how we view referring. However, in many cases, the evidence has been experimental, computational, philosophical, and/or based on small excerpts from discourse. We explained in Chapter 1 that our main interest in this volume was to present an integrated perspective on referring with particular attention given to what we have called atypical uses of reference. This chapter, then, sets out the distinction we would like to make on the division between typical and atypical uses. In our original plans, we anticipated having a chapter dedicated to each topic, but it became quickly obvious that it would be almost impossible to talk about one without the other. In a sense, we can think of typical and atypical as contrastive in that one gets its meaning from not being the other, but we would not like to conceptualise the two terms as antonyms. We have a good sense of the general use of ‘typical’, which means ‘having the qualities of a type or specimen; serving as a representative specimen of a class or kind’ (‘typical’, adj2). If someone drinks coffee regularly, we might say that coffee is this person’s typical drink. If one day they refuse a cup of coffee, we might notice that this is not very typical of that person. And we might even wonder if it means they are not feeling well. We have already suggested in various places so far, especially in Chapter 2, that some referring expressions are not typical, or that they are doing something different (e.g. my Canada). In this chapter, we will set the stage for both our set of chapters on typical reference (Chapters 5 and 6) and for the third section of this volume, which includes four chapters dedicated to more atypical uses of reference. In the next section, we review what is meant in the literature by the concepts of typicality and atypicality by considering related terms such as prototypicality, markedness, core versus periphery, and norms versus exploitations. Having established how we will be using the terms in this work, Section 4.3 presents a brief overview of two cognitive theories of reference: the Accessibility Scale (Ariel, 1990) and the Givenness Hierarchy (Gundel et al., 1993). These models are acknowledged in the field of reference as setting the standards for typical uses of reference, and they have established the importance of the need for a ‘common ground’ between interlocutors. Finally, Section 4.4 takes a first step towards discussing atypical instances of reference. It critically examines what it means for a referring expression to be atypical, and what functions such atypical expressions might have in the discourse. 4.2

What Is Typicality and Atypicality?

As will become clear throughout the following discussion, what seems to underpin all notions of typicality is the concept of frequency. Frequency, in

4.2 What Is Typicality and Atypicality?

75

turn, is intrinsically linked to usage. To better understand how typicality stems from frequency of usage, we will first briefly consider the role of usage. According to Bybee (2013: 50), ‘[i]t is repetition that leads to conventionalization of categories and associations, as well as to the automation of sequences’. The more a specific feature or pattern is repeated in use, the more conventional it becomes. Thus, by taking a usage-based approach to language, it is possible to make claims about ‘the extent to which linguistic patterns are common or rare’ (Biber, 2000: 287; emphasis added), that is, what is typical or atypical in language. By looking at linguistic phenomena in use, we can identify association patterns, that is, ‘the systematic ways in which linguistic features are used in association with other linguistic and non-linguistic features’ (Biber, 2000: 289). With ‘linguistic features’, Biber is referring to both lexical and grammatical ones. In other words, an individual word can either frequently co-occur with other words (its collocations), or it can have specific structural preferences, such as ‘whether a particular adjective typically occurs with attributive or predicative functions, or whether a particular verb typically occurs with transitive or intransitive functions’ (ibid.: 290). Furthermore, in addition to the variability of a linguistic feature in terms of its linguistic associations, research into its typicality also needs to consider its non-linguistic associations, that is, its distribution across registers, dialects, and time (Biber, 2000: 289). This flexibility of language is expressed as follows by Kemmer and Barlow (2000: ix), who discuss linguistic dynamicity and language users’ influence on each other’s linguistic systems: [U]nits of language (from phonemes to constructions) are not fixed but dynamic, subject to creative extension and reshaping with use. [. . .] Usage events are crucial to the ongoing structuring and operation of the linguistic system. Language productions are not only products of the speaker’s linguistic system, but they also provide input for other speakers’ systems (as well as, reflexively, for the speaker’s own), not just in initial acquisition but in language use throughout life. Thus, usage events play a double role in the system: they both result from, and also shape, the linguistic system itself in a kind of feedback loop.

Hence, everyday language input and usage shape the frequency patterns of individual language units and thus determine what language users consider a typical or an atypical use of language in a given context. Biber (2000: 287–288) states that ‘a use perspective is required to investigate the stylistic preferences of individuals, the differing linguistic preferences of groups of speakers, and the ways in which “registers” (or “genres”) favour some words and structures over others’. It is thus evident that the notion of typicality is tightly interlinked with usage events and is dependent on the frequency of occurrence of individual linguistic units, which, in turn, is dependent on different registers of language.

76

Typicality and Atypicality

In what follows, we first look at the concept of register as defined by Halliday in order to consider how its parameters interact with typicality. In the remaining sub-sections, we will discuss how (a)typicality relates to the concepts of prototypicality, markedness, and finally norms and exploitations. This section concludes with a summary of our position that both typicality and atypicality are best interpreted as usage-based and register-dependent properties of an instance of language use. 4.2.1

Register

Based on the discussion above, when we talk about typicality and atypicality, we are really talking about a kind of variation with a given context. Halliday (1968: 141) defines register variation as ‘variation according to use’, which he distinguishes from dialectal variation (variation according to user). Language in use is always in a particular context. Halliday refers to this as ‘context of situation’. Context is organised by the parameters of field, tenor, and mode of discourse, commonly referred to as the variables of register (see Lukin et al., 2011). Field represents the social context, seen as ‘a field of significant social action’ (Halliday, 1977/2002: 55), in other words, what is going on. Tenor includes social relationships and roles, where text is viewed as ‘an intersubjective event’ (Halliday 1981/2002: 245). Finally, mode concerns the symbolic organisation of language, involving ‘the role assigned to the text, including both medium and rhetorical function’ (ibid.: 226). We will be discussing these parameters in more detail in Chapter 5 and Chapter 6 when we consider how they contribute to our understanding of referring in spontaneous and prepared discourse, respectively. Different constellations of field, tenor, and mode in different situations interact with the lexicogrammar through a relationship of construal in the sense that context of situation is realised in and construed by language. This relationship is mediated by the semantic layer, which sits between context and lexicogrammar in Halliday’s model. As Taverniers (2011: 1122) explains, it is the theoretical concept of construal that accounts for ‘the relationship between language and extralinguistic reality’. This relationship is not unidirectional. As Lukin et al. (2011: 207) explain, ‘[r]egister is perhaps best understood as a dialectic – between system and instance – since the two are never actually possible without each other’. According to Bowcher (2019: 146), ‘[l]anguage varies according to the occasions on which it is being used, and speakers of a language typically have no difficulty recognizing this fact and managing their “repertoires” of language use accordingly’. What we can glean from this is that speakers are aware, to some extent, what the expectations are of a given register and they normally are aware of any deviation from those norms. This view of register is not

4.2 What Is Typicality and Atypicality?

77

deterministic, nor is it static or stable; in other words, it is not the context that determines language use. O’Donnell (2021) argues that ‘at each point of a text/ interaction, we as interactants have the choice to affirm the contextual expectation, or to vary from it, either using novel means to achieve some situational goal, or by shifting to a distinct Context of Situation’. Due to this dynamicity of language and due to register variation, what can be considered as typical or atypical differs from one linguistic situation to another. ‘[M]ost linguistic phenomena are not distributed in a simple binary opposition of “frequent” versus “rare”’ (Biber, 2012: 11). Frequency and typicality of linguistic patterns should not be described across language use in general (Biber, 2012). Some might even say that language is best viewed as a collection of registers, questioning whether a description of a language as a whole is even possible. Instead, register-specific accounts are seen as more informative, or at least, we could take the view that consideration of register parameters will lead to a richer description. For example, in differentiating between the patterns of use in speech and those in writing, which mostly involve differences in mode, we know that pronoun use increases in frequency considerably in spoken language. We will discuss pronoun use in spontaneous discourse in more detail in Chapter 5, and in Chapter 6 we consider referential choice in written genres. 4.2.2

Typicality and Prototypicality

As discussed above, typicality is usage-based and register-dependent. In addition, it is important to note that any notion of typicality in linguistic research would have to be based on empirical evidence, for example, from investigating a large corpus. Not only are speaker intuitions unreliable (Hanks, 2013), but ‘a corpus offers information that a native speaker cannot replicate: an indication of “central and typical” usage’ (Hunston, 2002: 42). We do want to be confident when we say something is a typical instance or typical use and, increasingly, big data is used to provide large-scale evidence. However, this is not to say that there is no place for individual perceptions of typicality. Intuitive ideas of what is typical in language are instead covered by the notion of ‘prototypicality’. The term originates from Rosch’s (1975) Prototype Theory, which showed that people have an idea of what counts as a typical member of a category and which members are more peripheral. For example, in some places, such as North America, a robin might be considered a more typical member of the category of bird1 than a penguin (also see Aitchison,

1

As a reminder, we use small caps to indicate the label of a category.

78

Typicality and Atypicality

1994: 55). Such prototypes are based on an individual’s judgement of what is typical and the idea that some instances are more typical than others. For some in corpus linguistics (cf. Hunston, 2002: 43), the term ‘prototypical’ is used to mean what language users believe to be the most typical but which does not necessarily (although it might) align with corpus evidence (see also Hanks, 2013). Thus, while ‘typicality’ is closely linked with frequency and is based on corpus evidence, ‘prototypicality’ is more associated with a sense of ‘belief’ and of what ‘ought to be’, based on an individual’s experience of the world. 4.2.3

Markedness

The terms ‘marked’ and ‘unmarked’ are often used to indicate that an instance or some feature stands out in some way. The term ‘marked’ is problematic for a number of reasons, the main one being that its use is not consistent, which has resulted in the term being highly polysemous. After examining twelve distinct senses of markedness across a variety of sub-fields of linguistics, Haspelmath (2006) concluded that the term is misleading and unhelpful. He explains that ‘the “intuitive” shared sense of “marked/unmarked” is not distinguishable from the sense of everyday words like uncommon/common, abnormal/normal, unusual/usual, unexpected/expected’ (ibid.: 63). Marked and unmarked are terms used by Halliday quite specifically to subcategorise the theme of a clause. For Halliday (1994: 33), an unmarked theme is one where ‘there is no prior context leading up to it and no positive reason for choosing anything else’. In other words, under normal circumstances, unmarked theme is the default pattern of the clause. A marked theme, according to Halliday and Matthiessen (2014: 97), adds ‘a feature of contrastiveness’. Marked in this sense means that the themes are ‘unusual enough to draw attention to themselves, and [. . .] they only occur when contextual reasons overrule the unmarked choice’ (Thompson, 2004: 145). The distinction is illustrated in examples (78) and (79), taken from Halliday and Matthiessen (2014: 92). The assumption is that theme will be expressed, typically, by the subject of the clause, unless some other element occurs first (at least in declarative clauses). (78)

in this job, Anne, we’re working with silver [marked theme]

(79)

the people that buy silver love it [unmarked theme]

This use of marked indicates a semantic distinction in the domain of a grammatical category. Although, by definition, marked themes are not necessarily ones that are rare, there is nevertheless an assumption that they are infrequent and/or that they stand out. Other uses of ‘marked’ in grammatical

4.2 What Is Typicality and Atypicality?

79

descriptions tend to carry a rarity or frequency distinction. In a sense, this is roughly what we are trying to capture with the terms typical and atypical. 4.2.4

Norms and Exploitations

The terms we have considered so far seem to be quite similar in terms of the roles played by frequency and, to some extent, expectancy. In Hanks’ (2013) theory of norms and exploitations, he uses corpus pattern analysis to identify different patterns of use. A norm is defined as ‘a pattern of ordinary usage in everyday language with which a particular meaning or implicature is associated’ (Hanks, 2013: 92). Norms are not differentiated by frequency; ‘rare patterns are recognized as patterns, despite their rarity, because they have distinctive meanings’ (ibid.). In this sense, a norm has an established meaning. Exploitations are creative choices on the part of the speaker. Hanks (2013: 212; emphasis in original) offers the following description of an exploitation: An exploitation is a deliberate departure from an established pattern of normal word use, either in order to talk about new or unusual things or in order to say old things in a new, interesting way. Exploitations are part of the natural human habit of playing with language, but they can also serve a serious purpose, especially in cases where a more well-established way of talking about a particular event, situation, or entity is not available. [. . .] exploitations are typically dynamic, creative, or graphic choices within the boundaries of possible language use.

Thus, exploitations are more deliberate and unusual uses of language, produced by individual speakers for creative purposes, including phenomena such as anomalous collocations, metaphors, metonymy, zeugmas, oxymorons, understatements, euphemisms, hyperboles, irony, sarcasm, and puns. We will return to the notion of exploitations in our discussion of atypical reference in subsequent chapters in order to illustrate that atypical use of reference is indeed often linked to a creative use of language. As is the case with all categories, the boundary between norms and exploitations is not clear-cut. Instead, it is possible for exploitations to become norms, or an alternation of a norm (see Hanks, 2013), in their own right when used more consistently by a larger number of language users as this entrenches the meaning of the use. Exploitations are regularly the source of new meanings, especially by metaphoric extension. ‘Metaphorical variants are probably always marked close to the time of their coming into the language, but there is no reason why they should stay marked after that period’ (Steiner, 2004: 142). By this, Steiner means that they carry more meaning, in the sense that both speaker and addressee are likely to recognise that a new meaning is construed. Exploitations are atypical instances of language in the sense discussed above in that they are infrequent, but infrequency does not help us distinguish them from norms, which can also be infrequent. Instead, it is their

80

Typicality and Atypicality

creativity and the way they exploit meaning potential that defines them. Given our discussion of register as ‘variation according to use’ rather than ‘user’, we might expect that, due to the nature of exploitations, they are not influenced by register; however, it may be that certain tenor relationships (e.g. more formal relationships, such as interviewer/interviewee or doctor/patient) constrain creative and novel uses of language. 4.2.5

Typicality and Atypicality as Usage-Based and Register-Dependent

To bring this discussion of (a)typicality to some sort of conclusion and to try to consolidate the various perspectives we have touched upon, let us return to Bybee’s (2013) point about repetition and its role in conventionalisation. It is like a well-trodden path in the forest; it is just easier than making a new trail. The more we are used to a way of speaking, the easier it is for us and for our collocutors. Automation, as Bybee (2013: 50) points out, ‘allows the units of language to be combined in conventional ways that lead to fluency in both production and perception’. However, the point we discussed above by O’Donnell (2021) is important; as speakers, we can opt out of conventionalised ways of speaking at any moment. And, as Hanks (2013) claims, if we want to express something new, a new meaning, and if our conventions are not fit for the purposes at hand, we can be creative and exploit existing conventions to do something different. How, then, can we settle on a view of typicality? While we acknowledge that any boundaries we try to place around a given term will likely end up overflowing, we propose that we view typicality as conventionalised language use which is reasonably expected in a given context of situation. The corollary, then, is that atypicality is infrequent, unconventionalised language use for a given context of situation. As Hanks’ (2013) theory of norms and exploitations states, norms may be infrequent; in other words, frequency is not a property of a norm. A low-frequency instance of a language pattern may be a norm (typical) or an exploitation (atypical). As a final point, and we would not like to put too much emphasis on the terminology here, we simply do not know if Hanks’ (2013) model is suited to reference, given that it is designed for lexical analysis. For this reason, we prefer to use ‘typical’ and ‘atypical’ as a means of identifying the types of reference that we are primarily interested in discussing in this volume. Therefore, there are a variety of factors which contribute to determining typicality. It may well be that the use of these terms is only of temporary value, for the specific purposes of this volume. There may not be any functional value to them in the broader aim of developing our understanding of reference. The two remaining sections of this chapter will focus, respectively, on what we understand by typical reference and atypical reference. This discussion will be necessarily brief since these concepts and the evidence for them will be considered in detail in Chapters 5 and 6, respectively.

4.3 Typical Reference

4.3

81

Typical Reference

In Chapter 1, we emphasised the importance of viewing reference from multiple perspectives, including discourse, functional, and cognitive approaches, and informing our understanding with empirical research, for example, from psycholinguistics. In this section, we will consider what we learn about typical reference from these areas. Traditionally, reference has been viewed as an anaphoric pairing of a referring expression with a lexical antecedent, that is, a previous lexical instance referring to the same referent (see Chapters 1 and 2). This traditional view of referring has been found to be unsatisfactory since a referent seldom remains static throughout discourse. Instead, a discourse-functional perspective has the advantage of being able to take into account the ‘evolving’ nature of a referent throughout the discourse (Cornish, 2010: 226) in ways in which the antecedent models cannot. As an illustration of what is meant by ‘evolving reference’, we can track the mental representation we have of the referent (potato) of the expressions highlighted in bold in the recipe given in example (80), from Jones (2014). (80)

i) Peel 1kg of Maris Piper potatoes and cut each into 4 even-sized pieces if they are medium size, 2–3 if Ø smaller (5cm pieces). ii) Drop the potatoes into a large pan of water. iii) Add salt, then wait for the water to boil. iv) Simmer the potatoes uncovered, reasonably vigorously, for 2 mins. v) Meanwhile, put your choice of fat into the hot roasting tin and heat it in the oven for a few mins, so it’s really hot. vi) Drain the potatoes in a colander. vii) Carefully put the potatoes into the hot fat and roast them in the oven until they are golden and crisp.

A reader of this recipe would very likely not hold the same mental representation for potato in vii) as the initial mention of potato, the original unpeeled potatoes in i). Throughout the recipe, potato has evolved as a referent. According to Jones (2014: 37), this example ‘clearly illustrates the need to take into account the discourse dimension of a text as the antecedent’s referent does not remain static’. The intended referent has developed from a non-edible vegetable to a delicious culinary delight. We saw in Chapters 1 and 2 that accounts from philosophy and from textuality (i.e. textual co-reference) are limited when it comes to a chain of referring expressions like those in (80), and yet anyone used to reading recipes, and especially anyone who is used to roasting potatoes, will find these references entirely typical. From psycholinguistic research on reference, we have discovered how interlocutors in any communicative event collaborate in order to refer successfully (see Clark and Wilkes-Gibbs, 1986; Schober and Clark, 1989; Brown, 1995; Brennan and Clark, 1996; Bezuidenhout, 2019). Specifically, Schober and Clark’s (1989) collaborative view of referring suggests that participants in a conversation actively work together to ensure that understanding takes place

82

Typicality and Atypicality

and do not proceed with the conversation until they are satisfied that they have understood each other. We will take a close look at examples where this happens in Chapter 7. Interrupting your speech partner when you have doubts about identifying the referent is also typical. In fact, speakers often catch themselves before the addressee even has a chance, and they rephrase the expression to make it more specific if they think the addressee might not be able to identify their intended referent. Finally, cognitive research on referring considers the interlocutors’ mental representations of a particular referent and whether they are matching at any given point in the discourse. Chafe (1976: 27) discusses the various statuses that nouns and referents may have in terms of ‘the speaker’s assessment of how the addressee is able to process what he is saying against the background of a particular context’. He suggests that not only are we able to store a great deal of knowledge, but we also have ‘temporary states’ with relation to that knowledge (ibid.). Therefore, speakers must adapt what they are saying to fit in with what they assume the addressee is thinking at that moment, and only then will the message be assimilated (unless they assume incorrectly of course, which might lead to a misunderstanding, requiring more collaborative work to reach a common ground). Two theories that deserve special attention in this field are Ariel’s Accessibility Theory (1990) and Gundel et al.’s Givenness Hierarchy (1993) as they will help us understand the form (i.e. the referring expression) that we conventionally expect most typically in order to successfully pick out an intended referent. These two theories are discussed in more detail in the following two sub-sections. 4.3.1

The Accessibility Scale

For Ariel (1990, 1994), the status of referring items is a matter of their degree of accessibility. In fact, at any given point in the discourse, the speaker or writer must assess how accessible the intended referent is to their addressee and then choose an appropriate referring expression that corresponds to the addressee’s mental representation of the discourse under construction. Indeed, given the dynamic nature of discourse referents, as we discussed above, the speaker must carry out a new assessment of the referent’s accessibility each time the same referent is referred to. Thus, the assumed accessibility of a referent will affect the type of referring expression a speaker chooses to use, and in turn, a chosen referring expression indicates how accessible the speaker deems the referent to be. Ariel (1990) proposes an Accessibility Marking Scale, which we illustrate here in Figure 4.1, where a referring expression of type (a), that is, a full name + modifier, indicates a low degree of accessibility, whereas pronouns and ellipses of type (k) to (o) suggest high accessibility.

4.3 Typical Reference

83

Low Accessibility (of the antecedent) (a) Full name + modifier (b) Full (‘namy’) name (c) Long definite description (d) Short definite description (e) Last name (f) First name (g) Distal demonstrative + modifier (h) Proximal demonstrative + modifier (i) Distal demonstrative (+ NP) (j) Proximal demonstrative (+ NP) (k) Stressed pronoun + gesture (l) Stressed pronoun (m) Unstressed pronoun (n) Cliticised pronoun (o) Extremely high accessibility markers (gaps, including pro, PRO and wh-traces, reflexives, and agreement)

High Accessibility (of the antecedent) Figure 4.1 Accessibility Marking Scale (Ariel, 1990: 73),

Thus, when choosing a referring expression, its degree of accessibility as given on Ariel’s scale should match the accessibility of the referent it is used to refer to. Ariel (1990: 32) argues that the degree of accessibility is dependent on three overlapping criteria: ‘informativity’ (the amount of lexical information), ‘rigidity’ (the ability to select a unique referent, based on the form), and ‘attenuation’ (phonological size). The more informative, rigid, and unattenuated a referring expression, the less accessibility it encodes. Examples (81) and (82), taken from Jones (2014: 162), illustrate this. (81)

There are two roads to eternity, a straight and narrow, and a broad and crooked.

(82)

After the news broke that comedian and actor Robin Williams had been found dead at his California home, tributes began to pour in from friends and colleagues.2

2

From The Guardian 12/08/14. Available at: www.theguardian.com/film/2014/aug/12/robin-wil liams-death-online-tributes (accessed 19 August 2019).

84

Typicality and Atypicality

The sentence in (81) is an example of a high degree of accessibility, where the gaps created by the unexpressed categorisation (e.g. one or road) in a straight and narrow [one/road] and a broad and crooked [one/road] are lexically uninformative, phonologically unattenuated, and lacking in rigidity (i.e. the use of ellipsis is flexible and can be applied to a limitless number of referents). In example (82), on the other hand, the noun phrase comedian and actor Robin Williams is an instance of a full name with pre-modification. It is lexically informative, highly rigid in that it can only apply to one individual, and therefore indicates that the speaker or writer does not consider the referent to be currently accessible to the addressee; instead, the expression is used to introduce the referent into the discourse. Interestingly, Ariel (1990) does not include indefinite noun phrases in the scale, even though they are probably the most typical means to introduce a new entity into the discourse. While the Accessibility Marker Scale may be more ‘reference’ oriented, indefinite expressions should still have a place on the scale as they can be used to refer to a specific entity, newly introduced into the discourse. Consider, for example, the utterance in example (83), which was said in a scenario where a mother was looking for her son at a supermarket. (83)

I’m looking for a little boy with red hair (from Jones, 2014: 164)

Here, the mother clearly has a specific referent in mind, even though she does not assume that the addressee knows the identity of her son (although she certainly hopes they can at least identify him). There is, then, a case to be made for indefinite noun phrases to have a place on the Accessibility Scale, although it should be noted that Ariel’s model only accounts for the accessibility status of antecedents (Reboul, 1997). Other similar scales do indeed include them (see Givón, 1983; Gundel et al., 1993), and we will turn to one such scale in the next sub-section. To sum up, speakers (including writers), in their choice of referring expression, are influenced by how accessible they deem the intended referent to be for their addressee (Vogels et al., 2019). Different types of referring expressions mark for a higher or lower degree of accessibility. Based on this scale, which type of referring expression is typical in a given context is dependent on the level of the intended referent’s accessibility. 4.3.2

The Givenness Hierarchy

Similar to the Accessibility Scale, the Givenness Hierarchy (Gundel et al., 1993, 2019) evaluates which type of referring expression is deemed appropriate based on the intended referent’s status in the interlocutors’ shared common ground. As opposed to Ariel’s mutually exclusive levels of accessibility,

4.3 Typical Reference in focus

>

activated

>

familiar

85 >

uniquely

>

referential

identifiable {it}

{that, this, this N}

{that N}

{the N}

>

type identifiable

{indefinite

{a N}

this N}

Figure 4.2 The Givenness Hierarchy (Gundel et al., 2019: 68)

however, this hierarchy comprises six cognitive statuses that are implicational, which means that each status entails all the lower statuses (Gundel et al., 1993: 275). The six statuses of the Givenness Hierarchy are displayed in Figure 4.2 alongside the hypothesised English forms they are assumed to encode (see also our discussion of Givón’s (1993a) continuum of referential intent in Chapter 2). As is explained by Gundel et al. (2019: 69), by using a particular form, a speaker/writer indicates that the intended referent matches the respective cognitive status of that form, and that the hearer/reader is able to correctly identify the referent. Moreover, because all lower cognitive statuses in the hierarchy are entailed in the higher ones, a speaker/writer also implies that all the lower cognitive statuses are met. For example, a referent that is in focus is necessarily also activated, familiar, uniquely identifiable, referential, and type identifiable. In the following, each of the six cognitive statuses will be considered in turn. The lowest cognitive status of a referent is type identifiable, which means that the addressee is able to understand what type of entity is being referred to (cf. type specification as discussed in Chapter 3). For instance, in example (84) the addressee is expected to know the denotation of dog and use that knowledge to categorise the instance of that class. This type of expression was discussed in Chapter 3 in terms of individuative indefinite reference by Radden and Dirven (2007). (84)

I couldn’t sleep last night. A dog (next door) kept me awake. (Gundel et al., 2019: 69)

For the second cognitive status, referential, the addressee is not only expected to know the type dog, but also to ‘either retrieve an existing representation of the speaker’s intended referent or construct a new representation by the time the sentence has been processed’ (Gundel et al., 2019: 69). The use of this dog in example (85) illustrates how the speaker signals to the addressee that they are talking about a particular dog, but that the addressee is not expected to be able to identify the referent as distinct from other members of the class. (85)

I couldn’t sleep last night. This dog (next door) kept me awake. (Gundel et al., 2019: 70)

86

Typicality and Atypicality

The referent, however, is only signalled as having the status, of uniquely identifiable when the definite article is used, as shown in (86), although the referent is not necessarily familiar to the addressee. They can make sense of the referring expression and identify it as one particular dog as long as enough descriptive material is provided to single out the referent, that is, next door in this case (see Gundel et al., 2019: 70). (86)

I couldn’t sleep last night. The dog next door kept me awake. (Gundel et al., 2019: 70)

In instances where the status is familiar, as in example (87), the addressee is assumed to be able to not just recognise the referent as uniquely identifiable, but also use knowledge the speaker believes they have previously gained in relation to this particular referent (e.g. perhaps as part of their long-term memory, or if the referent has recently been mentioned). The demonstrative pronoun that indicates that the speaker assumes that the addressee already knows of their neighbour’s dog. In this model, the status of familiar is a necessary condition for all personal pronouns and definite demonstratives (Gundel et al., 2019: 71). (87)

I couldn’t sleep last night. That dog (next door) kept me awake. (Gundel et al., 2019: 71)

Building on this shared common ground between the interlocutors even further, the next stage in the hierarchy is ‘activated’. As we read example (88), we might be left wondering what that refers to. We feel like some information is missing. However, if we imagine that the speaker and addressee are together at the speaker’s home and they can both hear the sound of a dog barking, then it becomes much easier to understand the connection between the two sentences. Here the speaker is referring to something from the context of situation (recall our discussion of register above), that is, exophoric reference, through the sound of dog barking. The activated status is assumed to always apply to the speech participants themselves, and it is also required for all pronominal forms (Gundel et al., 2019: 71). (88)

I couldn’t sleep last night. That kept me awake. (Gundel et al., 2019: 72)

Finally, the highest cognitive status in the Givenness Hierarchy is when the referent is in focus. In other words, the referent is not just activated ‘in shortterm memory, but is also at the current center of attention’ (Gundel et al., 2019: 72), that is, it is the current focal topic in a conversation. The Givenness Hierarchy identifies the speaker’s assessment of the addressee’s level of activation for the referent in question and associates the various levels with certain lexical forms. Such a correspondence does assume a one-to-one relation, which is an appealing idea, but in practice we often find

4.4 Towards an Account of Atypicality in Reference

87

that linguistic forms are recycled for more than one function. Nevertheless, these form predictions do provide us with a general description of typical reference, depending on the activation level. In the next section, we will turn our attention to what it means for a reference to be considered atypical, although based on what we have said here, we can expect that this will involve a mismatch of one kind or another. 4.4

Towards an Account of Atypicality in Reference

This section examines uses of reference that deviate from their expected forms as predicted by the Accessibility Scale or the Givenness Hierarchy. Both of these theories predict the form of a referring expression that is needed in order to make an intended referent identifiable for the addressee. We first discuss reasons why identifiability might not be the only function of a referring expression, and that when such other functions are at play, the form of the expression may deviate from what is typically expected in the context. Following this, we examine what it means for a referring expression to be atypical, drawing on the discussion of typicality and atypicality in earlier sections. We also consider various examples of referring expressions that deviate from the norm in order to fulfil further pragmatic functions. 4.4.1

The (In)Stability of Identifiability

As we have seen in previous sections, one of the functions of referring expressions is to map the information in the current utterance onto the antecedent-trigger in the mental discourse representation of the interlocutors. The referring expressions should enable the addressee to identify the referent onto which the current information needs to be attached. There are several linguistic means to do this: in English, these means typically involve zero anaphors, pronouns, proper nouns, and definite noun phrases, all of which are considered to be definite. These expressions differ in their lexical specificity, which consequently impacts on their level of identificational explicitness (Vonk et al., 1992: 302; Schiffrin, 1994: 199). Which device is chosen by the speaker seems to depend largely on how effective it is considered to be in terms of fulfilling its identificational role, which in turn is related to various factors, such as the interlocutors’ representation of the ongoing discourse event, the co-text and context, the assumed degree of accessibility, or the cognitive status of the entity in question and the intentions of the speaker. The pragmatic principle of the maxim of quantity can be applied here too; the choice of referring expression seems to be a matter of locating it on a continuum which balances providing sufficient information for identification with not providing too much.

88

Typicality and Atypicality

However, perhaps identification of a referent is not the only function of referring expressions. The wordings of referring expressions are not intended to purely assist in the identification of the referent, but rather also to fulfil a different discourse function (cf. Hanks, 2019). Fox (1987) found that anaphora is also governed by rhetorical organisation. For example, the principal determinant of whether a pronoun is used rather than a repeated noun phrase is whether the referring expression is in the same structural unit as the initial mention. Further, a full NP is often used at the beginning of ‘a new rhetorical unit’ (Fox, 1987: 136). Fox also suggests that non-structural factors play a role in the choice of referring expression, such as ‘categorisation of the referent, further information about the referent, and comparison and contrast of the referent with other people’ (ibid.). In addition, Vonk et al. (1992: 303) found that referring expressions that are more specific than necessary for identification of the antecedent indicate ‘an episode boundary’. They use the following text to illustrate this point (Vonk et al., 1992: 303): 1. 2. 3. 4. 5. 6. 7.

Sally Jones got up early this morning. She wanted to clean the house. Her parents were coming to visit her. She was looking forward to seeing them. She weighs 80 kilograms. She had to lose weight on her doctor’s advice. So she planned to cook a nice but sober meal.

The use of the pronoun she in sentence 5 above may not cause any identificational problems, but the more specific Sally would ‘[make] the sentence sound better’ (Vonk et al., 1992: 304). Vonk et al. (1992) showed empirically that when a device is used that is more specific than needed for the recovery of the intended entity, it also has a discourse structuring function. That is, it marks the beginning of a new theme concerning the same discourse referent. So, it seems that referring expressions do not function merely as identificational devices, but discourse restraints also play a role in the choice of expression. It is generally accepted that once a speaker introduces a referent for the first time, subsequent mentions of the referent continue in a fairly predictable way. That is, an initial mention is typically introduced or presented using an indefinite noun phrase (but if the speaker assumes that the addressee is able to recover the identity of the referent, a definite NP may be used), and subsequent mentions (which are therefore recoverable) tend to be encoded as definite (as we discussed in Chapter 2). But as Fries (2001: 89) notes, ‘it is not always the case that speakers match actual recoverability with presented recoverability’, and in such cases the referential identities of the referents are perhaps not what is most important in the discourse.

4.4 Towards an Account of Atypicality in Reference

4.4.2

89

Atypical Reference as an Exploitation of the Norm

Earlier in this chapter, atypicality was defined as a low-frequency, unconventionalised occurrence of a pattern in a given register. Similar to Hanks’ (2013) exploitations, such uses bring additional meaning, often making a further pragmatic implicature. In referring, atypical uses of expressions can be identified when there is a mismatch between the form of the expression and the form expected by the givenness status of the intended referent. If typical referring expressions follow the expectancies of the Givenness Hierarchy in order to allow the addressee to identify the referent, then atypical references will deviate from this hierarchy, either giving more or less information than is necessary to identify the intended referent. It is this under- and overspecification of the referent which then leads to pragmatic implicatures with functions that go beyond the simple identification of the referent. We will come back to this point in Chapter 8, where we examine formally indefinite expressions which are being used to refer to a highly accessible referent. For now, let us consider the extract in example (89), which has been taken from the first page of the novel All the Pretty Horses by Cormac McCarthy (1992). (89)

The candleflame and the image of the candleflame caught in the pierglass twisted and righted when he entered the hall and again when he shut the door. He took off his hat and came slowly forward. The floorboards creaked under his boots. In his black suit he stood in the dark glass where the lilies leaned so palely from their waisted cut-glass vase. Along the cold hallway behind him hung the portraits of forebears only dimly known to him all framed in glass and dimly lit above the narrow wainscotting. He looked down at the guttered candlestub. He pressed his thumbprint in the warm wax pooled on the oak veneer. Lastly he looked at the face so caved and drawn among the folds of funeral cloth, the yellowed moustache, the eyelids paper thin. That was not sleeping. That was not sleeping.

Several paragraphs later, the reader is introduced to another referent by means of a pronoun. Note also that the identity of he has still not been revealed. (90)

She looked up from the stove when he came in and looked him up and down in his suit. Buenos días, guapo, she said

It is not until eleven paragraphs later that we learn the name of the protagonist, John Grady Cole. The she turns out to be a character only known as Abuela (grandmother in Spanish), an old Mexican woman who had lived on the ranch since the turn of the century and who helped raise John Grady. We never find out her name. According to the Givenness Hierarchy, use of the personal pronouns he and she in the introductory paragraphs to All the Pretty Horses would mark

90

Typicality and Atypicality

these referents as ‘in focus’, that is, the highest givenness status. Due to the implicational nature of the hierarchy, this means that the two referents are also ‘activated’, ‘familiar’, ‘uniquely identifiable’, etc. However, at this point in the story, the reader is not yet able to identify the two referents at all. As the narrative unfolds, the discourse representation of the referent develops in that we learn about his actions upon arriving home, his clothing, and the time in which the scene was set. The reader, therefore, has a mental representation of the referent, but identification is still not possible in terms of referent resolution, and the referential content of the discourse representation of the individual is not ratified until we encounter a lexical expression. Hence, the reader is here presented with an atypical use of reference. Usually, when introducing a new entity into the discourse, one might expect the use of a proper name (e.g. John Grady Cole), or an indefinite noun phrase with added descriptive material (e.g. a man in a black suit and hat). Instead, the writer has chosen to refer to the referent simply by pronouns for the extent of several paragraphs. This is an atypical use for a pronoun, which should, in theory, signal identifiability (i.e. being able to pick out who or what we are talking about and not simply a type of entity with particular qualities). As we will see in Chapter 7, this type of incongruency would make it difficult for the addressee to accept the referring expression. In a conversation, if a speaker uses a personal pronoun and the addressee does not share the level of activation assumed by the speaker, the addressee will request clarification and the conversation will not continue until the identity of the referent is resolved. However, if we consider this use of pronouns within the register of fictional writing, we find that this use is a literary device that allows the writer to ‘trigger the interpretation that a discourse entity is highly prominent’ (Epstein, 2002: 349). In using this literary strategy, the writer chooses ‘a definite description to introduce an important entity at the start of a narrative, for the purpose of calling the reader’s attention to that entity’ (ibid.). Because this choice of expression for a referent that is not in focus is deliberate and creative, we may want to think of it as an exploitation (Hanks, 2013), although perhaps it has become a norm, to some extent, as a conventional literary device. We consider this use to be atypical since the personal pronouns in example (89) are intended to indicate not that the referent should be viewed as in focus but rather that the entity will be the primary topic of concern in the immediately following discourse. Holding back on the identity of the individual forces the question of identity to be addressed by detail and inference, which contributes an element of suspense. Clearly, here, the function of definite descriptions is not a straightforward matter of identifiability.

4.5 Summary of the Chapter

4.5

91

Summary of the Chapter

The aim of this chapter was two-fold. First, we wanted the chapter to serve as an introduction to the current section on typical reference and the third section of the volume on atypical reference and extensions. Second, we explored and developed our understanding of the concept of typicality in relation to atypicality. We settled on a view of typicality as conventionalised language use which can be said to fall within the norms of a given register, and atypicality as an instance of language which can be said to be infrequent and unconventionalised for a given register. The following two chapters explore typical reference in two different modes: spontaneous discourse (Chapter 5) and written discourse (Chapter 6). The remainder of this book, in addition to providing an empirical account of reference in the atypical uses of reference in Chapters 7 to 10, aims to show that these atypical uses serve the needs of the speaker, expressing functions which go beyond the central function of identifiability of referring expressions. As a result, we will argue that atypical reference, despite being atypical by our definition, is often actually quite normal and common. It is more a matter of perspective.

5

Referring in Spontaneous Discourse

I don’t outline at all; I don’t find it useful, and I don’t like the way it boxes me in. I like the element of surprise and spontaneity, of letting the story find its own way. — Khaled Hosseini

5.1

Introducing Spontaneity

In the previous chapter, we contrasted the concepts of typical and atypical expressions in an attempt to capture a sense of norms and deviations from norms in terms of both forms of expressions and their use. This chapter aims to present a picture of typical referring expressions; however, we recognise that what is typical is dependent on a variety of factors. As we noted, register, or how language varies according to use, is an important consideration when making a distinction between what can or cannot be considered as typical language. In what follows, we will use the concept of context, as originally set out by Halliday (1970, 1978, 1991), to explain why we have opted for the term ‘spontaneous’ to describe the context of reference we will be discussing. The literature on referring expressions uses the classifier ‘spontaneous’ to capture a type of dialogue or conversational interaction (e.g. Gundel and Johnson, 2013 or Hughes and Allen, 2015) or to refer to instances or uses of specific linguistic features without prompting or priming (e.g. Branigan et al., 2016). Our intention here is to make a distinction between what we can claim about typical referring expressions that approximate spontaneous production in different communicative modes and what we can claim about typical referring expressions when more planning and revising are involved, as we will see in Chapter 6. To help define how we think of spontaneous language, we will first consider the concept of register and its relationship to context, and we will use this to provide some parameters for how we are viewing spontaneous discourse. Following this, we will have a look at what features of referring we find in spontaneously produced language. Here, we will try to capture the most common or frequent types of referring expressions found in spontaneous discourse. 92

5.2 Mode: Defining Spontaneous Language

5.2

93

Mode: Defining Spontaneous Language The plumber came yesterday to service the boiler. He was supposed to just give it the once-over and be on his way. But the bloody incompetent idiot ended up knobbling the boiler completely and now we don’t have any hot water.

The above quote could easily represent a written transcript of part of a spoken conversation or a written message such as an email or a post on social media. This instance of language use has a spontaneous feel to it, in part because of the type of communicative act (a simple recount) but also in part because of the lexical choices (knobbling, bloody, idiot). The example also displays certain referential features that we would recognise as not belonging to more formal, written registers of language. For example, the use of definite expressions such as the plumber and the boiler signal that the speaker assumes the addressee already knows that a plumber was expected and that there was a problem with the boiler. Without that assumption, the text would have had to be organised differently, perhaps beginning with the problem with the boiler to introduce the need for the plumber. We assume here that the speaker and addressee know each other and that the interlocutors are likely co-present, although this short text could also have been sent by text message to someone who was very far away and who might not read the message for hours hence. It is becoming increasingly difficult to maintain a simple dichotomy between spoken and written language. The past twenty years have seen an exponential increase in the awareness of different modes of production and different registers. As a result, this has opened up the space between spoken and written language, which we can think of, metaphorically, in terms of dimensions of language use. It seems reasonable to assume that there are a variety of features affecting the wording of a referring expression. Corpus-based studies have shown that noun phrases are important indexes of text type. For example, Quirk et al. (1985: 1352) made claims about ‘how sensitive the noun phrase is as an index of style and how responsive it can be to the basic purpose and subject matter in varying types of discourse’. Biber et al. (1998: 108) have shown that ‘texts from different registers often differ dramatically in the use of these referring expressions’. This type of linguistic variation, that is, variation according to use, or register, has been described by Halliday (1970: 324; see also Halliday, 1978) as ‘context’, or more technically ‘context of situation’. As we saw in the previous chapter, for Halliday, register consists of three variables or parameters. To briefly recap, the three register variables include the field of discourse, the tenor of discourse, and the mode of discourse. These three variables are described by Halliday (1981/2002: 227) as follows: field represents the social context, seen as ‘the nature of the social process in which

94

Referring in Spontaneous Discourse

the text is embedded – “what is going on”’; tenor includes social relationships and roles, where a text is viewed as ‘the interpersonal relationships among the participants – “who are taking part”’; and mode concerns the symbolic organisation of language, including ‘the role assigned to the text, including both medium and rhetorical function – “what part the language is playing”’. However, it is important to situate this view of register as part of a broader theoretical framework which integrates language within a model of context (see Wegener and Fontaine, in press). For Halliday, language serves a fundamental social function, and it is shaped as it is because of the functions it has come to serve (Halliday, 1970: 324). Given that we want to talk about different types of referring expressions, we do need to consider what the options are and how these expressions may vary according to their use. Berry (2017: 43) explains that ‘it is impossible to make sense of linguistic choices without reference to context and, also that it is impossible to make sense of context without reference to linguistic choices. Linguistic choices are strongly influenced by the context of the situation in which they occur’. It is this context of situation that is captured by register, allowing us to account for variation in language, whether that is, for example, developmentally as with language learning, in different social uses, or in terms of the mode of production. While the general sense of ‘context’ can include the material setting, that is, the immediate environment, the relationship between context and language is one of meaning (or semantics). Context, in this sense, can be modelled as a type of knowledge (or information) which is held by the speaker. Fawcett (1980) identifies three types of context with different sources of information, which can be quite useful for our discussion of reference: ‘context of situation’ (recoverable from the observable), ‘context of co-text’ (recoverable from preceding text), and ‘context of culture’ (recoverable from long-term memory). This view of context as information held by the speaker is the generally accepted one, although despite the use of this term being prevalent in the literature, it is rarely defined and, in some cases, its use suggests a meaning similar to ‘circumstances’. For example, Abbott (2019: 169) uses ‘context’ as follows: ‘In these contexts (i.e. in sentences with other operators), definite descriptions can pattern with the quantificational NPs rather than the referential ones’. In this use, context means co-occurrence (i.e. local linguistic environment). This sense of context is not the one we are focussing on here. Instead, we are establishing a term to capture the various influences which the speaker is either conscious of or accessing unconsciously through some automated reflex. Roberts (2019: 302–303) proposes the following definition of context: We take an utterance to be an ordered pair of a linguistic constituent s and a context of utterance c. As is common, we take c to specify a great deal of information not only about the concrete situation in which the utterance was made – the speaker,

5.2 Mode: Defining Spontaneous Language

95

addressee, location, time, etc. – but also about the interlocutors’ Common Ground and perhaps other information that is systematically tracked in a conversational record.

This approach to context is more aligned with the situational context we have been discussing. The difference is that our discussion has distributed the features of context along the registerial variables of field, tenor, and mode, all of which need to be included. For example, lack of visual co-presence has a significant effect on the use of deictics, and we would expect a reduction in certain types of deictics, and certainly in the use of gestural deictics, where it is known that visual access is not shared. Anyone who has had a telephone conversation with a two-year-old will have figured this out quite quickly. As noted above, ‘mode of discourse’, or simply ‘mode’, includes both medium and rhetorical function. The distinction between mode, as it is generally understood, and medium is difficult to capture, as is the case for the terms channel and medium. We are not helped by the fact that, as O’Halloran (2004) points out, these terms are not used consistently in the literature. For example, we find mode used to capture modes of communication in studies of multimodality. Prior to computer-mediated communication, specifically digital rather than analogue communication, the separation of channel and medium was, to some extent, easier. Traditionally, spoken language and written language would have been considered two different modes of communication, or indeed two different codes. This is largely due to the concept of mode being developed to account for visual (image-based) communication as well as spoken communication and written communication. For O’Halloran (2004: 20), mode refers to ‘the channel (auditory, visual, or tactile, for example) through which semiotic activity takes place, medium for the material resources of the channel, and genre for the text types’. With digital writing, for example, email, the channel is visual (although arguably electronic light on a screen could be classed as a different channel to physical print and indeed to sign language, and yet these all involve the visual sensory system). Kress and van Leeuwen (2001: 2) explain that ‘in the age of digitisation, different modes have technically become the same at some level of representation’. It is important to note here that Kress and van Leeuwen’s use of mode differs from O’Halloran since they use it to refer to a semiotic resource, whereas O’Halloran draws a distinction between mode as channel and as a semiotic resource. The examples listed above involving a visual channel can be differentiated by medium, whereby the computer is the medium for digital writing, paper and pen/pencil for handwriting, and hands and space for sign language1. For our purposes, we

1

For accounts of reference in Auslan, Australian Sign Language, see Johnston and Schembri (2007), and for a detailed description of the noun phrase in British Sign Language, see Rudge (2022).

96

Referring in Spontaneous Discourse

will simply consider that our use of mode includes both, combined with rhetorical function (or purpose), which includes, for example, persuasive versus descriptive language. Mode covers all aspects of language use that relate to its symbolic organisation and the role language plays in the ongoing activity.2 Normally, an auditory channel requires interlocutor co-presence, but so does a visual channel where sign language is used. Written visual language typically does not occur with interlocutor co-presence because its very nature displaces the need for the interactant’s co-presence. However, in the case of synchronous digital language such as online chat conversations, the synchronous nature of the technology requires co-presence. Even in the case of near-synchronous text messaging, where the interactants are not required to be co-present but could be, there is a feeling of interaction. Hasan (1998: 242–245) explains that ‘the value of dialogue in graphic channels is different from that of dialogue in the phonic channel where it is a genuine option’. Since digital channels and other visual channels (e.g. sign language) must also include dialogue as a genuine option, we cannot think of spoken language and written language as being distinguished on this basis. In other words, dialogue or interactivity as a genuine option is part of any channel. Synchronous and near-synchronous interaction needs to be separated from channel and medium. We can assume that the more asynchronous linguistic interaction is, the greater the potential and availability of planning. Another way to think about degrees of planning in language production is to consider how much conscious choice is involved. The degree of consciousness is different from, but related to, the degree of interactivity discussed in the previous paragraph. Berry (1975, 2013) developed the concept of ‘a cline of consciousness’, which she explains as ranging from fully subconscious to fully conscious. This cline allows us to distinguish between, for example, a conversation with a friend and a job interview, or indeed between online messaging, which would tend towards a lower level of consciousness, and a Wikipedia entry, which would tend towards a greater level of consciousness. It may be more useful to interpret Berry’s use of ‘conscious’ in terms of amount of planning and revising involved. For example, greater consciousness on the cline for Berry may mean more planning and/or revising and/or more deliberate choice. While we have no reason to assume that the speaker of one type of text is more or less conscious than another, it may be useful to consider how

2

In this view of mode, all aspects of text/discourse are included in terms of how the discourse is organised; the type of mode of production, including channel and medium; the purpose of the text/discourse (speaker’s intentions); and whether language is itself the activity (e.g. interview, speech, essay) or whether it is ancillary to another activity (e.g. talk while playing chess or a sport).

5.2 Mode: Defining Spontaneous Language

97

effortful the production of text is. An informal, conversation-like text message may be thought of as ‘less conscious’ in the sense that it is less effortful and involves less planning and revision than other types of texts. However, the topic might be sensitive (e.g. someone has had some bad news), or the relationship might be a difficult one, and consequently, even a short text may be drafted and redrafted with a lot of consideration going into the wording of the text. Similarly, a work-related email may be spontaneous (on the part of the writer) but formal (in terms of tenor), and a personal email may be more planned depending on the nature of the communicative exchange (e.g. arranging to meet for coffee vs expressing sympathy). Therefore, it is not possible to categorise individual texts/discourses such as email or word processing along these clines, since the factors influencing the linguistic choices relate to register. To return to the issue of synchronicity, aside from chat messaging, for example, writing and typing are typically carried out with little to no immediate feedback, which allows the writer time to consider the reader, the purpose of the text, and the amount of detail and clarity needed (see Chapter 6). Less interactivity and greater asynchronicity generally provides greater opportunity to plan and possibly rehearse language before speaking, writing, or signing it. A good counterexample of planning in synchronous interaction is an oral exam which takes place with co-present interlocutors, but where the questions asked are planned ahead of time and the candidate taking the exam has studied and, to some extent, prepared answers. While comments such as they spoke without thinking are commonly heard in spoken communication, a comment such as they wrote/typed without thinking is much less likely to occur. By using the term ‘spontaneous’, we are trying to capture uses of language that are produced ‘without thinking’, as a feature of mode. Mode does have an effect on language production, but not one that is easily separated from field and tenor. For example, the field of discourse plays an important role because the topic may be very familiar or completely unfamiliar either to the speaker or addressee or both. For example, the more technical or specialised the topic, the more information is needed in referring to discourse referents (e.g. see example (91)3, expression highlighted in bold). Especially in the case of scientific discourse, we may expect to find high lexical density even when the mode is spoken and spontaneous. While shared common ground might be viewed as a feature of mode and related to the textual or discoursal organisation of information, it is fundamentally about shared knowledge of the field of discourse, which is expressed by the types of expressions used to represent participants and the processes in which they participate. 3

Source: www.investmentexecutive.com/news/research-and-markets/funds-banks-face-challengesfrom-digital-dollar/ (accessed 6 April 2022).

98

Referring in Spontaneous Discourse

(91)

Traditional money market funds and other kinds of cash investments could be disrupted by the introduction of a digital dollar in the U.S.

Tenor also overlaps with mode since it concerns the nature of the relationship between collocutors. It is probably fair to say that the vast majority of our daily interactions with family and friends are at the lower end of Berry’s cline of consciousness, which was discussed above. We are, to some extent, on autopilot in these interactions, which are largely automatic, frequency based and, to some extent, formulaic. This type of interaction is most likely to occur with people we know well and with whom we have an informal relationship. ‘Degree of formality depends on the kind of social relationship between discourse participants, including their relative status, gender, age, etc.’ (Kibrik, 2011: 13). Greater social distance and increased formality in the social setting will naturally lead speakers to give more (or less) thought to what they are saying. However, as Berry (2017: 43) points out, ‘in certain circumstances linguistic choices may bring about a change in the context, for instance leading it to become more formal or more informal’. For example, in a parent–child interaction, if the parent suddenly addresses the child with their full name, for example, Lise Margaret Fontaine, a shift in context will be signalled, or created, by the language act itself. According to O’Donnell (2021: 14), context is dynamic; ‘we should be taking the approach such that every act creates its own context, which sometimes is coherent to the context created by prior acts of the interactants’. Shifts of this nature can be sudden and unexpected or gradual and anticipated. The excerpt in example (92) illustrates how, despite the same physical (or material) setting, a shift in tenor is identifiable. The comments made by Gordon Brown during the 2010 election in the United Kingdom provide a good example of clearly spontaneously produced language. It was not scripted nor planned. His off-record comments to an aide were captured by his microphone, which had not been turned off. In this exchange, he refers to a member of the public, who had just been speaking to him, as a sort of bigoted woman, comments which many claimed at the time cost him the election. (92) brown: aide: brown: aide: brown: aide: brown:

Exchange between UK Prime Minister Gordon Brown and his aide (2010): That was a disaster. Well I just . . . should never have put me in with that woman. Whose idea was that? I don’t know, I didn’t see. It was Sue, I think. It was just ridiculous. I’m not sure if they will go with that. They will go with that. What did she say? Oh everything, she was just a sort of bigoted woman. She said she used to be Labour. I mean it’s just ridiculous.

5.3 Overview of Referring Expressions in Spontaneous Communication 99

There is some information that is striking about this example as compared to the one mentioned at the beginning of this chapter. The plumber example could almost stand alone in the sense that, even if we do not know which plumber is being discussed, we can easily follow the meaning of the message. Here, we really do not know what Gordon Brown is talking about from the text itself. We know something was a disaster, but we do not know what. We do not know who Sue is and we do not know the identity of they, although we can assume they refers to the press/media. What might not have been immediately obvious is that each and every clause in the exchange has a pronoun in subject position and one that is either definite in the sense of anaphorically referring to something already said or known to the addressee (e.g. it, she, they), or deictically referring to the speaker or exophorically to something in the material context (e.g. I, that). As will be discussed in the next section, this high frequency of pronominal reference in subject position is indeed a feature of spontaneous language production. Before moving on to our description of referring in spontaneous language, we want to establish what we mean by spontaneous in relation to referring expressions. Contrary to most uses of the term ‘spontaneous’, we are not restricting this term to conversational dialogue. Spontaneous language can be written or signed, and in this sense, it is not an inherent feature of channel or medium. It is a variable of register, and in particular mode, because of its influence on production. However, it is dominantly a feature of the individual speaker, who has control over how spontaneous their language will or should be. We can say that the more spontaneous the language production is, the more we can expect to find high-frequency items, more phoric items, more formulaicity, and more repetition. The reason for this is ease of processing, and the evidence for this will be discussed in the next section. 5.3

Overview of Referring Expressions in Spontaneous Communication

It is actually very difficult to provide a description of typical referring expressions in spontaneously produced communication, in part for some of the reasons mentioned above. Many factors influence the nature of referring. For example, distinctions need to be made between given and new information in the expression of a referent. Of course, this applies irrespective of the degree of spontaneity of the language being produced. However, as we will see, for English but also for most languages (Kibrik, 2011), the clausal position of the referring expression will influence the lexicogrammatical realisation of the expression. ‘Referential choice is made by the speaker on the basis of a variety of factors’ (ibid.: 15), including discourse structure, discourse type, and

100

Referring in Spontaneous Discourse

lexicogrammatical choice. In Chapter 9, we will discuss this point in relation to children’s use of reference. In his study of spoken discourse,4 Du Bois (2003b) has shown that lexical NPs are relatively infrequent. He proposes a hypothesis called Preferred Argument Structure,5 which makes the claim that, in spontaneous discourse, ‘certain configurations of arguments are systematically preferred over other grammatically possible alternatives’ (ibid.: 33). The connection between arguments expressed by lexical NPs and pronouns or names is directly influenced by the accessibility of the referent and reflects the connection between the clause grammar and discourse referents. In English, there are basically three configurations of argument structure: clauses with only one argument, that is, intransitive clauses; clauses with two arguments, that is, transitive clauses; and clauses with three arguments, that is, ditransitive clauses. In each case, the argument is most typically expressed by some type of nominal expression (although in the case of the ditransitive construction, the third argument may be expressed by a prepositional phrase which includes a noun phrase or a pronoun, as in He gave the parcel to the girl vs He gave her the parcel vs He gave it to her). The degree of lexicality of the arguments in a clause is directly related to the degree of information encoded in the expression. In this sense, any so-called new information will almost necessarily be expressed by a lexical NP. One of the key features of spontaneous communication is a greater proportion of given information or accessible referents as expressed most typically by pronouns. As Du Bois (2003b: 20) shows, most transitive clauses in spontaneous speech will include at least one lexical NP, but ‘pronoun-only clauses’, as shown in example (93) taken from Du Bois (ibid.) ‘are reasonably common’. (93)

He kissed her.

Du Bois proposes four constraints or preferences on the expression of arguments along two related dimensions: grammar and pragmatics. The implication of these preferences is that for any transitive clause, the subject argument is much more likely to express an accessible, given referent by the use of a pronoun or proper name and, given English clausal configurations, any new information will be expressed by the complement or direct object. This position is indeed the case put forward by Halliday (1967), who very early made the claim that, in English, given information will map onto initial clause position (e.g. Subject-Theme) and new information will occur in the Rheme (post-subject position).

4 5

Santa Barbara Corpus of Spoken American English (Du Bois, 2000, 2003a). This term was originally proposed in Du Bois (1987).

5.3 Overview of Referring Expressions in Spontaneous Communication 101

In a sample of 5326 clauses of spontaneous discourse, Du Bois (2003b: 35) found that 47.4 per cent of these clauses had no lexical NPs in argument positions and 45.3 per cent had only one argument expressed by a lexical NP. Only 7.3 per cent of all instances involved two lexical NPs. These figures are influenced, of course, by the frequency of intransitive clauses; however, we know that transitive clauses are by far more frequent in any discourse in English (Chrispin, 2022). If we turn now, as Du Bois (2003b: 37) does, to the clause distribution of the lexical expressions, we find a total of 257 lexical NPs where 57 per cent occur as complements (direct objects) and only 8 per cent in a transitive subject position. The remaining lexical NPs occurred as subject in intransitive clauses (35 per cent). These results lead Du Bois (ibid.) to conclude that ‘[t]he grammatical realization of arguments in a clause does not take place in a functional vacuum. It is tied to cognitive and pragmatic factors like information management, which influence the realization of arguments as lexical or pronominal’. The tendency described above suggests a dominance of pronouns in spontaneous discourse. It is important to note that this description is not suggesting that it is not possible to have more than one lexical NP in any given clause in this context, or that there is anything ungrammatical about it, but rather that it is less frequent. Example (94), taken from Du Bois (ibid.), illustrates this point well as there is clearly nothing atypical about these clauses in spontaneous discourse. However, the general preference for pronoun reference in spontaneous discourse is clear, and where lexical NPs occur, they tend to express complement arguments, as illustrated in example (95), which was discussed earlier in Chapter 2 (example (23)). (94)

My wife would write a check for ten dollars.

(95)

John went to see the doctor and he did some tests and he also said he had a sinus infection. I’m glad he finally went.

The use of pronouns is a key feature of cohesion according to Halliday and Hasan (1976). They use the term reference to refer to lexical items which have a semantic interpretation ‘in their own right’ from those items which are ‘directives indicating that information is to be retrieved from elsewhere’ (ibid.: 31). To avoid any confusion with the more general use of reference, for the purposes of this discussion we will refer to Halliday and Hasan’s (1976) type of reference as ‘cohesive reference’ (see Fontaine, 2013: 169). In Figure 5.1 we illustrate the categories of cohesive reference, where we find a fundamental split between referents which are exophoric, that is, part of the context of 6

There is an error in Du Bois’ (2003b: 35) table. The total given is 535 but the sum of the categories is 532.

102

Referring in Spontaneous Discourse Cohesive reference

Exophoric

Endophoric

[situational]

[textual]

Anaphora

Cataphora

[to preceding text]

[to following text]

Figure 5.1 Categories of cohesive reference (adapted from Halliday and Hasan, 1976: 33) speaker only speaker speech roles

speaker plus addressee(s)

/ we

you

person

male human

singular specific other roles

plural they generalized human one

non-human

he

female she it

Figure 5.2 Classification of person reference (Halliday and Hasan, 1976: 44)

situation, and those which are endophoric, that is, part of the ongoing discourse (text). In this view, there is a difference to be made, for example, between speaker role reference (i.e. speaker or addressee) and what we might call thirdparty reference. The speaker’s self-reference and reference to the addressee are situational and classed as exophoric. We might think of this as textindependent. Those references to discourse referents are classed as endophoric and may either be used to refer anaphorically to something already mentioned or to signal an upcoming referent cataphorically. An alternative way to look at pronoun use is presented by Halliday and Hasan’s (1976: 44) system for person reference, as given in Figure 5.2. While speech roles (i.e. speaker-reference and addressee-reference) are not the only type of exophoric reference, they are dominant in spontaneous discourse. Other types of exophoric reference, such as the use of demonstrative pronouns, for example, that, as in That is my coffee (invented example), are not considered part of the person reference system but nevertheless involve situational (deictic) reference, typically only used when collocutors share the same material setting, to some extent, at the same time. Figure 5.3 highlights the difference between exophoric and endophoric person reference, including what

5.3 Overview of Referring Expressions in Spontaneous Communication 103

Figure 5.3 Types of phoric relations (adapted from Martin, 1992: 124)

Halliday and Hasan (1976: 53) call generalised exophoric reference, where ‘the referent is treated as being as if it were immanent in all contexts of situation’. If we try to combine Figures 5.1–5.3, we will see that speech roles will all be exophoric and therefore situation dependent, and all specific ‘other roles’ will be endophoric and therefore text (or discourse) dependent; in other words, in Halliday and Hasan’s classification of person reference, reference is either to a speech role or co-referential (e.g. by anaphor) within the text, much like the examples we discussed in Chapter 1. We are then left with pronouns that serve a generic or non-specific referential function. References that sit outside the situational context are referred to by Halliday and Hasan (1976) and also Martin (1992) as homophoric reference since this type of reference ‘is used when interlocutors’ membership in a particular community means that certain participants can be treated as inherently “given”’ (Martin, 1992: 122). Proper names fall into this category, as do referents such as the sun or the dog, as in Have you fed the dog? However, whereas for Halliday and Hasan (1976) homophoric reference is a type of exophoric reference, Martin (1992) argues that homophora should sit outside this division because these types of expressions are outside context of situation, unlike exophoric and endophoric references, which are both inherently situational. In this view, as presented in Figure 5.3, both endophoric and exophoric reference are viewed as dependent on situational context, and all other phoric types are captured by homophora. According to Martin (2019: 361), ‘[p]horic entities presume information which has to be recovered from the co-text or nonverbal context, for which a range of recovery strategies is proposed (anaphora, cataphora, exophora, homophora, etc.)’. These different phoric relations offer a way of capturing the relationship between a phoric item (e.g. pronoun, definite article) and its context ‘by way of presumed information’ (Martin, 1992: 121). Although we are focussing here on pronoun use in spontaneous discourse, it is perhaps worth noting that, according to Martin (1992: 142), ‘first mentions

104

Referring in Spontaneous Discourse

of participants in discourse are commonly phoric. Exophoric and homophoric first mention is quite common’. It would be very surprising to have an endophoric first mention given how the recoverability of the intended referent would be assumed and, if it is a first mention, it would have to be cataphoric, which we can reasonably assume would be more challenging for the addressee to process. We saw an example of this in (89) in Chapter 4, where a personal pronoun was used for a referent that was not in focus in terms of the Givenness Hierarchy. Normally, however, first mentions, where they do occur, do not occur in subject position. Kärkkäinen (1996: 694) explains that, where that does happen, the referent is ‘almost always New + Identifiable’, noting that ‘these tend to be particular or specific referents that are well known to the participants (in the case of proper names and kinship terms), or general NPs that are semantically bland’. In other words, ‘even though they are clearly New, [they] do not seem to contribute a great deal of information’ (ibid.: 693). This view of certain new referents to the discourse aligns with Martin’s (1992) work on first mentions, although Martin did not specifically consider clause structure in relation to homophoric and exophoric first mentions. In their large corpus-based study of noun phrases, Quirk et al. (1985: 1351)7 analysed over 4,000 NPs from a corpus of informal speech (N = 4,192), where they found that proper names and pronouns made up 46.3 per cent of expressions in subject position, while 49.24 per cent of subjects were simple NPs, which included bare nominals (e.g. water or trees) and determiner + noun combinations without any modifiers (e.g. the water or a tree). Quirk et al. (1985: 1351) also examined NP distribution in their corpus of serious talk (N = 4,599), and while perhaps still spontaneous to some degree, it can be assumed to be less spontaneous than informal speech. We do find that proper nouns and pronouns have a lower frequency (32.14 per cent) in subject position in serious talk, as do simple NPs (37.94 per cent). NPs with modifiers were much more notable in this position (a frequency of 7.46 as compared to 3.53 per cent in the corpus of informal speech). Similar results were found in Fontaine’s (2008) corpus-based study of informal email conversations, which we consider as spontaneous discourse. In total, 2,473 referring expressions were analysed across two speakers. The distribution of these expressions is given in Table 5.1. The most frequent type of expression found in the corpus involved proper names and pronouns (61.5 per cent). Of all pronoun uses, 46.5 per cent involved interactant reference, that is, speech roles following Figure 5.2, which represents 22.7 per cent (N = 561) 7

The frequency data provided in Quirk et al. (1985) concerning NPs include ‘some blatant mistakes’ (Jucker, 2012: 94), and therefore we have used the figures provided by Jucker (2012: 95).The percentages given here represent the number of NPs of the total number of NPs in the subcorpus.

5.3 Overview of Referring Expressions in Spontaneous Communication 105

Table 5.1 Distribution of type of referring expressions Type of referring expression

N

%

Lexical expressions Non-interactant pronouns Interactant pronouns Proper names Interrogative pronouns Other (date, time, address) Total

929 646 561 313 8 16 2,473

37.6 26.1 22.7 12.7 0.3 0.6 100.0

of all expressions in the corpus. The frequency of speaker self-reference among interactant reference was 75 per cent, which indicates a very high prevalence of speaker self-reference. Pronoun use in non-interactant roles made up 26.1 per cent of all expressions and 33.9 per cent of all noninteractant referring expressions. The vast majority of these pronouns were anaphoric, making up 71.7 per cent of all such pronouns (18.7 per cent of all 2,473 expressions). While we cannot make claims that this sample of spontaneous discourse is representative, the frequency distribution presented in Table 5.1 is very much in line with the evidence from the literature discussed above. The relatively high frequency of pronouns and proper names as compared to lexical expressions is illustrated in example (96), taken from Fontaine (2008), where we find only one lexical referring expression, the book, which is underscored in the example. (96)

I haven’t talked to Lois so I don’t know if she got the book, you could have sent it from here it would be cheaper.

What the distribution in Table 5.1 shows is that only 37.6 per cent of expressions are lexical and the remainder, 62.4 per cent, which constitutes the vast majority, are made up of pronouns and proper names. While this corpus was not analysed for given versus new informational status, we can rely on previous work to make some assumptions. According to Kärkkäinen (1996: 683), ‘conversational discourse contains not only fewer lexical NPs but also fewer New arguments than narrative data’. Narrative discourse may be quite spontaneous or more prepared or planned, but by its very nature, it typically involves introducing a variety of new referents to the addressee. Du Bois (1987) offers an explanation of the relatively low frequency of lexical referring expressions in such spontaneous discourse by considering the influence of what he refers to as information pressure. Information pressure is often low – such as intimate conversation between family members or long friends, where interlocutors may refer to each other with 1st and

106

Referring in Spontaneous Discourse

2nd person pronouns, and otherwise share large amounts of currently active background information. Especially when one looks at conversation, new issues arise, because of the much larger amount of background information that is typically shared between participants, and the concomitant rarity of new entity references; introductions of new human participants may be especially rare. (ibid.: 835)

Information pressure may explain the relative infrequency of new entity references in the email discourse presented in Table 5.1. As Du Bois suggests, greater shared background information will typically require less new information to be brought in to the discourse. Indeed, Stegemann-Philipps et al. (2021) show, in an experimental context, that the more surprising the referent, the more likely the speaker is to use a full lexical NP. According to the authors, ‘an unexpected turn in the unfolding of an event progression calls for a referring expression that could boost the salience of the agent to the listener’ (ibid.: 2902). 5.4

Concluding Remarks

While it is difficult to capture, in any definitive sense, what spontaneous discourse is, it is perhaps most useful to think of it in terms of a fluid range from spontaneous to more deliberate and prepared language production. All language use is embedded in context; we cannot label one type of text as spontaneous or not, but we can identify features of language that is more conscious, deliberate, and rehearsed and/or revised. Context, as we have seen, is dynamic. Even within the same conversation, the topic could shift in such a way that results in a parallel shift from more spontaneous to more effortful, for example, if some bad news is shared or the conversation becomes difficult for any reason. We should therefore avoid a strict spontaneous versus planned division since it is generally accepted that all language production is planned to one extent or another and these distinctions tend rather to be a matter of degree. What we have tried to capture in this chapter on spontaneous language is language that is on the lower end of Berry’s (1975, 2013) cline of consciousness (mode), informal in terms of social interaction (tenor), and familiar in terms of topic (field). While this description is overly simplistic, it does capture a basic account of what we mean when we discuss spontaneous discourse. In terms of reference, what we find from the evidence presented above is that the consequence of this combination of features results in the dominance of a referential choice of pronouns, especially in subject position. There is very little use of the kind of nominal complexity we saw in Chapter 3 in the qualification zone. Typical reference in spontaneous discourse, then, relies very strongly on highly accessible and in focus referents, which makes it much easier for acts of reference to be successful and reduces the cognitive load of

5.4 Concluding Remarks

107

the speaker and addressee. In Chapter 7 we will return to the features discussed here as we discuss some problems in the referential process. In the next chapter, we continue with our discussions of typical reference but here, rather than spontaneous discourse, we will be considering reference in written discourse, which is generally much less spontaneous and more planned, with various revision and editing stages.

6

Referring in Prepared Discourse

Writers with time to plan, edit, and rewrite, . . . should satisfy their responsibilities to readers by eliminating everything but elementary proposals, and many writers do. (Clark and Wilkes-Gibb, 1986: 36)

6.1

Time and a Word

In the previous chapter, we made the claim that there are features of register that can be important to understand in terms of how they influence our understanding of referring expressions. In our discussion of Halliday’s three parameters of context of situation, or registerial variables, we noted that while field and tenor are important, it is really mode that has a dominant role to play. In particular, we argued that differences in mode of production in relation to the degree/absence of spontaneity in language production was one important factor, and we showed that even when language is written, as in the case of informal email conversation, the types of expressions used are very similar to informal spoken conversation. Spontaneous discourse, then, contrasts with more planned, deliberated language production. Chukharev-Hudilainen (2014) makes a similar distinction in his work using the terms spontaneity and preparedness. Prepared discourse is perhaps a better term than planned discourse, given that all language is planned at one level or another. What we are essentially trying to capture is the idea that this more prepared written language is, as Halliday noted, ‘in essence a more conscious process than speaking’ (Halliday 1992/2002: 336). Writers have time, time to choose their words carefully, and this is a valuable resource which speakers typically do not have. One interesting feature of spoken language is that, other than memorising a speech, it is very difficult to plan our text. Once a text is memorised, the spoken text is almost a copying task, translated from memory and not produced for the immediate purposes of the discourse. Written texts are ideal for examining language produced with a high degree of preparedness for a variety of reasons which we will discuss below. As we said in the previous chapter, we cannot categorise any specific 108

6.2 Mode and Prepared Writing

109

mode of discourse or instance of language as spontaneous or not spontaneous, but we do know that writing for publication (e.g. articles, fiction) or for academic purposes (e.g. essays, reports) generally follows a model of written language production that includes some degree of planning and revising and/or editing. However, writing can be spontaneous, as shown by Fontaine and Aldridge (2015), who used keystroke logging methods to observe differences in mode. Even without technology, we can imagine students in a classroom sending notes back and forth behind the teacher’s back. These notes will most likely simulate spoken conversation, and we would expect the students to write quickly without too much thought and get the message sent as quickly as possible. Of course, it is also possible that the students in question might want to send a message that is very carefully planned. What we want to do in this chapter is provide a contrastive perspective to the spontaneous discourse we considered in Chapter 5. For this purpose, we will concentrate on what we might think of as institutional writing (writing text in academic, professional, and social settings) due to the high degree of preparedness involved. In what follows, we will first consider the parameters of register, with a focus on mode, in relation to prepared writing as we did for spontaneous discourse in Chapter 5. Then, in Section 6.3, we consider the interactive aspect of referring when it comes to prepared writing, given that we emphasised in earlier chapters how important collaboration is to referring. Section 6.4 then focusses on the typical profile of referring expressions in prepared writing. As we saw in Chapter 5, the high frequency of pronominal reference that we find in spontaneous discourse is substantially reduced, and we will look at what replaces those types of referential expressions. The question of the relationship between reference and genre is raised in Section 6.5, where we consider the evidence for reference to be viewed as genrespecific. The chapter concludes in Section 6.5 with a summary of both this chapter and the set of three chapters making up this part on typical reference. 6.2

Mode and Prepared Writing

All instances of language use are embedded in a particular contextual situation. The parameters of this situation have been described within Halliday’s (1981/ 2002) model of register, as we saw in Chapter 5. To remind ourselves, we will consider examples (97) and (98), each taken from Halliday (1992/2002: 331), which he identifies respectively as ‘more spoken’ and ‘more written’. This distinction (i.e. spoken vs written) is based on the channel and medium rather than on features of mode related to planning and preparing, and whether or not the addressee is co-present at the point of production. We have added emphasis to these examples to indicate, in bold, each clausal subject and to identify the verbal groups associated to each subject, which are underscored. What is quite

110

Referring in Prepared Discourse

striking about these two examples is how most nominal expressions in (97) are pronouns and that example (98) has far fewer verbal groups than (97). This observation about pronoun use is in line with what we would expect based on what we have learnt about typical reference in spontaneous discourse. (97)

Whenever I’d visited there before I’d end up feeling that other people might get hurt if I tried to do anything more.

(98)

Every previous visit had left me with a sense of the risk to others in further attempts at action on my part.

While Halliday notes that spoken English differs from written in terms of clause complexity, as we see in these two examples, the key distinguishing feature is that ‘written English is marked by complexity in the nominal group’ (Halliday 1992/2002: 343), as with a sense of the risk to others in further attempts at action on my part in (98). This complexity typically features in the qualification zone of the noun phrase, as we saw in Chapter 3. We will return to this issue in Section 6.4 when we consider what typical referring expressions look like in prepared written texts. In what follows, we will provide a brief overview of some of the key features about writing, including the most well-known model of writing, to help us understand the importance of mode in the study of reference. Outside of experimental methods, such as those used in psycholinguistic research, the study of writing has traditionally focussed on writing as a product rather than a process. The same is true for the study of speech. There has been, however, far more interest in writing processes from outside psycholinguistics than there has been in speech processes. The writing process is typically modelled by Flower and Hayes (1981), who posit that writing comprises three main phases (planning, translating, and reviewing), which interact with each other in a cyclic fashion and not necessarily in any particular order. The planning phase involves setting goals, generating ideas, and planning how best to organise them. The translating phase concerns language production, that is, getting ideas and concepts from the planning phase into linguistic form, for example, writing sentences and paragraphs. Finally, in the reviewing phase, which can be subconscious or conscious, writers are reading what has been written, and evaluating and revising it. Importantly, reviewing does not take place in a linear sequence only once a text has been written, but rather ‘[t]he sub-processes of revising and evaluating, along with generating, share the special distinction of being able to interrupt any other process and occur at any time in the act of writing’ (ibid.: 374). All three phases are governed by the Monitor (ibid.: 367), which monitors (as the name suggests) all aspects of active writing and progress. Our main two mediums for writing are handwriting with pen and paper and digital writing with a computer. Irrespective of the medium, writing requires

6.2 Mode and Prepared Writing

111

significant cognitive effort (Piolat et al., 1996). The phases of the writing process are always competing for limited cognitive resources (i.e. working memory) (Kellogg, 2001), which is why the level of difficulty of the task is important; the more difficult the task, the more challenging the process of writing becomes. In terms of the role of medium, typing is a bi-manual, discrete activity at the level of the keyboard, while handwriting is a unimanual, continuous activity. In addition, when handwriting, the visual focus of the writer is the tip of the pen, but while typing, visual attention is detached from the haptic input because the eyes typically switch from keyboard to screen. When writing by hand, planning is a critical and important element, and the writer invests more in these stages because revision of a final text can be costly (which is a similar issue if the text is being typed on a typewriter). The use of a word processor makes editing text easy, and the writer can spend less time planning. There is a similar trade-off with transcription (cf. Sharples, 1996; Hartley, 2007), where we find that handwriting is a relatively complex process which includes (a) retrieving orthographic representations from longterm memory, (b) parsing those representations into graphemes, (c) retrieving the forms for each grapheme, and (d) activating appropriate motor sequences. In contrast, typing involves simpler graphemic processing and motor sequences and so may impose less transcription load on text generation, all else being equal. What seems apparent from this brief discussion is that while medium may play an important role in varying the writing process and how different stages of the writing process are affected, there is no reason to think that a handwritten text will have any difference in terms of referring expressions. Selfmonitoring in language production is not restricted to the writing process. It is well-established that speakers self-monitor (Levelt, 1989), and it is widely agreed that monitoring serves the purposes of ‘correcting errors already made and preventing such errors from recurring’ (Nozari and Novick, 2017). What happens in prepared written language is that time affords greater monitoring, revising, and re-translating (i.e. revising the plans, wording, and/or organisation). As we noted above, one key distinction between written and spoken texts concerns complexity in the noun phrase. Halliday (1992/2002) explains this in terms of lexical density; written language will generally have more content words per clause (i.e. content dense) as compared to spoken language which will have fewer content words for the same text length, which means more frequent function words. In their study of nominal expressions in a geology textbook as compared to those from a classroom lecture, Biber and Conrad (2009: 56–57) found that per 100 words, the textbook used 29.1 nouns compared to 10.8 nouns in the spoken lecture, and 2.5 pronouns compared to 15.3 pronouns in the lecture (2009: 65). We know that pronouns are used for things that the addressee is

112

Referring in Prepared Discourse

familiar with or can easily identify (i.e. an entity that is ‘in focus’ (Gundel et al., 1993) or highly accessible (Ariel, 2001)), or for things that are present in the immediate environment. As Biber and Conrad (2009: 64) note, the classroom provides more opportunities to refer exophorically (i.e. to things observable in the situational context), whereas a textbook which is produced does not. Personal pronouns such as I, you, and we were noted in the lecture text, demonstrating potential interaction between speaker and audience. This interaction also provides the possibility for clarification if the referring expression is ambiguous. As this feature is missing from the written text, the textbook writers therefore need to ensure that their referring expressions are explicit enough to avoid any ambiguity, thus the use of full noun phrases. Now that we can observe writing during the production of text with the use of keystroke logging software, we can gain some insight into the types of revisions writers make which evidence their awareness of how successful their referring expressions are likely to be. In a study of writing processes using keystroke logging methods, Bowen (2016) found that the noun phrase played a particularly important role in revisions and the management of information in the samples of academic writing he studied. The most frequent types of nominal revisions involved the determination zone and the categorisation zone. Determiners such as the and a were often exchanged, typically to revise the identifiability of the referent to a lower level of accessibility, as illustrated in example (99), which we will return to in a moment. Changes in the categorisation zone mostly involved a change in type specification (see Chapter 7). One of the key findings of this study was that ‘nominal structures were made more informationally dense, and that part of the clause which was likely to be more salient to the reader also became more salient to the writer come reviser’ (ibid.: 213). From this perspective then, it is not simply about creating a lexically denser text but rather about attempting to address the needs of the reader. During revisions, the writer becomes the reader, and if they do so well enough, they become more aware of the accessibility and identifiability of the referents in their text. If we now return to example (99) and compare it with a later revision of the same part of the text, which is given here as example (100), we can see that the nominal the command is changed to a command, that is, the definite article the is deleted and the indefinite article a is inserted. In these examples, the revisions are indicated as follows: deleted text is in italics and bounded by curly brackets, and insertions are indicated in bold surrounded by square brackets. As noted by Bowen (2016), writers were regularly revising their noun phrases from those presented as recoverable to those which had a lower status of accessibility. This direction to a lower accessibility status may seem surprising but, as Bowen suggests, it may

6.2 Mode and Prepared Writing

113

well indicate the shift from the writer who is familiar with all the referents in the text to the reader who is not. Bowen notes that the change we see in example (99) was surprising since this was not a first mention and indeed the writer had already introduced command as a non-specific expression. Further changes were made to this sentence after other revisions had been made in the text (the revision in example (99) is the 73rd revision in this text and the one in example (100) is the 75th revision). The changes in example (100) provide some clarity. As Bowen (2016: 158) explains, the writer in this case seems to have noticed a potential problem with the expression and consequently adds a second type (order). (99)

{The} [A] command is threatening to the hearer, J. (Bowen, 2016: 158)

(100)

{A}[An order or] command is threatening to the hearer, J[‘s negative face]. (ibid.: 159)

Additional examples of similar types of nominal revisions are given in (101) and (102), where we can see how writers revise nominal expressions in ways which anticipate the reader’s needs. In example (101) the shift is from a broad but definite generic reference to one that is indefinite due to the use of a secondary (partitive) determiner (some sections). Although the use of this in example (102) is typically seen as an instance of a definite expression, and indeed the writer will have had a specific referent in mind as this sentence was written, during revisions the reader-writer recognises that perhaps the reader of the text will not be able to identify the referent. The expression, the Norman invasion and consequent transition into Modern English, is not only a lexically dense nominal expression, but it includes important lexical information, such as the type specification, which this does not. (101)

Bernstein claimed that while the middle-class were likely to have access to both codes, [some sections of] the working-class were likely to have access only to the restricted code. (Bowen, 2016: 158)

(102)

Although{this isg [the Norman invasion and consequent transition into Modern English] is a specific example of where the Chronicle can illustrate changes in English. (ibid.: 168)

There are clearly differences between written language and spoken language in terms of medium; one is visual, and the other is aural. However, as we have tried to show here, it is not so much these differences that influence acts of referring, but rather features related to spontaneous versus prepared discourse, and whether or not the speaker and addressee are co-present (e.g. share the same immediate context of situation). The most significant way that this difference is manifested is in relation to greater lexical density, which shows up mainly in the complexity of the noun phrase. We will look at the form of

114

Referring in Prepared Discourse

referring expressions in writing in Section 6.4, but for now we want to pick up on the point we made above about how the writer views the reader and how we can still see referring as a collaborative, interactive act. 6.3

How We Make a Jointly Constructed Text ‘Remotely’

In any written text, the reader must keep track of incoming information and construct some kind of organised representation of the entities present in the text by relating new information to old. It is, of course, the writer’s job to provide sufficient cues to ensure that the reader is able to do this efficiently and with the least effort, at least in English (cf. Weil, 1844). As such, the writer’s choice of wording and skill in signalling referents through nominal expressions can make it more, or less, difficult for their readers to make the intended connections and suppress the wrong ones. In the written mode, this skill is even more critical than in the spoken mode because the writer and reader are not ‘co-present’ (cf. Clark and Wilkes-Gibbs, 1986) and misunderstandings cannot be repaired ‘live’. On the other hand, the writer, in prepared writing, has time to plan, revise, and rewrite their message, as we saw in the previous section. In principle, this advantage should ensure that referring expressions are clear, unambiguous, and easily ratified by the addressee. Normally, readers will have no problems keeping track of discourse referents, although of course this is not always the case. As we saw in Chapter 1, experimental research into referring expressions has devoted considerable attention to the notion of collaboration between speaker and addressee. It is generally agreed that collaboration between interlocutors helps participants refer successfully, as was discussed in previous chapters. Schober and Clark (1989: 211–232) propose that the traditional view of gathering evidence about a speaker’s intention, which they call the ‘autonomous view’, is flawed. This view places emphasis on the listener’s role in decoding each utterance and interpreting it against what they assume to be the common ground of the interlocutors. Instead, the ‘collaborative view’ (Schober and Clark, 1989) suggests that participants in a conversation actively work together to ensure that understanding takes place, and they do not proceed with the conversation until they are satisfied that they have a mutual understanding. In our discussion of revisions to nominal expressions in the previous section, we saw evidence to suggest that writers are attending to this need to ensure understanding between them and their readers. However, the lack of copresence means that the reader cannot stop the text and ask the writer to clarify. We will see examples of this in Chapter 7. The question we must ask at this point, then, is how this collaboration is possible in the written mode. Indeed, in written modes, the writer and reader are not co-present when the text is written

6.4 Referring Expressions in Written Texts

115

or read, so there is no opportunity for readers to signal a problem or ask a question, nor is there any opportunity for the writer to revise any of the expressions – it is too late because they are not interacting in a face-to-face setting and cannot see if the other is confused or has arrived at an unintended interpretation. Further, how we understand referring in writing depends on how we view the writing process. If we view writing as a static, monologic process, then it is, of course, difficult to imagine how this jointly constructed process could occur. However, if writing is considered to be a dynamic, intertextual, interpersonal, and dialogic process, then we can start to think about the relationship between the writer and the reader as allowing for some kind of joint construction, albeit at a distance. Even though the writer cannot directly or actively respond to the reader’s level of understanding, ‘each feels the other’s presence in a way that strongly influences their behaviour towards the text’ (Fox, 1987: 93). In the written mode, it is all about the writer’s anticipation of the reader’s (somewhat incomplete) understanding of the text; at the same time, the reader must speculate about the writer’s intentions, form understandings about them, and abandon them if they no longer fit (ibid.: 94). As we mentioned in Chapter 1, Clark and Wilkes-Gibbs (1986: 34) proposed a weakened or modified version of their principle of mutual responsibility, which they call the principle of ‘distant responsibility’. This principle means that the writer has to ensure that the reader understands their meaning ‘to a criterion sufficient for current purposes’ (ibid.: 36). So, the writer and the reader still need to be on the same wavelength to some extent, but the onus is more on the writer to make sure they have provided enough support for the reader to be able to interpret the referring expression appropriately.1 In this situation, it could be that the writer needs to provide other pointers to help the reader, for example, certain cotextual or semantic clues, such as the ongoing sequence of references to the same referent or the fact that the heading or title is the discourse topic of the text. 6.4

Referring Expressions in Written Texts

We have already seen that English has a rich set of forms for achieving reference, including articles (e.g. a command/the command), pronouns (e.g. it), demonstratives (e.g. this/that command), and full lexical noun phrases

1

We are only making this claim for English. There is debate in the literature as to whether or not different languages do or should shift the responsibility to the reader. English is generally, however, viewed as a writer-responsible language, which means the onus of responsibility for the coherence of the text is on the writer (see Hinds (1987) for discussion of writer versus reader responsibility and MacKenzie (2015) for an alternative perspective).

116

Referring in Prepared Discourse

(e.g. a command that is threatening to the hearer). The factors which influence the choice of referring expression in writing are, to a large extent, the same as those in the spoken mode. However, as we noted above, the writer has a bigger share of the responsibility to ensure clarity in referring because if the reader encounters any problems, there is no opportunity to resolve them. For example, a referring expression must be sufficiently informative to disambiguate it from potential competitors. We know that a pronoun is likely to be used to refer back to an antecedent in the immediately preceding sentence or clause because the referent is at the current centre of attention or ‘in focus’ (Gundel et al., 2006: 1). However, if there are two potential competitors for the interpretation of the pronoun (i.e. two potential referents), there could be some ambiguity, particularly in the written mode when any contextual (exophoric) information is missing. As seen in example (102), the use of a pronoun can be a high-risk referential choice unless the writer is very certain the reader will be able to ratify the expression, that is, the reader will be able to accept the referring expression and believe they have a sufficiently clear understanding of the identity (or lack of identifiability) of the referent. The reason pronouns are particularly challenging is because most personal and demonstrative pronouns used by speakers to refer are in subject position and are typically anaphorically co-referential with a referent that has already been mentioned in the text, that is, endophoric reference. Subject pronoun use can, however, refer back to a previous mention of a referent, or they can be used anaphorically to refer to an event expressed by a clause. For example, the use of this in examples (103) and (104), taken from Bevacqua et al. (2021: 193),2 illustrates the potential problem of an expression that does not provide sufficient information for the addressee (in this case the reader). In both cases, this could potentially be used to refer to wine as the referent (as in (103)) or to the event of ageing, as in (104). In both cases, the referent is clarified by the predication, that is, attributing a quality versus a causative use of make. (103)

The Northern Italian wine aged well. This was one of the best wines I’d ever tasted.

(104)

The Northern Italian wine aged well. This made the wine more marketable.

The use of a demonstrative pronoun, such as this or these, as a pronoun rather than a determiner followed by a noun is commonly referred to as ‘unattended this’ (Gray, 2010). An example of unattended this is given in (105), where the expression of interest is highlighted in bold.

2

The sentences given here are experimental and not naturally occurring.

6.4 Referring Expressions in Written Texts (105)

117

The most common reasons given were that computers made writing “easier,” facilitated the correction of spelling and punctuation mistakes, were “modern,” “useful,” “interesting,” and “faster” than writing by hand. This corroborates conclusions from other researchers, who report that CMC in the second language classroom facilitates general communication and the revision process . . . (Gray, 2006: 35)

While pronominal uses of this/these (i.e. unattended this/these) are infrequent compared to uses of this/these as determiners (i.e. attended this/these), Gray (2010: 181) notes that they ‘overwhelmingly refer to antecedents that are complete clauses (but not extended discourse that spans sentence boundaries) [and] when the demonstratives are followed by a noun, shell nouns and abstract nouns are used most of the time’. If viewed from the Givenness Hierarchy, as explained by Gundel (1996: 146), the use of this or this+N ‘explicitly signals that the referent is activated (since this is a necessary and sufficient condition for appropriate use of the proximal demonstrative determiner) [and] anything which is activated is also familiar, uniquely identifiable, referential and type identifiable’. What this tells us is that the writer, at least, believes all of these things about the referent and assumes the reader will also be able to interpret the expression as the writer intended. In terms of our discussion in Chapter 4 concerning what is or is not typical, both types of expression are typical, although with different relative frequencies. However, the pronominal use of this/these will force the reader to assume the antecedent is evident, and as we all know in our role as readers, whether for work or pleasure, this is not always the case.3 All of the above can apply to both spoken and written language, of course. But given that there are established differences between the spoken and written modes of production, there must also be some differences between the use of referring expressions in the two modalities. While much has been written about referring in spoken language, as we saw in the previous chapter, less attention has been given to the written mode. So, what features of referring are specific to the written mode? In order to answer this question, we must bear in mind that the written modality, like the spoken, is not made up of a homogenous group of texts; within each mode, there are many different genres or registers, all of which are likely to vary in their ‘typical’ features of referring. We will now briefly consider some differences in the forms of the expressions themselves before we turn to differences in written text types in Section 6.5. In Chapter 5, we examined typical referring expressions from a broad and general perspective in spontaneous discourse. We can draw on that description to contrast it with what we find in prepared discourse. Returning again to Quirk

3

In the spirit of certainty, this this refers to the propositional antecedent, the antecedent is evident.

118

Referring in Prepared Discourse

et al.’s (1985) study of noun phrases from the Survey of English Usage, rather than focussing on informal speech and serious talk, we can turn to fiction writing and scientific writing. As a reminder, a simple noun phrase in their account includes those NPs where the head is not modified, which includes those NPs expressed by a bare noun or by a determiner + noun combination. Also, as we did in Chapter 5, we will follow the data as presented in Jucker (2012) since they provide a corrected version of the findings given in Quirk et al. (1985: 1351). As noted by Quirk et al. (1985: 1350–1351), scientific writing differs most significantly from other text types due to a much higher frequency of complex noun phrases and, specifically, a much higher frequency of noun phrases with multiple modification. In subject position, NPs in scientific writing are most frequently expressed by simple NPs, as is the case for all four corpora in the study (informal conversation, formal talk, fiction, and scientific writing). However, the complex NPs in subject position have a frequency of 15.22 per cent, which is more than double what was found in serious talk, fiction, or informal speech (informal speech has the lowest frequency of complex NPs of all, as we saw in Chapter 5). We see the same trend with NPs expressing more than one modifier and/or qualifier. If we consider the frequency of NPs with multiple modifiers, as a total of all NPs in scientific writing, Quirk et al. (1985: ibid.) report a frequency of 23.81 per cent, whereas, for comparison, the frequency of complex NPs with multiple modifiers in fiction writing is much lower (10.05 per cent). Biber et al. (1999) show similar results in their corpus of academic writing, where 60 per cent of all noun phrases have some type of modifier. Fontaine (2008) found that 76.3 per cent of NPs (excluding those expressed by a proper name or a pronoun) included a modifier and/or a qualifier. However, we cannot make accurate comparisons without knowing how the NPs were analysed. As we saw in Chapter 3, one person’s determiner is another person’s qualifier. For example, for Biber et al. (1999), the first lexical noun of the NP was identified as the head. Therefore, any expressions following this head would necessarily be coded as a postmodifier. The problem with this approach is illustrated by the following two examples: a piece of cake or a lot of trouble (from ibid.: 635). Both NPs would be considered complex for Biber et al. because they consider the underscored items as postmodifiers. In the functionally oriented description we proposed in Chapter 3, both cake and trouble would be considered as expressing the type element (i.e. in the categorisation zone), and any grammatical complexity, if we want to think of it that way, would be in the determination zone. Examples (106) and (107) illustrate complex NPs, expressed by modifiers and/or qualifiers (categorisation zone indicated in bold, modification and qualification zones in italics).

6.4 Referring Expressions in Written Texts

119

(106)

a living laboratory where people who had a direct interest in the forest supported by the latest science and technology would become partners in sustainable forest management decisions (Fontaine, 2008)

(107)

fossils of extinct sea creatures called ammonites that lived in coiled shells resembling the modern coiled nautilus (Biber and Conrad, 2009: 56)

Formal writing contains fewer deictic references than informal conversation because formal writing aims to ‘approach absolute intelligibility, that is to create its own context so that it can be read in any situation’ (Smith Jr, 1986: 111). Thus, the presence of deictics in writing may have the effect of making the text seem less formal. In example (108), we can see how the use of personal deictics creates a level of informality which closes the distance between writer and reader and is thus more interpersonal. (108)

With a general election imminent we need to ensure every Labour supporter is registered to vote. You can find the link below, please ensure family and friends are registered. @labourfuture 17/08/2019

Compare this to example (109) from an academic text. (109)

In this paper, it is argued that the standard decision theoretic axioms provide core principles action-guiding preferences of rational agents should be required to satisfy. This, however, does not involve a commitment to consequentialism. [. . .] This is possible because these principles should be interpreted as an expression of coherentism, not of consequentialism. (Taken from the abstract of Rationality: Coherence and Structure by Julian Nida-Rümelin (2000))

The use of passive structures rather than personal deictic reference (I and you, respectively) in the above text creates a distance between the writer and the reader in order to present a more objective stance. What this overview has shown is that typical referring expressions, in terms of their form, are quite similar, irrespective of how spontaneous the discourse is and whether or not it is produced through a written medium or a spoken (aural) medium. This similarity is due to the fact that it is the same noun phrase structure (as a potential) that is at the basis of referring expressions, that is, all referring expressions draw on the same set of resources. However, what is different is the distribution of the various types of expressions. Notably, the use of pronouns, proper names, and other ‘simple’ NPs is dominant, irrespective of mode, as a general pattern. There is considerable potential for encoding a higher density of information and specificity in the noun phrase, which is something writers need to exploit in ways which will ideally ensure the reader’s understanding. The two most significant ways for this to be done are through the choices available in the modification and qualification zones of the noun phrase. We see evidence of this in revisions that writers make as they

120

Referring in Prepared Discourse

produce their text, at least in academic writing, as shown by Bowen and Thomas (2020: 13), who explain that ‘Classifiers (pre-modifiers) and Qualifiers (post-modifiers) seem to have the greatest potential for fine-tuning a referent’s complexity and identity’. The greater lexical density that we see in prepared written texts is due in part to the writer’s need to support the reader, but also in cases due to factors related to the subject domain (e.g. greater use of sub-classification and nominal compounds) or to word-length constraints of certain types of texts (e.g. essays, published academic articles, news editorials). 6.5

Written Genres

In Chapter 5, and in the current chapter, we have been primarily discussing the influence of certain aspects of mode in our account of the main influences between more spontaneous and more prepared discourses, and in discourses we do include all forms of language production. We have not so far given much consideration to the structural and organisational aspects of discourse, which is another aspect of mode, at least in Halliday’s (1981) account of register. Some texts share registerial features to a significantly similar extent that we can classify them in terms of genre. There is some debate about differences and similarities between the use of the terms register and genre (cf. Lukin et al., 2011), but we do not really need to be concerned with that for our purposes. In this section, we will simply consider whether or not referential choice is genre-specific, that is, by the conventions of the text type. Based on our discussion in the previous section, it seems reasonable to assume that different types of text will have different distributions of simple versus complex noun phrases. We saw some evidence for this in the different type of texts considered above. However, only looking at the form of expressions does not tell us much about their meaning and use unless some kind of semantic or referential annotation is used. After all, it is the same set of lexicogrammatical resources which are available to writers (and speakers), and these resources are limited, at least in terms of the functional elements (or zones, as we described them in Chapter 3). This issue has led to some debate in the literature, which remains either resolved or not, depending on your perspective. Although perhaps others had begun to consider this problem earlier, the first major account of the relationship between genre and referential choice is found in work done by Fox (1987) in her study of third person anaphora in written and spoken English. Her approach was to specifically evaluate the relationship between discourse structure and anaphora. Her analysis of different genres in conversation and expository prose focussed specifically on anaphora. She found that structural aspects of the discourse could be used to explain choice

6.5 Written Genres

121

of referring expression, and she concludes that ‘there is no single rule for anaphora that can be specified for all genres’ (1987: 152). While there has been some support for this position (e.g. Clancy, 1980; Tomlin, 1987; Vonk et al., 1992), there has also been some criticism (Ariel, 1989; Cornish, 1989; Toole, 1996; Kronrod and Engel, 2001). In particular, Toole (1996: 263) argued not only that Fox’s approach was not valid due to criticisms of the methods used, but more significantly that the approach was unnecessary since Accessibility Theory (see Section 4.3.1) can ‘account for the use of referring expressions in all genres’. Toole (ibid.: 285–286) concludes that ‘there is a strong and significant association between accessibility and lexical choice in the four genres examined’ and that, therefore, referential choice is not genre-specific. However, there may be some relation between the organisation of text, that is, the register parameter of mode, which involves how given and new information is structured in text. Work by Jucker (2012: 252) on newspaper texts suggests that from a structural perspective ‘noun phrases can potentially carry very heavy informational loads. But their structure and complexity varies systematically, not only according to the three categories of newspapers, but also according to the individual newspaper sections’. Identifying noun phrase types and their frequency is quite straightforward; however, determining their informational value and referential role in discourse is more challenging. The problem with many of these studies is the overreliance on nominal types. As Schnedecker (2018: 40) explains, ‘reference theories – whether French or Anglo-Saxon – have hardly ever stressed the role played by these lexical categories in differentiating between discourse genres, and this includes Accessibility Theory (Ariel, 1990), even though it is based on a corpus of authentic texts’. In other words, reference theories have not fully taken ‘a semantico-referential viewpoint’ (ibid.: 64). The basic division in the debate is whether reference theory can account for reference in all genres or whether different accounts are needed for different genres. It is likely that the reality is somewhere between these two extremes. In part, it depends on the theoretical model of context (including genre) that you have in your framework. It is surely language use that construes genre (or text type) and not genre, as a language-external force, that determines referential choice. We will come back to this point in a moment after we consider two recent studies which have adopted a more semantico-referential approach. A recent study by Sarda and Carter-Thomas (2021) contrasted reference chains and referential choice in news articles in English and French. Their data consisted of twenty news articles, ten in English and ten in French. With respect to referential choice, they found that English journalists had a clear preference for proper names and pronouns, while the French journalists favoured what Sarda and Carter-Thomas call ‘anaphores infidèles’

122

Referring in Prepared Discourse

(inconsistent anaphors, our translation), which involve lexical anaphors that do not conserve type specification.4 Examples (110) and (111), reproduced here from Sarda and Carter-Thomas (2021: 50), illustrate respectively expressions (in bold) used to refer to French President Macron. In (110), we see the repetition of a proper name (his surname) and in (111), a series of lexical expressions including the apprentice-philosopher and the young graduate.5 (110)

Macron has used his luck smartly. His critics keep repeating that centre mainstream policies never win elections. Macron strongly disagrees, even if he still sees himself as on the left, though [. . .].

(111)

Animé par le goût de l’action publique et le souci d’être en prise avec les décisions politico-administratives, l’apprenti-philosophe quitte alors l’université pour entrer à l’ENA. On connaît la suite de l’histoire : le cursus de l’ENA terminé en 2004, le jeune énarque rejoint l’Inspection des finances et les hautes administrations [. . .].

The advantage of using lexical anaphors in this way is that they allow the writer to draw on a wider range of properties associated to the referent, such as ‘stable or “permanent” properties, for example those relating to professional or geographical relations (e.g. the Parisian), or “temporary” properties, linked to a particular context or point of view (e.g. the cyclist)’ (Sarda and CarterThomas, 2021: 63, our translation).6 While this study has not contrasted different genres, the comparison is nevertheless enlightening. Referential choice is not exclusively about accessibility. What we see here is evidence that for a referent that is in focus, we find that different expressions serve the writer’s purpose differently. See Mulkern (1996) for a discussion of referential variation in proper name use. The second study we wanted to mention is by Xu (2023), who uses Prince’s (1981) taxonomy of given-new information, or assumed familiarity,7 to analyse the information values of nominal expressions across four written genres (newspaper, travel guide, essay, and government document) from the 4

5 6

7

The French original is: ‘Alors que les journalistes anglais affichent une préférence nette pour les anaphores fidèles, les noms propres et les pronoms sans variation, leurs confrères français font un plus grand usage des anaphores infidèles, avec un noyau lexical qui varie. Les journalistes français profitent aussi de ces constructions pour apporter des informations subjectives (non factuelles) au sujet des référents’. The French word énarque refers to a student or graduate of the École nationale d’administration (ENA), the National School of Administration. The original quote in French is: ‘les propriétés stables ou “permanentes”, comme celles qui ont trait aux relations professionnelles, ou géographiques (le chef, le Parisien), et les propriétés “provisoires”, liées à un contexte ou à un point de vue particulier (le passager, le cycliste)’ (Sarda and Carter-Thomas, 2021: 63). Prince’s (1981) framework is different from those discussed earlier in this volume (e.g. Ariel, 1990; Gundel, 1996), but it is not so dissimilar that we cannot map aspects of each onto the other. See also Lambrecht (1994).

6.6 Summary

123

American National Corpus. Xu analysed 3,095 nominal expressions in a corpus of 13,243 words. Using a measure she calls dominance, she was able to compare given-new chains across different text types. A given-new chain with a lower dominance value would indicate that there are fewer repeated mentions of certain information components within the chain. For example, Xu found that the given-new chains in the travel guide texts had on average a lower dominance value than those of the newspaper texts, despite the fact that the two texts expressed similar amounts of information. Xu suggests that the dominance of information in newspaper text offers a reasonable explanation for the larger capacity for new information (including both Brand-new (unanchored) and Brand-new anchored in Prince’s terms) in travel guide texts, since components of given information were not repeated as often as they were in the newspaper texts. What this suggests is that it is not nominal expressions that are genre-specific per se, but rather the distribution of given-new information. In other words, the macro goals of the text suggest a preference for more or less new information, and this is then reflected through nominal expressions. We might say that the noun phrase mediates between the writer’s goals in producing the text and the writer’s distant responsibility to the reader with respect to reference. To bring this section to a close, we do not have a definitive answer about whether or not reference is genre-specific. However, as suggested above, we do suspect that there is a connection. In an attempt to work out whether it is possible to predict referential choice, Kibrik et al. (2016) combined approaches from theory, corpus analysis, machine learning, and experimentation in their study of referential choice in written text. They conclude that referential choice is not deterministic, or at least not fully deterministic: ‘There are certain conditions in which more than one referential option is appropriate and, in fact, each one would fare well enough. Under such conditions human language users may act differently on different occasions’ (Kibrik et al., 2016: 10). The findings from Kibrik et al. (2016) together with those discussed above suggest that while there is a strong correspondence between the form of an expression and its referential function, and that this varies to some extent according to the demands of the text (i.e. the writer’s purpose), the relationship is not predictive or deterministic in any strict sense. 6.6

Summary

In this chapter, we have focussed on referring in prepared written texts. In terms of register, the parameter of mode was argued to be particularly significant in defining some of the features of reference. While the principles of referring are quite similar in spontaneous and prepared discourse, some differences are noteworthy. Referring expressions in prepared discourse tend to be

124

Referring in Prepared Discourse

more frequently information-rich; the modification and qualification zones are more active in this type of discourse. Irrespective of mode and/or other register parameters, successful reference relies on the ability of the speaker/writer to produce an expression which adequately reflects their assumptions about the addressee’s knowledge, memory, and attention state. In other words, for successful referring, the expression has to be at an appropriate level of accessibility (Ariel, 1990), and it must signal the assumed cognitive status of the referent for the addressee (Gundel et al., 1993). We saw evidence that writers attend to this in our discussion of revisions in the writing process. We also considered variation in referential choice across various text types, but it was clear that comparisons are fraught with problems. It is difficult to compare reference when different frameworks are used. It is also difficult to know if comparison is appropriate as we cannot assume that we are comparing like with like. Stubbs (1993) stressed that not only do we need to concern ourselves with complete texts, but that we need to be aware that in a given text, linguistic features will not be evenly distributed. It might not make sense to compare genres without consideration of the structures that contribute to defining the rhetorical structure. As we know from studies of type-token ratio in text, the length of a text is also important. For example, how many new referents are introduced to a short text as compared to a long one? Do different texts and different sections of texts all have the same referential chaining (cf. identity chains (Martin, 1992))? We will not take up these questions in any detail here, but they are important ones to consider. See, for example, Kunz et al. (2021) or Amoia et al. (2012), who have carried out detailed studies. This chapter and the previous two have provided an overview of what we might reasonably consider typical reference. These three chapters establish the basis upon which to interpret the chapters in the next part, which explores atypical reference from a variety of perspectives.

Part III

Atypical Reference

7

Referring in Non-collaborative Contexts

When speakers refer to something, be it an everyday object or an exotic, unfamiliar one, they are faced with many choices. One person’s valuable antique may be another’s yard sale fodder. (Metzing and Brennan, 2003: 201)

7.1

Introduction

The perspective on reference which we have presented to this point might suggest that referring, while quite a complex process, works very well between collocutors. It is probably fair to say that this is true in the vast majority of cases. What we know about the production of referring expressions is that the speaker knows who or what the referent is and has a conceptual representation for it. They know whether the reference is specific and identifiable or not, and if there is doubt, the speaker has the lexicogrammatical resources to manage this, as does the addressee. Regardless of the field within which work is being done on referring expressions, one recurring theme is consistent; speakers producing these expressions either know or anticipate how able the addressee is in identifying the discourse referent specified. Since an essential part of referring consists in invoking in the hearer’s mind a representation of an object, the recognition by the hearer of such an intention automatically produces such a representation: as soon as the hearer recognizes that a noun phrase is being used as a referring expression he [they know] that the speaker is referring to a specific object [they have] in mind. (Searle, 1990: xvi)

When referring is successful, this anticipation on the part of the speaker is more or less on the mark; the speaker and addressee are, to some extent, of the same mind when it comes to matching the speaker’s intention to the discourse world. Schiffrin (2006: 36) offers the following explanation: ‘The referring expression used by the speaker would then (again, ideally) allow a hearer to recognize the speaker’s intention: to identify a referent sufficiently similar to what the speaker intends so that each can then say (and understand) something about that referent’. This meeting of minds, somewhat like synchronising 127

128

Referring in Non-collaborative Contexts

watches to be at the same moment of time, is essential, and where it is absent, discourse breaks down. Metzing and Brennan (2003: 201) summarise this position very nicely when they say that ‘successful referring requires that the speaker and addressee be able to take, at least for the moment, the same perspective on a referent’. While this may sound straightforward, and given that referring is something we do with ease thousands of times throughout the day, when we stop to think about it, there is quite a bit of complex information that the speaker must access in order to be successful in referring. Once a speaker forms the intention to mention a referent, the next stage involves getting the expression itself right. ‘When the speaker’s/writer’s decision to mention a referent is in place, another discourse phenomenon becomes relevant: referential choice that is the process of choosing an appropriate linguistic expression for the referent in question’ (Kibrik et al., 2016: 1; see also Kibrik, 2011: 69). Referential choice is influenced by various factors, which are referred to as activation factors by Kibrik (2011: 69). These activation factors involve ‘either properties of a discourse context, or of a referent itself’ (ibid.). We will be discussing the influence of both types of properties later in the chapter when we consider referring in more adversarial contexts. Let us return to the idea that for successful referring the speaker has to adopt a perspective that they believe to match that of the addressee. As we saw in earlier chapters, part of this perspective involves the accessibility of the referent and its information value (e.g. whether it is given or recoverable, or new or re-introduced). However, a key feature of successful referring involves implicit or explicit collaboration. Perhaps the most significant finding from psycholinguistics is the understanding that the process of referring is collaborative (Clark and Wilkes-Gibbs, 1986), that is, speech partners work together to accept an expression. Whether explicitly acknowledged or not, any initial referring expression is seen as provisional, only deemed to have been satisfied when the addressee has accepted, or ratified, the expression (Brennan, 2000: 4). In other words, expressions are assumed to be accepted by the addressee unless the expression is rejected or questioned. There is evidence that once a given expression is ratified, speakers tend to converge on the use of a particular expression. In other words, if one speaker uses one way of referring to something, unless there are reasons to change it, the other speaker will also use the same expression. Brennan and Clark (1996) have shown that, in their experimental work, referring expressions vary considerably more across conversations than they do within a given conversation. This has led researchers to suggest that there is a kind of contract that is developed between two speakers in conversation, whereby there is a kind of established agreement as to how a referent will be referred to. This type of agreement is commonly referred to as lexical entrainment (Metzing and Brennan, 2003: 202; see also Brennan and

7.1 Introduction

129

Clark, 1996). There is a sense in which maintaining the same lexical expression as the one established between speakers is not only advantageous to both speaker and addressee, but also that not doing so can cause problems. If precedent use is speaker specific, this benefit should be greater when the speaker who repeats the expression is the same one who established it in the first place, a prediction we refer to as the same speaker advantage for maintained precedents. Speakers are said to break a precedent when they refer to something using an expression that differs from how they referred to it before; for example, referring to something you called a sportscar as the Ferrari. Breaking an established precedent violates listeners’ expectations, and thus should elicit confusion and delay identification of the intended referent. To the extent precedent use is speaker-specific, breaking a precedent should be most confusing when the person who breaks it is the same one who established it in the first place. (Kronmüller and Barr, 2015: 2–3)

Despite the strong evidence for same speaker maintenance of referent type classification, or for ‘referential precedents’ (Kronmüller and Barr, 2015), there is some doubt about why this might be the case. The key question they raise is whether this is due to episodic priming or common ground. While most work in this area relates to memory and processing and is dealt with using psycholinguistic methods, very little is known about the more social and discourse implications of establishing a referential precedent (Kronmüller and Barr, 2015). One advantage to establishing a precedent is that such expressions ‘not only reduce linguistic variability, making speech more predictable, but also provide a powerful basis for reducing ambiguity in reference resolution. When speakers repeat a description they used earlier in the discourse, they are likely to be referring to the same thing they referred to on the last occasion they used that description’ (ibid.: 2). Later in this chapter, we will examine some evidence which suggests that, in some contexts, shifting the type specification of the referent may be advantageous to the speaker or to the addressee, although not both. Establishing a referential precedent involves some degree of lexical entrainment, which is often determined by the selection of a noun which classifies the referent as a type. As we noted at the beginning of this chapter, ‘one person’s valuable antique may be another’s yard sale fodder’ (Metzing and Brennan, 2003: 201). The choice of antique or junk in an expression to refer to an object will offer a different classification irrespective of the accessibility status of the object in the mind of the speaker. The nouns that represent the class or type of the referent were captured in Chapter 3 by the type element of the noun phrase (i.e. how the referent is categorised by the speaker). There is a denotational relationship between the intended referent and its expression with ‘a lexical noun that appropriately evokes a referent’ (Schiffrin, 2006: 36). This use of a lexical noun allows the speaker to classify the entity being referred to, representing ‘a type of thing to which the particular referent belongs’ (Radden and

130

Referring in Non-collaborative Contexts

Dirven, 2007: 88; see also Langacker, 2004: 78). It is the expression in its entirety that is used by the speaker to refer to a particular object, as in John, he, and the man from upstairs. An important point must be made here: the referent of any noun phrase is in the mind of the speaker rather than in the external world. As Rijkhoff (2002: 27) explains, ‘referents of NPs are rather mental representations of entities as they are created, stored, and retrieved in the minds of the speech participants’. In relation to the noun and its use in the lexicalisation of the expression, Brennan and Clark (1996) observed that, when speakers are referring, they are not describing a class of object. Instead, they are making choices, both lexical and syntactic, based on the current purposes of the referring expression; they are ‘conceptualizing an object’ (ibid.: 1482). They point out that, in terms of lexical choices, ‘there is still a great deal of variability [. . .] the problem is that most objects can be conceptualized in indefinitely many ways’ (ibid.). Referential choice is not simply about ‘choosing an appropriate linguistic expression’ (Kibrik et al., 2016: 1), it is also about choosing a conceptualisation of the referent. It is this conceptualisation that gets entrenched in the mind of the addressee. The shared perspective required for successful reference involves many factors, as we have been discussing, but we have so far in this discussion only considered the perspective of the speaker, who, upon the intention to refer, must both choose an appropriate expression and choose a conceptualisation that will allow for ratification of the expression and for discourse to continue. As has been mentioned above, successful reference is also assumed to be collaborative, and this is where the role of the addressee becomes important. ‘Once a referring expression has been presented by a speaker, it may be accepted and taken up by the addressee, or it may be adjusted, depending on whether the addressee understands and accepts the perspective it expresses’ (Metzing and Brennan, 2003: 201). The addressee’s role is to signal to the speaker the extent to which they find the expression acceptable. If it is not acceptable, for any number of reasons, then the collocutors will need to fix the problem. Problems with referring expressions have been very well studied in the literature, but one area, that of issues related to type lexicalisation, is relatively understudied, and these issues are precisely what we want to explore in this chapter. In what follows, we first provide a summary of the key problems encountered in referring, including various types of issues and how they get resolved. Following this, in Section 7.3, we examine various factors in relation to producing and ratifying a referring expression when the context is discursively non-collaborative, despite shared common ground. Most of our understanding of referring expressions and problematic reference (e.g. rejections and repairs) comes from experimental approaches or narrative discourse; there is very little

7.2 Problematic Reference

131

work, if any, that considers how referring works when the speaker and addressee do not share goals and have no invested interest in a shared conceptualisation of the referent. We look at various examples from primarily legal contexts to extend our understanding of referring in these atypical contexts. 7.2

Problematic Reference

In this section, we focus on problems in referring, or what Schiffrin (2006) calls problematic referrals. Most of the time, referring expressions are acceptable to the addressee. If we weren’t good at this, even the simplest conversation would become quite laborious and frustrating. Here, we will be taking a closer look at what happens when a referring expression needs to be fixed. Certainly, the addressee can be the one to signal a problem, but speakers often catch the issue even before the addressee has a chance to, and indeed the addressee might not even realise immediately that there is a divergence in the intended perspective. One of the difficulties in studying problematic referrals is that we (as speakers) are not always aware that we have chosen the ‘wrong’ word; nor do we always discover that our hearers’ knowledge is not sufficient for the successful resolution of a pronoun. The problems for us as listeners are different: we may not realize that we have interpreted the ‘wrong’ referent until it seems too late to backtrack and too awkward to begin the appropriate adjustments. Luckily problematic referrals are often self-initiated relatively quickly through false starts, cut-offs, unfilled pauses, uh and um, restarts, and/or repetitions that are followed (although sometimes after a delay) by another referral. (Schiffrin, 2006: 39)

As noted by Schiffrin in the quote above, the problem can be identified by the speaker or the addressee, and similarly the repair can be made by the speaker or addressee. When it comes to pronominal reference, the referring expression can be problematic for various reasons, but most commonly when ‘speakers overestimate common ground’ (Svennevig, 2010: 179). In these cases, there is usually more than one possible referent, and misjudging the potential for competing referents or the potential for underspecification of the referent can ‘cause referential conflict’ (Kibrik, 2011: 63). In example (112), taken from Schiffrin (2006: 56), we can see that the use of the pronoun they does not always refer to the same referent. In fact, we infer from this exchange that there are four distinct discourse referents. The first use of they is unidentifiable from the excerpt, but we know this expression is being used to refer to a set of people who have reported this information, and this same discourse referent is referred to again in the final line. The second use of they is co-referential with this bank and clearly distinct from the initial they in that same line. Based on the argument structure of the clause But they caught

132

Referring in Non-collaborative Contexts

them the minute they walked out the door, the subject they is clearly the people who caught the robbers. These people could be bank security or police officers. It is also clear that the second use of they (the ones who walked out the door) is being used to refer to the robbers, who had walked out the door and were then caught. In this example, the same referring expression is used to refer to different discourse referents. Based on the potential for referential conflict, we might have expected the speaker to use a lexical expression, for example, one with a noun such as the robbers or the guys in they caught the robbers as soon as they walked in. However, as Kibrik (2011: 63) explains, ‘referential conflict as such occurs if nothing in the linguistic structure discriminates between the activated referents’. Given that the clausal structure does discriminate among the activated referents, there was no referential problem for either speaker or addressee to resolve. (112)

c: a: c: a: c:

They1 said this bank2 got robbed yesterday. I don’t know whether they2 did or not. What? Really? Yeh. Wow. But they3 caught them4 the minute they4 walked out the door.

That’s what they1 claim.

As mentioned above, when a referring expression is problematic for any reason, it can be either the speaker or the addressee who signals the problem. It can be useful to differentiate between self-initiated or other-initiated repairs based on whether the speaker signals the need for a repair or not (e.g. Geluykens, 1994: 18; Schiffrin, 2006: 25). In addition, the repair itself can be offered by the speaker (self-repair) or by the addressee (other-repair). There were no repairs in example (112), but in example (113), taken from Geluykens (1994: 54), the pronominal reference it (in bold and italics) was used, and then the speaker initiates the repair themself and, after a brief pause, decides to add a full lexical expression to ensure the addressee will be able to be certain about what it is that works very well. The problematic reference is thus resolved by a self-initiated self-repair. (113)

A: I think it looks as though it works very well . . . their method of reproduction

However, the speaker is not always aware that there is a problem, and in example (114), also taken from Geluykens (1994: 136), the referential conflict is not picked up by the speaker but by the addressee. The use of he (in bold and italics) on the sixth line of the excerpt is not informative enough for the addressee, who signals the problem and proposes a repair using a proper name, Ken, which is then confirmed. As Geluykens (ibid.) explains, ‘the fact that the

7.2 Problematic Reference

133

intended antecedent, Ken, is the last mentioned preceding referent does not appear to be a sufficient condition for establishing unambiguous coreference’. (114)

a: b: a: b: a: b: a: b:

(. . .) we went to dine with Ken and Carlotta oh I believe James was entertaining Michael yes that’s right. he’s entertaining us on Friday I think hah - everybody individually don’t tell Ken this but rumour has it that he’s about to go Ken is yeah

Although pronouns are much more frequently involved in repairs than lexical expressions (Schiffrin, 2006: 67), problems with the use of a given noun in an expression are very interesting. When it comes to lexical expressions, that is, the use of a lexical NP, we find similar sorts of problems with similar kinds of repairs. One important distinction, however, is the role played by the typespecifying noun. The selection of the nominal which classifies the referent is a key aspect of referential choice. Lexical reference can also be ambiguous. For example, the type-specifying noun can be problematic. According to Schiffrin (2006: 112), ‘speakers sometimes choose a referring expression that assumes more recipient familiarity than is warranted, i.e. a form that is too “minimal”’. In these cases, the speaker adapts the referring expression (referral in Schiffrin’s terms) based on the speaker’s updated understanding of the addressee’s knowledge. Awareness of this knowledge is not part of a linear process. It is not the case that speakers work out the addressee’s level of knowledge prior to producing the referring expression; this awareness can happen at any point. As we saw above, most of the time the speaker’s understanding of the addressee’s knowledge is good enough for successful referring, but we have also seen examples, such as in (113), where this awareness comes into play after the expression has been uttered. An example of this occurring with a lexical expression is given in (115), taken from Geluykens (1994: 178). In this case, the noun year is deemed not specific enough by the speaker who, after producing the referring expression, anticipates some lexical ambiguity given the different definitions of what a year is. The speaker realises the potential ambiguity (self-initiated) and selfrepairs the expression. When an NP is repaired with another NP, the selfinitiated self-repair is the dominant repair type with a frequency of 73 per cent of all lexical (nominal) repairs (ibid.: 161). (115)

and then next year, next real year not academic year I shall go back to East Africa for another year

There is some evidence that addressees do cope well with ambiguous expressions since, as explained by Arnold et al. (2000: B25), ‘referent accessibility is

134

Referring in Non-collaborative Contexts

influenced by the comprehender’s estimation of the probability that a given referent will be mentioned again in the discourse, and that this estimation is built probabilistically on the basis of multiple cues, such as order-of-mention and recency’. In other words, if there is no indication that identifying the referent specifically is essential at the point of utterance, addressees will generally allow the discourse to continue under the assumption that they will be able to resolve the ambiguity at a later stage of the discourse. However, addressees will only tolerate the lack of specificity for so long, and if there is no resolution within a reasonable amount of time, the addressee will signal a problem and negotiate a repair. Although the repair in example (116) is immediate, it does show how attentive addressees are to ratifying expressions when a speaker indicates that the referent is identifiable (e.g. the training college). In this example, taken from Geluykens (1994: 164), we find that the addressee initiates the repair and the speaker provides a more specific expression (i.e. other-initiated, speaker-repair), in this case a proper name. The speaker who mentions the training college uses an expression that signals to the addressee that they should know which training college they are referring to. Perhaps this college has been mentioned in the past and the speaker assumes the addressee will recall this information, or it may be the case that the speaker has assumed more knowledge than the addressee has. (116)

a: c: a: c:

she’s been in this steady good job at . . . lecturing at the training college now for oh it must be ten or twelve years what training college? Rosary Mill mhm ah yeah

In some cases, where the lexical expression, often a simple NP, is insufficient, ad hoc descriptions are required, that is, additional descriptions that were not part of the original planning for the referring expression. The additional information added in the repair is often related to characteristics of the referent, including its location. According to Svennevig (2010: 187), ‘in a situation where there is a potential reference problem, speakers may do other things than adding information about the referent. They may check whether or not the interlocutor knows the referent or the referring expression’. This kind of checking can be done whether the repair is speaker-initiated or other-initiated. In example (117), from Geluykens (1994: 165), the simple NP, the bookcase, underspecifies the referent for the addressee, which is clear from the response given in the second line, bookcase?, and this signals the referential problem (other-initiated). The addressee clearly does not know which bookcase is being referred to, so the speaker offers a location (in our room), a material description (the big glass fronted bookcase), and since this was still insufficient information, the speaker finally adds the new one, which does allow the

7.2 Problematic Reference

135

addressee to ratify the referent (speaker-repair). It is clear that the speaker and addressee are working together to converge on the discourse referent. Here, it is not clear whether the problem is due to the use of bookcase or to the lack of information in the NP. For example, the addressee might not recognise the piece of furniture as a bookcase; perhaps there are no books on it. Alternatively, perhaps there are other pieces of furniture in the house that could be classified as a bookcase and the addressee is not sure which one is being referred to. This example shows clearly how the speaker and addressee work together, collaboratively, to reach a point where they share a sufficiently similar conceptualisation of the discourse referent. As noted by Geluykens (1994: 165), this back-and-forth process ‘could, in principle, go on for a very long time, until successful reference is established. In practice, however, since Speaker and hearer can be expected to be co-operative, the Speaker will try to give a sufficiently informational form as soon as possible’. (117)

c: b: c: b: c: b: c: b:

have you tried the bookcase bookcase? in our room — eh? you know the big glass fronted bookcase? yeah the new one oh that

Before moving on to our main section in this chapter, we will consider one final example which illustrates how closely aligned the mental representations between speaker and addressee become for successful reference. Example (118), taken from Geluykens (1994: 172), begins, much like what we saw in examples (116) and (117), with a simple NP. In this case, the speaker assumes that the addressee has the clock in question sufficiently in focus as to be able to ratify the reference. In this case, even the addressee seems to think they know which clock they are discussing. It is only when further detail is provided that the addressee begins to think they might be talking about different clocks. As soon as this becomes clear, the addressee signals the problem to the speaker (other-initiated repair) by asking which clock do you mean? What is particularly interesting here is that both the speaker and addressee offer the repair simultaneously. The overlap in speech is indicated by ‘{’ in the example. Geluykens (1994: 172) suggests that this type of repair merits a category of its own, labelling it an ‘other-initiated, self- and other-correcting repair’. This example may even suggest, as discussed above (cf. Arnold et al., 2000), that the addressee may have had some doubts initially but did not signal a repair was needed because they assumed it would become clear as the discourse continued, and that the ambiguity would be resolved. It is clear that when a repair is signalled, especially by the addressee, it detracts from the discourse,

136

Referring in Non-collaborative Contexts

somewhat like a detour. It is as if the ongoing discourse has to be suspended until the problematic reference is resolved. (118)

b:

a: b: a: b:

listen mother’s clock is very, is running quarter of an hour slow. I’ve slightly moved it to a faster position but how do I move the hands? Does that, does that face spring off? no no you look on the back . . . if you look the the three little brown . . . the screw things no no if you just turn, you . . . which clock do you mean? { the electric one? { the electric yeah

As we saw above, referential conflict, that is, ‘an irresolvably ambiguous referential expression’ (Kibrik, 2011: 64), forms the large majority of problematic reference. According to Kibrik (ibid.: 65), it is ‘quite real in language users’ everyday practice, as anyone can verify from [their] actual experience’. The fact that pronouns are more likely to be involved in a repair than a noun may be related to differences between endophoric (text internal) and exophoric (text external) reference, as suggested by Schiffrin (2006: 67), who states that ‘internal relations in the text cause more problems than external matching between word and world’. The examples discussed above involving ambiguous pronoun use illustrate this well. In some cases, the argument roles in which the reference occurs provide some disambiguating information. These endophoric types of reference may be easier to resolve since they are text internal. However, lexical expressions generally require collocutors to share the same conceptualisation of the referent, and the means for doing so relies on the lexicogrammatical resources of the NP, including, importantly, the type specification provided by the noun. In the remainder of this chapter, we will be focussing on a different type of conflict, almost the inverse of what happens with referential conflict. Instead of there being more than one activated referent which needs to be reduced to one, there is only one such activated referent, but the classification of the referent is in conflict for any number of reasons. As was mentioned above, there are ‘indefinitely many ways’ (Brennan and Clark, 1996: 1482) to conceptualise the referent. The problematic instances that we will explore in the next section involve conflict over how invested the speaker is in establishing a referential type and setting a referential precedent. 7.3

Referential Choice

In early experimental work on referring, Brennan and Clark (1996: 1487) showed that addressees will generally accept the lexical classification put forward by their speech partner even if they would normally use another one. Many of us have different terms for certain entities. These differences

7.3 Referential Choice

137

Figure 7.1 ‘A kitchen tool’ by Andrew Mason (2005) licensed under CC BY 2.0. www.flickr.com/photos/a_mason/7251819

could be due to family-specific naming practices, dialect, or differences in technical or domain specialism. For example, consider the item in Figure 7.1. If you ask ten different people what this item is called, you might not get ten different answers, but you might get three or four. We could say that it is a kitchen utensil. Although utensil is a hypernym, unlike hypernyms such as dog, utensil is unlikely to be used to specify this utensil during conversation. The reason for this is mostly likely because in any kitchen there are many utensils and utensil is not a commonly used word. If you did need to refer to this object, what would you call it? Would you say spatula, lifter, flipper, or fish slice? Even if you normally say spatula, when talking to someone else who classifies it with a different noun, for example, lifter, it would be unlikely to cause a referential problem. In collaborative or mutually supportive contexts, speech partners very readily engage in a conceptual pact in terms of lexical selection (see Brennan and Clark, 1996). In fact, they tend to try to reach that point as quickly as possible because not doing so disrupts the goals of the exchange. Of course, we can all imagine having a conversation about how many different words we have for this object, but then that becomes the topic of the conversation. The typical negotiation, then, of reaching a conceptual pact and setting a referential precedent, is something that speech partners try to complete quickly and collaboratively, but this is only done when there is a good reason to do so. Speech partners typically have no problem adopting a classification that they do not normally use. This type of variation might happen, for example, when different family members or friends use a different word for the spatula. At one family gathering you might use spatula and then at another you might use fish slice. The collaborative model of referring accounts nicely for referring and repairs in contexts which are collaborative in nature. In this section, we will explore referring when the context is more adversarial than collaborative. As Clark and Wilkes-Gibbs (1986: 37) showed in their work, ‘social factors govern the collaborative mode’. In Chapter 5, we considered some of these factors in terms of the context of situation. As an example of social factors, we will consider the statement made by Boris Johnson in response to questions about an event that took place in the garden of his official residence on 20 May

138

Referring in Non-collaborative Contexts

2020. As is now well known at the international level, in March 2020, the World Health Organization declared the novel coronavirus (COVID-19) a global pandemic. During the month of May 2020, the UK COVID-19 rules prohibited people from gathering, not even to visit lonely, elderly family members or to attend a funeral or indeed to be with a loved one who was dying. Thus, the discovery of an event at the official residence which involved over forty people caused outrage among the public and among many members of parliament. After this discovery was made public, Mr Johnson apologised to the House of Commons, and during an interview broadcast on 18 January 2022, he responded to a question about the 20 May 2020 event with the statement given in example (119). (119)

When I went out into that garden, I thought that I was attending a work event. (Boris Johnson, 18 January 2022)

What is interesting about this example in view of the discussions above about problematic referring is that he has selected a classifier + noun expression that he feels distances him from the accusations of hosting a party. The use of work as a classifier in the NP a work event construes a sub-category of event. He has been accused of attending a party, and many find his use of work event unacceptable. In this sense, the problem with the expression is signalled by a wide range of people, from the general public, to journalists, to government officials. There is not, strictly speaking, a repair initiated in the way we saw above. However, there is certainly considerable social pressure for Boris Johnson to avoid classifying the referent with the noun party. As we mentioned above, once the speaker has a referent in mind, the next stage of referring involves referential choice, which concerns how the referent is mentioned (Kibrik, 2011: 69). These choices relate to the expression of reference, including the type of referring expression and lexical choice. In addition to ensuring appropriate identification of the referent, the lexical expression, and in particular the noun, can in some cases bias the conceptualisation of the referent. In example (118), discussed above, the object in question was a clock. It is difficult to think of any other way to specify the type of referent; even the use of timekeeping device would be curious. It may be that the more concrete the object being referred to is, the less scope there is for varying the conceptualisation of the referent. The use of event in Boris Johnson’s statement (example (119)) is more abstract than concrete, even though an event is something you can observe. As was suggested in the discussion above, there are many ways to conceptualise a gathering of people, including event, party, gathering, or get-together. The choice of one noun over another is not neutral since there is a semantic effect that comes along with each due to the other words with which the noun typically collocates and its associated connotative meaning.

7.3 Referential Choice

7.3.1

139

Type Specification and Rephrasing

The semantic effect that a word picks up from frequent collocations is a type of attitudinal meaning referred to as semantic prosody. While the term itself is somewhat contentious (Hunston, 2007), for our purposes here we will follow Hunston’s (ibid.: 266) suggestion of ‘semantic preference’ as a term that can be used ‘to refer to the frequent co-occurrence of a lexical item with items expressing a particular evaluative meaning’. As Cotterill (2001: 311) has shown, consideration of this type of semantic preference (although she prefers the term semantic prosody) can be useful in revealing ‘the role played by lexical choice in constructing representations of reality’. What this suggests is that the lexicalisation of the referential type is not entirely objective; it is both cultural1 and personal in the sense that type identification is a kind of cultural norm, and it is a kind of personal stance in relation to the referent. Cotterill’s work (2001) provides a nice example of how type specification can shift, despite the speaker referring to the same discourse referent. She studied the transcripts from the O. J. Simpson double homicide trial, where Mr Simpson was accused of killing his wife, Nicole Brown Simpson, and her friend, Ronald Goldman. Her data focus on the opening statements from the prosecution and defence, which took place on 24 January 1995. As she explains (ibid.: 293), ‘[b]y carefully controlling and constraining the ways in which the actors and acts constituting the crime are formulated in court, through particular lexical choices, it is possible for lawyers to convey a whole set of meanings beyond the strictly referential, denotational ones’. We will consider just one example from her data, see (120), which she uses to demonstrate how the defence attorney uses shifting type specification in order to construe the intended discourse referent, in this case an event, as a verbal disagreement and not physical violence (ibid.: 306). (120)

Discussion > dispute > conversation They did have some discussion apparently maybe in ’86 or in ’84, some damage was done to a car and she was not in that car. She was not struck on that occasion, so I think you will find that incident of not great consequence. ... And there was a dispute because when Miss Brown Simpson would ever see Paula Barbieri’s picture, she was very upset about that, and Mr Simpson, on this date, I believe, saw a picture of somebody that she had dated during the time they had been separated and he said, gee, what is fair for me will be fair to you, and they go into the conversation that went over a period of

1

Fawcett (2000) refers to the type-specifying noun (thing in his terminology) as the cultural classification of the referent. See also Fontaine (2017) for a discussion of how this relates to the grammar of the noun phrase.

140

Referring in Non-collaborative Contexts time. At some point there was just a verbal conversation, it escalated, and tempers flare and that sometimes happens with married couples. That doesn’t make it right but sometimes it happens. (Cotterill, 2001: 306)

Based on what we have discussed so far, the highlighted nouns used in example (120) suggest shifting type specification from discussion to dispute and finally to conversation. This type of change in the expression is not a repair in the sense we saw above. It is rather a step-wise reconstrual of the conceptualisation of the referent which plants, in the addressees’ minds (i.e. the jury), a certain attitude towards the event. The change in the type specification is atypical (cf. Chapter 4) since, as we saw above, it is more collaborative to reach a conceptual pact quickly and to set a referential precedent. These strategies help make processing the reference much easier. It is important to note that here we are only looking at an excerpt; we do not have access to the full transcript, so we cannot read too much into the example. However, Cotterill (2001: 310) concludes that ‘[t]he lexical choices made by lawyers in their opening arguments are [. . .] potentially of great significance in orienting the jury in their consideration of the evidence’. The opening statements at a jury trial, such as the example we have just discussed, are monologic and planned. They are not open to other-initiated repairs or challenges. In the next section we look at more spontaneous and dialogic interaction in a legal setting in order to consider the nature of referring and, in particular, challenges to type specification. 7.3.2

Rejected Referring Expressions

The data2 under examination in this section come from transcripts from a public hearing for an unlawful reprisal application under the Ontario Ministry of Labour’s Occupational Health and Safety Act. This type of hearing occurs when a worker believes they have been penalised in an act of retaliation in response to something they have or have not done in accordance with the law and when, as a consequence, the worker has filed a complaint with the Ontario Labour Relations Board. The hearing is a legal proceeding which is led by a vice-chair, who is appointed by the Labour Relations Board. At the hearing, normally the applicant (the worker) and the employer are each represented by a lawyer of their choice. In the hearing under study here, the applicant is a former employee who has brought a claim of unlawful reprisal against his employer. In this case, the applicant is self-representing because of the cost of legal representation (which is a common occurrence); therefore, the only 2

An analysis and discussion of these data was published in part in Fontaine (2018).

7.3 Referential Choice

141

qualified lawyer involved is the employer’s legal representative. The representatives of the employer and the applicant or the applicant themself (as is the case here) call on witnesses during the hearing. Hearings at the Ontario Labour Relations Board are public, and the final decision is available in the public domain. In this case, the hearing was transcribed, but the names of the individuals participating and/or mentioned in the transcripts as well as the company have been made anonymous in order to prevent any individual person or company from being identified. We will be focussing our attention here on referring expressions which have been challenged, leading to some degree of negotiation between collocutors. In contrast to the examples we have been discussing earlier in the chapter, the repairs initiated in these data are not about referent identification, but rather type specification. Due to the complexity of the locutors involved, these data give us the opportunity to consider both collaborative and non-collaborative (or adversarial) contexts within the larger context of the hearing. As we will see in the discussion below, it matters whether the person being asked questions is on the same side as the person asking the questions. The number of referring expressions in any text will generally exceed 50 per cent of the text, and with a very long text, such as a hearing, the number of such expressions is very large. Restricting the data to a specific type of exchange, for example, the negotiation of an expression, means that there is a more manageable number to work with. The data analysis here is not meant to be exhaustive or offer replicable levels of statistical significance. However, it does provide an illustrative perspective of the nature of referring in contexts which can be said to be less cooperative than the ones provided in typical literature on referring. The particular excerpt of the transcript under examination here was selected because it includes examination in chief (direct examination or primary questioning of a witness) and cross examination of the same witness over one afternoon. This gives a more balanced sample to work with given that the witness was the same person being questioned, and both the lawyer (during examination in chief ) and the applicant (during cross examination) had the opportunity to ask this witness questions on the same general topic of concern for that portion of the hearing. The total word count for this excerpt was 134,692 words (52,347 words and 82,345 words in the examination in chief and cross examination, respectively). A total of twenty-eight expressions which were identified as problematic reference were selected for analysis. Of these, eleven were during examination in chief with the witness representing the company and seventeen were during cross examination by the applicant. The low frequency of problematic reference is unsurprising and to be expected. Clark and Wilkes-Gibbs (1986) use the term ‘reject’ rather than ‘repair’ in order to include a variety of types of rephrasing of the referring expression.

142

Referring in Non-collaborative Contexts

Figure 7.2 Rephrasing referring expressions (based on Clark and WilkesGibbs, 1986)

According to Clark and Wilkes-Gibbs (1986: 20), ‘an initial noun phrase that isn’t acceptable must be refashioned’. They identified three different ways a referring expression (RE) can be refashioned: repair, expansion, or replacement (see Figure 7.2). For our purposes, we will expand the notion of refashioning to any referring expression (not just ones that are not acceptable) that is rephrased, either for particular speaker motivations, as we saw in example (120), or because of a particular problem with the expression. In the data under examination here, it will become clear that the social context is complex. The excerpts from the hearing under study here are taken from a naturally occurring (not scripted) dialogue in an adversarial context. The fact that there is an applicant (a kind of plaintiff ) and a defendant means that the two parties do not share the same goals, although the lawyer and witnesses representing the company normally do. Ehrlich (2002: 738) explains that ‘whereas the adversarial nature of cross examination means that the purpose of a cross-examiner’s questions will be to challenge and undermine the truth of witnesses’ evidence, in examination in chief the purpose of questions, by contrast, will be to presuppose the truth and adequacy of witnesses’ evidence’. This point is significant because, by comparing instances of rejected referring expressions between examination in chief and cross examination, we can gain an understanding about whether these different roles have an impact on the nature of collaboration in referring. Although the number of rejected expressions identified in the excerpt is relatively small, the context of refashioning was not the same in examination in chief as compared to those in cross examination. The majority of expressions that were rejected (i.e. ones that then had to be repaired, expanded, or replaced) were ones initially expressed by the witness in examination in chief; in other words, the lawyer would interrupt his witness leading to a revised expression (or ‘refashioned’ in Clark and Wilkes-Gibbs’ (1986) terms). In cross examination, we find the inverse, where most rejected expressions were initially given by the applicant, who is in the role of a lawyer. Further, in examination in

7.3 Referential Choice

143

chief, all rejections were lawyer-initiated, whereas in cross examination, most rejections were witness-initiated. The most common reason for initiating a repair was due to lack of sufficient specification. Of the twenty-eight expressions examined, fifteen (53.6 per cent) involved problems with identification. For instance, in example (121), the lawyer is questioning his witness (W) during examination in chief (EIC), and when the witness says the order number five, the lawyer either does not know or would like it clarified for the record that the specific order being referred to is the compliance plan. Clark and Wilkes-Gibbs (1986) refer to expressions such as the compliance plan? as trial noun phrases since they are tentatively offered and then have to be ratified by the speaker (here confirmed by yes). We saw a similar instance in example (114) in resolving the ambiguous use of he when the addressee offered a trial noun phrase, Ken in Ken is. In both examples, the entire NP is replaced. (121)

Replaced Referring Expression eic. Okay. Were these . . . Have these orders been complied with? w. All except the order number five. eic. The compliance plan? w. Yes.

In addition to the typical problems in referring as discussed above, we also find some expressions are so problematic that there is no agreement, and after some attempts to repair, the intended referent and associated expression are abandoned. An example of this is given in Excerpt 3. This lengthy excerpt illustrates an abandoned referring expression, a type of rejection which is not accounted for in Clark and Wilkes-Gibbs (1986) and which we assume is quite infrequent. The excerpt is taken from cross examination (CE) with the witness (W) for the company and includes an interruption by the lawyer responsible for examination in chief. The expression in question, and the one rejected by the witness, is supervisors in lockout violations.3 The witness rejects the expression and signals the problem. Despite admitting that he knew of one incident of a supervisor in a lockout violation, the witness rejects the wording, and with the objection raised by the opposing lawyer (EIC), the intended referent is abandoned and the discussion moves on. This example illustrates the challenges that can occur between collocutors who do not share the same goals and the complexity involved when the stakes are high in an adversarial context. 3

A lockout is the term used to refer to a procedure where a machine is locked, with a physical lock, so that the machine cannot be connected to the power supply in order to ensure that the machine cannot function. All operational supplies to the machine are inhibited by the lock. The lock ensures a safe working environment if there is doubt about the safe operation of the machine, normally for repairs. A lockout violation occurs when someone removes such a lock.

144

Referring in Non-collaborative Contexts

Based on this evidence, we suggest a revised account of Clark and WilkesGibbs’ (1986) negotiation phase in their model of collaboratively building referring expressions, which includes abandoning the referring expression. All referring expressions have to be either accepted, refashioned, or abandoned, but clearly there is a cycle to the negotiation of the referring expression. As we saw above, in the vast majority of cases, a given referring expression is simply implicitly acknowledged. However, when an expression must be refashioned for whatever reason, the outcome of one or more attempts at refashioning will either be to accept or abandon the referring expression. Not doing so puts the discourse aims at risk. Excerpt 3 Abandoned Referring Expression ce.

w. ce. w. ce. w. ce. w. ce. w. ce. w. ce. eic. ce. eic. ce. eic. ce.

Okay. And can you understand today, you know, while these things were going on, and we got all these orders in June for supervisors, excuse me, for supervisors in lockout violations in June, that when July came . . . I don’t . . . Maybe I’m missing the point, but supervisors in lockout violations, I don’t know of any supervisor in a lockout violation. You don’t know about the guy that cut the lock off, and put three . . . Okay, that one. Yes. Supervisors in lockout . . . Okay, you don’t . . . . . . violations, . . . About . . . . . . I don’t. There’s the, the removal of the lock by Mr. S, yes. And the lock box. But it doesn’t say ‘supervisor’. It says we have to do a better job of education, but there was no supervisor in a lockout violation. It doesn’t say, ‘Provide supervision to ensure the . . .’ Mr. Jones. ‘. . . safety of the worker?’ Okay, I’m sorry. The order . . . I’m reading . . . . . . says what it says. . . . the same document. Okay. At any rate, you understand that the applicant had an obligation under the Act and that that could include contacting the Ministry of Labour to ensure the Health and Safety of the workers. Did you understand that?

We now turn our attention to rejected referring expressions where the noun used to specify the type of referent is challenged. In these cases, the problem with the expression is not that the referent cannot be identified, but that the way in which it is being represented is objectionable. For instance, in example (120), the shift in the use of dispute to conversation to specify the type in reference to the same discourse referent can be said to be a relexification of type specification. Such instances are not true repairs in the sense that this term

7.3 Referential Choice

145

is used in the literature on reference; the accessibility and specificity of the expression is not being challenged or rejected. Instead, it is the wording itself that is being negotiated. Since example (120) comes from an opening statement with a jury, the term ‘rejection’ or ‘repair’ would not be appropriate for capturing why these expressions are revised since the expression has not been rejected. Instead, the lawyer is skilfully shifting the conceptualisation of the referent for his intended addressee, in this case the jury. It is perhaps useful to use Clark and Wilkes-Gibbs’ (1986) terms and say that the speaker refashions the referring expression by replacing all or part of the NP. In the legal context, the type of referential precedent set during testimony may have potentially serious implications in relation to the referential choice of noun. In Excerpt 4, the purpose of the questioning relates to whether or not a dangerous incident had been formally recorded. The witness offers the expression a very long training session to explain where and how the incident had been recorded, but the applicant (CE) rejects the type specification. It is worth noting that both the witness and applicant know exactly what event they are referring to, so it is only the classification of the referent that is being questioned. The rejection of the expression is in fact emphasised by the applicant’s response, that is, a very long what?, making it clear that it is the use of the term training session that is not being accepted. It is also clear that in both their minds, a supervisors’ meeting is considered less significant than a training session when it concerns addressing serious safety violations. They both know that the supervisors’ meetings cover various topics and that these events are therefore a more general exchange of information, rather than a dedicated training session. This excerpt also provides evidence to suggest the witness and applicant knew of a referential precedent for this event, that is, a supervisors’ meeting. Excerpt 4 Training Session versus Meeting c. w. ce. w. ce. w.

But this supervisor, with 15 years of experience, puts no less than three lives at risk and, in your opinion, it’s not serious enough to record. That’s incorrect. The incident, itself, was recorded in great detail in a very long training session held with all of the supervisors. A very long what? A training session with all of the supervisors. Oh, I thought . . . I thought you said it was just reviewed at a supervisors’ meeting. Exactly. And that supervisors’ meeting generally covers a variety of topics. This meeting was – this was the focus.

The next example, presented in Excerpt 5, illustrates how challenging the process of negotiating a referring expression can be when the context itself is uncollaborative. In this excerpt, again, we see that the intended type specification put forward by the applicant is rejected by the addressee, the

146

Referring in Non-collaborative Contexts

witness in this case. This shift involves replacing the type of the referent; it is a semantic shift in type specification, which we refer to as the relexification of type specification. In Excerpt 5, the applicant (CE) and the witness (W) are discussing why there are protective guards on certain machinery and what the implications are of having damaged or malfunctioning guards. The vice-chair (VC) and the company lawyer (EIC) also take part in this exchange. The specific incident being discussed relates to an occasion where the applicant stopped a certain type of machine called a ‘patcher’ due to unsafe work conditions. What we find here is a tension between wanting to minimise the effort of collaboration (i.e. reducing the number of steps to reaching agreement) and the implications of the classification of the referent in question, which can make a big difference when considering the severity of the incident. The discourse referent in question is the entity which could be lost if the guards on the machine did not protect the worker. The initiating expression is a finger or a hand or something. As we will see, until the referring expression is ratified, the discourse is sidelined. Excerpt 5 Hand versus Limb versus Body Parts ce.

Now, I believe you testified that these guards were particularly serious because it’s a punching machine; you could lose a finger or a hand or something. w. I don’t know if I testified that, but I would agree with you that they’re serious. ce. Okay. Now, was that the first time you were made aware of problems with patcher guards? w. I don’t recall any other incident, no. ce. Okay. Do you recall an incident where the applicant did exactly the same thing just prior to this order being issued? w. No, I don’t. eic. I’m sorry. Did exactly the same what? w. Stop. Well, I ... Okay, I’m sorry. ce. Were you aware of an occasion in late May or early June, I would say lat e May – in May, where the applicant shut down all the patchers because the guards were improper? w. No, I do not recall that. ce. Okay. If I said to you ... If I put to you that I had done that, and when Mr. A. came back the guards were bent up and we had to make guards really heavy duty so that the workers couldn’t bend them up? What would you say to that? Would you remember that? w. Yes, I do. ce. Okay. vc.: Just hold on. w. I remember that the workers themselves would modify the guards. ce. Why? w. I don’t know why. ce. Did anybody ask? w. Maybe. I didn’t.

7.3 Referential Choice ce. w. ce. w. ce. w. ce. w. ce. w. ce.

ce.

147

Well, you agree it’s, it’s a potential for a loss of limb. I’m not sure I understand. You agree there’s a potential for loss of limb... Finger. ... or hand. Yeah, fair enough. Okay. Finger or hand but ... Okay. ... I don’t think you can get your arm in there. No. Okay. Okay, I’m sorry. Limb would be arm, I’m sorry. Okay. So we knew we could lose some body parts. And we knew that the workers were bending the guards up. We knew that was happening at the time of this inspection because of the stop order.

Despite the clear adversarial relationship between the speakers in this exchange, the example illustrates a type of collaboration in ratifying the referring expression. Both speech partners are working together, and the discourse does not progress until this is resolved. However, it is also clear that, in this case, there is some reluctance to ratify the expression. The applicant first mentions a finger or a hand, but the witness refuses to accept that without the safety guards on the machines, a worker could lose a finger or a hand. At this point, the applicant then begins a detour discussion to establish that the guards protect workers from injury, which includes the loss of a finger or a hand. This marks a kind of aside in the discourse, a temporary suspension of the general aims, in order to establish the nature of the referring expression. Oddly, the witness appears to offer, as a trial NP, the exact expression that he initially rejected, that is, finger or hand, but he rejects the idea that a worker could lose their entire arm. The applicant proposes body parts, which the witness implicitly accepts. Once there is agreement on the type specification of the referent, the discourse then gets back on course. While we cannot discuss each of the twenty-eight examples of a rejected referring expression from the data under study here, the three excerpts presented above serve a useful illustrative purpose in capturing the noted difference in referring refashioning found in this data sample when comparing collaborative and uncollaborative contexts. As we might expect, we find refashioned expressions in both examination in chief and in cross examination, where for this data set, the former is a collaborative context and the latter is not. However, one important difference is that the witness never rejected an expression that was produced by his lawyer during examination in chief, and this same witness rejected the most expressions during cross examination. A much larger data sample would need to be considered before we could

148

Referring in Non-collaborative Contexts

confirm these claims about the role of these two micro-contexts within the legal setting. Even in this relatively small sample, we find that, although the need for rejecting and refashioning a referring expression arises in both examination in chief and cross examination, the nature of the two different contexts suggests different strategies of referring are involved in mutually collaborative contexts (examination in chief ) as compared to more adversarial contexts (cross examination). The refashioning strategy that was most common in both settings concerns the identification of the referent. In these cases, the speaker did not make the expression specific enough for the addressee to identify the referent. However, there were some differences between examination in chief and cross examination. For example, a rejected referring expression which was refashioned by expanding the NP was exclusively found in examination in chief, whereas repairs which relexified the type specification of the referent, that is, replaced REs (Clark and Wilkes-Gibbs, 1986), were exclusive to cross examination. This finding suggests that during examination in chief, the collaborative context, the lawyer encouraged the witness to refashion certain expressions in order to make the expressions more informative (e.g. by adding modifiers) without changing the classification of the referent. However, in cross examination, both the applicant and the witness rejected expressions and required the expression to be replaced. In other words, the refashioning did not add more information but instead reclassified the referent (e.g. training session vs meeting from Excerpt 4 and limb vs body part from Excerpt 5). In this more adversarial (i.e. not mutually supportive) context, while the two speakers are collaborating to a certain extent on referring, it seems they are only doing so because not doing so would stop the flow of discourse. They are in fact picking on each other’s lexical choice because there is a lot at stake. 7.4

Concluding Remarks

All referring is collaborative because it requires ‘actions by both speakers and interlocutors’ (Clark and Wilkes-Gibbs, 1986: 2). The collaborative nature of successful referring includes non-interactive discourse, such as opening statements at trial or written texts, since speakers and writers anticipate the addressees’ knowledge and ability to identify a given referent (or not). The collaboration in these cases is virtual, often with no direct feedback on the expression. Collaboration is needed even when the social relationship is adversarial. The strategies for ratifying an expression can be different and, in some cases, as shown in the excerpts presented here, some of the strategies may even be antagonistic because of a resistance to converge on a conceptualisation of the referent in a given lexical representation. The refashioning strategy of replacing the NP, or part of it, seems a particularly

7.4 Concluding Remarks

149

relevant feature for uncollaborative, or adversarial, contexts. We can easily imagine a teenager trying to negotiate permission to attend a party preferring to classify it as a get-together, as the nouns get-together and house party do not construe the same type of event for most parents. Most evidence we have on the negotiation phase of consolidating a conceptual pact or setting a referential precedent has come from experimental work. What we have focussed on here is a non-guided, non-experimental, and adversarial setting. We have seen that speakers can be quite concerned with the specific detail of the wording of the expression. Especially in the crossexamination interaction, the choice of noun was negotiated, and there was some reluctance to agree on the lexical expression. The reason for this may be that to accept the expression would somehow imply responsibility or other consequences which would have been undesirable. What is clear, however, is that any disagreement or lack of consensus on the expression will stop the flow of discourse. An agreement on the classification of the referent is critically important. We have suggested that, even within the same legal proceeding, the immediate interaction constitutes a micro-context which is evidenced through the referring strategies. As explained by Berry (2013: 376), ‘in certain types of situation, notably face-to-face situations, the context can in fact be negotiated, making the contextual configuration a joint creation of the interactants’. Examining the negotiation of rejected referring expressions has revealed that the context of referring is different in examination in chief as compared to cross examination. It is not simply the tenor which accounts for the differentiation, but also field and mode. The participants are construed differently, and referential precedents are more challenging to establish. ‘When speakers refer to something, be it an everyday object or an exotic, unfamiliar one, they are faced with many choices’ (Metzing and Brennan, 2003: 201). It really is quite remarkable that with all the demands on speakers when referring, including memory constraints and lexical, grammatical, and contextual choices, we manage to get it right most of the time. It is, nevertheless, quite fascinating when we take a closer look and see what the strategies are and how we collaborate on referring despite quite adversarial contexts.

8

Indefinite Expressions for Referring Definitely

oh he was brilliant in that no one in the whole world could have played that part except David Bowie (Toole, 1996: 283)

8.1

Introduction1

In a sense this chapter follows on more from Chapter 6 than Chapter 7, in that it will focus on referring expressions in a mode which is best described as prepared writing. However, as in Chapter 7, we will be focussing on expressions which are atypical in use; in other words, expressions which deviate in some way from the typical conventions for the mode. The quote at the beginning of the chapter, taken from Toole’s (1996) study of the effect of genre on referential choice, is a curious one. The pronoun he is clearly used to refer to David Bowie, and in the second mention to this same referent, we would expect an anaphoric pronoun (i.e. him). The use of a full proper name instead seems to provide emphasis (ibid.: 283), as though we get more from this expression than we would from a pronoun. According to Toole (ibid.) ‘[t]his expressive sense of referential selection appears to override normal accessibility considerations’. In this chapter, we will be focussing on other types of referential selection which similarly appear to override normal accessibility considerations. We have already discussed the widely accepted set of conventions in relation to definiteness and indefiniteness, notably the use of an indefinite expression for something the addressee is not expected to be able to identify and a definite expression to refer to something identifiable (see Chapter 2). We have argued throughout this volume that atypical uses of referring expressions remain largely unexplored. This volume provides concentrated attention on this phenomenon, and once you start looking for them, it turns out that the

1

An earlier version of this work was published as Jones (2018).

150

8.1 Introduction

151

conventional rules of reference are often exploited for effect. To illustrate this point, we will take a look at the indefinite expression (in bold) in example (122), from an opinion article printed in The Guardian. For people living in the United Kingdom and possibly for fans of David Beckham, it is possible that even without knowing that the title of the piece is David Beckham: how this crock of a footballer can still woo the French, mention of non-novelty underwear might be specific enough to single out David Beckham. (122)

It seems that France, like Spain and the United States before it, is poised to be charmed by a man who, with his un-British attention to grooming, muscle tone and non-novelty underwear, may become an honorary Frenchman before his six months in Paris are up (01/02/13)2

In principle, the use of the indefinite article signifies that the writer does not expect the reader to be able to identify the man with these very specific traits. However, in reality, a reader with an interest in British football (and we can assume that someone reading an article about David Beckham does) would be able to identify the man referred to in (122) as David Beckham, especially if we consider that the expression is part of an ongoing narrative about Beckham, where reference has been made to him eighteen times previously. This process of identification perhaps works because, in the spirit of cooperation, the speaker intends to convey additional meaning by means of an implicature (Grice, 1975), which needs to be inferred by the addressee. Of course, the use of the indefinite article in such expressions could be to indicate that the speaker is talking about a type of entity with particular qualities. This use is common, as can be seen in the expression in bold in example (123), from an opinion piece titled Why is Theresa May still in No 10? Because the Tories need a human shield. (123)

Why is she still here? The Tories, after all, have a well-publicised talent for regicide: a leader who has failed is generally found standing in front of No 10 or Tory HQ within 24 hours, making a statement to the press. John Major, William Hague, David Cameron: all departed in this way. (The Guardian 20/06/17)

The expression in bold in example (123) has a similar structure to the one in example (122) and can be described as follows: A(n)+NOUN+RESTRICTIVE RELATIVE CLAUSE (A(n)+N+RRC forthwith). This expression is undoubtedly referring to a type of leader, made clear by the use of the adverb generally to introduce what typically happens in this scenario. The subsequent sentence then provides instances of the type (i.e. a list of leaders with the qualities

2

Texts reproduced under Open Licence terms and conditions. Copyright owned by The Guardian News & Media Ltd and Independent Digital News and Media Ltd.

152

Indefinite Expressions for Referring Definitely

presented in the relative clause). However, the A(n)+N+RRC expressions in examples (124) and (125), identified, respectively, as [i], [ii], and [ii] and in bold, appear to be signalling something different. (124)

Today the United States stands not as a source of inspiration to the rest of the world but as a source of fear. Instead of hailing its first female president, it seems poised to hand the awesome power of its highest office to [i] a man who revels in his own ignorance, racism and misogyny . . . [ii] A man with no control of his impulses will be unrestrained, the might of a superpower at the service of his ego and his id. (The Guardian 09/11/2016)

(125)

But you have to watch out for those Trumpites who pop up to call you “fake news” and who frighten radio station editors. The media’s continuing respect for “fair play” when discussing [iii] a president who is self-evidently a dangerous and racist xenophobe (as opposed, for example, to the Arab variety) should one day be examined. (The Independent 20/01/18)

Each of these three expressions is singling out a particular man (i.e. Donald Trump), one who is readily identifiable in the given context and who has been mentioned directly several times in the earlier sections of the texts; for example prior to the excerpt in (124), the text included reference to Donald Trump such as the new president elect, Trump, he, and similarly, with the excerpt given in (125), previous referring expressions included Donald Trump and the booby who thinks he’s running the United States. The chain of reference to this individual is carried on in the subsequent co-text. For example, following the excerpt in (124), we find his impulses, his ego, his id, and in (125), Trump and this infantile person, which all serve to reinforce how highly accessible the referent is. In each of these examples, it would be felicitous (semantically and syntactically, at least) to replace the indefinite expression with a proper name. At this point, we would like to draw on a few points discussed in earlier chapters. At a glance, these expressions seem similar to the ‘anaphores infidèles’ (Sarda and Carter-Thomas, 2021) that were discussed in Chapter 6. In a similar way, the use of lexical expressions for a referent that is in focus allows the writer (or speaker) to raise to the attention of the addressee certain properties, permanent or temporary, associated to the referent. One key difference with the expression examples in bold in (122), (124), and (125) is that, unlike those in Sarda and Carter-Thomas (2021), they are introduced by indefinite determiners. As we saw in earlier chapters, in particular Chapters 2 and 4, expressions introduced by an indefinite determiner are generally considered to be indefinite. However, the large-scale study by Kibrik et al. (2016) has shown, as we saw in Chapter 6, that referring expressions are not categorical; there is no one-to-one relationship between referential role and formal expression. What we do find are strong tendencies and correlations, but in

8.1 Introduction

153

these conventions we also find opportunities to exploit them (cf. Hanks (2013), as discussed in Chapter 4). While we might not be able to know the writer’s intentions in producing these expressions, we can get some insight into the accessibility status of the reader. In a study based on a corpus of forty journalistic opinion articles, Jones (2014) identified several key features of this particular use of the A(n)+N+RRC expression. The expressions selected for her study include those which occur in non-initial position in a reference or identity chain (cf. Martin, 1992; Halliday, 1994); in other words, there was a potential previous mention of the expression in question. The results of Jones’ (2014) reader interpretation experiment showed that readers largely do not interpret these expressions, when seen in context, as referring to a type or a newly introduced referent, as convention would suggest, but rather to the previously mentioned, fully identified entity. The results of this experiment also showed that the amount and level of detail of lexical information in the relative clause plays a role in how the expressions are interpreted. That is, the more the lexical detail in the relative clause, the more likely it is that the reader interprets the expression as referring to the definite referent. For example, readers were almost equally divided about whether the expression in (126) referred to the specific MP Nadine Dorries or a type of MP with those qualities, but almost all readers interpreted the expression in example (127) as referring to the referent, Louis Kahn. (126)

an MP who can spread such inaccuracies

(127)

a man who died in a public lavatory in a low-grade public building, whose corpse lay unrecognised in a New York City morgue for three days, and who flitted from one family home to another

Using Toole’s (1996) framework, Jones (2014) found that 93 per cent of the expressions in her corpus achieved a high or mid degree of accessibility. These results show that these expressions are comparable to demonstratives, pronouns, and zero anaphora in terms of how accessible the text producer assumes the referent is in the mind of the reader. What this means is that the A(n)+N +RRC expressions in Jones’ (2014) study are not produced as if the referents are new in the discourse, but rather as given and highly accessible. What we want to explore in this chapter is how expressions introduced with an indefinite determiner become functionally identifiable. To do so, we will discuss the use of referring expressions, such as those highlighted in bold above, with the exception of example (123), and consider how the writer creates an adequately definite shared space with the reader to allow for an indefinite expression to be understood as referring to the established entity. In what follows, we will briefly discuss the notions of reference and (in)definiteness to complement the view we presented on this in Chapter 2. In Section 8.3

154

Indefinite Expressions for Referring Definitely

we expand upon existing accounts of ‘late’ or ‘second-mention’ indefinites, as well as Hasan’s (1985) co-extensional framework. Building on these concepts, Section 8.4 further develops the notions of co-extension and similarity chains (ibid.) through an analysis of cohesive ties in four journalistic opinion articles. We show that sufficient co-textual and contextual scaffolding is put in place by the writer for the indefinite expression to be understood as being used to refer to a specific identifiable referent. Finally, in Section 8.5, some consideration will be given to the reasons for overriding normal accessibility considerations, when we would expect definiteness encoding to be maintained through the use of a definite expression. 8.2

Referring and (In)Definiteness

We know that, conventionally, only formally definite expressions are used to refer to entities that are considered to be identifiable to the addressee, at least in English (see Chapter 2). These expressions can be realised by an NP with a definite or demonstrative determiner (the/this apple), a pronoun (it), or a proper noun (Bardsey Island Apple). Indefinite expressions are assumed to be used for referents which the addressee is not expected to be able to identity and are typically realised by NPs with indefinite determiners and pronouns (e.g. an apple; some apples; any apples; any). It is generally accepted that the conventional pattern of article use is an indefinite first mention to introduce the referent into the discourse, with subsequent mentions being signalled by a definite marker, as illustrated in the italicised expressions in (128). However, it has also been acknowledged that there are exceptions to this pattern (e.g. Du Bois, 1980; Martin, 1992; Givón, 1993a), as we shall see below. (128)

A German policewoman was shot in the head when a man grabbed a police gun at a suburban station in Munich. The woman, 26, was critically wounded . . . (BBC 13/06/17)

Du Bois (1980: 219) suggests that definiteness involves ‘a tracing of the constant idea (referent) through links with the shifting words (references) used to refer to the idea’, which speakers have control over. What is interesting here, though, is that even if the identity of an object is known to both interlocutors, they are under no obligation to use the available definite expression. For example, if a parent says to their small child, someone left their bike out in the rain last night, both speaker and addressee (the culpable ‘someone’) are well aware of the fully identifiable culprit. However, for some pragmatic purpose, the speaker chooses not to explicitly identify them and uses an indefinite expression instead. This obfuscation is done, however, while remaining in the spirit of cooperation; that is, the speaker is well aware that the addressee is able to identify the referent.

8.2 Referring and (In)Definiteness

155

In Gricean terms, the speaker is flouting the maxim of quantity, which requires the speaker to make their contribution as informative as required; in this case, the speaker is not being as informative as required (which would involve explicit identification of the wrongdoer), but the addressee, assuming speaker cooperation, is able to work out the implied meaning. Even though the conditions for the use of a definite expression are in place, the speaker can choose whether or not to oblige. Du Bois (1980: 219) concludes that ‘speakers have facultative control of definiteness’. This conclusion is echoed by Fries (2001: 83, emphasis in original), who states: ‘Of course, speakers always have the choice of how to present new information so while information which is presented as structurally New is usually new in fact to the listener, individual speakers may choose to present as New, information which is obvious to the listener’. The notion of choice is crucial here; ultimately is it under the speaker’s control how a referent is presented. Speakers make a choice about how they wish to depict the referent, whether consciously or not; they have to ensure that it fits with the assumed discourse representation held by the addressee. These claims suggest there are other parameters beyond identifiability which govern the choice of an (in)definite expression. The ‘wordings’ of referring expressions are not always necessarily intended to assist in the identification of the referent, but rather to fulfil a discourse function. Several scholars have examined this idea. As was discussed in Chapter 4, Vonk et al. (1992: 303) suggested that some expressions serve a discourse structuring function (e.g. the repeated use of pronominal reference in the Sally Jones example). By diverging from conventional associations between the form of the expression and what it signals in terms of accessibility, there is generally some additional meaning that is construed. Fiction, as we might expect, seems to offer some good examples of creative referential choice. In the excerpt in example (129), from 10 Days in a Mad House by Nelly Bly (1887), we find a scenario that is the inverse, to some extent, to the one described in Chapter 4 from All the Pretty Horses (Macarthy, 1992), where we discussed the use of repeated pronominal reference despite the referent not being in focus. In example (129), the writer introduces two women in a canonical way, that is, with an indefinite expression. References to the women continue using definite descriptions (anaphorically), as is conventionally the case. However, when the time comes to learn the name of the second woman, the writer chooses to use an indefinite article with a proper name, a Mrs Caine. (129)

I watched two women, who seemed of all the crowd to be the most sociable, and I selected them as the ones to work out my salvation, or, more properly speaking, my condemnation and conviction. Excusing myself and saying that I felt lonely, I asked if I might join their company. They graciously consented [. . .]. One said her name was Mrs. King and that she was a Southern woman.

156

Indefinite Expressions for Referring Definitely Then she said that I had a Southern accent. She asked me bluntly if I did not really come from the South. I said “Yes.” The other woman got to talking about the Boston boats and asked me if I knew at what time they left. For a moment I forgot my role of assumed insanity, and told her the correct hour of departure. [. . .] She said in reply that she had been unfortunate and had come to New York, where she had worked at correcting proofs on a medical dictionary for some time, but that her health had given way under the task, and that she was now going to Boston again. [. . .] Here I must introduce a new personage by name into my narrative. It is the woman who had been a proofreader, and was about to return to Boston. She was a Mrs. Caine, who was as courageous as she was good-hearted. She came into my room, and sat and talked with me a long time, taking down my hair with gentle ways.

Proper nouns presuppose identifiability without needing to be introduced by the and can be considered to represent ‘definite concepts’ (Chafe, 1972: 57). It is interesting that here there is some kind of anti-definiteness strategy at play. Du Bois suggests that the use of the indefinite article with a proper name marks nonidentifiability, that the particular individual is known to the speaker, but that the name ‘may mean nothing to the addressee’ (1980: 218). However, in this extract, as readers, we have already been introduced to Mrs Caine as one member of a pair by means of a plural indefinite expression, and continued references to her were made with definite descriptions. The reader can presumably identify the individual at the point of the occurrence of the indefinite expression. Contrasted against a series of typical reference in written text, this instance makes us wonder why the indefinite article is used, as it must surely do more than indicate the status of type identifiable. In fact, we can imagine that the author might have also used a certain Mrs Caine or somebody named Mrs Caine. However, the author has the narrator formally state her intention to introduce this new character, despite having already done so: here I must introduce a new personage by name into my narrative. She feels that it is at this point that she is making the formal introduction. Even though, in reality, the reader is quite capable of connecting the name Mrs Caine to the many previous definite mentions, and thus can identify her. And yet, the writer still feels that until the name has been formally introduced, an element of indefiniteness remains. This element, though, is perhaps to do with providing the reader with new information about the referent rather than helping with the identification process, hence the strategy to remove the definiteness from the expression. 8.2.1

Late Indefinites

We saw above that indefinite expressions are typically used for nonidentifiable referents and definite expressions for something that the addressee

8.2 Referring and (In)Definiteness

157

is expected to be able to identify, but also that there are exceptions to this tendency. We have also seen examples of a particular exception to this convention: the use of A(n)+N+RRC to refer an entity previously established in the discourse. The purpose of this sub-section is to review extant accounts of these ‘late’ indefinite expressions and show that while they provide a useful framework from which to guide an analysis of the targeted A(n)+N+RRC expressions in this chapter, they cannot fully account for their use and so we offer alternative explanations. There have been a few attempts to account for the use of indefinite expressions which refer to a previously mentioned entity (Du Bois, 1980, 1997; Ushie, 1986; Epstein, 1994; Schouten and Vonk, 1995; Jones, 2014). In the case of these ‘late’ or ‘second-mention’ indefinites, the speaker chooses a structurally indefinite expression for a non-initial mention, the referent of which is potentially uniquely identifiable to the addressee. The speaker does this in defiance of the typical conventions governing the use of the indefinite article, thereby creating an implicature which conveys something different from what has, strictly speaking, been expressed. Schouten and Vonk (1995: 4) view the use of indefinite expressions for known referents as ‘marked’, although as we argued in Chapter 4, we are using the term atypical. This refers to expressions which function in ways that are inconsistent with the conventional use of indefinite expression (e.g. to introduce a new referent into the discourse). The forms of referring expressions are part of a more general pragmatic principle based on speaker–addressee knowledge, and in the case of atypical referring expressions, the speaker violates the principle in order to convey additional information, which needs to be inferred by the addressee (i.e. flouting an implicature (Grice, 1975), which conveys that the speaker expects the hearer to identify their true intentions (i.e. resolve the implicature)). The speaker relies on the addressee to accept that the expression is intended to be atypical, and thus infers an interpretation that is meaningful in the specific discourse context. Further, for successful interpretation, atypical uses of indefinite expressions necessitate the establishment of a relationship between an existing referent and the referent of the indefinite expression. It is the combination of the indefinite form and the current mental discourse representation which influences the addressee’s interpretation of the form of the expression (e.g. any items that signal definiteness or indefiniteness). Schouten and Vonk (1995: 6) also argue that the interpretation of atypical referential expressions is derived from the conventional function of the socalled marked form when observed in a typical way (cf. Hanks’ (2013) concept of exploitation). That is, the choice of an indefinite or definite form is based on whether the entity is or can be inferred to be a unique member of a set (i.e. a set in the current referential domain). The set may be present in (or inferable from) the surrounding discourse, situational or world knowledge (Schouten and

158

Indefinite Expressions for Referring Definitely

Vonk, 1995: 12). So, the choice of form of a referring expression is based on the value for unique identifiability and the accessibility of the referent in the existing discourse representation. Once a non-uniquely identifiable referent has been introduced into the discourse, it then becomes uniquely identifiable and accessible (ibid.: 13). If either of these factors is transgressed (or flouted, in Gricean terms), an expression is being used in an atypical way. At first glance, this general explanation works for the expressions we are discussing in this chapter. But when we delve deeper into the types of expressions that these scholars discuss, we will see that there remains something exceptional about the A(n)+N+RRC expressions used in their specific context, which is unaccounted for in these previous descriptions. Ushie (1986) characterises late indefinites as corepresentational. That is, indefinite expressions which identify known referents have underlying representations which are identical to an already established referent. An indefinite expression which refers to an already established referent relates to ‘a certain degree of “newness”’ (ibid.: 440) that results from the particular way in which the text producer organises and presents the content of a text. This presentation can be triggered by the speaker’s (re)interpretation of an event or ‘a shift/ discontinuity in the point of view’ (ibid). In terms of the re-interpretation of an event, this could be to emphasise a new feature of an established referent because of its relevance to the particular situation, representing the entity in a new light. Ushie (ibid.: 430) cites the text in (130) to illustrate this reinterpretation of a referent, where the two expressions in bold re-interpret Jan Morris and Rolling Stone magazine, respectively. Here, the new features highlighted show the referents in contrast to one another – traditionalist as opposed to lively and innovative, an effect that would not have been achieved had pronouns been used (which would be the expected form given the salience of the referents). (130)

In 1974, the American magazine Rolling Stone invited Jan Morris to write a series of travel articles. The fruits of that collaboration between a romantic traditionalist Welsh author and a lively and innovative American paper appears in Destinations

While this account may capture the use of some A(n)+N+RRC expressions, we will see in the co-extensional analyses in Section 8.3 that newness cannot account for the use of these particular expressions because the information in the RRCs is either ‘discourse-old’ or ‘hearer-old’ (Prince, 1992), that is, the information is either given in the discourse or in the knowledge of the addressee. There is also another type of indefinite expression which is used when ‘the point of view is shifted’ (Du Bois, 1980: 259) from the speaker to someone else involved in the text. In such ‘reintroductory noun phrases’ (Schouten and Vonk, 1995: 14), the indefinite expression reflects a new, subjective

8.2 Referring and (In)Definiteness

159

interpretation of the referent or a shift in perspective from which the referent is based. In example (131), an indefinite expression (in bold) is used to refer to an entity which has been previously introduced (underscored expressions) and is uniquely identifiable to the addressee. (131)

A 35-year-old citizen of Nijmegen, caught in the act during a break-in in a pizzeria on St. Annastraat, appears to have much more on his plate. The man has confessed to being an accomplice of the armed robbery [. . .]. The man was being sought by the police for several days [. . .]. Yesterday night the police suddenly received a call from a pizzeria owner, who had caught a burglar. It turned out to be the 35-year-old citizen of Nijmegen who was on the wanted list of the police. (from Schouten and Vonk, 1995: 15)

When readers get to the indefinite expression, a burglar, they might expect a definite expression to be used, based on the current context, and indeed a definite expression would work well here (. . . a pizzeria owner, who had caught the burglar/the man), even though the coherence may be slightly disrupted. However, the referent has been introduced ‘into the subordinated perspective of a character in the text’ (Schouten and Vonk, 1995: 15). From the point of view of the pizzeria owner, the identity of the burglar is unknown, and the use of an indefinite expression conveys this fact. Because of the clash between the expected expression (a definite expression) and the actual expression used (an indefinite expression), the expression is atypical. The class of thing denoted in the indefinite expression (burglar) is not uniquely identifiable to the pizzeria owner; thus, the intended referent is reintroduced, despite the fact that the reader would be able to identify the referent, thus creating a narrative effect. In reintroductory indefinites the unique identifiability of the referent is not measured against the knowledge structures of the addressee, but rather the knowledge structures of another participant in the text. However, the expressions we are discussing do not fit into this explanation; for example, even though Trump (in (124) and (125)) and Beckham (in (122)) have been reintroduced into the text from another’s perspective, it would still be difficult to argue that the particular participants in the text do not know the identity of the referent singled out in the expression. To illustrate this, let us look again at the expressions with their immediate co-text, which we have renumbered for ease of referral as (132), (133), and (134), respectively. (132)

Today the United States stands not as a source of inspiration to the rest of the world but as a source of fear. Instead of hailing its first female president, it seems poised to hand the awesome power of its highest office to a man who revels in his own ignorance, racism and misogyny . . .

(133)

But you have to watch out for those Trumpites who pop up to call you “fake news” and who frighten radio station editors. The media’s continuing respect

160

Indefinite Expressions for Referring Definitely for “fair play” when discussing a president who is self-evidently a dangerous and racist xenophobe (as opposed, for example, to the Arab variety) should one day be examined.

(134)

It seems that France, like Spain and the United States before it, is poised to be charmed by a man who, with his un-British attention to grooming, muscle tone and non-novelty underwear, may become an honorary Frenchman before his six months in Paris are up.

In (132), the people of the United States know the identity of the man they elected. They did, after all, do the electing. In (133), the writers of the media know who they are discussing as they are the ones doing the discussing. In (134), sports writers know who may become an honorary Frenchman because they know all about David Beckham and know how charming he is perceived to be. In addition, it is likely that the reader also relies on the discourse representation of the referent they have already built up to guide their interpretations as well as the general principle of relevance to infer that the indefinite expression refers to the identifiable entity. Once all these factors are combined, it is hardly surprising that the indefinite expression allows for identification of the referent. On the other hand, certain predicative indefinites express properties of an established referent which do not uniquely identify that particular referent. The properties are, however, attributed to it (Schouten and Vonk, 1995). Such use of non-identifying properties can be seen in opposition to definite expressions, which are used when the properties are uniquely tied to a referent, as illustrated in the expression in bold in example (135). (135)

Erik has been totally out of it since he of all people, on Monday, had to find that horrible couple: his old friend Robert an insane murderer and Magda, a woman who in my opinion never meant very much to him, in the most absurd state a human being can be in, dead. (in Schouten and Vonk, 1995: 24)

This type of expression appears to be ‘a curious mixture of the referential function and the predicative function for which indefinites can be used’ (Schouten and Vonk, 1995: 23), where the expressions refer to already identifiable referents (the referential function) but also attribute information which is not uniquely tied to the referent (the predicative function). However, if we return to the expression under discussion here, the properties in the relative clauses can be uniquely tied to the referents in the contexts in which they appear. The semantic content of the relative clause contains information which is either discourse-old or hearer-old (Prince, 1992), and so newness cannot be the sole explanation for the use of the indefinite article. We will see how this connection is achieved in the co-extensional analyses in Section 8.3.

8.2 Referring and (In)Definiteness

161

Some classifying indefinites, on the other hand, can be used to introduce the intended referent as ‘a member of a subclass of a basic category, with certain properties’ (Schouten and Vonk, 1995: 26), where new information is conveyed in the comment part of the topic–comment structure of the sentence (see also Epstein, 1994: 223). At first glance, this kind of atypical indefinite shares some features with the A(n)+N+RRC expression, as is evident in the example in bold in (136), which is from W. H. Auden’s foreword to Dag Hammarskjöld’s book Markings.3 (136)

In addition to the spiritual suffering of loneliness, of having to leave behind him ‘the world which had made him what he was’, Hammarskjöld had to endure [...] the plain physical suffering of constant nervous strain and overwork. If, as the reader goes through the entries between 1953 and 1957, he finds himself impatient [...] with their relentless earnestness and not infrequent repetitiousness, let him remember that most of them must have been written by a man at the extreme limits of mental and physical exhaustion.

The expression at the extreme limits of mental and physical exhaustion is both hearer-new and discourse-new (Prince, 1992). However, these classifying indefinites differ from the A(n)+N+RRC expressions we have been discussing because, as we will see in Section 8.3, the information in the relative clauses in the A(n)+N+RRC expressions is not new (i.e. the content can be traced back to the preceding text or ongoing discourse about the entity in question), and so the use of the indefinite article cannot be explained solely through the newness function of the indefinite article. Our discussion here has covered existing accounts of late indefinites and has shown that, while they provide a useful background from which to guide an analysis of our A(n)+N+RRC expressions, they cannot fully account for their use. The A(n)+N+RRC expressions do not contain new information, and the content in the RRCs can be tied uniquely to the referent. For a more detailed review of the literature discussing late indefinites, see Jones (2014). 8.2.2

Lexical Cohesion

It was noted above that the results of Jones’ (2014) reader interpretation experiments suggested that the content of the relative clauses in the A(n)+N +RRC expressions appears to lead the reader to either a referring or nonreferring interpretation. Therefore, a closer examination of what ties the relative clauses to the established referent would be useful to determine how the reader makes this interpretation. To do this, we extend Halliday and

3

This example was originally cited in Ushie (1986) and later discussed in Schouten and Vonk (1995: 26).

162

Indefinite Expressions for Referring Definitely

Hasan’s well-established approach to lexical cohesion (1976), which identifies semantic fields and the logical relations between words within these fields. This particular approach is employed because it provides a transparent way of linking together various elements of the text and discourse. After a critical overview of Halliday and Hasan’s approach to lexical cohesion, a detailed analysis of the semantic ties of four A(n)+N+RRC expressions is carried out. Hasan (1985) merges Halliday and Hasan’s (1976) two principle aspects of lexical cohesion, reiteration and collocation, into one broad meaning relation: co-extension. Co-extension, however, is best understood if contrasted with Halliday and Hasan’s two other meaning relations: co-reference and coclassification. Co-reference for Halliday and Hasan (1976: 31) is the relationship of situational identity that exists between members of a cohesive tie. In a cohesive tie, the two (or more) terms are anchored together through a meaning relation (Hasan, 1985: 73). Halliday and Hasan (1976: 31) maintain that co-referential items are not interpreted semantically in their own right but rather are ‘directives’ which signal that the information is to be recovered from elsewhere. Their example is reproduced here as (137). (137)

Three blind mice, three blind mice. See how they run! See how they run!

The expression they refers anaphorically to three blind mice, and as a directive, signposts that the meaning of they needs to be retrieved from three blind mice. Halliday and Matthiessen (2014: 624) broaden this definition and comment that the identity of the entity being referred to is recoverable from ‘the instantial system of meanings that is built up by the speaker and listener as the text unfolds’. Thus, the characteristics of the entity being referred to are not static, and the representation of the entity in the mind of the addressee evolves as the discourse develops. Co-classification, on the other hand, relates to the things, processes, or circumstances which belong to an identical class of items, but in which each tie refers to a distinct member of the class. A relatable example of coclassification is given in (138), uttered in parental exasperation by one of the authors. (138)

There’s Lego® everywhere! It’s in the kitchen, in the hallway, and it’s even in the bathroom!

The relationship between each member of the tie in (138), the series of expressions in italics, is not that of referential identity as each instantiation of lego (throughout the house) refers to a distinct member of the class of items lego. The pile of Lego in the kitchen is different to that in the hallway, and both are different to that in the bathroom. The meaning relations of

8.2 Referring and (In)Definiteness

163

co-reference and co-classification are different, but what ties them together is that in both cases the meaning of the second item (and any subsequent items) in the chain is implicit and needs to be retrieved from elsewhere. So, in the case of (138), the referent of the pronoun it in it’s in the kitchen, the nominal ellipsis in in the hallway, and the second pronoun it can only be retrieved from the initial mention of Lego. The various sprawlings of Lego in each case may include different block sizes and colours, that is, they are unlikely to be identical sets. Co-extension, the third meaning relation, and the one we are most concerned with here, is a semantic relation between members of a cohesive tie, where members refer to something within the same general field of meaning (Hasan, 1985: 74). For example, lunch, restaurant, and meal are all cohesively tied to the semantic field of food and if presented in a text would form a ‘similarity chain’ (as opposed to an identity chain). Co-extension differs from co-classification and co-reference in that the meaning of one member is not elicited only by reference to its relation to another, but rather the relationship arises from some contiguity of meaning. Co-extension is usually realised by lexical items or content words, which form a semantic grouping (a similarity chain). Within the context of a specific text, items can refer to related actions, events, and objects and their attributes. The semantic grouping they form is genre-specific (ibid.: 85). The field of discourse determines, to a large extent, what items may form the semantic groupings which make up the similarity chains. Following Halliday and Hasan (1976: 22), we are defining discourse as ‘the total event, in which the text is functioning, together with the purposive activity of the speaker or writer’. It is important to note here that grammatical cohesion and lexical cohesion are in a relationship which is interdependent and reciprocal, and in a typical text they work ‘hand-in-hand, the one supporting the other’ (Hasan, 1985: 83). Hasan (ibid.) provides a brief text to illustrate this point, which we have reproduced as example (139), where items of interest are highlighted in bold. (139)

once upon a time there was a little girl and she went out for a walk and she saw a lovely little teddybear and so she took it home and when she got it home she washed it

Hasan then analyses this simple example in terms of the threads of continuity or chains highlighted: girl, walk, teddybear, and home in Figure 8.1. The four separate chains (marked with different kinds of lines) link together items which are related to each other in some way, and demonstrate the simultaneity of cohesive chains. Girl and she, and teddybear and it are part of two separate

164

Indefinite Expressions for Referring Definitely

Figure 8.1 Cohesive links (Hasan, 1985)

identity chains, and the relation between the members of each chain is that of co-reference. The chain initiated by walk is an example of a ‘similarity chain’ (Hasan, 1985: 84), a chain composed of items that refer to ‘non-identical members of the same class of things, events etc., or to members of non-identical but related classes of things events etc.’ (ibid). Items in a similarity chain are related by co-classification or co-extension. So, with the example of went-walk-got, the items lie within the same general area of meaning; that is, ‘walking is a kind of going, and going is an important part of getting anywhere’ (ibid.: 85), and the relation is that of co-extension. Whereas in identity chains the relationship between items is identity of reference, the relationship between items in a similarity chain is that of similarity of reference (ibid.). As noted above, similarity chains are formed by items in a text which are in the same ‘general field of meaning’. However, to delimit the vagueness of ‘general field of meaning’, Hasan (1985: 80) proposes that cohesive ties between items are only established when they stand in the sense-relations of synonymy, antonymy, meronymy, and hyponymy. Whilst repetition could be argued to be absolute synonymy, Hasan (ibid.: 81) claims that the cohesive tie is not ‘strictly-speaking’ established on the basis of a sense-relation, but nevertheless contributes to the cohesion of a text as a ‘similar experiential meaning is encoded in each repeated occurrence of the lexical unit’. For the purposes of our discussion, repetition of a lexical unit is considered to be one of the ways of realising co-extension. According to Hasan (1985: 80), the sense-relation restriction prevents the formation of chains such as the following: flower, petal, stem, stalk, twig, branch, trunk, tree, wood, log, faggot, tinder, fire, and flame. In this chain, all members of the chain are semantically related to the immediately preceding item, but not necessarily to those further away. That is, we can see the meaning relation between flower and petal (i.e. that of hyponymy), but the connection between flower and flame is unclear. Thus, by restricting the ‘general field of meaning’ to items which have sense-relations, this problem is avoided. However, delimiting the notion of ‘general field of meaning’ by relying on sense-relations between ‘content words’ or ‘lexical items’ creates a problem

8.3 Co-extensional Analyses

165

for texts in which the co-extensional ties are provided by units larger than content words. In fact, support for including larger lexical elements comes from Halliday (1994: 311), who includes ‘wordings having more than one lexical item in them, such as maintaining an express locomotive at full steam’ (italics in the original) in his description of what constitutes reiteration and collocation, both of which are ‘relations between lexical elements’, that is, features of lexical cohesion. As we will see in Co-extensional Analysis 3 in Section 8.3.4 (a text titled Israel’s Royal Welcome), the textual cohesion comes from co-extensional ties between chunks of text which refer to events or circumstances (in the case of Co-extensional Analysis 3, events or circumstances which demonstrate the overall ‘leitmotif’ of the text, i.e. Israel’s discrimination against Palestinians and non-Jews). In texts such as this, the sense-relation restriction is problematic since cohesion is formed by items which cannot be categorised into sense-relations. In fact, the cohesion here goes beyond lexical cohesion; these chunks (which we will call ‘propositions’) are made up of highly complex noun phrases, clauses, and sentences, and so the cohesion can be argued to be more ‘thematic’ (see analysis in Section 8.3.4). So, perhaps, rather than delimiting the ‘general field of meaning’ by senserelations, it would be more logical for items to form a similarity chain if items in a text are in the same semantic field; this may well correspond to the initial item in the chain or the main theme of a text. So, in the flower, petal, etc. example above, the semantic field would be flowers and the unrelated items of fire, tinder, etc. would simply not be accepted as part of that grouping. The formation of similarity chains in this way would preclude the need for the sense-relation restriction and therefore allow for larger units of meaning to be included. Through this discussion of Halliday and Hasan’s (1985) approach to lexical cohesion, we suggested that the co-extensional framework should be broadened to include utterances larger than content words or lexical items because cohesion can be created by units (i.e. propositions) which illustrate events and circumstances pertaining to the overall leitmotif of a text. This change would, however, mean doing away with the sense-relation delimitation and instead require that similarity chains are formed by items which are in the same semantic field. In the following section, we will see these changes in practice. 8.3

Co-extensional Analyses

Out of the meaning relations discussed above, co-extension is the most relevant to the analysis of the A(n)+N+RRC expressions. Therefore, to explore the potential and parameters of Halliday and Hasan’s approach to lexical

166

Indefinite Expressions for Referring Definitely

cohesion, and to show that the use of these expressions cannot be fully captured by existing accounts of late indefinites, four examples of A(n)+N+RRC expressions and their texts are analysed. Each expression and its co-text offers something unique to the analysis or they highlight a distinct aspect of these expressions. These special features are discussed in the analyses below, but first, we need to establish how the similarity chains in the texts were identified for this analysis. In any text there are manifold similarity chains running through it, but what we are concerned with in this discussion are the similarity chains which show how the writer provides a shared definite context in the text so that the indefinite expression can be understood as referring to an established referent. Key similarity chains were identified in the texts by dividing the relative clause of each A(n)+N+RRC expression into component ‘chunks’ of information, which hold some kind of semantic content in their own right. In some cases, this corresponds to an entire clause. The result of this was that sometimes the relative clauses were not divided at all as there was only one semantic unit and therefore just one similarity chain (see Figure 8.4), and others are very complex relative clauses and contain several semantic units, and thus several similarity chains (see Figure 8.3). Members identified as being part of the similarity chains are based on information related to the RRC given in the text preceding the occurrence of the A(n)+N+RRC expression. 8.3.1

Background to Co-extensional Analyses and Texts

To better understand the following analyses, we provide descriptions of the terms used below as well as the rationale for the selection of the texts. Identity chain: the referent being tracked and the number of times the referent has been mentioned in the text preceding the occurrence of the A(n)+N+RRC expression. Similarity chain: the members of the chain which create lexical cohesion through mentions of items in the same general field of meaning as the items identified in the RRCs. For this analysis, only items which have occurred in the text preceding the occurrence of the A(n)+N+RRC expression are traced because how the writer creates a sufficiently ‘shared’ context for the indefinite expression to get interpreted as definite is of interest. In this way, these items are ‘old’ in the discourse model of the addressee (Prince, 1992: 303) and textually traceable. Shared cultural knowledge: the information in the relative clauses can also be ‘hearer-old’ (Prince, 1992: 301) or ‘copresent’ (Clark and Marshall, 1981: 38), which means that the information is considered old with respect to the text

8.3 Co-extensional Analyses

167

producer’s assumptions about what the hearer already knows (Prince, 1992: 301). To put this into context, it is necessary to consider the kind of texts being analysed here, as well as the assumed readership. The texts are opinion pieces from the British broadsheet newspaper The Guardian. They are about a specific, well-known entity in a particular field of discourse. We can assume that readers of such articles have an interest in the particular entity being discussed and are familiar with the field of discourse, and therefore have some previous, culturally shared knowledge about it (or at the very least, the writer is behaving as if they do). Thus, the connection is outside the text, in the broader field of discourse, but nevertheless helps add to the overall cohesion which ties the attributes in the RRCs to the specific participant in the text (cf. co-reference and co-classification, which are not considered cohesive if they occur outside the text (Halliday and Hasan, 1976)). A note about the expressions and texts: The co-extensional analysis presented below is based on four A(n)+N+RRC expressions as used in texts published in The Guardian as mentioned above. The reasons for selecting these particular expressions and texts are various, which we will outline here. The first expression (shown in Figure 8.2) comes from an editorial text, In Praise of . . . Reader’s Digest, and the referent we are interested in is the Reader’s Digest. This expression was selected because the members of its similarity chain correspond most closely to Hasan’s notion of ‘lexical items’ or ‘content words’ forming co-extensional ties, which allows us to illustrate an analysis which closely reflects Hasan’s original framework since we do not need to extend the parameters of the framework to include larger units of text. The second A(n)+N+RRC expression (shown in Figure 8.3) refers to David Beckham, and is from an opinion piece called David Beckham: how this crock of a footballer can still woo the French. The expression in Figure 8.3 demonstrates how complex these expressions can be. It contains several similarity chains, each of which has several members, and as well as illustrating cotextual cohesive ties, the expression contains a link to shared cultural knowledge. The third A(n)+N+RRC expression (shown in Figure 8.4) comes from an opinion article called Israel’s Royal Welcome and has as its main participant The Jewish National Fund. The expression in Figure 8.4 provides a good example of why we need to extend the field of lexical items to whole propositions, and thus provides an opportunity to show how Hasan’s framework can be expanded. The final expression we analyse here (see Figure 8.5) comes from a text called The man who came in from the cold and refers to Greg Dyke, a British media executive, shortly after he resigned from the BBC in 2004. The A(n)+N+RRC expression in Figure 8.5 was chosen because it does not have any co-textual ties, but instead the writer relies on shared cultural

168

Indefinite Expressions for Referring Definitely

knowledge for identification of the referent. What is particularly interesting about this example is that the identity chain of reference is continued in the RRC, which provides the reader with additional support to identify the referent. 8.3.2

Co-extensional Analysis 1: The Reader’s Digest Text

The A(n)+N+RRC expression in Figure 8.2 corresponds most closely to Hasan’s notion of how content words contribute to forming co-extensional ties. It illustrates a relatively simple co-extensional meaning relationship between the A(n)+N+RRC expression and the text. The members of the similarity chains (SCs) appear in the text before the occurrence of the A(n) +N+RRC expression and are considered to be in the same general field of meaning as the similarity chain. SC1 ‘part of the furniture’ has two members, one of which, so familiar, relates to the notion that if something is part of the furniture, it is therefore familiar. The remaining member, comforting anecdotes, semantically and cohesively links the idea that something that is part of the furniture, and is therefore familiar, is also usually comforting. SC2 ‘for so many years’ has three members (its foundation in 1922, in the 1950s, in the 1970s), which all relate to past time. But not only does the expression have cotextual cohesive ties, it also contains shared cultural knowledge. The Reader’s Digest is a well-known general interest ‘family’ magazine, and even if readers of the article have not read it personally, they would likely be aware of its existence as an established ‘institution’ in the United Kingdom and that it has been ‘part of the furniture for so many years’.

Figure 8.2 Similarity chain for the Reader’s Digest text

8.3 Co-extensional Analyses

8.3.3

169

Co-extensional Analysis 2: The David Beckham Text

The David Beckham text shown in Figure 8.3 demonstrates how complex A(n) +N+RRC expressions can be. This expression contains three similarity chains, each of which has several members, and as well as illustrating co-textual cohesive ties, it shows a link to shared cultural knowledge that the writer assumes the reader brings with them to the text. SC1, ‘appearance’, has two textual members and is also shared cultural knowledge: Beckham’s looks and physique get considerable media attention in the United Kingdom. The second, ‘charming the French’ has five members and ties together the core argument of the text, that although Beckham is old (‘age’ being another similarity chain, which is dealt with and dismissed as an argument in paragraph 3), his personal attributes may still help him win over the French. SC3, ‘move to France’, has five members and makes repeated reference to the situation described in the text, that Beckham has been signed to play for Paris St-Germain. Two of the members of SC3, a British cultural ambassador to France and an astute diplomatic move, also relate cohesively to SC2 as they suggest that one of the reasons for his move to France is his charm. We are also provided with some textually new information, non-novelty underwear, which can certainly be considered as culturally shared knowledge (in the United Kingdom at least). At the time this article was written, it was difficult to step outside one’s house without seeing images of David Beckham modelling his own underwear range. Thus, the connection is outside the text, to the reader’s assumed cultural knowledge of Beckham, but it nevertheless creates cohesion. Something that

Figure 8.3 Similarity chain for the David Beckham text

170

Indefinite Expressions for Referring Definitely

is particularly striking in this example is that the A(n)+N+RRC expression contains two direct definite references. That is, the possessive pronoun his occurs twice in the expression, which can only be interpreted as referring to the previously mentioned participant in the text, David Beckham. 8.3.4

Co-extensional Analysis 3: The Jewish National Fund Text

Hasan’s (1985: 80–81) claim that co-extension is usually realised by lexical items is problematic for the example we see in Figure 8.4. The textual cohesion in this article is created by units larger than individual lexical items or grammatical units. It is clear that these larger chunks of text (i.e. propositions) about discrimination against Palestinians and other people who are not Jewish are crucial to the continuity of the argument running through the article and therefore to the links of the preceding text to the A(n)+N+ RRC expression. It would be impossible to create a similarity chain without them because it is reference to events and circumstances in the preceding text which creates the textual cohesion. More specifically, the members of the similarity chain in this text are all illustrations of actions taken by the Jewish National Fund which discriminate against Palestinians and others; the relative clause sums up the overall ‘leitmotif’ of the text by providing an overview of what all these occurrences exemplify. Without the inclusion of propositions as items which are able to create cohesive ties, it would not have been possible to demonstrate how the preceding text is cohesively linked to the RRC in this text.

Figure 8.4 Similarity chain for the Jewish National Fund text

8.3 Co-extensional Analyses

171

Figure 8.5 Similarity chain for the Greg Dyke text

8.3.5

Co-extensional Analysis 4: The Greg Dyke Text

The text in Figure 8.5 illustrates an A(n)+N+RRC expression without textual co-extensional ties. The expression in this text occurs in the second paragraph, after five previous mentions to the participant, Greg Dyke. There are no lexical links to the preceding co-text, so the question is: how can the connection between the entity in the indefinite expression and the definite referent be made? There seem to be several strategies. First, the identity chain of references to the participant is continued in the RRC, with references to his defence and he took. Furthermore, this RRC contains other features of definiteness, a specific episodic event and a definite past time. Moreover, the content of the RRC can be considered shared cultural knowledge, but the time of writing and context must be taken into account. This article was written on 29 January 2004, the day that Greg Dyke resigned from the BBC after an extremely well-publicised enquiry (The Hutton Report) into errors of judgement made by the BBC when checking news stories. The media furore surrounding this affair was intense, and much of what was discussed was about the government’s interference in the BBC and Dyke’s resistance to this. So, even though there are no preceding textual ties between the text and the RRC in this instance, the connection to the referent is made by continuity of reference and cohesive ties to shared cultural knowledge of the time. There are ties, not to the text but to the ongoing discourse surrounding the events described in the text. So again, there is no newness in the expression which can account for the use of the indefinite article. 8.3.6

Concluding Comments on the Analyses

The analyses in this section have illustrated various features of the A(n)+N +RRC expressions we have been discussing. We have seen how these features operate within their texts to create cohesion so that the indefinite expressions can be interpreted as referring to the definite referent. More specifically, the

172

Indefinite Expressions for Referring Definitely

analyses have shown that the content of the RRCs is connected to the preceding discourse via textual co-extensional ties and/or to the shared cultural knowledge that the reader is assumed to bring with them to the text. The following section considers how and why this exploitation of traditional referring conventions might occur. 8.4

Discussion: ‘Dual’ Reference

This analysis has revealed the limitations of existing accounts of late indefinites and challenged and extended Hasan’s co-extensional framework. Coextensional analysis can provide a means of illustrating how the referents of these complex indefinite expressions are cohesively tied to the preceding text and ongoing discourse. The writer does not simply rely on textual cohesion, but also taps into the cultural knowledge assumed as shared by the readership, that is, the writer makes out that the information is ‘hearer-old’, given the likely readership of this of kind of article. Another strategy to guide interpretation is the continuation of the identity chain of reference. In each of the expressions, there is something to link the expression to the specific participant, whether it is textual ties, contextual/discoursal ties, or continuity of reference. Thus, there is an assumption of cooperation and adherence to the principle of distant responsibility towards the reader (Clark and Wilkes-Gibbs, 1986: 34), through the provision of a platform for the interpretation of the expression as referring to the established participant. These expressions can be highly complex and have manifold ties to the preceding text and surrounding discourse. These ties allow the expressions to function in a definite manner, but perhaps it is not only the ties that perform this role. The reader may be happy to be persuaded by the presence of specific detail in the relative clause that the expression is referring to the previously mentioned entity. The amount of detail in the relative clause may well indicate how much the writer intends for the reader to interpret the expression as referring to the definite entity. We will now consider why writers might do this. One explanation for why the chain of definite reference is broken and a switch to indefinite reference is made is that there is some kind of duality in this type of reference: reference to the fully identified referent and to a virtual referent with the same qualities (cf. Chapter 10). Rather than the expression being understood as a generalisation to a type or sub-class of entity or, on the other hand, definite reference to the identified individual, the two factors are considered simultaneously. Thus, the writer intends to refer to both the individual and a type, and the type is the generalisable form associated with the entity, creating what Epstein (1994: 226) calls a ‘generalising effect’. For Epstein (ibid.: 227), these expressions have two simultaneous functions: they are reclassified in light of the new information in the relative clause, and the ascribed property is presented as both characteristic of the whole class and

8.4 Discussion: ‘Dual’ Reference

173

implicitly portrayed as characteristic of a specific uniquely identifiable entity. The addressee recognises that these expressions are not simply referring to an arbitrary member of the particular category and is aware of the identity of the referent ‘thanks to the (virtual) link between it and its previous mention’ (ibid). However, Epstein, like Schouten and Vonk (1995) and Ushie (1986), explains the use of the indefinite article through some element of newness in the expression, but as we have seen from the co-extensional analyses, the expressions do not contain new information, whether it is discourse-new or hearernew (cf. Prince, 1992), so the newness hypothesis does not hold. However, the duality explanation should not be abandoned because of this. An explanation for why the switch from definite to indefinite reference occurs in these particular texts comes from journalism academic Howard Barrell (personal communication, 2012),4 who suggests that [A]n important feature of our Western rational tradition is the assumption that we best explain life by identifying and referring to regularities. [. . .] Journalists who recognise this [. . .] often wish they were involved in something less random in its collection of data. They find themselves dealing [. . .] with individual stories – stories that are saleable to audiences precisely because they are individual, a departure from the norm, or sensational in some way. Yet journalists simultaneously yearn to be able to extrude some kind of ‘pattern’, ‘regularity’ or ‘rule’ from one or a collection of these stories. Indeed, ‘analysis’ (and perhaps ‘comment’ as well) would seem to demand that they do so.

One way journalists are able to achieve this is by using an indefinite expression to refer to something already established. Using a definite expression initially allows the journalist to establish the parameters and relationships which characterise their story and make it unique. When the journalist wishes to argue that the story they have told about a specific individual ‘may be governed by some rule or regularity of politics or of existence’ (Barrell, 2012, personal communication) or that their story may enable the reader to identify some previously unrecognised rule or regularity, the journalist ‘may very well start referring to “a prime minister who does this/fails to do that”’. So, the intention of the journalist here is to abstract from the specificities and limitations of their story a more general truth. The use of an indefinite expression makes of the story an instance or example of some truth or regularity beyond itself, suggesting that the journalist intends to refer to the specific individual and at the same time establish an abstract principle: ‘dual’ reference. The effect of this categorical principle is more powerful when the expression is less explicit: the broader the regularity, the more abstract and generalisable it becomes, and therefore the stronger the impact. But the more explicit the details in the RRC (and therefore the more ties), the stronger the association 4

Interview with Dr Howard Barrell at the School of Journalism, Media and Cultural Studies, Cardiff University, November 2012.

174

Indefinite Expressions for Referring Definitely

is with the specific individual, but as a consequence of this explicitness, the regularity or general truth is less powerful. This explanation is certainly reasonable for the use of this expression in the context of opinion journalism, but further exploration into other genres or registers may reveal data that leads to different explanations. Future research should investigate whether the A(n)+N+RRC expression occurs in spoken data and other written genres, and if so, what function it has. 8.5

Conclusion

This chapter, like the previous one, has explored an atypical use of reference. The type of expression considered in this chapter is complex in terms of the noun phrase structure and in that sense, somewhat typical of prepared writing. However, due to the indefinite article which to some extent determines its indefinite status, its use for definite reference suggests a deviation from conventional accessibility marking. The use and function of the expression A(n)+N +RRC which we have seen in journalistic opinion writing presents challenges to existing accounts of late indefinites and Hasan’s co-extensional framework (1985). Explanations of late indefinites tend to rely on the newness factor, but the co-extensional analyses showed that the information in the expressions is not actually new, as evidenced by the context and readership of the texts. It was also suggested that the co-extensional framework be broadened to accommodate linguistic units larger than lexical items because cohesive ties can be formed by clauses and sentences depicting events and actions. English does not have an explicit linguistic way of exhibiting simultaneously the relationship between old and identifiable and new and nonidentifiable, so the relation has to operate outside of language, perhaps in cognition only. It is difficult to engineer a way in which language is responsive to this relation (in English anyway), as it only responds to the construal of the ‘virtual’ referent as non-identifiable (cf. Du Bois, 1997). However, the writers of the texts analysed have found a way of ensuring that the association to the old, identifiable referent remains explicit, through the use of an RRC which contains already established information which is tied cohesively to the cotext, context, or ongoing discourse about the specific entity. There are three components carrying meaning in the A(n)+N+RRC expressions which guide interpretation: the indefinite article; the detail and level of specificity of the information in the relative clause; and the surrounding cotext, context, and ongoing discourse about the specific entity. To interpret the expressions in a meaningful way, the three components cannot be separated. Only when they are considered as three interrelated parts which make up the whole can we begin to understand the use and function of the A(n)+N +RRC expression.

9

Child’s Play

In the development of the child as a social being, language has the central role. Language is the main channel through which the patterns of living are transmitted to [children], through which [they] learn to act as a member of a ‘society’ . . . and to adopt its ‘culture’. (Halliday, 1978: 9)

9.1

Introduction

We have so far in this volume only discussed reference from the perspective of adult discourse. The previous two chapters focussed on examples of atypical reference. While the previous two chapters focussed on examples of atypical reference, in this chapter we turn our attention to children by looking at the normal development of referring and at how creative and playful reference can be for them. We will first provide an overview of how children develop their use of reference along with their general language skills. Following this, in Section 9.3, we focus on degrees of specification since we know adults sometimes overspecify their intended referent, but with children we generally find that they underspecify their intended referent. Our attention here will be on children’s use of referring expression from a broad range of examples, which will complement our discussion of how reference develops generally. We will show that children quickly get to a stage where they can exploit reference for their own purposes or just to have fun, which is when their expressions become atypical. The playful nature of referring expressions is then explored in Section 9.4, drawing on examples from children’s literature. Whereas the variations we find in children’s use of referring are different from those of adults, referring expression use in children is still typical even if evolving through their development. The literary examples we will look at are deliberate exploitations of known conventions in relation to reference. 9.2

Children and Reference

The quote by Halliday (1978) at the beginning of this chapter captures the perspective we have been taking throughout the volume, which is that 175

176

Child’s Play

language is meaning-driven. Speakers are fundamentally oriented to exchange meanings, and language is our tool to do so. For Halliday (e.g. 1975, 1978), language is not acquired, but rather children are learning how to mean; in other words, they are learning how to make meaning with others. In this view, language is a system of meaning making or of meaning potential. According to Halliday (1975: 33), children are making meaning well before they have what we think of as words or grammar; they are ‘capable of expressing a range of meanings that at first seem difficult to pin down, because they do not translate easily into adult language, but that become quite transparent when interpreted functionally, in the light of the question, “What has the child learnt to do by means of language?”’. What children learn to do with reference is what we will focus on in this section, beginning first with protolanguage, moving to the development of the concept of common ground, and finally to the stage where children have mastered the conventional uses of reference and begin to master language (i.e. they can play with it). 9.2.1

The Earliest Stage: Protolanguage

[T]he infant’s protolanguage is a system of simple signs, or value-&-token pairs, where the value is the semantic content and the token is the phonetic and/or kinetic expression (i.e. sound and/or gesture). There is no lexicogrammar in it, and no reference. This kind of semiotic system is typical of humans in their first 12 to 15 months of life; it develops along with the human body, where it tracks the various stages of bodily control: raising the head, rolling over, sitting up, and crawling, and it starts to be superseded when the child begins walking upright. (Halliday, 2011: 73)

The precursors to reference begin long before any adult can recognise something akin to language as we understand it as adults. As noted by Serratrice and Allen (2015: 2), ‘even before the appearance of first words, children can track and identify referents through gesture and eye gaze’. Gaze-focussed attention is an important step to linguistic reference. According to Salazar-Orvig et al. (2021: 11), by the time children are uttering their first words, they ‘have already mastered two “proto-referential” non-verbal functions (1) attracting their interlocutor’s attention to an object or an event, and thus establishing joint attention, (2) and taking part in a joint-attention episode’. In other words, there is content in the sense of a referent and an awareness that someone else is also focussed on that referent. From very early on, children may be able to identify objects and draw someone else’s attention to them simply by the use of eye gaze, gesture, and/or sound. According to Hannah and Brennan (2007: 596), ‘the ability to use eye gaze cues develops very early in life. Infants not only are highly sensitive to another person’s head position, but attend to the other’s eyes by 2 months of age, to the direction of gaze by around 6 months, and, with accuracy, to the object that the other is fixating by around 12 months’.

9.3 The Earliest Stage: Protolanguage

177

Adults begin to notice words in infant babbling around the age of 9 to 15 months, many of which are onomatopoeic (Laing, 2019). Lexical development begins relatively slowly, over a 6 to 8-month period, until the child has evidence of roughly fifty words. We are using ‘word’ here rather loosely (cf. Wray, 2015) and not considering whether for the child a given utterance is one or two words. In terms of lexical development, we understand that these utterances are holophrastic, for example ‘allgone’ (Wray, 1999: 220), and in this stage there is no evidence that the child has a ‘grammar’, that is, not producing novel combinations of words. Once the child has enough of a lexical base, new words come into use at a much higher speed (see Stromswold, 2000: 910; also Levelt, 2013: ch 10; McCabe, 2021: 42 & 61).1 This word spurt is associated, in part, with a realisation that words can be used to refer to things, whereas before they might just be uttered as part of a routine. Aitchison (2011: 103–104) explains that The child may shriek BA delightedly when she reaches the page with the round blue blob in the middle, but may not for some weeks realize that this sequence of sounds is actually the ‘name’ of a certain type of round object, a ball. . . . The sudden realization that things have names appears to lead to a surge of ‘labelling’ everyday objects such as CAR, MILK, BALL, APPLE, followed by a surge of ‘describing’, with the use of words such as BLUE, GONE, BROKE, HIT.

The awareness of naming things is an important developmental milestone, one that effectively enables children to make sense of their world by categorising instances. Many parents of English-speaking children will no doubt have been slightly embarrassed when their child refers to some unknown man in the street or in a shop as dada, sometimes to the point that makes the parent say ‘it’s her word for male adult’. When the expression is uttered in the home, the parents would likely gleefully assume that their child had learnt to refer to their father, but in fact the child is still working out that there is only one member of the class called dada (i.e. father). We can appreciate the child’s reasoning since it worked for dog (in whatever phonological form, e.g. /da/). As Halliday (1998: 11) explains, With memory, the child is able to construe classes of phenomena out of repeated instances, using a re-entrant mapping to impose categories on its experience of the world. Linguistically, this means that the child is now able to construe “common” terms ± generalized common nouns as opposed to individuated proper nouns; and this is the beginning of reference, referential meaning. (The early individuated signs of the protolanguage are not yet referential; a protolinguistic mama means something like “I want (you) mummy”.) 1

We are grateful to Anne McCabe for pointing out that this rapid development was noted very early in child language studies by Stern and Stern (1907), which is discussed in Levelt (2013: 338–339).

178

Child’s Play

This phase corresponds to Wray and Perkins’ (2000) phase 2, which lays the foundation for the more operational or generative stage of language development that follows. The move out of the protolanguage stage and into a more adult-like language system, that is, patterns of language which construe meanings, is, according to Wray (1999: 222), ‘supported by the relative stability of operating within a “socio-interactional bubble”, in which most linguistic communication occurs within a restricted number of familiar child-careroriented situations’. For Wray (2002), the more or less holophrastic nature of the child’s lexicon at this stage is the beginning of the development of the heteromorphic lexicon, which sees variation in the various components of the lexicon. With holophrastic utterances (or formulaic expressions, as Wray calls them), there is not yet reference as we understand it. We find, at this stage, that an utterance expresses a meaningful act (e.g. a declaration, a request). Towards the end of this stage, the child starts to recognise patterns and becomes analytical, recognising that parts of the expressions can be useful in expanding the linguistic repertoire available. For example, allgone, which might be used as a declaration of disappearance, might end up in an utterance such as birdie gone, and when this happens, we begin to see the child’s language as an emerging complex system of meaning potential. According to Wray and Perkins (2000: 21), phase 2 begins sometime in the child’s second year (between 20 and 30 months of age) and continues to develop until around 8 years of age. During this stage, we find ‘a preference for analytic over formulaic language processing’ (ibid.). As we will see below, by the end of this stage, reference is entirely (or nearly) adult-like. The key developments as concerns reference, then, occur in the analytic phase (i.e. phase 2) from years 2 to 8, when children are developing their own language system. Structural language skills are, according to Davies et al. (2016: 109), very important in supporting the child’s development of reference. As pointed out by SalazarOrvig (2019: 284), ‘full referring expressions progressively emerge with grammatical and syntactical development’. It is important to note that what we have described above relates to a very general view of language development in the first two years of life, and we must acknowledge that individual variation occurs, given that there are many factors that can influence this development. For example, an infant without sight is not going to develop gaze as a means of directing attention. 9.2.2

Developing Awareness of Common Ground

We have emphasised throughout the volume, but in particular in our discussion above in relation to child language development, that language is inherently social and interactive, while also fundamentally cognitive. Reference in particular is collaborative, and we saw that this is true even in the protolanguage

9.3 The Earliest Stage: Protolanguage

179

phase and as part of proto-reference. As Castillo et al. (2019: 6) explain, the factors involved in interaction also require continual evaluation of a variety of issues in relation to linguistic choice and the nature of the task at hand. ‘[T]hese factors may sometimes exert conflicting pressures: A perceptually salient choice might not be communicatively efficient, and repeating a partner’s linguistic choice does not ensure optimal adaptation to the task. Thus, the emergence, establishment, and evolution of linguistic conventions are likely to be the product of multiple competing pressures’ (ibid.). Successful reference, therefore, is complex, involving not only many different aspects, but some which present conflicting influences on referential choice. We will not be able to explore such interactions in what follows, but we would like to keep this in mind as we consider the development of reference in children during the analytical stage of language development. None of the contributing factors operate in isolation. 9.2.2.1 Common Ground The term common ground is ubiquitous in discussions of reference, and we have used it throughout this volume without formally defining it. We assume that it represents ‘the knowledge that [speakers] believe themselves to share with their listeners’ (Branigan, Bell, and McLean, 2016: 2; see also Clark et al., 1983; Allan, 2013; and many others). In addition, as pointed out by Stegenwallner-Schütz and Adani (2020: 277), common ground information is not only shared but also ‘modified between interlocutors during a dialogue’. It is generally assumed that this is the main area influencing the development of reference in children, that is, that they, for whatever reason, may sometimes not effectively estimate shared information with their interlocutors. We will take a closer look at this in what follows, including how well children manage the concepts we have covered in previous chapters, such as givenness, accessibility, and definiteness. In a study on the use of repetition, Clark and Bernicot (2008) found that adults and children often use repetition to share information. In their study, French-speaking children between the average ages of 2;3 to 3;6 years were observed in interaction with their parents. They found that ‘repetition as a conversational device allows both expert and novice speakers to add information to common ground’ (ibid.: 364). An example of a repetition from their data is presented as (140), where speaker repetitions are indicated by > (for the speaker) and underscored for the repeated expression; glosses are given in inverted commas, and interpretation of what the child intended to say in parentheses. As the authors point out, these repetitions are often to establish that the right information is shared, in this case whether or not the child is saying they want a drink of something. What happens later in the child’s development is that as they develop more skill, ‘they can make more use of pronouns and deictics, as alternatives to repeats, when they add information to

180

Child’s Play

common ground’ (ibid.: 363). These results are echoed elsewhere, for example, by Salazar-Orvig (2019: 299), who highlights the significance of repetition: ‘in dialogues, repetition and answer can account for a significant proportion of children’s referring expressions’. (140)

MOTHER:

Tu veux pas boire? Mange alors ‘You don’t want to drink? Eat then.’ > CORENTHIN (2;3): Hum. Eu boi (hum. Veux boire) ‘hum want drink’ > MOTHER: Tu veux boire? ‘You want to drink?’ > CORENTHIN (2;3): Hum ‘hum’

Repetition is, however, an important feature of language development generally and serves a variety of functions, such as cohesion and ratification (Salazar-Orvig, 2019: 300), in addition to establishing common ground, as we have just seen. For example, as explained by Clark (2007), children use repetition when encountering new words, which allows them to show the following three key aspects of their knowledge development: ‘(i) that they have recognized the adult’s X as a new term or expression; (ii) that they are ratifying the adult’s use of X on this occasion; and (iii) that they are adding this use of X to common ground’ (ibid.: 158). Repetition, then, allows both adult and child to verify they share the same understanding of an utterance and also to confirm or correct the information. In addition, according to Hughes and Allen (2015: 56–57), not only is the repetition of full noun phrases ‘less cognitively demanding’, it also provides children with ‘a way to participate in conversation and assume a role in turn-taking before they have fully mastered the more complex system of pronouns and demonstratives typically used to maintain continuity of reference’. In terms of contributions to the development of reference among children, the awareness of the need to converge on information and to explicitly share this information with an interlocutor as common ground is essential to successful reference. Children demonstrate an awareness of shared knowledge (or the lack of it) much earlier than they seem able to incorporate the perspective of their addressee in their use of language, as we will see below. According to O’Neill (1996: 672), children as young as 2 years old modify their referential expression depending on what they know their addressee knows, suggesting that they ‘can tailor their requests to a communicative partner’s knowledge’. Grigoroglou (2018: 85) showed the level of engagement of the addressee (listener) was more influential than the addressee’s visual access. The 4- and 5-year-old children in her study ‘were more likely to be informative when communicating with an interactive, as opposed to a more passive, listener,

9.3 The Earliest Stage: Protolanguage

181

with other aspects of the interaction remaining unchanged’ (ibid.).2 However, even at age 8, experimental evidence from work by Branigan, Bell, and McLean (2016) suggests that reaching an adult-like understanding of common ground is still in development. As they explain, ‘although children have some understanding that linguistic common ground accumulates differently according to distinctions in listeners’ participant roles, their understanding is not yet fully adultlike’ (ibid.: 11). The concept of common ground, or shared information, involves many factors that reach beyond linguistic choice, including, as mentioned above, shared awareness of the material context (whether visual, auditory, etc.) and a variety of cognitive processes (e.g. working memory capacity, shared information). Part of what we have discussed here involves the use of repetitions as a tool to confirm information and acknowledge that it is shared. Successful referring involves another important cognitive process, which involves taking an alternative perspective than that of your own (see Bezuidenhout, 2019 for a discussion of egocentric views of reference). We will consider perspective taking and accessibility of the referent in Section 9.3.1.3. 9.2.2.2 Definiteness Having discussed (in)definiteness throughout this volume, we now take a brief look at how this conceptual continuum establishes itself in children’s use of referring expressions. According to Salazar-Orvig (2019: 288) ‘[i]n languages that have the definite/indefinite contrast (as English and French), definite and indefinite determiners make up the majority of determiners used by the youngest children (Coene, 2006; Rozendaal and Baker, 2008; Salazar-Orvig et al., 2013) as well as by adults. In the first stages definite determiners precede indefinite ones by far’. Based on a set of experiments testing children’s accuracy rate of using the definite or indefinite article to make reference to fictitious referents, Maratsos (1974) found that, while children aged 3 and 4 years still made mistakes in their use of definite and indefinite articles, both groups used the articles correctly at a rate above chance. By the age of 3, then, children already seem to have a good grasp of differentiating between specific and non-specific reference. These findings were supported by Gundel and Johnson (2013: 43), who found that ‘children use definite and indefinite articles, demonstrative determiners and personal pronouns appropriately by age 3 or earlier’. However, we know that part of using markers of definiteness involves being able to adopt a perspective that is different from your own, and children are not always able to manage this, as we will see in the next sub-section. SalazarOrvig (2019: 296) explains that these results, and other similar studies, suggest 2

See McCabe (2021: 214–215) for a useful discussion of the role of engagement during interactions with children in their language development.

182

Child’s Play

‘young children tend to make “egocentric” errors, that is, they do not seem to take their interlocutor’s knowledge into account and overuse definites in contexts where the referent is new for the interlocutor’. Campbell et al. (2000) investigated factors affecting children’s use of pronouns as referring expressions. In a pair of experiments with children between 2.5 and 3.5 years as well as school-age children, she examined the frequency of children’s response options to two different questions: ‘(a) a generic question: “What happened?”, indicating the questioner’s general lack of knowledge; and (b) a specific question: “What did the motorcycle do?”’ (ibid.: 1339). In response to these questions, children could choose, for example, a null reference such as driving; a full NP, as in The motorcycle was driving; or a pronoun, as in It was driving. The children’s answers (whether to a generic or a specific question type) provided evidence of the children’s different levels of insight into the interlocutor’s knowledge of the situation, that is, their common ground. As a general result, Campbell et al. (2000: 1347) found that children’s use of pronouns was affected by the type of question asked by the experimenter. Children below the age of 4 years and school-age children would answer specific questions with either a null reference or a pronoun use most of the time. This makes sense because a specific situational question such as What did the motorcycle do? inherently indicates the questioner’s knowledge about elements on the situational context, and thus allows children of both ages to accurately judge the common ground they share with their interlocutor. In response to generic questions, on the other hand, the younger cohort would use pronouns most of the time, indicating a lack of insight into the questioner’s knowledge of the situation as they would not be able to decode the referent of the pronoun used. In contrast, school-age children would use full nouns most frequently, ‘perhaps indicating their growing sensitivity to the questioner’s needs in this particular discourse context’ (ibid.: 1347) and thus their growing awareness of information that is not shared with their interlocutor. As children progress through their language development, the frequencies with which they use specific linguistic features, units, or patterns in specific linguistic contexts or registers start to align with those of an adult speaker. According to Burgoon et al. (2002: 120), ‘as communicators mature, they learn not only the mechanics of language but also what to say and when to say it’. The same applies to the use of definite and indefinite expressions. As was mentioned above, while children from age 3 are generally using definite and indefinite articles appropriately for their interlocutor, definite article use is much more frequent and, according to Salazar-Orvig (2019: 295–296), ‘definite determiners shift to exophoric values’. In a third stage, indefinites are used to introduce new specific referents and anaphoric values appear for definites. Cross-linguistic evidence suggests that in all languages, being aware that

9.3 The Earliest Stage: Protolanguage

183

newness must be marked is a late development, according to Hickmann (2003: 319), who explains that ‘[i]ndefinite determiners are used systematically at around seven years in the Indo-European languages’. Gundel and Johnson (2013: 54–55) discuss children’s atypical use of the definite article (as compared to adults’) as found in experimental data by Schaeffer and Matthewson (2005). The children and adult participants were given an image of Mickey Mouse, who had just finished drawing a house, and were then asked the question What did Mickey Mouse do? (see Gundel and Johnson, 2013: 55). All adults, but only about half of the children, responded using an indefinite article, for example, He drew a house. The other half of the children responded with a definite expression, for example, He drew the house. Schaeffer and Matthewson (2005) attributed this difference in article use to the children’s inability to accurately assess the common ground, that is, they assume that because Mickey’s drawing of a house is identifiable to them, it is also identifiable to their interlocutor. However, as Gundel and Johnson (2013: 55) point out, this reasoning does not fully explain why some children answered with a definite article. In fact, in the given experimental situation, the children would be correct in assuming that the house is indeed identifiable to both themselves and their interlocutor as the latter has just given them the picture and has asked them what Mickey was doing. Clearly, both the child and the researcher knew of the existence of the house, which would justify the use of the definite article. Yet, all adult participants chose to answer with an indefinite article, which does not identify the house as being part of the common ground. Gundel and Johnson (2013: 55) explain that ‘in the context of this experiment, [. . .] it is the property of being a house and not the particular identity of the house that is relevant. Adults thus use the indefinite article when conditions for using the definite article are met, if token/unique identifiability is irrelevant’. In this sense, then, there is no need to use a definite article to refer to Mickey’s drawing of a house because it is not important to specify which house he drew. Instead, it is only important that Mickey drew a house rather than, say, a tiger. Using a definite article in this situation is overspecifying the referent. Indeed, Mickey drew the house creates an inference that there are two drawings – one of a house and a second of something else – and that Mickey did not draw the second drawing, but only the house. It is possible that the children who responded using a definite article might not yet have been fully aware of the pragmatic inferences that arose from overspecifying the referent. In Section 9.4 we will discuss how children typically underspecify their referents, although there is considerable variation. It is very difficult to get a clear sense of how reference develops in children’s language development. The difficulty stems from the nature of the data, how they are obtained, and the varying methodological approaches, notably differences in focus and methods from the psycholinguistic and discourse traditions,

184

Child’s Play

for example. However, not only are these factors preventing a consolidated account, but we also find conflicting results and interpretations. As put forward by Salazar-Orvig (2019: 298), ‘[c]onflicting results could be due to the differences of dialogical contexts. It has often been stressed that dialogues provide children with a scaffolding context in which they are not faced with challenges such as the absence of mutual knowledge and diverging perspectives which would reveal cognitive difficulties’. Sometimes, it truly is a wonder that we ever learn how to use language so effortlessly and in such a relatively short space of time. 9.2.2.3 Learning to Take Another Perspective In the previous sub-section, we suggested that part of the challenge for children as they develop their use of reference is that they may struggle to manage the information that is shared between them and their interlocutors. However, Bezuidenhout (2019: 54) explains that ‘a good deal of evidence has accumulated that children are not egocentric and that their poor performance on referential communication tasks can be attributed to other causes’. Earlier in this chapter, we also briefly discussed just how much cognitive demand is involved in successful reference. It is understandably the case that not all children will have the same cognitive resources available to them in any given interactional setting, for any number of reasons. Recent work by Meir and Novogrodsky (2021: 26), who studied the use of definiteness in monolingual and bilingual Russian–Hebrew speaking children, suggests that bilingualism may compromise ‘the encoding of definiteness [. . .] in monolingual children with ASD [Autism Spectrum Disorder] and in bilingual children with and without ASD due to their lower morpho-syntactic skills. Among bilingual children this is due to reduced language exposure and among children with ASD, to language deficits’. The task, then, for children in mastering reference is rife with challenges, and we cannot assume that all children will face the same difficulties or have the same kinds of contexts and discourses, or any number of other influencing factors. As we saw in earlier chapters, becoming skilled with reference necessarily involves not only the ‘ability to assess the mental states of others’ (Gundel and Johnson, 2013: 44), but also to be aware of and to update as needed the common ground; given that common ground is dynamic not static. The knowledge that children need to develop has been described by Gundel and Johnson (2013: 46) as follows: In addition to this linguistic knowledge, children must be able to assess if something has a particular cognitive status, for example is the addressee’s attention focused on it, just as they must be able to assess whether something is a dog or not. They must also assess how much information about cognitive status (or conceptual content) is necessary and relevant, i.e. when to use the most informative form and when a weaker form is sufficient or would result in an implicature that the information encoded by the stronger

9.4 Under/Overspecification

185

form does not apply. These non-linguistic abilities involve being able to assess an interlocutor’s mental state in relation to the object referred to and understanding that this may be different from their own.

In short, children must not only learn what different words mean and how to use them in different contexts, they must also learn to understand the common ground they share with their interlocutor and to be aware of information that might differ from that of their interlocutor. Consequently, they must learn how to efficiently bridge any potential information gaps. As was discussed above, most children use many aspects of referring accurately to a high degree by the age of 3 years. In their paper on children’s use of referring expressions in spontaneous discourse, however, Gundel and Johnson (2013) show that even though children make successful use of reference most of the time, there are some ways in which they diverge, understandably, from how adults specify their intended referents. Gundel and Johnson (2013: 55) explain that referential choice involves a level of ‘conscious reasoning about the epistemic states of others, and as such corresponds to a kind of mind-reading ability that is typically not fully developed until after the age of 4’. The authors suggest this may explain some of the high frequencies of certain types of referring expressions, such as demonstratives. While these expressions may seem difficult to explain, we should recall that there may be structural factors which conflict with referential status and/or the ability to reason about epistemic states. For example, Theakston (2012: 710) states that sentence position can affect accessibility and, we presume, whether an expression encodes newness or not. As she explains, ‘[i]n naturalistic studies, although accessibility, when defined as prior mention in the discourse, does not exclusively determine children’s choice of referring expression, this is typically because there are other discourse factors at play to indicate the intended referent’ (ibid.). Theakston (ibid.: 711) further proposes that by age 5, children already conform to preferred argument structure (PAS), that is, pronouns are preferred in subject position in spontaneous discourse. This distribution pattern directs where pronouns (or contentlight nominal expressions) should be placed in the utterance, that is, they should be first. It may be the case, as suggested by Theakston (2012: 714), that ‘[c]hildren’s apparently sophisticated abilities to appropriately select referring expressions in naturalistic discourse may mask an immature understanding of how information status determines choice of referring expression in circumstances in which information status competes with patterns of PAS, and in particular, under more complex discourse conditions’. 9.3

Under/Overspecification

Throughout this chapter, we have considered various aspects of referring, such as the precursors to referential language, the concept of definiteness, the extent

186

Child’s Play

to which children manage shared information, and how they develop their ability to take the perspective of their addressee. These discussions have not been conclusive on how these factors interact, although the picture we gain from the above discussions is that children generally function very well referentially by 3 years, and they continue to develop this skill along with greater grammatical dexterity until the age of 8 or more. There is still one aspect we wish to isolate for the purposes of discussion, and this concerns what it means for children’s expressions to under- or overspecify their intended referent. What we mean by this is whether the expression sufficiently encodes the ‘right amount’ of information, according to referential conventions. This idea of producing an expression that provides sufficient information generally stems from Gricean theory (Grice, 1975), which, as Tourtouri et al. (2021: 2) explain, means that ‘optimal referring expressions should convey only necessary information (minimal specifications)’ and in particular, ‘speaker’s utterances should carry no less information than required for the purposes of the exchange (first maxim), but also no more information than required (second maxim)’. As we saw in Chapter 7, while it is very common for the addressee to decide that an expression is not informative enough, sometimes the speaker realises this and adds or updates the information provided in the initial referring expression. We noted in Chapter 6 that when it comes to prepared writing, getting the informational content of expressions right is important because there is no opportunity to add or remove information if the reader either needs more information or is struggling because of too much information. It is not the case that more or less information is better but that there is an expectation set up by convention with respect to referring that speakers rely on. As we saw in Chapter 4, we consider adhering to these conventions as typical, but these conventions are not deterministic; there is variation. While we will not say that variations on these conventions are marked, we would say that they add information because they are, to some extent, surprising in some way. Givón’s (1985: 197) principle of quantity explains this in the following terms: ‘the less predictable/accessible/continuous a topic is, the more coding material is used to represent it in language’. As we can see in Figure 9.1, which looks on the surface similar to other models we have discussed in this volume, the full NPs are viewed as coding the most information, which is typically associated with less accessible referents and/or less given (more new) referents. This model is particularly relevant to examining under- and overspecification, precisely because it allows us to consider how much information is encoded. This correspondence has been noted to be quite strong (see Perniss and Özyürek, 2015: 39); the more predictable the information, the less

9.4 Under/Overspecification

187

Most continuous/predictable Zero anaphora Unstressed pronouns/verb agreement Stressed pronouns Full NPs Modified full NPs

Least continuous/predictable Figure 9.1 Principle of quantity (Givón, 1985)

that needs to be said. As concerns referent accessibility and referential choice, if reference to a referent is unpredictable (in informational terms), then more markers of information are needed, for example, a full NP. The ability to adequately specify the referent is an important skill as underspecifications run the risk of leaving the addressee with few resources, other than inference, to ratify the expression. If we consider the principle of quantity, underspecifications do not encode enough information for the addressee and, consequently, they signal to the addressee that the speaker believes they should have this information. As we have seen, zero anaphora and pronouns do not specify information but rather signal that the information is elsewhere and is evidently retrievable. These forms have particular significance for discourse maintenance and topic management in particular. In a comparison of referential choice in children and elderly adults, Hendricks et al. (2014: 404) have shown that both populations overuse pronouns, that is, they underspecify referents, but for different reasons. Whereas children seemed less able to take the perspective of their addressee, this was not the problem for elderly adults, who for example introduced new characters with more full NPs. Instead, they suggest that ‘elderly speakers have difficulty keeping track of the structure of the discourse and determining the prominence of the referents in the discourse’ (ibid.). Overspecification seems to be less problematic. Expressions with more information than required for the identification of the referent may include redundant information, or it may be the case that with more information encoded, we find, in addition to identifiability, more surprising content or more evaluative content, for example. Tourtouri et al. (2021: 31) explain that the effect of surplus information, or ‘referential redundancy’, is dependent on various factors which relate to the context of situation; ‘[a]spects of the common ground, such as the complexity of the visual scene, can both affect the speakers’ use of redundant adjectives and adjust the listeners’ interpretation

188

Child’s Play

Table 9.1 Specification values of initial referring expressions for children and adults (adapted from Uzundag and Küntay, 2018: 84)

Child participants (pretest)

Exact specification Overspecification Underspecification Misspecification Total

Adult participants

Demonstrative noun phrase condition group

Relative clause condition group

Adults (no conditions)

21% 1% 73% 5% 100%

13% 78% 8% 100%

63% 28% 6% 3% 100%

of these adjectives’ (ibid.). We saw above that a child’s ability to accurately assess common ground is something that develops over time. It is therefore not surprising that we may see instances of overspecification in children’s language as they try to work out what knowledge is shared with their interlocutors. Research into Turkish children’s specification for initial referring expressions (Uzundag and Küntay, 2018) has shown that while children are able to provide the right amount of information, they rarely overspecify the referent, and most frequently their expressions suggest underspecification. The results from Uzundag and Küntay’s (2018: 84) study are presented in Table 9.1. In this study, 59 typically developing, native Turkish-speaking children (age range was from 4 years to 5 years 9 months) took part in an experimental study to investigate whether hearing more informative and structurally more complex expressions would influence children’s referential expressions. The adult comparison group consisted of 11 university students with an average age of 24.3. The children were divided into two groups as indicated in Table 9.1, according to which NP structure they were going to be exposed to (29 children in the demonstrative noun phrase condition and 30 in the relative clause condition). For our purposes here, we are only considering the pretask data as we are interested in how much information was specified in their referring expressions and not the effect of the intervention. Therefore in Table 9.1, the specification values of initial referring expressions for the children are given separately according to the condition they would be exposed to following the pretest. At this stage, both children and adults were exposed to the same pretest items in the same order. The authors found that ‘children were usually uninformative in their referential expressions; mostly,

9.4 Under/Overspecification

189

they either just named the referent or mentioned its location with respect to distractors’ (ibid.: 88). By comparison, we find the inverse among the adults in the study since they rarely underspecified the intended referent, and although they most frequently provided an exact specification (63 per cent), they nevertheless overspecified the referent in almost a third of instances (28 per cent). To conclude this section on specification, we will consider two examples from the personal diaries of two of the authors. The first is an example of spontaneous discourse and the second from prepared writing. Both examples involve boys, Jack (4;10) and Tom (5;2). The situation in which example (141) occurs is as follows, Jack, looking a bit shifty, walks into the living room where his mother is seated. He informs her that something was an accident. The pronoun it in (141) is being used to refer to an event which resulted in ink getting on Jack’s bed. We will not try to read his mind, but if we interpret the use of the pronoun following the discussion above, the addressee, in this case his mother, would have to interpret this as continuous information since the use of it should mean the referent is recoverable. If we consider the utterance in terms of preferred argument structure, then the clause is well-formed. The pronominal expression here overspecifies the cognitive status of the intended referent, but is an underspecification of lexical information. In Givón’s (1985) terms, we might interpret this as Jack suggesting that this information is not very significant. Perhaps that is the intention; that is, something happened, but it’s OK because it was an accident. The important thing for Jack is that it is an accident and an accident is the expression encoding the most information, so we should (possibly in his view) pay attention to that. We cannot be certain of Jack’s intentions, but the use of it is atypical in the sense that he would know that it was not part of the common ground he shares with his mother, although typical if we consider that at this age, some children are still developing their ability to take the perspective of another, especially when there are conflicting pressures (e.g. no doubt he was worried he’d be in trouble). (141)

It was an accident

The next example, given in (142), is a short story written by Tom (see Figure 9.2 for an image of his original writing). This story was written at a time when Tom was fascinated by knights and castles. The reference to Caerphilly in the story refers to Caerphilly Castle in Wales. Tom is a Welsh–English bilingual child who was attending a Welsh-medium school at the time. Some of the grapheme–phoneme correspondences in his writing reflect Welsh orthography (e.g. use of ‘y’ for the indefinite article a; note that this is phonological since Welsh does not have an indefinite article).

190

Child’s Play

Figure 9.2 Tom’s story as originally written (142)

Once upon a time there was a big big tower. and in the tower there was a sleeping dragon and the dragon captured the princess. and there was a brave knight and he dared to fight the dragon and they challenged and challenged and challenged and eventually the knight defeated the dragon. and that’s how there was a legend of Caerphilly.

In the story, we can see that Tom has mastered, to a large extent, how to introduce new referents into his story and to maintain them (e.g. ‘there was a big big tower’ > ‘in the tower’; ‘there was a brave knight’ > ‘and he dared to fight the dragon’). However, the expression of interest to us here is the princess. As we have seen elsewhere, this type of expression should signal that the referent is in focus and activated. It is reasonable to assume that for Tom, given the nature of the story, there is obviously a princess (i.e. why else would the knight be there if not to rescue a princess?).3 It is clearly not the case that Tom has not reached a stable use of appropriate expressions for introducing a new referent, but here it is possible that the referent was so highly activated for him, he was unable to take a moment to check whether the reader might be able to identify the princess. In a sense, however, identification is not relevant. 3

We appreciate that the knight-castle-dragon stories are sexist but see Robert Munsch’s (1980) The Paper Bag Princess for an excellent, more interesting take on this type of story.

9.5 Creative Licence: Playing with Reference

191

It does not matter which princess it is, but it is one princess, not any princess. This example has been used with undergraduate students, who all fail to identify the princess as being new. While this is only an anecdotal perspective, it suggests that readers do not feel that they are ‘missing’ information. Having considered how the use of reference develops for children as their language skills develop, we will now turn to how adults use reference creatively, through atypical uses of reference in stories written for young children. 9.4

Creative Licence: Playing with Reference Hey! Diddle, Diddle, The cat and the fiddle, The cow jumped over the moon; The little dog laughed To see such fun, And the dish ran away with the spoon.

If we wanted to be strict with our interpretation of the simple noun phrases in the nursery rhyme cited above, we would be forced to say that their use is atypical and that the [the + Noun] expressions suggest an identifiable referent in each case. For those of us who have recited this rhyme more times than they can count with a youngster, we know that young children find it funny, but they do not ask which cat? Or which spoon? Often the rhyme is being read with accompanying illustrations, and when this is the case, the can serve exophorically to signal which cat, fiddle, cow, etc. is involved in the various actions. According to Poix (2018, §44), this traditional rhyme is ‘designed to delight children with impossible images and to develop their sense of fantasy. In fact, much of English children’s literature, including nursery rhymes and fairy tales, also belongs to the fantasy genre, which explores the realms of imaginary or magical worlds’. In this section we will briefly discuss three examples of atypical reference taken from children’s literature that illustrate how literature written for children plays with referential convention. Authors of children’s literature seem particularly adept at playing with reference in ways that work very well despite these creative instances being used atypically. In some cases, this involves type specification by lexical items that are completely novel, for example as is famously the case with authors such as Dr Seuss or Roald Dahl. Poix (2018) comments that ‘[c]oinages in the context of children’s literature range from transparent formations to opaque creations which do not necessarily have a clear extralinguistic referent’. In example (143), we are introduced to a character who is a ‘Bee-Watcher’ from Hawtch-Hawtch. Hawtch-Hawtch is clearly a nonce formation, as is ‘Bee-Watcher’, but the grammatical pattern of the noun phrase is easily described in terms of the functional account presented in

192

Child’s Play

Chapter 3. The agentive -er suffix is highly productive in English, and with the following sentence explaining that the Bee-Watcher’s job is to watch, there is no doubt about the meaning. As Poix (2018, §108) explains, ‘it is standard practice to create occasionalisms to name entities which have no existence in the real world’. (143)

Out west near Hawtch-Hawtch there’s a Hawtch-Hawtcher Bee-Watcher. His job is to watch . . .

This example comes from Dr Seuss’ Did I Ever Tell You How Lucky You Are?, which is a storybook originally published in 1973 and designed for children aged 4 and above. If we jump ahead a bit in the story, we see that another character is introduced, as shown in (144). (144)

Our old bee-watching man just isn’t bee watching as hard as he can, he ought to be watched by another Hawtch-Hawtcher! The thing that we need is a BeeWatcher-Watcher!

The first time any child heard the story of Hawtch-Hawtch, we can be fairly certain that they had never heard of a Bee-Watcher-Watcher. As we can see in example (144), this Bee-Watcher is deemed to be not watching hard enough, and so a new type of worker is needed, not a Bee-Watcher, but a Bee-WatcherWatcher. However, note that the subject of the first sentence in (144) presents a different type specification of the referent, now man and not Bee-Watcher. This NP is a more complex expression; it therefore encodes more information, despite having the discourse function of maintaining the earlier mentioned referent. We also find repetition with the use of ‘bee-watching’. As anyone who has read this story to a child can attest, it is a fun story (also philosophically deep!) and brings laughter because, like most stories by Dr Seuss, it is very silly. All silliness aside, though, the children listening to this story, or reading it for themselves, have no trouble with the new referent introduced, despite the fact that there is no class of thing in the world such as a Bee-Watcher, nor a Bee-Watcher-Watcher, or a Bee-Watcher-WatcherWatcher. Even the complex grammar of the expression is not a problem, in that there are multiple sub-classifiers, which would be bracketed as follows: (((Bee-Watcher)-Watcher)-Watcher).4 Older children will begin to notice the use of initial capital letters, which will signal proper name status, and indeed the illustrations in the book show each individual Hawtch-Hawtch citizen who has a ‘watching’ role is named by their role, that is, there is no class of ‘beewatcher’ or ‘bee-watcher-watcher’ because each role is (n-1) recursively defined (i.e. by the previous role of whatever is being watched). 4

Children’s rhymes and stories are full of this type of complexity. Famous examples include The House That Jack Built and There Was an Old Lady Who Swallowed a Fly.

9.5 Creative Licence: Playing with Reference

193

The next two examples both involve atypical uses of definite articles. We will first consider an example from Roald Dahl’s Danny the Champion of the World, a short novel first published in 1975 and listed as appropriate for ages 7 and older. The excerpt from this story which is presented in example (145) describes an encounter between Danny, the main character, and the main antagonist, Mr Victor Hazell. All the NPs introduced with a definite article seem unremarkable except for the expression indicated in bold. (145)

The big shiny silver Rolls-Royce had braked suddenly and come to a stop right alongside the filling-station. Behind the wheel I could see the enormous pink beery face of Mr Victor Hazell staring at the pheasants. I could see the mouth hanging open, the eyes bulging out of his head like toadstools and the skin of his face turning from pink to bright scarlet.

As has been well established throughout this volume, a simple NP with a definite article (i.e. the + noun) should signal a ‘uniquely identifiable’ status as if we know which mouth is being referred to. However, the moment we read the mouth, we wonder which mouth is being referred to. We reject Mr Victor Hazell’s because a typical referring expression for that would be ‘his mouth’. If we do not know whose mouth it is, we then guess it must be the pheasants’, but we reject this option also because the noun is in the singular (e.g. not their mouths). What we would like to emphasise here is that at the point of reading the, we do not know which mouth is being referred to and the reader is required to make an inference. The effect of the rather than a possessive pronoun seems to be to dehumanise Victor Hazell. In Givón’s (1985) terms, this type of highly informative NP is typically associated with less accessible referents, but in this case Mr Hazell is clearly in focus and highly accessible. We do find similar expressions in discourse, but their use does seem particularly adult-like. In Chapter 1, we briefly discussed such an example, where we suggested that this atypical use of the provided a connotation of objectifying the person being referred to. For ease of reference, we have reproduced it here as example (146). (146)

That’s not to say I’m not feeling some inflation; as a consumer I’ve begun really noticing that my money isn’t going as far as it did, say, about five years ago when the wife and I relocated to California.

Our final example comes from a more recent contribution to children’s literature. Troll and the Oliver, by Adam Stower, is a picture book published in 2013 and suggested for children 3–7 years of age. In what follows, we take the liberty of reproducing the first four lines of the story below as Excerpt 6, which appears together with an illustration (Shallow, n.d.) that accompanies the wording. The emphasis in the text has been added to illustrate the expressions of interest.

194

Child’s Play

Excerpt 6 Troll and the Oliver

This is the Troll. And this is an Oliver.

Every day around lunch time . . . . . . Troll tries to eat the Oliver. But catching an Oliver is a tricky business. No matter how hard Troll tried, he could never quite manage it. And the Oliver was never any help at all. Instead of standing nice and still, the Oliver dashed about all over the place, which made grabbing it very difficult.

We find repeated use of demonstrative pronouns in subject position (e.g. this). The expressions in bold introduce the two main characters (the Troll and an Oliver). As we can see, the demonstrative pronouns are used exophorically to direct attention to the image of each referent. If we recall from our discussion above, children in the age group for this story have just emerged from a stage where, as the literature suggests, they overuse demonstrative pronouns (Gundel and Johnson, 2013). What is a bit curious is that one referent is introduced with a definite article and the other with an indefinite article (e.g. the Troll and an Oliver). Children at this age may not notice the spelling conventions used in the text, but they certainly would know that Oliver is a boy’s name, and there is no such thing as a class of ‘Olivers’ as there is a class of boys. The chain of reference for each character in this short extract is as follows: This > the Troll > Troll > Troll > he This > an Oliver > the Oliver > an Oliver > the Oliver > the Oliver > it

9.5 Creative Licence: Playing with Reference

195

We have included an Oliver in the ‘Oliver’ chain despite the fact that its accessibility status is only type identifying. Given the repetition in the chain, we would argue that there is some ambiguity about this status. The reference chain for ‘Troll’ is much clearer. In this short opening to the story, it is clear that Troll is a he and Oliver is an it. Everyone knows (i.e. all children) that in English, it is reserved for non-human animate creatures and inanimate objects. While still operating within the referential conventions for this age group, the writer is exploiting the definite and indefinite articles and the third person pronouns for playful effect. The unusual use (and non-use) of the definite article humanises (the) Troll and dehumanises (the) Oliver, making him a unique instance of a type. This pattern of article and pronoun use continues while the story is being told from the perspective of (the) Troll. After a turn of events in the story, the perspective shifts and the story is then told from the point of view of Oliver. Reference becomes strategies become inverted, so to speak, and the author returns to the more conventional pattern of article use (i.e. the Troll, it and Oliver, he). [spoiler alert] After the final turn of events in the story, the perspective changes one last time, this time to a joint perspective, where, after a failed attempt to eat Oliver, Troll and Oliver become friends, as illustrated by example (147). (147)

Every day we sing and bake.

The atypical uses of reference described in Troll and the Oliver challenge readers to see the world from a different perspective and to mark shifts in perspective. An example of one transition is given in examples (148) and (149), which are adjacent sentences in the story but on different pages. At the turn of the page, there are no more Olivers, only Oliver, who, in (149), is referred to by a proper name, and Troll becomes the Troll. Reference to Troll continues with the pronoun it and mirrors the referential choice used for Oliver in the first part of the story. (148)

He’d had enough of pesky Olivers.

(149)

The next morning, Oliver fetched his basket and set off to go shopping as usual. He took extra care through the woods, expecting the Troll to jump out at any moment. But it didn’t.

Playing with proper names and type specification comes naturally to children. To illustrate this point, we will consider two examples from the personal diary of one of the authors in an exchange with her son on two different occasions. The first example, given in (150), is a casual conversation between Jack, age 6;4, and his mother which illustrates a developing understanding of the referential difference between a name and a type. In (151), Jack, age 7;2, uses his brother’s name as a type to reassure his mother that he is not becoming

196

Child’s Play

more like his brother. Whereas in (150), Jack is not fully aware of what the relationship is between Nutella and chocolate spread, his utterance in (151) demonstrates the creative, atypical use of reference. (150)

jack: Why is it called Nutella and not chocolate spread? katy: Why are you called Jacob and not boy? jack: Huh? katy: Why don’t we just call you boy? jack: Cos it’s rude!

(151)

Don’t worry mum I’m not turning into a Sam!

9.5

Conclusion

Children develop skill at referring from a very young age and continue to refine and develop more sophisticated referring until around 8 years of age or later, when they are able to demonstrate creative and playful uses of reference. When trying to capture how children use and develop their use of referring expressions, we find ‘a paradoxical landscape’ (Salazar-Orvig, 2019: 304). As explained by Salazar-Orvig (ibid.), ‘[m]ost studies of young children show a rather precocious competence in the use of referring expressions, whereas studies on narratives and other experimental settings tend to show that mastering the values and use of referring expressions is not achieved before age nine’. It is difficult to paint a full picture of their use of reference because evidence is patchy and there are many different approaches to the study of children’s referring strategies. In order to use reference (and language in general) creatively, in other words, consciously for a specific purpose, such as to make an inference or to hedge, children must be aware of the effect their use of language has on their interlocutor, and vice versa. They must not only be able to correctly evaluate the shared common ground, but they also need to be aware of the pragmatic inferences resulting from under- or overspecifying a referent, that is, from giving too little or too much information. According to Gundel and Johnson (2013: 55), this ability ‘is typically not fully developed until after the age of 4’. Therefore, we assume that the creative type of atypical use of reference only starts to develop in children after the age of 4. The literary examples discussed above demonstrate how well children from age 4 can recognise atypical uses and find them funny rather than confusing. Of course, we cannot be certain that they are in fact processing the articles themselves and not relying on other features of the texts (e.g. the alliteration in the Hawtch-Hawtch example), but they are clearly able to understand what the referent is doing in the story. It may be, as suggested above, that a certain level of grammatical skill is needed before they can themselves begin to produce such creative uses of reference.

10

Reference and Metonymy

When a woorde hath a proper signification of the owne, & beyng referred to an other thyng, hath an other meanyng, the Grecians cal it Metonymia. T. Wilson (1553)

10.1

A Name by Any Other . . .

This chapter is the last in the series of four chapters that makes up the volume’s section on atypical reference. Since we would not like to pick favourites, we will not say we have kept the best for last, but we will say that we have kept the most atypical reference to the end. This chapter examines uses of referential metonymy, which to paraphrase Mr Wilson’s quotation above means roughly an expression used to refer to something that differs from the signification of the words. Since, in the last chapter, we looked at the development of reference, we will illustrate metonymy with an example from the personal diary of one of the authors. The metonymic expression is highlighted in bold in example (152), where clearly the speaker, Sam, age 13;4, is engaged in the motion event of coming through, no doubt carrying a chocolate cake. (152)

Watch out - the chocolate cake’s coming through!

This example is from a young teenager, and given our account of language development in the previous chapter, we can assume Sam has already developed an adult-like use of this creative form of reference. However, use and comprehension of metonymy begins very early in language development. Falkum et al. (2017: 89) found that metonymy was productive among young children; for example, in expressions such as let’s play (the) N, where the child is suggesting a specific game but rather than using the name of the game, they pick a salient feature of the game, as in let’s play zebras, a game where zebras are salient. The authors suggest (ibid.) that ‘[m]etonymy, then, could well serve early on as a productive referential strategy for young children’. Experimental studies with young children have provided evidence that children from as young as 3 years of age can both understand and produce metonymy. Köder and Falkum (2020: 202) use eye-tracking data to show that 197

198

Reference and Metonymy

gaze provides evidence of ‘an early sensitivity to metonymy already from the age of 3 and a general improvement with age’. As we will see, there are different types of metonymy, and there is still considerable debate about exactly how it works. We also find extensive overlap between metonymy and other regular features of language, such as noun–noun compounds (e.g. screwdriver, as an action for instrument metonymy) and noun to verb conversions (or transcategorisations), such as the use of hammer as a verb, which many consider to be metonymically motivated (see Davies, 2004; Bauer, 2018). In fact, there could be an argument made for viewing the Dr Seuss example we discussed in Chapter 9 as metonymy, repeated here as example (153). On the surface, this is a simple NP, but what it expresses is a person from Hawtch-Hawtch who watches bees. The -er nominal watcher is metonymic (see Panther and Thornburg (2009) on the metonymy of -er nominals), and the compound is metonymic as well. While we will not debate this issue, we agree with Bowerman (2019: 40), who explains that ‘there also appears to be a common conceptual principle underlying young children’s lexical innovations. Zero derivations, in particular, exploit the kind of salient associative relations of (often part-whole) contiguity between entities which also underpin referential metonymy’. (153)

a Hawtch-Hawtcher Bee-Watcher

Metonymy is ubiquitous, and we do not have to look very far to find examples. It does, however, have slippery boundaries, and if you push the concept far enough, sometimes it feels as though every instance of language is metonymic to some extent. In this chapter, we ask the question about where we should draw the line around what constitutes reference through our exploration of referential metonymy. In previous chapters, we have emphasised the importance of type specification, for example, but what happens to this aspect of reference when the lexical expression does not specify the type of referent but rather some feature or property related to it? In what follows, we discuss what we mean by referential metonymy and why we consider it among our examples of atypical reference. With referential metonymy, the criteria for referring cannot be found solely in the referring expression itself, especially given that the lexical noun in some instances of referential metonymy cannot be said to classify the referent (e.g. a teenager and a chocolate cake are not the same kind of thing). In Section 10.3, we will discuss examples of metonymy from medical discourse contexts, including nurse handover discourse and medical students’ self-reflections of their clinical experiences. Based on the discussions in Sections 10.2 and 10.3, we situate metonymy as a type of reference in Section 10.4, that is, metonymic reference, and consider what it is about this type of metonymy that justifies thinking about it as atypical reference. The chapter ends with brief concluding remarks in Section 10.5.

10.2 Metonymy, of the Referential Kind

10.2

199

Metonymy, of the Referential Kind

It has become somewhat of a custom in the literature on metonymy to begin with the well-known example, the ham sandwich is waiting for his check, which is attributed to Nunberg (1979), also discussed by Lakoff and Johnson (1980) and many others. While we know that ham sandwich specifies a type of food (a subclassification of sandwich), we also know that it is not ‘the ham sandwich’ that is waiting to pay; instead, the expression is being used by the speaker to refer to a person, presumably someone who has just eaten a ham sandwich. The use of his is co-referential with this person, so we know that this person is male, but nothing more. This type of metonymy is termed referential metonymy because the expression is a referring expression, and it uses metonymy for the reference. In this sense, it is a referential choice. Metonymy is often characterised as an ad hoc description which carries an associative or relational meaning with respect to the intended discourse referent. While relatively more work has been done on metonymy from a cognitive perspective (e.g. Langacker, 1999; Croft, 2002), very little work has considered its textual role or its place in ongoing discourse (although see Al-Sharafi, 2004; Denroche, 2018). Closer investigation into the literature on metonymy opens up a can of worms, where all of a sudden the phenomenon’s defining features are brought into question, where the lines between metonymy and metaphor are blurred, where different types of metonymy overlap, and an array of different terminology causes havoc in an already confusing storm of arguments and counterarguments. As an illustration, we will consider the three definitions presented as examples (154) to (156). The common theme across these three definitions is indirect reference. (154)

‘[Metonymy is] a phenomenon of indirect reference in which a linguistic sign refers not to its default referent Ri, but to another referent Rj.’ (Hilpert, 2006: 126)

(155)

‘Metonymy thus involves speaking about a salient reference point which allows us to access another conceptual entity, the target.’ (Radden and Dirven, 2007: 14)

(156)

‘[M]etonymy is seen as indirect reference, or reference shift, in which a linguistic sign refers not [only] to its default concept A, but to another concept B, within a single semantic field.’ (Ädel, 2014: 74)

Let us return to the the ham sandwich example. This example has been discussed everywhere1 in the literature on metonymy. For our purposes here,

1

See Littlemore (2015) for a useful discussion of the distinction between metonymy and hyperbole.

200

Reference and Metonymy

we will draw on the explanation offered by Hurford et al. (2007: 339), which they give as follows: In the particular context in which this sentence would be uttered, presumably a café or restaurant, the person uttering the sentence would know that there was a close relationship between the thing ordered and the person who ordered it. Because this relationship is so obvious in the context, it is permissible to refer to the person by what [they] ordered. As a matter of fact, this might be the preferred way of referring to the person, because the people who wait on customers in a diner typically don’t learn the names of their customers, but they are well aware of what their customers ordered.

Interestingly, whenever this example is considered in the literature, it seems that the addressee is not considered, but as we know from our discussion of reference in this volume, the addressee plays a critical role in any act of referring. Perhaps we can add to the detail of the context and imagine who the speaker might have been speaking to. In a busy restaurant setting, we could easily assume that not only is the referent impatient, but the speaker is probably getting impatient as well. The speaker could be the person who writes the check and is waiting for a server to give it to the customer, or perhaps the speaker is the server who is waiting for the cashier to write the check. It might not matter too much who in the restaurant context is the speaker. We also do not know why the statement has been said. Is it a statement which simply gives information to the addressee or is it a request for action? We have highlighted the need for a model of context in any approach to reference, especially in Chapter 5. What Hurford et al. (2007) are saying is that context is one of the principal motivating factors in the use of referential metonymy. As we saw in Chapter 5, Halliday views language as embedded within context, both situational and cultural. Martin (1992: 121) explains these two aspects of context as follows: context of situation includes ‘relevant information that can be perceived (seen, heard, felt, tasted, smelled), including text’, whereas context of culture concerns ‘relevant information which cannot be perceived, but which can be assumed because of shared knowledge among interlocutors deriving from their membership in some definable community’. In the ‘ham sandwich’ example, we can infer, as Hurford et al. (2007) seem to do, that the identification of the intended referent is anchored in the context of situation, that is, the person who is waiting to pay is visible to the addressee. However, this is not always the case. The use of metonymy in example (157) comes from an online forum where participants can ask for advice on employment law. The metonymic expression is highlighted in bold. (157)

2

hey i am working in a restaurant and recently one of my tables left without paying. I was given a warning but told that next time the difference will be taken from my wages. I know that other employees have to pay for walkouts regularly. Is this legal? (Sketchengine enTenTen152)

For details of SketchEngine, an online resource for research using corpus linguistics, see Kilgarriff et al. (2014).

10.2 Metonymy, of the Referential Kind

201

It is worth noting that had the noun customer been used instead of table, it would not be an instance of metonymy. As Lakoff (1987: 78) explains, metonymy can be thought of as ‘a “stands-for” relation’. Barcelona (2019: 171) provides a broad but very useful definition, highlighting the importance of a pragmatic function by viewing metonymy as ‘an asymmetric mapping of a conceptual entity, the source, onto another conceptual entity, the target. Source and target are in the same frame and their roles are linked by a pragmatic function, so that the target is mentally activated’. Metonymy is not strictly a feature of language but rather an ‘essentially conceptual process, which is reflected in various types of semiotic modes, particularly human language (both oral and sign language), but also gesture, art (music, painting, sculpture), etc.’ (Barcelona, 2019: 168). This conceptual aspect is found throughout discussions of metonymy. Slabakova et al. (2016: 176) describe metonymy as ‘a well-established mental process, whereby the mention of some entity (activity, person, thing, time period, etc.) is interpreted to stand for a related entity’. This can be seen in example (157), where it is clear that the speaker is not specifying a particular table as the intended referent but rather a person, that is, a customer, at a specific restaurant. It should also be clear from this example that, as noted by Bowerman (2016: 17), this type of metonymy cannot be seen as a neologism. In other words, we do not have a new lexeme table that denotes a kind of human (as we find in, for example, the use of board to mean ‘food is provided’ rather than ‘plank of wood’). The use of walkouts in example (157)3 is, however, best explained in these terms; even though it is also a metonymy (similar to Bee-Watcher), it has become established as a lexeme in the English language. With walkout, we have a lexeme which is intended to denote a class of humans that leaves a restaurant without paying. While our discussion of metonymy is restricted to its use in referring expressions, it is worth noting that there are other types of metonymy which are not referential, such as instrument for action metonymy (e.g. everytime I try to hammer a nail. . . I hit my finger instead).4 In addition to referential metonymy, Littlemore and Tagg (2018) include propositional metonymy, illocutionary metonymy, and situational metonymy; however, space does not permit a discussion of these different types. For a useful overview of various ways in which metonymies can be classified by type, see Ruiz de Mendoza Ibáñez and Díez Velasco (2003). The definition of referential metonymy that we adopt is from Bowerman (2016: 2), who states that it involves ‘a speaker’s intentional use of an expression to refer to an entity that does not fall under the literal denotation of the expression’. As shown in example (157), the denotation used is that of table, but the entity referred to by the speaker is not a

3 4

This example is given in original orthography. Example from SketchEngine enTenTen15 (Kilgarriff et al., 2014).

202

Reference and Metonymy

member of that class. Referential metonymy, therefore, involves the use of metonymy in a referring expression. How this differs from metaphor will be discussed later in this section. The role of context of situation, which as we saw in Chapter 5 includes the material setting, is very important since it provides a source for the metonymy. In example (157), the context involves a restaurant with staff serving food to a customer at a table. The intended referent is the anonymous person who sat at a table and who ordered and ate food at the table. This situation is very similar to the ‘ham sandwich’ example discussed above, with the following two exceptions: at the point of utterance, the speaker and addressee are not in the restaurant and the NP the ham sandwich is definite but one of my tables is not. As Bowerman explains (2016: 13), ‘the referent in the “ham sandwich” case belongs to the class of restaurant customers; for all members of that class, their food order is (in the restaurant context) a useful identifying property’. In contrast, the use of table in (157) is not, that is, every customer eating in a restaurant is seated at a table. Therefore, context, either situational or cultural, provides regular and efficient sources for the purposes of referring; however, the options here tend to be in contrast to the wider, or typical cultural classification of the entity (i.e. table or sandwich rather than person). Metonymy is generally seen as efficient, something like a shortcut, since the referring process is being facilitated rather than being inhibited. A very basic addressee model (cf. Wilkes-Gibbs and Clark, 1992) must play a role here, and the speaker’s anticipated knowledge about what information the addressee has access to is important. In example (157), had the speaker used a typical denotation such as person, man, woman, or customer, for example, recently a person I served left without paying, the addressee would have managed to ratify the referent without difficulty, that is, an unidentifiable person did not pay. However, the main issue in the text is about not being paid and who is responsible. It is not about that one person. Furthermore, given that this happens regularly, the concern is that it might happen again. The use of metonymy here has multiple advantages, textually speaking. It allows the speaker to use an expression that brings certain features, aspects, or information into focus (cf. Warren’s (2004) work on the focussing role of metonymy). This view would suggest that the use of metonymy serves a textual function. An important question, then, is what feature or property should be selected? Clearly, in the text given in example (157), using the food item eaten would not have worked, as shown in invented examples (158) to (160). (158)

i am working in a restaurant and recently the ham sandwich left without paying.

(159)

i am working in a restaurant and recently one of my ham sandwiches left without paying.

10.2 Metonymy, of the Referential Kind (160)

203

i am working in a restaurant and recently a ham sandwich left without paying.

The interesting aspect of using table rather than anything else is that it makes salient the set of potential referents and allows the speaker to select an indefinite instance from that set (cf. reference mass, e.g. Radden, 2009). Bowerman (2016: 4) highlights this point as follows: [T]he inherent salience of a property does not guarantee its usefulness in every situation. For example, if the situation demands economical-efficient identification of a target entity, the most relevant property of the entity will be one which, regardless of whether or not it is inherently salient, is individuating, i.e. in the context, uniquely picks out the entity without introducing any referential ambiguity which would require extra processing effort to resolve.

Therefore, the fundamental difference between the ‘ham sandwich’ example and example (157) is that the referring expression, the ham sandwich, is used by the speaker to be uniquely referring (i.e. specific and identifiable), whereas the expression one of my tables is not. We will return to this point in the discussion below. Metonymies involving proper names differ from noun-based metonymic reference. To illustrate some of these differences, we will consider examples of the use of the proper name Canada. Typically, the use of Canada as a proper name would be expected to refer to the land mass, which is geographically and politically identifiable. Proper names are, typically, used by speakers to uniquely refer, that is, to refer to a uniquely identifiable referent. However, Dancygier (2009: 168) presents a convincing case for the rich contextual frames represented by proper names. In examples (161) to (163),5 the highlighted instance of Canada is being used to refer to different referents in each case. In example (161), it cannot be the case that Canada, as a land mass, signed the deal, since only people can sign a deal. This type of metonymy is different from what was discussed above because it does not involve a shift in type (i.e. it involves a proper name) but rather in type of referring expression (i.e. proper name rather than lexical NP). Examples (162) and (163) are very similar in this regard. The same proper name is used, but again, the discourse referent is different in each case. However, what is shared in these cases is an opening up of the referential potential. As explained by Cislaru (2007: 108), ‘metonymy more than doubles the number of possible referential and topic developments’. In example (161), we can see that both spatial reference and institutional meanings are accessible. In example (162), we also find both place and institution construed. For Cislaru (2007: 102) the use of place-name metonymy can be seen as ‘a cognitive and pragmatic tool, putting reference

5

Recall that the symbol ø indicates zero anaphor, as seen in previous chapters.

204

Reference and Metonymy

at the service of the topic’. We also find multiple meanings construed in example (163), with place and team meanings but not institutional. The team semantics are reinforced textually by the anaphoric use of them later in the sentence. (161)

With little time left ahead of a deadline to agree to a renewed NAFTA trade deal, Canadian and US officials on Sunday tried to settle differences on tough issues such as protection against US tariffs. The Trump administration said Canada must sign onto the text of the updated North American Free Trade Agreement by midnight ET on Sunday or ø face exclusion from the pact, which includes Mexico. (The Guardian, 30 September 2018)

(162)

Just four months ago in Quebec, a runaway train carrying 72 cars of crude oil killed 47 people and demolished the town of Lac-Megantic. . . . In the aftermath of this accident, Canada issued an emergency directive requiring a minimum of two-person crews for trains carrying hazardous materials. (Transportation Trades Department 29 October 2013)

(163)

But in the 19 Olympic and World Championship final games since womens (sic) hockey arrived on the international stage in 1999, it has been Canada against the United States 18 times. Canada begins by ø facing Switzerland on Saturday at 8:00 a.m. in their first game of the tournament – watch it live on CBC TV and streaming at cbc.ca/olympics. The mere fact Canada has won the past three of four gold medals at the Olympics – ø losing only in Nagano in 1998 – would make them the favorites in the eyes of many. (SketchEngine, enTenTen15)

Hanks’ (2013) approach to lexical semantics may be relevant here. A word does not have a meaning; it has meaning potential. The semantic features associated with a given lexeme constitute its meaning potential (i.e. a network of various semantic features, including properties). Any instance of a word will activate some, but unlikely all, of the meaning potential. For Hanks (ibid.: 73), meaning is best viewed as a dynamic event, that is, ‘dynamic events that involve cooperation between speakers and hearers or writers and readers, rather than static, finitely bounded abstract objects’. As we saw in Chapter 2, proper names are lexemes like any other, and in this sense, even they have meaning potential. If we adopt Dancygier’s (2009) use of contextual frames, then knowing a proper name means knowing what features and/or properties are associated with it within the contextual frame. No doubt the same can be said for all lexemes. For example, the verbs watch and see may share some features (e.g. visual perception), but not others (e.g. durativity).6 For Hanks (2000: 215) ‘[a] word’s unique contribution is some combination of the 6

See Chrispin and Fontaine (2023) for a detailed discussion of the semantic profile of these two verbs.

10.2 Metonymy, of the Referential Kind

205

components that make up its meaning potential, activated by contextual triggers. Components that are not triggered do not even enter the lists in the hypothetical disambiguation tournament. They do not even get started, because the context has already set a semantic frame into which only certain components will fit’. This perspective could be extended to the kinds of words we find in metonymic expressions. With this in mind, we will complete this overview of referential metonymy by considering the features that distinguish it from metaphor. 10.2.1

Metonymy and Metaphor

There is an important distinction to be made between metonymy and metaphor. As Peirsman and Geeraerts (2006: 271) explain, metonymy is seen as ‘a shift within one domain; metaphor is a shift across domains’. While clearly there is a shift with referential metonymy, there is no transfer of domain. As Barnden (2010: 12) explains, ‘[i]t might be claimed that in metaphorical similarity there is no real source-side entity corresponding to the target-side entities, whereas in metonymy there is. For example, metaphorically casting a person Richard as a lion does not involve a particular, real lion, whereas metonymically referring to some real artworks via an artist does involve the artist being real as well’. In other words, in the examples we have been considering, there really is a ham sandwich, a table, and a country named Canada. In the instances we have been discussing, intra-domain shift seems to be an important feature. We might say that it is dependent on it and that it is very much anchored in its context of use. The use of the concept of domain to distinguish between metonymy and metaphor is not as straightforward as it may seem (Panther, 2006; Dancygier, 2009). The question seems to rely on definitions of source and target domains and whether these constitute truly distinct domains (or not); see, for example, Peirsman and Geeraerts (2006: 271). According to Taylor (2002: 196–197), we cannot assume domains are mutually distinct; domains typically overlap and interact in numerous and complex ways. Various proposals have been put forward to account for the differences, but for our purposes we will simply consider what is meant by the concept of contiguity and whether it offers any clarity in distinguishing between metonymy and metaphor. As we saw above, Barnden (2010) differentiates metaphor and metonymy based on similarity as opposed to what we might call actuality. In other words, ‘metaphor is a matter of similarity between source and target items, and metonymy a matter of contiguity between them’ (ibid.: 6). The account of contiguity proposed by Peirsman and Geeraerts (2006: 273) follows Dirven (2002: 88) in preferring the concept of contiguity over the notion of domain

206

Reference and Metonymy

since, as they argue, the relations underpinning metonymy are not objective but ‘the result of a construal by a culturally-conditioned language user’. For Dirven (2002.: 91), ‘contiguity must be taken to mean “conceptual contiguity”’, which means that referential metonymy should work even if the addressee does not share the same conceptual contiguity as the speaker. For Barnden (2010: 10), ‘contiguity is itself to some extent partly in the eye of the beholder’. Perhaps the fact that metonymy is ‘a mentally, socially and culturally constituted matter’ is why it is so well suited to acts of referring. 10.3

Metonymic Expressions in Healthcare

Based on our discussion above, we have seen that metonymy provides an efficient, economical, and effective means of referring that is used in all contexts, and despite its apparent complexity, it is a linguistic resource that comes naturally to young children (see Falkum et al., 2017). Its use becomes particularly handy in busy clinical settings. If we consider example (164), taken from Al-Sharafi (2000: 162), we can imagine how easy it would be to understand the multiple metonymic expressions in this brief dialogue between two nurses. Without a bit more information, we might assume the first nurse is looking at someone and asking who the person is, but in fact this nurse is holding a report. The demonstrative pronoun this is clearly referring to the patient who is the subject of the report and not the report itself (compare with what’s this?, which would be a question about the report or document). In the response the second nurse, repeating the demonstrative pronoun, identifies the patient as the hernia, and the follow-up mention to this referent uses the anaphoric pronoun he. As Al-Sharafi (ibid.) explains, ‘“this” has an exophoric reference and its reference is understood only via metonymic reasoning’. (164)

n1: n2:

Who’s this? This is the hernia. You know he has to be seen by the doctor today.

Given that this is used simultaneously to refer to the report and the patient, there is a kind of semiotic doubling. Halliday (2006: 99) claims that with certain types of expression,7 ‘the grammar is creating virtual phenomena, phenomena which exist purely on the semiotic plane. This is achieved by a process of semantic junction, whereby two category meanings combine’. We would not like to suggest that the blend of meanings we are discussing with metonymy are categorical, but perhaps, as suggested by Fontaine (2019: 220), the referent is a virtual one, ‘existing on the semiotic plane and being realized

7

Here Halliday is discussing his concept of ‘grammatical metaphor’, which could be better interpreted by the concept of metonymy as there are considerable similarities.

10.3 Metonymic Expressions in Healthcare

207

in textually motivated ways’, which allows, as Littlemore (2015: 105) explains, ‘the speaker to say two things at once’. In this section, we will discuss the use of metonymic reference in healthcare settings. The first set of examples comes from nurse shift-change handover meetings in a hospital setting. Following this we will consider how student doctors talk about using metonymy to refer to their patients in reflective essays following a clinical placement. 10.3.1

Nurse Shift-Change Handover Discourse

Nurse-to-nurse shift handover communication consists of a primarily informationgiving meeting where the departing head nurse provides incoming nurses with important patient information at the beginning of their shift. This information exchange forms a crucial part of patient care since the handover (or handoff ) is where nurses get important details of each patient that will be in their care. Nurse handover discourse is characterised by language that is cryptic to outsiders and full of medical jargon and abbreviations (Athanasakis, 2013: 307). These handover sessions are important to patient care and represent an understudied area of the healthcare profession, as noted by the American College of Obstetricians and Gynecologists (2012): Patient handoffs are a necessary component of current medical care. Accurate communication of information about a patient from one member of the health care team to another is a critical element of patient care and safety; it is also one of the least studied and taught elements of daily patient care. One of the leading causes of medical errors is a breakdown in communication.

Referential metonymy has the potential to both facilitate the accurate communication of patient information and impede effective patient support. One of the potential negative consequences of uses of metonymy is associated to stigma and negative connotations, as we will see below when we consider diary entries from new doctors. However, ‘[m]etonymic language can be used to promote economic information exchanges, particularly valuable when a clinical environment is busy. Nursing language richly illustrates the variety of attributes that can be selected by speakers for attention’ (Moors, 2015: 59). Despite the serious nature of the handover sessions, they also offer nurses the opportunity to socialise and talk about things they are doing outside of work (Bangerter et al., 2011: 211), and therefore nurse handover discourse sometimes also includes informal conversational discourse about personal matters. As noted above, handover discourse can be quite opaque to those not involved in this discourse, as illustrated in (165), taken from Bartlett et al. (2020: 404). Note that [FNLN] is used to replace a patient’s name (first name last name) and an F or M following FNLN indicates male or female patient.

208

Reference and Metonymy

As we can see in this example, the handover discourse follows a template as the head nurse runs through the status of the patients. There is a set list of items that must be covered (Spilioti et al., 2019), and as we see here, the organisation of the discourse is focussed on identifying patients by their location, and this is done within the larger organising unit of type of medical issue. Patients are identified at first as D5, D6, and C5, which are the patients’ respective locations, and also by name to identify patients who are not for resuscitation (‘Not For Resus’). For patients at risk of falling, the specific patients are only referred to by bed location. One patient within this category is identified by name along with his location (gentleman on trolley 3 [FNLNM]) since important information about his circumstances is given to the nurses who will be caring for him. There are multiple instances of metonymy throughout this brief example, including referential metonymy (e.g. at risk of falls A bay bed 3) and predicative metonymy, which was briefly mentioned above (e.g. we’ve had no cardiac arrests within the last 12 hours). This example provides a clear sense of the type of language used in nurse shift-change handover discourse when referencing patients. (165)

we’ve had no cardiac arrests within the last 12 hours. Not For Resus D5 [FNLNM]. D6 [FNLNM]. C5 [FNLNF]. No falls. at risk of falls A bay bed 3 and 4, B bay bed 3 and 4, C bay 1 3 and 5 and all of D bay. Trolleys 1 2 3 and 5, 7 and 13, gentleman on trolley 3 [FNLNM], he’s had a POVA initiated against his lodger. His lodger lives with him and takes care of his finances

The use of location for person metonymy in nurse handover discourse is common, as shown, for example, in (166) with the use of D1 and D4, each expressing a location, to refer to two patients. It is worth noting that the use of BMs is also metonymic since BM is an initialism for ‘Boehringer Mannheim, a German pharmaceutical company that used to make by far the most commonly used test strips for blood glucose’ (Grant, 2009: 339). When a literal denotation is used for type specification, as in our lady in B1 that had a fall yesterday, location is given along with relevant information about the patient’s medical condition through the use of qualifiers (in B1 and that had a fall yesterday). Referring to patients through their location allows nursing staff to identify medical conditions that nurses starting their shift would need to be aware of, such as a particular concern or treatment. Example (165) above provides a striking example of this where we see patients listed simply by bed for standard areas requiring attention (e.g. ‘at risk of falls’). (166)

D1 and D4 have both had high BMs. so that’s a [FNLNM] and a [FNLNM]. and our lady in B1 that had a fall yesterday, urm, she’s been a little bit drowsy today] (Lloyd et al., 2021: 228)

10.3 Metonymic Expressions in Healthcare

209

Although to date no study has examined metonymic reference in nurse handover data specifically, in the examples reported in the literature we have been discussing, there appears to be a low frequency of illness for person metonymies. The reason for this is likely because these examples are from an assessment ward, so any diagnosis may not yet be known. However, one such instance is given in example (167) (from Lloyd et al., 2021: 228–229), with the metonymic reference highlighted in bold. What is interesting in this example is that this patient is just arriving and so does not yet have a location. One patient (male) is being moved out but is still under the care of the nurses while awaiting transfer, and a new patient (female) is arriving but has not yet been put in a stable location. The male patient is identified by location (on trolley 10) and named (‘[FNLNM]’). The instruction given to the nurses is to move him to another trolley and to move the new patient to trolley 10. There is some uncertainty about which patient is being discussed, and to clarify, the patient’s medical treatment is used to uniquely identify the patient (the dialysis). (167)

nurse 1:

nurse 3: nurse 1:

We’ve got a gentleman on trolley 10. He’s going to go down to UHW now, [FNLNM]. They just brought somebody over now. Put him on the extra trolley for now to do the move and then we’ll get her on to number 10. Is that alright? Yeh. Is that the one that was just coming through the door? the dialysis oh yeh that one that’s just come through the door.

The dominant use of patient location to refer to patients ensures that patients will be identified accurately, which is crucial in hospital settings where there is no room for error. Nurse handover discourse is a high-risk communicative context where precision is key. According to Raeisi et al. (2019: 1), Failure of effective handover is a major preventable cause of patient harm, which is also the most important step in ensuring the patient’s safe handover. Patient handover is in fact a valuable affair and an essential part of processes and workflows in hospitals. In other words, one of the most important steps in ensuring the continuity of care for [patients] is the transfer of professional responsibility and accountability for some or all aspects of patient care or a group of patients to an individual or a professional group temporarily or permanently.

What we can see from these examples is that patient location is a valuable referential aid in nurse handover discourse. Although we have only discussed a selection of examples, it seems clear that no information about a patient is given without first establishing where that patient is and then indicating any additional information if that is deemed essential for medical reasons or for patient well-being. However, metonymy is so commonplace, it gets used without notice, and it appears to be used and understood with incredible ease. However, while ‘metonymy can pass unnoticed in clinical practice’ (Moors, 2015: 58),

210

Reference and Metonymy

literature on nursing discourse rarely includes mention of the term.8 It is difficult to get access to this type of data, and as was mentioned above, nurse handover discourse is unfortunately understudied and insufficiently taught, which is surprising since it is an essential part of daily practice and every nurse in a hospital setting will take part in this important exchange of information. What we will consider now are staff attitudes about the use of metonymy and how it potentially affects their feelings about their patients. 10.3.2

Staff Attitudes towards Metonymy

While the use of referential metonymy is efficient, there is often a distancing effect which takes the human, animate person out of focus. Camp et al. (2020: 24) note that labelling, including uses of metonymy can influence social distance, in particular ‘the desire of individuals to separate themselves from specific individuals or a specific group’. Moors (2015: 60) explains how language on the ward in psychiatric discourse can influence staff attitudes and potentially the care that patients receive. For example, he notes (ibid.) instances of patients diagnosed with a personality disorder being referred to as ‘a P.D.’ (pronounced /pi:di:/): ‘[a] frequent use of the term is “Just a P.D.”, with the term “just” highlighting the assessment that the person has no attributes beyond the putative diagnosis’ (ibid.). Language use in medical and healthcare contexts can lead to the creation of stigma, ‘including the creation of “in” and “out” groups, the “otherization” of “out” individuals, and the application of labels and stereotypes to reinforce “otherization” and marginalization’ (Camp et al., 2020: 23). In a study of reflective essays written by medical students, Camp et al. (2020) considered how doctors in training9 refer to patients and what, if any, labels they use. A significant finding was that students’ use of metonymy could flag an important teaching opportunity for aspects of professional development that may otherwise be missed. In addition, they noted that ‘the use of metonymy was statistically associated with moral distress’ (ibid.: 31). Unlike in the nurse handover discourse, the uses of metonymy in the reflective essays in the Camp et al. (2020) study are not related to the pressures of handover discourse and the need to uniquely identify a patient. It is unsurprising then that there were relatively infrequent instances of metonymy

8

9

Use of this type of metonymy should be avoided if the patient and/or family can hear as it is disempowering and will de-personalise the patient. We are grateful to Michelle Aldridge for discussions of this point. The students in this study were studying at the University of Texas Southwestern Medical Center at Dallas.

10.3 Metonymic Expressions in Healthcare

211

throughout the essays. There was one notable exception in relation to patients with substance use disorders and patients who were seen as ‘ethically or interpersonally difficult’ (ibid.: 31). It is worth noting, as the authors point out, that these findings were based on one class of medical students, and, therefore, they may not be reflective of student doctors more generally or of the illnesses that were mentioned, since this would depend on the individual clinical clerkship of each student. In total, only sixty instances of metonymy were identified in the essays.10 Metonymy use was categorised into thematic clusters, based on what the metonymic reference was being used to describe (for details, see Camp et al., 2020: 25). Results from this study showed the following distribution by thematic category: substance abuse (N = 27, e.g. ‘addict’), illness (N = 4; e.g. ‘diabetic’), body part (N = 4), clinical status (N = 6; ‘new admit’), reproductive health (N = 5; ‘primagravida’), challenging clinical situations (N = 6; ‘malingerer’), and other thoughts on patients as people (N = 3), which is illustrated in example (168). (168)

I think that everyone comes to medical school thinking how they will be different and think of the patient first and say, “I won’t just see them as a disease.” I think the truth is that you may think of the patients well being [sic] first, but you will always see the disease. (Camp et al., 2020: 27)

There is a clear sense in (168) that medical students are concerned about how they will see their patient and that a person-first approach will be challenging. We can also glean from this that if as a doctor you ‘see’ the disease, then, as the most salient feature (to the medical professional), there must be a kind of cognitive battle to suppress the ease with which metonymy is used. A clearer sense of this feeling is captured well by example (169). (169)

All morning, I was immensely thrilled about “the colpo” as everyone called it . . . but as I walked into the room, my heart sank. “The colpo” was not just a procedure but actually a 38 year old Hispanic female in worn out jeans and a baggy t-shirt. Sitting alone in the corner hugging herself, the patient was shivering in the frigid exam-room. She slowly lifted her head and as I looked into her eyes, the implications of the colposcopy hit me like a punch in the stomach. I was immediately ashamed at how easily—only 3 weeks in clinic—I had fallen prey to dehumanizing patients. (Camp et al., 2020: 29)

The medical student seems to have interpreted ‘the colpo’ (the colposcopy) as a procedure, that is, something that needs to be done. However, the realisation that

10

A total of 802 essays were analysed, and only 46 included instances of metonymy. See Camp et al. (2020) for details.

212

Reference and Metonymy

this referred to the person who had to undergo the procedure came as quite a shock, and this student was taken by surprise at how quickly this had happened, in part, no doubt, due to the use of metonymy. This type of metonymy is much closer to the neologisms discussed above (e.g. walkouts), where there is, in addition to the use of metonymy in the word formation process, a sense of classification, that is, a type specification. In other words, there is now a class of person which shares certain properties. While colpo is not a word that would be recognised in general English discourse, English has many such lexemes which would be (e.g. breadwinner). Camp et al. (2020: 31) explain this as follows: Each use of metonymy in the challenging clinical situations category was used by someone other than the student. In most cases (“a talker,” “fatigue problem,” “malingerer”), the metonymy happens before the student actually meets the patient, and these references fuel the students’ anxieties about who they might encounter during that clinical visit. The student later gets to know the patient as a person, and this becomes a rewarding encounter. In this sense, these essays are similar to “Room 13” and “the colpo” in that metonymy was used to connote an anonymous idea of a stereotypical patient and that this develops into a conceptualisation of a real person once the student has an actual encounter with the patient.

From this we can see that metonymy is not always the best referential choice for all contexts of use. While it seems highly effective and efficient in nurse handover discourse, repeated use in other contexts may lead to establishing damaging effects due to its role in stigmatisation. Camp et al. (2020: 24) note that, in relation to describing individuals with substance use disorder, ‘the Office of National Drug Control Policy issued a memorandum in 2017 encouraging the use of person-first language’. There will, therefore, be contexts where the use of metonymy needs to be discouraged or avoided entirely. 10.4

Metonymic Reference

Based on the discussion in the previous section, we see that referential metonymy is prevalent in healthcare settings. Its frequency, purpose, and appropriateness vary considerably from context to context. We have, to this point, more or less assumed that referential metonymy is indeed referential. As is often the case when taking a theoretical position, there is room for debate, but our interest in this volume has been squarely focussed on referring expressions, which is why we prefer the term metonymic reference rather than referential metonymy. We see metonymy as a type of referential choice, and whether or not it is successful in an act of referring is governed by the same factors that we have been discussing throughout this volume. In this section, we will discuss the similarities of reference and referential metonymy in order to explicitly make the case for this phenomenon to be seen as a type of reference rather than a type of metonymy.

10.4 Metonymic Reference

213

Most literature on metonymy comes from the fields of semantics and/or pragmatics, although increasingly from areas of linguistics which are more concerned with lexicogrammar, including using methods from corpus linguistics (cf. Littlemore and Tagg, 2018). Work on reference, as was mentioned in Chapter 1, has a long tradition in philosophy and semantics. Very few studies of both topics have taken an integrated approach, drawing on functional (usage-based) approaches to language, including cognitive and text-based perspectives, within a framework that includes a model of context. As will be discussed in this section, there are key considerations which underpin metonymy and which overlap with those for reference. According to Jiang (2013: 1), ‘[w]hat metonymy involves is not about the meaning of words, but about what people do with the words to refer to an intended object or person’. This position aligns very much with our perspective. Our interest is in acts of referring and, in this chapter, we have been considering how metonymy fits in with that perspective. We have stressed throughout this volume the role of the speaker’s intention to refer to a referent, the role of shared knowledge, and the role of the parameters of context. Therefore, a position on metonymy that aligns with these aspects of reference is attractive. However, Jiang (ibid.) also specifies certain constraints on referential uses of metonymy, which we will consider now. Jiang (2013) discusses three cognitive constraints on the use of metonymy; these are salience, mutuality, and effects and effort, which we will consider in turn. Salience refers to how speakers perceive the world and how, based on our perception, we might give some concepts preference over others. ‘[T]he choice of metonymic expressions is not randomly made, but it is linked to the way in which people perceive and conceptualize the world. Among the factors that can make an entity suitable to serve as a metonymic reference point are the cognitive principles of relative salience’ (ibid.: 7). Similar to our discussion of meaning potential above, Jiang (2013: 9) argues that ‘metonymic expressions usually exploit particularly salient relations (including salient properties) which can be readily used for their identification’. Consideration of semantic features and referent properties is an inherent part of referring, as we have seen in earlier chapters. The second constraint proposed by Jiang (2013), mutuality, involves the role of context and shared knowledge. As he explains, ‘a large set of assumptions can be economically communicated without its total content being linguistically encoded, since the hearer may use manifest (or mutually manifest) contextual assumptions in deriving the intended message’ (ibid.: 9). We have been discussing this aspect of reference in terms of common ground or shared information. Finally, the third cognitive constraint, which Jiang (2013: 10) calls ‘effects and effort’, relates to the trade-off between processing effort and cognitive

214

Reference and Metonymy

effect. This constraint is somewhat similar to the principle of quantity (Givòn, 1985) which was discussed in Chapter 9, although there are some differences. Jiang (2013: 10) claims that ‘the human cognitive system automatically aims at maximizing relevance, or deriving the greatest cognitive effects possible for the smallest possible effort’. He uses the examples presented here as (170) and (171) to illustrate this point. The cognitive effort, according to Jiang (ibid.), is not the same in processing money as compared to a sheep, that is, ‘some metonymies are more conventional or more creative than others’. Most adults in the English-speaking world will know immediately what it means to marry money (e.g. to marry a person who is rich), but many would not know that a sheep refers to a person who was born in the year of the sheep in the Chinese lunar calendar; even if they knew about the lunar calendar, sheep is not the most commonly known animal (the year of the dragon, for example). The more conventional the expression, the more typical it is and the more automated its processing is, which results in a lower cognitive demand. The inverse can also be said to apply here: the less conventional the expression, the more atypical it is and, consequently, it may require more effort to process. In Chapter 4 we explored various aspects of typical and atypical expressions, although recall that, as Hanks (2013) pointed out, some typical (i.e. conventional) expressions are infrequent but nevertheless well-established in the language. (170)

She finally married money

(171)

She should avoid marrying a sheep at all costs

These constraints are not independent of each other and work together to ‘help determine what metonymic expression will be selected on a particular occasion’. Bowerman (2019: 26–27) provides an example of where extra effort pays off. In her example, presented here as (172), the speaker and addressee both know the name of the referent (i.e. the person who is dancing). As she points out, it would have been far easier (more efficient and less effortful) to use the person’s name to identify them, but that would not have allowed the speaker to draw attention to what the referent was wearing (green trousers). The additional information encoded is going to be interpreted as relevant and significant (cf. Givón, 1985), in this case leading to some laughter between speaker and addressee. As Bowerman explains (2019: 26–27), ‘providing extra, useful information will also increase the chances that the audience will pay attention, fully process the utterance, and remember the message that the speaker wants to convey’. (172)

Look at the green trousers dance!

When we consider metonymy, what we find is a type of reference, a choice in the expression to be used to refer to a referent. Metonymic reference differs in

10.5 Typical and Atypical Metonymic Reference

215

two main ways from non-metonymic reference. The first is that the speaker is clearly not processing lexical selection for a type specification that corresponds to the typical denotations for that referent. In this sense, metonymic reference is inherently atypical. The second differs in terms of referential potential and topic development, as discussed above (e.g. Cislaru, 2007). As we saw with our brief discussion of this in who’s this? at the beginning of Section 10.3, metonymic reference activates a kind of referential doubling, that is, there is a report and a patient at the same time. Normally, nominal expressions which reference more than one referent are seen as complex noun phrases where the additional referent(s) is expressed through the qualification zone, as we saw in Chapter 3. However, with metonymic reference, it is as though some aspect of what would have been expressed ‘literally’ by one or more qualifiers is instead expressed in the categorisation zone and somehow the intended referent and some aspect of the semantics of the metonymic wording are both active conceptually at the same time. In the case of stigmatisation and social distancing, as was discussed in the medical student essays, the human referent becomes less and less frequently activated until it is almost never brought into focus. To some extent, this view is in line with the position of Bowerman (2019: 51), who argues that ‘referential metonymy should be understood as the literal meaning of the metonymically-used referring expression plus (i) the speaker’s intention as to the target referent, and (ii) contextual assumptions about relevant contiguous links between the linguistically-specified referent of the metonymically-used referring expression and the target referent’. 10.5

Typical and Atypical Metonymic Reference

Now that we have reviewed the concept of metonymy from the perspective of reference and examined some examples from different contexts, we are in a position to draw the chapter to a close with a discussion of how typical, or atypical, metonymic reference is. There is clearly considerable variation, and just as is the case with non-metonymic reference, the answer depends on various factors. We need to remind ourselves of the position we adopted in Chapter 4, where we established our view of typicality as conventionalised language use, which can be said to fall within the norms of a given register. Atypicality, then, would include instances of language which can be said to be infrequent and unconventionalised for a given register. Proper name metonymies are, for Hanks (2013), conventional, that is, mostly well-established conventionally in the language, and therefore typical. Spontaneous, never before uttered, ad hoc metonymies are clearly atypical. However, what we find with non-metonymic reference is that the accessibility status of the referent holds irrespective of referential choice (i.e. the expression used by the speaker to refer). A uniquely identifiable referent will be specified

216

Reference and Metonymy

as such by the speaker whether metonymy is used or not (e.g. D1 and D4 have both had high BMs. vs Mrs Jones has had . . .). Equally, if identifiability is not at stake, an indefinite expression can be used (e.g. one of my tables left without paying vs someone left . . .). Most metonymic reference will fall in between these two extremes, but in principle, as we have seen throughout this volume, the conventional choice for type specification is a noun which can be reasonably accepted as a culturally informed classification of the referent. Where this is not the case, there is an argument for the expression to be an atypical referential choice, but one which most likely brings additional interesting and/or important information. While work is still needed at the two ends of the a-/typicality scale, those in the middle merit particular attention in future work, since expressions which appear to be conventional, ‘may well encode a particular perspective or evaluative slant that is so subtle that it will be taken at face value and not questioned by the interlocutor. Such uses of metonymy are likely to be among the most powerful forms of persuasive language’ (Littlemore, 2015: 205).

11

Epilogue

Now at length are you come to the Epiloge (as it were) or full conclusion of your worke. Briefe Examination (1564)

11.1

Synthesis

We have indeed come to the end of our work. In preparing this volume, we hoped to address the need for an integrated approach to referring, one that brings together functional grammar and cognitive linguistics within a model of context and one that is informed by empirical evidence. We wanted to offer our perspective on building bridges between the traditional literature on reference and more usage-based functional approaches to language. Most of all, we felt there was a need to provide an account of atypical uses of reference given that most literature on reference concentrates on typical reference, frequently using isolated expressions or experimental tasks. What we hope we have shown throughout the volume is that referring expressions matter. They are a fundamental part of our language development, and they align with grammatical skill, to some extent, which means that any study of reference needs an account of lexicogrammar. More than that, however, we wanted to demonstrate that naturally occurring uses of reference not only deviate in certain conditions from their conventional uses, but that both the nature of reference and the referential choice of the expression truly matters in the world. The way we categorise our experience of the world, and the people and things in it, is not neutral. There are real consequences, and we have illustrated the impact of referring through our account of atypical uses of reference, whether that is more creative and playful (e.g. Chapters 8 and 9) or more manipulative (e.g. Chapters 7 and 10). Before closing this chapter, we will briefly summarise the contribution made by this volume and then we will give some consideration to future perspectives. The first part of the volume comprised three chapters which established the foundation to the book by introducing readers to the broad and multidisciplinary field of referring and reference, and by presenting a grammatical 217

218

Epilogue

description of the main linguistic resource for reference, the noun phrase. Chapter 1 presented perspectives from philosophy, psychology, linguistics, and computer science to explain the main background of influences that are still relevant to our understanding of reference. It also covered the main text and discourse approaches to reference, allowing us to highlight the important role of the cognitive perspective, for example, the role of shared information and givenness, which has enriched our understanding of reference. Chapter 2 tackled the complex array of issues related to the concept of definiteness, where we argued that the lexicogrammatical realisation of the referring expression (i.e. referential choice) and its referential function are not related in a one-to-one relationship, but rather in terms of typical and atypical uses of the expressions for a particular referential intent. In Chapter 3, we provided our curated description of the noun phrase by drawing on a cognitive-functional lexicogrammar of referring expressions. The four zones categorised various contributions to the function of the expression in an act of referring. These three chapters formed the basis for the remainder of the volume and provided the explanation for why an integrated approach to reference is needed. The second part of the volume aimed to establish a general view of typical uses of reference. Chapter 4 provided the motivation for the distinction we have made between typical and atypical reference. We argued for a view of typicality as conventionalised language use, which can be said to fall within the norms of a given register and, for atypicality, conversely, to be viewed as infrequent and unconventionalised uses of language for a given register. In Chapter 5, we examined typical uses of reference in spontaneous discourse by building on the concept of context of situation and considering how parameters of register establish a framework for reference variation. Focussing specifically on mode of production, we defined spontaneous discourse in terms of contextual parameters and then described typical reference using evidence from examples attested in spontaneous discourse. We showed that typical reference in spontaneous discourse relies on highly accessible and in focus referents, which makes it easier for acts of reference to be successful and which reduces the cognitive load of the speaker and addressee. Chapter 6 examined reference in planned language production. Our portrait of typical reference in prepared discourse drew on evidence of institutional writing (written text in academic, professional, and social settings) and the effects of the written mode of production on referential choice, including evidence from keystroke logging methods which give us insight into revisions during the writing process. Referential choice in prepared discourse contrasts with that of spontaneous discourse, with a relatively low frequency of pronominal reference and a higher occurrence of more complex referential expressions. We concluded that referring expressions in prepared discourse tend to be more frequently

11.1 Synthesis

219

information-rich and, consequently, the modification and qualification zones of the noun phrase are more active in this type of discourse. The third and final part of the volume was set against the background of the first two part. Adding on to this background through four chapters, the major contribution of this volume has been to provide a new perspective to the study of reference, with each chapter presenting a different perspective on atypical uses of referring expressions and focussing on different situational contexts, or registers. Chapter 7 examined reference in a non-collaborative context. This chapter began with an overview of problems in referring, including how they get resolved. The main aim was to examine uses of reference in contexts which are discursively non-collaborative. Drawing on examples from legal discourse, we showed that referential collaboration is indeed needed, even when the social relationship is adversarial, but that the strategies for ratifying an expression differ depending on whether or not the speaker and addressee share the same discourse goals. Chapter 8 detailed the atypical use of a certain type of indefinite expression. Using written texts from opinion writing in newspaper discourse, we demonstrated how writers can create an adequately definite shared space with their readers to allow an indefinite expression to be understood as referring to an established entity. We concluded that writers use these atypical expressions to ensure the association to an old, identifiable referent remains explicit by exploiting contextually and co-textually established information which is tied to the context or ongoing discourse. Chapter 9 focussed on reference as used by and with children. The chapter first provided an overview of the main developments in uses of reference by children as they learn language. Drawing on a review of relevant literature, we discussed the significance of the tendency for children to underspecify their intended referent in the use of referring expressions. We showed, using a broad range of examples, how quickly children get to a stage where they can exploit reference creatively for their own purposes or just to have fun. We also examined atypical uses of referring expressions in children’s literature, and we considered the reasons for the authors’ deliberate exploitations of known conventions in relation to reference. Finally, in Chapter 10, we flipped the perspective on referential metonymy, arguing for it to be viewed as a type of reference, that is, metonymic reference. Using examples of metonymy from medical discourse contexts, including nurse handover discourse and medical students’ self-reflections, we showed that metonymic reference can be seen as typical or atypical, depending on a variety of factors. While metonymic reference is essential in some contexts for safe and efficient reference, it is clear that, in other contexts, its use can signal stigmatisation of patients and/or difficulties of medical professions in maintaining morale. In these cases, referential choice becomes especially significant. In summary, then, building on the wealth of knowledge that we find in the study of reference, this volume has forged new ground by focussing on

220

Epilogue

atypical uses of reference. The expressions under study here have allowed us to shine some light on aspects of reference that have been significantly understudied. The range of contexts we have explored has helped to bring the theory to life and has shown how language use can inform theory and how theory can, in turn, develop our understanding of these essential expressions. The integrated approach we have developed provides those interested in reference with a comprehensive framework which brings together the lexicogrammar of the expressions with insights from cognitive linguistics, functional grammar, and a theoretical model of context. 11.2

Beyond This Volume

Despite the complexity involved with referring, it seems to happen so easily. Children cope well with reference from a very early age, as we have seen. It can be quite subtle, and small adjustments to an expression will construe additional, and sometimes surprising, information. However, when referring goes wrong, for any reason, interlocutors are very quick to stop ongoing discourse and make sure they align with the speaker on the mental conceptualisation of the referent, even if they do not share the same goals. Metonymy, in particular, is elusive; as a referential choice, it is very efficient, but this efficiency sometimes comes at a cost. Entrenched metonymic reference, as we have seen, can be detrimental, making it difficult for the literal specification of the referent to remain cognitively active, as addressed in Chapter 10. As we have argued throughout this volume, all reference is embedded in context. ‘[W]e never just refer, we always do it as part of a broader speech act’ (Hanks, 2019: 30). The referential dance involves a very complex set of steps, steps which are almost impossible to explain. Referring is an activity; we learn it by doing it. There does not appear to be any evidence that any single (existing) theory or method will provide all the answers we need concerning our understanding of referring. According to Polguère (2015: 68), ‘the right question to ask ourselves when faced with a linguistic expression should be: “How did the Speaker do it?”’. If we genuinely want to answer this question, we need a meaning-driven framework that is able to be complemented with other disciplinary approaches. We are very far from introspective, armchair linguistics. ‘The empirical age of linguistics does need for its investigations to turn on itself’ (Fontaine and Wegener, in press). The next question we have to ask is ‘how can we reconcile different disciplinary approaches?’ Perhaps, most importantly, we need to develop effective ways to take advantage of experimental methods without losing the important insights we gain from usage-based approaches. Embedded in this question is the important role of the parameters of register. We showed in various discussions throughout the

11.2 Beyond This Volume

221

volume, but notably in Chapters 5 and 6, how particularly significant the parameter of mode is in influencing reference, although, of course, field and tenor also play a role. A promising example of what can be achieved in developing methods is demonstrated by Trevisan and Garcia (2019: 44), who show that register variables can be used effectively in experimental methods to resolve ‘a long-standing trade-off between ecological validity and experimental control, [given that] most experiments [are] built on the assumption that the latter can be achieved only at the expense of the former’. Each of the four chapters in the third part of this volume offers fertile ground for future research. They equally provide an example of an integrated approach to referring which can be used to forge new ground in other contexts; for example, to address different issues with respect to referring, and/or to gain new insights into old problems. The volume as a whole, then, is more of a starting point than an end point on the journey to discover the fascinating field of reference.

References

Abbott, B. (2019). The indefiniteness of definiteness. In Abbott, B. and Gundel, J. (eds.), Oxford Handbook of Reference. Oxford: Oxford University Press, 158–175. Ädel, A. (2014). Metonymy in the semantic field of verbal communication: A corpusbased analysis of WORD. Journal of Pragmatics, 67, 72–88. Aitchison, J. (1994). Words in the Mind: An Introduction to the Mental Lexicon. 2nd ed. Oxford: Blackwell. (2011). The Articulate Mammal: An Introduction to Psycholinguistics. London: Routledge. Al-Sharafi, A. (2000). Towards a textual theory of metonymy: A semiotic approach to the nature and role of metonymy in text. PhD dissertation, University of Durham. (2004). Textual Metonymy: A Semiotic Approach. Basingstoke: Palgrave. Allan, K. (2013). What is common ground? In Capone, A., Lo Piparo, F. and Carapezza, M. (eds.), Perspectives on Linguistic Pragmatics. Cham, Switzerland: Springer, 285–310. American College of Obstetricians and Gynecologists. (2012). Communication strategies for patient handoffs: Committee Opinion No. 517. Obstetrics & Gynecology, 119, 408–411. Amoia, M., Kunz, K. and Lapshinova-Koltuns, K. (2012). Coreference in spoken vs. written texts: A corpus-based analysis. In LREC (ed.), Proceedings of the Eighth International Conference in Language Resources and Evaluation. Istanbul: European Language Resource Association, 158–164. Anderson, J. (2003). On the structure of names. Folia Linguistica, 37, 347–398. Appelt, D. (1985/1988). Planning English referring expressions. In McDonald, D. and Bolc, L. (eds.), Natural Language Generation Systems. Berlin: Springer, 69–97. Appelt, D. and Kronfeld, A. (1987). A computational model of referring. In Proceedings of the 10th IJCAI, Milan, Italy, 640–647. Ariel, M. (1989). Review of Discourse Structure and Anaphora, by B. A. Fox. Journal of Linguistics, 25(2), 493–498. (1990). Accessing Noun-Phrase Antecedents. London: Routledge. (1994). Interpreting anaphoric expressions: A cognitive versus a pragmatic approach. Journal of Linguistics, 30, 3–42. (1996). Referring expressions and the +/ coreference distinction. In Gundel, J. and Fretheim, T. (eds.), Referent and Referent Accessibility. Amsterdam: Benjamins, 13–35.

222

References

223

(2001). Accessibility theory: An overview. In Sanders, T., Schilperoord, J. and Spooren, W. (eds.), Text Representation: Linguistic and Psycholinguistic Aspects. Amsterdam: Benjamins, 29–87. Arnold, J., Eisenband, J., Brown-Schmidt, S. and Trueswell, J. (2000). The rapid use of gender information: Evidence of the time course of pronoun resolution from eyetracking. Cognition, 76, B13–B26. Athanasakis, E. (2013). Synthesizing knowledge about nursing shift handovers: Overview and reflections from evidence-based literature. International Journal of Caring Sciences. 6(3), 300–313. Bach, K. (1987). Thought and Reference. Oxford: Oxford University Press. (2008). On referring and not referring. In Gundel, J. K. and Hedberg, N. (eds.), Reference: Interdisciplinary Perspectives. Oxford and New York: Oxford University Press, 13–58. Bangerter, A., Mayor, E. and Pekarek Doehler, S. (2011). Reported speech in conversational storytelling during nursing shift handover meetings. Discourse Processes, 48(3), 183–214. Barcelona, A. (2019). Metonymy. In Dąbrowska, E. and Divjak, D. (eds.), Cognitive Linguistics: Foundations of Language. Berlin: De Gruyter, 167–194. Barnden, J. A. (2010). Metaphor and metonymy: Making their connections more slippery. Cognitive Linguistics, 21(1), 1–34. Bartlett, T., Ylanne, V., Spilioti, T. and Aldridge-Waddon, M. (2020). Nursing handovers as unbounded and scalar events. Applied Linguistics Review, 12(3), 401–418. Bauer, L. (2018). Conversion as metonymy. Word Structure, 11(2), 175–184. Berry, M. (1975). An Introduction to Systemic Linguistics: I. Structures and Systems. London: Batsford. (2013). Towards a study of the differences between formal written English and informal spoken English. In Fontaine, L., Bartlett, T. and O’Grady, G. (eds.), Systemic Functional Linguistics: Exploring Choice. Cambridge: Cambridge University Press, 365–384. (2017). Stratum, delicacy, realisation and rank. In Bartlett, T. and O’Grady, G. (eds.), The Routledge Handbook of Systemic Functional Linguistics. London: Routledge, 42–55. Bevacqua, L., Loáiciga, S., Rohde, H. and Hardmeier, C. (2021). Event and entity coreference across five European languages: Effects of context and referring expression. Dialogue and Discourse, 21(2), 192–226. Bezuidenhout, A. (2019). Joint reference. In Gundel, J. K. and Abbott, B. (eds.), The Oxford Handbook of Reference. Oxford: Oxford University Press, 45–66. Biber, D. (2000). Investigating language use through corpus-based analyses of association patterns. In Kemmer, S. and Barlow, M. (eds.), Usage-Based Models of Language. Stanford: CSLI, 287–313. (2012). Register as a predictor of linguistic variation. Corpus Linguistics and Linguistic Theory, 8(1), 9–37. Biber, D. and Clark, V. (2002). Historical shifts in modification patterns with complex noun phrase structures: How long can you go without a verb? In Fanego, T., Lopez-Couso, M. J. and Perez-Guerra, J. (eds.), English Historical Syntax and Morphology. Amsterdam: Benjamins, 43–66.

224

References

Biber, D. and Conrad, S. (2009). Register, Genre and Style. Cambridge: Cambridge University Press. Biber, D., Conrad, S. and Reppen, R. (1998). Corpus Linguistics: Investigating Language Structure and Use. Cambridge: Cambridge University Press. Biber, D., Conrad, S., Finegan, E., Johansson, S. and Leech, G. (1999). Longman Grammar of Spoken and Written English. Harlow: Longman. Birner, B. and Ward, G. (1994). Uniqueness, familiarity, and the definite article in English. Berkeley Linguistics Society, 20, 93–102. Bischoff, S. and Jany, C. (2013). Introduction. In Bischoff, S. and Jany, C. (eds.), Functional Approaches to Language. Berlin: De Gruyter, 1–8. Bowcher, W. L. (2019). Context and register. In Thompson, G., Bowcher, W. L., Fontaine, L. and Schönthal, D. (eds.), The Cambridge Handbook of Systemic Functional Linguistics. Cambridge: Cambridge University Press, 142–170. Bowen, N. (2016). Modelling choice in digital writing: Functional revisions and ‘texture’. PhD thesis, Cardiff University. Bowen, N. and Thomas, N. (2020). Manipulating texture and cohesion in academic writing: A keystroke logging study. Journal of Second Language Writing, 50, 100773. Bowerman, J. (2016). Examining the nature of referential metonymy. UCL Working Papers in Linguistics, 28, 1–19. (2019). What’s really going on with the ham sandwich? An investigation into the nature of referential metonymy. International Review of Pragmatics, 11, 22–55. Branigan, H., Bell, J. and McLean, J. (2016). Do you know what I know? The impact of participant role in children’s referential communication. Frontiers in Psychology, 7, Article 213. Branigan, H., Tosi, A. and Gillespie-Smith, K. (2016). Spontaneous lexical alignment in children with an autistic spectrum disorder and their typically developing peers. Journal of Experimental Psychology: Learning, Memory, and Cognition, 42(11), 1821–1831. Breban, T. (2010). English Adjectives of Comparison: Lexical and Grammaticalized Uses. Berlin: De Gruyter. (2011). Secondary determiners as markers of generalized instantiation in English Noun Phrases. Cognitive Linguistics, 22(3), 511–533. Breban, T. and Davidse, K. (2003). Adjectives of comparison: The grammaticalization of their attribute uses into postdeterminer and classifier uses. Folia Linguistica, 37 (3/4), 269–317. Brennan, S. (2000). Processes that shape conversation and their implications for Computational Linguistics. In Proceedings, 38th Annual Meeting of the Association for Computational Linguistics. Hong Kong: ACL. Available at: www.aclweb.org/anthology/P00-1001. Accessed 04/08/14 Brennan, S. and Clark, H. (1996). Conceptual pacts and lexical choice in conversation. Journal of Experimental Psychology: Learning, Memory and Cognition, 22, 1482–1493. https://doi.org/10.1037/0278-7393.22.6.1482. Brown, G. (1995). Speakers, Listeners and Communication. Explorations in Discourse Analysis. Cambridge: Cambridge University Press. Brown, G. and Yule, G. (1983). Discourse Analysis. Cambridge: Cambridge University Press.

References

225

Burgoon, M., Denning, P. and Roberts, L. (2002). Language expectancy theory. In Dillard, J. and Pfau, M. (eds.), The Persuasion Handbook: Developments in Theory and Practice. Thousand Oaks: Sage, 117–137. Butler, C. (2005). Functional approaches to language. In Butler, C., Gómez González, M. d. L. A. and Doval-Suárez, S. (eds.), The Dynamics of Language Use: Functional and Contrastive Perspectives. Amsterdam: Benjamins, 3–17. Butler, C. and Gonzálvez-García, F. (2014). Exploring Functional-Cognitive Space. Amsterdam: Benjamins. Bybee, J. (2013). Usage-based theory and exemplar representations of constructions. In Hoffmann, T. and Trousdale, G. (eds.), The Oxford Handbook of Construction Grammar. Oxford: Oxford University Press, 49–69. Camp, M., Cole, A. and Sadler, J. (2020). Addicts and admits: Metonymy in medical students’ reflective writing. Teaching and Learning in Medicine, 32(1), 23–33. https://doi:10.1080/10401334.2019.1607742. Campbell, A., Brooks, P. and Tomasello, M. (2000). Factors affecting young children’s use of pronouns as referring expressions. Journal of Speech, Language, and Hearing Research, 43, 1337–1349. Castillo, L., Smith, K. and Branigan, H. (2019). Interaction promotes the adaptation of referential conventions to the communicative context. Cognitive Science, 43(8), Article e12780. https://doi.org/10.1111/cogs.12780. Chafe, W. (1972). Discourse structure and human knowledge. In Freedle, R. and Carrol, J. (eds.), Language Comprehension and the Acquisition of Knowledge. New York: Academic Press, 41–69. (1976). Givenness, contrastiveness, definiteness, subjects, topics, and point of view. In Li, C. N. (ed.), Subject and Topic. New York: Academic Press, 25–55. (1994). Discourse, Consciousness and Time. Chicago: The University of Chicago Press. Chrispin, L. (2022). A cognitive functional account of pure intransitive verbal categories and their intransitive patterning. PhD thesis, Cardiff University. Chrispin, L. and Fontaine, L. (2023). A cognitive-functional approach to ‘watch’ as a verb of perception. In McGregor, W., Van linden, A., Gentens, C. and Ghesquière, L. (eds.), Reconnecting Form and Meaning: Lexis and Grammar from CognitiveFunctional and Usage-Based Perspectives. Amsterdam: Benjamins, 209–236. Christopherson, P. (1939). The Articles: A Study of their Theory and Use in English. London: Einar Munksgaard and Oxford University Press. Chukharev-Hudilainen, E. (2014). Pauses in spontaneous written communication: A keystroke logging study. Journal of Writing Research, 6(1), 61–84. CiF Editorial (2009). In praise of... Reader’s Digest. The Guardian. Available at: www .theguardian.com/commentisfree/2009/aug/21/readers-digest-journalism-news. Accessed 6 December 2022. Cislaru, G. (2007). Metonymic modelling of discourse, discourse modelling of metonymy. The case of the place-name based metonymies. Culture, Language and Representation. Revista de Estudios Culturales de la Universitat Jaume I, 5, 93–110. Clancy, P. (1980). Referential choice in English and Japanese discourse. In Chafe, W. (ed.), The Pear Stories: Cognitive, Cultural, and Linguistic Aspects of Narrative Production. Norwood: Ablex, 127–201.

226

References

Clapp, L., Reimer, M. and Spire, A. (2019). Negative existentials. In Gundel, J. and Abbott, B. (eds.), The Oxford Handbook of Reference. Oxford: Oxford University Press, 203–235. Clark, E. (2007). Young children’s uptake of new words in conversation. Language in Society, 36(2), 157–182. Clark, E. and Bernicot, J. (2008). Repetition as ratification: How parents and children place information in common grounds. Journal of Child Language, 35(2), 349–371. Clark, H. and Marshall, C. (1981). Definite reference and mutual knowledge. In Joshi, A., Webber, B. and Sag, I. (eds.), Elements of Discourse Understanding. Cambridge: Cambridge University Press, 10–63. Clark, H. and Wilkes-Gibbs, D. (1986). Referring as a collaborative process. Cognition, 22, 1–39. Clark, H., Schreuder, R. and Buttrick, S. (1983). Common ground and the understanding of demonstrative reference. Journal of Verbal Learning and Verbal Behaviour, 22, 245–258. Coene, M. (2006). On the acquisition of the indefinite article: A cross-linguistic study of French, Italian, Romanian and Spanish child speech. In S. Vogeleer (ed.) Bare Plurals, Indefinites and Weak-Strong Distinction. Amsterdam: John Benjamins, 121–146. Cornish, F. (1989). Discourse structure and anaphora: Written and conversational English, Lingua, 79(2–3), 229–243. (1999). Anaphora, Discourse and Understanding. Oxford: Oxford University Press. (2010). Anaphora: Text-based or discourse- dependent? Functionalist vs. formalist accounts. Functions of Language, 17, 207–241. Cotterill, J. (2001). Domestic discord, rocky relationships: Semantic prosodies in representations of marital violence in the O. J. Simpson Trial. Discourse and Society, 12(3), 291–312. https://doi:10.1177/0957926501012003002. Cozens, C. and O’Carroll, L. (2004).The man who came in from the cold. The Guardian. 29 January 2004. Available at: www.theguardian.com/media/2004/jan/ 29/broadcasting.huttoninquiry1. Accessed 6th Dec 2022 Croft, W. (2001). Radical Construction Grammar: Syntactic Theory in Typological Perspective. Oxford: Oxford University Press. (2002). The role of domains in the interpretation of metaphors and metonymies. In Dirven, R. and Pörings, R. (eds.), Metaphor and Metonymy in Comparison and Contrast. Berlin: De Gruyter, 161–205. Croft, W. and Cruise, A. (2004). Cognitive Linguistics. Cambridge: Cambridge University Press. Crompton, P. (2008). Definiteness and indefiniteness in theme. In LewandowskaTomaszczyk, B. (ed.), Corpus Linguistics, Computer Tools, and Applications – State of the Art. Frankfurt am Main: Peter Lang, 309–328. Dale, R. (1989). Cooking up referring expressions. In Proceedings of the 27th Annual Meeting of the Association for Computational Linguistics, Vancouver, Canada. Dale, R. and Reiter, E. (1995). Computational interpretations of the Gricean maxims in the generation of referring expressions. Cognitive Science, 19, 233–263. Dale, R., Geldof, S. and Prost, J-P. (2005). Using natural language generation in automatic route description. Journal of Research and Practice in Information Technology, 37(1), 89–105.

References

227

Dancygier, B. (2009). Genitives and proper names in constructional blends. In Evans, V. and Pourcel, S. (eds.), New Directions in Cognitive Linguistics. Amsterdam: Benjamins, 161–184. (2011). Cognitive linguistics and the study of textual meaning. Constructions and Frames, 3(2), 208–235. Davidse, K., Breban, T. and Van linden, A. (2008). Deictification: The development of secondary deictic meanings by adjectives in the English NP. English Language and Linguistics, 12, 475–503. Davies, C. (2004). A corpus-based investigation of noun to verb conversion in English. Unpublished PhD dissertation, University of Liverpool. Davies, C., Andrés-Roqueta, C. and Norbury, C. (2016). Referring expressions and structural language abilities in children with specific language impairment: A pragmatic tolerance account. Journal of Experimental Child Psychology, 144, 98–113. de Haan, P. (1987). Relative clauses in indefinite noun phrases. English Studies, 68, 171–189. (1989). Postmodifying Clauses in the English Noun Phrase: A Corpus-Based Study. Amsterdam: Rodopi. Denroche, C. (2018). Text metaphtonymy: The interplay of metonymy and metaphor in discourse. Metaphor and the Social World, 8(1), 1–24. Dirven, R. (2002). Metonymy and metaphor: Different mental strategies of conceptualization. In Dirven, R and Pörings, R. (eds.), Metaphor and Metonymy in Comparison and Contrast. Berlin: De Gruyter, 75–111. Dryer, M. (2004). Noun phrases without nouns. Functions of Language, 11(1), 43–76. Du Bois, J. (1980). Beyond definiteness. The trace of identity in discourse. In Chafe, W. (ed.), The Pear Stories: Cognitive, Cultural, and Linguistic Aspects of Narrative Production. Norwood: Ablex, 203–274. (1987). The discourse basis of ergativity. Language, 63, 805–855. (1997). Definiteness, reference and analogues. University of California, Santa Barbara. Unpublished paper. (ed.) (2000). Santa Barbara Corpus of Spoken American English, Part 1. Philadelphia: Linguistic Data Consortium. (ed.) (2003a). Santa Barbara Corpus of Spoken American English, Part 2. Philadelphia: Linguistic Data Consortium. (2003b). Argument structure: Grammar in use. In Du Bois, J., Kumpf, L. and Ashby, W. (eds.), Preferred Argument Structure: Grammar as Architecture for Function. Amsterdam: Benjamins, 11–60. Ehrlich, S. (2002). Legal institutions, nonspeaking recipiency and participants’ orientations. Discourse and Society, 13(6), 731–747. https://doi.org/10.1177/ 0957926502013006753. Epstein, R. (1994). Discourse and definiteness: Synchronic and diachronic perspectives. PhD dissertation, University of California. (2002). The definite article, accessibility, and the construction of discourse referents. Cognitive Linguistics, 12(4), 333–378. Evans, G. (1982). The Varieties of Reference (published posthumously, McDowell, J. (ed.)). Oxford: Oxford University Press. Falkum, I., Recasens, M. and Clark, E. (2017). ‘The moustache sits down first’: On the acquisition of metonymy. Journal of Child Language, 43, 87–119.

228

References

Fawcett, R. (1980). Cognitive Linguistics and Social Interaction: Towards an Integrated Model of a Systemic Functional Grammar and the Other Components of an Interacting Mind. Heidelberg: Julius Groos and Exeter University. (1993). Language as program: A reassessment of the nature of descriptive linguistics. Language Sciences, 14(4), 623–657. (2000). A Theory of Syntax for Systemic Functional Linguistics. Amsterdam: Benjamins. (2004). Extracts from THE SYSTEM NETWORK AND REALIZATION RULES FOR ‘THING’ showing systems that enter into the generation of MODIFIERS and QUALIFIERS in the ENGLISH NOMINAL GROUP. Unpublished manuscript, Cardiff University. (2007). Modelling ‘Selection’ between Referents in the English Nominal Group. In Butler, C., Hidalgo Downing, R. and Lavid, J. (eds.), Functional Perspectives on Grammar and Discourse: In Honour of Angela Downing. Amsterdam: Benjamins, 165–204. Flower, L. and Hayes. J. (1981). A cognitive process theory of writing. College Composition and Communication, 32(4), 365–387. Fontaine, L. (2008). A systemic functional approach to referring expressions: Reconsidering postmodification in the nominal group. PhD thesis, Cardiff University. (2012). Analyzing English Grammar: A Systemic-Functional Introduction. Cambridge: Cambridge University Press. (2013). Semantic options and complex functions: A recursive view of choice. In Fontaine, L., Bartlett, T. and O’Grady, G. (eds.), Systemic Functional Linguistics: Exploring Choice. Cambridge: Cambridge University Press, 95–114. (2017). The English Nominal Group: The centrality of the Thing element. In Bartlett, T. and O’Grady, G. (eds.), Routledge Systemic Functional Linguistic Handbook. London: Routledge, 267–283. (2018). The noun, grammar and context. Linguistics and the Human Sciences, 11, 2–3. (2019). A textual perspective on referential metonymy. Estudios de Lingüística Aplicada, 37(70), 199–225. Fontaine, L. and Aldridge, M. (2015). The impact of mode on writing processes: A cognitive functional perspective on student writing. LyCE Estudios, 17, 9–34. Fontaine, L. and Jones, K. (2020). We are all one: Shifting reference in reconciliation talk. In Zappavigna, M. and Dreyfus, S. (eds.), Discourses of Hope and Reconciliation. London: Bloomsbury Publishing, 185–204. Fontaine, L. and Schönthal, D. (2020). Referring and the nominal group: A closer look at the selector element. In Tucker, G., Huang, G., Fontaine, L. and McDonald, E. (eds.), Approaches to Systemic Functional Grammar. London: Equinox, 174–190. Fontaine, L. and Wegener, R. (2022). Epilogue. In Aldridge, M. and Asp, E. (eds.), Theories and Evidences in Functional Linguistics. London: Routledge, 170–179. Fox, B. (1987). Discourse, Structure and Anaphora. Cambridge: Cambridge University Press. Frege, G. (1892/1993). On sense and reference. In Moore, A. (ed.), Meaning and Reference. Oxford: Oxford University Press, 23–42.

References

229

Fries, P. (1970). Tagmeme Sequences in the English Noun Phrase. Oklahoma: Benjamin F. Elson. (1983). On the status of theme in English: Arguments from discourse. In Petöfi, J. and Sözer, E. (eds.), Micro and Macro Connexity of Texts. Hamburg: Helmut Buske Verlag, 116–152. (2001). Issues in modeling the textual metafunction: A constructive approach. In Scott, M. and Thompson, G. (eds.), Patterns of Text: In Honour of Michael Hoey. Amsterdam: Benjamins, 83–107. Garnham, A. (2001). Mental Models and the Interpretation of Anaphora. New York: Psychology Press. Geluykens, R. (1994). The Pragmatics of Discourse Anaphora in English: Evidence from Conversational Repair. Berlin: De Gruyter. Gernsbacher, M. (1991). Comprehending conceptual anaphors. Language and Cognitive Processes, 6, 81–105. Ghesquière, L. (2014). The Directionality of (Inter)subjectification in the English Noun Phrase: Pathways of Change. Berlin: De Gruyter. Givón, T. (1983). Topic continuity in discourse: An introduction. In Givón, T. (ed.), Topic Continuity in Discourse: A Quantitative Cross-Language Study. Amsterdam: Benjamins, 4–41. (1985). Iconicity, isomporphism and non-arbitrary coding in syntax. In Haiman, J. (ed.), Iconicity in Syntax. Amsterdam: Benjamins, 187–219. (1989). Mind, Code and Context. Essays in Pragmatics. London: Lawrence Erlbaum. (1993a). English Grammar 1. A Functional-Based Introduction. Amsterdam: Benjamins. (1993b). English Grammar 2. A Functional-Based Introduction. Amsterdam: Benjamins. (1995). Functionalism and Grammar. Amsterdam: Benjamins. (2015). The Diachrony of Grammar. Amsterdam: Benjamins. Grant, P. (2009). What’s the BM? BMJ, 339, a2539. Gray, B. (2006). This and these as pronouns and determiners: A corpus-based study of published academic research articles. PhD dissertation, Iowa State University. (2010). On the use of demonstrative pronouns and determiners as cohesive devices: A focus on sentence-initial this/these in academic prose. Journal of English for Academic Purposes, 9(3), 167–183. Greenstein, T. (2008). Israel’s Royal welcome. The Guardian. 25 March 2008. Available at: www.guardian.co.uk/commentisfree/2008/mar/25/ israelsroyalwelcome. Accessed 6 December 2022 Grice, H. (1975). Logic and conversation. In Cole, P. and Morgan, J. (eds.), Syntax and Semantics, Vol. 3: Speech Acts. New York: Seminar Press, 225–242. Grigoroglou, M. (2018). Informativeness in children’s event descriptions: Effects of task, typicality and listener needs. PhD dissertation, University of Delaware. Gundel, J. K. (1996). Relevance theory meets the givenness hierarchy. An account of inferrables. In Gundel, J. K. and Fretheim, T. (eds.), Reference and Referent Accessibility. Amsterdam: Benjamins, 141–153. Gundel, J. K. and Johnson, K. (2013). Children’s use of referring expressions in spontaneous discourse: Implications for theory of mind development. Journal of Pragmatics, 56, 43–57.

230

References

Gundel, J. K., Hedberg, N. and Zacharski, R. (1993). Cognitive status and the form of referring expressions in discourse. Language, 69(2), 274–307. (2019). Cognitive status and the form of referring expressions in discourse. In Gundel, J. K. and Abbott, B. (eds.), The Oxford Handbook of Reference. Oxford: Oxford University Press, 67–99. Gundel, J. K., Hedberg, N., Zacharski, R., Mulkern, A., Custis, T., Swierzbin, B. and Watters, S. (2006). Coding protocol for statuses on the givenness hierarchy. Manuscript, University of Minnesota. Available at: www.sfu.ca/~hedberg/ Coding_for_Cognitive_Status.pdf. Accessed 16 May 2013. Halliday, M. A. K. (1967). Notes on transitivity and theme in English: Part 1. Journal of Linguistics, 3(1), 37–81. (1968). Users and uses of language. In Fishman, J. (ed.), Readings in the Sociology of Language. The Hague: Mouton, 139–169. (1970). Functional diversity in language as seen from a consideration of modality and mood in English. Foundations of Language, 6(3), 322–361. (1973). Explorations in the Functions of Language. London: Edward Arnold. (1975/2004). Learning how to mean. In Webster, J. (ed.), The Language of Early Childhood. Collected Works of M. A. K Halliday, vol 3. London: Continuum, 28–59. (1977/2002). Text as semantic choice in social contexts (1977). In Webster, J. (ed.), Linguistic Studies of Text and Discourse. Collected Works of M. A. K Halliday, vol 2. London: Continuum, 23–84. (1978). Language as a Social Semiotic: The Social Interpretation of Language and Meaning. London: Edward Arnold. (1981/2002). Text semantics and clause grammar: How is a text like a clause? In Webster, J. (ed.), On Grammar, Collected Works of M. A. K. Halliday, vol 1. London: Continuum, 219–260. (1985). An Introduction to Functional Grammar. London: Arnold. (1991/2007). The notion of ‘context’ in language education. In Webster, J. (ed.), Language and Education. Collected Works of M. A. K Halliday, vol 9. London: Continuum, 269–290. (1992/2002). Spoken and written modes of meaning. In Webster, J. (ed.), On Grammar. Collected Works of M. A. K Halliday, vol 1. London: Continuum, 323–351. (1994). An Introduction to Functional Grammar. 2nd ed. London: Arnold. (1997/2003). Linguistics as metaphor. In Webster, J. (ed.), On Language and Linguistics. Collected Works of M. A. K. Halliday, vol 3. London: Continuum, 248–270. (1998/2004). Representing the child as a semiotic being (one who means). In Webster, J. (ed.), The Language of Early Childhood. Collected Works of M. A. K. Halliday, vol 3. London: Continuum, 6–27. (2000/2002). Grammar and daily life: Concurrence and complementarity. In Webster, J. (ed.), On Grammar. Collected Works of M. A. K. Halliday, vol 1. London: Continuum, 369–383. (2006/2013). Written language, standard language, global language. In Webster, J. (ed.), Halliday in the 21st Century. Collected Works of M. A. K. Halliday, vol 11. London: Continuum, 87–104.

References

231

(2011/2013). Why do we need to understand about language? In Webster, J. (ed.), Halliday in the 21st Century. Collected Works of M. A. K. Halliday, vol 11. London: Continuum, 71–82. Halliday, M. A. K. and Hasan, R. (1976). Cohesion in English. London: Longman. Halliday, M. A. K. and Matthiessen, C. (1999). Construing Experience through Meaning: A Language-Based Approach to Cognition. London: Continuum. (2004). An Introduction to Functional Grammar. 3rd ed. London: Arnold. (2014). Halliday’s Introduction to Functional Grammar. 4th ed. London: Routledge. Hanks, P. (2000). Do word meanings exist? Computers and the Humanities, 34(1/2), 205–215. (2013). Lexical Analysis: Norms and Exploitations. Cambridge: The MIT Press. (2019). Reference as a speech act. In Gundel, J. K. and Abbott, B. (eds.), The Oxford Handbook of Reference. Oxford: Oxford University Press, 11–18. Hannah, J. and Brennan, S. (2007). Speakers’ eye gaze disambiguates referring expressions during face-to-face conversation. Journal of Memory and Language, 57, 596–615. Hartley. J. (2007). Longitudinal studies of the effects of new technologies on writing: Two case studies. In Torrance, M., Van Waes, L. and Galbraith, D. (eds.), Writing and Cognition: Research and Applications. Oxford: Elsevier, 293–306. Hasan, R. (1985). The texture of a text. In Halliday, M. A. K. and Hasan, R. (eds.), Language, Context and Text. Oxford: Oxford University Press, 52–69. (1998). Speaking with reference to context. In Ghadessy, M. (ed.), Text and Context in Functional Linguistics. Amsterdam: Benjamins, 219–328. Haspelmath, M. (2006). Against markedness (and what to replace it with). Journal of Linguistics, 42, 25–70. Haviland, S. and Clark, H. (1974). What’s new? Acquiring new information as a process in comprehension. Journal of Verbal Learning and Verbal Behavior, 13, 512–521. Hawkins, J. (1978). Definiteness and Indefiniteness. London: Humanities Press/Croom Helm. (1991). On (in)definite articles: Implicatures and (Un)grammaticality, Prediction. Journal of Linguistics, 27, 405–442. Hawkins, R. (1981). Towards an account of the possessive constructions: NP’s N and N of NP. Journal of Linguistics, 17, 247–269. (1984). A note on referent identifiability and co-presence. Journal of Pragmatics, 8, 649–659. Hendriks, P., Koster, C. and Hoeks, J. (2014). Referential choice across the lifespan: Why children and elderly adults produce ambiguous pronouns. Language, Cognition and Neuroscience, 29(4), 391–407. https://doi:10.1080/01690965.2013 .766356. Hickmann, M. (2003). Children’s Discourse: Person, Space, and Time across Languages. Cambridge: Cambridge University Press. Hilpert, M. (2006). Keeping an eye on the data: Metonymies and their patterns. In Stefanowitsch, A. and Gries, S. Th. (eds.), Corpus-Based Approaches to Metaphor and Metonymy. Berlin: De Gruyter, 123–151. Hinds, J. (1987). Reader versus writer responsibility: A new typology. In Connor, U. and Kaplan, R. (eds.), Writing across Languages: Analysis of L2 Text. Reading, MA: Addison-Wesley, 141–152.

232

References

Horacek, H. (1997). An algorithm for generating referential descriptions with flexible interfaces. In Proceedings of the 35th Annual Meeting of the Association for Computational Linguistics (ACL97/EACL97), Madrid, Spain, 206–213. Hudson, R. (2004). Are determiners heads? Functions of Language, 11(1), 7–42. Hughes, M. and Allen, S. (2015). The incremental effect of discourse-pragmatic sensitivity on referential choice in the acquisition of a first language. Lingua, 155, 43–61. Hunston, S. (2002). Corpora in Applied Linguistics. Cambridge: Cambridge University Press. (2007). Semantic prosody revisited. International Journal of Corpus Linguistics, 12(2), 49–268. Hurford, J., Heasley, B. and Smith, M. (2007). Semantics: A Coursebook. 2nd ed. Cambridge: Cambridge University Press. Jeffries, S. (2013). David Beckham: How this crock of a footballer can still woo the French. The Guardian. 1 February 2013. Available at: www.guardian.co.uk/ commentisfree/2013/feb/01/david-beckham-woo-french-football. Accessed 6 December 2022. Jiang, X. (2013). Referential metonymy: Reference transfer and pragmatic motivations. UCL Working Papers in Linguistics, 25, 1–19. Johnston, T. and Schembri, A. (2007). Australian Sign Language (Auslan): An Introduction to Sign Language Linguistics. Cambridge: Cambridge University Press. Jones, K. (2014). Towards an understanding of the use of indefinite expressions for definite reference in English discourse. PhD thesis, Cardiff University. (2018). ‘A man who revels in his own ignorance, racism and misogyny’: Identifiable referents trump indefinite grammar. Functional Linguistics, 5, article number: 11. Jucker, A. (2012). Social Stylistics: Syntactic Variation in British Newspapers. Berlin: De Gruyter Mouton. Kärkkäinen, E. (1996). Preferred argument structure and subject role in American English conversational discourse. Journal of Pragmatics, 25, 675–701. Kartunnen, L. (1976). Discourse Referents. Reprinted in Syntax and Semantics. London: Academic Press Inc. Kavalir, M. (2017). Post-deictic and grounding. In Neumann, S., Wegener, R., Fest, J., Niemietz, P. and Hützen, N. (eds.), Challenging Boundaries in Linguistics: Systemic Functional Perspectives. Aachen British and American Studies Series. Frankfurt am Main: Peter Lang Verlag, 53–72. Kellogg, R. (2001). Competition for working memory among writing processes. The American Journal of Psychology, 114(2), 175–191. Kemmer, S. and Barlow, M. (2000). Introduction: A usage-based conception of language. In Kemmer, S. and Barlow, M. (eds.), Usage-Based Models of Language. Stanford: CSLI, vii–xxviii. Kibrik, A. (2011). Reference in Discourse. Oxford: Oxford University Press. Kibrik, A., Khudyakova M., Dobrov, G., Linnik, A. and Zalmanov, D. (2016). Referential choice: Predictability and its limits. Frontiers in Psychology, 7, 1–21. https://doi:10.3389/fpsyg.2016.01429. Kilgarriff, A., Baisa, V., Bušta, J., Jakubíček, M., Kovář, V., Michelfeit, J., Rychlý, P. and Suchomel, V. (2014). The sketch engine: Ten years on. Lexicography, 1, 7–36.

References

233

Klassen, K. (2018). Investigating the lexical load of proper names for L2 English readers. PhD thesis, Cardiff University. (2021). Proper name theory and implications for second language reading. Language Teaching, 55(2), 149–155. https://doi:10.1017/S026144482100015X. Köder, F. and Falkum, I. L. (2020). Children’s metonymy comprehension: Evidence from eye-tracking and picture selection. Journal of Pragmatics, 156, 191–205. Kress, G. and van Leeuwen, T. (2001). Multimodal Discourse: The Modes and Media of Contemporary Communication. London: Arnold. Kronfeld, A. (1990). Reference and Computation: An Essay in Applied Philosophy of Language. Cambridge: Cambridge University Press. Kronmüller, E. and Barr, D. (2015). Referential precedents in spoken language comprehension: A review and meta-analysis. Journal of Memory and Language, 83, 1–19. https://doi:10.1016/j.jml.2015.03.008. Kronrod, A. and Engel, O. (2001). Accessibility theory and referring expressions in newspaper headlines. Journal of Pragmatics, 33, 683–699. Kunz, K., Lapshinova-Koltunski, E., Martínez-Martínez, J., Menzel, K. and Steiner, E. (2021). GECCo – German-English Contrasts in Cohesion: Insights from Corpus-Based Studies of Languages, Registers and Modes. Berlin, Boston: De Gruyter. Laing, C. (2019). Phonological motivation for the acquisition of onomatopoeia: An analysis of early words. Language Learning and Development, 15(2), 177–197. https://doi:10.1080/15475441.2019.1577138. Lakoff, G. (1987). Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: The University of Chicago Press. Lakoff, G. and Johnson, M. (1980). Metaphors We Live By. Chicago: The University of Chicago Press. Lambrecht, K. (1994). Information Structure and Sentence Form: Topic, Focus, and the Mental Representations of Discourse Referents. Cambridge: Cambridge University Press. Langacker, R. (1987). Foundations of Cognitive Grammar: Descriptive Application, vol 2. Stanford: Stanford University Press. (1991). Foundations of Cognitive Grammar: Volume II: Descriptive Application. Stanford: Stanford University Press. (1999). Grammar and Conceptualization. Berlin: De Gruyter. (2004). Remarks on nominal grounding. Functions of Language, 11(1), 77–113. (2008). Cognitive Grammar: A Basic Introduction. Oxford: Oxford University Press. Levelt, W. (1989). Speaking: From Intention to Articulation. Cambridge: MIT Press. (2013). A History of Psycholinguistics : The Pre-Chomskyan Era. Oxford: Oxford University Press. Littlemore, J. (2015). Metonymy: Hidden Shortcuts in Language, Thought and Communication. Cambridge: Cambridge University Press. Littlemore, J. and Tagg, C. (2018). Metonymy and text messaging: A framework for understanding creative uses of metonymy. Applied Linguistics, 39(4), 481–507. Lloyd, H., Bartlett, T., Aldridge-Waddon, M., Spilioti, T., and Ylanne, V. (2021). Opening up space for compassion in nurses’ handover meetings. Communication and Medicine, 16(3), 224–237.

234

References

Lukin, A., Moore, A., Herke, M., Wegener, R. and Wu, C. (2011). Halliday’s model of register revisited and explored. Linguistics and the Human Sciences, 4(2), 187–213. Lyons, C. (1999). Definiteness. Cambridge: Cambridge University Press. Lyons, J. (1977). Semantics. Cambridge: Cambridge University Press. Macarthy, C. (1992). All the Pretty Horses. New York: Random House. MacKenzie, I. (2015). Rethinking reader and writer responsibility in academic English. Applied Linguistics Review, 6(1), 1–21. https://doi.org/10.1515/applirev-20150001 Maratsos, M. (1974). Children’s use of definite and indefinite articles. Child Development, 45(2), 446–455. Martin, J. (1992). English Text – System and Structure. Amsterdam: Benjamins. (2019). Discourse semantics. In Thompson, G., Bowcher, W., Fontaine, L. and Schönthal, D. (eds.), The Cambridge Handbook of Systemic Functional Linguistics Cambridge: Cambridge University Press, 358–381. https://doi:10 .1017/9781316337936.016. Martínez-Insua, A. E. and Pérez-Guerra, J. (2011). An open-sesame approach to English noun phrases: Defining the NP. English Language and Linguistics, 15(2), 201–221. https://doi:10.1017/S1360674311000013. Matthews, D., Lieven, E. and Tomasello, M. (2007). How toddlers and preschoolers learn to uniquely identify referents for others: A training study. Child Development, 78(6), 1744–1759. McCabe, A. (2021). A Functional Linguistic Perspective on Developing Language. London: Routledge. Meir, N. and Novogrodsky, R. (2021). Referential expressions in monolingual and bilingual children with and without Autism Spectrum Disorder (ASD): A study of informativeness and definiteness. Journal of Child Language, 1–30. https://doi:10 .1017/S0305000921000702. Metzing, C. and Brennan, S. (2003). When conceptual pacts are broken: Partnerspecific effects on the comprehension of referring expressions. Journal of Memory and Language, 49, 201–213. Moors, A. (2015). Metonymy and pragmatic inferencing in mental health. PhD dissertation, University of Technology Sydney. Mulkern, A. (1996). The game of the name. In Gundel, J. K. and Fretheim, T. (eds.), Reference and Referent Accessibility. Amsterdam: Benjamins, 235–250. Munsch, R. (1981). The Paper Bag Princess. Toronto: Annick Press. Nida-Rümelin, J. (2000). Rationality: Coherence and structure. In Nida-Rümelin, J. and Spohn, W. (eds.), Rationality, Rules, and Structure. Theory and Decision Library, vol 28. Dordrecht: Springer. https://doi.org/10.1007/978-94-015-9616-9_1. Nozari, N. and Novick, J. (2017). Monitoring and control in language production. Current Directions in Psychological Science, 26(5), 403–410. Nunberg, G. (1979). The non-uniqueness of semantic solutions. Polysemy, Linguistics and Philosophy, 3, 143–184. Oakhill, J., Garnham, A., Gernsbacher, M. and Cain, K. (1992). How natural are conceptual anaphors? Language and Cognitive Processes, 7(3/4), 257–280. O’Halloran, K. (2004). Mathematical Discourse: Language, Symbolism and Visual Images. London: Continuum.

References

235

O’Donnell, M. (2021). Dynamic modelling of context: Field, Tenor and Mode revisited. Lingua. https://doi.org/10.1016/j.lingua.2020.102952. O’Neill, D. (1996). Two-year-old children’s sensitivity to a parent’s knowledge state when making requests. Child Development, 67(2), 659–677. https://doi:10.2307/ 1131839. Panther, K.-U. (2006). Metonymy as usage event. In Kristiansen, G., Achard, M., Dirven, R. and de Mendoza Ibañez, R. (eds.), Cognitive Linguistics: Current Applications and Future Perspectives. Berlin: De Gruyter, 147–186. Panther, K.-U. and Thornburg, L. (2009). The roles of metaphor and metonymy in English -er nominals. In Dirven, R. and Pörings, R. (eds.), Metaphor and Metonymy in Comparison and Contrast. Berlin: De Gruyter Mouton, 279–322. Paris, C. and McKeown, K. (1987). Discourse strategies for describing complex physical objects. In Kempen, G. (ed.), Natural Language Generation. NATO ASI Series, vol 135. Springer: Dordrecht. https://doi.org/10.1007/978-94-009-3645-4_8. Peirsman, Y. and Geeraerts, D. (2006). Metonymy as prototypical category. Cognitive Linguistics, 17(3), 269–316. Perniss, P. and Özyürek, A. (2015). Visible cohesion: A comparison of reference tracking in sign, speech, and co-speech gesture. Topics in Cognitive Science, 7(1), 36–60. https://doi: 10.1111/tops.12122. Epub 4 December 2014. PMID: 25472492. Piolat, A., Roussey, J.-Y., Olive, T. and Farioli, F. (1996). Charge mentale et mobilisation des processus rédactionnels: examen de la procédure de Kellogg. Psychologie Française, 41(4), 339–354. Poix, C. (2018). Neology in children’s literature: A typology of occasionalisms. Lexis [Online], 12. URL: http://journals.openedition.org/lexis/2111; doi:10.4000/ lexis.2111 Polguère, A. (2015). Lexical contextualism: The Abélard Syndrome. In Gala, N., Rapp, R. and Bel-Enguix, G. (eds.), Language Production, Cognition, and the Lexicon, 48. London: Springer, 53–73. Power, R. (1999). Generating referring expressions with a unification grammar. In Proceedings of the Nth Conference of the European Association for Computation Linguistics, Bergen, Norway. Prince, E. (1981). Toward a taxonomy of given–new information. In Cole, P. (ed.), Radical Pragmatics. New York: Academic Press, 223–255. (1992). Subjects, definiteness and information status. In Mann, W. C. and Thompson, S. A. (eds.), Discourse Description: Diverse Linguistic Analyses of a FundRaising Text. Amsterdam/Philadelphia: Benjamins, 295–325. Quirk, R., Greenbaum, S., Leech, S. and Svartvik, J. (1985). A Comprehensive Grammar of the English Language. London: Longman. Radden, G. (2009). Generic reference in English: A metonymic and conceptual blending analysis. In Panther, K.-U., Thornburg, L. and Barcelona, A. (eds.), Metonymy and Metaphor in Grammar. Amsterdam: Benjamins, 199–228. Radden, G. and Dirven, R. (2007). Cognitive English Grammar. Amsterdam: Benjamins. Raeisi, A., Rarani, M. and Soltani, F. (2019). Challenges of patient handover process in healthcare services: A systematic review. Journal of Education and Health Promotion, 8, 173. https://doi.org/10.4103/jehp.jehp_460_18. Reboul, A. (1997). What (if anything) is accessibility? A relevance-oriented criticism of Ariel’s Accessibility Theory of referring expressions. In Connolly, J., Vismans, R.,

236

References

Butler, C. and Gatward, R. (eds.), Discourse and Pragmatics in Functional Grammar. Berlin: De Gruyter, 91–108. Reiter, E. and Dale, R. (1997). Building applied natural language generation systems. Journal of Natural Language Engineering, 3, 57–87. Reiter, E., Sripada, S., Hunter, J., Yu, J. and Davy, I. (2005). Choosing words in computer-generated weather forecasts. Artificial Intelligence, 67, 137–169. Rigby, B. (2022). Interview with B. Johnson. Sky News. www.youtube.com/watch?v= 7upXsB4ztkE. Rijkhoff, J. (2002). The Noun Phrase. Oxford: Oxford University Press. Roberts, C. (2019). Contextual influences on reference. In Abbott, B. and Gundel, J. (eds.), Oxford Handbook of Reference. Oxford: Oxford University Press, 302–324. Rosch, E. (1975). Cognitive representations of semantic categories. Journal of Experimental Psychology: General, 104, 192–233. Rozendaal, M. and Baker, A. (2008). A cross-linguistic investigation of the acquisition of the pragmatics of indefinite and definite reference in two-year olds. Journal of Child Language, 35, 773–807. Rudge, L. A. (2022). The nominal group in British Sign Language: A preliminary description. Word, 68(2), 83–108. https://doi.org/10.1080/00437956.2021 .2024351. Ruiz de Mendoza Ibáñez, F. and Díez Velasco, O. (2003). Patterns of conceptual interaction. In Dirven, R. and Pöings, R. (eds.), Metaphor and Metonymy in Comparison and Contrast. Berlin: De Gruyter, 489–553. Russell, B. (1930). Introduction to Mathematical Philosophy. London: G. Allen and Unwin. Sachs, S. (1967). Recognition memory for syntactic and semantic aspects of connected discourse. Perception and Psychophysics, 2(9), 437–442. Sainsbury, M. (2005). Reference without Referents. Oxford: Oxford University Press. Salazar-Orvig, A. (2019). Reference and referring expressions in first language acquisition. In Abbott, B. and Gundel, J. (eds.), Oxford Handbook of Reference. Oxford: Oxford University Press, 283–308. Salazar-Orvig, A., Haydée, M., Caët, S., Corlateanu, C., da Silva, C., Hassan, R., Heurdier, J., Mené, M., Leber-Marin, J., and Morgenstern, A. (2013). Definite and indefinite determiners in French-speaking toddlers: Distributional features and pragmatic-discursive factors. Journal of Pragmatics, 56, 88–112. Salazar-Orvig, A., de Weck, G., Hassan, R. and Rialland, A. (2021). A dialogical approach to the acquisition and usage of referring expressions: Theoretical challenges and methodological issues. In Salazar-Orvig, A., de Weck, G., Hassan, R. and Rialland, A. (eds.), The Acquisition of Referring Expressions: A Dialogical Approach. Amsterdam: Benjamins, 2–38. Sanford, A. and Garrod, S. (1981). Understanding Written Language: Explorations of Comprehension beyond the Sentence. New York: John Wiley & Sons. Sarda, L. and Carter-Thomas, S. (2021). Référence, qualification et distribution de l’information : approche contrastive français-anglais. Travaux de Linguistique : Revue Internationale de Linguistique Française, De Boeck Université, 82, 45–76. Schaeffer, J. and Matthewson, L. (2005). Grammar and pragmatics in the acquisition of article systems. Natural Language and Linguistic Theory, 23, 53–101. Schiffrin, D. (1994). Approaches to Discourse. Massachusetts/Oxford: Blackwell.

References

237

(2006). In Other Words: Variation in Reference and Narrative. Cambridge: Cambridge University Press. Schnedecker, C. (2018). Reference chains and genre identification: From Discrete to Non-Discrete Units. In Legallois, D., Charnois, T. and Larjavaara, M. (eds.), The Grammar of Genres and Styles. Berlin: De Gruyter, 39–66. Schober, M. and Clark, H. (1989). Understanding addressees and overhearers. Cognitive Psychology, 21, 211–232. Schönthal, D. (2016). On the multifaceted nature of English of-NPs: A theoretical, corpus, cotextual and cognitive approach. PhD thesis, Cardiff University. (2022). The cohesive landscape of English of-NPs: An empirical, expression-centred approach to coherence. In Aldridge, M. and Asp, E. (eds.), Theories and Evidences in Functional Linguistics. London: Routledge. Schouten, C. and Vonk, W. (1995). On the use of marked indefinite noun phrases. The Perspectival Nature of Indefinite Noun Anaphors. Technical Report, ESPRIT Project 6665 DANDELION. Tilburg / Nijmegen etc.: DANDELION Consortium and CEC. Searle, J. (1969). Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press. (1990). Foreword. In A. Kronfeld (ed.), Reference and Computation: An Essay in Applied Philosophy of Language. Cambridge: Cambridge University Press, xiii–xviii. Serratrice, L. and Allen, S. (2015). Introduction: An overview of the acquisition of reference. In Serratrice, L. and Allen, S. (eds.), The Acquisition of Reference. Amsterdam: Benjamins, 1–24. Shallow, J. (n.d.). The way i see things [illustration]. Sharples, M. (1996). An account of writing as creative design. In Levy, C. M. and Ransdell, S. (eds.), The Science of Writing: Theories, Methods, Individual Differences, and Applications. Mahwah: Lawrence Erlbaum Associates, 127–148. Slabakova, R., Cabrelli Amaro, J. and Kyun Kang, S. (2016). Regular and novel metonymy: Can you curl up with a Good Agatha Christie in Your Second Language? Applied Linguistics, 37(2), 175–197. Smith Jr, E. (1986). Achieving impact through the interpersonal component. In Couture, B. (ed.), Functional Approaches to Writing: Research Perspectives. London: Frances Pinter, 108–119. Spilioti, T., Aldridge-Waddon, M., Bartlett, T. and Ylanne, V. (2019). Conceptualizing language awareness in healthcare communication: The case of nurse shift-change handover meetings. Language Awareness, 28(3), 207–226. Stegemann-Philipps, C., Butz, M., Winkler, S. and Achimova, A. (2021). Speakers use more informative referring expressions to describe surprising events. In Proceedings of the Annual Meeting of the Cognitive Science Society, 43. Retrieved from https://escholarship.org/uc/item/7374p5xq. Stegenwallner-Schütz, M. and Adani, F. (2020). Production of referring expressions by children with ASD: Effects of referent accessibility and working memory capacity. Language Acquisition, 27(3), 276–305. https://doi:10.1080/10489223.2020 .1769625. Steiner, E. (2004). Ideational grammatical metaphor: Exploring some implications for the overall model. Languages in Contrast, 4(1), 139–166. Stern, C. and Stern, W. (1907). Die Kindersprache. Eine psychologische und sprachtheoretische Untersuchung. Leipzig: Johann Ambrosius Barth.

238

References

Strawson, P. (1950). On referring. Mind, 59(235), 320–344. Stromswold, K. (2000). The cognitive neuroscience of language acquisition. In Gazzaniga, M. (ed.), The New Cognitive Neurosciences. Cambridge: MIT Press, 909–932. Stubbs, M. (1993). British traditions in text analysis: From Firth to Sinclair. In Baker, M., Francis, G. and Tognini-Bonelli, E. (eds.), Text and Technology: In Honour of John Sinclair. Amsterdam: Benjamins, 1–33. Svennevig, J. (2010). Pre-empting reference problems in conversation. Language in Society, 39, 173–202. https://doi:10.1017/S0047404510000060. Taverniers, M. 2011. The syntax–semantics interface in systemic functional grammar: Halliday’s interpretation of the Hjelmslevian model of stratification. Journal of Pragmatics, 43, 1100–1126. Taylor, J. (2002). Cognitive Grammar. Oxford: Oxford University Press. Theakston, A. (2012). ‘The spotty cow tickled the pig with a curly tail’: How do sentence position, preferred argument structure, and referential complexity affect children’s and adults’ choice of referring expression? Applied Psycholinguistics, 33(4), 691–724. https://doi:10.1017/S0142716411000531. Thompson, G. (2004). Introducing Functional Grammar. 2nd ed. London: Hodder Education. Tomlin, R. (1987). Linguistic reflections of cognitive events. In Tomlin, R. (ed.) Coherence and Grounding in Discourse. Amsterdam: Benjamins, 455–479. Toole, J. (1996). The effect of genre on referential choice. In Gundel, J. and Fretheim, T. (eds.), Reference and Referent Accessibility. Amsterdam: Benjamins, 263–290. Tourtouri, E., Delogu, F. and Crocker, M. (2021). Rational redundancy in referring expressions: Evidence from event-related potentials. Cognitive Science, 45(12), e13071. https://doi:10.1111/cogs.13071. Trevisan, P. and García, A. (2019). Systemic functional grammar as a tool for experimental stimulus design: New appliable horizons in psycholinguistics and neurolinguistics. Language Sciences, 75, 35–46. “typical, adj.” OED Online. Oxford University Press, March 2022. Accessed 17 March 2022. Ushie, Y. (1986). ‘Corepresentation’ – A textual function of the indefinite expression. Text, 6(4), 427–446. Uzundag, B. and Küntay, A. (2018). Children’s referential communication skills: The role of cognitive abilities and adult models of speech. Journal of Experimental Child Psychology, 172, 73–95. https://doi.org/10.1016/j.jecp.2018.02.009. Van Langendonck, W. (2007). Theory and Typology of Proper Names. Berlin: De Gruyter. van Lier, E. and Rijkhoff, J. (2013). Flexible word classes in linguistic typology and grammatical theory. In Rijkhoff, J. and van Lier, E. (eds.), Flexible Word-Classes: A Typological Study of Underspecified Parts of-Speech. Oxford: Oxford University Press, 1–30. Vogels, J., Krahmer, E. and Maes, A. (2019). Accessibility and reference production: The interplay between linguistic and non-linguistic factors. In Abbott, B. and Gundel, J. (eds.), Oxford Handbook of Reference. Oxford: Oxford University Press, 337–364. Vonk, W., Hustinx, L. and Simons, W. (1992). The use of referential expressions in structuring discourse. Language and Cognitive Processes, 7(3–4), 301–333.

References

239

Warren, B. (2004). Anaphoric pronouns of metonymic expressions. metaphorik.de 07. Wegener, R. and Fontaine, L. (2023, in press). A functional approach to context. In Romero-Trillo, J. (ed.), The Cambridge Handbook of Language and Context. Cambridge: Cambridge University Press. Weil, H. (1844). De L’ordre des Mots dans les Langues Anciennes Comparées aux Langues Modernes. Paris: Didier Erudition. Wilkes-Gibbs, D. and Clark, H. H. (1992). Coordinating beliefs in conversation. Journal of Memory and Language, 31(2), 183–194. Wilson, T. (1553). The arte of rhetorique for the vse of all soche as are studious of eloquence. London: R. Graftonus. Wray, A. (1999). Formulaic language in learners and native speakers. Language Teaching, 32(4), 213–231. https://doi:10.1017/S0261444800014154. (2002). Formulaic Language and the Lexicon. Cambridge: Cambridge University Press. (2015). Why are we so sure we know what a word is?. In Taylor, J. (ed.), The Oxford Handbook of the Word. Oxford: Oxford University Press, 725–750. Wray, A. and Perkins, M. (2000). The functions of formulaic language: An integrated model. Language and Communication, 20(1), 1–28. Wright, P. (1990). Using constraints and reference in task-oriented dialogue. Journal of Semantics, 7, 65–79. Xu, H. (2023). The informational role of nominal expressions in structuring English Text. PhD thesis, RWTH Aachen University.

Index

accessibility · 15, 18–20, 82, 100, 112, 121, 124, 129, 133, 145, 150, 153–155, 158, 185, 187, 195, 215 accessibility scale · 82, 84, 87 attenuation · 83, 84 informativity · 83 rigidity · 83, 84 addressee perspective · 39, 82, 180, 186–187 adjective function of · 59 adjective-intensifiers · 55 anaphora · 13–17, 19, 28, 50, 81, 88, 99, 102–103, 105, 116, 120, 122, 150, 182, 204, 206 anaphoric predication · 17, 17 inconsistent anaphors · 122 and indefiniteness · 36 metonymic anaphora · 15 zero anaphora · 18, 87, 153, 187 anaphores infidèles · 121 antecedent · 14–16 antecedent trigger’ · 15 atypicality · 74–77, 79, 89, 91, 157, 215 indefinites · 161 automation · 80 bare noun · 118 bridging reference · 15 cataphora · 102, 103 categorisation (of concepts) · 73 chain of reference · 17, 28, 152, 171–172, 194 channel · 95, 99, 109 classifier · 55–59, 92, 120, 138, 192 function of · 57 cline of consciousness · 96–98, 106 co-classification · 162–163 co-extension · 162–165, 163, 168–172, 174 co-extensional ties · 165, 168 cognitive status · 20, 85–86, 124, 184, 189 coherence · 13–14, 159

240

cohesion · 19, 101, 163–167, 165, 170–172, 180 cohesive ties · 169–170, 174 collaboration · 10, 12, 82, 114, 128, 130, 137, 142, 146–148, 148 common ground · 82, 84, 86, 95, 97, 114, 129–131, 176, 179, 179–185, 187, 196, 213 complexity · 23, 35, 63–64, 106, 110–111, 113, 118, 120–121, 141, 143, 187 compound nouns · 57–58 conceptualisation · 12, 27, 31, 52–53, 62, 130–131, 135–136, 138, 140, 145, 148, 212 reference to · 27, 31, 52–53, 130, 138 construal · 76 context · 7, 10, 14–20, 27, 34, 38, 41–43, 75, 77–78, 92–94, 98, 103, 106, 128–129, 137, 141–142, 147–149, 200, 212–213 contextual frames · 45, 203–204 contiguity · 205, 206 conventional · 75, 80, 157, 176, 214–216 co-reference · 163 corpora · 34, 67, 77–78, 104–105, 118, 123, 153 co-textual approaches to reference · 13, 115 cultural classification · 139 definite article · 3–4, 7, 12, 41, 45, 50, 52, 60, 86, 112, 161, 181–183, 193–195, See the definite noun phrase · 43, 87, 193 definiteness · 12, 23, 32–34, 36, 40–41, 43, 47, 49, 150, 154, 156, 181, 184 ambiguity · 23, 61, 133 in child language acquisition · 181, 183 vs identifiability · 33, 42, 50 vs indefiniteness · 32, 35, 41, 150 deixis · 54, 59, 61–62, 65, 95, 119 demonstrative pronouns · 33, 60, 102, 116, 194 unattended this · 116 denotation · 17, 85, 129, 201–202, 208, 215

Index determiner · 49, 54, 59–64, 104, 112, 116 article · 60, 154 definite · 42, 47, 60, 62, 181–182 deictic determiner · 50 demonstrative determiner · 117, 154, 181 function of · 63 indefinite · 42, 60, 62, 152, 154, 181–182 possessive determiner · 45, 60–61 primary determiner · 60 quantifiers · 60–61 secondary determiner · 54, 60–64, 62, 113 determiner phrase · 48 discourse · 14, 16–19, 26, 30, 39, 43, 81–82, 88, 90, 93, 97, 99–100, 147 narrative discourse · 105 discourse event · 16 discourse representation of the referent · 15, 90 discourse-functional approach · 16 domain entities · 8 dual reference · 172 effects and effort · 213 egocentricity · 182, 184 endophora · 28, 101–104, 136 entrainment · 128 episode boundary · 88 -er nominal · 198 esphora · 4 existence · 26 exophora · 28, 101–103, 136, 206 experiential function · 59 explicitness · 40–41, 87, 174 exploitations · 79–80 eye gaze · 176, 198 familiarity · 32, 122, 133 field · 76, 93, 97, 149, 163, 167 Finnish language · 33 formulaic language · 178 French language · 121, 179 frequency · 74–79, 101, 118 function vs form · 47 functional-cognitive model of the English NP · 54 genre · 95, 120–123 gesture · 95, 176, 201 givenness · 18, 34, 38, 89–90 givenness hierarchy · 20, 84, 86, 89 activated, status of · 18, 20, 85–87 bare nominals · 104 familiar, status of · 20, 34, 85–86, 90, 117 in focus, status of · 85, 104 referential, status of · 20, 85, 90, 117 type identifiable, status of · 20, 85, 117

241 uniquely identifiable, status of · 85–86, 90, 117 Grice’s maxims maxim of quantity · 87 Grice’s maxims · 9, 155, 186 grounding · 15, 51 holophrastic utterance · 178 homophora · 103, 103 identifiability · 33, 41–42, 90, 112, 154, 156, 158 of indefinite expressions · 41, 154, 158 identification · 41, 43, 45, 87–88, 90, 143, 148, 160 identified referent · 172 identity chain · 153, 164, 166 implicature · 17, 79, 89, 151, 157, 184 indefinite article · 12, 23–24, 33, 36, 42, 50, 112, 151, 155–156, 181–183, 189, 194–195 indefinite determiners · 152, 154, 181 indefinite noun phrase · 37, 84 indefiniteness · 23, 32, 35–36, 38, 42–43, 47, 150, 156 and identifiability · 158–159, 216 indefinites · 23, 25, 32, 160–161 corepresentational · 158 vs definites · 23, 25, 182 identifiable · 153–154, 156–157, 160 late · 161, 166, 174 , See late indefinites non-identifiable · 34, 37, 156, 174 reintroductory · 158–159 second-mention · 154, 157 indexicals · 19 infant babbling · 177 inferables · 15, 34 inference · 15, 90, 183, 187, 193, 196 information · 18, 34, 38, 62, 67, 100, 106, 114, 122, 134, 155, 186 information pressure · 105 instantiation · 51 intentions · 10, 25, 41, 87, 115, 153, 157, 189 Japanese language · 33 joint construction · 115 journalistic writing · 153–154, 174 keystroke logging · 109, 112 language acquisition · See language development language development · 175–183, 197 late indefinites · 156, 158, 161, 166, 174 lexical cohesion · 162, 163, 165–166

242

Index

lexical explicitness · 41 lexicalisation · 130, 139 map task · 12, 12 markedness · 78, 183 meaning potential · 204 medium · 95, 95, 110, 113, 119 mental representations · 16, 19, 29, 82, 130, 135 metaphor · 79, 205 metonymy · 197–203, 207–208, 210–216 instrument for action · 201 and metaphor · 205 place-name metonymy · 203 referential metonymy · 198, 199, 200, 201, 206–208, 212, 215 mode · 76, 93–96, 96, 97, 108–110, 114–117, 120 mode of production · 94, 108, 218 modifier · 54, 58–59, 104, 118 classifying · 59 degree modification · 59 descriptive modification · 59 epithets · 59 objective · 59 restrictive modifiers · 44 subjective · 59 mutual responsibility · 11, 11, 115 mutuality · 213 naming · See proper names newness · 38, 158, 161, 173–174, 183, 185 nominal group · See noun phrase non-referential expressions · 29 norms · 79–80 noun phrase · 4, 9, 48, 50–51, 53–68, 88, 93, 100, 112, 121, 123, 129 categorisation zone · 55, 58, 69, 112, 118, 215 definite vs indefinite · 41 determination zone · 54, 59, 63, 65, 112, 118 full nominal · 51 function of · 57 functional zones · 54 headedness · 49, 57 modification zone · 55, 58, 124 qualification zone · 55, 110, 124, 215 semantically bland · 104 noun, function of · 51 noun-intensifiers · 55 numerative element · 61 overspecification · 9, 11, 89, 186, 188 in child language · 183, 186, 188–189

passive structures · 119 person reference · 102 person-first language · 212 phoric relations · 103, See anaphora, cataphora, endophora, esphora, exophora, homophora possessive pronouns · 61 postmodifier · 54, 63–64, 67, 118 predicative function · 160 preferred argument structure · 100, 185, 189 principle of quantity · 186, 187 problematic reference · 130, 132, 136, 141 pronominal reference · 132 pronouns · 30, 35, 39, 89–90, 102–106, 110–111, 116, 131, 185, 187, 194 ambiguity · 136 in child language development · 180–184 in spontaneous discourse · 100–101, 103, 189 in written language · 90, 195 proper names · 6, 41, 43–46, 103–105, 119, 121, 195, 203 in additional language learning · 46 attitudinal perspective · 46 as descriptors · 44 hearer-old · 45 not uniquely referring · 45 plural names · 45 referential function of · 46 as words · 46 protolanguage · 176–178 prototypicality · 77 qualification · 65–66, 68, 118–119 by intrinsic relation · 65 by property · 65 by schematic relation · 66 by situation · 65–67 ratification · 130, 180 reconciliation · 35 recoverability · 33, 41, 88, 104 refashioning · 141–142, 148 reference · 6, 10, 13, 16, 23, 25–30, 25–30, 31, 31, 36, 101 atypical reference · 191, 197–198 chains · 34, 121, 123, 163–164 cognitive research on · 82 cohesive reference · 27–28, 101 collaborative nature of · 49, 130, 148, 178 co-reference · 13, 19, 27–28, 164 creative use of · 90, 155, 196–197 definite · 4, 33, 35, 46–47, 50, 170, 172–173 discourse reference · 16, 30 dual reference · 172–173

Index evolving reference · 81 first mentions · 32, 103–104, 154 generic · 50, 53 indefinite · 36, 47, 50, 113, 172–173 indirect · 199 individuative · 50, 53 literary model · 10 metonymic · 212–216 phoricity · See phoric relations psycholinguistic research on · 9–10, 81 shifting reference · 35 specific reference · 37, 40, 181 successful reference · 124 textual reference · 28, 30 typical reference · 81, 106, 110, 156 referent · 29–31, 176, 215 virtual · 172, 174, 206 referential choice · 50, 87, 99, 106, 116, 120–123, 128, 130, 136, 138, 145, 155, 179, 185, 187, 195, 199, 212, 215 referential conflict · 132 referential function · 160 referential intent · 39–40 referential precedents · 129, 149 referral · 133 referring expression · 6, 8–9, 17–19, 23, 31, 34, 48, 67–68, 81–82, 84, 86–88, 92–94, 99, 104, 112, 116, 119, 121, 127–133, 138, 142, 144–145, 148, 158, 175, 185, 203, 215 A(n)+N+RRC · 151 abandoned referring expression · 140 provisional · 12, 128 rejected referring expression · 142, 144, 147–148 repair · 134–135 referring expression generation · 8–9 planning · 9 register · 75–78, 80, 90, 92–97, 108–109, 117, 120, 182, 215 registerial variables · 95 relative clauses · 44, 64–67, 151, 160–161, 166, 171 repetition · 75, 80, 164, 179–181, 195 rephrasing · 139, 141 salience · 213 self-monitor · 111 semantic field · 162–163, 165, 199 semantic junction · 206 semantic preference · 139 semantic prosody · 139

243 sense · 6 sense-relation restrictions · 164 sense-relations · 164–165 shared cultural knowledge · 34, 166–172 shared information · See common ground sign language · 95, 201 similarity chain · 164, 165–171 Spanish language · 36 specificity · 33, 87, 119, 134, 145, 174 spontaneous discourse · 92, 100, 185, 189 stigma · 207, 210, 212, 215 subject · 30, 34, 78, 99–101, 104, 106, 109, 116, 118, 132, 185, 192, 194, 206 tenor · 76, 80, 93, 95, 97–98 text · 16, 16 textual antecedent · 14–16 the · 3–5, 24, 27, 32, 41, 45, 50, 59–60, 62, 86, 112, 191, 193 thing · 57, 60, 62, 64–65 topic · 33, 36, 97, 106, 115, 137, 141, 145, 161, 186, 203, 215 Turkish language · 188 type · 51, 55, 57, 59–60, 85, 129, 138, 144, 146, 151, 172, 203 function of · 57 type element · 57, 118 type identification · 139 type specification · 51–58, 113, 139–140, 144–145, 147–148, 191–192, 195, 208 type-specifying noun · 133 typicality · 74, 80 underspecification · 11, 62, 131, 134, 186, 188–189 in child language · 186, 188–189 uniquely referring use · 7, 203 uniqueness · 6, 33, 44–45, 203 unmarkedness · 78 usage-based approaches · 75, 80, 213 vagueness · 58, 164 Welsh language · 33 word formation · 212 word spurt · 177 word, meaning of · 177 writing · 95, 97, 109–119, 174, 189 formal writing · 119 handwriting · 95, 110 model of · 109–110